- 05 May, 2020 8 commits
-
-
Arnaldo Carvalho de Melo authored
To register that an option was set, like with the upcoming 'perf record --switch-output-option' one. Acked-by:
Jiri Olsa <jolsa@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Link: http://lore.kernel.org/lkml/20200429131106.27974-7-acme@kernel.orgSigned-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com>
-
Arnaldo Carvalho de Melo authored
I.e. so far we had just one event in that side band thread, a dummy one with attr.bpf_event set, so that 'perf record' can go ahead and ask the kernel for further information about BPF programs being loaded. Allow for more than one event to be there, so that we can use it as well for the upcoming --switch-output-event feature. Acked-by:
-
Arnaldo Carvalho de Melo authored
To avoid dragging more stuff into the perf python binding in the following csets. Reported-by:
Jiri Olsa <jolsa@kernel.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Signed-off-by:
-
Arnaldo Carvalho de Melo authored
For the upcoming --switch-output-event option we want to create the side band event, populate it with the specified events and then, if it is present multiple times, go on adding to it, then, if the BPF tracking is required, use the first event to set its attr.bpf_event to get those PERF_RECORD_BPF_EVENT metadata events too. Acked-by:
-
Arnaldo Carvalho de Melo authored
Renaming bpf_event__add_sb_event() to evlist__add_sb_event() and requiring that the evlist be allocated beforehand. This will allow using the same side band thread and evlist to be used for multiple purposes in addition to react to PERF_RECORD_BPF_EVENT soon after they are generated. Acked-by:
-
Arnaldo Carvalho de Melo authored
Where state related to a 'perf top' session is grouped. Acked-by:
-
Arnaldo Carvalho de Melo authored
Where state related to a 'perf record' session is grouped. Acked-by:
-
Arnaldo Carvalho de Melo authored
Trying to disentangle this a bit further, unfortunately it uses parse_events(), its interesting to have it separated anyway, so do it. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Signed-off-by:
-
- 30 Apr, 2020 13 commits
-
-
Kajol Jain authored
The hv_24×7 feature in IBM
® POWER9™ processor-based servers provide the facility to continuously collect large numbers of hardware performance metrics efficiently and accurately. This patch adds hv_24x7 metric file for different Socket/chip resources. Result: power9 platform: command:# ./perf stat --metric-only -M Memory_RD_BW_Chip -C 0 -I 1000 1.000096188 0.9 0.3 2.000285720 0.5 0.1 3.000424990 0.4 0.1 command:# ./perf stat --metric-only -M PowerBUS_Frequency -C 0 -I 1000 1.000097981 2.3 2.3 2.000291713 2.3 2.3 3.000421719 2.3 2.3 4.000550912 2.3 2.3 Signed-off-by:Kajol Jain <kjain@linux.ibm.com> Acked-by:
-
Kajol Jain authored
Commit 54b50916 ("perf stat: Implement --metric-only mode") added function 'valid_only_metric()' which drops "Hz" or "hz", if it is part of "ScaleUnit". This patch enable it since hv_24x7 supports couple of frequency events. Signed-off-by:
-
Kajol Jain authored
Added test case for parsing "?" in metric expression. Signed-off-by:
-
Kajol Jain authored
Patch enhances current metric infrastructure to handle "?" in the metric expression. The "?" can be use for parameters whose value not known while creating metric events and which can be replace later at runtime to the proper value. It also add flexibility to create multiple events out of single metric event added in JSON file. Patch adds function 'arch_get_runtimeparam' which is a arch specific function, returns the count of metric events need to be created. By default it return 1. This infrastructure needed for hv_24x7 socket/chip level events. "hv_24x7" chip level events needs specific chip-id to which the data is requested. Function 'arch_get_runtimeparam' implemented in header.c which extract number of sockets from sysfs file "sockets" under "/sys/devices/hv_24x7/interface/". With this patch basically we are trying to create as many metric events as define by runtime_param. For that one loop is added in function 'metricgroup__add_metric', which create multiple events at run time depend on return value of 'arch_get_runtimeparam' and merge that event in 'group_list'. To achieve that we are actually passing this parameter value as part of `expr__find_other` function and changing "?" present in metric expression with this value. As in our JSON file, there gonna be single metric event, and out of which we are creating multiple events. To understand which data count belongs to which parameter value, we also printing param value in generic_metric function. For example, command:# ./perf stat -M PowerBUS_Frequency -C 0 -I 1000 1.000101867 9,356,933 hv_24x7/pm_pb_cyc,chip=0/ # 2.3 GHz PowerBUS_Frequency_0 1.000101867 9,366,134 hv_24x7/pm_pb_cyc,chip=1/ # 2.3 GHz PowerBUS_Frequency_1 2.000314878 9,365,868 hv_24x7/pm_pb_cyc,chip=0/ # 2.3 GHz PowerBUS_Frequency_0 2.000314878 9,366,092 hv_24x7/pm_pb_cyc,chip=1/ # 2.3 GHz PowerBUS_Frequency_1 So, here _0 and _1 after PowerBUS_Frequency specify parameter value. Signed-off-by: -
