Greeting,
FYI, we noticed a -7.6% regression of netperf.Throughput_total_tps due to commit:
commit: f0996bc2978e02d2ea898101462b960f6119b18f ("ubsan: Fix nasty -Wbuiltin-declaration-mismatch GCC-9 warnings")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
in testcase: netperf
on test machine: 8 threads Intel(R) Core(TM) i7-4770 CPU @ 3.40GHz with 8G memory
with following parameters:
ip: ipv4
runtime: 300s
nr_threads: 200%
cluster: cs-localhost
test: TCP_RR
cpufreq_governor: performance
ucode: 0x27
test-description: Netperf is a benchmark that can be use to measure various aspect of networking performance.
test-url: http://www.netperf.org/netperf/
Details are as below:
-------------------------------------------------------------------------------------------------->
To reproduce:
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
bin/lkp install job.yaml # job file is attached in this email
bin/lkp run job.yaml
=========================================================================================
cluster/compiler/cpufreq_governor/ip/kconfig/nr_threads/rootfs/runtime/tbox_group/test/testcase/ucode:
cs-localhost/gcc-7/performance/ipv4/x86_64-rhel-7.6/200%/debian-x86_64-2019-05-14.cgz/300s/lkp-hsw-d01/TCP_RR/netperf/0x27
commit:
v5.1
f0996bc297 ("ubsan: Fix nasty -Wbuiltin-declaration-mismatch GCC-9 warnings")
v5.1 f0996bc2978e02d2ea898101462
---------------- ---------------------------
%stddev %change %stddev
\ | \
529578 -7.6% 489241 netperf.Throughput_total_tps
33098 -7.6% 30577 netperf.Throughput_tps
1.526e+08 -9.6% 1.379e+08 ± 2% netperf.time.voluntary_context_switches
1.589e+08 -7.6% 1.468e+08 netperf.workload
26007 ± 11% -31.7% 17760 ± 10% cpuidle.POLL.usage
1.80 ± 40% -0.7 1.14 ± 19% mpstat.cpu.all.idle%
1038227 -7.1% 965017 vmstat.system.cs
4174 ± 12% +15.0% 4799 ± 5% slabinfo.anon_vma.active_objs
4212 ± 11% +14.0% 4799 ± 5% slabinfo.anon_vma.num_objs
329.81 -100.0% 0.00 uptime.boot
194.38 ± 9% -100.0% 0.00 uptime.idle
60071 +1.4% 60889 proc-vmstat.nr_active_anon
58101 +1.8% 59122 proc-vmstat.nr_anon_pages
4248 -1.7% 4175 proc-vmstat.nr_kernel_stack
60069 +1.4% 60886 proc-vmstat.nr_zone_active_anon
3.186e+09 -6.8% 2.968e+09 perf-stat.i.branch-instructions
88559276 -6.7% 82586895 perf-stat.i.branch-misses
1048131 -7.0% 975134 perf-stat.i.context-switches
1.82 +7.0% 1.94 perf-stat.i.cpi
4.744e+09 -7.0% 4.414e+09 perf-stat.i.dTLB-loads
11699752 ± 2% -7.6% 10814910 perf-stat.i.dTLB-store-misses
2.956e+09 -7.0% 2.748e+09 perf-stat.i.dTLB-stores
63.54 +6.4 69.97 perf-stat.i.iTLB-load-miss-rate%
17352165 ± 3% -12.4% 15198695 ± 4% perf-stat.i.iTLB-load-misses
9871780 -34.6% 6452252 perf-stat.i.iTLB-loads
1.61e+10 -6.9% 1.5e+10 perf-stat.i.instructions
