mirror of
https://github.com/torvalds/linux.git
synced 2024-12-15 07:33:56 +00:00
685439a7a0
77 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
Florian Fischer
|
b03b89b350 |
perf stat: Add user_time and system_time events
It bothered me that during benchmarking using 'perf stat' (to collect
for example CPU cache events) I could not simultaneously retrieve the
times spend in user or kernel mode in a machine readable format.
When running 'perf stat' the output for humans contains the times
reported by rusage and wait4.
$ perf stat -e cache-misses:u -- true
Performance counter stats for 'true':
4,206 cache-misses:u
0.001113619 seconds time elapsed
0.001175000 seconds user
0.000000000 seconds sys
But 'perf stat's machine-readable format does not provide this information.
$ perf stat -x, -e cache-misses:u -- true
4282,,cache-misses:u,492859,100.00,,
I found no way to retrieve this information using the available events
while using machine-readable output.
This patch adds two new tool internal events 'user_time' and
'system_time', similarly to the already present 'duration_time' event.
Both events use the already collected rusage information obtained by
wait4 and tracked in the global ru_stats.
Examples presenting cache-misses and rusage information in both human
and machine-readable form:
$ perf stat -e duration_time,user_time,system_time,cache-misses -- grep -q -r duration_time .
Performance counter stats for 'grep -q -r duration_time .':
67,422,542 ns duration_time:u
50,517,000 ns user_time:u
16,839,000 ns system_time:u
30,937 cache-misses:u
0.067422542 seconds time elapsed
0.050517000 seconds user
0.016839000 seconds sys
$ perf stat -x, -e duration_time,user_time,system_time,cache-misses -- grep -q -r duration_time .
72134524,ns,duration_time:u,72134524,100.00,,
65225000,ns,user_time:u,65225000,100.00,,
6865000,ns,system_time:u,6865000,100.00,,
38705,,cache-misses:u,71189328,100.00,,
Signed-off-by: Florian Fischer <florian.fischer@muhq.space>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Xing Zhengjun <zhengjun.xing@linux.intel.com>
Link: https://lore.kernel.org/r/20220420102354.468173-3-florian.fischer@muhq.space
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
John Garry
|
864bc8c905 |
perf parse-events: Support event alias in form foo-bar-baz
Event aliasing for events whose name in the form foo-bar-baz is not
supported, while foo-bar, foo_bar_baz, and other combinations are, i.e.
two hyphens are not supported.
The HiSilicon D06 platform has events in such form:
$ ./perf list sdir-home-migrate
List of pre-defined events (to be used in -e):
uncore hha:
sdir-home-migrate
[Unit: hisi_sccl,hha]
$ sudo ./perf stat -e sdir-home-migrate
event syntax error: 'sdir-home-migrate'
\___ parser error
Run 'perf list' for a list of valid events
Usage: perf stat [<options>] [<command>]
-e, --event <event>event selector. use 'perf list' to list available events
To support, add an extra PMU event symbol type for "baz", and add a new
rule in the bison file.
Signed-off-by: John Garry <john.garry@huawei.com>
Acked-by: Ian Rogers <irogers@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Qi Liu <liuqi115@huawei.com>
Cc: Shaokun Zhang <zhangshaokun@hisilicon.com>
Cc: linuxarm@huawei.com
Link: https://lore.kernel.org/r/1642432215-234089-2-git-send-email-john.garry@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
Ian Rogers
|
ec5c5b3d2c |
perf metric: Encode and use metric-id as qualifier
For a metric like IPC a group of events like {instructions,cycles}:W
would be formed.
If the events names were changed in parsing then the metric expression
parser would fail to find them.
This change makes the event encoding be something like:
{instructions/metric-id=instructions/, cycles/metric-id=cycles/}
and then uses the evsel's stable metric-id value to locate the events.
This fixes the case that an event is restricted to user because of the
paranoia setting:
$ echo 2 > /proc/sys/kernel/perf_event_paranoid
$ perf stat -M IPC /bin/true
Performance counter stats for '/bin/true':
150,298 inst_retired.any:u # 0.77 IPC
187,095 cpu_clk_unhalted.thread:u
0.002042731 seconds time elapsed
0.000000000 seconds user
0.002377000 seconds sys
Adding the metric-id as a qualifier has a complication in that
qualifiers will become embedded in qualifiers.
For example, msr/tsc/ could become msr/tsc,metric-id=msr/tsc// which
will fail parse-events.
To solve this problem the metric is encoded and decoded for the
metric-id with !<num> standing in for an encoded value.
Previously ! wasn't parsed.
With this msr/tsc/ becomes msr/tsc,metric-id=msr!3tsc!3/
The metric expression parser is changed so that @ isn't changed to /,
instead this is done when the ID is encoded for parse events.
metricgroup__add_metric_non_group() and metricgroup__add_metric_weak_group()
need to inject the metric-id qualifier, so to avoid repetition they are
merged into a single metricgroup__build_event_string with error codes
more rigorously checked.
stat-shadow's prepare_metric() uses the metric-id to match the metricgroup
code.
As "metric-id=..." is added to all events, it is adding during testing
with the fake PMU.
This complicates pmu_str_check code as PE_PMU_EVENT_FAKE won't match as
part of a configuration.
The testing fake PMU case is fixed so that if a known qualifier with an
! is parsed then it isn't reported as a fake PMU.
This is sufficient to pass all testing but it and the original mechanism
are somewhat brittle.
Signed-off-by: Ian Rogers <irogers@google.com>
Acked-by: Andi Kleen <ak@linux.intel.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Antonov <alexander.antonov@linux.intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andrew Kilroy <andrew.kilroy@arm.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Changbin Du <changbin.du@intel.com>
Cc: Denys Zagorui <dzagorui@cisco.com>
Cc: Fabian Hemmer <copy@copy.sh>
Cc: Felix Fietkau <nbd@nbd.name>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jacob Keller <jacob.e.keller@intel.com>
Cc: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Joakim Zhang <qiangqing.zhang@nxp.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Kees Kook <keescook@chromium.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nicholas Fraser <nfraser@codeweavers.com>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Riccardo Mancini <rickyman7@gmail.com>
Cc: Sami Tolvanen <samitolvanen@google.com>
Cc: ShihCheng Tu <mrtoastcheng@gmail.com>
Cc: Song Liu <songliubraving@fb.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Sumanth Korikkar <sumanthk@linux.ibm.com>
Cc: Thomas Richter <tmricht@linux.ibm.com>
Cc: Wan Jiabing <wanjiabing@vivo.com>
Cc: Zhen Lei <thunder.leizhen@huawei.com>
Link: https://lore.kernel.org/r/20211015172132.1162559-17-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Ian Rogers
|
2b62b3a611 |
perf parse-events: Add new "metric-id" term
Add a new "metric-id" term to events so that metric parsing can set an
ID that can be reliably looked up.
Metric parsing currently will turn a metric like "instructions/cycles"
into a parse events string of "{instructions,cycles}:W".
However, parse-events may change "instructions" into "instructions:u" if
perf_event_paranoid=2.
When this happens expr__resolve_id currently fails as stat-shadow adds
the ID "instructions:u" to match with the counter value and the metric
tries to look up the ID just "instructions".
A later patch will use the new term.
An example of the current problem:
$ echo -1 > /proc/sys/kernel/perf_event_paranoid
$ perf stat -M IPC /bin/true
Performance counter stats for '/bin/true':
1,217,161 inst_retired.any # 0.97 IPC
1,250,389 cpu_clk_unhalted.thread
0.002064773 seconds time elapsed
0.002378000 seconds user
0.000000000 seconds sys
$ echo 2 > /proc/sys/kernel/perf_event_paranoid
$ perf stat -M IPC /bin/true
Performance counter stats for '/bin/true':
150,298 inst_retired.any:u # nan IPC
187,095 cpu_clk_unhalted.thread:u
0.002042731 seconds time elapsed
0.000000000 seconds user
0.002377000 seconds sys
Note: nan IPC is printed as an effect of "perf metric: Use NAN for
missing event IDs." but earlier versions of perf just fail with a parse
error and display no value.