Shaokun Zhang authored
get_cpuid_str() is used in tools/perf/arch/xxx/util/header.c, fix the name in comment. Signed-off-by:
Shaokun Zhang <zhangshaokun@hisilicon.com> Cc: Andi Kleen <ak@linux.intel.com> Link: http://lore.kernel.org/lkml/1588141992-48382-1-git-send-email-zhangshaokun@hisilicon.comSigned-off-by:
-
Zou Wei authored
Fixes coccicheck warning: tools/perf/builtin-report.c:1403:2-34: WARNING: Assignment of 0/1 to bool variable Reported-by:
Hulk Robot <hulkci@huawei.com> Signed-off-by:
Zou Wei <zou_wei@huawei.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lore.kernel.org/lkml/1587904683-3510-1-git-send-email-zou_wei@huawei.comSigned-off-by:
-
Zou Wei authored
Fixes coccicheck warnings: tools/perf/builtin-diff.c:1565:2-3: Unneeded semicolon tools/perf/builtin-lock.c:778:2-3: Unneeded semicolon tools/perf/builtin-mem.c:126:2-3: Unneeded semicolon tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c:555:2-3: Unneeded semicolon tools/perf/util/ordered-events.c:317:2-3: Unneeded semicolon tools/perf/util/synthetic-events.c:1131:2-3: Unneeded semicolon tools/perf/util/trace-event-read.c:78:2-3: Unneeded semicolon Reported-by:
-
Zou Wei authored
Fixes coccicheck warnings: tools/perf/builtin-c2c.c:1712:2-3: Unneeded semicolon tools/perf/builtin-c2c.c:1928:2-3: Unneeded semicolon tools/perf/builtin-c2c.c:2962:2-3: Unneeded semicolon Reported-by:
-
Zou Wei authored
Fixes coccicheck warning: tools/lib/traceevent/kbuffer-parse.c:441:2-3: Unneeded semicolon Reported-by:
Steven Rostedt (VMware) <rostedt@goodmis.org> Link: http://lore.kernel.org/lkml/1588065121-71236-1-git-send-email-zou_wei@huawei.comSigned-off-by:
-
Stephane Eranian authored
When printing iregs, there was a double newline printed because perf_sample__fprintf_regs() was printing its own and then at the end of all fields, perf script was adding one. This was causing blank line in the output: Before: $ perf script -Fip,iregs 401b8d ABI:2 DX:0x100 SI:0x4a8340 DI:0x4a9340 401b8d ABI:2 DX:0x100 SI:0x4a9340 DI:0x4a8340 401b8d ABI:2 DX:0x100 SI:0x4a8340 DI:0x4a9340 401b8d ABI:2 DX:0x100 SI:0x4a9340 DI:0x4a8340 After: $ perf script -Fip,iregs 401b8d ABI:2 DX:0x100 SI:0x4a8340 DI:0x4a9340 401b8d ABI:2 DX:0x100 SI:0x4a9340 DI:0x4a8340 401b8d ABI:2 DX:0x100 SI:0x4a8340 DI:0x4a9340 Committer testing: First we need to figure out how to request that registers be recorded, so we use: # perf record -h reg Usage: perf record [<options>] [<command>] or: perf record [<options>] -- <command> [<options>] -I, --intr-regs[=<any register>] sample selected machine registers on interrupt, use '-I?' to list register names --buildid-all Record build-id of all DSOs regardless of hits --user-regs[=<any register>] sample selected machine registers on interrupt, use '--user-regs=?' to list register names # Ok, now lets ask for them all: # perf record -a --intr-regs --user-regs sleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 4.105 MB perf.data (2760 samples) ] # Lets look at the first 6 output lines: # perf script -Fip,iregs | head -6 ffffffff8a06f2f4 ABI:2 AX:0xffffd168fee0a980 BX:0xffff8a23b087f000 CX:0xfffeb69aaeb25d73 DX:0xffff8a253e8310f0 SI:0xfffffff9bafe7359 DI:0xffffb1690204fb10 BP:0xffffd168fee0a950 SP:0xffffb1690204fb88 IP:0xffffffff8a06f2f4 FLAGS:0x4e CS:0x10 SS:0x18 R8:0x1495f0a91129a R9:0xffff8a23b087f000 R10:0x1 R11:0xffffffff R12:0x0 R13:0xffff8a253e827e00 R14:0xffffd168fee0aa5c R15:0xffffd168fee0a980 ffffffff8a06f2f4 ABI:2 AX:0x0 BX:0xffffd168fee0a950 CX:0x5684cc1118491900 DX:0x0 SI:0xffffd168fee0a9d0 DI:0x202 BP:0xffffb1690204fd70 SP:0xffffb1690204fd20 IP:0xffffffff8a06f2f4 FLAGS:0x24e CS:0x10 SS:0x18 R8:0x0 R9:0xffffd168fee0a9d0 R10:0x1 R11:0xffffffff R12:0xffffffff8a23e480 R13:0xffff8a23b087f240 R14:0xffff8a23b087f000 R15:0xffffd168fee0a950 ffffffff8a06f2f4 ABI:2 AX:0x0 BX:0x0 CX:0x7f25f334335b DX:0x0 SI:0x2400 DI:0x4 BP:0x7fff5f264570 SP:0x7fff5f264538 IP:0xffffffff8a06f2f4 FLAGS:0x24e CS:0x10 SS:0x2b R8:0x0 R9:0x2312d20 R10:0x0 R11:0x246 R12:0x22cc0e0 R13:0x0 R14:0x0 R15:0x22d0780 # Reproduced, apply the patch and: [root@five ~]# perf script -Fip,iregs | head -6 ffffffff8a06f2f4 ABI:2 AX:0xffffd168fee0a980 BX:0xffff8a23b087f000 CX:0xfffeb69aaeb25d73 DX:0xffff8a253e8310f0 SI:0xfffffff9bafe7359 DI:0xffffb1690204fb10 BP:0xffffd168fee0a950 SP:0xffffb1690204fb88 IP:0xffffffff8a06f2f4 FLAGS:0x4e CS:0x10 SS:0x18 R8:0x1495f0a91129a R9:0xffff8a23b087f000 R10:0x1 R11:0xffffffff R12:0x0 R13:0xffff8a253e827e00 R14:0xffffd168fee0aa5c R15:0xffffd168fee0a980 ffffffff8a06f2f4 ABI:2 AX:0x0 BX:0xffffd168fee0a950 CX:0x5684cc1118491900 DX:0x0 SI:0xffffd168fee0a9d0 DI:0x202 BP:0xffffb1690204fd70 SP:0xffffb1690204fd20 IP:0xffffffff8a06f2f4 FLAGS:0x24e CS:0x10 SS:0x18 R8:0x0 R9:0xffffd168fee0a9d0 R10:0x1 R11:0xffffffff R12:0xffffffff8a23e480 R13:0xffff8a23b087f240 R14:0xffff8a23b087f000 R15:0xffffd168fee0a950 ffffffff8a06f2f4 ABI:2 AX:0x0 BX:0x0 CX:0x7f25f334335b DX:0x0 SI:0x2400 DI:0x4 BP:0x7fff5f264570 SP:0x7fff5f264538 IP:0xffffffff8a06f2f4 FLAGS:0x24e CS:0x10 SS:0x2b R8:0x0 R9:0x2312d20 