| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <20161013013849.GB24130@yexl-desktop>
Date: Thu, 13 Oct 2016 09:38:49 +0800
From: kernel test robot <xiaolong.ye@...el.com>
To: Chao Yu <yuchao0@...wei.com>
Cc: Jaegeuk Kim <jaegeuk@...nel.org>,
LKML <linux-kernel@...r.kernel.org>,
Linus Torvalds <torvalds@...ux-foundation.org>, lkp@...org
Subject: [lkp] [f2fs] 6ca56ca429: fsmark.files_per_sec +16.4% improvement
FYI, we noticed a +16.4% improvement of fsmark.files_per_sec due to commit:
commit 6ca56ca429aa94399534ec00598f7f9847c4cae2 ("f2fs: don't submit irrelevant page")
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
in testcase: fsmark
on test machine: 48 threads Intel(R) Xeon(R) CPU E5-2697 v2 @ 2.70GHz with 64G memory
with following parameters:
iterations: 1x
nr_threads: 64t
disk: 1BRD_48G
fs: f2fs
filesize: 4M
test_size: 40G
sync_method: fsyncBeforeClose
cpufreq_governor: performance
The fsmark is a file system benchmark to test synchronous write workloads, for example, mail servers workload.
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.
Details are as below:
-------------------------------------------------------------------------------------------------->
To reproduce:
git clone git://git.kernel.org/pub/scm/linux/kernel/git/wfg/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/filesize/fs/iterations/kconfig/nr_threads/rootfs/sync_method/tbox_group/test_size/testcase:
gcc-6/performance/1BRD_48G/4M/f2fs/1x/x86_64-rhel-7.2/64t/debian-x86_64-2016-08-31.cgz/fsyncBeforeClose/ivb44/40G/fsmark
commit:
3f5f4959b1 ("f2fs: fix to commit bio cache after flushing node pages")
6ca56ca429 ("f2fs: don't submit irrelevant page")
3f5f4959b144d9ba 6ca56ca429aa94399534ec0059
---------------- --------------------------
%stddev %change %stddev
\ | \
386639 ± 5% -25.5% 288161 ± 1% fsmark.app_overhead
239.25 ± 0% +16.4% 278.55 ± 0% fsmark.files_per_sec
43.20 ± 0% -10.2% 38.79 ± 6% fsmark.time.elapsed_time
43.20 ± 0% -10.2% 38.79 ± 6% fsmark.time.elapsed_time.max
46992 ± 1% -80.4% 9199 ± 1% fsmark.time.involuntary_context_switches
46078 ± 9% -25.7% 34237 ± 2% fsmark.time.minor_page_faults
3395 ± 0% -45.8% 1839 ± 6% fsmark.time.percent_of_cpu_this_job_got
1466 ± 0% -51.6% 710.26 ± 0% fsmark.time.system_time
2200312 ± 3% -44.6% 1218799 ± 1% fsmark.time.voluntary_context_switches
57255 ± 12% -37.2% 35955 ± 7% interrupts.CAL:Function_call_interrupts
8946 ± 4% -10.4% 8012 ± 3% numa-meminfo.node0.KernelStack
6085 ± 6% +15.5% 7031 ± 4% numa-meminfo.node1.KernelStack
126393 ± 44% +64.2% 207479 ± 31% numa-numastat.node0.numa_miss
127512 ± 47% +62.4% 207048 ± 28% numa-numastat.node1.numa_foreign
12923 ± 5% +8.5% 14017 ± 5% slabinfo.kmalloc-512.active_objs
12968 ± 5% +9.1% 14142 ± 5% slabinfo.kmalloc-512.num_objs
8960 ± 4% -10.0% 8061 ± 3% numa-vmstat.node0.nr_kernel_stack
6110 ± 6% +15.7% 7069 ± 4% numa-vmstat.node1.nr_kernel_stack
87648 ± 28% -41.4% 51367 ± 16% numa-vmstat.node1.numa_miss
343170 ± 5% -63.6% 125068 ± 3% softirqs.RCU
265499 ± 2% -55.0% 119492 ± 1% softirqs.SCHED
932841 ± 0% -36.8% 589681 ± 0% softirqs.TIMER