Signed-off-by: Ian Rogers <irogers@google.com>
Acked-by: Andi Kleen <ak@linux.intel.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Antonov <alexander.antonov@linux.intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andrew Kilroy <andrew.kilroy@arm.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Changbin Du <changbin.du@intel.com>
Cc: Denys Zagorui <dzagorui@cisco.com>
Cc: Fabian Hemmer <copy@copy.sh>
Cc: Felix Fietkau <nbd@nbd.name>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jacob Keller <jacob.e.keller@intel.com>
Cc: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Joakim Zhang <qiangqing.zhang@nxp.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Kees Kook <keescook@chromium.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nicholas Fraser <nfraser@codeweavers.com>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Riccardo Mancini <rickyman7@gmail.com>
Cc: Sami Tolvanen <samitolvanen@google.com>
Cc: ShihCheng Tu <mrtoastcheng@gmail.com>
Cc: Song Liu <songliubraving@fb.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Sumanth Korikkar <sumanthk@linux.ibm.com>
Cc: Thomas Richter <tmricht@linux.ibm.com>
Cc: Wan Jiabing <wanjiabing@vivo.com>
Cc: Zhen Lei <thunder.leizhen@huawei.com>
Link: https://lore.kernel.org/r/20211015172132.1162559-15-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Ian Rogers
|
cb7bfb1da6 |
perf parse-events: Remove unnecessary #includes
Minor cleanup motivated by trying to separately fuzz test parse-events. Signed-off-by: Ian Rogers <irogers@google.com> Acked-by: Namhyung Kim <namhyung@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Leo Yan <leo.yan@linaro.org> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lore.kernel.org/lkml/20210127184629.516169-1-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |
||
Namhyung Kim
|
fb6c79d726 |
perf tools: Add 'cgroup-switches' software event
It counts how often cgroups are changed actually during the context
switches.
# perf stat -a -e context-switches,cgroup-switches -a sleep 1
Performance counter stats for 'system wide':
11,267 context-switches
10,950 cgroup-switches
1.015634369 seconds time elapsed
Committer notes:
The kernel patches landed in v5.13, but this entry wasn't filled in
perf's parse-events tables, which was leading to a segfault when running
'perf list' on a kernel with that feature, as reported by Thomas
Richter.
Also removed the part touching tools/include/uapi/linux/perf_event.h as
it was updated in the usual sync with the kernel UAPI headers, in a
previous, already upstream, patch.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Richter <tmricht@linux.ibm.com>
Link: http://lore.kernel.org/lkml/20210210083327.22726-3-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
Song Liu
|
01bd8efcec |
perf stat: Introduce ':b' modifier
Introduce 'b' modifier to event parser, which means use BPF program to manage this event. This is the same as --bpf-counters option, but only applies to this event. For example, perf stat -e cycles:b,cs # use bpf for cycles, but not cs perf stat -e cycles,cs --bpf-counters # use bpf for both cycles and cs Suggested-by: Jiri Olsa <jolsa@kernel.org> Signed-off-by: Song Liu <song@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Link: https://lore.kernel.org/r/20210425214333.1090950-5-song@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |
||
Kan Liang
|
2a57d40832 |
perf tools: Support the auxiliary event
On the Intel Sapphire Rapids server, an auxiliary event has to be
enabled simultaneously with the load latency event to retrieve complete
Memory Info.
Add X86 specific perf_mem_events__name() to handle the auxiliary event.
- Users are only interested in the samples of the mem-loads event.
Sample read the auxiliary event.
- The auxiliary event must be in front of the load latency event in a
group. Assume the second event to sample if the auxiliary event is the
leader.
- Add a weak is_mem_loads_aux_event() to check the auxiliary event for
X86. For other ARCHs, it always return false.
Parse the unique event name, mem-loads-aux, for the auxiliary event.
Committer notes:
According to
|
||
Andi Kleen
|
0997a2662f |
perf tools: Add support for exclusive groups/events
Peter suggested that using the exclusive mode in perf could avoid some
problems with bad scheduling of groups. Exclusive is implemented in the
kernel, but wasn't exposed by the perf tool, so hard to use without
custom low level API users.
Add support for marking groups or events with :e for exclusive in the
perf tool. The implementation is basically the same as the existing
pinned attribute.
Committer testing:
# perf test "parse event"
6: Parse event definition strings : Ok
# perf test -v "parse event" |& grep :u*e
running test 56 'instructions:uep'
running test 57 '{cycles,cache-misses,branch-misses}:e'
#
#
# grep "model name" -m1 /proc/cpuinfo
model name : AMD Ryzen 9 3900X 12-Core Processor
#
# perf stat -a -e '{cycles,cache-misses,branch-misses}:e' sleep 1
Performance counter stats for 'system wide':
<not counted> cycles (0.00%)
<not counted> cache-misses (0.00%)
<not counted> branch-misses (0.00%)
1.001269893 seconds time elapsed
Some events weren't counted. Try disabling the NMI watchdog:
echo 0 > /proc/sys/kernel/nmi_watchdog
perf stat ...
echo 1 > /proc/sys/kernel/nmi_watchdog
# echo 0 > /proc/sys/kernel/nmi_watchdog
# perf stat -a -e '{cycles,cache-misses,branch-misses}:e' sleep 1
Performance counter stats for 'system wide':
1,298,663,141 cycles
30,962,215 cache-misses
5,325,150 branch-misses
1.001474934 seconds time elapsed
#
# The output for asking for precise events on AMD needs to improve, it
# supposedly works only for system wide or per CPU
#
# perf stat -a -e '{cycles,cache-misses,branch-misses}:uep' sleep 1
Error:
The sys_perf_event_open() syscall returned with 22 (Invalid argument) for event (cycles).
/bin/dmesg | grep -i perf may provide additional information.
# perf stat -a -e '{cycles,cache-misses,branch-misses}:ue' sleep 1
Performance counter stats for 'system wide':
746,363,126 cycles
16,881,611 cache-misses
2,871,259 branch-misses
1.001636066 seconds time elapsed
#
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20201014144255.22699-1-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
Jiri Olsa
|
4929e95a14 |
perf tools: Fix term parsing for raw syntax
Jin Yao reported issue with possible conflict between raw events and
term values in pmu event syntax.
Currently following syntax is resolved as raw event with 0xead value:
uncore_imc_free_running/read/
instead of using 'read' term from uncore_imc_free_running pmu, because
'read' is correct raw event syntax with 0xead value.
To solve this issue we do following:
- check existing terms during rXXXX syntax processing
and make them priority in case of conflict
- allow pmu/r0x1234/ syntax to be able to specify conflicting
raw event (implemented in previous patch)
Also add automated tests for this and perf_pmu__parse_cleanup call to
parse_events_terms, so the test gets properly cleaned up.
Fixes:
|
||
|
Jiri Olsa
|
c33cdf5411 |
perf tools: Allow r0x<HEX> event syntax
Add support to specify raw event with 'r0<HEX>' syntax within pmu term syntax like: -e cpu/r0xdead/ It will be used to specify raw events in cases where they conflict with real pmu terms, like 'read', which is valid raw event syntax, but also a possible pmu term name as reported by Jin Yao. Reported-by: Jin Yao <yao.jin@linux.intel.com> Signed-off-by: Jiri Olsa <jolsa@kernel.org> Acked-by: Ian Rogers <irogers@google.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Michael Petlan <mpetlan@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Link: http://lore.kernel.org/lkml/20200725121959.1181869-1-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |
||
|
Jiri Olsa
|
387ad33fe7 |
perf tools: Add fake pmu support
Add a way to create a pmu event without the actual PMU being in place. This way we can test metrics defined for any processor. The interface is to define fake_pmu in struct parse_events_state data. It will be used only in tests via special interface function added in following changes. Signed-off-by: Jiri Olsa <jolsa@kernel.org> Acked-by: Ian Rogers <irogers@google.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Michael Petlan <mpetlan@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lore.kernel.org/lkml/20200602214741.1218986-2-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |
||
|
Jiri Olsa
|
1244a32736 |
perf parse: Add 'struct parse_events_state' pointer to scanner
We need to pass more data to the scanner so let's start with having it to take pointer to 'struct parse_events_state' object instead of just start token. Signed-off-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Michael Petlan <mpetlan@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: http://lore.kernel.org/lkml/20200524224219.234847-4-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |
||
|
Jiri Olsa
|
ae7626418d |
perf stat: Fail on extra comma while parsing events
Ian reported that we allow to parse following:
$ perf stat -e ,cycles true
which is wrong and we should fail, like we do with this fix:
$ perf stat -e ,cycles true
event syntax error: ',cycles'
\___ parser error
The reason is that we don't have rule for ',' in 'event' start condition
and it's matched and accepted by default rule.
Add scanner debug support (that Ian already added for expr code),
which was really useful for finding this. It's enabled together with
bison debug via 'make PARSER_DEBUG=1'.
Reported-by: Ian Rogers <irogers@google.com>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Ian Rogers <irogers@google.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Michael Petlan <mpetlan@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20200520074050.156988-1-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
Jiri Olsa
|
3a6c51e4d6 |
perf parser: Add support to specify rXXX event with pmu
The current rXXXX event specification creates event under PERF_TYPE_RAW
pmu type. This change allows to use rXXXX within pmu syntax, so it's
type is used via the following syntax:
-e 'cpu/r3c/'
-e 'cpum_cf/r0/'
The XXXX number goes directly to perf_event_attr::config the same way as
in '-e rXXXX' event. The perf_event_attr::type is filled with pmu type.
