| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
[ Upstream commit b63fd11cced17fcb8e133def29001b0f6aaa5e06 ]
When using 'perf stat' with repeat and interval option, it shows wrong
values for events.
The wrong values will be shown for the first interval on the second and
subsequent repetitions.
Without the fix:
# perf stat -r 3 -I 2000 -e faults -e sched:sched_switch -a sleep 5
2.000282489 53 faults
2.000282489 513 sched:sched_switch
4.005478208 3,721 faults
4.005478208 2,666 sched:sched_switch
5.025470933 395 faults
5.025470933 1,307 sched:sched_switch
2.009602825 1,84,46,74,40,73,70,95,47,520 faults <------
2.009602825 1,84,46,74,40,73,70,95,49,568 sched:sched_switch <------
4.019612206 4,730 faults
4.019612206 2,746 sched:sched_switch
5.039615484 3,953 faults
5.039615484 1,496 sched:sched_switch
2.000274620 1,84,46,74,40,73,70,95,47,520 faults <------
2.000274620 1,84,46,74,40,73,70,95,47,520 sched:sched_switch <------
4.000480342 4,282 faults
4.000480342 2,303 sched:sched_switch
5.000916811 1,322 faults
5.000916811 1,064 sched:sched_switch
#
prev_raw_counts is allocated when using intervals. This is used when
calculating the difference in the counts of events when using interval.
The current counts are stored in prev_raw_counts to calculate the
differences in the next iteration.
On the first interval of the second and subsequent repetitions,
prev_raw_counts would be the values stored in the last interval of the
previous repetitions, while the current counts will only be for the
first interval of the current repetition.
Hence there is a possibility of events showing up as big number.
Fix this by resetting prev_raw_counts whenever perf stat repeats the
command.
With the fix:
# perf stat -r 3 -I 2000 -e faults -e sched:sched_switch -a sleep 5
2.019349347 2,597 faults
2.019349347 2,753 sched:sched_switch
4.019577372 3,098 faults
4.019577372 2,532 sched:sched_switch
5.019415481 1,879 faults
5.019415481 1,356 sched:sched_switch
2.000178813 8,468 faults
2.000178813 2,254 sched:sched_switch
4.000404621 7,440 faults
4.000404621 1,266 sched:sched_switch
5.040196079 2,458 faults
5.040196079 556 sched:sched_switch
2.000191939 6,870 faults
2.000191939 1,170 sched:sched_switch
4.000414103 541 faults
4.000414103 902 sched:sched_switch
5.000809863 450 faults
5.000809863 364 sched:sched_switch
#
Committer notes:
This was broken since the cset introducing the --interval feature, i.e.
--repeat + --interval wasn't tested at that point, add the Fixes tag so
that automatic scripts can pick this up.
Fixes: 13370a9b5bb8 ("perf stat: Add interval printing")
Signed-off-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Tested-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: stable@vger.kernel.org # v3.9+
Link: http://lore.kernel.org/lkml/20190904094738.9558-2-srikar@linux.vnet.ibm.com
[ Fixed up conflicts with libperf, i.e. some perf_{evsel,evlist} lost the 'perf' prefix ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
|
|
[ Upstream commit 443f2d5ba13d65ccfd879460f77941875159d154 ]
Observe a segmentation fault when 'perf stat' is asked to repeat forever
with the interval option.
Without fix:
# perf stat -r 0 -I 5000 -e cycles -a sleep 10
# time counts unit events
5.000211692 3,13,89,82,34,157 cycles
10.000380119 1,53,98,52,22,294 cycles
10.040467280 17,16,79,265 cycles
Segmentation fault
This problem was only observed when we use forever option aka -r 0 and
works with limited repeats. Calling print_counter with ts being set to
NULL, is not a correct option when interval is set. Hence avoid
print_counter(NULL,..) if interval is set.
With fix:
# perf stat -r 0 -I 5000 -e cycles -a sleep 10
# time counts unit events
5.019866622 3,15,14,43,08,697 cycles
10.039865756 3,15,16,31,95,261 cycles
10.059950628 1,26,05,47,158 cycles
5.009902655 3,14,52,62,33,932 cycles
10.019880228 3,14,52,22,89,154 cycles
10.030543876 66,90,18,333 cycles
5.009848281 3,14,51,98,25,437 cycles
10.029854402 3,15,14,93,04,918 cycles
5.009834177 3,14,51,95,92,316 cycles
Committer notes:
Did the 'git bisect' to find the cset introducing the problem to add the
Fixes tag below, and at that time the problem reproduced as:
(gdb) run stat -r0 -I500 sleep 1
<SNIP>
Program received signal SIGSEGV, Segmentation fault.
print_interval (prefix=prefix@entry=0x7fffffffc8d0 "", ts=ts@entry=0x0) at builtin-stat.c:866
866 sprintf(prefix, "%6lu.%09lu%s", ts->tv_sec, ts->tv_nsec, csv_sep);
(gdb) bt
#0 print_interval (prefix=prefix@entry=0x7fffffffc8d0 "", ts=ts@entry=0x0) at builtin-stat.c:866
#1 0x000000000041860a in print_counters (ts=ts@entry=0x0, argc=argc@entry=2, argv=argv@entry=0x7fffffffd640) at builtin-stat.c:938
#2 0x0000000000419a7f in cmd_stat (argc=2, argv=0x7fffffffd640, prefix=<optimized out>) at builtin-stat.c:1411
#3 0x000000000045c65a in run_builtin (p=p@entry=0x6291b8 <commands+216>, argc=argc@entry=5, argv=argv@entry=0x7fffffffd640) at perf.c:370
#4 0x000000000045c893 in handle_internal_command (argc=5, argv=0x7fffffffd640) at perf.c:429
#5 0x000000000045c8f1 in run_argv (argcp=argcp@entry=0x7fffffffd4ac, argv=argv@entry=0x7fffffffd4a0) at perf.c:473
#6 0x000000000045cac9 in main (argc=<optimized out>, argv=<optimized out>) at perf.c:588
(gdb)
Mostly the same as just before this patch:
Program received signal SIGSEGV, Segmentation fault.