918887 ± 0% +10.7% 1016759 ± 5% vmstat.io.bo
36.00 ± 7% -50.7% 17.75 ± 8% vmstat.procs.r
91317 ± 2% -33.1% 61113 ± 6% vmstat.system.cs
4.78 ± 24% -47.7% 2.50 ± 47% sched_debug.cpu.cpu_load[0].avg
327.26 ± 26% -43.8% 183.96 ± 25% sched_debug.cpu.curr->pid.avg
4634 ± 23% -37.8% 2881 ± 26% sched_debug.cpu.load.avg
7074 ± 7% -13.4% 6123 ± 10% sched_debug.cpu.load.stddev
0.32 ± 24% -41.0% 0.19 ± 26% sched_debug.cpu.nr_running.avg
34586 ± 13% -34.6% 22617 ± 4% proc-vmstat.numa_hint_faults
25428 ± 15% -36.9% 16055 ± 8% proc-vmstat.numa_hint_faults_local
1643 ± 13% -32.9% 1103 ± 7% proc-vmstat.numa_pages_migrated
38023 ± 12% -31.8% 25918 ± 3% proc-vmstat.numa_pte_updates
27488 ± 10% -30.2% 19195 ± 10% proc-vmstat.pgactivate
160759 ± 3% -15.4% 135954 ± 2% proc-vmstat.pgfault
70.35 ± 0% -44.6% 38.99 ± 5% turbostat.%Busy
2103 ± 0% -44.6% 1165 ± 5% turbostat.Avg_MHz
25.32 ± 0% +70.0% 43.06 ± 1% turbostat.CPU%c1
1.09 ± 5% +16.4% 1.27 ± 4% turbostat.CPU%c3
3.24 ± 0% +415.0% 16.69 ± 16% turbostat.CPU%c6
144.66 ± 0% -14.2% 124.19 ± 3% turbostat.CorWatt
176.91 ± 0% -11.5% 156.56 ± 2% turbostat.PkgWatt
90937705 ± 2% +15.0% 1.046e+08 ± 2% cpuidle.C1-IVT.time
899621 ± 3% -10.1% 808719 ± 3% cpuidle.C1-IVT.usage
71892754 ± 4% -83.3% 11995117 ± 8% cpuidle.C1E-IVT.time
494948 ± 4% -76.3% 117427 ± 2% cpuidle.C1E-IVT.usage
1.291e+08 ± 4% -66.6% 43150497 ± 3% cpuidle.C3-IVT.time
716877 ± 4% -69.8% 216236 ± 2% cpuidle.C3-IVT.usage
3.416e+08 ± 2% +196.0% 1.011e+09 ± 11% cpuidle.C6-IVT.time
677099 ± 1% +101.7% 1365821 ± 9% cpuidle.C6-IVT.usage
308611 ± 34% +765.0% 2669407 ±106% cpuidle.POLL.time
1053 ± 2% +57.9% 1662 ± 3% cpuidle.POLL.usage
4.36e+11 ± 0% -40.5% 2.592e+11 ± 2% perf-stat.branch-instructions
0.12 ± 3% +72.4% 0.21 ± 4% perf-stat.branch-miss-rate%
43.93 ± 1% +8.6% 47.70 ± 3% perf-stat.cache-miss-rate%
2.256e+09 ± 0% +10.1% 2.482e+09 ± 1% perf-stat.cache-misses
4170241 ± 3% -39.6% 2518308 ± 1% perf-stat.context-switches
4.498e+12 ± 0% -45.8% 2.436e+12 ± 1% perf-stat.cpu-cycles
375218 ± 4% -54.8% 169686 ± 0% perf-stat.cpu-migrations
0.43 ± 0% +66.4% 0.72 ± 1% perf-stat.dTLB-load-miss-rate%
1.909e+09 ± 1% -7.5% 1.766e+09 ± 1% perf-stat.dTLB-load-misses
4.392e+11 ± 1% -44.5% 2.436e+11 ± 2% perf-stat.dTLB-loads
0.07 ± 6% -13.2% 0.06 ± 3% perf-stat.dTLB-store-miss-rate%
32271724 ± 6% -25.0% 24187747 ± 6% perf-stat.dTLB-store-misses
4.764e+10 ± 1% -13.7% 4.112e+10 ± 3% perf-stat.dTLB-stores
25.37 ± 6% +56.4% 39.67 ± 4% perf-stat.iTLB-load-miss-rate%
74032060 ± 2% -52.4% 35249761 ± 1% perf-stat.iTLB-loads
1.609e+12 ± 0% -38.2% 9.945e+11 ± 2% perf-stat.instructions
64272 ± 7% -33.1% 43021 ± 7% perf-stat.instructions-per-iTLB-miss
0.36 ± 0% +14.1% 0.41 ± 0% perf-stat.ipc
149838 ± 2% -15.1% 127269 ± 2% perf-stat.minor-faults
40.38 ± 0% -2.1% 39.56 ± 0% perf-stat.node-load-miss-rate%
6.614e+08 ± 0% -5.4% 6.255e+08 ± 0% perf-stat.node-load-misses
9.763e+08 ± 0% -2.1% 9.558e+08 ± 0% perf-stat.node-loads
149847 ± 2% -15.1% 127289 ± 2% perf-stat.page-faults
245414 ± 9% +317.5% 1024622 ±114% latency_stats.avg.max
