A new branch sample type PERF_SAMPLE_BRANCH_HW_INDEX has been introduced
in latest kernel.
Enable HW_INDEX by default in LBR call stack mode.
If kernel doesn't support the sample type, switching it off.
Add HW_INDEX in attr_fprintf as well. User can check whether the branch
sample type is set via debug information or header.
Committer testing:
First collect some samples with LBR callchains, system wide, for a few
seconds:
# perf record --call-graph lbr -a sleep 5
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.625 MB perf.data (224 samples) ]
#
Now lets use 'perf evlist -v' to look at the branch_sample_type:
# perf evlist -v
cycles: size: 120, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|CALLCHAIN|CPU|PERIOD|BRANCH_STACK, read_format: ID, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol: 1, bpf_event: 1, branch_sample_type: USER|CALL_STACK|NO_FLAGS|NO_CYCLES|HW_INDEX
#
So the machine has the kernel feature, and it was correctly added to
perf_event_attr.branch_sample_type, for the default 'cycles' event.
If we do it in another machine, where the kernel lacks the HW_INDEX
feature, we get:
# perf record --call-graph lbr -a sleep 2s
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 1.690 MB perf.data (499 samples) ]
# perf evlist -v
cycles: size: 120, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|CALLCHAIN|CPU|PERIOD|BRANCH_STACK, read_format: ID, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol: 1, bpf_event: 1, branch_sample_type: USER|CALL_STACK|NO_FLAGS|NO_CYCLES
#
No HW_INDEX in attr.branch_sample_type.
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexey Budankov <alexey.budankov@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Pavel Gerasimov <pavel.gerasimov@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vitaly Slobodskoy <vitaly.slobodskoy@intel.com>
Link: http://lore.kernel.org/lkml/20200228163011.19358-3-kan.liang@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The low level index of raw branch records for the most recent branch can
be recorded in a sample with PERF_SAMPLE_BRANCH_HW_INDEX
branch_sample_type. Extend struct branch_stack to support it.
However, if the PERF_SAMPLE_BRANCH_HW_INDEX is not applied, only nr and
entries[] will be output by kernel. The pointer of entries[] could be
wrong, since the output format is different with new struct
branch_stack. Add a variable no_hw_idx in struct perf_sample to
indicate whether the hw_idx is output. Add get_branch_entry() to return
corresponding pointer of entries[0].
To make dummy branch sample consistent as new branch sample, add hw_idx
in struct dummy_branch_stack for cs-etm and intel-pt.
Apply the new struct branch_stack for synthetic events as well.
Extend test case sample-parsing to support new struct branch_stack.
Committer notes:
Renamed get_branch_entries() to perf_sample__branch_entries() to have
proper namespacing and pave the way for this to be moved to libperf,
eventually.
Add 'static' to that inline as it is in a header.
Add 'hw_idx' to 'struct dummy_branch_stack' in cs-etm.c to fix the build
on arm64.
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexey Budankov <alexey.budankov@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Pavel Gerasimov <pavel.gerasimov@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vitaly Slobodskoy <vitaly.slobodskoy@intel.com>
Link: http://lore.kernel.org/lkml/20200228163011.19358-2-kan.liang@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When we put an event with multiple probes, perf-probe fails to delete
with filters. This comes from a failure to list up the event name
because of overwrapping its name.
To fix this issue, skip to list up the event which has same name.
Without this patch:
# perf probe -l \*
probe_perf:map__map_ip (on perf_sample__fprintf_brstackoff:21@
probe_perf:map__map_ip (on perf_sample__fprintf_brstackoff:25@
probe_perf:map__map_ip (on append_inlines:12@util/machine.c in
probe_perf:map__map_ip (on unwind_entry:19@util/machine.c in /
probe_perf:map__map_ip (on map__map_ip@util/map.h in /home/mhi
probe_perf:map__map_ip (on map__map_ip@util/map.h in /home/mhi
# perf probe -d \*
"*" does not hit any event.