R10:0x0 R11:0x246 R12:0x22cc0e0 R13:0x0 R14:0x0 R15:0x22d0780 ffffffff8a24074b ABI:2 AX:0xcb BX:0xcb CX:0x0 DX:0x0 SI:0xffffb1690204ff58 DI:0xcb BP:0xffffb1690204ff58 SP:0xffffb1690204ff40 IP:0xffffffff8a24074b FLAGS:0x24e CS:0x10 SS:0x18 R8:0x0 R9:0x0 R10:0x0 R11:0x0 R12:0x0 R13:0x0 R14:0x0 R15:0x0 ffffffff8a310600 ABI:2 AX:0x0 BX:0xffffffff8b8c39a0 CX:0x0 DX:0xffff8a2503890300 SI:0xffffb1690204ff20 DI:0xffff8a23e4080000 BP:0xffff8a23e4080000 SP:0xffffb1690204fec0 IP:0xffffffff8a310600 FLAGS:0x28e CS:0x10 SS:0x18 R8:0x0 R9:0x0 R10:0x0 R11:0x0 R12:0xffffffffffffffea R13:0xffff8a23e4080020 R14:0x0 R15:0x0 ffffffff8a11b688 ABI:2 AX:0x0 BX:0xffff8a237b7c8800 CX:0xffffb1690204fae0 DX:0x78 SI:0xffff8a237b7c8800 DI:0xffffb1690204fa10 BP:0xffffb1690204fb00 SP:0xffffb1690204fa00 IP:0xffffffff8a11b688 FLAGS:0x8a CS:0x10 SS:0x18 R8:0x1495f0a917eba R9:0xffffd168fde19a48 R10:0xffffb1690204fd98 R11:0xffff8a253e82afb0 R12:0xffff8a237b7c8800 R13:0xffffb1690204fb00 R14:0x0 R15:0xffff8a237b7c8800 [root@five ~]# To see it more clearly, lets get just two of those registers by sample: # perf record -a --intr-regs=ax,bx --user-regs=cx,dx sleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 3.502 MB perf.data (1653 samples) ] # Extra info, lets see what gets setup in that 'struct perf_event_attr': # perf evlist -v cycles: size: 120, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD|REGS_USER|REGS_INTR, read_format: ID, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 2, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol: 1, bpf_event: 1, sample_regs_user: 0xc, sample_regs_intr: 0x3 # Cook, some PERF_SAMPLE_REGS_USER|PERF_SAMPLE_REGS_INTR + attr.sample_regs_user and attr.sample_regs_intr register masks, now lets see if those newlines are gone in a more compact fashion: # perf script -Fip,iregs,uregs ffffffff8a56df78 ABI:2 AX:0xffff8a25137b6028 BX:0xffff8a2502f18000 ABI:2 CX:0x7f204460e49b DX:0xf42920 ffffffff8a56df78 ABI:2 AX:0xffff8a25137b6028 BX:0xffff8a2502f18000 ABI:2 CX:0x7f204460e49b DX:0xf42920 ffffffff8a56df78 ABI:2 AX:0xffff8a25137b6028 BX:0xffff8a2502f18000 ABI:2 CX:0x7f204460e49b DX:0xf42920 ffffffff8a56df78 ABI:2 AX:0xffff8a25137b6028 BX:0xffff8a2502f18000 ABI:2 CX:0x7f204460e49b DX:0xf42920 ffffffff8a56df78 ABI:2 AX:0xffff8a25137b6028 BX:0xffff8a2502f18000 ABI:2 CX:0x7f204460e49b DX:0xf42920 ffffffff8a56df78 ABI:2 AX:0xffff8a25137b6028 BX:0xffff8a2502f18000 ABI:2 CX:0x7f204460e49b DX:0xf42920 ffffffff8a29b78d ABI:2 AX:0x2a20ffcd6000 BX:0x2ec7d9000 ABI:2 CX:0x7f204460e49b DX:0xf42920 # And where was that? # perf script -Fip,iregs,uregs,sym,dso ffffffff8a56df78 strrchr (/lib/modules/5.7.0-rc2/build/vmlinux) ABI:2 AX:0xffff8a25137b6028 BX:0xffff8a2502f18000 ABI:2 CX:0x7f204460e49b DX:0xf42920 ffffffff8a56df78 strrchr (/lib/modules/5.7.0-rc2/build/vmlinux) ABI:2 AX:0xffff8a25137b6028 BX:0xffff8a2502f18000 ABI:2 CX:0x7f204460e49b DX:0xf42920 ffffffff8a56df78 strrchr (/lib/modules/5.7.0-rc2/build/vmlinux) ABI:2 AX:0xffff8a25137b6028 BX:0xffff8a2502f18000 ABI:2 CX:0x7f204460e49b DX:0xf42920 ffffffff8a56df78 strrchr (/lib/modules/5.7.0-rc2/build/vmlinux) ABI:2 AX:0xffff8a25137b6028 BX:0xffff8a2502f18000 ABI:2 CX:0x7f204460e49b DX:0xf42920 ffffffff8a56df78 strrchr (/lib/modules/5.7.0-rc2/build/vmlinux) ABI:2 AX:0xffff8a25137b6028 BX:0xffff8a2502f18000 ABI:2 CX:0x7f204460e49b DX:0xf42920 ffffffff8a56df78 strrchr (/lib/modules/5.7.0-rc2/build/vmlinux) ABI:2 AX:0xffff8a25137b6028 BX:0xffff8a2502f18000 ABI:2 CX:0x7f204460e49b DX:0xf42920 ffffffff8a29b78d __vma_link_rb (/lib/modules/5.7.0-rc2/build/vmlinux) ABI:2 AX:0x2a20ffcd6000 BX:0x2ec7d9000 ABI:2 CX:0x7f204460e49b DX:0xf42920 # Signed-off-by:Stephane Eranian <eranian@google.com> Tested-by:
-
Ian Rogers authored
The synthesize benchmark, run on a single process and thread, shows perf_event__synthesize_mmap_events as the hottest function with fgets and sscanf taking the majority of execution time. fscanf performs similarly well. Replace the scanf call with manual reading of each field of the /proc/pid/maps line, and remove some unnecessary buffering. This change also addresses potential, but unlikely, buffer overruns for the string values read by scanf. Performance before is: $ sudo perf bench internals synthesize -m 16 -M 16 -s -t \# Running 'internals/synthesize' benchmark: Computing performance of single threaded perf event synthesis by synthesizing events on the perf process itself: Average synthesis took: 102.810 usec (+- 0.027 usec) Average num. events: 17.000 (+- 0.000) Average time per event 6.048 usec Average data synthesis took: 106.325 usec (+- 0.018 usec) Average num. events: 89.000 (+- 0.000) Average time per event 1.195 usec Computing performance of multi threaded perf event synthesis by synthesizing events on CPU 0: Number of synthesis threads: 16 Average synthesis took: 68103.100 usec (+- 441.234 usec) Average num. events: 30703.000 (+- 0.730) Average time per event 2.218 usec And after is: $ sudo perf bench internals synthesize -m 16 -M 16 -s -t \# Running 'internals/synthesize' benchmark: Computing performance