1850877 ± 4% -97.1% 54167 ± 3% latency_stats.hits.call_rwsem_down_read_failed.get_node_info.[f2fs].write_data_page.[f2fs].do_write_data_page.[f2fs].f2fs_write_data_page.[f2fs].f2fs_write_cache_pages.[f2fs].f2fs_write_data_pages.[f2fs].do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range.f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs]
65.00 ± 5% +15866.5% 10378 ± 2% latency_stats.hits.call_rwsem_down_write_failed.f2fs_submit_page_mbio.[f2fs].do_write_page.[f2fs].write_data_page.[f2fs].do_write_data_page.[f2fs].f2fs_write_data_page.[f2fs].f2fs_write_cache_pages.[f2fs].f2fs_write_data_pages.[f2fs].do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range.f2fs_do_sync_file.[f2fs]
14504 ± 4% -96.3% 535.00 ± 6% latency_stats.hits.call_rwsem_down_write_failed.set_node_addr.[f2fs].f2fs_write_node_page.[f2fs].fsync_node_pages.[f2fs].f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range.do_fsync.SyS_fsync.entry_SYSCALL_64_fastpath
265151 ± 3% +313.9% 1097572 ± 1% latency_stats.hits.do_write_page.[f2fs].write_data_page.[f2fs].do_write_data_page.[f2fs].f2fs_write_data_page.[f2fs].f2fs_write_cache_pages.[f2fs].f2fs_write_data_pages.[f2fs].do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range.f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range
4803 ± 83% -99.9% 7.00 ± 30% latency_stats.max.call_rwsem_down_write_failed.f2fs_submit_page_mbio.[f2fs].do_write_page.[f2fs].write_node_page.[f2fs].f2fs_write_node_page.[f2fs].fsync_node_pages.[f2fs].f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range.do_fsync.SyS_fsync.entry_SYSCALL_64_fastpath
515558 ± 21% +204.5% 1570079 ± 91% latency_stats.max.f2fs_sync_fs.[f2fs].f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range.do_fsync.SyS_fsync.entry_SYSCALL_64_fastpath
515558 ± 21% +407.2% 2614921 ± 75% latency_stats.max.max
0.00 ± -1% +Inf% 108643 ± 14% latency_stats.sum.call_rwsem_down_read_failed.f2fs_submit_merged_bio_cond.[f2fs].f2fs_write_cache_pages.[f2fs].f2fs_write_data_pages.[f2fs].do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range.f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range.do_fsync.SyS_fsync
1353580 ± 5% -99.8% 2678 ± 11% latency_stats.sum.call_rwsem_down_read_failed.get_node_info.[f2fs].f2fs_write_node_page.[f2fs].fsync_node_pages.[f2fs].f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range.do_fsync.SyS_fsync.entry_SYSCALL_64_fastpath
1419760 ± 6% -82.8% 244115 ± 39% latency_stats.sum.call_rwsem_down_read_failed.get_node_info.[f2fs].new_node_page.[f2fs].get_dnode_of_data.[f2fs].f2fs_map_blocks.[f2fs].f2fs_preallocate_blocks.[f2fs].f2fs_file_write_iter.[f2fs].__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
502303 ± 3% -99.8% 760.75 ± 41% latency_stats.sum.call_rwsem_down_read_failed.get_node_info.[f2fs].new_node_page.[f2fs].new_inode_page.[f2fs].init_inode_metadata.[f2fs].f2fs_add_regular_entry.[f2fs].__f2fs_do_add_link.[f2fs].__f2fs_add_link.[f2fs].f2fs_create.[f2fs].path_openat.do_filp_open.do_sys_open
