[lkp-robot] [x86/entry/64] 3e3b9293d3: will-it-scale.per_thread_ops -6.3% regression
by kernel test robot
Greeting,
We noticed a -6.3 regression of will-it-scale.per_thread_ops due to commit:
commit: 3e3b9293d392c577b62e24e4bc9982320438e749 ("x86/entry/64: Return to userspace from the trampoline stack")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
in testcase: will-it-scale
on test machine: 256 threads Phi with 96G memory
with following parameters:
nr_task: 100%
mode: thread
test: futex3
cpufreq_governor: performance
test-description: Will It Scale takes a testcase and runs it from 1 through to n parallel copies to see if the testcase will scale. It builds both a process and threads based test in order to see any differences between the two.
test-url: https://github.com/antonblanchard/will-it-scale
Details are as below:
-------------------------------------------------------------------------------------------------->
=========================================================================================
tbox_group/testcase/rootfs/kconfig/compiler/nr_task/mode/test/cpufreq_governor:
lkp-knl-f1/will-it-scale/debian-x86_64-2016-08-31.cgz/x86_64-kexec/gcc-7/100%/thread/futex3/performance
commit:
7f2590a110b837af5679d08fc25c6227c5a8c497
3e3b9293d392c577b62e24e4bc9982320438e749
7f2590a110b837af 3e3b9293d392c577b62e24e4bc
---------------- --------------------------
fail:runs %reproduction fail:runs
| | |
635:4 -2117% 551:4 dmesg.timestamp:last
:4 25% 1:4 kmsg.do_IRQ:#No_irq_handler_for_vector
%stddev %change %stddev
\ | \
763588 -6.3% 715282 will-it-scale.per_thread_ops
9821 -1.2% 9706 will-it-scale.time.maximum_resident_set_size
64640 -24.7% 48651 will-it-scale.time.system_time
12268 +130.4% 28260 will-it-scale.time.user_time
93710 ± 22% -16.0% 78721 ± 2% meminfo.Shmem
93563 ± 21% -16.0% 78626 ± 2% numa-meminfo.node0.Shmem
23457 ± 21% -16.0% 19695 ± 2% numa-vmstat.node0.nr_shmem
11213 ± 2% -12.5% 9811 vmstat.system.cs
69.31 -17.0 52.33 mpstat.cpu.sys%
13.18 +17.2 30.40 mpstat.cpu.usr%
847737 +69.6% 1437473 softirqs.RCU
8457683 ± 12% +14.4% 9674578 ± 4% softirqs.TIMER
64640 -24.7% 48651 time.system_time
12268 +130.4% 28260 time.user_time
6398 ± 17% -12.8% 5580 ± 3% proc-vmstat.nr_inactive_anon
9817 ± 18% -13.9% 8451 ± 4% proc-vmstat.nr_mapped
23492 ± 21% -16.2% 19693 ± 2% proc-vmstat.nr_shmem
6398 ± 17% -12.8% 5580 ± 3% proc-vmstat.nr_zone_inactive_anon
27596 ± 26% -19.0% 22345 ± 2% proc-vmstat.pgactivate
67.06 ± 10% -22.8% 51.78 ± 10% sched_debug.cfs_rq:/.util_avg.stddev
1007010 ± 3% +7.3% 1080219 ± 5% sched_debug.cpu.avg_idle.avg
1953781 ± 27% +101.9% 3944146 ± 41% sched_debug.cpu.avg_idle.max
176220 ± 43% +147.9% 436915 ± 48% sched_debug.cpu.avg_idle.stddev
667.73 ± 19% -29.9% 468.27 ± 17% sched_debug.cpu.curr->pid.stddev
587552 ± 15% +59.5% 937023 ± 29% sched_debug.cpu.max_idle_balance_cost.max
8681 ±127% +601.6% 60908 ± 61% sched_debug.cpu.max_idle_balance_cost.stddev
8187 ± 17% -25.1% 6131 ± 2% sched_debug.cpu.nr_switches.avg
186981 ± 6% -23.6% 142817 ± 3% sched_debug.cpu.nr_switches.max
27312 ± 8% -21.2% 21528 ± 5% sched_debug.cpu.nr_switches.stddev
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
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,
Xiaolong
4 years, 4 months
- 1
- 0
[lkp-robot] [x86/kernel] 5bf300774c: WARNING:at_kernel/jump_label.c:#jump_label_test
by kernel test robot
FYI, we noticed the following commit (built with gcc-7):
commit: 5bf300774c2da4e2b3cf8891ad5a9fa104a429be ("x86/kernel: jump_table: use relative references")
url: https://github.com/0day-ci/linux/commits/Ard-Biesheuvel/add-support-for-r...
in testcase: trinity
with following parameters:
runtime: 300s
test-description: Trinity is a linux system call fuzz tester.
test-url: http://codemonkey.org.uk/projects/trinity/
on test machine: qemu-system-x86_64 -enable-kvm -cpu host -smp 2 -m 1G
caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
+-----------------------------------------------------------------------------------------------+------------+------------+
| | 2421cce8ea | 5bf300774c |
+-----------------------------------------------------------------------------------------------+------------+------------+
| boot_successes | 0 | 0 |
| boot_failures | 8 | 8 |
| BUG:kernel_reboot-without-warning_in_early-boot_stage,last_printk:early_console_in_setup_code | 5 | |
| BUG:kernel_hang_in_early-boot_stage,last_printk:early_console_in_setup_code | 3 | |
| WARNING:at_kernel/jump_label.c:#jump_label_test | 0 | 8 |
| RIP:jump_label_test | 0 | 8 |
| invoked_oom-killer:gfp_mask=0x | 0 | 8 |
| Mem-Info | 0 | 8 |
| Kernel_panic-not_syncing:Out_of_memory_and_no_killable_processes | 0 | 8 |
+-----------------------------------------------------------------------------------------------+------------+------------+
[ 0.040547] WARNING: CPU: 0 PID: 1 at kernel/jump_label.c:755 jump_label_test+0x3db/0x533
[ 0.042493] Modules linked in:
[ 0.043089] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.15.0-rc5-00008-g5bf3007 #1
[ 0.044494] RIP: 0010:jump_label_test+0x3db/0x533
[ 0.045428] RSP: 0000:ffff8800002dfce8 EFLAGS: 00010202
[ 0.046405] RAX: dffffc0000000001 RBX: 1ffff1000005bf9e RCX: 0000000000000000
[ 0.047830] RDX: 1ffffffff3b009f3 RSI: 0000000000000001 RDI: ffffffff9d804f98
[ 0.049218] RBP: dffffc0000000000 R08: 0000000000000001 R09: ffffffff9ec11c60
[ 0.050000] R10: 0000000000000000 R11: 0000000000000000 R12: ffffed000005bfa2
[ 0.050000] R13: ffffffff9ec5a940 R14: ffffffff9d06f0a0 R15: 0000000000000002
[ 0.050000] FS: 0000000000000000(0000) GS:ffff88001d000000(0000) knlGS:0000000000000000
[ 0.050000] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 0.050000] CR2: 00000000ffffffff CR3: 000000002a013001 CR4: 00000000001606f0
[ 0.050000] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 0.050000] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 0.050000] Call Trace:
[ 0.050000] ? jump_label_init_module+0x11/0x11
[ 0.050000] ? set_cpus_allowed_common+0x6c/0xc0
[ 0.050000] ? blocking_notifier_chain_register+0x22a/0x490
[ 0.050000] ? register_die_notifier+0x20/0x20
[ 0.050000] ? early_enable_events+0xa6/0x175
[ 0.050000] ? event_trace_init+0x3d5/0x3d5
[ 0.050000] ? __tasklet_schedule+0x400/0x400
[ 0.050000] ? jump_label_init_module+0x11/0x11
[ 0.050000] do_one_initcall+0x7d/0x267
[ 0.050000] ? initcall_blacklisted+0x1a0/0x1a0
[ 0.050000] ? extfrag_debug_init+0x78/0x78
[ 0.050000] ? workqueue_init+0x33a/0x341
[ 0.050000] kernel_init_freeable+0x2d6/0x4c9
[ 0.050000] ? rest_init+0x120/0x120
[ 0.050000] kernel_init+0xf/0x120
[ 0.050000] ? rest_init+0x120/0x120
[ 0.050000] ret_from_fork+0x24/0x30
[ 0.050000] Code: 31 c9 31 d2 44 0f b6 c0 48 c7 c7 78 4f 80 9d 88 44 24 04 44 89 c6 44 89 04 24 e8 56 72 87 fb 8a 44 24 04 44 8b 04 24 84 c0 74 02 <0f> ff 31 c9 31 d2 44 89 c6 48 c7 c7 48 4f 80 9d e8 35 72 87 fb
[ 0.050000] ---[ end trace 206c70f65ccc5d90 ]---
To reproduce:
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
bin/lkp qemu -k <bzImage> job-script # job-script is attached in this email
Thanks,
Xiaolong
4 years, 4 months
- 1
- 0
6e5f4c4bb6 ("irqdomain: Convert to XArray"): WARNING: possible irq lock inversion dependency detected
by kernel test robot
Greetings,
0day kernel testing robot got the below dmesg and the first bad commit is
git://git.infradead.org/users/willy/linux-dax.git xarray-2017-12-11
commit 6e5f4c4bb6069d905ccedd74c4aefe80aab35fb5
Author: Matthew Wilcox <mawilcox(a)microsoft.com>
AuthorDate: Wed Dec 13 10:37:50 2017 -0500
Commit: Matthew Wilcox <mawilcox(a)microsoft.com>
CommitDate: Mon Jan 1 19:38:59 2018 -0500
irqdomain: Convert to XArray
In a non-critical path, irqdomain wants to know how many entries are
stored in the xarray, so add xa_count(). This is a pretty straightforward
conversion; mostly just removing now-redundant locking. The only thing
of note is just how much simpler irq_domain_fix_revmap() becomes.
