- 12 Sep, 2023 25 commits
-
-
Namhyung Kim authored
The --lock-cgroup option shows lock contention stats break down by cgroups. Add LOCK_AGGR_CGROUP mode and use it instead of use_cgroup field. $ sudo ./perf lock con -ab --lock-cgroup sleep 1 contended total wait max wait avg wait cgroup 8 15.70 us 6.34 us 1.96 us / 2 1.48 us 747 ns 738 ns /user.slice/.../app.slice/app-gnome-google\x2dchrome-6442.scope 1 848 ns 848 ns 848 ns /user.slice/.../session.slice/org.gnome.Shell@x11.service 1 220 ns 220 ns 220 ns /user.slice/.../session.slice/pipewire-pulse.service For now, the cgroup mode only works with BPF (-b). Committer notes: Remove -g as it is used in the other tools with a clear meaning of collect/show callchains. As agreed with Namhyung off list. Reviewed-by:Ian Rogers <irogers@google.com> Signed-off-by:
Namhyung Kim <namhyung@kernel.org> Tested-by:
Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Hao Luo <haoluo@google.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Song Liu <song@kernel.org> Cc: bpf@vger.kernel.org Link: https://lore.kernel.org/r/20230906174903.346486-4-namhyung@kernel.orgSigned-off-by:
-
Namhyung Kim authored
Save cgroup info and display cgroup names if requested. This is a preparation for the next patch. Reviewed-by:
-
Namhyung Kim authored
The read_all_cgroups() is to build a tree of cgroups in the system and users can look up a cgroup using __cgroup_find(). Committer notes: Had to do this to cover that #else block: -static inline u64 __read_cgroup_id(const char *path) { return -1ULL; } +static inline u64 __read_cgroup_id(const char *path __maybe_unused) { return -1ULL; } Reviewed-by: -
Yang Jihong authored
Use BPF to collect statistics on softirq events based on perf BPF skeletons. Example usage: # perf kwork top -b Starting trace, Hit <Ctrl+C> to stop and report ^C Total : 135445.704 ms, 8 cpus %Cpu(s): 28.35% id, 0.00% hi, 0.25% si %Cpu0 [|||||||||||||||||||| 69.85%] %Cpu1 [|||||||||||||||||||||| 74.10%] %Cpu2 [||||||||||||||||||||| 71.18%] %Cpu3 [|||||||||||||||||||| 69.61%] %Cpu4 [|||||||||||||||||||||| 74.05%] %Cpu5 [|||||||||||||||||||| 69.33%] %Cpu6 [|||||||||||||||||||| 69.71%] %Cpu7 [|||||||||||||||||||||| 73.77%] PID SPID %CPU RUNTIME COMMMAND ------------------------------------------------------------- 0 0 30.43 5271.005 ms [swapper/5] 0 0 30.17 5226.644 ms [swapper/3] 0 0 30.08 5210.257 ms [swapper/6] 0 0 29.89 5177.177 ms [swapper/0] 0 0 28.51 4938.672 ms [swapper/2] 0 0 25.93 4223.464 ms [swapper/7] 0 0 25.69 4181.411 ms [swapper/4] 0 0 25.63 4173.804 ms [swapper/1] 16665 16265 2.16 360.600 ms sched-messaging 16537 16265 2.05 356.275 ms sched-messaging 16503 16265 2.01 343.063 ms sched-messaging 16424 16265 1.97 336.876 ms sched-messaging 16580 16265 1.94 323.658 ms sched-messaging 16515 16265 1.92 321.616 ms sched-messaging 16659 16265 1.91 325.538 ms sched-messaging 16634 16265 1.88 327.766 ms sched-messaging 16454 16265 1.87 326.843 ms sched-messaging 16382 16265 1.87 322.591 ms sched-messaging 16642 16265 1.86 320.506 ms sched-messaging 16582 16265 1.86 320.164 ms sched-messaging 16315 16265 1.86 326.872 ms sched-messaging 16637 16265 1.85 323.766 ms sched-messaging 16506 16265 1.82 311.688 ms sched-messaging 16512 16265 1.81 304.643 ms sched-messaging 16560 16265 1.80 314.751 ms sched-messaging 16320 16265 1.80 313.405 ms sched-messaging 16442 16265 1.80 314.403 ms sched-messaging 16626 16265 1.78 295.380 ms sched-messaging 16600 16265 1.77 309.444 ms sched-messaging 16550 16265 1.76 301.161 ms sched-messaging 16525 16265 1.75 296.560 ms sched-messaging 16314 16265 1.75 298.338 ms sched-messaging 16595 16265 1.74 304.390 ms sched-messaging 16555 16265 1.74 287.564 ms sched-messaging 16520 16265 1.74 295.734 ms sched-messaging 16507 16265 1.73 293.956 ms sched-messaging 16593 16265 1.72 296.443 ms sched-messaging 16531 16265 1.72 299.950 ms sched-messaging 16281 16265 1.72 301.339 ms sched-messaging <SNIP> Reviewed-by:Yang Jihong <yangjihong1@huawei.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ravi Bangoria <ravi.bangoria@amd.com> Cc: Sandipan Das <sandipan.das@amd.com> Link: https://lore.kernel.org/r/20230812084917.169338-17-yangjihong1@huawei.comSigned-off-by:
-
Yang Jihong authored
Use BPF to collect statistics on hardirq events based on perf BPF skeletons. Example usage: # perf kwork top -k sched,irq -b Starting trace, Hit <Ctrl+C> to stop and report ^C Total : 136717.945 ms, 8 cpus %Cpu(s): 17.10% id, 0.01% hi, 0.00% si %Cpu0 [||||||||||||||||||||||||| 84.26%] %Cpu1 [||||||||||||||||||||||||| 84.77%] %Cpu2 [|||||||||||||||||||||||| 83.22%] %Cpu3 [|||||||||||||||||||||||| 80.37%] %Cpu4 [|||||||||||||||||||||||| 81.49%] %Cpu5 [||||||||||||||||||||||||| 84.68%] %Cpu6 [||||||||||||||||||||||||| 84.48%] %Cpu7 [|||||||||||||||||||||||| 80.21%] PID SPID %CPU RUNTIME COMMMAND ------------------------------------------------------------- 0 0 19.78 3482.833 ms [swapper/7] 0 0 19.62 3454.219 ms [swapper/3] 0 0 18.50 3258.339 ms [swapper/4] 0 0 16.76 2842.749 ms [swapper/2] 0 0 15.71 2627.905 ms [swapper/0] 0 0 15.51 2598.206 ms [swapper/6] 0 0 15.31 2561.820 ms [swapper/5] 0 0 15.22 2548.708 ms [swapper/1] 13253 13018 2.95 513.108 ms sched-messaging 13092 13018 2.67 454.167 ms sched-messaging 13401 