of single threaded perf event synthesis by synthesizing events on the perf process itself: Average synthesis took: 50.388 usec (+- 0.031 usec) Average num. events: 17.000 (+- 0.000) Average time per event 2.964 usec Average data synthesis took: 52.693 usec (+- 0.020 usec) Average num. events: 89.000 (+- 0.000) Average time per event 0.592 usec Computing performance of multi threaded perf event synthesis by synthesizing events on CPU 0: Number of synthesis threads: 16 Average synthesis took: 45022.400 usec (+- 552.740 usec) Average num. events: 30624.200 (+- 10.037) Average time per event 1.470 usec On a Intel Xeon 6154 compiling with Debian gcc 9.2.1. Committer testing: On a AMD Ryzen 5 3600X 6-Core Processor: Before: # perf bench internals synthesize --min-threads 12 --max-threads 12 --st --mt # Running 'internals/synthesize' benchmark: Computing performance of single threaded perf event synthesis by synthesizing events on the perf process itself: Average synthesis took: 267.491 usec (+- 0.176 usec) Average num. events: 56.000 (+- 0.000) Average time per event 4.777 usec Average data synthesis took: 277.257 usec (+- 0.169 usec) Average num. events: 287.000 (+- 0.000) Average time per event 0.966 usec Computing performance of multi threaded perf event synthesis by synthesizing events on CPU 0: Number of synthesis threads: 12 Average synthesis took: 81599.500 usec (+- 346.315 usec) Average num. events: 36096.100 (+- 2.523) Average time per event 2.261 usec # After: # perf bench internals synthesize --min-threads 12 --max-threads 12 --st --mt # Running 'internals/synthesize' benchmark: Computing performance of single threaded perf event synthesis by synthesizing events on the perf process itself: Average synthesis took: 110.125 usec (+- 0.080 usec) Average num. events: 56.000 (+- 0.000) Average time per event 1.967 usec Average data synthesis took: 118.518 usec (+- 0.057 usec) Average num. events: 287.000 (+- 0.000) Average time per event 0.413 usec Computing performance of multi threaded perf event synthesis by synthesizing events on CPU 0: Number of synthesis threads: 12 Average synthesis took: 43490.700 usec (+- 284.527 usec) Average num. events: 37028.500 (+- 0.563) Average time per event 1.175 usec # Signed-off-by:Ian Rogers <irogers@google.com> Tested-by:
Namhyung Kim <namhyung@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andrey Zhizhikin <andrey.z@gmail.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Kefeng Wang <wangkefeng.wang@huawei.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Petr Mladek <pmladek@suse.com> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lore.kernel.org/lkml/20200415054050.31645-4-irogers@google.comSigned-off-by:
-
Ian Rogers authored
The synthesize benchmark shows the majority of execution time going to fgets and sscanf, necessary to parse /proc/pid/maps. Add a new buffered reading library that will be used to replace these calls in a follow-up CL. Add tests for the library to perf test. Committer tests: $ perf test api 63: Test api io : Ok $ Signed-off-by:
-
Ian Rogers authored
By default this isn't run as it reads /proc and may not have access. For consistency, modify the single threaded benchmark to compute an average time per event. Committer testing: $ grep -m1 "model name" /proc/cpuinfo model name : Intel(R) Core(TM) i7-8650U CPU @ 1.90GHz $ grep "model name" /proc/cpuinfo | wc -l 8 $ $ perf bench internals synthesize -h # Running 'internals/synthesize' benchmark: Usage: perf bench internals synthesize <options> -I, --multi-iterations <n> Number of iterations used to compute multi-threaded average -i, --single-iterations <n> Number of iterations used to compute single-threaded average -M, --max-threads <n> Maximum number of threads in multithreaded bench -m, --min-threads <n> Minimum number of threads in multithreaded bench -s, --st Run single threaded benchmark -t, --mt Run multi-threaded benchmark $ $ perf bench internals synthesize -t # Running 'internals/synthesize' benchmark: Computing performance of multi threaded perf event synthesis by synthesizing events on CPU 0: Number of synthesis threads: 1 Average synthesis took: 65449.000 usec (+- 586.442 usec) Average num. events: 9405.400 (+- 0.306) Average time per event 6.959 usec Number of synthesis threads: 2 Average synthesis took: 37838.300 usec (+- 130.259 usec) Average num. events: 9501.800 (+- 20.469) Average time per event 3.982 usec Number of synthesis threads: 3 Average synthesis took: 48551.400 usec (+- 225.686 usec) Average num. events: 9544.000 (+- 0.000) Average time per event 5.087 usec Number of synthesis threads: 4 Average synthesis took: 29632.500 usec (+- 50.808 usec) Average num. events: 9544.000 (+- 0.000) Average time per event 3.105 usec Number of synthesis threads: 5 Average synthesis took: 33920.400 usec (+- 284.509 usec) Average num. events: 9544.000 (+- 0.000) Average time per event 3.554 usec Number of synthesis threads: 6 Average synthesis took: 27604.100 usec (+- 72.344 usec) Average num. events: 9548.000 (+- 0.000) Average time per event 2.891 usec Number of synthesis threads: 7 Average synthesis took: 25406.300 usec (+- 933.371 usec) Average num. events: 9545.500 (+- 0.167) Average time per event 2.662 usec Number of synthesis threads: 8 Average synthesis took: 24110.400 usec (+- 73.229 usec) Average num. events: 9551.000 (+- 0.000) Average time per event 2.524 usec $ Signed-off-by:
-
- 23 Apr, 2020 3 commits
-
-
Stephane Eranian authored
To control degree of parallelism of the synthesize_mmap() code which is scanning /proc/PID/task/PID/maps and can be time consuming. Mimic perf top way of handling the option. If not specified will default to 1 thread, i.e. default behavior before this option. On a desktop computer the processing of /proc/PID/task/PID/maps isn't slow enough to warrant parallel processing and the thread creation has some cost - hence the default of 1. On a loaded server with >100 cores it is possible to see synthesis times in the order of seconds and in this case having the option is desirable. As the processing is a synchronization point, it is legitimate to worry if Amdahl's law will apply to this patch. Profiling with this patch in place: https://lore.kernel.org/lkml/20200415054050.31645-4-irogers@google.com/ shows: ... - 32.59% __perf_event__synthesize_threads - 32.54% __event__synthesize_thread + 22.13% perf_event__synthesize_mmap_events + 6.68% perf_event__get_comm_ids.constprop.0 + 1.49% process_synthesized_event + 1.29% __GI___readdir64 + 0.60% __opendir ... That is the processing is 1.49% of execution time and there is plenty to make parallel. This is shown in the benchmark in this patch: https://lore.kernel.org/lkml/20200415054050.31645-2-irogers@google.com/ Computing performance of multi threaded perf event synthesis by synthesizing events on CPU 0: Number of synthesis threads: 1 Average synthesis took: 127729.000 usec (+- 3372.880 usec) Average num. events: 21548.600 (+- 0.306) Average time per event 5.927 usec Number of synthesis threads: 2 Average synthesis took: 88863.500 usec (+- 385.168 usec) Average num. events: 21552.800 (+- 0.327) Average time per event 4.123 usec Number of synthesis threads: 3 Average synthesis took: 83257.400 usec (+- 348.617 usec) Average num. events: 21553.200 (+- 0.327) Average time per event 3.863 usec Number of synthesis threads: 4 Average synthesis took: 75093.000 usec (+- 422.978 usec) Average num. events: 21554.200 (+- 0.200) Average time per event 3.484 usec Number of synthesis threads: 5 Average synthesis took: 64896.600 usec (+- 353.348 usec) Average num. events: 21558.000 (+- 0.000) Average time per event 3.010 usec Number of synthesis threads: 6 Average synthesis took: 59210.200 usec (+- 342.890 usec) Average num. events: 21560.000 (+- 0.000) Average time per event 2.746 usec Number of synthesis threads: 7 Average synthesis took: 54093.900 usec (+- 306.247 usec) Average num. events: 21562.000 (+- 0.000) Average time per event 2.509 usec Number of synthesis threads: 8 Average synthesis took: 48938.700 usec (+- 341.732 usec) Average num. events: 21564.000 (+- 0.000) Average time per event 2.269 usec Where average time per synthesized event goes from 5.927 usec with 1 thread to 2.269 usec with 8. This isn't a linear speed up as not all of synthesize code has been made parallel. If the synthesis time was about 10 seconds then using 8 threads may bring this down to less than 4. Signed-off-by:
-
Tommi Rantala authored
Commit 2d4f2799 ("perf data: Add global path holder") missed path conversion in tests/topology.c, causing the "Session topology" testcase to "hang" (waits forever for input from stdin) when doing "ssh $VM perf test". Can be reproduced by running "cat | perf test topo", and crashed by replacing cat with true: $ true | perf test -v topo 40: Session topology : --- start --- test child forked, pid 3638 templ file: /tmp/perf-test-QPvAch incompatible file format incompatible file format (rerun with -v to learn more) free(): invalid pointer test child interrupted ---- end ---- Session topology: FAILED! Committer testing: Reproduced the above result before the patch and after it is back working: # true | perf test -v topo 41: Session topology : --- start --- test child forked, pid 19374 templ file: /tmp/perf-test-YOTEQg CPU 0, core 0, socket 0 CPU 1, core 1, socket 0 CPU 2, core 2, socket 0 CPU 3, core 3, socket 0 CPU 4, core 0, socket 0 CPU 5, core 1, socket 0 CPU 6, core 2, socket 0 CPU 7, core 3, socket 0 test child finished with 0 ---- end ---- Session topology: Ok # Fixes: 2d4f2799 ("perf data: Add global path holder") Signed-off-by:
Tommi Rantala <tommi.t.rantala@nokia.com> Tested-by:
-
Jin Yao authored
