Greeting, FYI, we noticed a -7.2% regression of blogbench.read_score due to commit: commit: 523983401644ebeb331c923c28c9591c07430a7d ("Btrfs: kill btrfs_clear_path_blocking") https://git.kernel.org/cgit/linux/kernel/git/kdave/linux.git for-4.20-part1 in testcase: blogbench on test machine: 88 threads Intel(R) Xeon(R) CPU E5-2699 v4 @ 2.20GHz with 64G memory with following parameters: disk: 1SSD fs: btrfs ucode: 0xb00002e cpufreq_governor: performance test-description: Blogbench is a portable filesystem benchmark that tries to reproduce the load of a real-world busy file server. test-url: https://www.pureftpd.org/project/blogbench In addition to that, the commit also has significant impact on the following tests: +------------------+-----------------------------------------------------------------------+ | testcase: change | fio-basic: | | test machine | 56 threads Intel(R) Xeon(R) CPU E5-2695 v3 @ 2.30GHz with 256G memory | | test parameters | bs=4k | | | cpufreq_governor=performance | | | disk=2pmem | | | fs=btrfs | | | ioengine=mmap | | | nr_task=50% | | | runtime=200s | | | rw=randwrite | | | test_size=100G | | | time_based=tb | | | ucode=0x3d | +------------------+-----------------------------------------------------------------------+ 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 ========================================================================================= compiler/cpufreq_governor/disk/fs/kconfig/rootfs/tbox_group/testcase/ucode: gcc-7/performance/1SSD/btrfs/x86_64-rhel-7.2/debian-x86_64-2018-04-03.cgz/lkp-bdw-ep3b/blogbench/0xb00002e commit: 9b142115ed ("btrfs: dev-replace: remove pointless assert in write unlock") 5239834016 ("Btrfs: kill btrfs_clear_path_blocking") 9b142115ed359348 523983401644ebeb331c923c28 ---------------- -------------------------- fail:runs %reproduction fail:runs | | | :4 25% 1:4 dmesg.WARNING:at#for_ip_interrupt_entry/0x %stddev %change %stddev \ | \ 2339806 ± 2% -7.2% 2171755 ± 2% blogbench.read_score 35999832 +102.7% 72956144 blogbench.time.file_system_outputs 9368 ± 6% +16.1% 10872 ± 12% blogbench.time.minor_page_faults 8033 +4.2% 8371 blogbench.time.percent_of_cpu_this_job_got 23833 +4.2% 24824 blogbench.time.system_time 14208581 -36.5% 9023264 ± 2% blogbench.time.voluntary_context_switches 3942 +90.7% 7519 blogbench.write_score 623343 ± 2% +40.2% 874008 ± 3% interrupts.CAL:Function_call_interrupts 5.32 ± 12% -3.5 1.80 ± 10% mpstat.cpu.idle% 0.00 ± 36% +0.0 0.01 ± 31% mpstat.cpu.iowait% 197893 ± 4% +115.2% 425773 softirqs.BLOCK 708383 ± 7% -29.7% 498259 ± 7% softirqs.SCHED 6513961 ± 5% +25.1% 8150114 numa-numastat.node0.local_node 6522556 ± 5% +25.0% 8154430 numa-numastat.node0.numa_hit 6186981 ± 3% +38.2% 8549312 ± 3% numa-numastat.node1.local_node 6195606 ± 3% +38.2% 8562194 ± 3% numa-numastat.node1.numa_hit 8.36 ± 15% +208.1% 25.76 ± 2% iostat.sda.avgqu-sz 117.35 ± 19% +72.9% 202.85 ± 5% iostat.sda.avgrq-sz 1337 ± 4% +117.9% 2914 iostat.sda.w/s 55899 +119.2% 122524 iostat.sda.wkB/s 1336 ± 6% +115.4% 2879 iostat.sda.wrqm/s 55754 +119.8% 122532 vmstat.io.bo 9410599 +84.4% 17349395 vmstat.memory.cache 55400592 -14.7% 47252921 vmstat.memory.free 135164 -29.3% 95527 vmstat.system.cs 226228 -2.1% 221394 vmstat.system.in 1.138e+08 ± 5% -72.8% 30993990 ± 4% cpuidle.C1.time 5684766 ± 5% -78.9% 1198851 ± 5% cpuidle.C1.usage 80465587 ± 58% -83.5% 13288839 ± 24% cpuidle.C1E.time 939509 ± 49% -80.4% 184165 ± 17% cpuidle.C1E.usage 8.707e+08 ± 31% -77.9% 1.925e+08 ± 45% cpuidle.C6.time 1114164 ± 21% -77.5% 251130 ± 47% cpuidle.C6.usage 799330 ± 18% -72.6% 218752 ± 4% cpuidle.POLL.time 43028 ± 6% -74.7% 10884 ± 5% cpuidle.POLL.usage 6383503 +91.0% 12190888 meminfo.Active 