4.994e+08 ± 5% -99.3% 3466489 ± 8% latency_stats.sum.call_rwsem_down_read_failed.get_node_info.[f2fs].write_data_page.[f2fs].do_write_data_page.[f2fs].f2fs_write_data_page.[f2fs].f2fs_write_cache_pages.[f2fs].f2fs_write_data_pages.[f2fs].do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range.f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs]
566230 ± 5% -99.8% 1017 ± 81% latency_stats.sum.call_rwsem_down_read_failed.is_checkpointed_node.[f2fs].f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range.do_fsync.SyS_fsync.entry_SYSCALL_64_fastpath
265023 ± 7% -99.9% 188.50 ± 59% latency_stats.sum.call_rwsem_down_read_failed.need_dentry_mark.[f2fs].fsync_node_pages.[f2fs].f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range.do_fsync.SyS_fsync.entry_SYSCALL_64_fastpath
812179 ± 5% -99.8% 1663 ± 16% latency_stats.sum.call_rwsem_down_read_failed.need_inode_block_update.[f2fs].f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range.do_fsync.SyS_fsync.entry_SYSCALL_64_fastpath
6243 ± 5% +11505.4% 724556 ± 2% latency_stats.sum.call_rwsem_down_write_failed.f2fs_submit_page_mbio.[f2fs].do_write_page.[f2fs].write_data_page.[f2fs].do_write_data_page.[f2fs].f2fs_write_data_page.[f2fs].f2fs_write_cache_pages.[f2fs].f2fs_write_data_pages.[f2fs].do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range.f2fs_do_sync_file.[f2fs]
4815 ± 83% -99.8% 10.25 ± 63% latency_stats.sum.call_rwsem_down_write_failed.f2fs_submit_page_mbio.[f2fs].do_write_page.[f2fs].write_node_page.[f2fs].f2fs_write_node_page.[f2fs].fsync_node_pages.[f2fs].f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range.do_fsync.SyS_fsync.entry_SYSCALL_64_fastpath
2235322 ± 2% -91.7% 185288 ± 12% latency_stats.sum.call_rwsem_down_write_failed.get_node_info.[f2fs].new_node_page.[f2fs].get_dnode_of_data.[f2fs].f2fs_map_blocks.[f2fs].f2fs_preallocate_blocks.[f2fs].f2fs_file_write_iter.[f2fs].__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
1626956 ± 3% -99.6% 6340 ± 28% latency_stats.sum.call_rwsem_down_write_failed.get_node_info.[f2fs].new_node_page.[f2fs].new_inode_page.[f2fs].init_inode_metadata.[f2fs].f2fs_add_regular_entry.[f2fs].__f2fs_do_add_link.[f2fs].__f2fs_add_link.[f2fs].f2fs_create.[f2fs].path_openat.do_filp_open.do_sys_open
3270833 ± 4% -99.6% 12180 ± 14% latency_stats.sum.call_rwsem_down_write_failed.set_node_addr.[f2fs].f2fs_write_node_page.[f2fs].fsync_node_pages.[f2fs].f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range.do_fsync.SyS_fsync.entry_SYSCALL_64_fastpath
1616277 ± 5% -96.8% 51120 ± 39% latency_stats.sum.call_rwsem_down_write_failed.set_node_addr.[f2fs].new_node_page.[f2fs].get_dnode_of_data.[f2fs].f2fs_map_blocks.[f2fs].f2fs_preallocate_blocks.[f2fs].f2fs_file_write_iter.[f2fs].__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
374018 ± 11% -82.0% 67194 ± 35% latency_stats.sum.call_rwsem_down_write_failed.set_node_addr.[f2fs].new_node_page.[f2fs].new_inode_page.[f2fs].init_inode_metadata.[f2fs].f2fs_add_inline_entry.[f2fs].__f2fs_do_add_link.[f2fs].__f2fs_add_link.[f2fs].f2fs_create.[f2fs].path_openat.do_filp_open.do_sys_open