For interval mode, the metric is printed after the '#' character if it exists. But it's not calculated by the counts generated in this interval. See the following examples: root@kbl-ppc:~# perf stat -M CPI -I1000 --interval-count 2 # time counts unit events 1.000422803 764,809 inst_retired.any # 2.9 CPI 1.000422803 2,234,932 cycles 2.001464585 1,960,061 inst_retired.any # 1.6 CPI 2.001464585 4,022,591 cycles The second CPI should not be 1.6 (4,022,591/1,960,061 is 2.1) root@kbl-ppc:~# perf stat -e cycles,instructions -I1000 --interval-count 2 # time counts unit events 1.000429493 2,869,311 cycles 1.000429493 816,875 instructions # 0.28 insn per cycle 2.001516426 9,260,973 cycles 2.001516426 5,250,634 instructions # 0.87 insn per cycle The second 'insn per cycle' should not be 0.87 (5,250,634/9,260,973 is 0.57). The current code uses a global variable 'rt_stat' for tracking and updating the std dev of runtime stat. Unlike the counts, 'rt_stat' is not reset for interval. While the counts are reset for interval. perf_stat_process_counter() { if (config->interval) init_stats(ps->res_stats); } So for interval mode, the 'rt_stat' variable should be reset too. This patch resets 'rt_stat' before read_counters(), so the runtime stat is only calculated by the counts generated in this interval. With this patch: root@kbl-ppc:~# perf stat -M CPI -I1000 --interval-count 2 # time counts unit events 1.000420924 2,408,818 inst_retired.any # 2.1 CPI 1.000420924 5,010,111 cycles 2.001448579 2,798,407 inst_retired.any # 1.6 CPI 2.001448579 4,599,861 cycles root@kbl-ppc:~# perf stat -e cycles,instructions -I1000 --interval-count 2 # time counts unit events 1.000428555 2,769,714 cycles 1.000428555 774,462 instructions # 0.28 insn per cycle 2.001471562 3,595,904 cycles 2.001471562 1,243,703 instructions # 0.35 insn per cycle Now the second 'insn per cycle' and CPI are calculated by the counts generated in this interval. Signed-off-by:Jin Yao <yao.jin@linux.intel.com> Acked-by:
-
- 22 Apr, 2020 6 commits
-
-
Jin Yao authored
As the code comments in perf_stat_process_counter() say, we calculate counter's data every interval, and the display code shows ps->res_stats avg value. We need to zero the stats for interval mode. But the current code only zeros the res_stats[0], it doesn't zero the res_stats[1] and res_stats[2], which are for ena and run of counter. This patch zeros the whole res_stats[] for interval mode. Fixes: 51fd2df1 ("perf stat: Fix interval output values") Signed-off-by:
-
Ian Rogers authored
al->sym may be NULL given current if conditions and may cause a segv. Fixes: d2bedb78 ("perf script: Allow --symbol to accept hexadecimal addresses") Signed-off-by:
-
Jagadeesh Pagadala authored
Code cleanup: Remove duplicate headers which are included twice. Signed-off-by:
Jagadeesh Pagadala <jagdsh.linux@gmail.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Link: http://lore.kernel.org/lkml/1587276836-17088-1-git-send-email-jagdsh.linux@gmail.comSigned-off-by:
-
Tommi Rantala authored
Fix div-by-zero if runtime is zero: $ perf bench futex hash --runtime=0 # Running 'futex/hash' benchmark: Run summary [PID 12090]: 4 threads, each operating on 1024 [private] futexes for 0 secs. Floating point exception (core dumped) Signed-off-by:
-
Tommi Rantala authored
Do not bother with close() if fd is not valid, just to silence valgrind: $ valgrind ./perf script ==59169== Memcheck, a memory error detector ==59169== Copyright (C) 2002-2017, and GNU GPL'd, by Julian Seward et al. ==59169== Using Valgrind-3.14.0 and LibVEX; rerun with -h for copyright info ==59169== Command: ./perf script ==59169== ==59169== Warning: invalid file descriptor -1 in syscall close() ==59169== Warning: invalid file descriptor -1 in syscall close() ==59169== Warning: invalid file descriptor -1 in syscall close() ==59169== Warning: invalid file descriptor -1 in syscall close() ==59169== Warning: invalid file descriptor -1 in syscall close() ==59169== Warning: invalid file descriptor -1 in syscall close() ==59169== Warning: invalid file descriptor -1 in syscall close() ==59169== Warning: invalid file descriptor -1 in syscall close() Signed-off-by: -
Ingo Molnar authored
Merge tag 'perf-core-for-mingo-5.8-20200420' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core Pull perf/core fixes and improvements from Arnaldo Carvalho de Melo: kernel + tools/perf: Alexey Budankov: - Introduce CAP_PERFMON to kernel and user space. callchains: Adrian Hunter: - Allow using Intel PT to synthesize callchains for regular events. Kan Liang: - Stitch LBR records from multiple samples to get deeper backtraces, there are caveats, see the csets for details. perf script: Andreas Gerstmayr: - Add flamegraph.py script BPF: Jiri Olsa: - Synthesize bpf_trampoline/dispatcher ksymbol events. perf stat: Arnaldo Carvalho de Melo: - Honour --timeout for forked workloads. Stephane Eranian: - Force error in fallback on :k events, to avoid counting nothing when the user asks for kernel events but is not allowed to. perf bench: Ian Rogers: - Add event synthesis benchmark. tools api fs: Stephane Eranian: - Make xxx__mountpoint() more scalable libtraceevent: He Zhe: - Handle return value of asprintf. Signed-off-by:
Ingo Molnar <mingo@kernel.org>
-
- 21 Apr, 2020 10 commits
-
-
Linus Torvalds authored