6076016 +95.6% 11882998 meminfo.Active(file) 8683414 +84.4% 16014311 meminfo.Cached 1644022 +92.6% 3166518 meminfo.Inactive 1625430 +93.7% 3147989 meminfo.Inactive(file) 55450824 -14.7% 47325613 meminfo.MemFree 678179 +86.4% 1264342 meminfo.SReclaimable 401687 +50.2% 603151 meminfo.SUnreclaim 1079867 +72.9% 1867493 meminfo.Slab 6814 ± 34% +458.0% 38021 ± 6% meminfo.Writeback 2656 +3.5% 2749 turbostat.Avg_MHz 5681184 ± 5% -79.0% 1194935 ± 5% turbostat.C1 0.42 ± 5% -0.3 0.12 ± 3% turbostat.C1% 935242 ± 50% -80.8% 179244 ± 17% turbostat.C1E 0.30 ± 58% -0.3 0.05 ± 27% turbostat.C1E% 1102813 ± 21% -77.7% 245393 ± 49% turbostat.C6 3.23 ± 31% -2.5 0.71 ± 45% turbostat.C6% 3.16 ± 9% -69.2% 0.97 ± 12% turbostat.CPU%c1 1.39 ± 47% -77.3% 0.32 ± 58% turbostat.CPU%c6 0.67 ± 49% -62.2% 0.25 ± 27% turbostat.Pkg%pc2 234.32 +1.6% 238.14 turbostat.PkgWatt 3.862e+12 +4.3% 4.027e+12 perf-stat.branch-instructions 1.684e+10 +2.3% 1.723e+10 perf-stat.branch-misses 30.36 -0.3 30.02 perf-stat.cache-miss-rate% 8.746e+10 -3.9% 8.406e+10 perf-stat.cache-misses 41358877 -29.9% 28984102 perf-stat.context-switches 4.32 -2.0% 4.23 perf-stat.cpi 7.074e+13 +2.7% 7.263e+13 perf-stat.cpu-cycles 0.02 ± 7% +0.0 0.02 perf-stat.dTLB-load-miss-rate% 9.149e+08 ± 7% +22.4% 1.12e+09 perf-stat.dTLB-load-misses 4.387e+12 +4.4% 4.58e+12 perf-stat.dTLB-loads 1.638e+13 +4.7% 1.716e+13 perf-stat.instructions 0.23 +2.0% 0.24 perf-stat.ipc 761956 -1.0% 754649 perf-stat.minor-faults 1.714e+10 -4.1% 1.643e+10 perf-stat.node-load-misses 1.202e+10 ± 2% -5.9% 1.131e+10 ± 2% perf-stat.node-store-misses 761958 -1.0% 754651 perf-stat.page-faults 1517674 +95.8% 2971175 proc-vmstat.nr_active_file 4752729 +109.2% 9944485 proc-vmstat.nr_dirtied 321416 ± 7% +9.1% 350782 ± 3% proc-vmstat.nr_dirty 1567569 -1.3% 1547670 proc-vmstat.nr_dirty_background_threshold 3138973 -1.3% 3099125 proc-vmstat.nr_dirty_threshold 2169155 +84.6% 4004236 proc-vmstat.nr_file_pages 13864507 -14.7% 11830502 proc-vmstat.nr_free_pages 405998 +93.9% 787215 proc-vmstat.nr_inactive_file 16065 +2.3% 16429 proc-vmstat.nr_kernel_stack 169409 +86.6% 316143 proc-vmstat.nr_slab_reclaimable 100393 +50.3% 150923 proc-vmstat.nr_slab_unreclaimable 1790 ± 35% +439.6% 9659 ± 6% proc-vmstat.nr_writeback 4229206 +120.2% 9311657 proc-vmstat.nr_written 1517674 +95.8% 2971175 proc-vmstat.nr_zone_active_file 405998 +93.9% 787215 proc-vmstat.nr_zone_inactive_file 323756 ± 7% +11.6% 361231 ± 2% proc-vmstat.nr_zone_write_pending 12740396 +31.3% 16724676 proc-vmstat.numa_hit 12723171 +31.3% 16707470 proc-vmstat.numa_local 3466918 +97.1% 6833346 proc-vmstat.pgactivate 21010209 +21.5% 25519008 proc-vmstat.pgalloc_normal 16906221 ± 2% +4.3% 17638056 proc-vmstat.pgfree 16888174 +120.2% 37189214 proc-vmstat.pgpgout 3439308 ± 12% +68.0% 5779423 ± 2% numa-meminfo.node0.Active 3292978 ± 13% +70.5% 5615415 ± 2% numa-meminfo.node0.Active(file) 4674100 ± 12% +62.9% 7614966 ± 2% numa-meminfo.node0.FilePages 900475 ± 16% +67.0% 1504140 ± 2% numa-meminfo.node0.Inactive 890125 ± 16% +67.0% 1486500 ± 2% numa-meminfo.node0.Inactive(file) 27300625 ± 2% -12.0% 24025498 numa-meminfo.node0.MemFree 5558153 ± 11% +58.8% 8826192 ± 2% numa-meminfo.node0.MemUsed 370938 ± 15% +60.9% 596960 ± 3% numa-meminfo.node0.SReclaimable 198127 ± 6% +45.2% 287679 ± 3% numa-meminfo.node0.SUnreclaim 569065 ± 11% +55.5% 884640 ± 3% numa-meminfo.node0.Slab 3534 ± 30% +439.6% 19072 ± 6% numa-meminfo.node0.Writeback 2936433 ± 16% +117.9% 6399389 ± 3% numa-meminfo.node1.Active 2775387 ± 16% +125.4% 6255600 ± 4% numa-meminfo.node1.Active(file) 590133 ± 5% +24.9% 737354 ± 5% numa-meminfo.node1.Dirty 3999720 ± 15% +109.6% 8383937 ± 4% numa-meminfo.node1.FilePages 741722 ± 18% +123.7% 1659082 ± 5% numa-meminfo.node1.Inactive 733558 ± 17% +126.0% 1658132 ± 5% numa-meminfo.node1.Inactive(file) 28161048 ± 2% -17.2% 23316049 numa-meminfo.node1.MemFree 4832002 ± 14% +100.4% 9684087 ± 3% numa-meminfo.node1.MemUsed 306477 ± 19% +117.4% 666212 ± 4% numa-meminfo.node1.SReclaimable 202862 ± 5% +55.7% 315954 ± 3% numa-meminfo.node1.SUnreclaim 509340 ± 13% +92.8% 982167 ± 4% numa-meminfo.node1.Slab 3309 ± 51% +500.9% 19885 ± 4% numa-meminfo.node1.Writeback 1238 ± 18% -72.5% 339.89 ± 4% sched_debug.cfs_rq:/.exec_clock.stddev 12674187 +10.6% 14013368 sched_debug.cfs_rq:/.min_vruntime.avg 13159010 +10.3% 14515696 sched_debug.cfs_rq:/.min_vruntime.max 284068 ± 3% +20.1% 341267 ± 8% sched_debug.cfs_rq:/.min_vruntime.stddev -1661029 +33.9% -2223696 sched_debug.cfs_rq:/.spread0.min 284027 ± 3% +20.1% 341256 ± 8% sched_debug.cfs_rq:/.spread0.stddev 261246 -34.5% 171104 sched_debug.cpu.nr_switches.avg 131794 ± 5% -25.3% 98505 ± 3% sched_debug.cpu.nr_switches.min 88301 ± 6% -20.2% 70424 ± 5% sched_debug.cpu.nr_switches.stddev -70.00 -9.8% -63.12 sched_debug.cpu.nr_uninterruptible.min 261990 -34.5% 171673 sched_debug.cpu.sched_count.avg 131829 ± 5% -25.3% 98478 ± 3% sched_debug.cpu.sched_count.min 89185 ± 7% -18.8% 72377 ± 7% sched_debug.cpu.sched_count.stddev 55278 ± 4% -78.3% 11971 ± 5% sched_debug.cpu.sched_goidle.avg 124430 ± 4% -73.5% 33031 ± 9% sched_debug.cpu.sched_goidle.max 9033 ± 24% -67.9% 2899 ± 27% sched_debug.cpu.sched_goidle.min 31192 ± 15% -80.3% 6157 ± 9% sched_debug.cpu.sched_goidle.stddev 163543 -33.2% 109236 sched_debug.cpu.ttwu_count.avg 559684 ± 9% +30.0% 727656 ± 5% sched_debug.cpu.ttwu_count.max 82509 ± 4% -30.6% 57266 ± 2% sched_debug.cpu.ttwu_count.min 60134 ± 4% +16.1% 69840 ± 5% sched_debug.cpu.ttwu_count.stddev 7548 +70.3% 12853 sched_debug.cpu.ttwu_local.avg 49292 ± 13% +133.5% 115118 ± 5% sched_debug.cpu.ttwu_local.max 4640 ± 5% +50.0% 6962 ± 2% sched_debug.cpu.ttwu_local.min 4665 ± 13% +140.2% 11204 ± 5% sched_debug.cpu.ttwu_local.stddev 823017 ± 13% +70.6% 1403862 ± 3% numa-vmstat.node0.nr_active_file 1291355 ± 14% +80.5% 2330828 ± 2% numa-vmstat.node0.nr_dirtied 1168265 ± 12% +63.0% 1903769 ± 2% numa-vmstat.node0.nr_file_pages 6825413 ± 2% -12.0% 6006516 numa-vmstat.node0.nr_free_pages 222478 ± 15% +67.0% 371641 ± 2% numa-vmstat.node0.nr_inactive_file 92708 ± 15% +61.0% 149233 ± 3% numa-vmstat.node0.nr_slab_reclaimable 49543 ± 5% +44.9% 71769 ± 3% numa-vmstat.node0.nr_slab_unreclaimable 906.00 ± 23% +412.1% 4639 ± 12% numa-vmstat.node0.nr_writeback 1053489 ± 13% +97.8% 2084007 ± 2% numa-vmstat.node0.nr_written 823017 ± 13% +70.6% 1403862 ± 3% numa-vmstat.node0.nr_zone_active_file 222478 ± 15% +67.0% 371640 ± 2% numa-vmstat.node0.nr_zone_inactive_file 3701754 ± 5% +23.3% 4562929 numa-vmstat.node0.numa_hit 3692968 ± 5% +23.4% 4558577 numa-vmstat.node0.numa_local 693808 ± 16% +125.4% 1563789 ± 4% numa-vmstat.node1.nr_active_file 1079325 ± 18% +142.5% 2617710 ± 4% numa-vmstat.node1.nr_dirtied 147501 ± 6% +25.1% 184488 ± 5% numa-vmstat.node1.nr_dirty 999853 ± 15% +109.6% 2095923 ± 4% numa-vmstat.node1.nr_file_pages 7040287 ± 2% -17.2% 5829252 numa-vmstat.node1.nr_free_pages 183348 ± 18% +126.1% 414583 ± 5% numa-vmstat.node1.nr_inactive_file 3287 ± 16% -19.7% 2639 ± 2% numa-vmstat.node1.nr_mapped 76615 ± 19% +117.4% 166540 ± 4% numa-vmstat.node1.nr_slab_reclaimable 50754 ± 5% +55.3% 78827 ± 3% numa-vmstat.node1.nr_slab_unreclaimable 837.00 ± 38% +475.9% 4820 ± 5% numa-vmstat.node1.nr_writeback 880188 ± 19% +164.8% 2330392 ± 4% numa-vmstat.node1.nr_written 693808 ± 16% +125.4% 1563789 ± 4% numa-vmstat.node1.nr_zone_active_file 183348 ± 18% +126.1% 414583 ± 