Signed-off-by: Matthew Wilcox <mawilcox(a)microsoft.com>
e5a49f4945 md: Convert raid5-cache to XArray
6e5f4c4bb6 irqdomain: Convert to XArray
7c7f1f88ff convert test suite
+---------------------------------------------------------+------------+------------+------------+
| | e5a49f4945 | 6e5f4c4bb6 | 7c7f1f88ff |
+---------------------------------------------------------+------------+------------+------------+
| boot_successes | 27 | 0 | 2 |
| boot_failures | 3 | 28 | 17 |
| kernel_BUG_at_mm/shmem.c | 1 | | |
| invalid_opcode:#[##] | 1 | | |
| EIP:shmem_undo_range | 1 | | |
| Kernel_panic-not_syncing:Fatal_exception | 1 | | |
| BUG:soft_lockup-CPU##stuck_for#s | 1 | 1 | |
| EIP:arch_local_irq_restore | 1 | 1 | 1 |
| Kernel_panic-not_syncing:softlockup:hung_tasks | 1 | 1 | |
| Mem-Info | 1 | 2 | 2 |
| WARNING:possible_irq_lock_inversion_dependency_detected | 0 | 27 | 17 |
| EIP:native_safe_halt | 0 | 25 | 15 |
| EIP:rcu_read_unlock | 0 | 1 | |
| BUG:workqueue_lockup-pool | 0 | 1 | |
| WARNING:at_drivers/pci/pci-sysfs.c:#pci_mmap_resource | 0 | 0 | 1 |
| EIP:pci_mmap_resource | 0 | 0 | 1 |
| EIP:arch_local_irq_enable | 0 | 0 | 1 |
+---------------------------------------------------------+------------+------------+------------+
[ 16.087487] audit: type=1326 audit(1514868349.000:2): auid=4294967295 uid=65534 gid=65534 ses=4294967295 pid=709 comm="trinity-c1" exe="/trinity" sig=9 arch=40000003 syscall=19 compat=0 ip=0x809af42 code=0x0
[ 16.192654] audit: type=1326 audit(1514868349.110:3): auid=4294967295 uid=65534 gid=65534 ses=4294967295 pid=706 comm="trinity-c2" exe="/trinity" sig=9 arch=40000003 syscall=19 compat=0 ip=0x809af42 code=0x0
[ 16.312623] audit: type=1326 audit(1514868349.230:4): auid=4294967295 uid=65534 gid=65534 ses=4294967295 pid=711 comm="trinity-c3" exe="/trinity" sig=9 arch=40000003 syscall=19 compat=0 ip=0x809af42 code=0x0
[ 17.005580]
[ 17.005924] ========================================================
[ 17.006865] WARNING: possible irq lock inversion dependency detected
[ 17.007542] 4.15.0-rc6-00087-g6e5f4c4 #1 Not tainted
[ 17.008048] --------------------------------------------------------
[ 17.008726] swapper/0/0 just changed the state of lock:
[ 17.009239] (&(&sighand->siglock)->rlock){-...}, at: [<2c8eea5c>] __lock_task_sighand+0x9b/0xe8
[ 17.010066] but this lock took another, HARDIRQ-unsafe lock in the past:
[ 17.010799] (&(&xa->xa_lock)->rlock){+.+.}
[ 17.010801]
[ 17.010801] and interrupts could create inverse lock ordering between them.
[ 17.010801]
[ 17.012396]
[ 17.012396] other info that might help us debug this:
[ 17.013128] Chain exists of:
[ 17.013128] &(&sighand->siglock)->rlock --> &(&(&sig->stats_lock)->lock)->rlock --> &(&xa->xa_lock)->rlock
[ 17.013128]
[ 17.014495] Possible interrupt unsafe locking scenario:
[ 17.014495]
[ 17.015164] CPU0 CPU1
[ 17.015624] ---- ----
[ 17.016075] lock(&(&xa->xa_lock)->rlock);
[ 17.016692] local_irq_disable();
[ 17.017420] lock(&(&sighand->siglock)->rlock);
[ 17.018134] lock(&(&(&sig->stats_lock)->lock)->rlock);
[ 17.019381] <Interrupt>
[ 17.019630] lock(&(&sighand->siglock)->rlock);
[ 17.020092]
[ 17.020092] *** DEADLOCK ***
[ 17.020092]
[ 17.020727] 2 locks held by swapper/0/0:
[ 17.021132] #0: (rcu_read_lock){....}, at: [<104f423d>] rcu_read_lock+0x0/0x5e
[ 17.021879] #1: (rcu_read_lock){....}, at: [<104f423d>] rcu_read_lock+0x0/0x5e
[ 17.022825]
[ 17.022825] the shortest dependencies between 2nd lock and 1st lock:
[ 17.023847] -> (&(&xa->xa_lock)->rlock){+.+.} ops: 81124 {
[ 17.024432] HARDIRQ-ON-W at:
[ 17.024764] lock_acquire+0x60/0x7c
[ 17.025405] _raw_spin_lock+0x22/0x31
[ 17.025986] xa_store+0x50/0x84
[ 17.026536] irq_domain_set_mapping+0x1d/0x20
[ 17.027226] __irq_domain_alloc_irqs+0x14f/0x18f
[ 17.027922] alloc_isa_irq_from_domain+0x86/0xac
[ 17.028680] mp_map_pin_to_irq+0x138/0x227
[ 17.029333] pin_2_irq+0x8f/0x9e
[ 17.030108] setup_IO_APIC+0xd3/0x62f
[ 17.031142] apic_bsp_setup+0x67/0x6e
[ 17.032169] apic_intr_mode_init+0x148/0x14d
[ 17.033044] x86_late_time_init+0xf/0x16
[ 17.033671] start_kernel+0x372/0x3e3
[ 17.034225] i386_start_kernel+0x95/0x99
[ 17.034833] startup_32_smp+0x15f/0x170
[ 17.035493] SOFTIRQ-ON-W at:
[ 17.035877] lock_acquire+0x60/0x7c
[ 17.036464] _raw_spin_lock+0x22/0x31
[ 17.037274] xa_store+0x50/0x84
[ 17.038019] irq_domain_set_mapping+0x1d/0x20
[ 17.039064] __irq_domain_alloc_irqs+0x14f/0x18f
[ 17.039955] alloc_isa_irq_from_domain+0x86/0xac
[ 17.040831] mp_map_pin_to_irq+0x138/0x227
[ 17.041779] pin_2_irq+0x8f/0x9e
[ 17.042631] setup_IO_APIC+0xd3/0x62f
[ 17.043540] apic_bsp_setup+0x67/0x6e
[ 17.044452] apic_intr_mode_init+0x148/0x14d
[ 17.045452] x86_late_time_init+0xf/0x16
[ 17.046417] start_kernel+0x372/0x3e3
[ 17.047392] i386_start_kernel+0x95/0x99
[ 17.048348] startup_32_smp+0x15f/0x170
[ 17.049348] INITIAL USE at:
[ 17.049870] lock_acquire+0x60/0x7c
[ 17.050755] _raw_spin_lock+0x22/0x31
[ 17.051656] xa_store+0x50/0x84
[ 17.052485] irq_domain_set_mapping+0x1d/0x20
[ 17.053634] __irq_domain_alloc_irqs+0x14f/0x18f
[ 17.054675] alloc_isa_irq_from_domain+0x86/0xac
[ 17.055479] mp_map_pin_to_irq+0x138/0x227
[ 17.056127] pin_2_irq+0x8f/0x9e
[ 17.056679] setup_IO_APIC+0xd3/0x62f
[ 17.057221] apic_bsp_setup+0x67/0x6e
[ 17.057782] apic_intr_mode_init+0x148/0x14d
[ 17.058445] x86_late_time_init+0xf/0x16
[ 17.059089] start_kernel+0x372/0x3e3
[ 17.059654] i386_start_kernel+0x95/0x99
[ 17.060225] startup_32_smp+0x15f/0x170
[ 17.060888] }
[ 17.061082] ... key at: [<24a7de73>] __key.14315+0x0/0xb84
[ 17.061697] ... acquired at:
[ 17.062013] _raw_spin_lock_irqsave+0x34/0x44
[ 17.062477] idr_alloc_ul+0x5e/0xc9
[ 17.062846] idr_alloc_cyclic+0x3f/0x7f
[ 17.063235] alloc_pid+0x80/0x19e
# HH:MM RESULT GOOD BAD GOOD_BUT_DIRTY DIRTY_NOT_BAD
git bisect start 8ee433aec7b1a46839f124bfe26b3d5d62d927dc 30a7acd573899fd8b8ac39236eff6468b195ac7d --
git bisect bad 37a9194ea88d636ac11001858872f86baa2b082c # 11:18 B 0 2 16 0 Merge 'm68k/master' into devel-catchup-201801021028
git bisect bad 50cf105bde938c6ed8cc0f0b0d02bed9d28f736f # 11:27 B 4 4 4 4 Merge 'tip/master' into devel-catchup-201801021028
git bisect good 8c39043ae750c5f93d15b99ab67116c94141592e # 11:47 G 12 0 8 8 0day base guard for 'devel-catchup-201801021028'
git bisect bad 4ab78e4764d67225a6fced4db0c20c7ae9b3bed1 # 11:57 B 1 1 1 1 Merge 'dax/xarray-2017-12-11' into devel-catchup-201801021028
git bisect good b00df7ac263327b60bae584f3d9735ec2d629437 # 12:14 G 12 0 5 5 shmem: Convert find_swap_entry to XArray
git bisect good 0f27d6177fb7a4b480d7b21ac487d1a37f083bab # 12:26 G 13 0 5 5 xfs: Convert m_perag_tree to XArray
git bisect bad 460a7c6f88f18183f97247094b6b05b7bfb0c842 # 12:37 B 0 1 15 0 ipc: Remove call to idr_preload
git bisect bad 6e5f4c4bb6069d905ccedd74c4aefe80aab35fb5 # 12:48 B 0 7 21 0 irqdomain: Convert to XArray
git bisect good 02171edd18243982e431712e64c5b500fec1c1f1 # 13:01 G 13 0 9 9 xfs: Convert mru cache to XArray
git bisect good e5a49f4945848cdaee779479be40e46e51b16346 # 13:12 G 13 0 9 9 md: Convert raid5-cache to XArray
# first bad commit: [6e5f4c4bb6069d905ccedd74c4aefe80aab35fb5] irqdomain: Convert to XArray
git bisect good e5a49f4945848cdaee779479be40e46e51b16346 # 13:18 G 39 0 16 25 md: Convert raid5-cache to XArray
# extra tests with debug options
git bisect bad 6e5f4c4bb6069d905ccedd74c4aefe80aab35fb5 # 13:29 B 0 2 16 0 irqdomain: Convert to XArray
# extra tests on HEAD of linux-devel/devel-catchup-201801021028
git bisect bad 8ee433aec7b1a46839f124bfe26b3d5d62d927dc # 13:30 B 0 24 49 7 0day head guard for 'devel-catchup-201801021028'
# extra tests on tree/branch dax/xarray-2017-12-11
git bisect bad 7c7f1f88ffc55b930efc4cac2fe4b3b4e26ef54f # 13:42 B 1 10 1 1 convert test suite
# extra tests with first bad commit reverted
git bisect good d7c52e4b7ce362f246385ca133bbae39ed50ca7c # 13:57 G 13 0 8 9 Revert "irqdomain: Convert to XArray"
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/lkp Intel Corporation
4 years, 4 months
- 1
- 0
[lkp-robot] [x86/entry] 863b74d09b: netperf.Throughput_tps 5.6% improvement
by kernel test robot
Greeting,
FYI, we noticed a 5.6% improvement of netperf.Throughput_tps due to commit:
commit: 863b74d09b1de0596de307e8a466a94be8d57407 ("x86/entry: Align entry text section to PMD boundary")
https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master
in testcase: netperf
on test machine: 4 threads Intel(R) Core(TM) i5-3317U CPU @ 1.70GHz with 4G memory
with following parameters:
ip: ipv4
runtime: 300s
nr_threads: 1
cluster: cs-localhost
test: UDP_RR
cpufreq_governor: performance
test-description: Netperf is a benchmark that can be use to measure various aspect of networking performance.