1258435 ± 4% -98.6% 18193 ± 8% latency_stats.sum.call_rwsem_down_write_failed.set_node_addr.[f2fs].new_node_page.[f2fs].new_inode_page.[f2fs].init_inode_metadata.[f2fs].f2fs_add_regular_entry.[f2fs].__f2fs_do_add_link.[f2fs].__f2fs_add_link.[f2fs].f2fs_create.[f2fs].path_openat.do_filp_open.do_sys_open
3.776e+08 ± 8% +312.4% 1.557e+09 ± 1% latency_stats.sum.do_write_page.[f2fs].write_data_page.[f2fs].do_write_data_page.[f2fs].f2fs_write_data_page.[f2fs].f2fs_write_cache_pages.[f2fs].f2fs_write_data_pages.[f2fs].do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range.f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range
20975 ±148% -100.0% 0.00 ± -1% latency_stats.sum.do_write_page.[f2fs].write_node_page.[f2fs].f2fs_write_node_page.[f2fs].fsync_node_pages.[f2fs].f2fs_do_sync_file.[f2fs].f2fs_sync_file.[f2fs].vfs_fsync_range.do_fsync.SyS_fsync.entry_SYSCALL_64_fastpath
87.20 ± 0% -16.4% 72.89 ± 4% perf-profile.calltrace.cycles-pp.__filemap_fdatawrite_range.filemap_write_and_wait_range.f2fs_do_sync_file.f2fs_sync_file.vfs_fsync_range
1.27 ± 1% +65.2% 2.10 ± 1% perf-profile.calltrace.cycles-pp.__generic_file_write_iter.f2fs_file_write_iter.__vfs_write.vfs_write.sys_write
78.05 ± 1% -24.2% 59.18 ± 5% perf-profile.calltrace.cycles-pp.__mutex_lock_slowpath.mutex_lock.do_write_page.write_data_page.do_write_data_page
0.00 ± -1% +Inf% 1.06 ± 10% perf-profile.calltrace.cycles-pp.__mutex_unlock_slowpath.mutex_unlock.do_write_page.write_data_page.do_write_data_page
4.34 ± 22% +48.5% 6.44 ± 15% perf-profile.calltrace.cycles-pp.__percpu_counter_sum.f2fs_balance_fs.f2fs_write_data_page.f2fs_write_cache_pages.f2fs_write_data_pages
1.21 ± 1% +50.2% 1.81 ± 7% perf-profile.calltrace.cycles-pp.__submit_merged_bio.f2fs_submit_page_mbio.do_write_page.write_data_page.do_write_data_page
2.81 ± 33% +52.7% 4.29 ± 21% perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.__percpu_counter_sum.f2fs_balance_fs.f2fs_write_data_page.f2fs_write_cache_pages
1.00 ± 1% +56.1% 1.57 ± 7% perf-profile.calltrace.cycles-pp.brd_do_bvec.brd_make_request.generic_make_request.submit_bio.__submit_merged_bio
1.20 ± 1% +50.7% 1.81 ± 7% perf-profile.calltrace.cycles-pp.brd_make_request.generic_make_request.submit_bio.__submit_merged_bio.f2fs_submit_page_mbio
9.74 ± 0% +125.9% 21.99 ± 5% perf-profile.calltrace.cycles-pp.call_cpuidle.cpu_startup_entry.start_secondary
10.37 ± 1% +117.9% 22.60 ± 6% perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary
9.73 ± 0% +125.9% 21.99 ± 5% perf-profile.calltrace.cycles-pp.cpuidle_enter.call_cpuidle.cpu_startup_entry.start_secondary
9.55 ± 0% +123.9% 21.38 ± 4% perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.call_cpuidle.cpu_startup_entry.start_secondary
87.41 ± 0% -35.7% 56.20 ± 57% perf-profile.calltrace.cycles-pp.do_fsync.sys_fsync.entry_SYSCALL_64_fastpath
81.97 ± 1% -20.5% 65.15 ± 5% perf-profile.calltrace.cycles-pp.do_write_data_page.f2fs_write_data_page.f2fs_write_cache_pages.f2fs_write_data_pages.do_writepages
80.55 ± 1% -21.1% 63.55 ± 5% perf-profile.calltrace.cycles-pp.do_write_page.write_data_page.do_write_data_page.f2fs_write_data_page.f2fs_write_cache_pages