Merge misc fixes from Andrew Morton: "15 fixes" * emailed patches from Andrew Morton <akpm@linux-foundation.org>: tools/vm: fix cross-compile build coredump: fix null pointer dereference on coredump mm: shmem: disable interrupt when acquiring info->lock in userfaultfd_copy path shmem: fix possible deadlocks on shmlock_user_lock vmalloc: fix remap_vmalloc_range() bounds checks mm/shmem: fix build without THP mm/ksm: fix NULL pointer dereference when KSM zero page is enabled tools/build: tweak unused value workaround checkpatch: fix a typo in the regex for $allocFunctions mm, gup: return EINTR when gup is interrupted by fatal signals mm/hugetlb: fix a addressing exception caused by huge_pte_offset MAINTAINERS: add an entry for kfifo mm/userfaultfd: disable userfaultfd-wp on x86_32 slub: avoid redzone when choosing freepointer location sh: fix build error in mm/init.c
-
git://git.kernel.org/pub/scm/virt/kvm/kvmLinus Torvalds authored
Pull kvm fixes from Paolo Bonzini: "Bugfixes, and a few cleanups to the newly-introduced assembly language vmentry code for AMD" * tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: KVM: PPC: Book3S HV: Handle non-present PTEs in page fault functions kvm: Disable objtool frame pointer checking for vmenter.S MAINTAINERS: add a reviewer for KVM/s390 KVM: s390: Fix PV check in deliverable_irqs() kvm: Handle reads of SandyBridge RAPL PMU MSRs rather than injecting #GP KVM: Remove CREATE_IRQCHIP/SET_PIT2 race KVM: SVM: Fix __svm_vcpu_run declaration. KVM: SVM: Do not setup frame pointer in __svm_vcpu_run KVM: SVM: Fix build error due to missing release_pages() include KVM: SVM: Do not mark svm_vcpu_run with STACK_FRAME_NON_STANDARD kvm: nVMX: match comment with return type for nested_vmx_exit_reflected kvm: nVMX: reflect MTF VM-exits if injected by L1 KVM: s390: Return last valid slot if approx index is out-of-bounds KVM: Check validity of resolved slot when searching memslots KVM: VMX: Enable machine check support for 32bit targets KVM: SVM: move more vmentry code to assembly KVM: SVM: fix compilation with modular PSP and non-modular KVM
-
git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhostLinus Torvalds authored
Pull virtio fixes and cleanups from Michael Tsirkin: - Some bug fixes - Cleanup a couple of issues that surfaced meanwhile - Disable vhost on ARM with OABI for now - to be fixed fully later in the cycle or in the next release. * tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost: (24 commits) vhost: disable for OABI virtio: drop vringh.h dependency virtio_blk: add a missing include virtio-balloon: Avoid using the word 'report' when referring to free page hinting virtio-balloon: make virtballoon_free_page_report() static vdpa: fix comment of vdpa_register_device() vdpa: make vhost, virtio depend on menu vdpa: allow a 32 bit vq alignment drm/virtio: fix up for include file changes remoteproc: pull in slab.h rpmsg: pull in slab.h virtio_input: pull in slab.h remoteproc: pull in slab.h virtio-rng: pull in slab.h virtgpu: pull in uaccess.h tools/virtio: make asm/barrier.h self contained tools/virtio: define aligned attribute virtio/test: fix up after IOTLB changes vhost: Create accessors for virtqueues private_data vdpasim: Return status in vdpasim_get_status ...
-
git://git.infradead.org/users/jjs/linux-tpmddLinus Torvalds authored
Pull tpm fixes from Jarkko Sakkinen: "A few bug fixes" * tag 'tpmdd-next-20200421' of git://git.infradead.org/users/jjs/linux-tpmdd: tpm/tpm_tis: Free IRQ if probing fails tpm: fix wrong return value in tpm_pcr_extend tpm: ibmvtpm: retry on H_CLOSED in tpm_ibmvtpm_send() tpm: Export tpm2_get_cc_attrs_tbl for ibmvtpm driver as module
-
git://github.com/ojeda/linuxLinus Torvalds authored
Pull clang-format fixlets from Miguel Ojeda: "Two trivial clang-format changes: - Don't indent C++ namespaces (Ian Rogers) - The usual clang-format macro list update (Miguel Ojeda)" * tag 'clang-format-for-linus-v5.7-rc3' of git://github.com/ojeda/linux: clang-format: Update with the latest for_each macro list clang-format: don't indent namespaces
-
Lucas Stach authored
Commit 7ed1c190 ("tools: fix cross-compile var clobbering") moved the setup of the CC variable to tools/scripts/Makefile.include to make the behavior consistent across all the tools Makefiles. As the vm tools missed the include we end up with the wrong CC in a cross-compiling evironment. Fixes: 7ed1c190 (tools: fix cross-compile var clobbering) Signed-off-by:
Lucas Stach <l.stach@pengutronix.de> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Cc: Martin Kelly <martin@martingkelly.com> Cc: <stable@vger.kernel.org> Link: http://lkml.kernel.org/r/20200416104748.25243-1-l.stach@pengutronix.deSigned-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Sudip Mukherjee authored
If the core_pattern is set to "|" and any process segfaults then we get a null pointer derefernce while trying to coredump. The call stack shows: RIP: do_coredump+0x628/0x11c0 When the core_pattern has only "|" there is no use of trying the coredump and we can check that while formating the corename and exit with an error. After this change I get: format_corename failed Aborting core Fixes: 315c6926 ("coredump: split pipe command whitespace before expanding template") Reported-by:Matthew Ruffell <matthew.ruffell@canonical.com> Signed-off-by:
Sudip Mukherjee <sudipm.mukherjee@gmail.com> Signed-off-by:
-
Yang Shi authored
Syzbot reported the below lockdep splat: WARNING: possible irq lock inversion dependency detected 5.6.0-rc7-syzkaller #0 Not tainted -------------------------------------------------------- syz-executor.0/10317 just changed the state of lock: ffff888021d16568 (&(&info->lock)->rlock){+.+.}, at: spin_lock include/linux/spinlock.h:338 [inline] ffff888021d16568 (&(&info->lock)->rlock){+.+.}, at: shmem_mfill_atomic_pte+0x1012/0x21c0 mm/shmem.c:2407 but this lock was taken by another, SOFTIRQ-safe lock in the past: (&(&xa->xa_lock)->rlock#5){..-.} and interrupts could create inverse lock ordering between them. other info that might help us debug this: Possible interrupt unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&(&info->lock)->rlock); local_irq_disable(); lock(&(&xa->xa_lock)->rlock#5); lock(&(&info->lock)->rlock); <Interrupt> lock(&(&xa->xa_lock)->rlock#5); *** DEADLOCK *** The full report is quite lengthy, please see: https://lore.kernel.org/linux-mm/alpine.LSU.2.11.2004152007370.13597@eggly.anvils/T/#m813b412c5f78e25ca8c6c7734886ed4de43f241d It is because CPU 0 held info->lock with IRQ enabled in userfaultfd_copy path, then CPU 1 is splitting a THP which held xa_lock and info->lock in IRQ disabled context at the same time. If softirq comes in to acquire xa_lock, the deadlock would be triggered. The fix is to acquire/release info->lock with *_irq version instead of plain spin_{lock,unlock} to make it softirq safe. Fixes: 4c27fe4c ("userfaultfd: shmem: add shmem_mcopy_atomic_pte for userfaultfd support") Reported-by: syzbot+e27980339d305f2dbfd9@syzkaller.appspotmail.com Signed-off-by:Yang Shi <yang.shi@linux.alibaba.com> Signed-off-by:
Hugh Dickins <hughd@google.com> Cc: Andrea Arcangeli <aarcange@redhat.com> Link: http://lkml.kernel.org/r/1587061357-122619-1-git-send-email-yang.shi@linux.alibaba.comSigned-off-by:
-
Hugh Dickins authored
Recent commit 71725ed1 ("mm: huge tmpfs: try to split_huge_page() when punching hole") has allowed syzkaller to probe deeper, uncovering a long-standing lockdep issue between the irq-unsafe shmlock_user_lock, the irq-safe xa_lock on mapping->i_pages, and shmem inode's info->lock which nests inside xa_lock (or tree_lock) since 4.8's shmem_uncharge(). user_shm_lock(), servicing SysV shmctl(SHM_LOCK), wants shmlock_user_lock while its caller shmem_lock() holds info->lock with interrupts disabled; but hugetlbfs_file_setup() calls user_shm_lock() with interrupts enabled, and might be interrupted by a writeback endio wanting xa_lock on i_pages. This may not risk an actual deadlock, since shmem inodes do not take part in writeback accounting, but there are several easy ways to avoid it. Requiring interrupts disabled for shmlock_user_lock would be easy, but it's a high-level global lock for which that seems inappropriate. Instead, recall that the use of info->lock to guard info->flags in shmem_lock() dates from pre-3.1 days, when races with SHMEM_PAGEIN and SHMEM_TRUNCATE could occur: nowadays it serves no purpose, the only flag added or removed is VM_LOCKED itself, and calls to shmem_lock() an inode are already serialized by the caller. Take info->lock out of the chain and the possibility of deadlock or lockdep warning goes away. Fixes: 4595ef88 ("shmem: make shmem_inode_info::lock irq-safe") Reported-by: syzbot+c8a8197c8852f566b9d9@syzkaller.appspotmail.com Reported-by: syzbot+40b71e145e73f78f81ad@syzkaller.appspotmail.com Signed-off-by:
-
Jann Horn authored
remap_vmalloc_range() has had various issues with the bounds checks it promises to perform ("This function checks that addr is a valid vmalloc'ed area, and that it is big enough to cover the vma") over time, e.g.: - not detecting pgoff<<PAGE_SHIFT overflow - not detecting (pgoff<<PAGE_SHIFT)+usize overflow - not checking whether addr and addr+(pgoff<<PAGE_SHIFT) are the same vmalloc allocation - comparing a potentially wildly out-of-bounds pointer with the end of the vmalloc region In particular, since commit fc970227 ("bpf: Add mmap() support for BPF_MAP_TYPE_ARRAY"), unprivileged users can cause kernel null pointer dereferences by calling mmap() on a BPF map with a size that is bigger than the distance from the start of the BPF map to the end of the address space. This could theoretically be used as a kernel ASLR bypass, by using whether mmap() with a given offset oopses or returns an error code to perform a binary search over the possible address range. To allow remap_vmalloc_range_partial() to verify that addr and addr+(pgoff<<PAGE_SHIFT) are in the same vmalloc region, pass the offset to remap_vmalloc_range_partial() instead of adding it to the pointer in remap_vmalloc_range(). In remap_vmalloc_range_partial(), fix the check against get_vm_area_size() by using size comparisons instead of pointer comparisons, and add checks for pgoff. Fixes: 83342314 ("[PATCH] mm: introduce remap_vmalloc_range()") Signed-off-by:Jann Horn <jannh@google.com> Signed-off-by:
-