5% numa-vmstat.node1.nr_zone_inactive_file 148581 ± 6% +27.7% 189690 ± 5% numa-vmstat.node1.nr_zone_write_pending 3503508 ± 5% +35.4% 4744281 ± 2% numa-vmstat.node1.numa_hit 3359384 ± 5% +36.8% 4595659 ± 2% numa-vmstat.node1.numa_local 412161 +75.9% 724947 slabinfo.Acpi-Namespace.active_objs 4040 +75.9% 7107 slabinfo.Acpi-Namespace.active_slabs 412179 +75.9% 725010 slabinfo.Acpi-Namespace.num_objs 4040 +75.9% 7107 slabinfo.Acpi-Namespace.num_slabs 14738 ± 7% +41.9% 20916 ± 5% slabinfo.Acpi-ParseExt.active_objs 377.50 ± 7% +42.5% 537.75 ± 5% slabinfo.Acpi-ParseExt.active_slabs 14747 ± 7% +42.3% 20985 ± 5% slabinfo.Acpi-ParseExt.num_objs 377.50 ± 7% +42.5% 537.75 ± 5% slabinfo.Acpi-ParseExt.num_slabs 338988 +97.8% 670518 slabinfo.Acpi-State.active_objs 6656 +98.0% 13180 slabinfo.Acpi-State.active_slabs 339477 +98.0% 672237 slabinfo.Acpi-State.num_objs 6656 +98.0% 13180 slabinfo.Acpi-State.num_slabs 17218 ± 6% +58.5% 27293 ± 4% slabinfo.avc_xperms_data.active_objs 17218 ± 6% +58.6% 27307 ± 4% slabinfo.avc_xperms_data.num_objs 1083 ± 2% +14.7% 1242 ± 2% slabinfo.biovec-128.active_objs 1083 ± 2% +14.7% 1242 ± 2% slabinfo.biovec-128.num_objs 852.25 ± 10% +158.9% 2206 ± 4% slabinfo.biovec-max.active_objs 222.25 ± 9% +153.1% 562.50 ± 4% slabinfo.biovec-max.active_slabs 890.75 ± 9% +152.7% 2251 ± 4% slabinfo.biovec-max.num_objs 222.25 ± 9% +153.1% 562.50 ± 4% slabinfo.biovec-max.num_slabs 328556 +94.7% 639780 slabinfo.btrfs_delayed_node.active_objs 6318 +94.7% 12303 slabinfo.btrfs_delayed_node.active_slabs 328560 +94.7% 639784 slabinfo.btrfs_delayed_node.num_objs 6318 +94.7% 12303 slabinfo.btrfs_delayed_node.num_slabs 24166 +69.7% 41014 slabinfo.btrfs_extent_buffer.active_objs 418.25 +71.2% 716.00 slabinfo.btrfs_extent_buffer.active_slabs 24292 +71.1% 41558 slabinfo.btrfs_extent_buffer.num_objs 418.25 +71.2% 716.00 slabinfo.btrfs_extent_buffer.num_slabs 274744 ± 2% +110.3% 577705 slabinfo.btrfs_extent_map.active_objs 4906 ± 2% +110.3% 10316 slabinfo.btrfs_extent_map.active_slabs 274769 ± 2% +110.3% 577758 slabinfo.btrfs_extent_map.num_objs 4906 ± 2% +110.3% 10316 slabinfo.btrfs_extent_map.num_slabs 327623 +95.1% 639035 slabinfo.btrfs_inode.active_objs 11702 +95.1% 22826 slabinfo.btrfs_inode.active_slabs 327679 +95.1% 639160 slabinfo.btrfs_inode.num_objs 11702 +95.1% 22826 slabinfo.btrfs_inode.num_slabs 431945 +80.1% 778072 slabinfo.dentry.active_objs 10288 +80.2% 18535 slabinfo.dentry.active_slabs 432133 +80.2% 778505 slabinfo.dentry.num_objs 10288 +80.2% 18535 slabinfo.dentry.num_slabs 6708 +25.7% 8431 ± 4% slabinfo.kmalloc-1024.active_objs 6770 +25.8% 8518 ± 4% slabinfo.kmalloc-1024.num_objs 19023 ± 6% +16.2% 22103 ± 3% slabinfo.kmalloc-128.active_objs 19181 ± 6% +16.5% 22340 ± 3% slabinfo.kmalloc-128.num_objs 22578 ± 4% +69.3% 38216 ± 5% slabinfo.kmalloc-192.active_objs 538.25 ± 4% +69.6% 913.00 ± 5% slabinfo.kmalloc-192.active_slabs 22622 ± 4% +69.6% 38367 ± 5% slabinfo.kmalloc-192.num_objs 538.25 ± 4% +69.6% 913.00 ± 5% slabinfo.kmalloc-192.num_slabs 1302 +9.4% 1424 slabinfo.kmalloc-4096.active_objs 29789 ± 5% +15.5% 34398 ± 3% slabinfo.kmalloc-96.active_objs 30023 ± 5% +16.2% 34881 ± 3% slabinfo.kmalloc-96.num_objs 222.50 ± 32% +36.4% 303.50 ± 19% slabinfo.nfs_read_data.active_objs 222.50 ± 32% +36.4% 303.50 ± 19% slabinfo.nfs_read_data.num_objs 315452 +91.1% 602801 slabinfo.radix_tree_node.active_objs 5633 +91.1% 10765 slabinfo.radix_tree_node.active_slabs 315485 +91.1% 602881 slabinfo.radix_tree_node.num_objs 5633 +91.1% 10765 slabinfo.radix_tree_node.num_slabs 4980 +31.5% 6549 ± 4% slabinfo.scsi_sense_cache.active_objs 4993 +31.5% 6564 ± 4% slabinfo.scsi_sense_cache.num_objs 567.00 ± 11% +25.9% 714.00 ± 2% slabinfo.sgpool-128.active_objs 573.75 ± 11% +27.0% 728.75 ± 2% slabinfo.sgpool-128.num_objs 9.71 ± 10% -1.0 8.66 ± 2% perf-profile.calltrace.cycles-pp.