0.56 -7.1% 0.52 perf-stat.i.ipc
1.80 +8.1% 1.95 perf-stat.overall.cpi
63.72 +6.4 70.17 perf-stat.overall.iTLB-load-miss-rate%
0.56 -7.5% 0.51 perf-stat.overall.ipc
3.177e+09 -6.9% 2.958e+09 perf-stat.ps.branch-instructions
88299894 -6.8% 82312493 perf-stat.ps.branch-misses
1045053 -7.0% 971889 perf-stat.ps.context-switches
4.73e+09 -7.0% 4.399e+09 perf-stat.ps.dTLB-loads
11665343 -7.6% 10778872 perf-stat.ps.dTLB-store-misses
2.947e+09 -7.1% 2.739e+09 perf-stat.ps.dTLB-stores
17301097 ± 3% -12.4% 15148078 ± 4% perf-stat.ps.iTLB-load-misses
9842789 -34.7% 6430777 perf-stat.ps.iTLB-loads
1.605e+10 -6.9% 1.495e+10 perf-stat.ps.instructions
4.91e+12 -7.6% 4.538e+12 perf-stat.total.instructions
27.60 -0.6 27.05 perf-profile.calltrace.cycles-pp.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe
27.36 -0.5 26.82 perf-profile.calltrace.cycles-pp.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe
25.94 -0.5 25.40 perf-profile.calltrace.cycles-pp.inet_recvmsg.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe
25.52 -0.5 24.98 perf-profile.calltrace.cycles-pp.tcp_recvmsg.inet_recvmsg.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64
3.53 ± 5% -0.4 3.17 ± 2% perf-profile.calltrace.cycles-pp.__skb_datagram_iter.skb_copy_datagram_iter.tcp_recvmsg.inet_recvmsg.__sys_recvfrom
3.57 ± 5% -0.4 3.21 ± 2% perf-profile.calltrace.cycles-pp.skb_copy_datagram_iter.tcp_recvmsg.inet_recvmsg.__sys_recvfrom.__x64_sys_recvfrom
2.66 ± 3% -0.3 2.39 ± 3% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64
1.19 ± 7% -0.3 0.92 ± 3% perf-profile.calltrace.cycles-pp.__inet_lookup_established.tcp_v4_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish.ip_local_deliver
2.39 ± 3% -0.3 2.14 ± 3% perf-profile.calltrace.cycles-pp.pick_next_task_fair.__schedule.schedule.schedule_timeout.wait_woken
0.95 ± 5% -0.2 0.76 ± 6% perf-profile.calltrace.cycles-pp.kmem_cache_alloc_node.__alloc_skb.sk_stream_alloc_skb.tcp_sendmsg_locked.tcp_sendmsg
2.17 ± 3% -0.2 1.98 ± 2% perf-profile.calltrace.cycles-pp.syscall_return_via_sysret
1.34 ± 5% -0.2 1.17 ± 8% perf-profile.calltrace.cycles-pp.selinux_ip_postroute_compat.selinux_ip_postroute.nf_hook_slow.ip_output.__ip_queue_xmit
1.68 ± 5% -0.2 1.52 ± 4% perf-profile.calltrace.cycles-pp.selinux_ip_postroute.nf_hook_slow.ip_output.__ip_queue_xmit.__tcp_transmit_skb
1.90 ± 4% -0.1 1.75 ± 2% perf-profile.calltrace.cycles-pp.nf_hook_slow.ip_output.__ip_queue_xmit.__tcp_transmit_skb.tcp_write_xmit
0.94 ± 5% -0.1 0.83 ± 7% perf-profile.calltrace.cycles-pp.sockfd_lookup_light.__sys_sendto.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.80 ± 8% -0.1 0.69 ± 8% perf-profile.calltrace.cycles-pp.__fget_light.sockfd_lookup_light.__sys_sendto.__x64_sys_sendto.do_syscall_64