13018 2.66 454.790 ms sched-messaging 13240 13018 2.64 454.587 ms sched-messaging 13251 13018 2.61 442.273 ms sched-messaging 13075 13018 2.61 438.932 ms sched-messaging 13220 13018 2.60 443.245 ms sched-messaging 13235 13018 2.59 443.268 ms sched-messaging 13222 13018 2.50 426.344 ms sched-messaging 13410 13018 2.49 426.191 ms sched-messaging 13228 13018 2.46 425.121 ms sched-messaging 13379 13018 2.38 409.950 ms sched-messaging 13236 13018 2.37 413.159 ms sched-messaging 13095 13018 2.36 396.572 ms sched-messaging 13325 13018 2.35 408.089 ms sched-messaging 13242 13018 2.32 394.750 ms sched-messaging 13386 13018 2.31 396.997 ms sched-messaging 13046 13018 2.29 383.833 ms sched-messaging 13109 13018 2.28 388.482 ms sched-messaging 13388 13018 2.28 393.576 ms sched-messaging 13238 13018 2.26 388.487 ms sched-messaging <SNIP> Reviewed-by: -
Yang Jihong authored
Use BPF to collect statistics on the CPU usage based on perf BPF skeletons. Example usage: # perf kwork top -h Usage: perf kwork top [<options>] -b, --use-bpf Use BPF to measure task cpu usage -C, --cpu <cpu> list of cpus to profile -i, --input <file> input file name -n, --name <name> event name to profile -s, --sort <key[,key2...]> sort by key(s): rate, runtime, tid --time <str> Time span for analysis (start,stop) # # perf kwork -k sched top -b Starting trace, Hit <Ctrl+C> to stop and report ^C Total : 160702.425 ms, 8 cpus %Cpu(s): 36.00% id, 0.00% hi, 0.00% si %Cpu0 [|||||||||||||||||| 61.66%] %Cpu1 [|||||||||||||||||| 61.27%] %Cpu2 [||||||||||||||||||| 66.40%] %Cpu3 [|||||||||||||||||| 61.28%] %Cpu4 [|||||||||||||||||| 61.82%] %Cpu5 [||||||||||||||||||||||| 77.41%] %Cpu6 [|||||||||||||||||| 61.73%] %Cpu7 [|||||||||||||||||| 63.25%] PID SPID %CPU RUNTIME COMMMAND ------------------------------------------------------------- 0 0 38.72 8089.463 ms [swapper/1] 0 0 38.71 8084.547 ms [swapper/3] 0 0 38.33 8007.532 ms [swapper/0] 0 0 38.26 7992.985 ms [swapper/6] 0 0 38.17 7971.865 ms [swapper/4] 0 0 36.74 7447.765 ms [swapper/7] 0 0 33.59 6486.942 ms [swapper/2] 0 0 22.58 3771.268 ms [swapper/5] 9545 9351 2.48 447.136 ms sched-messaging 9574 9351 2.09 418.583 ms sched-messaging 9724 9351 2.05 372.407 ms sched-messaging 9531 9351 2.01 368.804 ms sched-messaging 9512 9351 2.00 362.250 ms sched-messaging 9514 9351 1.95 357.767 ms sched-messaging 9538 9351 1.86 384.476 ms sched-messaging 9712 9351 1.84 386.490 ms sched-messaging 9723 9351 1.83 380.021 ms sched-messaging 9722 9351 1.82 382.738 ms sched-messaging 9517 9351 1.81 354.794 ms sched-messaging 9559 9351 1.79 344.305 ms sched-messaging 9725 9351 1.77 365.315 ms sched-messaging <SNIP> # perf kwork -k sched top -b -n perf Starting trace, Hit <Ctrl+C> to stop and report ^C Total : 151563.332 ms, 8 cpus %Cpu(s): 26.49% id, 0.00% hi, 0.00% si %Cpu0 [ 0.01%] %Cpu1 [ 0.00%] %Cpu2 [ 0.00%] %Cpu3 [ 0.00%] %Cpu4 [ 0.00%] %Cpu5 [ 0.00%] %Cpu6 [ 0.00%] %Cpu7 [ 0.00%] PID SPID %CPU RUNTIME COMMMAND ------------------------------------------------------------- 9754 9754 0.01 2.303 ms perf # # perf kwork -k sched top -b -C 2,3,4 Starting trace, Hit <Ctrl+C> to stop and report ^C Total : 48016.721 ms, 3 cpus %Cpu(s): 27.82% id, 0.00% hi, 0.00% si %Cpu2 [|||||||||||||||||||||| 74.68%] %Cpu3 [||||||||||||||||||||| 71.06%] %Cpu4 [||||||||||||||||||||| 70.91%] PID SPID %CPU RUNTIME COMMMAND ------------------------------------------------------------- 0 0 29.08 4734.998 ms [swapper/4] 0 0 28.93 4710.029 ms [swapper/3] 0 0 25.31 3912.363 ms [swapper/2] 10248 10158 1.62 264.931 ms sched-messaging 10253 10158 1.62 265.136 ms sched-messaging 10158 10158 1.60 263.013 ms bash 10360 10158 1.49 243.639 ms sched-messaging 10413 10158 1.48 238.604 ms sched-messaging 10531 10158 1.47 234.067 ms sched-messaging 10400 10158 1.47 240.631 ms sched-messaging 10355 10158 1.47 230.586 ms sched-messaging 10377 10158 1.43 234.835 ms sched-messaging 10526 10158 1.42 232.045 ms sched-messaging 10298 10158 1.41 222.396 ms sched-messaging 10410 10158 1.38 221.853 ms sched-messaging 10364 10158 1.38 226.042 ms sched-messaging 10480 10158 1.36 213.633 ms sched-messaging 10370 10158 1.36 223.620 ms sched-messaging 10553 10158 1.34 217.169 ms sched-messaging 10291 10158 1.34 211.516 ms sched-messaging 10251 10158 1.34 218.813 ms sched-messaging 10522 10158 1.33 218.498 ms sched-messaging 10288 10158 1.33 216.787 ms sched-messaging <SNIP> Reviewed-by: -
Yang Jihong authored
Provide the following options for perf kwork top: 1. -C, --cpu <cpu> list of cpus to profile 2. -i, --input <file> input file name 3. -n, --name <name> event name to profile 4. -s, --sort <key[,key2...]> sort by key(s): rate, runtime, tid 5. --time <str> Time span for analysis (start,stop) Example usage: # perf kwork top -h Usage: perf kwork top [<options>] -C, --cpu <cpu> list of cpus to profile -i, --input <file> input file name -n, --name <name> event name to profile -s, --sort <key[,key2...]