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe 9.39 ± 10% -1.0 8.37 ± 2% perf-profile.calltrace.cycles-pp.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe 9.16 ± 11% -1.0 8.15 ± 2% perf-profile.calltrace.cycles-pp.__vfs_read.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe 9.00 ± 11% -1.0 8.00 ± 2% perf-profile.calltrace.cycles-pp.generic_file_read_iter.__vfs_read.vfs_read.ksys_read.do_syscall_64 6.80 ± 11% -0.8 6.02 ± 2% perf-profile.calltrace.cycles-pp.copy_page_to_iter.generic_file_read_iter.__vfs_read.vfs_read.ksys_read 6.56 ± 12% -0.8 5.79 ± 2% perf-profile.calltrace.cycles-pp.copyout.copy_page_to_iter.generic_file_read_iter.__vfs_read.vfs_read 6.50 ± 12% -0.8 5.74 ± 2% perf-profile.calltrace.cycles-pp.copy_user_enhanced_fast_string.copyout.copy_page_to_iter.generic_file_read_iter.__vfs_read 1.48 ± 5% -0.5 0.93 ± 3% perf-profile.calltrace.cycles-pp.__x64_sys_rename.do_syscall_64.entry_SYSCALL_64_after_hwframe 1.47 ± 5% -0.5 0.93 ± 3% perf-profile.calltrace.cycles-pp.do_renameat2.__x64_sys_rename.do_syscall_64.entry_SYSCALL_64_after_hwframe 1.08 ± 5% -0.4 0.65 ± 3% perf-profile.calltrace.cycles-pp.btrfs_rename.vfs_rename.do_renameat2.__x64_sys_rename.do_syscall_64 1.09 ± 5% -0.4 0.67 ± 2% perf-profile.calltrace.cycles-pp.vfs_rename.do_renameat2.__x64_sys_rename.do_syscall_64.entry_SYSCALL_64_after_hwframe 2.29 ± 8% -0.3 2.03 ± 3% perf-profile.calltrace.cycles-pp.__softirqentry_text_start.run_ksoftirqd.smpboot_thread_fn.kthread.ret_from_fork 2.29 ± 8% -0.3 2.03 ± 3% perf-profile.calltrace.cycles-pp.run_ksoftirqd.smpboot_thread_fn.kthread.ret_from_fork 2.28 ± 8% -0.3 2.02 ± 3% perf-profile.calltrace.cycles-pp.rcu_process_callbacks.__softirqentry_text_start.run_ksoftirqd.smpboot_thread_fn.kthread 3.34 ± 6% -0.3 3.08 ± 3% perf-profile.calltrace.cycles-pp.do_filp_open.do_sys_open.do_syscall_64.entry_SYSCALL_64_after_hwframe 2.30 ± 8% -0.3 2.04 ± 3% perf-profile.calltrace.cycles-pp.smpboot_thread_fn.kthread.ret_from_fork 3.31 ± 6% -0.3 3.06 ± 3% perf-profile.calltrace.cycles-pp.path_openat.do_filp_open.do_sys_open.do_syscall_64.entry_SYSCALL_64_after_hwframe 2.06 ± 7% -0.2 1.82 ± 3% perf-profile.calltrace.cycles-pp.file_free_rcu.rcu_process_callbacks.__softirqentry_text_start.run_ksoftirqd.smpboot_thread_fn 0.59 ± 3% -0.0 0.54 ± 3% perf-profile.calltrace.cycles-pp.__alloc_file.alloc_empty_file.path_openat.do_filp_open.do_sys_open 0.61 ± 4% -0.0 0.57 ± 3% perf-profile.calltrace.cycles-pp.alloc_empty_file.path_openat.do_filp_open.do_sys_open.do_syscall_64 94.63 +0.7 95.34 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe 94.58 +0.7 95.30 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe 48.35 +0.8 49.15 perf-profile.calltrace.cycles-pp.do_sys_open.do_syscall_64.entry_SYSCALL_64_after_hwframe 39.07 +1.0 40.06 perf-profile.calltrace.cycles-pp._raw_spin_lock.__alloc_fd.do_sys_open.do_syscall_64.entry_SYSCALL_64_after_hwframe 38.86 +1.0 39.85 perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.__alloc_fd.do_sys_open.do_syscall_64 39.37 +1.0 40.36 perf-profile.calltrace.cycles-pp.__alloc_fd.do_sys_open.do_syscall_64.entry_SYSCALL_64_after_hwframe 33.62 +1.5 35.09 perf-profile.calltrace.cycles-pp._raw_spin_lock.__close_fd.__x64_sys_close.do_syscall_64.entry_SYSCALL_64_after_hwframe 33.98 +1.5 35.45 perf-profile.calltrace.cycles-pp.__x64_sys_close.do_syscall_64.entry_SYSCALL_64_after_hwframe 33.44 +1.5 34.91 perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.__close_fd.__x64_sys_close.do_syscall_64 33.78 +1.5 35.25 perf-profile.calltrace.cycles-pp.