test-url: http://www.netperf.org/netperf/
In addition to that, the commit also has significant impact on the following tests:
+------------------+-------------------------------------------------------------------+
| testcase: change | netperf: netperf.Throughput_tps 5.5% improvement |
| test machine | 4 threads Intel(R) Core(TM) i5-3317U CPU @ 1.70GHz with 4G memory |
| test parameters | cluster=cs-localhost |
| | cpufreq_governor=performance |
| | ip=ipv4 |
| | nr_threads=25% |
| | runtime=300s |
| | test=UDP_RR |
+------------------+-------------------------------------------------------------------+
| testcase: change | netperf: netperf.Throughput_tps 2.9% improvement |
| test machine | 4 threads Intel(R) Core(TM) i5-3317U CPU @ 1.70GHz with 4G memory |
| test parameters | cluster=cs-localhost |
| | cpufreq_governor=performance |
| | ip=ipv4 |
| | nr_threads=1 |
| | runtime=300s |
| | test=TCP_CRR |
+------------------+-------------------------------------------------------------------+
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:
cs-localhost/gcc-7/performance/ipv4/x86_64-rhel-7.2/1/debian-x86_64-2016-08-31.cgz/300s/lkp-u410/UDP_RR/netperf
commit:
dacae0a34b ("x86/mm/pti: Share cpu_entry_area with user space page tables")
863b74d09b ("x86/entry: Align entry text section to PMD boundary")
dacae0a34b101a69 863b74d09b1de0596de307e8a4
---------------- --------------------------
%stddev %change %stddev
\ | \
71553 +5.6% 75553 netperf.Throughput_tps
60.00 +1.7% 61.00 netperf.time.percent_of_cpu_this_job_got
12.13 ± 2% +4.3% 12.65 ± 2% netperf.time.user_time
21458458 +5.6% 22659135 netperf.time.voluntary_context_switches
132824 ± 4% -6.3% 124510 ± 5% meminfo.Active
927.25 ± 2% +7.9% 1000 ± 3% slabinfo.kmalloc-1024.active_objs
309593 +5.1% 325391 vmstat.system.cs
841604 ± 12% +92.0% 1616130 ± 20% cpuidle.POLL.time
54502 +44.2% 78607 ± 14% cpuidle.POLL.usage
974.75 +2.3% 997.25 turbostat.Avg_MHz
6.96 +1.5% 7.06 turbostat.CorWatt
18877 ±114% -69.7% 5713 ± 8% sched_debug.cpu.avg_idle.min
455.00 ± 2% +9.6% 498.58 ± 4% sched_debug.cpu.cpu_load[3].max
15123 ± 13% -29.4% 10677 ± 13% sched_debug.cpu.nr_load_updates.stddev
8621576 ± 4% +18.1% 10179775 ± 3% sched_debug.cpu.nr_switches.min
2339881 ± 13% -25.6% 1740106 ± 15% sched_debug.cpu.nr_switches.stddev
6.667e+09 ± 4% +8.4% 7.23e+09 ± 3% perf-stat.cache-references
94092065 +5.1% 98897208 perf-stat.context-switches
1.77 -2.0% 1.73 perf-stat.cpi
0.26 ± 10% +0.2 0.42 ± 18% perf-stat.dTLB-load-miss-rate%
5.223e+08 ± 9% +71.1% 8.936e+08 ± 20% perf-stat.dTLB-load-misses
1.97e+11 ± 2% +7.1% 2.111e+11 ± 2% perf-stat.dTLB-loads
20.95 +47.2 68.18 perf-stat.iTLB-load-miss-rate%
5.576e+08 ± 2% +72.2% 9.6e+08 perf-stat.iTLB-load-misses
2.103e+09 ± 2% -78.7% 4.481e+08 perf-stat.iTLB-loads
1372 ± 5% -39.9% 825.63 ± 5% perf-stat.instructions-per-iTLB-miss
0.57 +2.0% 0.58 perf-stat.ipc
10.77 ± 6% -10.8 0.00 perf-profile.calltrace.cycles-pp.__do_softirq.do_softirq_own_stack.do_softirq.__local_bh_enable_ip.ip_finish_output2
10.23 ± 6% -10.2 0.00 perf-profile.calltrace.cycles-pp.net_rx_action.__do_softirq.do_softirq_own_stack.do_softirq.__local_bh_enable_ip
9.71 ± 7% -9.7 0.00 perf-profile.calltrace.cycles-pp.process_backlog.net_rx_action.__do_softirq.do_softirq_own_stack.do_softirq
9.32 ± 7% -9.3 0.00 perf-profile.calltrace.cycles-pp.__netif_receive_skb_core.process_backlog.net_rx_action.__do_softirq.do_softirq_own_stack
8.64 ± 7% -8.6 0.00 perf-profile.calltrace.cycles-pp.ip_rcv.__netif_receive_skb_core.process_backlog.net_rx_action.__do_softirq
13.29 ± 13% -1.4 11.91 ± 29% perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_kernel.secondary_startup_64
13.29 ± 13% -1.4 11.91 ± 29% perf-profile.calltrace.cycles-pp.start_kernel.secondary_startup_64
13.27 ± 13% -1.4 11.90 ± 29% perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_kernel.secondary_startup_64
10.24 ± 15% -0.8 9.48 ± 33% perf-profile.calltrace.cycles-pp.cpuidle_enter_state.do_idle.cpu_startup_entry.start_kernel.secondary_startup_64
35.77 ± 3% -0.7 35.05 ± 9% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_fastpath
10.03 ± 16% -0.7 9.31 ± 33% perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.do_idle.cpu_startup_entry.start_kernel
11.62 ± 2% -0.7 10.90 ± 8% perf-profile.calltrace.cycles-pp.SYSC_recvfrom.entry_SYSCALL_64_fastpath
8.06 ± 6% -0.6 7.50 ± 10% perf-profile.calltrace.cycles-pp.ip_local_deliver_finish.ip_local_deliver.ip_rcv.__netif_receive_skb_core.process_backlog
8.09 ± 7% -0.5 7.54 ± 10% perf-profile.calltrace.cycles-pp.ip_local_deliver.ip_rcv.__netif_receive_skb_core.process_backlog.net_rx_action
11.09 ± 7% -0.5 10.58 ± 11% perf-profile.calltrace.cycles-pp.do_softirq.__local_bh_enable_ip.ip_finish_output2.ip_output.ip_send_skb
11.25 ± 7% -0.5 10.74 ± 11% perf-profile.calltrace.cycles-pp.__local_bh_enable_ip.ip_finish_output2.ip_output.ip_send_skb.udp_send_skb
10.92 ± 7% -0.5 10.41 ± 11% perf-profile.calltrace.cycles-pp.do_softirq_own_stack.do_softirq.__local_bh_enable_ip.ip_finish_output2.ip_output
10.25 ± 4% -0.5 9.75 ± 7% perf-profile.calltrace.cycles-pp.inet_recvmsg.SYSC_recvfrom.entry_SYSCALL_64_fastpath
8.06 ± 5% -0.5 7.57 ± 7% perf-profile.calltrace.cycles-pp.__skb_recv_udp.udp_recvmsg.inet_recvmsg.SYSC_recvfrom.entry_SYSCALL_64_fastpath
5.86 ± 6% -0.5 5.40 ± 7% perf-profile.calltrace.cycles-pp.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_udp.udp_recvmsg.inet_recvmsg
7.67 ± 7% -0.4 7.27 ± 10% perf-profile.calltrace.cycles-pp.__udp4_lib_rcv.ip_local_deliver_finish.ip_local_deliver.ip_rcv.__netif_receive_skb_core
6.70 ± 6% -0.4 6.32 ± 8% perf-profile.calltrace.cycles-pp.__skb_wait_for_more_packets.__skb_recv_udp.udp_recvmsg.inet_recvmsg.SYSC_recvfrom
10.07 ± 4% -0.3 9.72 ± 7% perf-profile.calltrace.cycles-pp.udp_recvmsg.inet_recvmsg.SYSC_recvfrom.entry_SYSCALL_64_fastpath
5.61 ± 6% -0.3 5.28 ± 7% perf-profile.calltrace.cycles-pp.schedule.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_udp.udp_recvmsg
14.07 ± 7% -0.3 13.74 ± 10% perf-profile.calltrace.cycles-pp.ip_finish_output2.ip_output.ip_send_skb.udp_send_skb.udp_sendmsg
16.21 ± 6% -0.3 15.92 ± 9% perf-profile.calltrace.cycles-pp.udp_send_skb.udp_sendmsg.sock_sendmsg.SYSC_sendto.entry_SYSCALL_64_fastpath
5.47 ± 6% -0.3 5.18 ± 7% perf-profile.calltrace.cycles-pp.__sched_text_start.schedule.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_udp
15.92 ± 6% -0.3 15.63 ± 9% perf-profile.calltrace.cycles-pp.ip_send_skb.udp_send_skb.udp_sendmsg.sock_sendmsg.SYSC_sendto
5.93 ± 7% -0.3 5.65 ± 9% perf-profile.calltrace.cycles-pp.__udp_enqueue_schedule_skb.udp_queue_rcv_skb.__udp4_lib_rcv.ip_local_deliver_finish.ip_local_deliver
5.53 ± 7% -0.3 5.27 ± 9% perf-profile.calltrace.cycles-pp.sock_def_readable.__udp_enqueue_schedule_skb.udp_queue_rcv_skb.__udp4_lib_rcv.ip_local_deliver_finish
6.90 ± 6% -0.2 6.66 ± 9% perf-profile.calltrace.cycles-pp.udp_queue_rcv_skb.__udp4_lib_rcv.ip_local_deliver_finish.ip_local_deliver.ip_rcv