> sort by key(s): rate, runtime, tid --time <str> Time span for analysis (start,stop) # perf kwork top -C 2,4,5 Total : 51226.940 ms, 3 cpus %Cpu(s): 92.59% id, 0.00% hi, 0.09% si %Cpu2 [| 4.61%] %Cpu4 [ 0.01%] %Cpu5 [||||| 17.31%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 0 99.98 17073.515 ms swapper/4 0 95.17 16250.874 ms swapper/2 0 82.62 14108.577 ms swapper/5 4342 21.70 3708.358 ms perf 16 0.13 22.296 ms rcu_preempt 75 0.02 4.261 ms kworker/2:1 98 0.01 2.540 ms jbd2/sda-8 61 0.01 3.404 ms kcompactd0 87 0.00 0.145 ms kworker/5:1H 73 0.00 0.596 ms kworker/5:1 41 0.00 0.041 ms ksoftirqd/5 40 0.00 0.718 ms migration/5 64 0.00 0.115 ms kworker/4:1 35 0.00 0.556 ms migration/4 353 0.00 1.143 ms sshd 26 0.00 1.665 ms ksoftirqd/2 25 0.00 0.662 ms migration/2 # perf kwork top -i perf.data Total : 136601.588 ms, 8 cpus %Cpu(s): 95.66% id, 0.04% hi, 0.05% si %Cpu0 [ 0.02%] %Cpu1 [ 0.01%] %Cpu2 [| 4.61%] %Cpu3 [ 0.04%] %Cpu4 [ 0.01%] %Cpu5 [||||| 17.31%] %Cpu6 [ 0.51%] %Cpu7 [||| 11.42%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 0 99.98 17073.515 ms swapper/4 0 99.98 17072.173 ms swapper/1 0 99.93 17064.229 ms swapper/3 0 99.62 17011.013 ms swapper/0 0 99.47 16985.180 ms swapper/6 0 95.17 16250.874 ms swapper/2 0 88.51 15111.684 ms swapper/7 0 82.62 14108.577 ms swapper/5 4342 33.00 5644.045 ms perf 4344 0.43 74.351 ms perf 16 0.13 22.296 ms rcu_preempt 4345 0.05 10.093 ms perf 4343 0.05 8.769 ms perf 4341 0.02 4.882 ms perf 4095 0.02 4.605 ms kworker/7:1 75 0.02 4.261 ms kworker/2:1 120 0.01 1.909 ms systemd-journal 98 0.01 2.540 ms jbd2/sda-8 61 0.01 3.404 ms kcompactd0 667 0.01 2.542 ms kworker/u16:2 4340 0.00 1.052 ms kworker/7:2 97 0.00 0.489 ms kworker/7:1H 51 0.00 0.209 ms ksoftirqd/7 50 0.00 0.646 ms migration/7 76 0.00 0.753 ms kworker/6:1 45 0.00 0.572 ms migration/6 87 0.00 0.145 ms kworker/5:1H 73 0.00 0.596 ms kworker/5:1 41 0.00 0.041 ms ksoftirqd/5 40 0.00 0.718 ms migration/5 64 0.00 0.115 ms kworker/4:1 35 0.00 0.556 ms migration/4 353 0.00 2.600 ms sshd 74 0.00 0.205 ms kworker/3:1 33 0.00 1.576 ms kworker/3:0H 30 0.00 0.996 ms migration/3 26 0.00 1.665 ms ksoftirqd/2 25 0.00 0.662 ms migration/2 397 0.00 0.057 ms kworker/1:1 20 0.00 1.005 ms migration/1 2909 0.00 1.053 ms kworker/0:2 17 0.00 0.720 ms migration/0 15 0.00 0.039 ms ksoftirqd/0 # perf kwork top -n perf Total : 136601.588 ms, 8 cpus %Cpu(s): 95.66% id, 0.04% hi, 0.05% si %Cpu0 [ 0.01%] %Cpu1 [ 0.00%] %Cpu2 [| 4.44%] %Cpu3 [ 0.00%] %Cpu4 [ 0.00%] %Cpu5 [ 0.00%] %Cpu6 [ 0.49%] %Cpu7 [||| 11.38%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 4342 15.74 2695.516 ms perf 4344 0.43 74.351 ms perf 4345 0.05 10.093 ms perf 4343 0.05 8.769 ms perf 4341 0.02 4.882 ms perf # perf kwork top -s tid Total : 136601.588 ms, 8 cpus %Cpu(s): 95.66% id, 0.04% hi, 0.05% si %Cpu0 [ 0.02%] %Cpu1 [ 0.01%] %Cpu2 [| 4.61%] %Cpu3 [ 0.04%] %Cpu4 [ 0.01%] %Cpu5 [||||| 17.31%] %Cpu6 [ 0.51%] %Cpu7 [||| 11.42%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 0 99.62 17011.013 ms swapper/0 0 99.98 17072.173 ms swapper/1 0 95.17 16250.874 ms swapper/2 0 99.93 17064.229 ms swapper/3 0 99.98 17073.515 ms swapper/4 0 82.62 14108.577 ms swapper/5 0 99.47 16985.180 ms swapper/6 0 88.51 15111.684 ms swapper/7 15 0.00 0.039 ms ksoftirqd/0 16 0.13 22.296 ms rcu_preempt 17 0.00 0.720 ms migration/0 20 0.00 1.005 ms migration/1 25 0.00 0.662 ms migration/2 26 0.00 1.665 ms ksoftirqd/2 30 0.00 0.996 ms migration/3 33 0.00 1.576 ms kworker/3:0H 35 0.00 0.556 ms migration/4 40 0.00 0.718 ms migration/5 41 0.00 0.041 ms ksoftirqd/5 45 0.00 0.572 ms migration/6 50 0.00 0.646 ms migration/7 51 0.00 0.209 ms ksoftirqd/7 61 0.01 3.404 ms kcompactd0 64 0.00 0.115 ms kworker/4:1 73 0.00 0.596 ms kworker/5:1 74 0.00 0.205 ms kworker/3:1 75 0.02 4.261 ms kworker/2:1 76 0.00 0.753 ms kworker/6:1 87 0.00 0.145 ms kworker/5:1H 97 0.00 0.489 ms kworker/7:1H 98 0.01 2.540 ms jbd2/sda-8 120 0.01 1.909 ms systemd-journal 353 0.00 2.600 ms sshd 397 0.00 0.057 ms kworker/1:1 667 0.01 2.542 ms kworker/u16:2 2909 0.00 1.053 ms kworker/0:2 4095 0.02 4.605 ms kworker/7:1 4340 0.00 1.052 ms kworker/7:2 4341 0.02 4.882 ms perf 4342 33.00 5644.045 ms perf 4343 0.05 8.769 ms perf 4344 0.43 74.351 ms perf 4345 0.05 10.093 ms perf # perf kwork top --time 128800, Total : 53495.122 ms, 8 cpus %Cpu(s): 94.71% id, 0.09% hi, 0.09% si %Cpu0 [ 0.07%] %Cpu1 [ 0.04%] %Cpu2 [|| 8.49%] %Cpu3 [ 0.09%] %Cpu4 [ 0.02%] %Cpu5 [ 0.06%] %Cpu6 [ 0.12%] %Cpu7 [|||||| 21.24%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 0 99.96 3981.363 ms swapper/4 0 99.94 3978.955 ms swapper/1 0 99.91 9329.375 ms swapper/5 0 99.87 4906.829 ms swapper/3 0 99.86 9028.064 ms swapper/6 0 98.67 3928.161 ms swapper/0 0 91.17 8388.432 ms swapper/2 0 78.65 7125.602 ms swapper/7 4342 29.42 2675.198 ms perf 16 0.18 16.817 ms rcu_preempt 4345 0.09 8.183 ms perf 4344 0.04 4.290 ms perf 4343 0.03 2.844 ms perf 353 0.03 2.600 ms sshd 4095 0.02 2.702 ms kworker/7:1 120 0.02 1.909 ms systemd-journal 98 0.02 2.540 ms jbd2/sda-8 61 0.02 1.886 ms kcompactd0 667 0.02 1.011 ms kworker/u16:2 75 0.02 2.693 ms kworker/2:1 4341 0.01 1.838 ms perf 30 0.01 0.788 ms migration/3 26 0.01 1.665 ms ksoftirqd/2 20 0.01 0.752 ms migration/1 2909 0.01 0.604 ms kworker/0:2 4340 0.00 0.635 ms kworker/7:2 97 0.00 0.214 ms kworker/7:1H 51 0.00 0.209 ms ksoftirqd/7 50 0.00 0.646 ms migration/7 76 0.00 0.602 ms kworker/6:1 45 0.00 0.366 ms migration/6 87 0.00 0.145 ms kworker/5:1H 40 0.00 0.446 ms migration/5 35 0.00 0.318 ms migration/4 74 0.00 0.205 ms kworker/3:1 33 0.00 0.080 ms kworker/3:0H 25 0.00 0.448 ms migration/2 397 0.00 0.057 ms kworker/1:1 17 0.00 0.365 ms migration/0 Reviewed-by: -