Committer testing:
So, lets see what goes in perf_event_attr::config for, say, the
'instructions' PERF_TYPE_HARDWARE (0) event, first we should look at how
to encode this event as a PERF_TYPE_RAW event for this specific CPU, an
AMD Ryzen 5:
# cat /sys/devices/cpu/events/instructions
event=0xc0
#
Then try with it _and_ the instruction, just to see that they are close
enough:
# perf stat -e rc0,instructions sleep 1
Performance counter stats for 'sleep 1':
919,794 rc0
919,898 instructions
1.000754579 seconds time elapsed
0.000715000 seconds user
0.000000000 seconds sys
#
Now we should try, before this patch, the PMU event encoding:
# perf stat -e cpu/rc0/ sleep 1
event syntax error: 'cpu/rc0/'
\___ unknown term
valid terms: event,edge,inv,umask,cmask,config,config1,config2,name,period,percore
#
Now with this patch, the three ways of specifying the 'instructions' CPU
counter are accepted:
# perf stat -e cpu/rc0/,rc0,instructions sleep 1
Performance counter stats for 'sleep 1':
892,948 cpu/rc0/
893,052 rc0
893,156 instructions
1.000931819 seconds time elapsed
0.000916000 seconds user
0.000000000 seconds sys
#
Requested-by: Thomas Richter <tmricht@linux.ibm.com>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Tested-by: Thomas Richter <tmricht@linux.ibm.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Michael Petlan <mpetlan@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Sumanth Korikkar <sumanthk@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Link: http://lore.kernel.org/lkml/20200416221405.437788-1-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
Adrian Hunter
|
47327f5667 |
perf events parser: Add missing Intel CPU events to parser
perf list expects CPU events to be parseable by name, e.g.
# perf list | grep el-capacity-read
el-capacity-read OR cpu/el-capacity-read/ [Kernel PMU event]
But the event parser does not recognize them that way, e.g.
# perf test -v "Parse event"
<SNIP>
running test 54 'cycles//u'
running test 55 'cycles:k'
running test 0 'cpu/config=10,config1,config2=3,period=1000/u'
running test 1 'cpu/config=1,name=krava/u,cpu/config=2/u'
running test 2 'cpu/config=1,call-graph=fp,time,period=100000/,cpu/config=2,call-graph=no,time=0,period=2000/'
running test 3 'cpu/name='COMPLEX_CYCLES_NAME:orig=cycles,desc=chip-clock-ticks',period=0x1,event=0x2/ukp'
-> cpu/event=0,umask=0x11/
-> cpu/event=0,umask=0x13/
-> cpu/event=0x54,umask=0x1/
failed to parse event 'el-capacity-read:u,cpu/event=el-capacity-read/u', err 1, str 'parser error'
event syntax error: 'el-capacity-read:u,cpu/event=el-capacity-read/u'
\___ parser error test child finished with 1
---- end ----
Parse event definition strings: FAILED!
This happens because the parser splits names by '-' in order to deal
with cache events. For example 'L1-dcache' is a token in
parse-events.l which is matched to 'L1-dcache-load-miss' by the
following rule:
PE_NAME_CACHE_TYPE '-' PE_NAME_CACHE_OP_RESULT '-' PE_NAME_CACHE_OP_RESULT opt_event_config
And so there is special handling for 2-part PMU names i.e.
PE_PMU_EVENT_PRE '-' PE_PMU_EVENT_SUF sep_dc
but no handling for 3-part names, which are instead added as tokens e.g.
topdown-[a-z-]+
While it would be possible to add a rule for 3-part names, that would
not work if the first parts were also a valid PMU name e.g.
'el-capacity-read' would be matched to 'el-capacity' before the parser
reached the 3rd part.
The parser would need significant change to rationalize all this, so
instead fix for now by adding missing Intel CPU events with 3-part names
to the event parser as tokens.
Missing events were found by using:
grep -r EVENT_ATTR_STR arch/x86/events/intel/core.c
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Link: http://lore.kernel.org/lkml/90c7ae07-c568-b6d3-f9c4-d0c1528a0610@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Adrian Hunter
|
eb7a52d46c |
perf record: Add aux-sample-size config term
To allow individual events to be selected for AUX area sampling, add aux-sample-size config term. attr.aux_sample_size is updated by auxtrace_parse_sample_options() so that the existing validation will see the value. Any event that has a non-zero aux_sample_size will cause AUX area sampling to be configured, irrespective of the --aux-sample option. Signed-off-by: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Link: http://lore.kernel.org/lkml/20191115124225.5247-8-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |
||
|
Adrian Hunter
|
1b9921546a |
perf tools: Add aux-output config term
Expose the aux_output attribute flag to the user to configure, by adding a config term 'aux-output'. For events that support it, selection of 'aux-output' causes the generation of AUX records instead of event records. This requires that an AUX area event is also provided. Signed-off-by: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20190806084606.4021-7-alexander.shishkin@linux.intel.com Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |
||
Jin Yao
|
064b4e82aa |
perf tools: Add a 'percore' event qualifier
Add a 'percore' event qualifier, like cpu/event=0,umask=0x3,percore=1/,
that sums up the event counts for both hardware threads in a core.
We can already do this with --per-core, but it's often useful to do
this together with other metrics that are collected per hardware thread.
So we need to support this per-core counting on a event level.
This can be implemented in only the user tool, no kernel support needed.
v4:
---
1. Add Arnaldo's patch which updates the documentation for
this new qualifier.
2. Rebase to latest perf/core branch
v3:
---
Simplify the code according to Jiri's comments.
Before:
"return term->val.percore ? true : false;"
Now:
"return term->val.percore;"
v2:
---
Change the qualifier name from 'coresum' to 'percore' according to
comments from Jiri and Andi.
Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
Tested-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jin Yao <yao.jin@intel.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1555077590-27664-2-git-send-email-yao.jin@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
Andi Kleen
|
f0fbb114e3 |
perf stat: Implement duration_time as a proper event
The perf metric expression use 'duration_time' internally to normalize
events. Normal 'perf stat' without -x also prints the duration time.
But when using -x, the interval is not output anywhere, which is
inconvenient for any post processing which often wants to normalize
values to time.
So implement 'duration_time' as a proper perf event that can be
specified explicitely with -e.
The previous implementation of 'duration_time' only worked for metric
processing. This adds the concept of a tool event that is handled by the
tool. On the kernel level it is still mapped to the dummy software
event, but the values are not read anymore, but instead computed by the
tool.
Add proper plumbing to handle this in the event parser, and display it
in 'perf stat'. We don't want 'duration_time' to be added up, so it's
only printed for the first CPU.
% perf stat -e duration_time,cycles true
Performance counter stats for 'true':
555,476 ns duration_time
771,958 cycles
0.000555476 seconds time elapsed
0.000644000 seconds user
0.000000000 seconds sys
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/20190326221823.11518-3-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
Arnaldo Carvalho de Melo
|
2fda5ada07 |
perf evsel: Introduce per event max_events property
This simply adds the field to 'struct perf_evsel' and allows setting
it via the event parser, to test it lets trace trace:
First look at where in a function that receives an evsel we can put a probe
to read how evsel->max_events was setup:
# perf probe -x ~/bin/perf -L trace__event_handler
<trace__event_handler@/home/acme/git/perf/tools/perf/builtin-trace.c:0>
0 static int trace__event_handler(struct trace *trace, struct perf_evsel *evsel,
union perf_event *event __maybe_unused,
struct perf_sample *sample)
3 {
4 struct thread *thread = machine__findnew_thread(trace->host, sample->pid, sample->tid);
5 int callchain_ret = 0;
7 if (sample->callchain) {
8 callchain_ret = trace__resolve_callchain(trace, evsel, sample, &callchain_cursor);
9 if (callchain_ret == 0) {
10 if (callchain_cursor.nr < trace->min_stack)
11 goto out;
12 callchain_ret = 1;
}
}
See what variables we can probe at line 7:
# perf probe -x ~/bin/perf -V trace__event_handler:7
Available variables at trace__event_handler:7
@<trace__event_handler+89>
int callchain_ret
struct perf_evsel* evsel
struct perf_sample* sample
struct thread* thread
struct trace* trace
union perf_event* event
Add a probe at that line asking for evsel->max_events to be collected and named
as "max_events":
# perf probe -x ~/bin/perf trace__event_handler:7 'max_events=evsel->max_events'
Added new event:
probe_perf:trace__event_handler (on trace__event_handler:7 in /home/acme/bin/perf with max_events=evsel->max_events)
You can now use it in all perf tools, such as:
perf record -e probe_perf:trace__event_handler -aR sleep 1
Now use 'perf trace', here aliased to just 'trace' and trace trace, i.e.