Error: Failed to delete events. Reason: No such file or directory (Code: -2)
With it:
# perf probe -d \*
Removed event: probe_perf:map__map_ip
#
Fixes: 72363540c0 ("perf probe: Support multiprobe event")
Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org>
Reported-by: He Zhe <zhe.he@windriver.com>
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: stable@vger.kernel.org
Link: http://lore.kernel.org/lkml/158287666197.16697.7514373548551863562.stgit@devnote2
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
clang warns:
util/block-info.c:298:18: error: result of comparison against a string
literal is unspecified (use an explicit string comparison function
instead) [-Werror,-Wstring-compare]
if ((start_line != SRCLINE_UNKNOWN) && (end_line != SRCLINE_UNKNOWN)) {
^ ~~~~~~~~~~~~~~~
util/block-info.c:298:51: error: result of comparison against a string
literal is unspecified (use an explicit string comparison function
instead) [-Werror,-Wstring-compare]
if ((start_line != SRCLINE_UNKNOWN) && (end_line != SRCLINE_UNKNOWN)) {
^ ~~~~~~~~~~~~~~~
util/block-info.c:298:18: error: result of comparison against a string
literal is unspecified (use an explicit string
comparison function instead) [-Werror,-Wstring-compare]
if ((start_line != SRCLINE_UNKNOWN) && (end_line != SRCLINE_UNKNOWN)) {
^ ~~~~~~~~~~~~~~~
util/block-info.c:298:51: error: result of comparison against a string
literal is unspecified (use an explicit string comparison function
instead) [-Werror,-Wstring-compare]
if ((start_line != SRCLINE_UNKNOWN) && (end_line != SRCLINE_UNKNOWN)) {
^ ~~~~~~~~~~~~~~~
util/map.c:434:15: error: result of comparison against a string literal
is unspecified (use an explicit string comparison function instead)
[-Werror,-Wstring-compare]
if (srcline != SRCLINE_UNKNOWN)
^ ~~~~~~~~~~~~~~~
Reviewer Notes:
Looks good to me. Some more context:
https://clang.llvm.org/docs/DiagnosticsReference.html#wstring-compare
The spec says:
J.1 Unspecified behavior
The following are unspecified:
.. Whether two string literals result in distinct arrays (6.4.5).
Signed-off-by: Nick Desaulniers <nick.desaulniers@gmail.com>
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Changbin Du <changbin.du@intel.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: John Keeping <john@metanate.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: clang-built-linux@googlegroups.com
Link: https://github.com/ClangBuiltLinux/linux/issues/900
Link: http://lore.kernel.org/lkml/20200223193456.25291-1-nick.desaulniers@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
To help in debugging, add this extra message:
detect_kbuild_dir: Couldn't find "/lib/modules/5.4.20-200.fc31.x86_64/build/include/generated/autoconf.h", missing kernel-devel package?.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We have supported the event modifier "percore" which sums up the event
counts for all hardware threads in a core and show the counts per core.
For example,
# perf stat -e cpu/event=cpu-cycles,percore/ -a -A -- sleep 1
Performance counter stats for 'system wide':
S0-D0-C0 395,072 cpu/event=cpu-cycles,percore/
S0-D0-C1 851,248 cpu/event=cpu-cycles,percore/
S0-D0-C2 954,226 cpu/event=cpu-cycles,percore/
S0-D0-C3 1,233,659 cpu/event=cpu-cycles,percore/
This patch provides a new option "--percore-show-thread". It is used
with event modifier "percore" together to sum up the event counts for
all hardware threads in a core but show the counts per hardware thread.
This is essentially a replacement for the any bit (which is gone in
Icelake). Per core counts are useful for some formulas, e.g. CoreIPC.
The original percore version was inconvenient to post process. This
variant matches the output of the any bit.
With this patch, for example,
# perf stat -e cpu/event=cpu-cycles,percore/ -a -A --percore-show-thread -- sleep 1
Performance counter stats for 'system wide':
CPU0 2,453,061 cpu/event=cpu-cycles,percore/
CPU1 1,823,921 cpu/event=cpu-cycles,percore/
CPU2 1,383,166 cpu/event=cpu-cycles,percore/
CPU3 1,102,652 cpu/event=cpu-cycles,percore/
CPU4 2,453,061 cpu/event=cpu-cycles,percore/
CPU5 1,823,921 cpu/event=cpu-cycles,percore/
CPU6 1,383,166 cpu/event=cpu-cycles,percore/
CPU7 1,102,652 cpu/event=cpu-cycles,percore/
We can see counts are duplicated in CPU pairs (CPU0/CPU4, CPU1/CPU5,
CPU2/CPU6, CPU3/CPU7).
The interval mode also works. For example,
# perf stat -e cpu/event=cpu-cycles,percore/ -a -A --percore-show-thread -I 1000
# time CPU counts unit events
1.000425421 CPU0 925,032 cpu/event=cpu-cycles,percore/
1.000425421 CPU1 430,202 cpu/event=cpu-cycles,percore/
1.000425421 CPU2 436,843 cpu/event=cpu-cycles,percore/
1.000425421 CPU3 1,192,504 cpu/event=cpu-cycles,percore/
1.000425421 CPU4 925,032 cpu/event=cpu-cycles,percore/
1.000425421 CPU5 430,202 cpu/event=cpu-cycles,percore/
1.000425421 CPU6 436,843 cpu/event=cpu-cycles,percore/
1.000425421 CPU7 1,192,504 cpu/event=cpu-cycles,percore/
If we offline CPU5, the result is:
# perf stat -e cpu/event=cpu-cycles,percore/ -a -A --percore-show-thread -- sleep 1
Performance counter stats for 'system wide':
CPU0 2,752,148 cpu/event=cpu-cycles,percore/
CPU1 1,009,312 cpu/event=cpu-cycles,percore/
CPU2 2,784,072 cpu/event=cpu-cycles,percore/
CPU3 2,427,922 cpu/event=cpu-cycles,percore/
CPU4 2,752,148 cpu/event=cpu-cycles,percore/
CPU6 2,784,072 cpu/event=cpu-cycles,percore/
CPU7 2,427,922 cpu/event=cpu-cycles,percore/
1.001416041 seconds time elapsed
v4:
---
Ravi Bangoria reports an issue in v3. Once we offline a CPU,
the output is not correct. The issue is we should use the cpu
idx in print_percore_thread rather than using the cpu value.