0.62 ± 6% +0.1 0.74 ± 4% perf-profile.calltrace.cycles-pp.__netif_receive_skb_core.__netif_receive_skb_one_core.process_backlog.net_rx_action.__softirqentry_text_start
2.68 ± 3% +0.1 2.81 perf-profile.calltrace.cycles-pp.dev_hard_start_xmit.__dev_queue_xmit.ip_finish_output2.ip_output.__ip_queue_xmit
1.05 ± 5% +0.2 1.27 ± 6% perf-profile.calltrace.cycles-pp.reweight_entity.dequeue_task_fair.__schedule.schedule.schedule_timeout
3.24 ± 2% +0.2 3.47 ± 2% perf-profile.calltrace.cycles-pp.tcp_clean_rtx_queue.tcp_ack.tcp_rcv_established.tcp_v4_do_rcv.tcp_v4_rcv
0.84 ± 6% +0.3 1.09 ± 5% perf-profile.calltrace.cycles-pp.ttwu_do_wakeup.try_to_wake_up.__wake_up_common.__wake_up_common_lock.sock_def_readable
0.71 ± 8% +0.3 0.97 ± 4% perf-profile.calltrace.cycles-pp.check_preempt_curr.ttwu_do_wakeup.try_to_wake_up.__wake_up_common.__wake_up_common_lock
4.09 ± 3% +0.4 4.48 perf-profile.calltrace.cycles-pp.__dev_queue_xmit.ip_finish_output2.ip_output.__ip_queue_xmit.__tcp_transmit_skb
4.13 ± 3% +0.4 4.57 ± 2% perf-profile.calltrace.cycles-pp.dequeue_task_fair.__schedule.schedule.schedule_timeout.wait_woken
0.34 ± 81% +0.4 0.78 ± 2% perf-profile.calltrace.cycles-pp.check_preempt_wakeup.check_preempt_curr.ttwu_do_wakeup.try_to_wake_up.__wake_up_common
90.77 +0.5 91.22 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe
0.16 ±152% +0.5 0.63 ± 4% perf-profile.calltrace.cycles-pp.validate_xmit_skb.__dev_queue_xmit.ip_finish_output2.ip_output.__ip_queue_xmit
90.23 +0.5 90.70 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe
32.13 +0.5 32.62 perf-profile.calltrace.cycles-pp.net_rx_action.__softirqentry_text_start.do_softirq_own_stack.do_softirq.__local_bh_enable_ip
32.62 +0.5 33.12 perf-profile.calltrace.cycles-pp.__softirqentry_text_start.do_softirq_own_stack.do_softirq.__local_bh_enable_ip.ip_finish_output2
0.39 ± 65% +0.5 0.90 ± 4% perf-profile.calltrace.cycles-pp.tcp_schedule_loss_probe.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg
32.89 +0.5 33.40 perf-profile.calltrace.cycles-pp.do_softirq_own_stack.do_softirq.__local_bh_enable_ip.ip_finish_output2.ip_output
31.17 +0.5 31.71 perf-profile.calltrace.cycles-pp.process_backlog.net_rx_action.__softirqentry_text_start.do_softirq_own_stack.do_softirq
41.58 +0.6 42.14 perf-profile.calltrace.cycles-pp.ip_output.__ip_queue_xmit.__tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames
47.15 +0.6 47.76 perf-profile.calltrace.cycles-pp.__tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg
30.12 +0.6 30.74 perf-profile.calltrace.cycles-pp.__netif_receive_skb_one_core.process_backlog.net_rx_action.__softirqentry_text_start.do_softirq_own_stack