Yang Jihong authored
Calculate the runtime of the softirq events and subtract it from the corresponding task runtime to improve the precision. Example usage: # perf kwork -k sched,irq,softirq record -- perf record -e cpu-clock -o perf_record.data -a sleep 10 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.467 MB perf_record.data (7154 samples) ] [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 2.152 MB perf.data (22846 samples) ] # perf kwork top Total : 136601.588 ms, 8 cpus %Cpu(s): 95.66% id, 0.04% hi, 0.05% si %Cpu0 [ 0.02%] %Cpu1 [ 0.01%] %Cpu2 [| 4.61%] %Cpu3 [ 0.04%] %Cpu4 [ 0.01%] %Cpu5 [||||| 17.31%] %Cpu6 [ 0.51%] %Cpu7 [||| 11.42%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 0 99.98 17073.515 ms swapper/4 0 99.98 17072.173 ms swapper/1 0 99.93 17064.229 ms swapper/3 0 99.62 17011.013 ms swapper/0 0 99.47 16985.180 ms swapper/6 0 95.17 16250.874 ms swapper/2 0 88.51 15111.684 ms swapper/7 0 82.62 14108.577 ms swapper/5 4342 33.00 5644.045 ms perf 4344 0.43 74.351 ms perf 16 0.13 22.296 ms rcu_preempt 4345 0.05 10.093 ms perf 4343 0.05 8.769 ms perf 4341 0.02 4.882 ms perf 4095 0.02 4.605 ms kworker/7:1 75 0.02 4.261 ms kworker/2:1 120 0.01 1.909 ms systemd-journal 98 0.01 2.540 ms jbd2/sda-8 61 0.01 3.404 ms kcompactd0 667 0.01 2.542 ms kworker/u16:2 4340 0.00 1.052 ms kworker/7:2 97 0.00 0.489 ms kworker/7:1H 51 0.00 0.209 ms ksoftirqd/7 50 0.00 0.646 ms migration/7 76 0.00 0.753 ms kworker/6:1 45 0.00 0.572 ms migration/6 87 0.00 0.145 ms kworker/5:1H 73 0.00 0.596 ms kworker/5:1 41 0.00 0.041 ms ksoftirqd/5 40 0.00 0.718 ms migration/5 64 0.00 0.115 ms kworker/4:1 35 0.00 0.556 ms migration/4 353 0.00 2.600 ms sshd 74 0.00 0.205 ms kworker/3:1 33 0.00 1.576 ms kworker/3:0H 30 0.00 0.996 ms migration/3 26 0.00 1.665 ms ksoftirqd/2 25 0.00 0.662 ms migration/2 397 0.00 0.057 ms kworker/1:1 20 0.00 1.005 ms migration/1 2909 0.00 1.053 ms kworker/0:2 17 0.00 0.720 ms migration/0 15 0.00 0.039 ms ksoftirqd/0 Reviewed-by: -
Yang Jihong authored
Calculate the runtime of the hardirq events and subtract it from the corresponding task runtime to improve the precision. Example usage: # perf kwork -k sched,irq record -- perf record -o perf_record.data -a sleep 10 [ perf record: Woken up 2 times to write data ] [ perf record: Captured and wrote 1.054 MB perf_record.data (18019 samples) ] [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 1.798 MB perf.data (16334 samples) ] # # perf kwork top Total : 139240.869 ms, 8 cpus %Cpu(s): 94.91% id, 0.05% hi %Cpu0 [ 0.05%] %Cpu1 [| 5.00%] %Cpu2 [ 0.43%] %Cpu3 [ 0.57%] %Cpu4 [ 1.19%] %Cpu5 [|||||| 20.46%] %Cpu6 [ 0.48%] %Cpu7 [||| 12.10%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 0 99.54 17325.622 ms swapper/2 0 99.54 17327.527 ms swapper/0 0 99.51 17319.909 ms swapper/6 0 99.42 17304.934 ms swapper/3 0 98.80 17197.385 ms swapper/4 0 94.99 16534.991 ms swapper/1 0 87.89 15295.264 ms swapper/7 0 79.53 13843.182 ms swapper/5 4252 36.50 6361.768 ms perf 4256 1.17 205.215 ms bash 151 0.53 93.298 ms systemd-resolve 4254 0.39 69.468 ms perf 423 0.34 59.368 ms bash 412 0.29 51.204 ms sshd 249 0.20 35.288 ms sd-resolve 16 0.17 30.287 ms rcu_preempt 153 0.09 17.266 ms systemd-timesyn 1 0.09 17.078 ms systemd 4253 0.07 12.457 ms perf 4255 0.06 11.559 ms perf 4234 0.03 6.105 ms kworker/u16:1 69 0.03 6.259 ms kworker/1:1H 4251 0.02 4.615 ms perf 4095 0.02 4.890 ms kworker/7:1 61 0.02 4.005 ms kcompactd0 75 0.02 3.546 ms kworker/2:1 97 0.01 3.106 ms kworker/7:1H 98 0.01 1.995 ms jbd2/sda-8 4088 0.01 1.779 ms kworker/u16:3 2909 0.01 1.795 ms kworker/0:2 4246 0.00 1.117 ms kworker/7:2 51 0.00 0.327 ms ksoftirqd/7 50 0.00 0.369 ms migration/7 102 0.00 0.160 ms kworker/6:1H 76 0.00 0.609 ms kworker/6:1 45 0.00 0.779 ms migration/6 87 0.00 0.504 ms kworker/5:1H 73 0.00 1.130 ms kworker/5:1 41 0.00 0.152 ms ksoftirqd/5 40 0.00 0.702 ms migration/5 64 0.00 0.316 ms kworker/4:1 35 0.00 0.791 ms migration/4 353 0.00 2.211 ms sshd 74 0.00 0.272 ms kworker/3:1 30 0.00 0.819 ms migration/3 25 0.00 0.784 ms migration/2 397 0.00 0.539 ms kworker/1:1 21 0.00 1.600 ms ksoftirqd/1 20 0.00 0.773 ms migration/1 17 0.00 1.682 ms migration/0 15 0.00 0.076 ms ksoftirqd/0 Reviewed-by: -
Yang Jihong authored
Add evsel__intval_common() helper to search for common_field in tracepoint format. Reviewed-by:
-
Yang Jihong authored
Some common tools for collecting statistics on CPU usage, such as top, obtain statistics from timer interrupt sampling, and then periodically read statistics from /proc/stat. This method has some deviations: 1. In the tick interrupt, the time between the last tick and the current tick is counted in the current task. However, the task may be running only part of the time. 2. For each task, the top tool periodically reads the /proc/{PID}/status information. For tasks with a short life cycle, it may be missed. In conclusion, the top tool cannot accurately collect statistics on the CPU usage and running time of tasks. The statistical method based on sched_switch tracepoint can accurately calculate the CPU usage of all tasks. This method is applicable to scenarios where performance comparison data is of high