5.13 ± 8% -0.2 4.89 ± 9% perf-profile.calltrace.cycles-pp.__wake_up_common_lock.sock_def_readable.__udp_enqueue_schedule_skb.udp_queue_rcv_skb.__udp4_lib_rcv
15.40 ± 6% -0.2 15.21 ± 9% perf-profile.calltrace.cycles-pp.ip_output.ip_send_skb.udp_send_skb.udp_sendmsg.sock_sendmsg
22.85 ± 4% -0.2 22.68 ± 9% perf-profile.calltrace.cycles-pp.udp_sendmsg.sock_sendmsg.SYSC_sendto.entry_SYSCALL_64_fastpath
23.14 ± 4% -0.1 23.08 ± 9% perf-profile.calltrace.cycles-pp.sock_sendmsg.SYSC_sendto.entry_SYSCALL_64_fastpath
23.86 ± 4% -0.1 23.81 ± 9% perf-profile.calltrace.cycles-pp.SYSC_sendto.entry_SYSCALL_64_fastpath
50.92 ± 3% +0.9 51.87 ± 8% perf-profile.calltrace.cycles-pp.secondary_startup_64
26.38 ± 5% +2.1 28.45 ± 8% perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.do_idle.cpu_startup_entry.start_secondary
27.00 ± 5% +2.3 29.30 ± 7% perf-profile.calltrace.cycles-pp.cpuidle_enter_state.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
37.64 ± 5% +2.3 39.95 ± 4% perf-profile.calltrace.cycles-pp.start_secondary.secondary_startup_64
37.63 ± 5% +2.3 39.95 ± 4% perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.secondary_startup_64
37.56 ± 5% +2.3 39.89 ± 4% perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
0.00 +8.1 8.05 ± 9% perf-profile.calltrace.cycles-pp.ip_rcv.__netif_receive_skb_core.process_backlog.net_rx_action.__softirqentry_text_start
0.00 +8.8 8.82 ± 10% perf-profile.calltrace.cycles-pp.__netif_receive_skb_core.process_backlog.net_rx_action.__softirqentry_text_start.do_softirq_own_stack
0.00 +9.2 9.22 ± 10% perf-profile.calltrace.cycles-pp.process_backlog.net_rx_action.__softirqentry_text_start.do_softirq_own_stack.do_softirq
0.00 +9.7 9.74 ± 11% perf-profile.calltrace.cycles-pp.net_rx_action.__softirqentry_text_start.do_softirq_own_stack.do_softirq.__local_bh_enable_ip
0.00 +10.3 10.27 ± 11% perf-profile.calltrace.cycles-pp.__softirqentry_text_start.do_softirq_own_stack.do_softirq.__local_bh_enable_ip.ip_finish_output2
10.91 ± 6% -10.9 0.00 perf-profile.children.cycles-pp.__do_softirq
13.29 ± 13% -1.4 11.91 ± 29% perf-profile.children.cycles-pp.start_kernel
11.73 ± 2% -0.8 10.97 ± 8% perf-profile.children.cycles-pp.SYSC_recvfrom
36.03 ± 3% -0.7 35.38 ± 9% perf-profile.children.cycles-pp.entry_SYSCALL_64_fastpath
8.71 ± 7% -0.6 8.13 ± 10% perf-profile.children.cycles-pp.ip_rcv
8.12 ± 7% -0.5 7.58 ± 10% perf-profile.children.cycles-pp.ip_local_deliver
11.04 ± 7% -0.5 10.50 ± 11% perf-profile.children.cycles-pp.do_softirq_own_stack
8.10 ± 7% -0.5 7.56 ± 10% perf-profile.children.cycles-pp.ip_local_deliver_finish
10.29 ± 4% -0.5 9.77 ± 7% perf-profile.children.cycles-pp.inet_recvmsg
11.15 ± 7% -0.5 10.64 ± 11% perf-profile.children.cycles-pp.do_softirq
9.34 ± 7% -0.5 8.84 ± 10% perf-profile.children.cycles-pp.__netif_receive_skb_core
9.77 ± 6% -0.5 9.28 ± 10% perf-profile.children.cycles-pp.process_backlog
10.27 ± 6% -0.5 9.78 ± 11% perf-profile.children.cycles-pp.net_rx_action
11.45 ± 7% -0.5 10.97 ± 10% perf-profile.children.cycles-pp.__local_bh_enable_ip
8.17 ± 5% -0.5 7.70 ± 7% perf-profile.children.cycles-pp.__skb_recv_udp
7.74 ± 7% -0.4 7.32 ± 10% perf-profile.children.cycles-pp.__udp4_lib_rcv
9.57 ± 2% -0.4 9.18 ± 5% perf-profile.children.cycles-pp.__sched_text_start
6.76 ± 6% -0.4 6.38 ± 8% perf-profile.children.cycles-pp.__skb_wait_for_more_packets
10.14 ± 4% -0.4 9.77 ± 7% perf-profile.children.cycles-pp.udp_recvmsg
5.67 ± 6% -0.4 5.31 ± 7% perf-profile.children.cycles-pp.schedule
14.10 ± 7% -0.3 13.76 ± 10% perf-profile.children.cycles-pp.ip_finish_output2
6.00 ± 7% -0.3 5.70 ± 9% perf-profile.children.cycles-pp.__udp_enqueue_schedule_skb
16.25 ± 6% -0.3 15.95 ± 10% perf-profile.children.cycles-pp.udp_send_skb
15.95 ± 6% -0.3 15.66 ± 9% perf-profile.children.cycles-pp.ip_send_skb
5.54 ± 7% -0.3 5.28 ± 9% perf-profile.children.cycles-pp.sock_def_readable
23.07 ± 4% -0.2 22.84 ± 9% perf-profile.children.cycles-pp.udp_sendmsg
6.99 ± 6% -0.2 6.76 ± 9% perf-profile.children.cycles-pp.udp_queue_rcv_skb
5.20 ± 8% -0.2 4.97 ± 9% perf-profile.children.cycles-pp.__wake_up_common_lock
15.53 ± 6% -0.1 15.38 ± 9% perf-profile.children.cycles-pp.ip_output
23.94 ± 4% -0.1 23.88 ± 9% perf-profile.children.cycles-pp.SYSC_sendto
23.22 ± 4% -0.1 23.16 ± 9% perf-profile.children.cycles-pp.sock_sendmsg
50.92 ± 3% +0.9 51.87 ± 8% perf-profile.children.cycles-pp.secondary_startup_64
51.05 ± 3% +1.0 52.01 ± 8% perf-profile.children.cycles-pp.do_idle
50.94 ± 3% +1.0 51.90 ± 8% perf-profile.children.cycles-pp.cpu_startup_entry
36.43 ± 4% +1.3 37.77 ± 13% perf-profile.children.cycles-pp.intel_idle
37.34 ± 4% +1.5 38.84 ± 12% perf-profile.children.cycles-pp.cpuidle_enter_state
37.64 ± 5% +2.3 39.95 ± 4% perf-profile.children.cycles-pp.start_secondary
0.00 +10.4 10.39 ± 10% perf-profile.children.cycles-pp.__softirqentry_text_start
36.43 ± 4% +1.3 37.77 ± 13% perf-profile.self.cycles-pp.intel_idle
perf-stat.iTLB-loads
3e+09 +-+---------------------------------------------------------------+
| .++.++ |
2.5e+09 +-++. .++. .+ + + +. .++.++.++.++.++ : |
| : ++ ++.++ : :+.+: : ++ : |
| : : : : : ++.++.++.++.+|
2e+09 +-+: : : : : |
|: : : : : : |
1.5e+09 +-+: : : : : |
|: : :: : : |
1e+09 +-+ :: :: |
|:: :: :: |
| : :: : |
5e+08 O-OO OO OO OO OO OO OO OO OO OO OO OO OO OO |
| : : : |
0 +-+---------------------------------------------------------------+
perf-stat.iTLB-load-misses
1.2e+09 O-+---------------------------------------------------------------+
| OO OO |
1e+09 +-+ O O OO |
| O O OO OO OO OO OO OO OO OO OO |
| |
8e+08 +-+ |
| |
6e+08 +-+ +. +. |
| +.++.++.++.++.+ ++.++ +.++.++.++.++.++.++. +. +.++.+ + ++.+|
4e+08 +-+: : : : : + + |
|: : : : : : |
|: : :: : : |
2e+08 +-+ :: :: |
| : : : |
0 +-+---------------------------------------------------------------+
perf-stat.iTLB-load-miss-rate_
80 +-+--------------------------------------------------------------------+
O OO OO |
70 +-+ O OO O OO OO O OO OO O OO O OO OO O OO |
60 +-+ |
| |
50 +-+ |
| |
40 +-+ |
| |
30 +-+ |
20 +-+ .++.+.++.++.|
| +.++.+.++.+.++.+ +.++.+ +.++.+.++.++.+.++.++. .+ .++.+ |
10 +-+: : : : : + + |
|:: :: :: |
0 +-+--------------------------------------------------------------------+
perf-stat.instructions-per-iTLB-miss
1800 +-+------------------------------------------------------------------+
| + :: + + ++.++ |
1600 +-+ :+ :: + + + :+ +.+ .++. + + .+ |
1400 +-++.+ ++.+.: :: +.+ : +.+ +.++. : + + ++. .++ +.+ |
| : + +: : : : + + +.|
1200 +-+: : : : : |
1000 +-+: : : : : |
|: : : : : : O O |
800 +-+: OO O OO OO:OO O OO:OO OO OO OO O O |
600 O-OO OO :: :: |
|:: :: :: |
400 +-+ : : |
200 +-+ : : |
| : : : |
0 +-+------------------------------------------------------------------+
[*] bisect-good sample
[O] bisect-bad sample
***************************************************************************************************
lkp-u410: 4 threads Intel(R) Core(TM) i5-3317U CPU @ 1.70GHz with 4G memory