0.00 +0.6 0.63 ± 3% perf-profile.calltrace.cycles-pp.tcp_cleanup_rbuf.tcp_recvmsg.inet_recvmsg.__sys_recvfrom.__x64_sys_recvfrom
5.72 +0.7 6.41 perf-profile.calltrace.cycles-pp.tcp_ack.tcp_rcv_established.tcp_v4_do_rcv.tcp_v4_rcv.ip_protocol_deliver_rcu
22.36 +0.8 23.11 perf-profile.calltrace.cycles-pp.tcp_rcv_established.tcp_v4_do_rcv.tcp_v4_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish
22.77 +0.8 23.57 perf-profile.calltrace.cycles-pp.tcp_v4_do_rcv.tcp_v4_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish.ip_local_deliver
61.36 +1.1 62.43 perf-profile.calltrace.cycles-pp.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe
61.07 +1.1 62.14 perf-profile.calltrace.cycles-pp.__sys_sendto.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe
50.16 +1.3 51.45 perf-profile.calltrace.cycles-pp.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg.sock_sendmsg
58.68 +1.3 59.98 perf-profile.calltrace.cycles-pp.tcp_sendmsg.sock_sendmsg.__sys_sendto.__x64_sys_sendto.do_syscall_64
59.36 +1.3 60.66 perf-profile.calltrace.cycles-pp.sock_sendmsg.__sys_sendto.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe
50.37 +1.3 51.71 perf-profile.calltrace.cycles-pp.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg.sock_sendmsg.__sys_sendto
57.27 +1.4 58.70 perf-profile.calltrace.cycles-pp.tcp_sendmsg_locked.tcp_sendmsg.sock_sendmsg.__sys_sendto.__x64_sys_sendto
27.64 -0.6 27.09 perf-profile.children.cycles-pp.__x64_sys_recvfrom
27.38 -0.5 26.83 perf-profile.children.cycles-pp.__sys_recvfrom
25.95 -0.5 25.41 perf-profile.children.cycles-pp.inet_recvmsg
3.55 ± 5% -0.4 3.18 ± 2% perf-profile.children.cycles-pp.__skb_datagram_iter
3.58 ± 5% -0.4 3.22 ± 2% perf-profile.children.cycles-pp.skb_copy_datagram_iter
2.68 ± 3% -0.3 2.40 ± 3% perf-profile.children.cycles-pp.entry_SYSCALL_64
1.20 ± 7% -0.3 0.93 ± 2% perf-profile.children.cycles-pp.__inet_lookup_established
2.55 ± 3% -0.2 2.33 ± 3% perf-profile.children.cycles-pp.pick_next_task_fair
0.98 ± 5% -0.2 0.77 ± 6% perf-profile.children.cycles-pp.kmem_cache_alloc_node
1.36 ± 4% -0.2 1.17 ± 2% perf-profile.children.cycles-pp.update_load_avg
1.35 ± 5% -0.2 1.19 ± 8% perf-profile.children.cycles-pp.selinux_ip_postroute_compat
1.25 ± 5% -0.2 1.08 ± 4% perf-profile.children.cycles-pp.release_sock
1.69 ± 4% -0.2 1.54 ± 4% perf-profile.children.cycles-pp.selinux_ip_postroute
1.21 ± 4% -0.1 1.07 ± 2% perf-profile.children.cycles-pp.ktime_get
1.11 ± 3% -0.1 0.99 ± 5% perf-profile.children.cycles-pp._raw_spin_lock_irqsave
2.19 ± 3% -0.1 2.08 ± 2% perf-profile.children.cycles-pp.sk_filter_trim_cap
1.00 ± 6% -0.1 0.89 ± 2% perf-profile.children.cycles-pp.tcp_mstamp_refresh
1.04 ± 3% -0.1 0.94 ± 2% perf-profile.children.cycles-pp._raw_spin_lock_bh