v3:
---
1. Fix the interval mode output error
2. Use cpu value (not cpu index) in config->aggr_get_id().
3. Refine the code according to Jiri's comments.
v2:
---
Add the explanation in change log. This is essentially a replacement
for the any bit. No code change.
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: Kan Liang <kan.liang@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20200214080452.26402-1-yao.jin@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
clang warns:
util/block-info.c:298:18: error: result of comparison against a string
literal is unspecified (use an explicit string comparison function
instead) [-Werror,-Wstring-compare]
if ((start_line != SRCLINE_UNKNOWN) && (end_line != SRCLINE_UNKNOWN)) {
^ ~~~~~~~~~~~~~~~
util/block-info.c:298:51: error: result of comparison against a string
literal is unspecified (use an explicit string comparison function
instead) [-Werror,-Wstring-compare]
if ((start_line != SRCLINE_UNKNOWN) && (end_line != SRCLINE_UNKNOWN)) {
^ ~~~~~~~~~~~~~~~
util/block-info.c:298:18: error: result of comparison against a string
literal is unspecified (use an explicit string
comparison function instead) [-Werror,-Wstring-compare]
if ((start_line != SRCLINE_UNKNOWN) && (end_line != SRCLINE_UNKNOWN)) {
^ ~~~~~~~~~~~~~~~
util/block-info.c:298:51: error: result of comparison against a string
literal is unspecified (use an explicit string comparison function
instead) [-Werror,-Wstring-compare]
if ((start_line != SRCLINE_UNKNOWN) && (end_line != SRCLINE_UNKNOWN)) {
^ ~~~~~~~~~~~~~~~
util/map.c:434:15: error: result of comparison against a string literal
is unspecified (use an explicit string comparison function instead)
[-Werror,-Wstring-compare]
if (srcline != SRCLINE_UNKNOWN)
^ ~~~~~~~~~~~~~~~
Reviewer Notes:
Looks good to me. Some more context:
https://clang.llvm.org/docs/DiagnosticsReference.html#wstring-compare
The spec says:
J.1 Unspecified behavior
The following are unspecified:
.. Whether two string literals result in distinct arrays (6.4.5).
Signed-off-by: Nick Desaulniers <nick.desaulniers@gmail.com>
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Changbin Du <changbin.du@intel.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: John Keeping <john@metanate.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: clang-built-linux@googlegroups.com
Link: https://github.com/ClangBuiltLinux/linux/issues/900
Link: http://lore.kernel.org/lkml/20200223193456.25291-1-nick.desaulniers@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The dso->kernel value is now set to everything that is in
machine->kmaps, but that was being used to decide if vmlinux lookup is
needed, which ended up making that lookup be made for kernel modules,
that now have dso->kernel set, leading to these kinds of warnings when
running on a machine with compressed kernel modules, like fedora:31:
[root@five ~]# perf record -F 10000 -a sleep 2
[ perf record: Woken up 1 times to write data ]
lzma: fopen failed on vmlinux: 'No such file or directory'
lzma: fopen failed on /boot/vmlinux: 'No such file or directory'
lzma: fopen failed on /boot/vmlinux-5.5.5-200.fc31.x86_64: 'No such file or directory'
lzma: fopen failed on /usr/lib/debug/boot/vmlinux-5.5.5-200.fc31.x86_64: 'No such file or directory'
lzma: fopen failed on /lib/modules/5.5.5-200.fc31.x86_64/build/vmlinux: 'No such file or directory'
lzma: fopen failed on vmlinux: 'No such file or directory'
lzma: fopen failed on /boot/vmlinux: 'No such file or directory'
lzma: fopen failed on /boot/vmlinux-5.5.5-200.fc31.x86_64: 'No such file or directory'
lzma: fopen failed on /usr/lib/debug/boot/vmlinux-5.5.5-200.fc31.x86_64: 'No such file or directory'
lzma: fopen failed on /lib/modules/5.5.5-200.fc31.x86_64/build/vmlinux: 'No such file or directory'
lzma: fopen failed on vmlinux: 'No such file or directory'
lzma: fopen failed on /boot/vmlinux: 'No such file or directory'
lzma: fopen failed on /boot/vmlinux-5.5.5-200.fc31.x86_64: 'No such file or directory'
lzma: fopen failed on /usr/lib/debug/boot/vmlinux-5.5.5-200.fc31.x86_64: 'No such file or directory'
lzma: fopen failed on /lib/modules/5.5.5-200.fc31.x86_64/build/vmlinux: 'No such file or directory'
lzma: fopen failed on vmlinux: 'No such file or directory'
lzma: fopen failed on /boot/vmlinux: 'No such file or directory'
lzma: fopen failed on /boot/vmlinux-5.5.5-200.fc31.x86_64: 'No such file or directory'
lzma: fopen failed on /usr/lib/debug/boot/vmlinux-5.5.5-200.fc31.x86_64: 'No such file or directory'
lzma: fopen failed on /lib/modules/5.5.5-200.fc31.x86_64/build/vmlinux: 'No such file or directory'
lzma: fopen failed on vmlinux: 'No such file or directory'
lzma: fopen failed on /boot/vmlinux: 'No such file or directory'
lzma: fopen failed on /boot/vmlinux-5.5.5-200.fc31.x86_64: 'No such file or directory'
lzma: fopen failed on /usr/lib/debug/boot/vmlinux-5.5.5-200.fc31.x86_64: 'No such file or directory'
lzma: fopen failed on /lib/modules/5.5.5-200.fc31.x86_64/build/vmlinux: 'No such file or directory'
[ perf record: Captured and wrote 1.024 MB perf.data (1366 samples) ]
[root@five ~]#
This happens when collecting the buildid, when we find samples for
kernel modules, fix it by checking if the looked up DSO is a kernel
module by other means.
Fixes: 02213cec64 ("perf maps: Mark module DSOs with kernel type")
Tested-by: Jiri Olsa <jolsa@redhat.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Kim Phillips <kim.phillips@amd.com>
Cc: Michael Petlan <mpetlan@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Link: http://lore.kernel.org/lkml/20200302191007.GD10335@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Make the code more compact by using asprintf() instead of malloc()+strncpy() which also uses
less memory and avoids these warnings with gcc 10:
CC /tmp/build/perf/util/cloexec.o
In file included from /usr/include/string.h:495,
from util/parse-events.h:12,
from util/parse-events.c:18:
In function ‘strncpy’,
inlined from ‘tracepoint_id_to_path’ at util/parse-events.c:271:5:
/usr/include/bits/string_fortified.h:106:10: error: ‘__builtin_strncpy’ offset [275, 511] from the object at ‘sys_dirent’ is out of the bounds of referenced subobject ‘d_name’ with type ‘char[256]’ at offset 19 [-Werror=array-bounds]
106 | return __builtin___strncpy_chk (__dest, __src, __len, __bos (__dest));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from /usr/include/dirent.h:61,
from util/parse-events.c:5:
util/parse-events.c: In function ‘tracepoint_id_to_path’:
/usr/include/bits/dirent.h:33:10: note: subobject ‘d_name’ declared here
33 | char d_name[256]; /* We must not include limits.h! */
| ^~~~~~
In file included from /usr/include/string.h:495,
from util/parse-events.h:12,
from util/parse-events.c:18:
In function ‘strncpy’,
inlined from ‘tracepoint_id_to_path’ at util/parse-events.c:273:5:
/usr/include/bits/string_fortified.h:106:10: error: ‘__builtin_strncpy’ offset [275, 511] from the object at ‘evt_dirent’ is out of the bounds of referenced subobject ‘d_name’ with type ‘char[256]’ at offset 19 [-Werror=array-bounds]
106 | return __builtin___strncpy_chk (__dest, __src, __len, __bos (__dest));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from /usr/include/dirent.h:61,
from util/parse-events.c:5:
util/parse-events.c: In function ‘tracepoint_id_to_path’:
/usr/include/bits/dirent.h:33:10: note: subobject ‘d_name’ declared here
33 | char d_name[256]; /* We must not include limits.h! */
| ^~~~~~
CC /tmp/build/perf/util/call-path.o
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: http://lore.kernel.org/lkml/20200302145535.GA28183@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
It is possible to return a pointer to a local variable when looking up
the architecture name for the running system and no normalization is
done on that value, i.e. we may end up returning the uts.machine local
variable.
While this doesn't happen on most arches, as normalization takes place,
lets fix this by making that a static variable and optimize it a bit by
not always running uname(), only the first time.
Noticed in fedora rawhide running with:
[perfbuilder@a5ff49d6e6e4 ~]$ gcc --version
gcc (GCC) 10.0.1 20200216 (Red Hat 10.0.1-0.8)
Reported-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf annotate --show-nr-samples does not really show number of samples.
The reason is we have two separate variables for the same purpose.
One is in symbol_conf.show_nr_samples and another is
annotation_options.show_nr_samples.
We save command line option in symbol_conf.show_nr_samples but uses
annotation_option.show_nr_samples while rendering tui/stdio2 browser.
Though, we copy symbol_conf.show_nr_samples to
annotation__default_options.show_nr_samples but that is not really
effective as we don't use annotation__default_options once we copy
default options to dynamic variable annotate.opts in cmd_annotate().
Instead of all these complication, keep only one variable and use it all
over. symbol_conf.show_nr_samples is used by perf report/top as well. So
let's kill annotation_options.show_nr_samples.
On a side note, I've kept annotation_options.show_nr_samples definition
because it's still used by perf-config code. Follow up patch to fix
perf-config for annotate will remove annotation_options.show_nr_samples.
Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexey Budankov <alexey.budankov@linux.intel.com>
Cc: Changbin Du <changbin.du@intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Taeung Song <treeze.taeung@gmail.com>
Cc: Thomas Richter <tmricht@linux.ibm.com>
Cc: Yisheng Xie <xieyisheng1@huawei.com>
Link: http://lore.kernel.org/lkml/20200213064306.160480-4-ravi.bangoria@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf annotate --show-total-period does not really show total period.
The reason is we have two separate variables for the same purpose.
One is in symbol_conf.show_total_period and another is
annotation_options.show_total_period.
We save command line option in symbol_conf.show_total_period but uses
annotation_option.show_total_period while rendering tui/stdio2 browser.
Though, we copy symbol_conf.show_total_period to
annotation__default_options.show_total_period but that is not really
effective as we don't use annotation__default_options once we copy
default options to dynamic variable annotate.opts in cmd_annotate().
Instead of all these complication, keep only one variable and use it all
over. symbol_conf.show_total_period is used by perf report/top as well.
So let's kill annotation_options.show_total_period.
On a side note, I've kept annotation_options.show_total_period
definition because it's still used by perf-config code. Follow up patch
to fix perf-config for annotate will remove
annotation_options.show_total_period.
Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexey Budankov <alexey.budankov@linux.intel.com>
Cc: Changbin Du <changbin.du@intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Taeung Song <treeze.taeung@gmail.com>
Cc: Thomas Richter <tmricht@linux.ibm.com>
Cc: Yisheng Xie <xieyisheng1@huawei.com>
Link: http://lore.kernel.org/lkml/20200213064306.160480-3-ravi.bangoria@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Before this patch:
# ./perf test 39 41
39: LLVM search and compile :
39.1: Basic BPF llvm compile : Ok
39.2: kbuild searching : FAILED!
39.3: Compile source for BPF prologue generation : Skip
39.4: Compile source for BPF relocation : Skip
41: BPF filter :
41.1: Basic BPF filtering : Ok
41.2: BPF pinning : Ok
41.3: BPF prologue generation : FAILED!
41.4: BPF relocation checker : Skip
#
Using 'perf test -v' for these tests shows that it is not finding
uapi/linux/fs.h, which ends up being because we don't setup the right header
path. Fix it.
After this patch:
# perf test 39 41
39: LLVM search and compile :
39.1: Basic BPF llvm compile : Ok
39.2: kbuild searching : Ok
39.3: Compile source for BPF prologue generation : Ok
39.4: Compile source for BPF relocation : Ok
41: BPF filter :
41.1: Basic BPF filtering : Ok
41.2: BPF pinning : Ok
41.3: BPF prologue generation : Ok
41.4: BPF relocation checker : Ok
#
Longer description:
In llvm-utils.c we use some techniques to obtain the kbuild make
directives and that recently stopped working as now 'ar' gets called and
expects to find the dummy.o used to echo these variables:
$(NOSTDINC_FLAGS) $(LINUXINCLUDE) $(EXTRA_CFLAGS)
Add the $(CC) line to satisfy that, making sure this works with all
kernels, i.e. preserving the temp directory and files in it used for
this technique we can see that it works everywhere:
# make -s -C /lib/modules/5.4.18-100.fc30.x86_64/build M=/tmp/tmp.qgaFHgxjZ4/ clean
# ls -la /tmp/tmp.qgaFHgxjZ4/
total 4
drwx------. 2 root root 80 Feb 14 09:42 .
drwxrwxrwt. 47 root root 1200 Feb 14 09:42 ..
-rw-r--r--. 1 root root 0 Feb 13 17:14 dummy.c
-rw-r--r--. 1 root root 121 Feb 13 17:14 Makefile
#
# cat /tmp/tmp.qgaFHgxjZ4/Makefile
obj-y := dummy.o
$(obj)/%.o: $(src)/%.c
@echo -n "$(NOSTDINC_FLAGS) $(LINUXINCLUDE) $(EXTRA_CFLAGS)"
$(CC) -c -o $@ $<
#
Then build with an old kernel Makefile:
# make -s -C /lib/modules/5.4.18-100.fc30.x86_64/build M=/tmp/tmp.qgaFHgxjZ4/ dummy.o
-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/9/include -I./arch/x86/include -I./arch/x86/include/generated -I./include -I./arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I./include/uapi -I./include/generated/uapi -include ./include/linux/kconfig.h
#
# ls -la /tmp/tmp.qgaFHgxjZ4/
total 8
drwx------. 2 root root 100 Feb 14 09:43 .
drwxrwxrwt. 47 root root 1200 Feb 14 09:43 ..
-rw-r--r--. 1 root root 0 Feb 13 17:14 dummy.c
-rw-r--r--. 1 root root 936 Feb 14 09:43 dummy.o
-rw-r--r--. 1 root root 121 Feb 13 17:14 Makefile
#
And a new one:
# make -s -C /lib/modules/5.4.18-100.fc30.x86_64/build M=/tmp/tmp.qgaFHgxjZ4/ clean
# ls -la /tmp/tmp.qgaFHgxjZ4/
total 4
drwx------. 2 root root 80 Feb 14 09:43 .
drwxrwxrwt. 47 root root 1200 Feb 14 09:43 ..
-rw-r--r--. 1 root root 0 Feb 13 17:14 dummy.c
-rw-r--r--. 1 root root 121 Feb 13 17:14 Makefile
# make -s -C /lib/modules/5.6.0-rc1+/build M=/tmp/tmp.qgaFHgxjZ4/ dummy.o
-nostdinc -isystem /usr/lib/gcc/x86_64-redhat-linux/9/include -I/home/acme/git/linux/arch/x86/include -I./arch/x86/include/generated -I/home/acme/git/linux/include -I./include -I/home/acme/git/linux/arch/x86/include/uapi -I./arch/x86/include/generated/uapi -I/home/acme/git/linux/include/uapi -I./include/generated/uapi -include /home/acme/git/linux/include/linux/kconfig.h
#
# ls -la /tmp/tmp.qgaFHgxjZ4/
total 16
drwx------. 2 root root 160 Feb 14 09:44 .
drwxrwxrwt. 47 root root 1200 Feb 14 09:44 ..
-rw-r--r--. 1 root root 158 Feb 14 09:44 built-in.a
-rw-r--r--. 1 root root 149 Feb 14 09:44 .built-in.a.cmd
-rw-r--r--. 1 root root 0 Feb 13 17:14 dummy.c
-rw-r--r--. 1 root root 936 Feb 14 09:44 dummy.o
-rw-r--r--. 1 root root 121 Feb 13 17:14 Makefile
-rw-r--r--. 1 root root 0 Feb 14 09:44 modules.order
#
Reported-by: Thomas Richter <tmricht@linux.ibm.com>
Tested-by: Thomas Richter <tmricht@linux.ibm.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Masahiro Yamada <masahiroy@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Sumanth Korikkar <sumanthk@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: Zefan Li <lizefan@huawei.com>
Link: https://www.spinics.net/lists/linux-perf-users/msg10600.html
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
For data collected on machines with front end stalled cycles supported,
such as found on modern AMD CPU families, commit 146540fb54 ("perf
stat: Always separate stalled cycles per insn") introduces a new line in
CSV output with a leading comma that upsets some automated scripts.
Scripts have to use "-e ex_ret_instr" to work around this issue, after
upgrading to a version of perf with that commit.
We could add "if (have_frontend_stalled && !config->csv_sep)" to the not
(total && avg) else clause, to emphasize that CSV users are usually
scripts, and are written to do only what is needed, i.e., they wouldn't
typically invoke "perf stat" without specifying an explicit event list.
But - let alone CSV output - why should users now tolerate a constant
0-reporting extra line in regular terminal output?:
BEFORE:
$ sudo perf stat --all-cpus -einstructions,cycles -- sleep 1
Performance counter stats for 'system wide':
181,110,981 instructions # 0.58 insn per cycle
# 0.00 stalled cycles per insn
309,876,469 cycles
1.002202582 seconds time elapsed
The user would not like to see the now permanent:
"0.00 stalled cycles per insn"
line fixture, as it gives no useful information.
So this patch removes the printing of the zeroed stalled cycles line
altogether, almost reverting the very original commit fb4605ba47
("perf stat: Check for frontend stalled for metrics"), which seems like
it was written to normalize --metric-only column output of common Intel
machines at the time: modern Intel machines have ceased to support the
genericised frontend stalled metrics AFAICT.
AFTER:
$ sudo perf stat --all-cpus -einstructions,cycles -- sleep 1
Performance counter stats for 'system wide':
244,071,432 instructions # 0.69 insn per cycle
355,353,490 cycles
1.001862516 seconds time elapsed
Output behaviour when stalled cycles is indeed measured is not affected
(BEFORE == AFTER):
$ sudo perf stat --all-cpus -einstructions,cycles,stalled-cycles-frontend -- sleep 1
Performance counter stats for 'system wide':
247,227,799 instructions # 0.63 insn per cycle
# 0.26 stalled cycles per insn
394,745,636 cycles
63,194,485 stalled-cycles-frontend # 16.01% frontend cycles idle
1.002079770 seconds time elapsed
Fixes: 146540fb54 ("perf stat: Always separate stalled cycles per insn")
Signed-off-by: Kim Phillips <kim.phillips@amd.com>
Acked-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Acked-by: Song Liu <songliubraving@fb.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20200207230613.26709-1-kim.phillips@amd.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
`tools/perf/util/map.c` has a function named `maps__insert` that
acquires a write lock if its in multithread context.
Even though this lock is released when function successfully completes,
there's a branch that is executed when `maps_by_name == NULL` that
returns from this function without releasing the write lock.
Added an `up_write` to release the lock when this happens.
Fixes: a7c2b572e2 ("perf map_groups: Auto sort maps by name, if needed")
Signed-off-by: Cengiz Can <cengiz@kernel.wtf>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: http://lore.kernel.org/lkml/20200120141553.23934-1-cengiz@kernel.wtf
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Kernel commit 88903c4643 ("tracing/probe: Add ustring type for user-space string")
adds support for user-space strings when type 'ustring' is specified.
Here is an example using sysfs command line interface
for kprobes:
Function to probe:
struct filename *
getname_flags(const char __user *filename, int flags, int *empty)
Setup:
# cd /sys/kernel/debug/tracing/
# echo 'p:tmr1 getname_flags +0(%r2):ustring' > kprobe_events
# cat events/kprobes/tmr1/format | fgrep print
print fmt: "(%lx) arg1=\"%s\"", REC->__probe_ip, REC->arg1
# echo 1 > events/kprobes/tmr1/enable
# touch /tmp/111
# echo 0 > events/kprobes/tmr1/enable
# cat trace|fgrep /tmp/111
touch-5846 [005] d..2 255520.717960: tmr1:\
(getname_flags+0x0/0x400) arg1="/tmp/111"
Doing the same with the perf tool fails.
Using type 'string' succeeds:
# perf probe "vfs_getname=getname_flags:72 pathname=filename:string"
Added new event:
probe:vfs_getname (on getname_flags:72 with pathname=filename:string)
....
# perf probe -d probe:vfs_getname
Removed event: probe:vfs_getname
However using type 'ustring' fails (output before):
# perf probe "vfs_getname=getname_flags:72 pathname=filename:ustring"
Failed to write event: Invalid argument
Error: Failed to add events.
#
Fix this by adding type 'ustring' in function
convert_variable_type().
Using ustring succeeds (output after):
# ./perf probe "vfs_getname=getname_flags:72 pathname=filename:ustring"
Added new event:
probe:vfs_getname (on getname_flags:72 with pathname=filename:ustring)
You can now use it in all perf tools, such as:
perf record -e probe:vfs_getname -aR sleep 1
#
Note: This issue also exists on x86, it is not s390 specific.
Signed-off-by: Thomas Richter <tmricht@linux.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: sumanthk@linux.ibm.com
Link: http://lore.kernel.org/lkml/20200120132011.64698-2-tmricht@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf with CoreSight fails to record trace data with command:
perf record -e cs_etm/@tmc_etr0/u --per-thread ls
failed to set sink "" on event cs_etm/@tmc_etr0/u with 21 (Is a
directory)/perf/
This failure is root caused with the commit 1dc925568f ("perf
parse: Add a deep delete for parse event terms").
The log shows, cs_etm fails to parse the sink attribution; cs_etm event
relies on the event configuration to pass sink name, but the event
specific configuration data cannot be passed properly with flow:
get_config_terms()
ADD_CONFIG_TERM(DRV_CFG, term->val.str);
__t->val.str = term->val.str;
`> __t->val.str is assigned to term->val.str;
parse_events_terms__purge()
parse_events_term__delete()
zfree(&term->val.str);
`> term->val.str is freed and assigned to NULL pointer;
cs_etm_set_sink_attr()
sink = __t->val.str;
`> sink string has been freed.
To fix this issue, in the function get_config_terms(), this patch
changes to use strdup() for allocation a new duplicate string rather
than directly assignment string pointer.
This patch addes a new field 'free_str' in the data structure
perf_evsel_config_term; 'free_str' is set to true when the union is used
as a string pointer; thus it can tell perf_evsel__free_config_terms() to
free the string.
Fixes: 1dc925568f ("perf parse: Add a deep delete for parse event terms")
Suggested-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Leo Yan <leo.yan@linaro.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Mike Leach <mike.leach@linaro.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Suzuki Poulouse <suzuki.poulose@arm.com>
Cc: linux-arm-kernel@lists.infradead.org
Link: http://lore.kernel.org/lkml/20200117055251.24058-2-leo.yan@linaro.org
[ Use zfree() in perf_evsel__free_config_terms ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
:# modified: tools/perf/util/evsel_config.h
Using .st_ctime clobbers the timestamp information in perf report header
whenever any operation is done with the file. Even tar-ing and untar-ing
the perf.data file (which preserves the file last modification timestamp)
doesn't prevent that:
[Michael@Diego tmp]$ ls -l perf.data
-> -rw-------. 1 Michael Michael 169888 Dec 2 15:23 perf.data
[Michael@Diego tmp]$ perf report --header-only
# ========
-> # captured on : Mon Dec 2 15:23:42 2019
[...]
[Michael@Diego tmp]$ tar c perf.data | xz > perf.data.tar.xz
[Michael@Diego tmp]$ mkdir aaa
[Michael@Diego tmp]$ cd aaa
[Michael@Diego aaa]$ xzcat ../perf.data.tar.xz | tar x
[Michael@Diego aaa]$ ls -l -a
total 172
drwxrwxr-x. 2 Michael Michael 23 Jan 14 11:26 .
drwxrwxr-x. 6 Michael Michael 4096 Jan 14 11:26 ..
-> -rw-------. 1 Michael Michael 169888 Dec 2 15:23 perf.data
[Michael@Diego aaa]$ perf report --header-only
# ========
-> # captured on : Tue Jan 14 11:26:16 2020
[...]
When using .st_mtime instead, correct information is printed:
[Michael@Diego aaa]$ ~/acme/tools/perf/perf report --header-only
# ========
-> # captured on : Mon Dec 2 15:23:42 2019
[...]
Signed-off-by: Michael Petlan <mpetlan@redhat.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
LPU-Reference: 20200114104236.31555-1-mpetlan@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The objdump utility has useful --prefix / --prefix-strip options to
allow changing source code file names hardcoded into executables' debug
info. Add options to 'perf report', 'perf top' and 'perf annotate',
which are then passed to objdump.
$ mkdir foo
$ echo 'main() { for (;;); }' > foo/foo.c
$ gcc -g foo/foo.c
foo/foo.c:1:1: warning: return type defaults to ‘int’ [-Wimplicit-int]
1 | main() { for (;;); }
| ^~~~
$ perf record ./a.out
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.230 MB perf.data (5721 samples) ]
$ mv foo bar
$ perf annotate
<does not show source code>
$ perf annotate --prefix=/home/ak/lsrc/git/bar --prefix-strip=5
<does show source code>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Tested-by: Jiri Olsa <jolsa@redhat.com>
LPU-Reference: 20200107210444.214071-1-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Use struct mmap_cpu_mask type for the tool's thread and mmap data
buffers to overcome current 1024 CPUs mask size limitation of cpu_set_t
type.
Currently glibc's cpu_set_t type has an internal mask size limit of 1024
CPUs.
Moving to the 'struct mmap_cpu_mask' type allows overcoming that limit.
The tools bitmap API is used to manipulate objects of 'struct mmap_cpu_mask'
type.
Committer notes:
To print the 'nbits' struct member we must use %zd, since it is a
size_t, this fixes the build in some toolchains/arches.
Reported-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Alexey Budankov <alexey.budankov@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/96d7e2ff-ce8b-c1e0-d52c-aa59ea96f0ea@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When a map is create to represent the main kernel area (vmlinux) with
map__new2() we allocate an extra area to store a pointer to the 'struct
maps' for the kernel maps, so that we can access that struct when
loading ELF files or kallsyms, as we will need to split it in multiple
maps, one per kernel module or ELF section (such as ".init.text").
So when map->dso->kernel is non-zero, it is expected that
map__kmap(map)->kmaps to be set to the tree of kernel maps (modules,
chunks of the main kernel, bpf progs put in place via
PERF_RECORD_KSYMBOL, the main kernel).
This was not the case when we were splitting the main kernel into chunks
for its ELF sections, which ended up making 'perf report --children'
processing a perf.data file with callchains to trip on
__map__is_kernel(), when we press ENTER to see the popup menu for main
histogram entries that starts at a symbol in the ".init.text" ELF
section, e.g.:
- 8.83% 0.00% swapper [kernel.vmlinux].init.text [k] start_kernel
start_kernel
cpu_startup_entry
do_idle
cpuidle_enter
cpuidle_enter_state
intel_idle
Fix it.
Reviewed-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/lkml/20191218190120.GB13282@kernel.org/
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