87.20 ± 0% -16.4% 72.89 ± 4% perf-profile.calltrace.cycles-pp.do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range.f2fs_do_sync_file.f2fs_sync_file
89.19 ± 0% -34.9% 58.08 ± 57% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_fastpath
4.62 ± 21% +48.2% 6.84 ± 14% perf-profile.calltrace.cycles-pp.f2fs_balance_fs.f2fs_write_data_page.f2fs_write_cache_pages.f2fs_write_data_pages.do_writepages
87.41 ± 0% -16.5% 72.99 ± 4% perf-profile.calltrace.cycles-pp.f2fs_do_sync_file.f2fs_sync_file.vfs_fsync_range.do_fsync.sys_fsync
1.55 ± 2% +51.4% 2.34 ± 1% perf-profile.calltrace.cycles-pp.f2fs_file_write_iter.__vfs_write.vfs_write.sys_write.entry_SYSCALL_64_fastpath
1.42 ± 1% +49.7% 2.13 ± 7% perf-profile.calltrace.cycles-pp.f2fs_submit_page_mbio.do_write_page.write_data_page.do_write_data_page.f2fs_write_data_page
87.41 ± 0% -16.5% 72.99 ± 4% perf-profile.calltrace.cycles-pp.f2fs_sync_file.vfs_fsync_range.do_fsync.sys_fsync.entry_SYSCALL_64_fastpath
87.11 ± 0% -16.3% 72.88 ± 4% perf-profile.calltrace.cycles-pp.f2fs_write_cache_pages.f2fs_write_data_pages.do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range
86.92 ± 0% -16.6% 72.52 ± 4% perf-profile.calltrace.cycles-pp.f2fs_write_data_page.f2fs_write_cache_pages.f2fs_write_data_pages.do_writepages.__filemap_fdatawrite_range
87.20 ± 0% -16.4% 72.89 ± 4% perf-profile.calltrace.cycles-pp.f2fs_write_data_pages.do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range.f2fs_do_sync_file
87.20 ± 0% -16.4% 72.89 ± 4% perf-profile.calltrace.cycles-pp.filemap_write_and_wait_range.f2fs_do_sync_file.f2fs_sync_file.vfs_fsync_range.do_fsync
1.21 ± 1% +50.2% 1.81 ± 7% perf-profile.calltrace.cycles-pp.generic_make_request.submit_bio.__submit_merged_bio.f2fs_submit_page_mbio.do_write_page
1.25 ± 1% +66.3% 2.08 ± 1% perf-profile.calltrace.cycles-pp.generic_perform_write.__generic_file_write_iter.f2fs_file_write_iter.__vfs_write.vfs_write
9.51 ± 0% +121.5% 21.06 ± 3% perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.cpuidle_enter.call_cpuidle.cpu_startup_entry
78.14 ± 1% -24.1% 59.32 ± 5% perf-profile.calltrace.cycles-pp.mutex_lock.do_write_page.write_data_page.do_write_data_page.f2fs_write_data_page
78.01 ± 1% -24.3% 59.02 ± 4% perf-profile.calltrace.cycles-pp.mutex_optimistic_spin.__mutex_lock_slowpath.mutex_lock.do_write_page.write_data_page
1.97 ± 1% +46.0% 2.88 ± 8% perf-profile.calltrace.cycles-pp.mutex_spin_on_owner.mutex_optimistic_spin.__mutex_lock_slowpath.mutex_lock.do_write_page
0.00 ± -1% +Inf% 1.22 ± 10% perf-profile.calltrace.cycles-pp.mutex_unlock.do_write_page.write_data_page.do_write_data_page.f2fs_write_data_page
2.63 ± 36% +51.6% 3.98 ± 22% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.__percpu_counter_sum.f2fs_balance_fs.f2fs_write_data_page
75.37 ± 1% -26.4% 55.51 ± 4% perf-profile.calltrace.cycles-pp.osq_lock.mutex_optimistic_spin.__mutex_lock_slowpath.mutex_lock.do_write_page
10.38 ± 1% +117.9% 22.62 ± 6% perf-profile.calltrace.cycles-pp.start_secondary