0.65 ± 6% -0.1 0.56 ± 3% perf-profile.children.cycles-pp.selinux_parse_skb
0.29 ± 9% -0.1 0.23 ± 4% perf-profile.children.cycles-pp.bictcp_acked
0.09 ± 15% -0.1 0.03 ±100% perf-profile.children.cycles-pp.should_failslab
0.17 ± 14% -0.0 0.12 ± 13% perf-profile.children.cycles-pp.__list_add_valid
0.18 ± 9% -0.0 0.14 ± 13% perf-profile.children.cycles-pp.tcp_rate_skb_sent
0.07 ± 27% +0.0 0.10 ± 19% perf-profile.children.cycles-pp.skb_clone_tx_timestamp
0.09 ± 16% +0.0 0.14 ± 9% perf-profile.children.cycles-pp.rb_first
0.22 ± 7% +0.0 0.27 ± 5% perf-profile.children.cycles-pp.clear_buddies
0.05 ± 52% +0.1 0.10 ± 22% perf-profile.children.cycles-pp.selinux_xfrm_sock_rcv_skb
0.19 ± 11% +0.1 0.24 ± 7% perf-profile.children.cycles-pp.tcp_stream_memory_free
0.06 ± 39% +0.1 0.12 ± 13% perf-profile.children.cycles-pp.ns_to_timespec64
0.36 ± 8% +0.1 0.42 ± 4% perf-profile.children.cycles-pp.__usecs_to_jiffies
0.21 ± 11% +0.1 0.28 ± 12% perf-profile.children.cycles-pp.native_load_tls
0.07 ± 23% +0.1 0.14 ± 3% perf-profile.children.cycles-pp.tcp_rbtree_insert
0.08 ± 21% +0.1 0.15 ± 8% perf-profile.children.cycles-pp.tcp_rate_skb_delivered
0.03 ±101% +0.1 0.10 ± 20% perf-profile.children.cycles-pp.tcp_rack_advance
0.29 ± 5% +0.1 0.38 ± 14% perf-profile.children.cycles-pp.tcp_rate_gen
0.02 ±152% +0.1 0.10 ± 7% perf-profile.children.cycles-pp.tcp_rack_update_reo_wnd
0.07 ± 27% +0.1 0.16 ± 7% perf-profile.children.cycles-pp.skb_clone
0.29 ± 6% +0.1 0.39 ± 7% perf-profile.children.cycles-pp.__tcp_select_window
0.20 ± 19% +0.1 0.30 ± 6% perf-profile.children.cycles-pp.tcp_send_delayed_ack
0.41 ± 8% +0.1 0.51 ± 9% perf-profile.children.cycles-pp.tcp_rearm_rto
0.22 ± 22% +0.1 0.32 ± 11% perf-profile.children.cycles-pp.netif_skb_features
0.01 ±300% +0.1 0.13 ± 15% perf-profile.children.cycles-pp.kmalloc_slab
0.63 ± 6% +0.1 0.75 ± 4% perf-profile.children.cycles-pp.__netif_receive_skb_core
2.69 ± 3% +0.1 2.82 perf-profile.children.cycles-pp.dev_hard_start_xmit
0.61 ± 8% +0.1 0.76 ± 9% perf-profile.children.cycles-pp.account_entity_dequeue
0.45 ± 19% +0.2 0.66 ± 3% perf-profile.children.cycles-pp.validate_xmit_skb
3.30 ± 3% +0.2 3.52 ± 2% perf-profile.children.cycles-pp.tcp_clean_rtx_queue
0.84 ± 6% +0.2 1.09 ± 4% perf-profile.children.cycles-pp.ttwu_do_wakeup
0.72 ± 7% +0.3 0.98 ± 4% perf-profile.children.cycles-pp.check_preempt_curr
0.54 ± 10% +0.3 0.81 ± 2% perf-profile.children.cycles-pp.check_preempt_wakeup
2.22 ± 3% +0.3 2.49 ± 3% perf-profile.children.cycles-pp.reweight_entity
0.35 ± 10% +0.3 0.64 ± 2% perf-profile.children.cycles-pp.tcp_cleanup_rbuf
4.13 ± 3% +0.4 4.53 perf-profile.children.cycles-pp.__dev_queue_xmit
0.71 ± 7% +0.4 1.13 ± 3% perf-profile.children.cycles-pp.tcp_schedule_loss_probe