the first 'trace' is tracing just that 'probe_perf:trace__event_handler' event,
while the traced trace is tracing all scheduler tracepoints, will stop at two
events (--max-events 2) and will just set evsel->max_events for all the sched
tracepoints to 9, we will see the output of both traces intermixed:
# trace -e *perf:*event_handler trace --max-events 2 -e sched:*/nr=9/
0.000 :0/0 sched:sched_waking:comm=rcu_sched pid=10 prio=120 target_cpu=000
0.009 :0/0 sched:sched_wakeup:comm=rcu_sched pid=10 prio=120 target_cpu=000
0.000 trace/23949 probe_perf:trace__event_handler:(48c34a) max_events=0x9
0.046 trace/23949 probe_perf:trace__event_handler:(48c34a) max_events=0x9
#
Now, if the traced trace sends its output to /dev/null, we'll see just
what the first level trace outputs: that evsel->max_events is indeed
being set to 9:
# trace -e *perf:*event_handler trace -o /dev/null --max-events 2 -e sched:*/nr=9/
0.000 trace/23961 probe_perf:trace__event_handler:(48c34a) max_events=0x9
0.030 trace/23961 probe_perf:trace__event_handler:(48c34a) max_events=0x9
#
Now that we can set evsel->max_events, we can go to the next step, honour that
per-event property in 'perf trace'.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Milian Wolff <milian.wolff@kdab.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-og00yasj276joem6e14l1eas@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
Alexey Budankov
|
f92da71280 |
perf record: Enable arbitrary event names thru name= modifier
Enable complex event names containing [.:=,] symbols to be encoded into Perf
trace using name= modifier e.g. like this:
perf record -e cpu/name=\'OFFCORE_RESPONSE:request=DEMAND_RFO:response=L3_HIT.SNOOP_HITM\',\
period=0x3567e0,event=0x3c,cmask=0x1/Duk ./futex
Below is how it looks like in the report output. Please note explicit escaped
quoting at cmdline string in the header so that thestring can be directly reused
for another collection in shell:
perf report --header
# ========
...
# cmdline : /root/abudanko/kernel/tip/tools/perf/perf record -v -e cpu/name=\'OFFCORE_RESPONSE:request=DEMAND_RFO:response=L3_HIT.SNOOP_HITM\',period=0x3567e0,event=0x3c,cmask=0x1/Duk ./futex
# event : name = OFFCORE_RESPONSE:request=DEMAND_RFO:response=L3_HIT.SNOOP_HITM, , type = 4, size = 112, config = 0x100003c, { sample_period, sample_freq } =
|
||
Agustin Vega-Frias
|
b2b9d3a3f0 |
perf pmu: Support wildcards on pmu name in dynamic pmu events
Starting on v4.12 event parsing code for dynamic pmu events already
supports prefix-based matching of multiple pmus when creating dynamic
events. E.g., in a system with the following dynamic pmus:
mypmu_0
mypmu_1
mypmu_2
mypmu_4
passing mypmu/<config>/ as an event spec will result in the creation of
the event in all of the pmus. This change expands this matching through
the use of fnmatch so glob-like expressions can be used to create events
in multiple pmus. E.g., in the system described above if a user only
wants to create the event in mypmu_0 and mypmu_1, mypmu_[01]/<config>/
can be passed.
Signed-off-by: Agustin Vega-Frias <agustinv@codeaurora.org>
Acked-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Timur Tabi <timur@codeaurora.org>
Change-Id: Icb25653fc5d5239c20f3bffdfdf4ab4c9c9bb20b
Link: http://lkml.kernel.org/r/1520454947-16977-1-git-send-email-agustinv@codeaurora.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
Linus Torvalds
|
31486372a1 |
Merge branch 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull perf updates from Ingo Molnar:
"The main changes in this cycle were:
Kernel:
- kprobes updates: use better W^X patterns for code modifications,
improve optprobes, remove jprobes. (Masami Hiramatsu, Kees Cook)
- core fixes: event timekeeping (enabled/running times statistics)
fixes, perf_event_read() locking fixes and cleanups, etc. (Peter
Zijlstra)
- Extend x86 Intel free-running PEBS support and support x86
user-register sampling in perf record and perf script. (Andi Kleen)
Tooling:
- Completely rework the way inline frames are handled. Instead of
querying for the inline nodes on-demand in the individual tools, we
now create proper callchain nodes for inlined frames. (Milian
Wolff)
- 'perf trace' updates (Arnaldo Carvalho de Melo)
- Implement a way to print formatted output to per-event files in
'perf script' to facilitate generate flamegraphs, elliminating the
need to write scripts to do that separation (yuzhoujian, Arnaldo
Carvalho de Melo)
- Update vendor events JSON metrics for Intel's Broadwell, Broadwell
Server, Haswell, Haswell Server, IvyBridge, IvyTown, JakeTown,
Sandy Bridge, Skylake, SkyLake Server - and Goldmont Plus V1 (Andi
Kleen, Kan Liang)
- Multithread the synthesizing of PERF_RECORD_ events for
pre-existing threads in 'perf top', speeding up that phase, greatly
improving the user experience in systems such as Intel's Knights
Mill (Kan Liang)
- Introduce the concept of weak groups in 'perf stat': try to set up
a group, but if it's not schedulable fallback to not using a group.
That gives us the best of both worlds: groups if they work, but
still a usable fallback if they don't. E.g: (Andi Kleen)
- perf sched timehist enhancements (David Ahern)
- ... various other enhancements, updates, cleanups and fixes"
* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (139 commits)
kprobes: Don't spam the build log with deprecation warnings
arm/kprobes: Remove jprobe test case
arm/kprobes: Fix kretprobe test to check correct counter
perf srcline: Show correct function name for srcline of callchains
perf srcline: Fix memory leak in addr2inlines()
perf trace beauty kcmp: Beautify arguments
perf trace beauty: Implement pid_fd beautifier
tools include uapi: Grab a copy of linux/kcmp.h
perf callchain: Fix double mapping al->addr for children without self period
perf stat: Make --per-thread update shadow stats to show metrics
perf stat: Move the shadow stats scale computation in perf_stat__update_shadow_stats
perf tools: Add perf_data_file__write function
perf tools: Add struct perf_data_file
perf tools: Rename struct perf_data_file to perf_data
perf script: Print information about per-event-dump files
perf trace beauty prctl: Generate 'option' string table from kernel headers
tools include uapi: Grab a copy of linux/prctl.h
perf script: Allow creating per-event dump files
perf evsel: Restore evsel->priv as a tool private area
perf script: Use event_format__fprintf()
...
|
||
|
Jiri Olsa
|
a271bfaf30 |
perf tools: Fix eBPF event specification parsing
Looks like I've reached the new level of stupidity, adding missing braces.
Committer testing:
Given the following eBPF C filter, that will add a record when it
returns true, i.e. when the tv_nsec variable is > 2000ns, should be
built and installed via sys_bpf(), but fails to do so before this patch:
# cat filter.c
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
SEC("func=hrtimer_nanosleep rqtp->tv_nsec")
int func(void *ctx, int err, long nsec)
{
return nsec > 1000;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
#
# perf trace -e nanosleep,filter.c usleep 1
invalid or unsupported event: 'filter.c'
Run 'perf list' for a list of valid events
Usage: perf trace [<options>] [<command>]
or: perf trace [<options>] -- <command> [<options>]
or: perf trace record [<options>] [<command>]
or: perf trace record [<options>] -- <command> [<options>]
-e, --event <event> event/syscall selector. use 'perf list' to list available events
#
And works again after it is applied, the nothing is inserted when the co
# perf trace -e *sleep,filter.c usleep 1
0.000 ( 0.066 ms): usleep/23994 nanosleep(rqtp: 0x7ffead94a0d0) = 0
# perf trace -e *sleep,filter.c usleep 2
0.000 ( 0.008 ms): usleep/24378 nanosleep(rqtp: 0x7fffa021ba50) ...
0.008 ( ): perf_bpf_probe:func:(ffffffffb410cb30) tv_nsec=2000)
0.000 ( 0.066 ms): usleep/24378 ... [continued]: nanosleep()) = 0
#
The intent of
|
||
|
Jiri Olsa
|
b6af53b7d6 |
perf tools: Add "reject" option for parse-events.l
Arnaldo reported broken builds in some distros using a newer flex release, 2.6.4, found in Alpine Linux 3.6 and Edge, with flex not spotting the REJECT macro: CC /tmp/build/perf/util/parse-events-flex.o util/parse-events.l: In function 'parse_events_lex': /tmp/build/perf/util/parse-events-flex.c:4734:16: error: \ 'reject_used_but_not_detected' undeclared (first use in this function) It's happening because we put the REJECT under another USER_REJECT macro in following commit: |
||
Ingo Molnar
|
15bcdc9477 |
Merge branch 'linus' into perf/core, to fix conflicts
Conflicts: tools/perf/arch/arm/annotate/instructions.c tools/perf/arch/arm64/annotate/instructions.c tools/perf/arch/powerpc/annotate/instructions.c tools/perf/arch/s390/annotate/instructions.c tools/perf/arch/x86/tests/intel-cqm.c tools/perf/ui/tui/progress.c tools/perf/util/zlib.c Signed-off-by: Ingo Molnar <mingo@kernel.org> |
||
|
Jiri Olsa
|
9445464bb8 |
perf tools: Unwind properly location after REJECT
We have defined YY_USER_ACTION to keep trace of the column location
during events parsing, but we need to clean it up when we call REJECT.