__close_fd.__x64_sys_close.do_syscall_64.entry_SYSCALL_64_after_hwframe 3.10 ± 27% -1.5 1.60 ± 2% perf-profile.children.cycles-pp.btrfs_search_slot 1.53 ± 34% -1.1 0.41 ± 2% perf-profile.children.cycles-pp.btrfs_tree_lock 1.52 ± 35% -1.1 0.40 ± 2% perf-profile.children.cycles-pp.btrfs_lock_root_node 9.72 ± 10% -1.0 8.68 ± 2% perf-profile.children.cycles-pp.ksys_read 9.41 ± 10% -1.0 8.38 ± 2% perf-profile.children.cycles-pp.vfs_read 9.16 ± 11% -1.0 8.16 ± 2% perf-profile.children.cycles-pp.__vfs_read 9.02 ± 11% -1.0 8.02 ± 2% perf-profile.children.cycles-pp.generic_file_read_iter 6.81 ± 11% -0.8 6.04 ± 2% perf-profile.children.cycles-pp.copy_page_to_iter 6.56 ± 12% -0.8 5.79 ± 2% perf-profile.children.cycles-pp.copyout 6.55 ± 12% -0.8 5.79 ± 2% perf-profile.children.cycles-pp.copy_user_enhanced_fast_string 1.26 ± 35% -0.7 0.59 ± 3% perf-profile.children.cycles-pp._raw_spin_lock_irqsave 1.48 ± 5% -0.5 0.93 ± 3% perf-profile.children.cycles-pp.__x64_sys_rename 1.48 ± 5% -0.5 0.93 ± 3% perf-profile.children.cycles-pp.do_renameat2 0.59 ± 40% -0.5 0.11 ± 4% perf-profile.children.cycles-pp.finish_wait 1.08 ± 5% -0.4 0.65 ± 3% perf-profile.children.cycles-pp.btrfs_rename 0.82 ± 7% -0.4 0.39 ± 3% perf-profile.children.cycles-pp.__btrfs_unlink_inode 1.09 ± 5% -0.4 0.67 ± 2% perf-profile.children.cycles-pp.vfs_rename 0.54 ± 30% -0.4 0.17 ± 4% perf-profile.children.cycles-pp.queued_write_lock_slowpath 0.56 ± 7% -0.3 0.26 perf-profile.children.cycles-pp.btrfs_lookup_dir_item 2.52 ± 8% -0.3 2.24 ± 3% perf-profile.children.cycles-pp.rcu_process_callbacks 2.29 ± 8% -0.3 2.03 ± 3% perf-profile.children.cycles-pp.run_ksoftirqd 2.30 ± 8% -0.3 2.04 ± 3% perf-profile.children.cycles-pp.smpboot_thread_fn 3.34 ± 6% -0.3 3.09 ± 3% perf-profile.children.cycles-pp.do_filp_open 2.27 ± 7% -0.3 2.01 ± 3% perf-profile.children.cycles-pp.file_free_rcu 3.31 ± 6% -0.3 3.06 ± 3% perf-profile.children.cycles-pp.path_openat 0.33 ± 27% -0.2 0.09 ± 7% perf-profile.children.cycles-pp.queued_read_lock_slowpath 0.50 ± 19% -0.2 0.29 ± 6% perf-profile.children.cycles-pp.btrfs_async_run_delayed_root 0.33 ± 9% -0.2 0.14 ± 5% perf-profile.children.cycles-pp.btrfs_del_inode_ref 0.50 ± 15% -0.2 0.32 ± 6% perf-profile.children.cycles-pp.__wake_up_common_lock 0.33 ± 13% -0.1 0.20 ± 5% perf-profile.children.cycles-pp.btrfs_evict_inode 0.33 ± 13% -0.1 0.20 ± 5% perf-profile.children.cycles-pp.evict 0.32 ± 25% -0.1 0.19 ± 5% perf-profile.children.cycles-pp.btrfs_lookup_inode 0.33 ± 11% -0.1 0.21 ± 5% perf-profile.children.cycles-pp.dentry_kill 0.33 ± 11% -0.1 0.21 ± 5% perf-profile.children.cycles-pp.__dentry_kill 0.56 ± 7% -0.1 0.45 ± 2% perf-profile.children.cycles-pp.dput 0.38 ± 14% -0.1 0.27 ± 6% perf-profile.children.cycles-pp.__wake_up_common 0.19 ± 15% -0.1 0.08 ± 5% perf-profile.children.cycles-pp.btrfs_delete_delayed_items 0.36 ± 13% -0.1 0.26 ± 4% perf-profile.children.cycles-pp.autoremove_wake_function 0.56 ± 9% -0.1 0.49 ± 4% perf-profile.children.cycles-pp.may_open 0.21 ± 17% -0.1 0.15 ± 12% perf-profile.children.cycles-pp.reschedule_interrupt 0.19 ± 17% -0.1 0.13 ± 14% perf-profile.children.cycles-pp.scheduler_ipi 0.40 ± 2% -0.1 0.34 ± 2% perf-profile.children.cycles-pp.btrfs_insert_empty_items 0.11 ± 9% -0.1 0.05 perf-profile.children.cycles-pp.btrfs_unlink_inode 0.49 ± 4% -0.0 0.44 ± 4% perf-profile.children.cycles-pp.generic_permission 0.62 ± 4% -0.0 0.57 ± 3% perf-profile.children.cycles-pp.alloc_empty_file 0.58 ± 4% -0.0 0.54 ± 4% perf-profile.children.cycles-pp.inode_permission 0.59 ± 4% -0.0 0.54 ± 3% perf-profile.children.cycles-pp.