90.85 +0.4 91.28 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
4.16 ± 3% +0.5 4.62 ± 2% perf-profile.children.cycles-pp.dequeue_task_fair
90.29 +0.5 90.75 perf-profile.children.cycles-pp.do_syscall_64
32.65 +0.5 33.15 perf-profile.children.cycles-pp.__softirqentry_text_start
32.19 +0.5 32.70 perf-profile.children.cycles-pp.net_rx_action
32.91 +0.5 33.42 perf-profile.children.cycles-pp.do_softirq_own_stack
31.19 +0.5 31.72 perf-profile.children.cycles-pp.process_backlog
41.60 +0.6 42.17 perf-profile.children.cycles-pp.ip_output
47.19 +0.6 47.80 perf-profile.children.cycles-pp.__tcp_transmit_skb
30.13 +0.6 30.76 perf-profile.children.cycles-pp.__netif_receive_skb_one_core
5.82 +0.7 6.52 perf-profile.children.cycles-pp.tcp_ack
22.41 +0.8 23.17 perf-profile.children.cycles-pp.tcp_rcv_established
22.77 +0.8 23.58 perf-profile.children.cycles-pp.tcp_v4_do_rcv
61.40 +1.1 62.46 perf-profile.children.cycles-pp.__x64_sys_sendto
61.09 +1.1 62.16 perf-profile.children.cycles-pp.__sys_sendto
58.74 +1.3 60.02 perf-profile.children.cycles-pp.tcp_sendmsg
59.36 +1.3 60.66 perf-profile.children.cycles-pp.sock_sendmsg
50.24 +1.3 51.55 perf-profile.children.cycles-pp.tcp_write_xmit
50.38 +1.4 51.74 perf-profile.children.cycles-pp.__tcp_push_pending_frames
57.40 +1.4 58.85 perf-profile.children.cycles-pp.tcp_sendmsg_locked
2.68 ± 3% -0.3 2.40 ± 3% perf-profile.self.cycles-pp.entry_SYSCALL_64
0.98 ± 8% -0.2 0.75 perf-profile.self.cycles-pp.__inet_lookup_established
0.90 ± 8% -0.2 0.70 ± 12% perf-profile.self.cycles-pp.ip_finish_output2
0.82 ± 9% -0.2 0.62 ± 8% perf-profile.self.cycles-pp.__check_object_size
1.09 ± 3% -0.1 0.96 ± 6% perf-profile.self.cycles-pp._raw_spin_lock_irqsave
0.61 ± 7% -0.1 0.51 ± 3% perf-profile.self.cycles-pp.selinux_parse_skb
0.57 ± 6% -0.1 0.50 perf-profile.self.cycles-pp.__sys_sendto
0.24 ± 10% -0.0 0.20 ± 10% perf-profile.self.cycles-pp.import_single_range
0.23 ± 10% -0.0 0.19 ± 10% perf-profile.self.cycles-pp.sk_filter_trim_cap
0.30 ± 6% -0.0 0.27 ± 7% perf-profile.self.cycles-pp.update_rq_clock
0.18 ± 11% +0.0 0.21 ± 8% perf-profile.self.cycles-pp._copy_from_iter_full
0.06 ± 25% +0.0 0.10 ± 23% perf-profile.self.cycles-pp.skb_clone_tx_timestamp
0.08 ± 24% +0.0 0.12 ± 16% perf-profile.self.cycles-pp.security_socket_recvmsg
0.18 ± 8% +0.0 0.22 ± 3% perf-profile.self.cycles-pp.clear_buddies
0.32 ± 5% +0.0 0.36 ± 4% perf-profile.self.cycles-pp.ip_rcv
0.07 ± 17% +0.0 0.12 ± 13% perf-profile.self.cycles-pp.rb_first
0.17 ± 6% +0.1 0.23 ± 7% perf-profile.self.cycles-pp.tcp_stream_memory_free
0.03 ± 82% +0.1 0.09 ± 21% perf-profile.self.cycles-pp.selinux_xfrm_sock_rcv_skb
0.21 ± 14% +0.1 0.27 ± 11% perf-profile.self.cycles-pp.__wake_up_common