precision. Example usage: # perf kwork Usage: perf kwork [<options>] {record|report|latency|timehist|top} -D, --dump-raw-trace dump raw trace in ASCII -f, --force don't complain, do it -k, --kwork <kwork> list of kwork to profile (irq, softirq, workqueue, sched, etc) -v, --verbose be more verbose (show symbol address, etc) # perf kwork -k sched record -- perf bench sched messaging -g 1 -l 10000 # Running 'sched/messaging' benchmark: # 20 sender and receiver processes per group # 1 groups == 40 processes run Total time: 14.074 [sec] [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 15.886 MB perf.data (129472 samples) ] # perf kwork top Total : 115708.178 ms, 8 cpus %Cpu(s): 9.78% id %Cpu0 [||||||||||||||||||||||||||| 90.55%] %Cpu1 [||||||||||||||||||||||||||| 90.51%] %Cpu2 [|||||||||||||||||||||||||| 88.57%] %Cpu3 [||||||||||||||||||||||||||| 91.18%] %Cpu4 [||||||||||||||||||||||||||| 91.09%] %Cpu5 [||||||||||||||||||||||||||| 90.88%] %Cpu6 [|||||||||||||||||||||||||| 88.64%] %Cpu7 [||||||||||||||||||||||||||| 90.28%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 4113 22.23 3221.547 ms sched-messaging 4105 21.61 3131.495 ms sched-messaging 4119 21.53 3120.937 ms sched-messaging 4103 21.39 3101.614 ms sched-messaging 4106 21.37 3095.209 ms sched-messaging 4104 21.25 3077.269 ms sched-messaging 4115 21.21 3073.188 ms sched-messaging 4109 21.18 3069.022 ms sched-messaging 4111 20.78 3010.033 ms sched-messaging 4114 20.74 3007.073 ms sched-messaging 4108 20.73 3002.137 ms sched-messaging 4107 20.47 2967.292 ms sched-messaging 4117 20.39 2955.335 ms sched-messaging 4112 20.34 2947.080 ms sched-messaging 4118 20.32 2942.519 ms sched-messaging 4121 20.23 2929.865 ms sched-messaging 4110 20.22 2930.078 ms sched-messaging 4122 20.15 2919.542 ms sched-messaging 4120 19.77 2866.032 ms sched-messaging 4116 19.72 2857.660 ms sched-messaging 4127 16.19 2346.334 ms sched-messaging 4142 15.86 2297.600 ms sched-messaging 4141 15.62 2262.646 ms sched-messaging 4136 15.41 2231.408 ms sched-messaging 4130 15.38 2227.008 ms sched-messaging 4129 15.31 2217.692 ms sched-messaging 4126 15.21 2201.711 ms sched-messaging 4139 15.19 2200.722 ms sched-messaging 4137 15.10 2188.633 ms sched-messaging 4134 15.06 2182.082 ms sched-messaging 4132 15.02 2177.530 ms sched-messaging 4131 14.73 2131.973 ms sched-messaging 4125 14.68 2125.439 ms sched-messaging 4128 14.66 2122.255 ms sched-messaging 4123 14.65 2122.113 ms sched-messaging 4135 14.56 2107.144 ms sched-messaging 4133 14.51 2103.549 ms sched-messaging 4124 14.27 2066.671 ms sched-messaging 4140 14.17 2052.251 ms sched-messaging 4138 13.81 2000.361 ms sched-messaging 0 11.42 1652.009 ms swapper/2 0 11.35 1641.694 ms swapper/6 0 9.71 1405.108 ms swapper/7 0 9.48 1372.338 ms swapper/1 0 9.44 1366.013 ms swapper/0 0 9.11 1318.382 ms swapper/5 0 8.90 1287.582 ms swapper/4 0 8.81 1274.356 ms swapper/3 4100 2.61 379.328 ms perf 4101 1.16 169.487 ms perf-exec 151 0.65 94.741 ms systemd-resolve 249 0.36 53.030 ms sd-resolve 153 0.14 21.405 ms systemd-timesyn 1 0.10 16.200 ms systemd 16 0.09 15.785 ms rcu_preempt 4102 0.06 9.727 ms perf 4095 0.03 5.464 ms kworker/7:1 98 0.02 3.231 ms jbd2/sda-8 353 0.02 4.115 ms sshd 75 0.02 3.889 ms kworker/2:1 73 0.01 1.552 ms kworker/5:1 64 0.01 1.591 ms kworker/4:1 74 0.01 1.952 ms kworker/3:1 61 0.01 2.608 ms kcompactd0 397 0.01 1.602 ms kworker/1:1 69 0.01 1.817 ms kworker/1:1H 10 0.01 2.553 ms kworker/u16:0 2909 0.01 2.684 ms kworker/0:2 1211 0.00 0.426 ms kworker/7:0 97 0.00 0.153 ms kworker/7:1H 51 0.00 0.100 ms ksoftirqd/7 120 0.00 0.856 ms systemd-journal 76 0.00 1.414 ms kworker/6:1 46 0.00 0.246 ms ksoftirqd/6 45 0.00 0.164 ms migration/6 41 0.00 0.098 ms ksoftirqd/5 40 0.00 0.207 ms migration/5 86 0.00 1.339 ms kworker/4:1H 36 0.00 0.252 ms ksoftirqd/4 35 0.00 0.090 ms migration/4 31 0.00 0.156 ms ksoftirqd/3 30 0.00 0.073 ms migration/3 26 0.00 0.180 ms ksoftirqd/2 25 0.00 0.085 ms migration/2 21 0.00 0.106 ms ksoftirqd/1 20 0.00 0.118 ms migration/1 302 0.00 1.440 ms systemd-logind 17 0.00 0.132 ms migration/0 15 0.00 0.255 ms ksoftirqd/0 Reviewed-by: -
Yang Jihong authored
Add a `struct rb_root_cached *root` parameter to work_sort() to sort the specified rb tree elements. No functional change. Reviewed-by:
-
Yang Jihong authored
The kwork_class type of sched is added to support recording and parsing of sched_switch events. As follows: # perf kwork -h Usage: perf kwork [<options>] {record|report|latency|timehist} -D, --dump-raw-trace dump raw trace in ASCII -f, --force don't complain, do it -k, --kwork <kwork> list of kwork to profile (irq, softirq, workqueue, sched, etc) -v, --verbose be more verbose (show symbol address, etc) # perf kwork -k sched record true [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.083 MB perf.data (47 samples) ] # perf evlist sched:sched_switch dummy:HG # Tip: use 'perf evlist --trace-fields' to show fields for tracepoint events Reviewed-by: -
Yang Jihong authored
Currently when no kwork event is specified, all events are configured by default. Now set to default event list string, which is more flexible and supports subsequent function extension. Also put setup_event_list() into each subcommand for different settings. Reviewed-by:
-