1.21 ± 1% +50.2% 1.81 ± 7% perf-profile.calltrace.cycles-pp.submit_bio.__submit_merged_bio.f2fs_submit_page_mbio.do_write_page.write_data_page
87.41 ± 0% -35.7% 56.20 ± 57% perf-profile.calltrace.cycles-pp.sys_fsync.entry_SYSCALL_64_fastpath
87.41 ± 0% -35.7% 56.20 ± 57% perf-profile.calltrace.cycles-pp.vfs_fsync_range.do_fsync.sys_fsync.entry_SYSCALL_64_fastpath
0.00 ± -1% +Inf% 0.97 ± 10% perf-profile.calltrace.cycles-pp.wake_up_q.__mutex_unlock_slowpath.mutex_unlock.do_write_page.write_data_page
81.53 ± 1% -20.8% 64.58 ± 5% perf-profile.calltrace.cycles-pp.write_data_page.do_write_data_page.f2fs_write_data_page.f2fs_write_cache_pages.f2fs_write_data_pages
0.66 ± 3% +54.4% 1.02 ± 7% perf-profile.children.cycles-pp.__alloc_pages_nodemask
87.21 ± 0% -16.4% 72.94 ± 4% perf-profile.children.cycles-pp.__filemap_fdatawrite_range
1.28 ± 1% +64.6% 2.12 ± 1% perf-profile.children.cycles-pp.__generic_file_write_iter
78.06 ± 1% -24.1% 59.23 ± 4% perf-profile.children.cycles-pp.__mutex_lock_slowpath
0.23 ± 11% +358.1% 1.06 ± 10% perf-profile.children.cycles-pp.__mutex_unlock_slowpath
4.36 ± 22% +48.6% 6.47 ± 15% perf-profile.children.cycles-pp.__percpu_counter_sum
1.29 ± 1% +50.8% 1.95 ± 7% perf-profile.children.cycles-pp.__submit_merged_bio
1.60 ± 2% +50.7% 2.42 ± 1% perf-profile.children.cycles-pp.__vfs_write
3.07 ± 31% +61.0% 4.94 ± 18% perf-profile.children.cycles-pp._raw_spin_lock_irqsave
0.68 ± 3% +52.9% 1.04 ± 7% perf-profile.children.cycles-pp.alloc_pages_current
0.60 ± 6% +50.0% 0.90 ± 39% perf-profile.children.cycles-pp.apic_timer_interrupt
1.08 ± 1% +56.5% 1.68 ± 7% perf-profile.children.cycles-pp.brd_do_bvec
1.29 ± 1% +51.0% 1.94 ± 7% perf-profile.children.cycles-pp.brd_make_request
9.93 ± 0% +136.1% 23.45 ± 13% perf-profile.children.cycles-pp.call_cpuidle
10.58 ± 1% +127.6% 24.07 ± 13% perf-profile.children.cycles-pp.cpu_startup_entry
9.93 ± 0% +136.1% 23.45 ± 13% perf-profile.children.cycles-pp.cpuidle_enter
9.72 ± 0% +134.4% 22.79 ± 12% perf-profile.children.cycles-pp.cpuidle_enter_state
87.42 ± 0% -16.4% 73.04 ± 4% perf-profile.children.cycles-pp.do_fsync
81.98 ± 1% -20.5% 65.19 ± 5% perf-profile.children.cycles-pp.do_write_data_page
80.56 ± 1% -21.1% 63.60 ± 5% perf-profile.children.cycles-pp.do_write_page
87.21 ± 0% -16.4% 72.94 ± 4% perf-profile.children.cycles-pp.do_writepages
89.27 ± 0% -15.2% 75.68 ± 4% perf-profile.children.cycles-pp.entry_SYSCALL_64_fastpath
4.62 ± 21% +48.4% 6.86 ± 13% perf-profile.children.cycles-pp.f2fs_balance_fs
87.42 ± 0% -16.4% 73.04 ± 4% perf-profile.children.cycles-pp.f2fs_do_sync_file
1.55 ± 2% +51.5% 2.34 ± 0% perf-profile.children.cycles-pp.f2fs_file_write_iter
1.43 ± 1% +49.7% 2.14 ± 7% perf-profile.children.cycles-pp.f2fs_submit_page_mbio
87.42 ± 0% -16.4% 73.04 ± 4% perf-profile.children.cycles-pp.f2fs_sync_file
87.11 ± 0% -16.3% 72.94 ± 4% perf-profile.children.cycles-pp.f2fs_write_cache_pages
86.94 ± 0% -16.5% 72.58 ± 4% perf-profile.children.cycles-pp.f2fs_write_data_page
87.21 ± 0% -16.4% 72.94 ± 4% perf-profile.children.cycles-pp.f2fs_write_data_pages