0.14 ± 11% +0.1 0.21 ± 8% perf-profile.self.cycles-pp.__tcp_push_pending_frames
0.18 ± 11% +0.1 0.25 ± 9% perf-profile.self.cycles-pp.__netif_receive_skb_one_core
0.07 ± 23% +0.1 0.13 ± 3% perf-profile.self.cycles-pp.tcp_rbtree_insert
0.21 ± 11% +0.1 0.28 ± 12% perf-profile.self.cycles-pp.native_load_tls
0.20 ± 11% +0.1 0.27 ± 14% perf-profile.self.cycles-pp.dequeue_entity
0.33 ± 9% +0.1 0.40 ± 9% perf-profile.self.cycles-pp.tcp_rearm_rto
0.03 ±122% +0.1 0.10 ± 22% perf-profile.self.cycles-pp.tcp_rack_advance
0.00 +0.1 0.07 ± 15% perf-profile.self.cycles-pp.netif_rx
0.07 ± 24% +0.1 0.15 ± 10% perf-profile.self.cycles-pp.tcp_rate_skb_delivered
0.16 ± 24% +0.1 0.24 ± 18% perf-profile.self.cycles-pp.netif_skb_features
0.02 ±152% +0.1 0.10 ± 9% perf-profile.self.cycles-pp.tcp_rack_update_reo_wnd
0.41 ± 17% +0.1 0.49 ± 12% perf-profile.self.cycles-pp.enqueue_entity
0.28 ± 5% +0.1 0.37 ± 13% perf-profile.self.cycles-pp.tcp_rate_gen
0.18 ± 20% +0.1 0.27 ± 4% perf-profile.self.cycles-pp.validate_xmit_skb
0.20 ± 20% +0.1 0.29 ± 5% perf-profile.self.cycles-pp.tcp_send_delayed_ack
0.06 ± 26% +0.1 0.15 ± 7% perf-profile.self.cycles-pp.skb_clone
0.28 ± 7% +0.1 0.38 ± 8% perf-profile.self.cycles-pp.__tcp_select_window
0.11 ± 18% +0.1 0.21 ± 15% perf-profile.self.cycles-pp.netif_rx_internal
0.42 ± 10% +0.1 0.54 ± 6% perf-profile.self.cycles-pp.__alloc_skb
0.00 +0.1 0.12 ± 13% perf-profile.self.cycles-pp.kmalloc_slab
0.53 ± 7% +0.1 0.65 ± 8% perf-profile.self.cycles-pp.account_entity_dequeue
0.62 ± 6% +0.1 0.75 ± 4% perf-profile.self.cycles-pp.__netif_receive_skb_core
0.60 ± 5% +0.2 0.77 ± 2% perf-profile.self.cycles-pp.try_to_wake_up
0.20 ± 10% +0.2 0.38 ± 5% perf-profile.self.cycles-pp.tcp_cleanup_rbuf
1.23 ± 6% +0.2 1.40 ± 3% perf-profile.self.cycles-pp.tcp_sendmsg_locked
0.90 ± 5% +0.2 1.12 ± 3% perf-profile.self.cycles-pp.reweight_entity
0.35 ± 13% +0.2 0.59 ± 3% perf-profile.self.cycles-pp.check_preempt_wakeup
1.14 ± 6% +0.3 1.39 ± 3% perf-profile.self.cycles-pp.tcp_write_xmit
0.37 ± 10% +0.3 0.64 ± 5% perf-profile.self.cycles-pp.tcp_schedule_loss_probe
1.10 ± 6% +0.3 1.45 ± 2% perf-profile.self.cycles-pp.tcp_ack
netperf.Throughput_tps
34000 +-+-----------------------------------------------------------------+
| + + .++ + |
33500 +-+.+.+. .+ .+ + +. .+.+. + .+. .+.+. .+ |
33000 +-+ ++.+ + .+ + +.+.+ + .+ + +.++ + .+ |
| + + + |
32500 +-+ |
32000 +-+ |
| |
31500 +-+ |
31000 +-+ O |
| O O O O O O O |
30500 +-+ O O O O O O O O
30000 +-+ O O |
O O O O O O O O O |
29500 +-O---O-O----------------O--O-O-O-O-O-O-----------------------------+
netperf.Throughput_total_tps
550000 +-+----------------------------------------------------------------+
| + .+ |
540000 +-+ + + .+.+ + |