=========================================================================================
cluster/compiler/cpufreq_governor/ip/kconfig/nr_threads/rootfs/runtime/tbox_group/test/testcase:
cs-localhost/gcc-7/performance/ipv4/x86_64-rhel-7.2/25%/debian-x86_64-2016-08-31.cgz/300s/lkp-u410/UDP_RR/netperf
commit:
dacae0a34b ("x86/mm/pti: Share cpu_entry_area with user space page tables")
863b74d09b ("x86/entry: Align entry text section to PMD boundary")
dacae0a34b101a69 863b74d09b1de0596de307e8a4
---------------- --------------------------
%stddev %change %stddev
\ | \
71338 +5.5% 75260 netperf.Throughput_tps
12.08 +6.1% 12.82 netperf.time.user_time
21394533 +5.5% 22572349 netperf.time.voluntary_context_switches
57288 ± 5% +22.6% 70207 ± 2% cpuidle.POLL.usage
308598 +5.0% 323970 vmstat.system.cs
2122 ± 4% +10.0% 2334 ± 2% slabinfo.anon_vma.active_objs
2160 ± 3% -8.3% 1980 ± 5% slabinfo.kmalloc-256.num_objs
975.00 +2.2% 996.75 turbostat.Avg_MHz
6.97 +1.4% 7.07 turbostat.CorWatt
194.66 ± 15% +21.9% 237.21 ± 13% sched_debug.cpu.cpu_load[0].avg
434603 ± 8% -24.1% 329920 ± 16% sched_debug.cpu.load.avg
781960 ± 9% -20.9% 618646 ± 12% sched_debug.cpu.load.max
1.31 ± 15% -29.4% 0.93 ± 20% sched_debug.cpu.nr_running.avg
93871893 +5.0% 98542678 perf-stat.context-switches
1.77 -2.2% 1.74 perf-stat.cpi
0.25 ± 8% +0.2 0.43 ± 15% perf-stat.dTLB-load-miss-rate%
4.962e+08 ± 8% +78.5% 8.86e+08 ± 15% perf-stat.dTLB-load-misses
1.973e+11 ± 2% +4.1% 2.054e+11 ± 3% perf-stat.dTLB-loads
1.21e+11 ± 4% +7.4% 1.3e+11 ± 2% perf-stat.dTLB-stores
21.11 +47.1 68.23 perf-stat.iTLB-load-miss-rate%
5.631e+08 +71.6% 9.665e+08 perf-stat.iTLB-load-misses
2.105e+09 -78.6% 4.501e+08 perf-stat.iTLB-loads
1346 ± 4% -39.1% 819.37 ± 3% perf-stat.instructions-per-iTLB-miss
0.56 +2.2% 0.58 perf-stat.ipc
325845 -0.9% 322993 perf-stat.minor-faults
325845 -0.9% 322994 perf-stat.page-faults
10.35 ± 4% -10.4 0.00 perf-profile.calltrace.cycles-pp.__do_softirq.do_softirq_own_stack.do_softirq.__local_bh_enable_ip.ip_finish_output2
9.87 ± 4% -9.9 0.00 perf-profile.calltrace.cycles-pp.net_rx_action.__do_softirq.do_softirq_own_stack.do_softirq.__local_bh_enable_ip
9.33 ± 4% -9.3 0.00 perf-profile.calltrace.cycles-pp.process_backlog.net_rx_action.__do_softirq.do_softirq_own_stack.do_softirq
8.96 ± 4% -9.0 0.00 perf-profile.calltrace.cycles-pp.__netif_receive_skb_core.process_backlog.net_rx_action.__do_softirq.do_softirq_own_stack
8.29 ± 4% -8.3 0.00 perf-profile.calltrace.cycles-pp.ip_rcv.__netif_receive_skb_core.process_backlog.net_rx_action.__do_softirq
10.18 ± 10% -2.8 7.35 ± 9% perf-profile.calltrace.cycles-pp.cpuidle_enter_state.do_idle.cpu_startup_entry.start_kernel.secondary_startup_64
9.98 ± 11% -2.8 7.17 ± 10% perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.do_idle.cpu_startup_entry.start_kernel
12.46 ± 11% -2.4 10.07 ± 11% perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_kernel.secondary_startup_64
12.47 ± 11% -2.4 10.09 ± 11% perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_kernel.secondary_startup_64
12.47 ± 11% -2.4 10.09 ± 11% perf-profile.calltrace.cycles-pp.start_kernel.secondary_startup_64
51.73 ± 5% -1.9 49.85 ± 4% perf-profile.calltrace.cycles-pp.secondary_startup_64
7.79 ± 4% -0.2 7.61 ± 3% perf-profile.calltrace.cycles-pp.ip_local_deliver_finish.ip_local_deliver.ip_rcv.__netif_receive_skb_core.process_backlog
7.83 ± 4% -0.2 7.65 ± 3% perf-profile.calltrace.cycles-pp.ip_local_deliver.ip_rcv.__netif_receive_skb_core.process_backlog.net_rx_action
7.38 ± 4% -0.0 7.37 ± 3% perf-profile.calltrace.cycles-pp.__udp4_lib_rcv.ip_local_deliver_finish.ip_local_deliver.ip_rcv.__netif_receive_skb_core
4.95 ± 4% +0.0 4.97 ± 4% perf-profile.calltrace.cycles-pp.__wake_up_common_lock.sock_def_readable.__udp_enqueue_schedule_skb.udp_queue_rcv_skb.__udp4_lib_rcv
5.28 ± 4% +0.0 5.33 ± 4% perf-profile.calltrace.cycles-pp.sock_def_readable.__udp_enqueue_schedule_skb.udp_queue_rcv_skb.__udp4_lib_rcv.ip_local_deliver_finish
5.64 ± 5% +0.1 5.73 ± 3% perf-profile.calltrace.cycles-pp.__udp_enqueue_schedule_skb.udp_queue_rcv_skb.__udp4_lib_rcv.ip_local_deliver_finish.ip_local_deliver
11.13 ± 5% +0.1 11.27 perf-profile.calltrace.cycles-pp.SYSC_recvfrom.entry_SYSCALL_64_fastpath
6.59 ± 4% +0.2 6.80 ± 3% perf-profile.calltrace.cycles-pp.udp_queue_rcv_skb.__udp4_lib_rcv.ip_local_deliver_finish.ip_local_deliver.ip_rcv
9.81 ± 4% +0.3 10.09 perf-profile.calltrace.cycles-pp.inet_recvmsg.SYSC_recvfrom.entry_SYSCALL_64_fastpath
7.47 ± 5% +0.3 7.82 ± 2% perf-profile.calltrace.cycles-pp.__skb_recv_udp.udp_recvmsg.inet_recvmsg.SYSC_recvfrom.entry_SYSCALL_64_fastpath
10.51 ± 4% +0.4 10.92 ± 4% perf-profile.calltrace.cycles-pp.do_softirq_own_stack.do_softirq.__local_bh_enable_ip.ip_finish_output2.ip_output
9.61 ± 4% +0.4 10.06 perf-profile.calltrace.cycles-pp.udp_recvmsg.inet_recvmsg.SYSC_recvfrom.entry_SYSCALL_64_fastpath
39.26 ± 3% +0.5 39.76 ± 4% perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.secondary_startup_64
39.26 ± 3% +0.5 39.76 ± 4% perf-profile.calltrace.cycles-pp.start_secondary.secondary_startup_64
39.19 ± 3% +0.5 39.70 ± 4% perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
10.68 ± 4% +0.5 11.20 ± 3% perf-profile.calltrace.cycles-pp.do_softirq.__local_bh_enable_ip.ip_finish_output2.ip_output.ip_send_skb
10.78 ± 4% +0.6 11.33 ± 3% perf-profile.calltrace.cycles-pp.__local_bh_enable_ip.ip_finish_output2.ip_output.ip_send_skb.udp_send_skb
5.17 ± 7% +0.6 5.75 ± 5% perf-profile.calltrace.cycles-pp.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_udp.udp_recvmsg.inet_recvmsg
5.98 ± 6% +0.6 6.59 ± 3% perf-profile.calltrace.cycles-pp.__skb_wait_for_more_packets.__skb_recv_udp.udp_recvmsg.inet_recvmsg.SYSC_recvfrom
4.96 ± 8% +0.6 5.60 ± 5% perf-profile.calltrace.cycles-pp.schedule.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_udp.udp_recvmsg
4.84 ± 7% +0.7 5.49 ± 5% perf-profile.calltrace.cycles-pp.__sched_text_start.schedule.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_udp
15.82 ± 5% +0.8 16.61 ± 4% perf-profile.calltrace.cycles-pp.udp_send_skb.udp_sendmsg.sock_sendmsg.SYSC_sendto.entry_SYSCALL_64_fastpath
13.62 ± 5% +0.8 14.43 ± 3% perf-profile.calltrace.cycles-pp.ip_finish_output2.ip_output.ip_send_skb.udp_send_skb.udp_sendmsg
15.49 ± 5% +0.8 16.29 ± 3% perf-profile.calltrace.cycles-pp.ip_send_skb.udp_send_skb.udp_sendmsg.sock_sendmsg.SYSC_sendto
15.03 ± 5% +0.8 15.86 ± 4% perf-profile.calltrace.cycles-pp.ip_output.ip_send_skb.udp_send_skb.udp_sendmsg.sock_sendmsg
27.46 ± 9% +1.1 28.52 ± 5% perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.do_idle.cpu_startup_entry.start_secondary
22.48 ± 5% +1.1 23.60 ± 6% perf-profile.calltrace.cycles-pp.udp_sendmsg.sock_sendmsg.SYSC_sendto.entry_SYSCALL_64_fastpath
28.08 ± 9% +1.2 29.24 ± 6% perf-profile.calltrace.cycles-pp.cpuidle_enter_state.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
23.62 ± 5% +1.2 24.85 ± 6% perf-profile.calltrace.cycles-pp.SYSC_sendto.entry_SYSCALL_64_fastpath