Yang Jihong authored
work_push_atom() supports nesting. Currently, all supported kworks are not nested. A `overwrite` parameter is added to overwrite the original atom in the list. Reviewed-by:
-
Yang Jihong authored
To support different types of reports, two parameters `struct perf_kwork * kwork` and `enum kwork_trace_type src_type` are added to work_init() of struct kwork_class for initialization in different scenarios. No functional change intended. Reviewed-by:
-
Yang Jihong authored
'perf kwork' processes data based on timestamps and needs to sort events. Fixes: f98919ec ("perf kwork: Implement 'report' subcommand") Reviewed-by:
-
Yang Jihong authored
Add missing report, latency and timehist subcommands to the document. Fixes: f98919ec ("perf kwork: Implement 'report' subcommand") Fixes: ad3d9f7a ("perf kwork: Implement perf kwork latency") Fixes: bcc8b3e8 ("perf kwork: Implement perf kwork timehist") Reviewed-by:
-
Yang Jihong authored
1. Atoms are managed in page mode and should be released using atom_free() instead of free(). 2. When the event does not match, the atom needs to free. Fixes: f98919ec ("perf kwork: Implement 'report' subcommand") Reviewed-by:
-
Yang Jihong authored
If only dummy event is recorded, tracking event is not needed. Add this test scenario. Test result: # ./perf test list 2>&1 | grep 'Setup struct perf_event_attr' 17: Setup struct perf_event_attr # ./perf test 17 -v 17: Setup struct perf_event_attr : --- start --- test child forked, pid 720198 <SNIP> running './tests/attr/test-record-dummy-C0' <SNIP> test child finished with 0 ---- end ---- Setup struct perf_event_attr: Ok Signed-off-by:
Ravi Bangoria <ravi.bangoria@amd.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Anshuman Khandual <anshuman.khandual@arm.com> Cc: Ian Rogers <irogers@google.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: James Clark <james.clark@arm.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Richter <tmricht@linux.ibm.com> Link: https://lore.kernel.org/r/20230904023340.12707-7-yangjihong1@huawei.comSigned-off-by:
-
Yang Jihong authored
Add a new test case to record sideband events for all CPUs when tracing selected CPUs Test result: # ./perf test list 2>&1 | grep 'perf record sideband tests' 95: perf record sideband tests # ./perf test 95 95: perf record sideband tests : Ok Signed-off-by:
-
Yang Jihong authored
User space tasks can migrate between CPUs, we need to track side-band events for all CPUs. The specific scenarios are as follows: CPU0 CPU1 perf record -C 0 start taskA starts to be created and executed -> PERF_RECORD_COMM and PERF_RECORD_MMAP events only deliver to CPU1 ...... | migrate to CPU0 | Running on CPU0 <----------/ ... perf record -C 0 stop Now perf samples the PC of taskA. However, perf does not record the PERF_RECORD_COMM and PERF_RECORD_MMAP events of taskA. Therefore, the comm and symbols of taskA cannot be parsed. The solution is to record sideband events for all CPUs when tracing selected CPUs. Because this modifies the default behavior, add related comments to the perf record man page. The sys_perf_event_open invoked is as follows: # perf --debug verbose=3 record -e cpu-clock -C 1 true <SNIP> Opening: cpu-clock ------------------------------------------------------------ perf_event_attr: type 1 (PERF_TYPE_SOFTWARE) size 136 config 0 (PERF_COUNT_SW_CPU_CLOCK) { sample_period, sample_freq } 4000 sample_type IP|TID|TIME|CPU|PERIOD|IDENTIFIER read_format ID|LOST disabled 1 inherit 1 freq 1 sample_id_all 1 exclude_guest 1 ------------------------------------------------------------ sys_perf_event_open: pid -1 cpu 1 group_fd -1 flags 0x8 = 5 Opening: dummy:u ------------------------------------------------------------ perf_event_attr: type 1 (PERF_TYPE_SOFTWARE) size 136 config 0x9 (PERF_COUNT_SW_DUMMY) { sample_period, sample_freq } 1 sample_type IP|TID|TIME|CPU|IDENTIFIER read_format ID|LOST inherit 1 exclude_kernel 1 exclude_hv 1 mmap 1 comm 1 task 1 sample_id_all 1 exclude_guest 1 mmap2 1 comm_exec 1 ksymbol 1 bpf_event 1 ------------------------------------------------------------ sys_perf_event_open: pid -1 cpu 0 group_fd -1 flags 0x8 = 6 sys_perf_event_open: pid -1 cpu 1 group_fd -1 flags 0x8 = 7 sys_perf_event_open: pid -1 cpu 2 group_fd -1 flags 0x8 = 9 sys_perf_event_open: pid -1 cpu 3 group_fd -1 flags 0x8 = 10 sys_perf_event_open: pid -1 cpu 4 group_fd -1 flags 0x8 = 11 sys_perf_event_open: pid -1 cpu 5 group_fd -1 flags 0x8 = 12 sys_perf_event_open: pid -1 cpu 6 group_fd -1 flags 0x8 = 13 sys_perf_event_open: pid -1 cpu 7 group_fd -1 flags 0x8 = 14 <SNIP> Signed-off-by: -
Yang Jihong authored