530000 +-+.+.+ .+. + + +. .++. .+. +. .+.+. +. |
| + +.+.+ +.+ + +.+.+ +.+ + +.+.+ +.+ |
520000 +-+ |
| |
510000 +-+ |
| |
500000 +-+ O |
490000 +-+ O O O O O O O |
| O O O O OO O O
480000 +-+ O O O O |
O O O O O O O O O O |
470000 +-O----O----------------O-------O-OO-O-----------------------------+
netperf.workload
1.65e+08 +-+--------------------------------------------------------------+
| .+ .+ |
| .+ + +.+ + |
1.6e+08 +-+.++. .+. .+ .+ + +. .++. .+. .+ .+.+. .++. .+. .+ .+ |
| + + + + + + + + + + + + |
| |
1.55e+08 +-+ |
| |
1.5e+08 +-+ |
| O O |
| OO O O O O O O OO O O
1.45e+08 +-+ O O O |
O O O O O O O O O |
| O O O O O O O OO O O |
1.4e+08 +-+--------------------------------------------------------------+
netperf.time.voluntary_context_switches
1.6e+08 +-+--------------------------------------------------------------+
| +.+ +. |
1.55e+08 +-+. + + : + .+. .+. |
| ++. +. .+ + + .+ ++. +. .+.+ ++. +. .+ |
| +. + ++ + +. + ++ +. + ++ |
1.5e+08 +-+ + + + |
| O |
1.45e+08 +-+ |
| O O O O O |
1.4e+08 +-+ O O O O O O O O |
| O O O O O O O O O O O O O O O
| OO O O |
1.35e+08 O-+ O O O |
| |
1.3e+08 +-+--------------------------------------------------------------+
[*] bisect-good sample
[O] bisect-bad sample
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.
Thanks,
Rong Chen
On 5/20/19 8:38 AM, kernel test robot wrote:
> Greeting,
>
> FYI, we noticed a -7.6% regression of netperf.Throughput_total_tps due to commit:
>
>
> commit: f0996bc2978e02d2ea898101462b960f6119b18f ("ubsan: Fix nasty -Wbuiltin-declaration-mismatch GCC-9 warnings")
> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
>
This can't be right. First of all the commit makes changes only lib/ubsan.c which is compiled only when CONFIG_UBSAN=y.
In your config:
# CONFIG_UBSAN is not set
But even in the case of enabled UBSAN that commit doesn't change the generated machine code at all.
On Mon, May 20, 2019 at 12:50:30PM +0300, Andrey Ryabinin wrote:
>
>
> On 5/20/19 8:38 AM, kernel test robot wrote:
> > Greeting,
> >
> > FYI, we noticed a -7.6% regression of netperf.Throughput_total_tps due to commit:
> >
> >
> > commit: f0996bc2978e02d2ea898101462b960f6119b18f ("ubsan: Fix nasty -Wbuiltin-declaration-mismatch GCC-9 warnings")
> > https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
> >
>
> This can't be right. First of all the commit makes changes only lib/ubsan.c which is compiled only when CONFIG_UBSAN=y.
> In your config:
> # CONFIG_UBSAN is not set
>
> But even in the case of enabled UBSAN that commit doesn't change the generated machine code at all.
Hi,
Sorry for bringing you inconvenience. We retested the commit and
couldn't reproduce the regression, please ignore the invalid report.
Best Regards,
Rong Chen