87.21 ± 0% -16.4% 72.94 ± 4% perf-profile.children.cycles-pp.filemap_write_and_wait_range
1.29 ± 1% +51.1% 1.95 ± 7% perf-profile.children.cycles-pp.generic_make_request
1.26 ± 1% +65.5% 2.09 ± 1% perf-profile.children.cycles-pp.generic_perform_write
9.69 ± 0% +120.5% 21.36 ± 3% perf-profile.children.cycles-pp.intel_idle
78.32 ± 1% -23.9% 59.63 ± 4% perf-profile.children.cycles-pp.mutex_lock
78.02 ± 1% -24.3% 59.08 ± 4% perf-profile.children.cycles-pp.mutex_optimistic_spin
1.97 ± 1% +46.3% 2.89 ± 8% perf-profile.children.cycles-pp.mutex_spin_on_owner
0.39 ± 6% +239.7% 1.32 ± 9% perf-profile.children.cycles-pp.mutex_unlock
2.90 ± 33% +64.2% 4.77 ± 19% perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath
75.38 ± 1% -26.3% 55.55 ± 4% perf-profile.children.cycles-pp.osq_lock
0.00 ± -1% +Inf% 1.30 ±160% perf-profile.children.cycles-pp.poll_idle
0.21 ± 3% +601.2% 1.47 ±130% perf-profile.children.cycles-pp.rest_init
0.21 ± 3% +601.2% 1.47 ±130% perf-profile.children.cycles-pp.start_kernel
10.38 ± 1% +117.9% 22.62 ± 6% perf-profile.children.cycles-pp.start_secondary
1.29 ± 1% +51.1% 1.95 ± 7% perf-profile.children.cycles-pp.submit_bio
87.42 ± 0% -16.4% 73.04 ± 4% perf-profile.children.cycles-pp.sys_fsync
1.64 ± 1% +50.5% 2.46 ± 1% perf-profile.children.cycles-pp.sys_write
0.64 ± 5% +62.4% 1.04 ± 6% perf-profile.children.cycles-pp.try_to_wake_up
87.42 ± 0% -16.4% 73.04 ± 4% perf-profile.children.cycles-pp.vfs_fsync_range
1.63 ± 2% +50.5% 2.45 ± 1% perf-profile.children.cycles-pp.vfs_write
0.65 ± 5% +63.7% 1.06 ± 7% perf-profile.children.cycles-pp.wake_up_q
81.53 ± 1% -20.7% 64.63 ± 5% perf-profile.children.cycles-pp.write_data_page
0.21 ± 3% +601.2% 1.47 ±130% perf-profile.children.cycles-pp.x86_64_start_kernel
0.21 ± 3% +601.2% 1.47 ±130% perf-profile.children.cycles-pp.x86_64_start_reservations
1.19 ± 5% +40.9% 1.68 ± 9% perf-profile.self.cycles-pp.__percpu_counter_sum
9.69 ± 0% +120.5% 21.36 ± 3% perf-profile.self.cycles-pp.intel_idle
1.97 ± 1% +46.3% 2.89 ± 8% perf-profile.self.cycles-pp.mutex_spin_on_owner
2.90 ± 33% +64.2% 4.77 ± 19% perf-profile.self.cycles-pp.native_queued_spin_lock_slowpath
75.38 ± 1% -26.3% 55.55 ± 4% perf-profile.self.cycles-pp.osq_lock
0.00 ± -1% +Inf% 1.30 ±160% perf-profile.self.cycles-pp.poll_idle
fsmark.files_per_sec
300 ++--------------------------------------------------------------------+
O O O OO O O O O O O O OO O O O O O O O O OO O O OO O
250 ++O OO O O O .*. O O O O O O |
*.*.**.*.*.**.*.*.**.*.*.**.*.*.** *.* *.*.*.**.*.*.**.*.*.**.*
| : * * |
200 ++ : : : |
| : : : |
150 ++ : : : : |
| : : : : |
100 ++ :: : : |
| :: : : |
| :: : : |
50 ++ : : |
| : : |
0 ++--------------------------------------*---*-------------------------+
[*] bisect-good sample
[O] bisect-bad sample
Thanks,
Xiaolong
View attachment "config-4.8.0-rc3-00094-g6ca56ca" of type "text/plain" (152801 bytes)
View attachment "job-script" of type "text/plain" (7218 bytes)
View attachment "job.yaml" of type "text/plain" (4748 bytes)
View attachment "reproduce" of type "text/plain" (1265 bytes)
Powered by blists - more mailing lists