User space tasks can migrate between CPUs, so when tracing selected CPUs, sideband for all CPUs is needed. In this case set the cpu map of the evsel to all online CPUs. This may modify the original cpu map of the evlist. Therefore, need to check whether the preceding scenario exists before record__init_thread_masks(). Dummy tracking has been set in record__open(), move it before record__init_thread_masks() and add a helper for unified processing. The sys_perf_event_open invoked is as follows: # perf --debug verbose=3 record -e cpu-clock -D 100 true <SNIP> Opening: cpu-clock ------------------------------------------------------------ perf_event_attr: type 1 (PERF_TYPE_SOFTWARE) size 136 config 0 (PERF_COUNT_SW_CPU_CLOCK) { sample_period, sample_freq } 4000 sample_type IP|TID|TIME|PERIOD|IDENTIFIER read_format ID|LOST disabled 1 inherit 1 freq 1 sample_id_all 1 exclude_guest 1 ------------------------------------------------------------ sys_perf_event_open: pid 10318 cpu 0 group_fd -1 flags 0x8 = 5 sys_perf_event_open: pid 10318 cpu 1 group_fd -1 flags 0x8 = 6 sys_perf_event_open: pid 10318 cpu 2 group_fd -1 flags 0x8 = 7 sys_perf_event_open: pid 10318 cpu 3 group_fd -1 flags 0x8 = 9 sys_perf_event_open: pid 10318 cpu 4 group_fd -1 flags 0x8 = 10 sys_perf_event_open: pid 10318 cpu 5 group_fd -1 flags 0x8 = 11 sys_perf_event_open: pid 10318 cpu 6 group_fd -1 flags 0x8 = 12 sys_perf_event_open: pid 10318 cpu 7 group_fd -1 flags 0x8 = 13 Opening: dummy:u ------------------------------------------------------------ perf_event_attr: type 1 (PERF_TYPE_SOFTWARE) size 136 config 0x9 (PERF_COUNT_SW_DUMMY) { sample_period, sample_freq } 1 sample_type IP|TID|TIME|IDENTIFIER read_format ID|LOST disabled 1 inherit 1 exclude_kernel 1 exclude_hv 1 mmap 1 comm 1 enable_on_exec 1 task 1 sample_id_all 1 exclude_guest 1 mmap2 1 comm_exec 1 ksymbol 1 bpf_event 1 ------------------------------------------------------------ sys_perf_event_open: pid 10318 cpu 0 group_fd -1 flags 0x8 = 14 sys_perf_event_open: pid 10318 cpu 1 group_fd -1 flags 0x8 = 15 sys_perf_event_open: pid 10318 cpu 2 group_fd -1 flags 0x8 = 16 sys_perf_event_open: pid 10318 cpu 3 group_fd -1 flags 0x8 = 17 sys_perf_event_open: pid 10318 cpu 4 group_fd -1 flags 0x8 = 18 sys_perf_event_open: pid 10318 cpu 5 group_fd -1 flags 0x8 = 19 sys_perf_event_open: pid 10318 cpu 6 group_fd -1 flags 0x8 = 20 sys_perf_event_open: pid 10318 cpu 7 group_fd -1 flags 0x8 = 21 <SNIP> 'perf test' needs to update base-record & system-wide-dummy attr expected values for test-record-C0: 1. Because a dummy sideband event is added to the sampling of specified CPUs. When evlist contains evsel of different sample_type, evlist__config() will change the default PERF_SAMPLE_ID bit to PERF_SAMPLE_IDENTIFICATION bit. The attr sample_type expected value of base-record and system-wide-dummy in test-record-C0 needs to be updated. 2. The perf record uses evlist__add_aux_dummy() instead of evlist__add_dummy() to add a dummy event. The expected value of system-wide-dummy attr needs to be updated. The 'perf test' result is as follows: # ./perf test list 2>&1 | grep 'Setup struct perf_event_attr' 17: Setup struct perf_event_attr # ./perf test 17 17: Setup struct perf_event_attr : Ok Signed-off-by:Adrian Hunter <adrian.hunter@intel.com> Tested-by:
-
Yang Jihong authored
Currently, intel-bts, intel-pt, and arm-spe may add tracking event to the evlist. We may need to search for the tracking event for some settings. Therefore, add evlist__findnew_tracking_event() helper. If system_wide is true, evlist__findnew_tracking_event() set the cpu map of the evsel to all online CPUs. Signed-off-by:
-
Yang Jihong authored
For dummy events that keep tracking, we may need to modify its cpu_maps. For example, change the cpu_maps to record sideband events for all CPUS. Add perf_evlist__go_system_wide() helper to support this scenario. Signed-off-by:
-
- 11 Sep, 2023 15 commits
-
-
Ian Rogers authored
Update perf JSON files with spelling fixes by Colin Ian King <colin.i.king@gmail.com> contributed in: https://github.com/intel/perfmon/pull/96 "Fix various spelling mistakes and typos as found using codespell #96" This is added on top of the spelling mistakes and release number updates in: https://github.com/intel/perfmon/pull/98 "EMR, SPR, CLX, SKX, BDX, HSX, BDW-DE, WSM-EP*, NHM-*, JKT, IVT : Release event updates" Some additional spelling fixes reported by Edward Baker <edward.baker@intel.com> are added on top of this. Signed-off-by:
-
Ian Rogers authored
Add emeraldrapids events that were added at intel's perfmon site in: https://github.com/intel/perfmon/pull/98 "EMR, SPR, CLX, SKX, BDX, HSX, BDW-DE, WSM-EP*, NHM-*, JKT, IVT : Release event updates" "Emerald Rapids (0xCF) was previously pointing to SPR core. In this pull request dedicated EMR files are introduced." Signed-off-by:
-
Ian Rogers authored
Add lunarlake events that were added at intel's perfmon site in: https://github.com/intel/perfmon/pull/97 "LNL: Release initial events" Reviewed-by:
Kan Liang <kan.liang@linux.intel.com> Signed-off-by:
-
Ian Rogers authored
parse_events_terms() existed in function names but was passed a 'struct list_head'. As many parse_events functions take an evsel_config list as well as a parse_event_term list, and the naming head_terms and head_config is inconsistent, there's a potential to switch the lists and get errors. Introduce a 'struct parse_events_terms', that just wraps a list_head, to avoid this. Add the regular init/exit functions and transition the code to use them. Reviewed-by:
James Clark <james.clark@arm.com> Signed-off-by:
-
Ian Rogers authored
When trying to add events to multiple PMUs the term list is copied first as adding the event will rewrite the event's name term into the sysfs and/or json encoding terms (see perf_pmu__check_alias). Change the parse events add API so the passed in term list is const, then copy the list when modification is necessary. Reviewed-by:
-
Ian Rogers authored
Add term_type to union of values returned by the lexer to avoid casts to and from an integer. Reviewed-by:
-
Ian Rogers authored
Add a const and rename str to event_name. Reviewed-by:
-
Ian Rogers authored
The parameter head_terms is always used in get_config_terms. Reviewed-by:
-
Ian Rogers authored
Allow metrics to expand for -M or --metrics options. Committer testing: # grep -m1 'model name' /proc/cpuinfo model name : AMD Ryzen 9 5950X 16-Core Processor # Before: Just expansion of files/directories in the pwd are expanded: # . tools/perf/perf-completion.sh # perf stat -M b block/ build/ # perf stat -M b After: # . tools/perf/perf-completion.sh # perf stat -M all_l2_cache_accesses all_remote_links_outbound data_fabric l1_itlb_misses l2_cache_misses_from_l2_hwpf macro_ops_dispatched tlb all_l2_cache_hits branch_misprediction_ratio decoder l2_cache l3_cache nps1_die_to_dram all_l2_cache_misses branch_prediction ic_fetch_miss_ratio l2_cache_accesses_from_l2_hwpf l3_read_miss_latency op_cache_fetch_miss_ratio # perf stat -M branch_ branch_misprediction_ratio branch_prediction # perf stat -M branch_prediction -a sleep 1 Performance counter stats for 'system wide': 115,079,765 ex_ret_brn # 4.0 % branch_misprediction_ratio 4,561,456 ex_ret_brn_misp 1.015925106 seconds time elapsed # Suggested-by: -
Ian Rogers authored
Use `perf list --raw-dump pfm` to support completion of libpfm4 events. Committer testing: # grep -m1 'model name' /proc/cpuinfo model name : AMD Ryzen 9 5950X 16-Core Processor Before: Files in the current directory are expanded when <tab> After: Only the PFM events are: # . tools/perf/perf-completion.sh # perf stat --pfm-events <tab> Becomes: # perf stat --pfm-events perf_raw::r0000 As apparently there are no other PFM events for this Ryzen 9 5950X machine. Signed-off-by:
-
Ian Rogers authored
'perf list' will list libpfm4 events and metrics which aren't valid options to the '-e' option. Restrict the events gathered so that invalid ones aren't shown. Before: $ perf stat -e <tab><tab> Display all 633 possibilities? (y or n) After: $ perf stat -e <tab><tab> Display all 375 possibilities? (y or n) Committer testing: # grep -m1 'model name' /proc/cpuinfo model name : AMD Ryzen 9 5950X 16-Core Processor # Before: # . tools/perf/perf-completion.sh # perf stat -e Display all 2672 possibilities? (y or n) After: # . tools/perf/perf-completion.sh # perf stat -e Display all 2648 possibilities? (y or n) Reported-by:
-
Ian Rogers authored
Generating metrics llc_code_read_mpi_demand_plus_prefetch, llc_data_read_mpi_demand_plus_prefetch, llc_miss_local_memory_bandwidth_read, llc_miss_local_memory_bandwidth_write, nllc_miss_remote_memory_bandwidth_read, memory_bandwidth_read, memory_bandwidth_write, uncore_frequency, upi_data_transmit_bw, C2_Pkg_Residency, C3_Core_Residency, C3_Pkg_Residency, C6_Core_Residency, C6_Pkg_Residency, C7_Core_Residency, C7_Pkg_Residency, UNCORE_FREQ and tma_info_system_socket_clks would trigger an address sanitizer heap-buffer-overflows on a SkylakeX. ``` ==2567752==ERROR: AddressSanitizer: heap-buffer-overflow on address 0x5020003ed098 at pc 0x5621a816654e bp 0x7fffb55d4da0 sp 0x7fffb55d4d98 READ of size 4 at 0x5020003eee78 thread T0 #0 0x558265d6654d in aggr_cpu_id__is_empty tools/perf/util/cpumap.c:694:12 #1 0x558265c914da in perf_stat__get_aggr tools/perf/builtin-stat.c:1490:6 #2 0x558265c914da in perf_stat__get_global_cached tools/perf/builtin-stat.c:1530:9 #3 0x558265e53290 in should_skip_zero_counter tools/perf/util/stat-display.c:947:31 #4 0x558265e53290 in print_counter_aggrdata tools/perf/util/stat-display.c:985:18 #5 0x558265e51931 in print_counter tools/perf/util/stat-display.c:1110:3 #6 0x558265e51931 in evlist__print_counters tools/perf/util/stat-display.c:1571:5 #7 0x558265c8ec87 in print_counters tools/perf/builtin-stat.c:981:2 #8 0x558265c8cc71 in cmd_stat tools/perf/builtin-stat.c:2837:3 #9 0x558265bb9bd4 in run_builtin tools/perf/perf.c:323:11 #10 0x558265bb98eb in handle_internal_command tools/perf/perf.c:377:8 #11 0x558265bb9389 in run_argv tools/perf/perf.c:421:2 #12 0x558265bb9389 in main tools/perf/perf.c:537:3 ``` The issue was the use of testing a cpumap with NULL rather than using empty, as a map containing the dummy value isn't NULL and the -1 results in an empty aggr map being allocated which legitimately overflows when any member is accessed. Fixes: 8a96f454 ("perf stat: Avoid SEGV if core.cpus isn't set") Signed-off-by: -
Kajol Jain authored
Update JSON/events for power10 platform with additional metrics. Signed-off-by:
Kajol Jain <kjain@linux.ibm.com> Cc: Athira Rajeev <atrajeev@linux.vnet.ibm.com> Cc: Disha Goel <disgoel@linux.ibm.com> Cc: Madhavan Srinivasan <maddy@linux.ibm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: linuxppc-dev@lists.ozlabs.org Link: https://lore.kernel.org/r/20230905114039.176645-3-kjain@linux.ibm.comSigned-off-by:
-
Kajol Jain authored
Update JSON/Events list with additional data-source events for power10 platform. Signed-off-by:
-
Kajol Jain authored
Update JSON/Events list with data-source events for power10 platform. Signed-off-by:
-