When REJECT is called, the lexer shrinks the text and re-runs the
matching, so we need to address it in resuming the previous location
value to keep it correct for error display, like:
Before:
$ perf stat -e 'cpu/uops_executed.core,krava/' true
event syntax error: '..38;5;9:mi=01;05;37;41:su=48;5;196;38;5;15:sg=48;5;1\
1;38;5;16:ca=48;5;196;38;5;226:tw=48;5;10;38;5;16:ow=48;5;10;38;5;21:st=48;5;\
21;38;50
�'
\___ unknown term
After:
$ ./perf stat -e 'cpu/uops_executed.core,krava/' true
event syntax error: '..cuted.core,krava/'
\___ unknown term
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Reported-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Tested-by: Andi Kleen <ak@linux.intel.com>
Cc: Changbin Du <changbin.du@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-vug2hchlny30jfsfrumbym26@git.kernel.org
Link: http://lkml.kernel.org/r/20171009140944.GD28623@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Ingo Molnar
|
ca4b9c3b74 |
Merge branch 'perf/urgent' into perf/core, to pick up fixes
Signed-off-by: Ingo Molnar <mingo@kernel.org> |
||
|
Jiri Olsa
|
29479bfe83 |
perf tools: Check wether the eBPF file exists in event parsing
Adding the check wether the eBPF file exists, to consider it
as eBPF input file. This way we can differentiate eBPF events
from events that end up with same suffix as eBPF file.
Before:
$ perf stat -e 'cpu/uops_executed.core/' true
bpf: builtin compilation failed: -95, try external compiler
WARNING: unable to get correct kernel building directory.
Hint: Set correct kbuild directory using 'kbuild-dir' option in [llvm]
section of ~/.perfconfig or set it to "" to suppress kbuild
detection.
event syntax error: 'cpu/uops_executed.core/'
\___ Failed to load cpu/uops_executed.c from source: 'version' section incorrect or lost
After:
$ perf stat -e 'cpu/uops_executed.core/' true
Performance counter stats for 'true':
181,533 cpu/uops_executed.core/:u
0.002795447 seconds time elapsed
If user makes type in the eBPF file, we prioritize the event syntax
and show following warning:
$ perf stat -e 'krava.c//' true
event syntax error: 'krava.c//'
\___ Cannot find PMU `krava.c'. Missing kernel support?
Reported-and-Tested-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Changbin Du <changbin.du@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/20171013083736.15037-9-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Andi Kleen
|
fd48aad9b0 |
perf stat: Support duration_time for metrics
Some of the metrics formulas (like GFLOPs) need to know how long the measurement period is. Support an internal event called duration_time, which reports time in second. It maps to the dummy event, but is special cased for statistics to report the walltime duration. So far it is not printed, but only used internally for metrics. Signed-off-by: Andi Kleen <ak@linux.intel.com> Acked-by: Jiri Olsa <jolsa@kernel.org> Link: http://lkml.kernel.org/r/20170831194036.30146-10-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |
||
|
Andi Kleen
|
5a5dfe4b85 |
perf tools: Support weak groups in 'perf stat'
Setting up groups can be complicated due to the complicated scheduling
restrictions of different PMUs.
User tools usually don't understand all these restrictions.
Still in many cases it is useful to set up groups and they work most of
the time. However if the group is set up wrong some members will not
report any value because they never get scheduled.
Add a concept of a 'weak group': try to set up a group, but if it's not
schedulable fallback to not using a group. That gives us the best of
both worlds: groups if they work, but still a usable fallback if they
don't.
In theory it would be possible to have more complex fallback strategies
(e.g. try to split the group in half), but the simple fallback of not
using a group seems to work for now.
So far the weak group is only implemented for perf stat, not for record.
Here's an unschedulable group (on IvyBridge with SMT on)
% perf stat -e '{branches,branch-misses,l1d.replacement,l2_lines_in.all,l2_rqsts.all_code_rd}' -a sleep 1
73,806,067 branches
4,848,144 branch-misses # 6.57% of all branches
14,754,458 l1d.replacement
24,905,558 l2_lines_in.all
<not supported> l2_rqsts.all_code_rd <------- will never report anything
With the weak group:
% perf stat -e '{branches,branch-misses,l1d.replacement,l2_lines_in.all,l2_rqsts.all_code_rd}:W' -a sleep 1
125,366,055 branches (80.02%)
9,208,402 branch-misses # 7.35% of all branches (80.01%)
24,560,249 l1d.replacement (80.00%)
43,174,971 l2_lines_in.all (80.05%)
31,891,457 l2_rqsts.all_code_rd (79.92%)
The extra event scheduled with some extra multiplexing
v2: Move fallback code to separate function.
Add comment on for_each_group_member
Adjust to new perf_evsel__close interface
v3: Fix debug print out.
Committer testing:
Before:
# perf stat -e '{branches,branch-misses,l1d.replacement,l2_lines_in.all,l2_rqsts.all_code_rd}' -a sleep 1
Performance counter stats for 'system wide':
<not counted> branches
<not counted> branch-misses
<not counted> l1d.replacement
<not counted> l2_lines_in.all
<not supported> l2_rqsts.all_code_rd
1.002147212 seconds time elapsed
# perf stat -e '{branches,l1d.replacement,l2_lines_in.all,l2_rqsts.all_code_rd}' -a sleep 1
Performance counter stats for 'system wide':
83,207,892 branches
11,065,444 l1d.replacement
28,484,024 l2_lines_in.all
12,186,179 l2_rqsts.all_code_rd
1.001739493 seconds time elapsed
After:
# perf stat -e '{branches,branch-misses,l1d.replacement,l2_lines_in.all,l2_rqsts.all_code_rd}':W -a sleep 1
Performance counter stats for 'system wide':
543,323,909 branches (80.01%)
27,100,512 branch-misses # 4.99% of all branches (80.02%)
50,402,905 l1d.replacement (80.03%)
67,385,892 l2_lines_in.all (80.01%)
21,352,885 l2_rqsts.all_code_rd (79.94%)
1.001086658 seconds time elapsed
#
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Link: http://lkml.kernel.org/r/20170831194036.30146-2-andi@firstfloor.org
[ Add a "'perf stat' only, for now" comment in the man page, suggested by Jiri ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Andi Kleen
|
77d0871c76 |
perf bpf: Tighten detection of BPF events
perf stat -e cpu/uops_executed.core,cmask=1/
would be detected as a BPF source event because the .c matches the .c
source BPF pattern.
v2:
Originally I tried to use lex lookahead, but it doesn't seem to work.
This now extends the BPF pattern to match longer events, but then does
an extra check in the C code to reject BPF matches that do not end with
.c/.o/.obj
This uses REJECT, which makes the flex scanner slower, but that
shouldn't be a big problem for the perf events.
Committer testing:
# perf trace -e write -e /home/acme/bpf/tracepoint.c cat /etc/passwd > /dev/null
0.000 ( 0.006 ms): cat/18485 write(fd: 1, buf: 0x7f59eebe1000, count: 3494 ) ...
0.006 ( ): raw_syscalls:sys_enter:NR 1 (1, 7f59eebe1000, da6, 22, 7f59eebe0010, 0))
0.008 ( ): perf_bpf_probe:_write:(ffffffff9626b2c0))
0.000 ( 0.010 ms): cat/18485 ... [continued]: write()) = 3494
#
It continues doing what was expected, i.e. identifying
/home/acme/bpf/tracepoint.c as a BPF event and activates the clang
machinery to build an eBPF object and then uses sys_bpf() to hook it up
to the raw_syscalls:sys_enter tracepoint, etc.
Andi forgot to add Wang to the CC list, fix it.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/20170811232634.30465-4-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
Wang Nan
|
2d470b62fa |
perf jevents: Handle events including .c and .o
This patch helps with Sukadev's vendor event tree where such events can happen.
>From Andi Kleen:
Any event including a .c/.o/.bpf currently triggers BPF compilation or loading
and then an error. This can happen for some Intel vendor events, which cannot
be used.