0x00000000005874a7 in print_interval (config=0xa1f2a0 <stat_config>, evlist=0xbc9b90, prefix=0x7fffffffd1c0 "`", ts=0x0) at util/stat-display.c:964
964 sprintf(prefix, "%6lu.%09lu%s", ts->tv_sec, ts->tv_nsec, config->csv_sep);
(gdb) bt
#0 0x00000000005874a7 in print_interval (config=0xa1f2a0 <stat_config>, evlist=0xbc9b90, prefix=0x7fffffffd1c0 "`", ts=0x0) at util/stat-display.c:964
#1 0x0000000000588047 in perf_evlist__print_counters (evlist=0xbc9b90, config=0xa1f2a0 <stat_config>, _target=0xa1f0c0 <target>, ts=0x0, argc=2, argv=0x7fffffffd670)
at util/stat-display.c:1172
#2 0x000000000045390f in print_counters (ts=0x0, argc=2, argv=0x7fffffffd670) at builtin-stat.c:656
#3 0x0000000000456bb5 in cmd_stat (argc=2, argv=0x7fffffffd670) at builtin-stat.c:1960
#4 0x00000000004dd2e0 in run_builtin (p=0xa30e00 <commands+288>, argc=5, argv=0x7fffffffd670) at perf.c:310
#5 0x00000000004dd54d in handle_internal_command (argc=5, argv=0x7fffffffd670) at perf.c:362
#6 0x00000000004dd694 in run_argv (argcp=0x7fffffffd4cc, argv=0x7fffffffd4c0) at perf.c:406
#7 0x00000000004dda11 in main (argc=5, argv=0x7fffffffd670) at perf.c:531
(gdb)
Fixes: d4f63a4741a8 ("perf stat: Introduce print_counters function")
Signed-off-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Tested-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Cc: stable@vger.kernel.org # v4.2+
Link: http://lore.kernel.org/lkml/20190904094738.9558-3-srikar@linux.vnet.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
|
|
[ Upstream commit fca32340a5e8b896f57d41fd94b8b1701df25eb1 ]
Executing command 'perf stat -T -- ls' dumps core on x86 and s390.
Here is the call back chain (done on x86):
# gdb ./perf
....
(gdb) r stat -T -- ls
...
Program received signal SIGSEGV, Segmentation fault.
0x00007ffff56d1963 in vasprintf () from /lib64/libc.so.6
(gdb) where
#0 0x00007ffff56d1963 in vasprintf () from /lib64/libc.so.6
#1 0x00007ffff56ae484 in asprintf () from /lib64/libc.so.6
#2 0x00000000004f1982 in __parse_events_add_pmu (parse_state=0x7fffffffd580,
list=0xbfb970, name=0xbf3ef0 "cpu",
head_config=0xbfb930, auto_merge_stats=false) at util/parse-events.c:1233
#3 0x00000000004f1c8e in parse_events_add_pmu (parse_state=0x7fffffffd580,
list=0xbfb970, name=0xbf3ef0 "cpu",
head_config=0xbfb930) at util/parse-events.c:1288
#4 0x0000000000537ce3 in parse_events_parse (_parse_state=0x7fffffffd580,
scanner=0xbf4210) at util/parse-events.y:234
#5 0x00000000004f2c7a in parse_events__scanner (str=0x6b66c0
"task-clock,{instructions,cycles,cpu/cycles-t/,cpu/tx-start/}",
parse_state=0x7fffffffd580, start_token=258) at util/parse-events.c:1673
#6 0x00000000004f2e23 in parse_events (evlist=0xbe9990, str=0x6b66c0
"task-clock,{instructions,cycles,cpu/cycles-t/,cpu/tx-start/}", err=0x0)
at util/parse-events.c:1713
#7 0x000000000044e137 in add_default_attributes () at builtin-stat.c:2281
#8 0x000000000044f7b5 in cmd_stat (argc=1, argv=0x7fffffffe3b0) at
builtin-stat.c:2828
#9 0x00000000004c8b0f in run_builtin (p=0xab01a0 <commands+288>, argc=4,
argv=0x7fffffffe3b0) at perf.c:297
#10 0x00000000004c8d7c in handle_internal_command (argc=4,
argv=0x7fffffffe3b0) at perf.c:349
#11 0x00000000004c8ece in run_argv (argcp=0x7fffffffe20c,
argv=0x7fffffffe200) at perf.c:393
#12 0x00000000004c929c in main (argc=4, argv=0x7fffffffe3b0) at perf.c:537
(gdb)
It turns out that a NULL pointer is referenced. Here are the
function calls:
...
cmd_stat()
+---> add_default_attributes()
+---> parse_events(evsel_list, transaction_attrs, NULL);
3rd parameter set to NULL
Function parse_events(xx, xx, struct parse_events_error *err) dives
into a bison generated scanner and creates
parser state information for it first:
struct parse_events_state parse_state = {
.list = LIST_HEAD_INIT(parse_state.list),
.idx = evlist->nr_entries,
.error = err, <--- NULL POINTER !!!
.evlist = evlist,
};
Now various functions inside the bison scanner are called to end up in
__parse_events_add_pmu(struct parse_events_state *parse_state, ..) with
first parameter being a pointer to above structure definition.
Now the PMU event name is not found (because being executed in a VM) and
this function tries to create an error message with
asprintf(&parse_state->error.str, ....)
which references a NULL pointer and dumps core.
Fix this by providing a pointer to the necessary error information
instead of NULL. Technically only the else part is needed to avoid the
core dump, just lets be safe...
Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Link: http://lkml.kernel.org/r/20180308145735.64717-1-tmricht@linux.vnet.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Sasha Levin <alexander.levin@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
commit 40c21898ba5372c14ef71717040529794a91ccc2 upstream.
When printing stats in CSV mode, 'perf stat' appends extra separators
when a counter is not supported:
<not supported>,,L1-dcache-store-misses,mesos/bd442f34-2b4a-47df-b966-9b281f9f56fc,0,100.00,,,,
Which causes a failure when parsing fields. The numbers of separators
should be the same for each line, no matter if the counter is or not
supported.
Signed-off-by: Ilya Pronin <ipronin@twitter.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Andi Kleen <ak@linux.intel.com>
Link: http://lkml.kernel.org/r/20180306064353.31930-1-xiyou.wangcong@gmail.com
Fixes: 92a61f6412d3 ("perf stat: Implement CSV metrics output")
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
When packaging the perf userland application into an AppImage, the
wait() call in perf stat returned too early. It turned out that some
other child process exited, but not the one perf stat launched:
$ sudo strace -e fork,execve,clone,wait4 -f ./perf-x86_64.AppImage stat sleep 1
execve("./perf-git.3a73b7f9-x86_64.AppImage", ["./perf-git.3a73b7f9-x86_64.AppIm"..., "stat", "sleep", "1"], 0x7ffec1bbf050 /* 18 vars */) = 0
clone(child_stack=NULL, flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, child_tidptr=0x7f6a6e7efe50) = 3912
strace: Process 3912 attached
[pid 3912] clone(child_stack=NULL, flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, child_tidptr=0x7f6a6e7efe50) = 3914
strace: Process 3914 attached
[pid 3912] +++ exited with 0 +++
[pid 3911] --- SIGCHLD {si_signo=SIGCHLD, si_code=CLD_EXITED, si_pid=3912, si_uid=0, si_status=0, si_utime=0, si_stime=0} ---
[pid 3914] clone(strace: Process 3915 attached
child_stack=0x7f6a6d9fefb0, flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID, parent_tidptr=0x7f6a6d9ff9d0, tls=0x7f6a6d9ff700, child_tidptr=0x7f6a6d9ff9d0) = 3915
[pid 3911] execve("/tmp/.mount_perf-g6VYMpl/AppRun", ["./perf-git.3a73b7f9-x86_64.AppIm"..., "stat", "sleep", "1"], 0x14aab70 /* 21 vars */) = 0
[pid 3911] clone(child_stack=NULL, flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, child_tidptr=0x7f4ae113c4d0) = 3916
strace: Process 3916 attached
[pid 3911] wait4(-1, [{WIFEXITED(s) && WEXITSTATUS(s) == 0}], 0, NULL) = 3912
[pid 3916] execve("/usr/libexec/perf-core/sleep", ["sleep", "1"], 0x27d3650 /* 22 vars */) = -1 ENOENT (No such file or directory)
[pid 3916] execve("/tmp/./sleep", ["sleep", "1"], 0x27d3650 /* 22 vars */) = -1 ENOENT (No such file or directory)
[pid 3916] execve("/home/milian/.bin/sleep", ["sleep", "1"], 0x27d3650 /* 22 vars */) = -1 ENOENT (No such file or directory)
[pid 3916] execve("/usr/lib/icecream/libexec/icecc/bin/sleep", ["sleep", "1"], 0x27d3650 /* 22 vars */) = -1 ENOENT (No such file or directory)
[pid 3916] execve("/ssd2/milian/projects/compiled/other/bin/sleep", ["sleep", "1"], 0x27d3650 /* 22 vars */) = -1 ENOENT (No such file or directory)
[pid 3916] execve("/home/milian/.bin/kf5/sleep", ["sleep", "1"], 0x27d3650 /* 22 vars */) = -1 ENOENT (No such file or directory)
[pid 3916] execve("/ssd2/milian/projects/compiled/kf5/bin/sleep", ["sleep", "1"], 0x27d3650 /* 22 vars */) = -1 ENOENT (No such file or directory)
[pid 3916] execve("/home/milian/projects/compiled/other/bin/sleep", ["sleep", "1"], 0x27d3650 /* 22 vars */) = -1 ENOENT (No such file or directory)
[pid 3916] execve("/home/milian/projects/compiled/kf5/bin/sleep", ["sleep", "1"], 0x27d3650 /* 22 vars */) = -1 ENOENT (No such file or directory)
[pid 3916] execve("/usr/local/sbin/sleep", ["sleep", "1"], 0x27d3650 /* 22 vars */) = -1 ENOENT (No such file or directory)
[pid 3916] execve("/usr/local/bin/sleep", ["sleep", "1"], 0x27d3650 /* 22 vars */) = -1 ENOENT (No such file or directory)
[pid 3916] execve("/usr/bin/sleep", ["sleep", "1"], 0x27d3650 /* 22 vars */
Performance counter stats for 'sleep 1':
<not counted> task-clock
<not counted> context-switches
<not counted> cpu-migrations
<not counted> page-faults
<not counted> cycles
<not counted> instructions
<not counted> branches
<not counted> branch-misses
0.000047194 seconds time elapsed
[pid 3916] --- SIGTERM {si_signo=SIGTERM, si_code=SI_USER, si_pid=3911, si_uid=0} ---
[pid 3916] +++ killed by SIGTERM +++
[pid 3911] --- SIGCHLD {si_signo=SIGCHLD, si_code=CLD_KILLED, si_pid=3916, si_uid=0, si_status=SIGTERM, si_utime=0, si_stime=0} ---
[pid 3915] --- SIGPIPE {si_signo=SIGPIPE, si_code=SI_USER, si_pid=3914, si_uid=0} ---
[pid 3911] +++ exited with 0 +++
[pid 3915] --- SIGHUP {si_signo=SIGHUP, si_code=SI_USER, si_pid=3914, si_uid=0} ---
[pid 3915] +++ exited with 0 +++
+++ exited with 0 +++
This patch uses waitpid instead to ensure the call waits for the
debuggee application launched by 'perf stat'. This fixes 'perf stat'
when launched from an AppImage:
$ ./perf-x86_64.AppImage stat sleep 1
Performance counter stats for 'sleep 1':
0.357235 task-clock (msec) # 0.000 CPUs utilized
1 context-switches # 0.003 M/sec
0 cpu-migrations # 0.000 K/sec
50 page-faults # 0.140 M/sec
1269602 cycles # 3.554 GHz
654278 instructions # 0.52 insn per cycle
129963 branches # 363.803 M/sec
7082 branch-misses # 5.45% of all branches
1.000633420 seconds time elapsed
Signed-off-by: Milian Wolff <milian.wolff@kdab.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20170912152523.4497-1-milian.wolff@kdab.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Peter reported that when he explicitely asked for multiple events with
the same name on the command line it got coalesced into just one line,
i.e.:
# perf stat -e cycles -e cycles -e cycles usleep 1
Performance counter stats for 'usleep 1':
3,269,652 cycles
0.000884123 seconds time elapsed
#
And while there is the --no-merges option to disable that auto-merging,
this is a blunt change in behaviour for such explicit request, so change
the code so that this auto merging is done only when handling the multi
PMU aliases with the same name that introduced this coalescing,
restoring the previous behaviour for the explicit case:
# perf stat -e cycles -e cycles -e cycles usleep 1
Performance counter stats for 'usleep 1':
1,472,837 cycles
1,472,837 cycles
1,472,837 cycles
0.001764870 seconds time elapsed
#
Reported-by: Peter Zijlstra <peterz@infradead.org>
Acked-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Fixes: 430daf2dc7af ("perf stat: Collapse identically named events")
Link: http://lkml.kernel.org/r/20170831184122.GK4831@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Make perf stat use group read if there are groups defined. The group
read will get the values for all member of groups within a single
syscall instead of calling read syscall for every event.
We can see considerable less amount of kernel cycles spent on single
group read, than reading each event separately, like for following perf
stat command:
# perf stat -e {cycles,instructions} -I 10 -a sleep 1
Monitored with "perf stat -r 5 -e '{cycles:u,cycles:k}'"
Before:
24,325,676 cycles:u
297,040,775 cycles:k
1.038554134 seconds time elapsed
After:
25,034,418 cycles:u
158,256,395 cycles:k
1.036864497 seconds time elapsed
The perf_evsel__open fallback changes contributed by Andi Kleen.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20170726120206.9099-4-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
To consolidate the error reporting facility.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-b41iot1094katoffdf19w9zk@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Implementing a new --smi-cost mode in perf stat to measure SMI cost.
During the measurement, the /sys/device/cpu/freeze_on_smi will be set.
The measurement can be done with one counter (unhalted core cycles), and
two free running MSR counters (IA32_APERF and SMI_COUNT).
In practice, the percentages of SMI core cycles should be more useful
than absolute value. So the output will be the percentage of SMI core
cycles and SMI#. metric_only will be set by default.
SMI cycles% = (aperf - unhalted core cycles) / aperf
Here is an example output.
Performance counter stats for 'sudo echo ':
SMI cycles% SMI#
0.1% 1
0.010858678 seconds time elapsed
Users who wants to get the actual value can apply additional
--no-metric-only.
Signed-off-by: Kan Liang <Kan.liang@intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Robert Elliott <elliott@hpe.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1495825538-5230-3-git-send-email-kan.liang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Only print the NMI watchdog hint when that watchdog it actually enabled.
This avoids printing these unnecessarily.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/n/tip-lnw7edxnqsphkmeew857wz1i@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Not needed in this header, added to the places that need poll(), wait()
and a few other prototypes.
Link: http://lkml.kernel.org/n/tip-i39c7b6xmo1vwd9wxp6fmkl0@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Not needed in this header, added to the places that need FILE,
putchar(), access() and a few other prototypes.
Link: http://lkml.kernel.org/n/tip-xxtdsl6nsna82j7puwbdjqhs@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
And remove it from util.h, disentangling it a bit more.
Link: http://lkml.kernel.org/n/tip-2zg9s5nx90yde64j3g4z2uhk@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Removing it from util.h, part of an effort to disentangle the includes
hell, that makes changes to util.h or something included by it to cause
a complete rebuild of the tools.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-ztrjy52q1rqcchuy3rubfgt2@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Moving them from util.h, where they don't belong. Since libc already
have string.h, name it slightly differently, as string2.h.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-eh3vz5sqxsrdd8lodoro4jrw@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
More stuff that came from git, out of the hodge-podge that is util.h
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-e3lana4gctz3ub4hn4y29hkw@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Needed to use the PRI[xu](32,64) formatting macros.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-wkbho8kaw24q67dd11q0j39f@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
(This is a patch has been sitting in the Intel CQM/CMT driver series for
a while, despite not depend on it. Sending it now independently since
the series is being discarded.)
When an event is in error state, read() returns 0 instead of sizeof()
buffer. In certain modes, such as interval printing, ignoring the 0
return value may cause bogus count deltas to be computed and thus
invalid results printed.
This patch fixes this problem by modifying read_counters() to mark the
event as not scaled (scaled = -1) to force the printout routine to show
<NOT COUNTED>.
Signed-off-by: Stephane Eranian <eranian@google.com>
Reviewed-by: David Carrillo-Cisneros <davidcc@google.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Paul Turner <pjt@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/20170412182301.44406-1-davidcc@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
To strip csv output, use ltrim() instead of just while loop and
isspace() at print_metric_{only}_csv().
Signed-off-by: Taeung Song <treeze.taeung@gmail.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1491575061-704-3-git-send-email-treeze.taeung@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
We got it from the git sources but never used it for anything, with the
place where this would be somehow used remaining:
static int run_builtin(struct cmd_struct *p, int argc, const char **argv)
{
prefix = NULL;
if (p->option & RUN_SETUP)
prefix = NULL; /* setup_perf_directory(); */
Ditch it.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-uw5swz05vol0qpr32c5lpvus@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Add generic infrastructure to perf stat to output ratios for
"MetricExpr" entries in the event lists. Many events are more useful as
ratios than in raw form, typically some count in relation to total
ticks.
Transfer the MetricExpr information from the alias to the evsel.
We mark the events that need to be collected for MetricExpr, and also
link the events using them with a pointer. The code is careful to always
prefer the right event in the same group to minimize multiplexing
errors. At the moment only a single relation is supported.
Then add a rblist to the stat shadow code that remembers stats based on
the cpu and context.
Then finally update and retrieve and print these values similarly to the
existing hardcoded perf metrics. We use the simple expression parser
added earlier to evaluate the expression.
Normally we just output the result without further commentary, but for
--metric-only this would lead to empty columns. So for this case use the
original event as description.
There is no attempt to automatically add the MetricExpr event, if it is
missing, however we suggest it to the user, because the user tool
doesn't have enough information to reliably construct a group that is
guaranteed to schedule. So we leave that to the user.
% perf stat -a -I 1000 -e '{unc_p_clockticks,unc_p_freq_max_os_cycles}'
1.000147889 800,085,181 unc_p_clockticks
1.000147889 93,126,241 unc_p_freq_max_os_cycles # 11.6
2.000448381 800,218,217 unc_p_clockticks
2.000448381 142,516,095 unc_p_freq_max_os_cycles # 17.8
3.000639852 800,243,057 unc_p_clockticks
3.000639852 162,292,689 unc_p_freq_max_os_cycles # 20.3
% perf stat -a -I 1000 -e '{unc_p_clockticks,unc_p_freq_max_os_cycles}' --metric-only
# time freq_max_os_cycles %
1.000127077 0.9
2.000301436 0.7
3.000456379 0.0
v2: Change from DivideBy to MetricExpr
v3: Use expr__ prefix. Support more than one other event.
v4: Update description
v5: Only print warning message once for multiple PMUs.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/20170320201711.14142-11-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
When any result that is being merged is bad, mark them all bad to give
consistent output in interval mode.
No before/after, because the issue was only found in theoretical review
and it is hard to reproduce
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/20170320201711.14142-4-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
The uncore PMU has a lot of duplicated PMUs for different subsystems.
When expanding an uncore alias we usually end up with a large
number of identically named aliases, which makes perf stat
output difficult to read.
Automatically sum them up in perf stat, unless --no-merge is specified.
This can be default because only the uncores generally have duplicated
aliases. Other PMUs have unique names.
Before:
% perf stat --no-merge -a -e unc_c_llc_lookup.any sleep 1
Performance counter stats for 'system wide':
694,976 Bytes unc_c_llc_lookup.any
706,304 Bytes unc_c_llc_lookup.any
956,608 Bytes unc_c_llc_lookup.any
782,720 Bytes unc_c_llc_lookup.any
605,696 Bytes unc_c_llc_lookup.any
442,816 Bytes unc_c_llc_lookup.any
659,328 Bytes unc_c_llc_lookup.any
509,312 Bytes unc_c_llc_lookup.any
263,936 Bytes unc_c_llc_lookup.any
592,448 Bytes unc_c_llc_lookup.any
672,448 Bytes unc_c_llc_lookup.any
608,640 Bytes unc_c_llc_lookup.any
641,024 Bytes unc_c_llc_lookup.any
856,896 Bytes unc_c_llc_lookup.any
808,832 Bytes unc_c_llc_lookup.any
684,864 Bytes unc_c_llc_lookup.any
710,464 Bytes unc_c_llc_lookup.any
538,304 Bytes unc_c_llc_lookup.any
1.002577660 seconds time elapsed
After:
% perf stat -a -e unc_c_llc_lookup.any sleep 1
Performance counter stats for 'system wide':
2,685,120 Bytes unc_c_llc_lookup.any
1.002648032 seconds time elapsed
v2: Split collect_aliases. Rename alias flag.
v3: Make sure unsupported/not counted is always printed.
v4: Factor out callback change into separate patch.
v5: Move check for bad results here
Move merged check into collect_data
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/20170320201711.14142-3-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
To be used in next patch to support automatic summing of alias events.
v2: Move check for bad results to next patch
v3: Remove trivial addition.
v4: Use perf_evsel__cpus instead of evsel->cpus
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/20170320201711.14142-2-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Make system wide (-a) the default option if no target was specified and
one of following conditions is met:
- there's no workload specified (current behaviour)
- there is workload specified but all requested
events are system wide ones
Mixed events core/uncore with workload:
$ perf stat -e 'uncore_cbox_0/clockticks/,cycles' sleep 1
Performance counter stats for 'sleep 1':
<not supported> uncore_cbox_0/clockticks/
980,489 cycles
1.000897406 seconds time elapsed
Uncore event with workload:
$ perf stat -e 'uncore_cbox_0/clockticks/' sleep 1
Performance counter stats for 'system wide':
281,473,897,192,670 uncore_cbox_0/clockticks/
1.000833784 seconds time elapsed
Committer note:
When testing I realized the default case for !root, i.e. no events
passed via -e, was broke by v2 of this patch, reported and after a
patch provided by Jiri it is back working:
[acme@jouet linux]$ perf stat usleep 1
Performance counter stats for 'usleep 1':
0.401335 task-clock:u (msec) # 0.297 CPUs utilized
0 context-switches:u # 0.000 K/sec
0 cpu-migrations:u # 0.000 K/sec
48 page-faults:u # 0.120 M/sec
458,146 cycles:u # 1.142 GHz
245,113 instructions:u # 0.54 insn per cycle
47,991 branches:u # 119.578 M/sec
4,022 branch-misses:u # 8.38% of all branches
0.001350029 seconds time elapsed
[acme@jouet linux]$
Suggested-and-Tested-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20170227094818.GA12764@krava
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
When using perf stat on an AMD F15h system with the default hw events
attributes, some of the events don't get counted:
Performance counter stats for 'sleep 1':
0.749208 task-clock (msec) # 0.001 CPUs utilized
1 context-switches # 0.001 M/sec
0 cpu-migrations # 0.000 K/sec
54 page-faults # 0.072 M/sec
1,122,815 cycles # 1.499 GHz
286,740 stalled-cycles-frontend # 25.54% frontend cycles idle
<not counted> stalled-cycles-backend (0.00%)
^^^^^^^^^^^^
<not counted> instructions (0.00%)
^^^^^^^^^^^^
<not counted> branches (0.00%)
<not counted> branch-misses (0.00%)
1.001550070 seconds time elapsed
The reason is that we have the HW watchdog consuming one PMU counter and
when perf tries to schedule 6 events on 6 counters and some of those
counters are constrained to only a specific subset of PMCs by the
hardware, the event scheduling fails.
So issue a hint to disable the HW watchdog around a perf stat session.
Committer note:
Testing it...
# perf stat -d usleep 1
Performance counter stats for 'usleep 1':
1.180203 task-clock (msec) # 0.490 CPUs utilized
1 context-switches # 0.847 K/sec
0 cpu-migrations # 0.000 K/sec
54 page-faults # 0.046 M/sec
184,754 cycles # 0.157 GHz
714,553 instructions # 3.87 insn per cycle
154,661 branches # 131.046 M/sec
7,247 branch-misses # 4.69% of all branches
219,984 L1-dcache-loads # 186.395 M/sec
17,600 L1-dcache-load-misses # 8.00% of all L1-dcache hits (90.16%)
<not counted> LLC-loads (0.00%)
<not counted> LLC-load-misses (0.00%)
0.002406823 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
#
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Robert Richter <rric@kernel.org>
Cc: Vince Weaver <vince@deater.net>
Link: http://lkml.kernel.org/r/20170211183218.ijnvb5f7ciyuunx4@pd.tnic
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
It now can have negative value to suppress the message entirely. So it
needs to check it being positive.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: kernel-team@lge.com
Link: http://lkml.kernel.org/r/20170217081742.17417-3-namhyung@kernel.org
[ Adjust fuzz on tools/perf/util/pmu.c, add > 0 checks in many other places ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Boris asked for default -a option in case we monitor only uncore events.
While implementing that I thought it might be actually useful to make it
overall default.
Running 'perf stat' will now collect system wide data.
Committer note:
Testing it:
# perf stat
^C
Performance counter stats for 'system wide':
3571.559178 cpu-clock (msec) # 4.000 CPUs utilized
3,346 context-switches # 0.937 K/sec
277 cpu-migrations # 0.078 K/sec
57,271 page-faults # 0.016 M/sec
4,535,633,835 cycles # 1.270 GHz
6,389,736,516 instructions # 1.41 insn per cycle
1,541,293,875 branches # 431.547 M/sec
14,526,396 branch-misses # 0.94% of all branches
0.892950118 seconds time elapsed
#
Requested-and-Acked-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20170217170034.GB15389@krava
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
cpu_topology_map
There are 2 problems wrt. cpu_topology_map on systems with sparse CPUs:
1. offline/absent CPUs will have their socket_id and core_id set to -1
which triggers:
"socket_id number is too big.You may need to upgrade the perf tool."
2. size of cpu_topology_map (perf_env.cpu[]) is allocated based on
_SC_NPROCESSORS_CONF, but can be indexed with CPU ids going above.
Users of perf_env.cpu[] are using CPU id as index. This can lead
to read beyond what was allocated:
==19991== Invalid read of size 4
==19991== at 0x490CEB: check_cpu_topology (topology.c:69)
==19991== by 0x490CEB: test_session_topology (topology.c:106)
...
For example:
_SC_NPROCESSORS_CONF == 16
available: 2 nodes (0-1)
node 0 cpus: 0 6 8 10 16 22 24 26
node 0 size: 12004 MB
node 0 free: 9470 MB
node 1 cpus: 1 7 9 11 23 25 27
node 1 size: 12093 MB
node 1 free: 9406 MB
node distances:
node 0 1
0: 10 20
1: 20 10
This patch changes HEADER_NRCPUS.nr_cpus_available from _SC_NPROCESSORS_CONF
to max_present_cpu and updates any user of cpu_topology_map to iterate
with nr_cpus_avail.
As a consequence HEADER_CPU_TOPOLOGY core_id and socket_id lists get longer,
but maintain compatibility with pre-patch state - index to cpu_topology_map is
CPU id.
perf test 36 -v
36: Session topology :
--- start ---
test child forked, pid 22211
templ file: /tmp/perf-test-gmdX5i
CPU 0, core 0, socket 0
CPU 1, core 0, socket 1
CPU 6, core 10, socket 0
CPU 7, core 10, socket 1
CPU 8, core 1, socket 0
CPU 9, core 1, socket 1
CPU 10, core 9, socket 0
CPU 11, core 9, socket 1
CPU 16, core 0, socket 0
CPU 22, core 10, socket 0
CPU 23, core 10, socket 1
CPU 24, core 1, socket 0
CPU 25, core 1, socket 1
CPU 26, core 9, socket 0
CPU 27, core 9, socket 1
test child finished with 0
---- end ----
Session topology: Ok
Signed-off-by: Jan Stancek <jstancek@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/d7c05c6445fca74a8442c2c73cfffd349c52c44f.1487146877.git.jstancek@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
When a tool can't open counters due to the kernel.perf_event_paranoit
sysctl setting, we inform how to tweak it to allow the operation to
succeed, in addition to that, suggest setting /etc/sysctl.conf to
make the setting permanent.
Suggested-by: Ingo Molnar <mingo@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-4gwe99k4a6p12d4u8bbyttj2@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
I.e. those parameters/functions _are_ used, so ditch that misleading attribute.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-13cqtjh0yojg5gzvpq1zzpl0@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Now that the required mechanic is there to deal with PMU specific
configuration, add the functionality to the tools where events can be
selected.
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>
Cc: linux-arm-kernel@lists.infradead.org
Link: http://lkml.kernel.org/r/1474041004-13956-7-git-send-email-mathieu.poirier@linaro.org
[ Fix the build on XSI-compliant systems, using str_error_r() to make sure we return a string, not an integer ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
To match how this is done in the kernel.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-gym6yshewpdegt153u8v2q5r@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
And remove it from tools/perf/{perf,util}.h, making code that needs
these macros to include linux/time64.h instead, to match how this is
used in the kernel sources.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-e69fc1pvkgt57yvxqt6eunyg@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
When we don't have a tracee (i.e. we're attaching to a task or CPU),
counters can still be running after our workload finishes, and can still
be running as we read their values. As we read events one-by-one, there
can be arbitrary skew between values of events, even within a group.
This means that ratios within an event group are not reliable.
This skew can be seen if measuring a group of identical events, e.g:
# perf stat -a -C0 -e '{cycles,cycles}' sleep 1
To avoid this, we must stop groups from counting before we read the
values of any constituent events. This patch adds and makes use of a new
disable_counters() helper, which disables group leaders (and thus each
group as a whole). This mirrors the use of enable_counters() for
starting event groups in the absence of a tracee.
Closing a group leader splits the group, and without a disabled group
leader the newly split events will begin counting. Thus to ensure counts
are reliable we must defer closing group leaders until all counts have
been read. To do so this patch removes the event closing logic from the
read_counters() helper, explicitly closes the events using
perf_evlist__close(), which also aids legibility.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1470747869-3567-1-git-send-email-mark.rutland@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
In create_perf_stat_counter, when a target CPU has not been provided, we
call __perf_evsel__open with empty_cpu_map, and open a single FD per
thread. However, in read_counter we assume that we opened events for the
product of threads and CPUs described in the evsel's cpu_map.
Thus, if an evsel has a cpu_map with more than one entry, we will
attempt to access FDs that we didn't open. This could result in a number
of problems (e.g. blocking while reading from STDIN if the fd memory
happened to be initialised to zero).
This is problematic for systems were a logical CPU PMU covers some
arbitrary subset of CPUs. The cpu_map of any evsel for that PMU will be
initialised based on the cpumask exposed through sysfs, even if the user
requests per-thread events.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1468577293-19667-2-git-send-email-mark.rutland@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
The tools so far have been using the strerror_r() GNU variant, that
returns a string, be it the buffer passed or something else.
But that, besides being tricky in cases where we expect that the
function using strerror_r() returns the error formatted in a provided
buffer (we have to check if it returned something else and copy that
instead), breaks the build on systems not using glibc, like Alpine
Linux, where musl libc is used.
So, introduce yet another wrapper, str_error_r(), that has the GNU
interface, but uses the portable XSI variant of strerror_r(), so that
users rest asured that the provided buffer is used and it is what is
returned.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-d4t42fnf48ytlk8rjxs822tf@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
To match the semantics for list.h in the kernel, that are used to
implement those macros.
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: Taeung Song <treeze.taeung@gmail.com>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-qbcjlgj0ffxquxscahbpddi3@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
When in CSV mode --metric-only outputs an header, unlike the other
modes. Previously it did not properly print headers for the aggregation
columns, so the headers were actually shifted against the real values.
Fix this here by outputting the correct headers for CSV.
v2: Indent array.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/1464119559-17203-4-git-send-email-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
When --metric-only is enabled there were no headers for the topology in
interval mode. Also when headers were printed they were on a separate
line.
Before:
$ perf stat --metric-only -A -I 1000 -a
1.001038376 frontend cycles idle insn per cycle stalled cycles per insn branch-misses of all branches
1.001038376 CPU0 123.54% 0.23 5.29 7.61%
1.001038376 CPU1 137.78% 0.24 5.13 10.07%
1.001038376 CPU2 64.48% 0.22 5.50 6.84%
After:
$ perf stat --metric-only -A -I 1000 -a
1.001111114 CPU0 82.46% 0.32 2.60 7.64%
1.001111114 CPU1 126.63% 0.02 42.83 0.15%
1.001111114 CPU2 193.54% 0.32 2.59 6.92%
v2: Move all headers on a single line
Reported-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/1464119559-17203-3-git-send-email-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
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>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core
Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:
User visible changes:
- Honour the kernel.perf_event_max_stack knob more precisely by not counting
PERF_CONTEXT_{KERNEL,USER} when deciding when to stop adding entries to
the perf_sample->ip_callchain[] array (Arnaldo Carvalho de Melo)
- Fix identation of 'stalled-backend-cycles' in 'perf stat' (Namhyung Kim)
- Update runtime using 'cpu-clock' event in 'perf stat' (Namhyung Kim)
- Use 'cpu-clock' for cpu targets in 'perf stat' (Namhyung Kim)
- Avoid fractional digits for integer scales in 'perf stat' (Andi Kleen)
- Store vdso buildid unconditionally, as it appears in callchains and
we're not checking those when creating the build-id table, so we
end up not being able to resolve VDSO symbols when doing analysis
on a different machine than the one where recording was done, possibly
of a different arch even (arm -> x86_64) (He Kuang)
Infrastructure changes:
- Generalize max_stack sysctl handler, will be used for configuring
multiple kernel knobs related to callchains (Arnaldo Carvalho de Melo)
Cleanups:
- Introduce DSO__NAME_KALLSYMS and DSO__NAME_KCORE, to stop using
open coded strings (Masami Hiramatsu)
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
Currently 'perf stat' always counts task-clock event by default. But
it's somewhat confusing for system-wide targets (especially with 'sleep
N' as the 'sleep' task just sleeps and doesn't use cputime). Changing
to cpu-clock event instead for that case makes more sense IMHO.
Before:
# perf stat -a sleep 0.1
Performance counter stats for 'system wide':
403.038603 task-clock (msec) # 4.001 CPUs utilized
150 context-switches # 0.372 K/sec
7 cpu-migrations # 0.017 K/sec
71 page-faults # 0.176 K/sec
23,705,169 cycles # 0.059 GHz
15,888,166 instructions # 0.67 insn per cycle
3,326,078 branches # 8.253 M/sec
87,643 branch-misses # 2.64% of all branches
0.100737009 seconds time elapsed
#
After:
# perf stat -a sleep 0.1
Performance counter stats for 'system wide':
404.271182 cpu-clock (msec) # 4.000 CPUs utilized
143 context-switches # 0.354 K/sec
13 cpu-migrations # 0.032 K/sec
73 page-faults # 0.181 K/sec
22,119,220 cycles # 0.055 GHz
13,622,065 instructions # 0.62 insn per cycle
2,918,769 branches # 7.220 M/sec
85,033 branch-misses # 2.91% of all branches
0.101073089 seconds time elapsed
#
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1463119263-5569-3-git-send-email-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
When the scaling factor is a full integer don't display fractional
digits. This avoids unnecessary .00 output for topdown metrics with
scale factors.
v2: Remove redundant check.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1462489447-31832-7-git-send-email-andi@firstfloor.org
[ Rename 'round' to 'stat_round' as 'round' is defined in math.h,
included by this patch, and this breaks the build on ubuntu 12.04 ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull perf updates from Ingo Molnar:
"Bigger kernel side changes:
- Add backwards writing capability to the perf ring-buffer code,
which is preparation for future advanced features like robust
'overwrite support' and snapshot mode. (Wang Nan)
- Add pause and resume ioctls for the perf ringbuffer (Wang Nan)
- x86 Intel cstate code cleanups and reorgnization (Thomas Gleixner)
- x86 Intel uncore and CPU PMU driver updates (Kan Liang, Peter
Zijlstra)
- x86 AUX (Intel PT) related enhancements and updates (Alexander
Shishkin)
- x86 MSR PMU driver enhancements and updates (Huang Rui)
- ... and lots of other changes spread out over 40+ commits.
Biggest tooling side changes:
- 'perf trace' features and enhancements. (Arnaldo Carvalho de Melo)
- BPF tooling updates (Wang Nan)
- 'perf sched' updates (Jiri Olsa)
- 'perf probe' updates (Masami Hiramatsu)
- ... plus 200+ other enhancements, fixes and cleanups to tools/
The merge commits, the shortlog and the changelogs contain a lot more
details"
* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (249 commits)
perf/core: Disable the event on a truncated AUX record
perf/x86/intel/pt: Generate PMI in the STOP region as well
perf buildid-cache: Use lsdir() for looking up buildid caches
perf symbols: Use lsdir() for the search in kcore cache directory
perf tools: Use SBUILD_ID_SIZE where applicable
perf tools: Fix lsdir to set errno correctly
perf trace: Move seccomp args beautifiers to tools/perf/trace/beauty/
perf trace: Move flock op beautifier to tools/perf/trace/beauty/
perf build: Add build-test for debug-frame on arm/arm64
perf build: Add build-test for libunwind cross-platforms support
perf script: Fix export of callchains with recursion in db-export
perf script: Fix callchain addresses in db-export
perf script: Fix symbol insertion behavior in db-export
perf symbols: Add dso__insert_symbol function
perf scripting python: Use Py_FatalError instead of die()
perf tools: Remove xrealloc and ALLOC_GROW
perf help: Do not use ALLOC_GROW in add_cmd_list
perf pmu: Make pmu_formats_string to check return value of strbuf
perf header: Make topology checkers to check return value of strbuf
perf tools: Make alias handler to check return value of strbuf
...
|
|
After 0161028b7c8a ("perf/core: Change the default paranoia level to 2")
'perf stat' fails for users without CAP_SYS_ADMIN, so just use
'perf_evsel__fallback()' to have the same behaviour as 'perf record',
i.e. set perf_event_attr.exclude_kernel to 1.
Now:
[acme@jouet linux]$ perf stat usleep 1
Performance counter stats for 'usleep 1':
0.352536 task-clock:u (msec) # 0.423 CPUs utilized
0 context-switches:u # 0.000 K/sec
0 cpu-migrations:u # 0.000 K/sec
49 page-faults:u # 0.139 M/sec
309,407 cycles:u # 0.878 GHz
243,791 instructions:u # 0.79 insn per cycle
49,622 branches:u # 140.757 M/sec
3,884 branch-misses:u # 7.83% of all branches
0.000834174 seconds time elapsed
[acme@jouet linux]$
Reported-by: Ingo Molnar <mingo@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-b20jmx4dxt5hpaa9t2rroi0o@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Add debug output of raw counter values per CPU when perf stat -v is
specified, together with their cpu numbers. This is very useful to
debug problems with per core counters, where we can normally only see
aggregated values.
v2: Make it depend on -vv, not -v
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1461787251-6702-12-git-send-email-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Add metric only support for -A too. This requires a new print function
that prints the metrics in the right order.
v2: Fix manpage
v3: Simplify nrcpus computation
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/1457049458-28956-7-git-send-email-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Add a new mode to only print metrics. Sometimes we don't care about the
raw values, just want the computed metrics. This allows more compact
printing, so with -I each sample is only a single line. This also
allows easier plotting and processing with other tools.
The main target is with using --topdown, but it also works with -T and
standard perf stat. A few metrics are not supported.
To avoiding having to hardcode all the metrics in the code it uses a two
pass approach: first compute dummy metrics and only print the headers in
the print_metric callback. Then use the callback to print the actual
values.
There are some additional changes in the stat printout code to handle
all metrics being on a single line.
One issue is that the column code doesn't know in advance what events
are not supported by the CPU, and it would be hard to find out as this
could change based on dynamic conditions. That causes empty columns in
some cases.
The output can be fairly wide, often you may need more than 80 columns.
Example:
% perf stat -a -I 1000 --metric-only
1.001452803 frontend cycles idle insn per cycle stalled cycles per insn branch-misses of all branches
1.001452803 158.91% 0.66 2.39 2.92%
2.002192321 180.63% 0.76 2.08 2.96%
3.003088282 150.59% 0.62 2.57 2.84%
4.004369835 196.20% 0.98 1.62 3.79%
5.005227314 231.98% 0.84 1.90 4.71%
v2: Lots of updates.
v3: Use slightly narrower columns
v4: Add comment
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/1457049458-28956-6-git-send-email-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
Add an extra check for frontend stalled in the metrics. This avoids an
extra column for the --metric-only case when the CPU does not support
frontend stalled.
v2: Add separate init function
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/1456858672-21594-8-git-send-email-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