__alloc_file 0.10 ± 15% -0.0 0.06 ± 7% perf-profile.children.cycles-pp.btrfs_truncate_inode_items 0.11 ± 11% -0.0 0.07 ± 7% perf-profile.children.cycles-pp.btrfs_commit_inode_delayed_inode 0.15 ± 3% -0.0 0.11 ± 6% perf-profile.children.cycles-pp.insert_with_overflow 0.11 ± 4% -0.0 0.08 ± 8% perf-profile.children.cycles-pp.btrfs_lookup 0.11 ± 4% -0.0 0.08 ± 8% perf-profile.children.cycles-pp.btrfs_lookup_dentry 0.16 ± 2% -0.0 0.14 ± 8% perf-profile.children.cycles-pp.btrfs_insert_dir_item 0.21 ± 4% -0.0 0.19 ± 3% perf-profile.children.cycles-pp.fput 0.16 ± 2% -0.0 0.14 ± 5% perf-profile.children.cycles-pp.btrfs_add_link 0.07 ± 5% -0.0 0.06 perf-profile.children.cycles-pp.btrfs_insert_inode_ref 0.01 ±173% +0.0 0.06 ± 14% perf-profile.children.cycles-pp.blk_done_softirq 0.01 ±173% +0.0 0.06 ± 14% perf-profile.children.cycles-pp.scsi_io_completion 0.01 ±173% +0.0 0.06 ± 14% perf-profile.children.cycles-pp.scsi_end_request 0.15 ± 15% +0.0 0.19 ± 2% perf-profile.children.cycles-pp.__vfs_write 0.15 ± 15% +0.0 0.20 ± 2% perf-profile.children.cycles-pp.vfs_write 0.13 ± 17% +0.1 0.18 ± 3% perf-profile.children.cycles-pp.btrfs_file_write_iter 0.01 ±173% +0.1 0.07 ± 13% perf-profile.children.cycles-pp.__filemap_fdatawrite_range 0.00 +0.1 0.05 ± 8% perf-profile.children.cycles-pp.btrfs_delayed_update_inode 0.15 ± 15% +0.1 0.20 ± 2% perf-profile.children.cycles-pp.ksys_write 0.00 +0.1 0.05 ± 9% perf-profile.children.cycles-pp.btrfs_release_path 0.05 ± 70% +0.1 0.11 ± 17% perf-profile.children.cycles-pp.__extent_writepage 0.05 ± 9% +0.1 0.11 ± 7% perf-profile.children.cycles-pp.setup_items_for_insert 0.00 +0.1 0.06 ± 14% perf-profile.children.cycles-pp.btrfs_set_token_32 0.00 +0.1 0.06 perf-profile.children.cycles-pp.btrfs_update_inode 0.06 ± 70% +0.1 0.12 ± 14% perf-profile.children.cycles-pp.extent_write_cache_pages 0.02 ±173% +0.1 0.08 ± 17% perf-profile.children.cycles-pp.btrfs_run_delayed_refs 0.02 ±173% +0.1 0.08 ± 17% perf-profile.children.cycles-pp.__btrfs_run_delayed_refs 0.04 ± 58% +0.1 0.10 ± 4% perf-profile.children.cycles-pp.read_block_for_search 0.00 +0.1 0.06 ± 6% perf-profile.children.cycles-pp.find_extent_buffer 0.03 ±100% +0.1 0.09 ± 7% perf-profile.children.cycles-pp.btrfs_get_token_32 0.09 ± 45% +0.1 0.15 ± 12% perf-profile.children.cycles-pp.do_writepages 0.01 ±173% +0.1 0.08 ± 17% perf-profile.children.cycles-pp.delayed_ref_async_start 0.00 +0.1 0.07 ± 6% perf-profile.children.cycles-pp.btrfs_del_items 0.06 ± 58% +0.1 0.13 ± 5% perf-profile.children.cycles-pp.btrfs_buffered_write 0.07 ± 68% +0.1 0.15 ± 14% perf-profile.children.cycles-pp.extent_writepages 0.00 +0.1 0.09 ± 4% perf-profile.children.cycles-pp.generic_bin_search 0.03 ±173% +0.1 0.12 ± 15% perf-profile.children.cycles-pp.run_one_async_start 0.03 ±173% +0.1 0.12 ± 15% perf-profile.children.cycles-pp.btrfs_submit_bio_start 0.03 ±173% +0.1 0.12 ± 15% perf-profile.children.cycles-pp.btrfs_csum_one_bio 0.03 ±173% +0.1 0.13 ± 14% perf-profile.children.cycles-pp.crc32c 0.00 +0.1 0.10 ± 4% perf-profile.children.cycles-pp.btrfs_try_tree_write_lock 0.00 +0.3 0.28 ± 4% perf-profile.children.cycles-pp.unlock_up 94.68 +0.7 95.39 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 94.63 +0.7 95.34 perf-profile.children.cycles-pp.do_syscall_64 48.36 +0.8 49.16 perf-profile.children.cycles-pp.do_sys_open 39.37 +1.0 40.36 perf-profile.children.cycles-pp.__alloc_fd 33.98 +1.5 35.45 perf-profile.children.cycles-pp.__x64_sys_close 33.78 +1.5 35.25 perf-profile.children.cycles-pp.__close_fd 79.40 +1.5 80.90 perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath 78.19 +2.5 80.73 perf-profile.children.cycles-pp._raw_spin_lock 6.48 ± 12% -0.7 5.74 ± 2% perf-profile.self.cycles-pp.copy_user_enhanced_fast_string 2.26 ± 7% -0.3 2.00 ± 3% perf-profile.self.cycles-pp.file_free_rcu 1.16 ± 10% -0.2 1.00 ± 2% perf-profile.self.cycles-pp.generic_file_read_iter 0.23 ± 24% -0.1 0.12 ± 3% perf-profile.self.cycles-pp.queued_write_lock_slowpath 0.09 ± 24% -0.1 0.03 ±100% perf-profile.self.cycles-pp._raw_spin_lock_irqsave 0.34 ± 5% -0.0 0.30 ± 4% perf-profile.self.cycles-pp.__alloc_file 0.17 ± 2% +0.0 0.20 ± 2% perf-profile.self.cycles-pp.__alloc_fd 0.00 +0.1 0.05 ± 8% perf-profile.self.cycles-pp.generic_bin_search 0.01 ±173% +0.1 0.08 ± 6% perf-profile.self.cycles-pp.btrfs_get_token_32 0.03 ±173% +0.1 0.12 ± 12% perf-profile.self.cycles-pp.__crc32c_le 78.84 +1.5 80.29 perf-profile.self.cycles-pp.native_queued_spin_lock_slowpath blogbench.time.system_time 30000 +-+-----------------------------------------------------------------+ | | 25000 OO+OO O O OO OO OO OO OOO OO OO O OO OOO | |+.++.++.O+.O++.++.++.++.+++.++.++.O+.++.+++.++.++.++.+++.++.++.++.+| | | 20000 +-+ | | | 15000 +-+ | | | 10000 +-+ | | | | | 5000 +-+ | | | 0 +-+----O------------------------------------------------------------+ blogbench.time.percent_of_cpu_this_job_got 9000 +-+------------------------------------------------------------------+ OO.OO.O+.OO.OO.OO.OO.OOO.OO.OO.OO.OO.OO.OO.OO.+++.+ .++.+ .++.++.+ | 8000 +-+ + + + + + + + + +.+| 7000 +-+ | | | 6000 +-+ | 5000 +-+ | | | 4000 +-+ | 3000 +-+ | | | 2000 +-+ | 1000 +-+ | | | 0 +-+----O-------------------------------------------------------------+ blogbench.time.voluntary_context_switches 1.6e+07 +-+---------------------------------------------------------------+ |+.+ .++ +.+ +.+ ++. .+ + .++. | 1.4e+07 +-+ +.+++.++.+++.++ +.+ ++.++.+ ++.++.+ ++.+++ +.+ + +| 1.2e+07 +-+ | | | 1e+07 +-+ | OO OO O O OO OO OO OO OO OOO OO OO OOO OO | 8e+06 +-+ O O | | | 6e+06 +-+ | 4e+06 +-+ | | | 2e+06 +-+ | | | 0 +-+----O----------------------------------------------------------+ blogbench.time.file_system_outputs 8e+07 +-+-----------------------------------------------------------------+ OO OO O OO OOO OO OO OO OOO OO OO OO O OOO | 7e+07 +-+ O | 6e+07 +-+ | | | 5e+07 +-+ | | | 4e+07 +-+ .+ .++. +. +. .+ +. .+ .++ +. | |+.++ + +++.++.++.++.+++.+ + ++.++ + ++.++ + +.+ ++.++.+| 3e+07 +-+ | 2e+07 +-+ | | | 1e+07 +-+ | | | 0 +-+----O------------------------------------------------------------+ blogbench.write_score 8000 OO+-O-O--OO----------O-----------------------------------------------+ | O OO OO OO OO OO OO OO OO OO OO OO | 7000 +-+ | 6000 +-+ | | | 5000 +-+ | | | 4000 +-+++.++.++.++.++.++.+++.++.++.++.++.++.++.++.+++.++.++.++.++.++.++.+| | | 3000 +-+ | 2000 +-+ | | | 1000 +-+ | | | 0 +-+----O-------------------------------------------------------------+ [*] bisect-good sample [O] bisect-bad sample *************************************************************************************************** lkp-hsw-ep6: 56 threads Intel(R) Xeon(R) CPU E5-2695 v3 @ 2.30GHz with 256G memory ========================================================================================= bs/compiler/cpufreq_governor/disk/fs/ioengine/kconfig/nr_task/rootfs/runtime/rw/tbox_group/test_size/testcase/time_based/ucode: 4k/gcc-7/performance/2pmem/btrfs/mmap/x86_64-rhel-7.2/50%/debian-x86_64-2018-04-03.cgz/200s/randwrite/lkp-hsw-ep6/100G/fio-basic/tb/0x3d commit: 9b142115ed ("btrfs: dev-replace: remove pointless assert in write unlock") 5239834016 ("Btrfs: kill btrfs_clear_path_blocking") 9b142115ed359348 523983401644ebeb331c923c28 ---------------- -------------------------- fail:runs %reproduction fail:runs | | | :2 6% 0:2 perf-profile.children.cycles-pp.error_entry :2 50% 1:2 kmsg.pstore:crypto_comp_decompress_failed,ret= :2 50% 1:2 kmsg.pstore:decompression_failed 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