This patch fixes this problem by forbidding BPF file patch containing '{', '}'
and ',', make sure flex consumes the leading '{', instead of matching it using
a BPF file path.
Tested result:
$ perf stat -e '{unc_p_clockticks,unc_p_power_state_occupancy.cores_c0}' -a -I 1000
invalid or unsupported event: '{unc_p_clockticks,unc_p_power_state_occupancy.cores_c0}'
Run 'perf list' for a list of valid events
(as expected, interperted as event)
$ perf stat -e 'aaa.c' -a -I 1000
ERROR: problems with path aaa.c: No such file or directory
(as expected, interpreted as BPF source)
$ perf stat -e 'aaa.ccc' -a -I 1000
invalid or unsupported event: 'aaa.ccc'
(as expected, interpreted as event)
$ perf stat -e '{aaa.c}' -a -I 1000
ERROR: problems with path aaa.c: No such file or directory
event syntax error: '{aaa.c}'
<SKIP>
(as expected, interpreted as BPF source)
$ perf stat -e '{cycles,aaa.c}' -a -I 1000
ERROR: problems with path aaa.c: No such file or directory
event syntax error: '{cycles,aaa.c}'
(as expected, interpreted as BPF source)
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Reported-by: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1475900185-37967-1-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
Mathieu Poirier
|
dd60fba732 |
perf tools: Add infrastructure for PMU specific configuration
This patch adds PMU driver specific configuration to the parser infrastructure by preceding any term with the '@' letter. As such doing something like: perf record -e some_event/@cfg1,@cfg2=config/ ... will see 'cfg1' and 'cfg2=config' being added to the list of evsel config terms. Token 'cfg1' and 'cfg2=config' are not processed in user space and are meant to be interpreted by the PMU driver. First the lexer/parser are supplemented with the required definitions to recognise the driver specific configuration. From there they are simply added to the list of event terms. The bulk of the work is done in function "parse_events_add_pmu()" where driver config event terms are added to a new list of driver config terms, which in turn spliced with the event's new driver configuration list. Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org> Acked-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1473179837-3293-4-git-send-email-mathieu.poirier@linaro.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |
||
|
Wang Nan
|
626a6b784e |
perf tools: Enable overwrite settings
This patch allows following config terms and option:
Globally setting events to overwrite;
# perf record --overwrite ...
Set specific events to be overwrite or no-overwrite.
# perf record --event cycles/overwrite/ ...
# perf record --event cycles/no-overwrite/ ...
Add missing config terms and update the config term array size because
the longest string length has changed.
For overwritable events, it automatically selects attr.write_backward
since perf requires it to be backward for reading.
Test result:
# perf record --overwrite -e syscalls:*enter_nanosleep* usleep 1
[ perf record: Woken up 2 times to write data ]
[ perf record: Captured and wrote 0.011 MB perf.data (1 samples) ]
# perf evlist -v
syscalls:sys_enter_nanosleep: type: 2, size: 112, config: 0x134, { sample_period, sample_freq }: 1, sample_type: IP|TID|TIME|CPU|PERIOD|RAW, disabled: 1, inherit: 1, mmap: 1, comm: 1, enable_on_exec: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, write_backward: 1
# Tip: use 'perf evlist --trace-fields' to show fields for tracepoint events
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-14-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Andi Kleen
|
44b1e60ab5 |
perf stat: Basic support for TopDown in perf stat
Add basic plumbing for TopDown in perf stat TopDown is intended to replace the frontend cycles idle/ backend cycles idle metrics in standard perf stat output. These metrics are not reliable in many workloads, due to out of order effects. This implements a new --topdown mode in perf stat (similar to --transaction) that measures the pipe line bottlenecks using standardized formulas. The measurement can be all done with 5 counters (one fixed counter) The result are four metrics: FrontendBound, BackendBound, BadSpeculation, Retiring that describe the CPU pipeline behavior on a high level. The full top down methology has many hierarchical metrics. This implementation only supports level 1 which can be collected without multiplexing. A full implementation of top down on top of perf is available in pmu-tools toplev. (http://github.com/andikleen/pmu-tools) The current version works on Intel Core CPUs starting with Sandy Bridge, and Atom CPUs starting with Silvermont. In principle the generic metrics should be also implementable on other out of order CPUs. TopDown level 1 uses a set of abstracted metrics which are generic to out of order CPU cores (although some CPUs may not implement all of them): topdown-total-slots Available slots in the pipeline topdown-slots-issued Slots issued into the pipeline topdown-slots-retired Slots successfully retired topdown-fetch-bubbles Pipeline gaps in the frontend topdown-recovery-bubbles Pipeline gaps during recovery from misspeculation These metrics then allow to compute four useful metrics: FrontendBound, BackendBound, Retiring, BadSpeculation. Add a new --topdown options to enable events. When --topdown is specified set up events for all topdown events supported by the kernel. Add topdown-* as a special case to the event parser, as is needed for all events containing -. The actual code to compute the metrics is in follow-on patches. v2: Use standard sysctl read function. v3: Move x86 specific code to arch/ v4: Enable --metric-only implicitly for topdown. v5: Add --single-thread option to not force per core mode v6: Fix output order of topdown metrics v7: Allow combining with -d v8: Remove --single-thread again v9: Rename functions, adding arch_ and topdown_. v10: Expand man page and describe TopDown better Paste intro into commit description. Print error when malloc fails. Signed-off-by: Andi Kleen <ak@linux.intel.com> Acked-by: Jiri Olsa <jolsa@kernel.org> Link: http://lkml.kernel.org/r/1464119559-17203-1-git-send-email-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |
||
|
Arnaldo Carvalho de Melo
|
792d48b4cf |
perf tools: Per event max-stack settings
The tooling counterpart, now it is possible to do:
# perf record -e sched:sched_switch/max-stack=10/ -e cycles/call-graph=dwarf,max-stack=4/ -e cpu-cycles/call-graph=dwarf,max-stack=1024/ usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.052 MB perf.data (5 samples) ]
# perf evlist -v
sched:sched_switch: type: 2, size: 112, config: 0x110, { sample_period, sample_freq }: 1, sample_type: IP|TID|TIME|CALLCHAIN|CPU|PERIOD|RAW|IDENTIFIER, read_format: ID, disabled: 1, inherit: 1, mmap: 1, comm: 1, enable_on_exec: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, sample_max_stack: 10
cycles/call-graph=dwarf,max-stack=4/: size: 112, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|CALLCHAIN|PERIOD|REGS_USER|STACK_USER|IDENTIFIER, read_format: ID, disabled: 1, inherit: 1, freq: 1, enable_on_exec: 1, sample_id_all: 1, exclude_guest: 1, exclude_callchain_user: 1, sample_regs_user: 0xff0fff, sample_stack_user: 8192, sample_max_stack: 4
cpu-cycles/call-graph=dwarf,max-stack=1024/: size: 112, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|CALLCHAIN|PERIOD|REGS_USER|STACK_USER|IDENTIFIER, read_format: ID, disabled: 1, inherit: 1, freq: 1, enable_on_exec: 1, sample_id_all: 1, exclude_guest: 1, exclude_callchain_user: 1, sample_regs_user: 0xff0fff, sample_stack_user: 8192, sample_max_stack: 1024
# Tip: use 'perf evlist --trace-fields' to show fields for tracepoint events
Using just /max-stack=N/ means /call-graph=fp,max-stack=N/, that should
be further configurable by means of some .perfconfig knob.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Milian Wolff <milian.wolff@kdab.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: Zefan Li <lizefan@huawei.com>
Link: http://lkml.kernel.org/n/tip-kolmn1yo40p7jhswxwrc7rrd@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Wang Nan
|
03e0a7df3e |
perf tools: Introduce bpf-output event
Commit
|
||
|
Wang Nan
|
e571e029bd |
perf tools: Enable indices setting syntax for BPF map
This patch introduces a new syntax to perf event parser:
# perf record -e './test_bpf_map_3.c/map:channel.value[0,1,2,3...5]=101/' usleep 2
By utilizing the basic facilities in bpf-loader.c which allow setting
different slots in a BPF map separately, the newly introduced syntax
allows perf to control specific elements in a BPF map.