22.84 ± 5% +1.2 24.09 ± 6% perf-profile.calltrace.cycles-pp.sock_sendmsg.SYSC_sendto.entry_SYSCALL_64_fastpath
35.01 ± 5% +1.5 36.47 ± 4% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_fastpath
0.00 +8.3 8.29 ± 3% perf-profile.calltrace.cycles-pp.ip_rcv.__netif_receive_skb_core.process_backlog.net_rx_action.__softirqentry_text_start
0.00 +9.1 9.10 ± 3% perf-profile.calltrace.cycles-pp.__netif_receive_skb_core.process_backlog.net_rx_action.__softirqentry_text_start.do_softirq_own_stack
0.00 +9.6 9.55 ± 4% perf-profile.calltrace.cycles-pp.process_backlog.net_rx_action.__softirqentry_text_start.do_softirq_own_stack.do_softirq
0.00 +10.1 10.13 ± 3% perf-profile.calltrace.cycles-pp.net_rx_action.__softirqentry_text_start.do_softirq_own_stack.do_softirq.__local_bh_enable_ip
0.00 +10.8 10.78 ± 4% perf-profile.calltrace.cycles-pp.__softirqentry_text_start.do_softirq_own_stack.do_softirq.__local_bh_enable_ip.ip_finish_output2
10.52 ± 5% -10.5 0.00 perf-profile.children.cycles-pp.__do_softirq
12.47 ± 11% -2.4 10.09 ± 11% perf-profile.children.cycles-pp.start_kernel
51.74 ± 5% -1.9 49.86 ± 4% perf-profile.children.cycles-pp.cpu_startup_entry
51.73 ± 5% -1.9 49.85 ± 4% perf-profile.children.cycles-pp.secondary_startup_64
51.84 ± 5% -1.9 49.97 ± 4% perf-profile.children.cycles-pp.do_idle
37.47 ± 8% -1.8 35.72 ± 6% perf-profile.children.cycles-pp.intel_idle
38.35 ± 8% -1.7 36.67 ± 6% perf-profile.children.cycles-pp.cpuidle_enter_state
7.88 ± 4% -0.2 7.70 ± 3% perf-profile.children.cycles-pp.ip_local_deliver
7.83 ± 4% -0.2 7.66 ± 3% perf-profile.children.cycles-pp.ip_local_deliver_finish
7.47 ± 5% -0.0 7.44 ± 3% perf-profile.children.cycles-pp.__udp4_lib_rcv
8.35 ± 4% +0.0 8.36 ± 3% perf-profile.children.cycles-pp.ip_rcv
5.01 ± 4% +0.0 5.03 ± 4% perf-profile.children.cycles-pp.__wake_up_common_lock
5.29 ± 4% +0.0 5.33 ± 4% perf-profile.children.cycles-pp.sock_def_readable
5.68 ± 5% +0.1 5.80 ± 3% perf-profile.children.cycles-pp.__udp_enqueue_schedule_skb
11.22 ± 5% +0.1 11.34 perf-profile.children.cycles-pp.SYSC_recvfrom
8.96 ± 4% +0.1 9.11 ± 3% perf-profile.children.cycles-pp.__netif_receive_skb_core
9.38 ± 4% +0.2 9.61 ± 4% perf-profile.children.cycles-pp.process_backlog
6.65 ± 4% +0.2 6.88 ± 3% perf-profile.children.cycles-pp.udp_queue_rcv_skb
9.86 ± 4% +0.3 10.15 perf-profile.children.cycles-pp.inet_recvmsg
9.92 ± 4% +0.3 10.20 ± 3% perf-profile.children.cycles-pp.net_rx_action
7.58 ± 5% +0.4 7.95 ± 2% perf-profile.children.cycles-pp.__skb_recv_udp
10.63 ± 4% +0.4 11.03 ± 4% perf-profile.children.cycles-pp.do_softirq_own_stack
9.65 ± 4% +0.5 10.11 perf-profile.children.cycles-pp.udp_recvmsg
39.26 ± 3% +0.5 39.76 ± 4% perf-profile.children.cycles-pp.start_secondary
9.00 ± 4% +0.5 9.52 perf-profile.children.cycles-pp.__sched_text_start
10.72 ± 4% +0.5 11.26 ± 3% perf-profile.children.cycles-pp.do_softirq
10.99 ± 4% +0.6 11.56 ± 3% perf-profile.children.cycles-pp.__local_bh_enable_ip
6.04 ± 6% +0.6 6.65 ± 3% perf-profile.children.cycles-pp.__skb_wait_for_more_packets
4.98 ± 7% +0.7 5.65 ± 5% perf-profile.children.cycles-pp.schedule
15.52 ± 5% +0.8 16.31 ± 3% perf-profile.children.cycles-pp.ip_send_skb
15.85 ± 5% +0.8 16.65 ± 4% perf-profile.children.cycles-pp.udp_send_skb
13.65 ± 5% +0.8 14.45 ± 3% perf-profile.children.cycles-pp.ip_finish_output2
15.18 ± 5% +0.8 16.00 ± 4% perf-profile.children.cycles-pp.ip_output
22.72 ± 5% +1.0 23.75 ± 6% perf-profile.children.cycles-pp.udp_sendmsg
23.69 ± 5% +1.2 24.92 ± 6% perf-profile.children.cycles-pp.SYSC_sendto
22.92 ± 5% +1.3 24.19 ± 6% perf-profile.children.cycles-pp.sock_sendmsg
35.36 ± 5% +1.4 36.75 ± 4% perf-profile.children.cycles-pp.entry_SYSCALL_64_fastpath
0.00 +10.9 10.88 ± 4% perf-profile.children.cycles-pp.__softirqentry_text_start
37.47 ± 8% -1.8 35.72 ± 6% perf-profile.self.cycles-pp.intel_idle
***************************************************************************************************
lkp-u410: 4 threads Intel(R) Core(TM) i5-3317U CPU @ 1.70GHz with 4G memory
=========================================================================================
cluster/compiler/cpufreq_governor/ip/kconfig/nr_threads/rootfs/runtime/tbox_group/test/testcase:
cs-localhost/gcc-7/performance/ipv4/x86_64-rhel-7.2/1/debian-x86_64-2016-08-31.cgz/300s/lkp-u410/TCP_CRR/netperf
commit:
dacae0a34b ("x86/mm/pti: Share cpu_entry_area with user space page tables")
863b74d09b ("x86/entry: Align entry text section to PMD boundary")
dacae0a34b101a69 863b74d09b1de0596de307e8a4
---------------- --------------------------
%stddev %change %stddev
\ | \
16127 +2.9% 16600 netperf.Throughput_tps
85.00 +1.2% 86.00 netperf.time.percent_of_cpu_this_job_got
11.54 -10.4% 10.34 netperf.time.user_time
9506374 +3.1% 9798407 netperf.time.voluntary_context_switches
148405 ± 3% +4.5% 155105 vmstat.system.cs
1242 ± 54% -67.2% 407.00 ±173% meminfo.Mlocked
1242 ± 54% -67.2% 407.00 ±173% meminfo.Unevictable
1230 ± 11% -18.3% 1004 ± 21% slabinfo.kmalloc-512.active_objs
1257 ± 10% -16.6% 1048 ± 16% slabinfo.kmalloc-512.num_objs
554903 ± 11% -20.7% 439944 ± 13% turbostat.C1E
1.09 ± 8% -0.2 0.86 ± 15% turbostat.C1E%
13798774 ± 12% -23.3% 10588105 ± 15% cpuidle.C1E.time
554906 ± 11% -20.7% 439945 ± 13% cpuidle.C1E.usage
2973540 ± 79% -56.2% 1302634 ± 10% cpuidle.POLL.time
292.96 ± 10% +18.7% 347.88 ± 6% sched_debug.cfs_rq:/.util_avg.min
284.51 ± 10% -14.4% 243.59 ± 6% sched_debug.cfs_rq:/.util_avg.stddev
225864 ± 20% +19.8% 270521 ± 4% sched_debug.cpu.avg_idle.avg
188340 ± 16% +26.1% 237525 ± 15% sched_debug.cpu.load.stddev
46688259 +2.5% 47873766 perf-stat.context-switches
1.88 -3.5% 1.81 perf-stat.cpi
21.83 ± 2% +44.6 66.46 perf-stat.iTLB-load-miss-rate%
3.624e+08 ± 4% +41.8% 5.139e+08 perf-stat.iTLB-load-misses
1.297e+09 -80.0% 2.594e+08 perf-stat.iTLB-loads
1911 ± 3% -27.2% 1391 ± 2% perf-stat.instructions-per-iTLB-miss
0.53 +3.6% 0.55 perf-stat.ipc
15.73 ± 2% -15.7 0.00 perf-profile.calltrace.cycles-pp.__do_softirq.do_softirq_own_stack.do_softirq.__local_bh_enable_ip.ip_finish_output2
15.36 ± 2% -15.4 0.00 perf-profile.calltrace.cycles-pp.net_rx_action.__do_softirq.do_softirq_own_stack.do_softirq.__local_bh_enable_ip
15.08 ± 2% -15.1 0.00 perf-profile.calltrace.cycles-pp.process_backlog.net_rx_action.__do_softirq.do_softirq_own_stack.do_softirq
14.73 ± 2% -14.7 0.00 perf-profile.calltrace.cycles-pp.__netif_receive_skb_core.process_backlog.net_rx_action.__do_softirq.do_softirq_own_stack
14.05 ± 2% -14.0 0.00 perf-profile.calltrace.cycles-pp.ip_rcv.__netif_receive_skb_core.process_backlog.net_rx_action.__do_softirq
46.34 ± 3% -5.6 40.72 ± 2% perf-profile.calltrace.cycles-pp.secondary_startup_64
32.84 ± 9% -5.6 27.25 ± 14% perf-profile.calltrace.cycles-pp.cpuidle_enter_state.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
32.21 ± 9% -5.4 26.77 ± 15% perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.do_idle.cpu_startup_entry.start_secondary
36.29 ± 9% -5.2 31.12 ± 13% perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.secondary_startup_64
36.29 ± 9% -5.2 31.12 ± 13% perf-profile.calltrace.cycles-pp.start_secondary.secondary_startup_64
36.25 ± 9% -5.2 31.08 ± 13% perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
10.05 ± 21% -0.4 9.61 ± 50% perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_kernel.secondary_startup_64
10.05 ± 21% -0.4 9.61 ± 50% perf-profile.calltrace.cycles-pp.start_kernel.secondary_startup_64
10.04 ± 21% -0.4 9.60 ± 50% perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_kernel.secondary_startup_64
8.31 ± 20% -0.3 7.99 ± 54% perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.do_idle.cpu_startup_entry.start_kernel
8.48 ± 20% -0.3 8.22 ± 53% perf-profile.calltrace.cycles-pp.cpuidle_enter_state.do_idle.cpu_startup_entry.start_kernel.secondary_startup_64
7.76 ± 2% +0.1 7.83 ± 7% perf-profile.calltrace.cycles-pp.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg.sock_sendmsg.SYSC_sendto
7.71 ± 2% +0.1 7.80 ± 7% perf-profile.calltrace.cycles-pp.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg.sock_sendmsg
5.61 +0.1 5.74 ± 6% perf-profile.calltrace.cycles-pp.ip_queue_xmit.tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked
5.81 ± 2% +0.2 6.04 ± 7% perf-profile.calltrace.cycles-pp.tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg
10.53 ± 2% +0.3 10.82 ± 5% perf-profile.calltrace.cycles-pp.sock_sendmsg.SYSC_sendto.entry_SYSCALL_64_fastpath
4.85 ± 4% +0.3 5.14 ± 3% perf-profile.calltrace.cycles-pp.inet_recvmsg.SYSC_recvfrom.entry_SYSCALL_64_fastpath
10.42 ± 2% +0.3 10.72 ± 5% perf-profile.calltrace.cycles-pp.tcp_sendmsg.sock_sendmsg.SYSC_sendto.entry_SYSCALL_64_fastpath
4.79 ± 4% +0.3 5.10 ± 3% perf-profile.calltrace.cycles-pp.tcp_recvmsg.inet_recvmsg.SYSC_recvfrom.entry_SYSCALL_64_fastpath
10.67 ± 2% +0.3 10.99 ± 6% perf-profile.calltrace.cycles-pp.SYSC_sendto.entry_SYSCALL_64_fastpath
5.15 ± 5% +0.3 5.48 ± 3% perf-profile.calltrace.cycles-pp.SYSC_recvfrom.entry_SYSCALL_64_fastpath
8.92 ± 2% +0.5 9.38 ± 6% perf-profile.calltrace.cycles-pp.tcp_sendmsg_locked.tcp_sendmsg.sock_sendmsg.SYSC_sendto.entry_SYSCALL_64_fastpath
4.26 ± 4% +0.6 4.88 ± 5% perf-profile.calltrace.cycles-pp.tcp_connect.tcp_v4_connect.__inet_stream_connect.inet_stream_connect.SYSC_connect
11.23 +0.7 11.97 ± 3% perf-profile.calltrace.cycles-pp.ip_finish_output2.ip_output.ip_queue_xmit.tcp_transmit_skb.tcp_write_xmit
5.05 ± 4% +0.7 5.79 ± 5% perf-profile.calltrace.cycles-pp.tcp_v4_connect.__inet_stream_connect.inet_stream_connect.SYSC_connect.entry_SYSCALL_64_fastpath
8.21 ± 2% +0.8 8.99 ± 5% perf-profile.calltrace.cycles-pp.tcp_v4_do_rcv.tcp_v4_rcv.ip_local_deliver_finish.ip_local_deliver.ip_rcv
11.52 +0.9 12.38 ± 4% perf-profile.calltrace.cycles-pp.ip_output.ip_queue_xmit.tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames
6.13 +0.9 7.01 ± 5% perf-profile.calltrace.cycles-pp.ip_queue_xmit.tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames.tcp_close
13.56 ± 3% +0.9 14.45 ± 3% perf-profile.calltrace.cycles-pp.ip_local_deliver.ip_rcv.__netif_receive_skb_core.process_backlog.net_rx_action
13.51 ± 3% +0.9 14.41 ± 3% perf-profile.calltrace.cycles-pp.ip_local_deliver_finish.ip_local_deliver.ip_rcv.__netif_receive_skb_core.process_backlog
13.30 ± 3% +0.9 14.23 ± 3% perf-profile.calltrace.cycles-pp.tcp_v4_rcv.ip_local_deliver_finish.ip_local_deliver.ip_rcv.__netif_receive_skb_core
6.38 +1.0 7.34 ± 5% perf-profile.calltrace.cycles-pp.tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames.tcp_close.inet_release
16.02 ± 2% +1.2 17.25 ± 3% perf-profile.calltrace.cycles-pp.__local_bh_enable_ip.ip_finish_output2.ip_output.ip_queue_xmit.tcp_transmit_skb
15.99 ± 2% +1.2 17.23 ± 3% perf-profile.calltrace.cycles-pp.do_softirq.__local_bh_enable_ip.ip_finish_output2.ip_output.ip_queue_xmit
7.45 ± 2% +1.2 8.69 ± 5% perf-profile.calltrace.cycles-pp.__tcp_push_pending_frames.tcp_close.inet_release.sock_release.sock_close
7.41 ± 2% +1.2 8.66 ± 5% perf-profile.calltrace.cycles-pp.tcp_write_xmit.__tcp_push_pending_frames.tcp_close.inet_release.sock_release
15.83 ± 2% +1.3 17.08 ± 3% perf-profile.calltrace.cycles-pp.do_softirq_own_stack.do_softirq.__local_bh_enable_ip.ip_finish_output2.ip_output
9.46 ± 4% +1.4 10.81 ± 5% perf-profile.calltrace.cycles-pp.__inet_stream_connect.inet_stream_connect.SYSC_connect.entry_SYSCALL_64_fastpath
9.51 ± 4% +1.4 10.86 ± 4% perf-profile.calltrace.cycles-pp.inet_stream_connect.SYSC_connect.entry_SYSCALL_64_fastpath
9.66 ± 4% +1.4 11.08 ± 4% perf-profile.calltrace.cycles-pp.SYSC_connect.entry_SYSCALL_64_fastpath
9.21 ± 2% +1.5 10.67 ± 6% perf-profile.calltrace.cycles-pp.inet_release.sock_release.sock_close.__fput.task_work_run
9.27 ± 2% +1.5 10.75 ± 6% perf-profile.calltrace.cycles-pp.sock_close.__fput.task_work_run.exit_to_usermode_loop.syscall_return_slowpath
8.99 ± 2% +1.5 10.47 ± 5% perf-profile.calltrace.cycles-pp.tcp_close.inet_release.sock_release.sock_close.__fput
9.25 ± 2% +1.5 10.73 ± 6% perf-profile.calltrace.cycles-pp.sock_release.sock_close.__fput.task_work_run.exit_to_usermode_loop
10.96 ± 2% +1.6 12.57 ± 5% perf-profile.calltrace.cycles-pp.task_work_run.exit_to_usermode_loop.syscall_return_slowpath.entry_SYSCALL_64_fastpath
10.76 ± 2% +1.6 12.38 ± 5% perf-profile.calltrace.cycles-pp.__fput.task_work_run.exit_to_usermode_loop.syscall_return_slowpath.entry_SYSCALL_64_fastpath
11.05 ± 2% +1.6 12.68 ± 5% perf-profile.calltrace.cycles-pp.exit_to_usermode_loop.syscall_return_slowpath.entry_SYSCALL_64_fastpath
11.12 ± 2% +1.6 12.75 ± 5% perf-profile.calltrace.cycles-pp.syscall_return_slowpath.entry_SYSCALL_64_fastpath
42.12 ± 2% +4.3 46.40 ± 2% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_fastpath
0.00 +15.0 15.00 ± 3% perf-profile.calltrace.cycles-pp.ip_rcv.__netif_receive_skb_core.process_backlog.net_rx_action.__softirqentry_text_start
0.00 +15.8 15.79 ± 3% perf-profile.calltrace.cycles-pp.__netif_receive_skb_core.process_backlog.net_rx_action.__softirqentry_text_start.do_softirq_own_stack
0.00 +16.2 16.16 ± 3% perf-profile.calltrace.cycles-pp.process_backlog.net_rx_action.__softirqentry_text_start.do_softirq_own_stack.do_softirq
0.00 +16.5 16.52 ± 3% perf-profile.calltrace.cycles-pp.net_rx_action.__softirqentry_text_start.do_softirq_own_stack.do_softirq.__local_bh_enable_ip
0.00 +16.9 16.93 ± 3% perf-profile.calltrace.cycles-pp.__softirqentry_text_start.do_softirq_own_stack.do_softirq.__local_bh_enable_ip.ip_finish_output2
16.23 ± 2% -16.2 0.00 perf-profile.children.cycles-pp.__do_softirq
41.37 ± 4% -5.9 35.52 ± 2% perf-profile.children.cycles-pp.cpuidle_enter_state
40.53 ± 4% -5.8 34.77 ± 2% perf-profile.children.cycles-pp.intel_idle
46.34 ± 3% -5.6 40.72 ± 2% perf-profile.children.cycles-pp.secondary_startup_64
46.35 ± 3% -5.6 40.74 ± 2% perf-profile.children.cycles-pp.cpu_startup_entry
46.37 ± 3% -5.6 40.80 ± 2% perf-profile.children.cycles-pp.do_idle
36.29 ± 9% -5.2 31.12 ± 13% perf-profile.children.cycles-pp.start_secondary
10.05 ± 21% -0.4 9.61 ± 50% perf-profile.children.cycles-pp.start_kernel
5.44 ± 4% -0.4 5.04 ± 6% perf-profile.children.cycles-pp.tcp_rcv_established
4.92 +0.2 5.17 ± 6% perf-profile.children.cycles-pp.__sched_text_start
10.56 ± 2% +0.3 10.85 ± 5% perf-profile.children.cycles-pp.sock_sendmsg
4.86 ± 4% +0.3 5.15 ± 3% perf-profile.children.cycles-pp.inet_recvmsg
10.43 ± 2% +0.3 10.73 ± 5% perf-profile.children.cycles-pp.tcp_sendmsg
4.81 ± 4% +0.3 5.12 ± 3% perf-profile.children.cycles-pp.tcp_recvmsg
5.19 ± 5% +0.3 5.52 ± 3% perf-profile.children.cycles-pp.SYSC_recvfrom
10.69 ± 2% +0.3 11.02 ± 6% perf-profile.children.cycles-pp.SYSC_sendto
8.96 ± 2% +0.5 9.42 ± 6% perf-profile.children.cycles-pp.tcp_sendmsg_locked
6.32 ± 4% +0.5 6.83 ± 4% perf-profile.children.cycles-pp.__release_sock
6.70 ± 3% +0.5 7.21 ± 4% perf-profile.children.cycles-pp.release_sock
4.28 ± 4% +0.6 4.90 ± 5% perf-profile.children.cycles-pp.tcp_connect
5.06 ± 4% +0.8 5.82 ± 5% perf-profile.children.cycles-pp.tcp_v4_connect
13.65 ± 3% +0.9 14.56 ± 3% perf-profile.children.cycles-pp.ip_local_deliver
13.56 ± 3% +0.9 14.47 ± 3% perf-profile.children.cycles-pp.ip_local_deliver_finish
13.40 ± 3% +0.9 14.34 ± 3% perf-profile.children.cycles-pp.tcp_v4_rcv
14.14 ± 2% +1.0 15.09 ± 3% perf-profile.children.cycles-pp.ip_rcv
14.76 ± 2% +1.1 15.82 ± 3% perf-profile.children.cycles-pp.__netif_receive_skb_core
15.14 ± 2% +1.1 16.23 ± 3% perf-profile.children.cycles-pp.process_backlog
15.41 ± 2% +1.2 16.59 ± 3% perf-profile.children.cycles-pp.net_rx_action
15.89 ± 2% +1.2 17.12 ± 3% perf-profile.children.cycles-pp.do_softirq_own_stack
16.04 ± 2% +1.2 17.29 ± 3% perf-profile.children.cycles-pp.do_softirq
14.34 ± 2% +1.3 15.60 ± 4% perf-profile.children.cycles-pp.tcp_v4_do_rcv
15.22 ± 2% +1.3 16.53 ± 4% perf-profile.children.cycles-pp.__tcp_push_pending_frames
15.18 ± 2% +1.3 16.52 ± 4% perf-profile.children.cycles-pp.tcp_write_xmit
9.46 ± 4% +1.4 10.82 ± 5% perf-profile.children.cycles-pp.__inet_stream_connect
9.51 ± 4% +1.4 10.87 ± 5% perf-profile.children.cycles-pp.inet_stream_connect
16.46 ± 2% +1.4 17.86 ± 3% perf-profile.children.cycles-pp.__local_bh_enable_ip
9.67 ± 4% +1.4 11.10 ± 4% perf-profile.children.cycles-pp.SYSC_connect
9.23 ± 2% +1.5 10.69 ± 5% perf-profile.children.cycles-pp.inet_release
9.03 ± 2% +1.5 10.50 ± 6% perf-profile.children.cycles-pp.tcp_close
9.28 ± 2% +1.5 10.76 ± 6% perf-profile.children.cycles-pp.sock_close
9.25 ± 2% +1.5 10.75 ± 6% perf-profile.children.cycles-pp.sock_release
17.93 ± 2% +1.5 19.45 ± 3% perf-profile.children.cycles-pp.ip_finish_output2
11.07 ± 2% +1.6 12.69 ± 5% perf-profile.children.cycles-pp.exit_to_usermode_loop
10.97 ± 2% +1.6 12.59 ± 5% perf-profile.children.cycles-pp.task_work_run
10.79 ± 2% +1.6 12.42 ± 5% perf-profile.children.cycles-pp.__fput
11.14 ± 2% +1.6 12.77 ± 5% perf-profile.children.cycles-pp.syscall_return_slowpath
8.83 ± 4% +1.7 10.51 ± 5% perf-profile.children.cycles-pp.tcp_rcv_state_process
18.77 ± 2% +1.8 20.59 ± 3% perf-profile.children.cycles-pp.ip_output
19.39 ± 2% +1.9 21.34 ± 3% perf-profile.children.cycles-pp.ip_queue_xmit
20.76 ± 2% +2.3 23.09 ± 3% perf-profile.children.cycles-pp.tcp_transmit_skb
42.48 ± 2% +4.2 46.67 ± 2% perf-profile.children.cycles-pp.entry_SYSCALL_64_fastpath
0.00 +17.5 17.45 ± 3% perf-profile.children.cycles-pp.__softirqentry_text_start
40.53 ± 4% -5.8 34.77 ± 2% perf-profile.self.cycles-pp.intel_idle
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,
Xiaolong
4 years, 4 months
- 1
- 0
[rtnl] b1585bdfb2: kernel_BUG_at_net/core/dev.c
by kernel test robot
FYI, we noticed the following commit (built with gcc-7):
commit: b1585bdfb20476e31f11c3fd44faac770325b8c1 ("rtnl: device allocation/free via rtnl_link_ops")
url: https://github.com/0day-ci/linux/commits/Denis-Drozdov/rtnl-device-alloca...
in testcase: trinity
with following parameters:
runtime: 300s
test-description: Trinity is a linux system call fuzz tester.
test-url: http://codemonkey.org.uk/projects/trinity/
on test machine: qemu-system-x86_64 -enable-kvm -m 512M
caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
+------------------------------------------+------------+------------+
| | ead68f2161 | b1585bdfb2 |
+------------------------------------------+------------+------------+
| boot_successes | 260 | 2 |
| boot_failures | 4 | 13 |
| BUG:kernel_hang_in_boot_stage | 3 | |
| BUG:kernel_hang_in_test_stage | 1 | |
| kernel_BUG_at_net/core/dev.c | 0 | 13 |
| invalid_opcode:#[##] | 0 | 13 |
| RIP:free_netdev | 0 | 13 |
| general_protection_fault:#[##] | 0 | 9 |
| RIP:put_page | 0 | 9 |
| Kernel_panic-not_syncing:Fatal_exception | 0 | 13 |
+------------------------------------------+------------+------------+
[ 35.401334] kernel BUG at net/core/dev.c:8231!
[ 35.402342] invalid opcode: 0000 [#1] SMP
[ 35.403055] Modules linked in:
[ 35.403636] CPU: 0 PID: 134 Comm: kworker/u2:2 Not tainted 4.15.0-rc4-00203-gb1585bd #74
[ 35.404979] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1 04/01/2014
[ 35.407023] Workqueue: netns cleanup_net
[ 35.407760] RIP: 0010:free_netdev+0xde/0xf7
[ 35.408491] RSP: 0018:ffffa49a00af7d48 EFLAGS: 00010202
[ 35.409356] RAX: 0000000000000004 RBX: ffff8b8ca7856000 RCX: 0000000000000001
[ 35.410473] RDX: 0000000000000001 RSI: ffffffff8e0e7e00 RDI: 0000000000000000
[ 35.411597] RBP: ffff8b8ca7855fd8 R08: 0000000000000080 R09: 0000000000000004
[ 35.412722] R10: 00000000ffffffd2 R11: 0000000000000000 R12: ffff8b8ca7856060
[ 35.413847] R13: 00000000fffefd9a R14: 00000000fffefd9a R15: dead000000000100
[ 35.414976] FS: 0000000000000000(0000) GS:ffff8b8cbf400000(0000) knlGS:0000000000000000
[ 35.416305] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 35.417242] CR2: 000000000325de08 CR3: 0000000034559000 CR4: 00000000000006f0
[ 35.418347] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 35.419455] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 35.420576] Call Trace:
[ 35.421081] netdev_run_todo+0x247/0x290
[ 35.421770] ip6_tnl_exit_batch_net+0x150/0x15f
[ 35.422549] cleanup_net+0x19b/0x234
[ 35.423195] process_one_work+0x2b3/0x4c3
[ 35.423894] ? worker_thread+0x1e0/0x25b
[ 35.424574] ? rescuer_thread+0x27a/0x27a
[ 35.425274] worker_thread+0x1a3/0x25b
[ 35.425940] ? rescuer_thread+0x27a/0x27a
[ 35.426631] kthread+0xf5/0xfa
[ 35.427199] ? kthread_create_worker_on_cpu+0x43/0x43
[ 35.428056] ? call_usermodehelper_exec_async+0x118/0x11f
[ 35.428954] ? do_group_exit+0xb2/0xb2
[ 35.429618] ret_from_fork+0x24/0x30
[ 35.430266] Code: 00 00 00 74 0c 48 c7 c7 d0 a6 4b 8e e8 4e 5f ff ff 8a 83 80 05 00 00 84 c0 75 0c 48 89 df 5b 5d 41 5c e9 12 ff ff ff 3c 03 74 02 <0f> 0b c6 83 80 05 00 00 04 48 8d bb b0 05 00 00 5b 5d 41 5c e9
[ 35.433215] RIP: free_netdev+0xde/0xf7 RSP: ffffa49a00af7d48
[ 35.434198] ---[ end trace def93e3513cf4702 ]---
To reproduce:
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
bin/lkp qemu -k <bzImage> job-script # job-script is attached in this email
Thanks,
lkp
4 years, 4 months
- 1
- 0