Test result:
# cat ./test_bpf_map_3.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(unsigned char),
.max_entries = 100,
};
SEC("func=hrtimer_nanosleep rqtp->tv_nsec")
int func(void *ctx, int err, long nsec)
{
char fmt[] = "%ld\n";
long usec = nsec * 0x10624dd3 >> 38; // nsec / 1000
int key = (int)usec;
unsigned char *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), (unsigned char)*pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
Normal case:
# echo "" > /sys/kernel/debug/tracing/trace
# ./perf record -e './test_bpf_map_3.c/map:channel.value[0,1,2,3...5]=101/' usleep 2
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep usleep
usleep-405 [004] d... 2745423.547822: : 101
# ./perf record -e './test_bpf_map_3.c/map:channel.value[0...9,20...29]=102,map:channel.value[10...19]=103/' usleep 3
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# ./perf record -e './test_bpf_map_3.c/map:channel.value[0...9,20...29]=102,map:channel.value[10...19]=103/' usleep 15
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
# cat /sys/kernel/debug/tracing/trace | grep usleep
usleep-405 [004] d... 2745423.547822: : 101
usleep-655 [006] d... 2745434.122814: : 102
usleep-904 [006] d... 2745439.916264: : 103
# ./perf record -e './test_bpf_map_3.c/map:channel.value[all]=104/' usleep 99
# cat /sys/kernel/debug/tracing/trace | grep usleep
usleep-405 [004] d... 2745423.547822: : 101
usleep-655 [006] d... 2745434.122814: : 102
usleep-904 [006] d... 2745439.916264: : 103
usleep-1537 [003] d... 2745538.053737: : 104
Error case:
# ./perf record -e './test_bpf_map_3.c/map:channel.value[10...1000]=104/' usleep 99
event syntax error: '..annel.value[10...1000]=104/'
\___ Index too large
Hint: Valid config terms:
map:[<arraymap>].value<indices>=[value]
map:[<eventmap>].event<indices>=[event]
where <indices> is something like [0,3...5] or [all]
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-9-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Wang Nan
|
a34f3be70c |
perf tools: Enable BPF object configure syntax
This patch adds the final step for BPF map configuration. A new syntax
is appended into parser so user can config BPF objects through '/' '/'
enclosed config terms.
After this patch, following syntax is available:
# perf record -e ./test_bpf_map_1.c/map:channel.value=10/ ...
It would takes effect after appling following commits.
Test result:
# cat ./test_bpf_map_1.c
/************************ BEGIN **************************/
#include <uapi/linux/bpf.h>
#define SEC(NAME) __attribute__((section(NAME), used))
struct bpf_map_def {
unsigned int type;
unsigned int key_size;
unsigned int value_size;
unsigned int max_entries;
};
static void *(*map_lookup_elem)(struct bpf_map_def *, void *) =
(void *)BPF_FUNC_map_lookup_elem;
static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
(void *)BPF_FUNC_trace_printk;
struct bpf_map_def SEC("maps") channel = {
.type = BPF_MAP_TYPE_ARRAY,
.key_size = sizeof(int),
.value_size = sizeof(int),
.max_entries = 1,
};
SEC("func=sys_nanosleep")
int func(void *ctx)
{
int key = 0;
char fmt[] = "%d\n";
int *pval = map_lookup_elem(&channel, &key);
if (!pval)
return 0;
trace_printk(fmt, sizeof(fmt), *pval);
return 0;
}
char _license[] SEC("license") = "GPL";
int _version SEC("version") = LINUX_VERSION_CODE;
/************************* END ***************************/
- Normal case:
# ./perf record -e './test_bpf_map_1.c/map:channel.value=10/' usleep 10
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.012 MB perf.data ]
- Error case:
# ./perf record -e './test_bpf_map_1.c/map:channel.value/' usleep 10
event syntax error: '..ps:channel:value/'
\___ Config value not set (missing '=')
Hint: Valid config term:
map:[<arraymap>]:value=[value]
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
# ./perf record -e './test_bpf_map_1.c/xmap:channel.value=10/' usleep 10
event syntax error: '..pf_map_1.c/xmap:channel.value=10/'
\___ Invalid object config option
[SNIP]
# ./perf record -e './test_bpf_map_1.c/map:xchannel.value=10/' usleep 10
event syntax error: '..p_1.c/map:xchannel.value=10/'
\___ Target map not exist
[SNIP]
# ./perf record -e './test_bpf_map_1.c/map:channel.xvalue=10/' usleep 10
event syntax error: '..ps:channel.xvalue=10/'
\___ Invalid object map config option
[SNIP]
# ./perf record -e './test_bpf_map_1.c/map:channel.value=x10/' usleep 10
event syntax error: '..nnel.value=x10/'
\___ Incorrect value type for map
[SNIP]
Change BPF_MAP_TYPE_ARRAY to '1' in test_bpf_map_1.c:
# ./perf record -e './test_bpf_map_1.c/map:channel.value=10/' usleep 10
event syntax error: '..ps:channel.value=10/'
\___ Can't use this config term to this type of map
Hint: Valid config term:
map:[<arraymap>].value=[value]
(add -v to see detail)
Signed-off-by: Wang Nan <wangnan0@huawei.com>
[for parser part]
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1456132275-98875-5-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Wang Nan
|
17cb5f84b8 |
perf tools: Create config_term_names array
config_term_names[] is introduced for future commits which will be able
to retrieve the config name through the config term.
Utilize this array in parse_events_formats_error_string() so the missing
'{,no-}inherit' terms are added.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1455882283-79592-10-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Wang Nan
|
d509db0473 |
perf tools: Compile scriptlets to BPF objects when passing '.c' to --event
This patch provides infrastructure for passing source files to --event
directly using:
# perf record --event bpf-file.c command
This patch does following works:
1) Allow passing '.c' file to '--event'. parse_events_load_bpf() is
expanded to allow caller tell it whether the passed file is source
file or object.
2) llvm__compile_bpf() is called to compile the '.c' file, the result
is saved into memory. Use bpf_object__open_buffer() to load the
in-memory object.
Introduces a bpf-script-example.c so we can manually test it:
# perf record --clang-opt "-DLINUX_VERSION_CODE=0x40200" --event ./bpf-script-example.c sleep 1
Note that '--clang-opt' must put before '--event'.
Futher patches will merge it into a testcase so can be tested automatically.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Alexei Starovoitov <ast@plumgrid.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: David Ahern <dsahern@gmail.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kaixu Xia <xiakaixu@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1444826502-49291-10-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Wang Nan
|
84c86ca12b |
perf tools: Enable passing bpf object file to --event
By introducing new rules in tools/perf/util/parse-events.[ly], this
patch enables 'perf record --event bpf_file.o' to select events by an
eBPF object file. It calls parse_events_load_bpf() to load that file,
which uses bpf__prepare_load() and finally calls bpf_object__open() for
the object files.
After applying this patch, commands like:
# perf record --event foo.o sleep
become possible.
However, at this point it is unable to link any useful things onto the
evsel list because the creating of probe points and BPF program
attaching have not been implemented. Before real events are possible to
be extracted, to avoid perf report error because of empty evsel list,
this patch link a dummy evsel. The dummy event related code will be
removed when probing and extracting code is ready.
Commiter notes:
Using it:
$ ls -la foo.o
ls: cannot access foo.o: No such file or directory
$ perf record --event foo.o sleep
libbpf: failed to open foo.o: No such file or directory
event syntax error: 'foo.o'
\___ BPF object file 'foo.o' is invalid
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
$ file /tmp/build/perf/perf.o
/tmp/build/perf/perf.o: ELF 64-bit LSB relocatable, x86-64, version 1 (SYSV), not stripped
$ perf record --event /tmp/build/perf/perf.o sleep
libbpf: /tmp/build/perf/perf.o is not an eBPF object file
event syntax error: '/tmp/build/perf/perf.o'
\___ BPF object file '/tmp/build/perf/perf.o' is invalid
(add -v to see detail)
Run 'perf list' for a list of valid events
Usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
$
$ file /tmp/foo.o
/tmp/foo.o: ELF 64-bit LSB relocatable, no machine, version 1 (SYSV), not stripped
$ perf record --event /tmp/foo.o sleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data ]
$ perf evlist
/tmp/foo.o
$ perf evlist -v
/tmp/foo.o: type: 1, size: 112, config: 0x9, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|PERIOD, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
$
So, type 1 is PERF_TYPE_SOFTWARE, config 0x9 is PERF_COUNT_SW_DUMMY, ok.
$ perf report --stdio
Error:
The perf.data file has no samples!
# To display the perf.data header info, please use --header/--header-only options.
#
$
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@plumgrid.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: David Ahern <dsahern@gmail.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kaixu Xia <xiakaixu@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1444826502-49291-4-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Wang Nan
|
374ce938aa |
perf tools: Enable pre-event inherit setting by config terms
This patch allows perf record setting event's attr.inherit bit by
config terms like:
# perf record -e cycles/no-inherit/ ...
# perf record -e cycles/inherit/ ...
So user can control inherit bit for each event separately.
In following example, a.out fork()s in main then do some complex
CPU intensive computations in both of its children.
Basic result with and without inherit:
# perf record -e cycles -e instructions ./a.out
[ perf record: Woken up 9 times to write data ]
[ perf record: Captured and wrote 2.205 MB perf.data (47920 samples) ]
# perf report --stdio
# ...
# Samples: 23K of event 'cycles'
# Event count (approx.): 23641752891
...
# Samples: 24K of event 'instructions'
# Event count (approx.): 30428312415
# perf record -i -e cycles -e instructions ./a.out
[ perf record: Woken up 5 times to write data ]
[ perf record: Captured and wrote 1.111 MB perf.data (24019 samples) ]
...
# Samples: 12K of event 'cycles'
# Event count (approx.): 11699501775
...
# Samples: 12K of event 'instructions'
# Event count (approx.): 15058023559
Cancel inherit for one event when globally enable:
# perf record -e cycles/no-inherit/ -e instructions ./a.out
[ perf record: Woken up 7 times to write data ]
[ perf record: Captured and wrote 1.660 MB perf.data (36004 samples) ]
...
# Samples: 12K of event 'cycles/no-inherit/'
# Event count (approx.): 11895759282
...
# Samples: 24K of event 'instructions'
# Event count (approx.): 30668000441
Enable inherit for one event when globally disable:
# perf record -i -e cycles/inherit/ -e instructions ./a.out
[ perf record: Woken up 7 times to write data ]
[ perf record: Captured and wrote 1.654 MB perf.data (35868 samples) ]
...
# Samples: 23K of event 'cycles/inherit/'
# Event count (approx.): 23285400229
...
# Samples: 11K of event 'instructions'
# Event count (approx.): 14969050259
Committer note:
One can check if the bit was set, in addition to seeing the result in
the perf.data file size as above by doing one of:
# perf record -e cycles -e instructions -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.911 MB perf.data (63 samples) ]
# perf evlist -v
cycles: size: 112, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|ID|CPU|PERIOD, read_format: ID, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
instructions: size: 112, config: 0x1, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|ID|CPU|PERIOD, read_format: ID, disabled: 1, inherit: 1, freq: 1, sample_id_all: 1, exclude_guest: 1
#
So, the inherit bit was set in both, now, if we disable it globally using
--no-inherit:
# perf record --no-inherit -e cycles -e instructions -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.910 MB perf.data (56 samples) ]
# perf evlist -v
cycles: size: 112, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|ID|CPU|PERIOD, read_format: ID, disabled: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
instructions: size: 112, config: 0x1, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|ID|CPU|PERIOD, read_format: ID, disabled: 1, freq: 1, sample_id_all: 1, exclude_guest: 1
No inherit bit set, then disabling it and setting just on the cycles event:
# perf record --no-inherit -e cycles/inherit/ -e instructions -a usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.909 MB perf.data (48 samples) ]
# perf evlist -v
cycles/inherit/: size: 112, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|ID|CPU|PERIOD, read_format: ID, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
instructions: size: 112, config: 0x1, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|ID|CPU|PERIOD, read_format: ID, disabled: 1, freq: 1, sample_id_all: 1, exclude_guest: 1
#
We can see it as well in by using a more verbose level of debug messages in
the tool that sets up the perf_event_attr, 'perf record' in this case:
[root@zoo ~]# perf record -vv --no-inherit -e cycles/inherit/ -e instructions -a usleep 1
------------------------------------------------------------
perf_event_attr:
size 112
{ sample_period, sample_freq } 4000
sample_type IP|TID|TIME|ID|CPU|PERIOD
read_format ID
disabled 1
inherit 1
mmap 1
comm 1
freq 1
task 1
sample_id_all 1
exclude_guest 1
mmap2 1
comm_exec 1
------------------------------------------------------------
sys_perf_event_open: pid -1 cpu 0 group_fd -1 flags 0x8
sys_perf_event_open: pid -1 cpu 1 group_fd -1 flags 0x8
sys_perf_event_open: pid -1 cpu 2 group_fd -1 flags 0x8
sys_perf_event_open: pid -1 cpu 3 group_fd -1 flags 0x8
------------------------------------------------------------
perf_event_attr:
size 112
config 0x1
{ sample_period, sample_freq } 4000
sample_type IP|TID|TIME|ID|CPU|PERIOD
read_format ID
disabled 1
freq 1
sample_id_all 1
exclude_guest 1
------------------------------------------------------------
sys_perf_event_open: pid -1 cpu 0 group_fd -1 flags 0x8
<SNIP>
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexei Starovoitov <ast@plumgrid.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1446029705-199659-2-git-send-email-wangnan0@huawei.com
[ s/u64/bool/ for the perf_evsel_config_term inherit field - jolsa]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Jiri Olsa
|
7f94af7a48 |
perf tools: Introduce 'P' modifier to request max precision
The 'P' will cause the event to get maximum possible detected precise
level.
Following record:
$ perf record -e cycles:P ...
will detect maximum precise level for 'cycles' event and use it.
Commiter note:
Testing it:
$ perf record -e cycles:P usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.013 MB perf.data (9 samples) ]
$ perf evlist
cycles:P
$ perf evlist -v
cycles:P: size: 112, { sample_period, sample_freq }: 4000, sample_type:
IP|TID|TIME|PERIOD, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1,
enable_on_exec: 1, task: 1, precise_ip: 2, sample_id_all: 1, mmap2: 1,
comm_exec: 1
$
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1444068369-20978-6-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
He Kuang
|
ffeb883e56 |
perf tools: Show proper error message for wrong terms of hw/sw events
Show proper error message and show valid terms when wrong config terms
is specified for hw/sw type perf events.
This patch makes the original error format function formats_error_string()
more generic, which only outputs the static config terms for hw/sw perf
events, and prepends pmu formats for pmu events.
Before this patch:
$ perf record -e 'cpu-clock/freqx=200/' -a sleep 1
invalid or unsupported event: 'cpu-clock/freqx=200/'
Run 'perf list' for a list of valid events
usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
After this patch:
$ perf record -e 'cpu-clock/freqx=200/' -a sleep 1
event syntax error: 'cpu-clock/freqx=200/'
\___ unknown term
valid terms: config,config1,config2,name,period,freq,branch_type,time,call-graph,stack-size
Run 'perf list' for a list of valid events
usage: perf record [<options>] [<command>]
or: perf record [<options>] -- <command> [<options>]
-e, --event <event> event selector. use 'perf list' to list available events
Signed-off-by: He Kuang <hekuang@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1443412336-120050-2-git-send-email-hekuang@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
Kan Liang
|
d457c96392 |
perf callchain: Per-event type selection support
This patchkit adds the ability to set callgraph mode (fp, dwarf, lbr) per
event. This in term can reduce sampling overhead and the size of the
perf.data.
Here is an example.
perf record -e 'cpu/cpu-cycles,period=1000,call-graph=fp,time=1/,cpu/instructions,call-graph=lbr/' sleep 1
perf evlist -v
cpu/cpu-cycles,period=1000,call-graph=fp,time=1/: type: 4, size: 112,
config: 0x3c, { sample_period, sample_freq }: 1000, sample_type:
IP|TID|TIME|CALLCHAIN|PERIOD|IDENTIFIER, read_format: ID, disabled: 1,
inherit: 1, mmap: 1, comm: 1, enable_on_exec: 1, task: 1, sample_id_all:
1, exclude_guest: 1, mmap2: 1, comm_exec: 1
cpu/instructions,call-graph=lbr/: type: 4, size: 112, config: 0xc0, {
sample_period, sample_freq }: 4000, sample_type:
IP|TID|TIME|CALLCHAIN|PERIOD|BRANCH_STACK|IDENTIFIER, read_format: ID,
disabled: 1, inherit: 1, freq: 1, enable_on_exec: 1, sample_id_all: 1,
exclude_guest: 1
Signed-off-by: Kan Liang <kan.liang@intel.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: http://lkml.kernel.org/r/1439289050-40510-1-git-send-email-kan.liang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
||
|
Namhyung Kim
|
09af2a5535 |
perf record: Support per-event freq term
Now perf can set per-event value of time and (sampling) period. But I guess most users like me just want to set frequency rather than period. So add the 'freq' term in the event parser. Signed-off-by: Namhyung Kim <namhyung@kernel.org> Acked-by: Jiri Olsa <jolsa@kernel.org> Cc: David Ahern <dsahern@gmail.com> Cc: Kan Liang <kan.liang@intel.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1439102724-14079-1-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |
||
|
Kan Liang
|
3206771239 |
perf tools: Per-event time support
This patchkit adds the ability to turn off time stamps per event. One usaful case for partial time is to work with per-event callgraph to enable "PEBS threshold > 1" (https://lkml.org/lkml/2015/5/10/196), which can significantly reduce the sampling overhead. The event samples with time stamps off will not be ordered. Signed-off-by: Kan Liang <kan.liang@intel.com> Acked-by: Jiri Olsa <jolsa@kernel.org> Cc: Andi Kleen <ak@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Link: http://lkml.kernel.org/r/1438677022-34296-2-git-send-email-kan.liang@intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |