Tail call optimizations can remove stack frames that are used in
unwinding tests. Add an attribute that can be used to disable the tail
call optimization. Tested on clang and GCC.
Committer notes:
Old versions of clang don't like that __attribute__((optimize)), so add
an ifdef to make it go away.
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: clang-built-linux@googlegroups.com
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Quentin Monnet <quentin@isovalent.com>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lore.kernel.org/lkml/20200530082015.39162-2-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
So that when one runs:
$ make -C tools/perf build-test
We make sure that recent changes don't break that opt-in build.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Alexey Budankov <alexey.budankov@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Andrii Nakryiko <andriin@fb.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Ian Rogers <irogers@google.com>
Cc: Igor Lubashev <ilubashe@akamai.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Jiwei Sun <jiwei.sun@windriver.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Martin KaFai Lau <kafai@fb.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Yonghong Song <yhs@fb.com>
Cc: yuzhoujian <yuzhoujian@didichuxing.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This patch links perf with the libpfm4 library if it is available and
LIBPFM4 is passed to the build. The libpfm4 library contains hardware
event tables for all processors supported by perf_events. It is a helper
library that helps convert from a symbolic event name to the event
encoding required by the underlying kernel interface. This library is
open-source and available from: http://perfmon2.sf.net.
With this patch, it is possible to specify full hardware events by name.
Hardware filters are also supported. Events must be specified via the
--pfm-events and not -e option. Both options are active at the same time
and it is possible to mix and match:
$ perf stat --pfm-events inst_retired:any_p:c=1:i -e cycles ....
One needs to explicitely ask for its inclusion by using the LIBPFM4 make
command line option, ie its opt-in rather than opt-out of feature
detection and build support.
Signed-off-by: Stephane Eranian <eranian@google.com>
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Alexey Budankov <alexey.budankov@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Andrii Nakryiko <andriin@fb.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Igor Lubashev <ilubashe@akamai.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Jiwei Sun <jiwei.sun@windriver.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Martin KaFai Lau <kafai@fb.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Yonghong Song <yhs@fb.com>
Cc: bpf@vger.kernel.org
Cc: netdev@vger.kernel.org
Cc: yuzhoujian <yuzhoujian@didichuxing.com>
Link: http://lore.kernel.org/lkml/20200505182943.218248-2-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This header is part of the jsmn JSON parser, introduced in 867a979a83.
Correct the SPDX tag to indicate that it is under the MIT license.
Signed-off-by: Ed Maste <emaste@freebsd.org>
Acked-by: Andi Kleen <ak@linux.intel.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: http://lore.kernel.org/lkml/20200528170858.48457-1-emaste@freefall.freebsd.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Fix an issue where addresses in the DWARF line table are offset by -0x40
(GEN_ELF_TEXT_OFFSET). This can be seen with `objdump -S` on the ELF
files after perf inject.
Committer notes:
Ian added this in his Acked-by reply:
---
Without too much knowledge this looks good to me. The original code came
from oprofile's jit support:
https://sourceforge.net/p/oprofile/oprofile/ci/master/tree/opjitconv/debug_line.c#l325
---
Signed-off-by: Nick Gasson <nick.gasson@arm.com>
Acked-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lore.kernel.org/lkml/20200528051916.6722-1-nick.gasson@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
For each PC/BCI pair in the JVMTI compiler inlining record table, the
jitdump plugin emits debug line table entries for every source line in
the method preceding that BCI. Instead only emit one source line per
PC/BCI pair. Reported by Ian Rogers. This reduces the .dump size for
SPECjbb from ~230MB to ~40MB.
Signed-off-by: Nick Gasson <nick.gasson@arm.com>
Acked-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lore.kernel.org/lkml/20200528054049.13662-1-nick.gasson@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We forgot to add it, so one would have to explicitely ask for it to be
run, fix that by adding it to the set of tests that are performed by
default when one does:
$ make -C tools/perf build-test
It was being exercised only in the make_minimal test, this patch makes
it be tested in isolation, i.e. disabling only this feature.
Fixes: e26e63be64 ("perf build: Add sdt feature detection")
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We forgot to add it, so one would have to explicitely ask for it to be
run, fix that by adding it to the set of tests that are performed by
default when one does:
$ make -C tools/perf build-test
It was being exercised only in the make_minimal test, this patch makes
it be tested in isolation, i.e. disabling only this feature.
Fixes: 8ee4646038 ("perf build: Add libcrypto feature detection")
Cc: Stephane Eranian <eranian@google.com>
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>
So that we make sure that even on x86-64 and other architectures where
that is the default method we test build the fallback to libaudit that
other architectures use.
I.e. now this line got added to:
$ make -C tools/perf build-test
<SNIP>
make_no_syscall_tbl_O: cd . && make NO_SYSCALL_TABLE=1 FEATURES_DUMP=/home/acme/git/perf/tools/perf/BUILD_TEST_FEATURE_DUMP -j12 O=/tmp/tmp.W0HtKR1mfr DESTDIR=/tmp/tmp.lNezgCVPzW
<SNIP>
$
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Ingo reported that the libaudit was always appearing as OFF:
Auto-detecting system features:
... dwarf: [ on ]
... dwarf_getlocations: [ on ]
... glibc: [ on ]
... gtk2: [ on ]
... libaudit: [ OFF ]
And everything seemed to work, i.e. we were checking for a feature that
we don't use, causing confusion for people building perf, so work to
remove that nuisance while making sure that it works when an arch
doesn't provide the alternative method to generate the syscall id/name
conversion tables.
Longer explanation of the new modus operandi:
$ make -C tools/perf O=/tmp/build/perf NO_SYSCALL_TABLE=1
<SNIP>
Auto-detecting system features:
... dwarf: [ on ]
... dwarf_getlocations: [ on ]
... glibc: [ on ]
... gtk2: [ on ]
... libbfd: [ on ]
... libcap: [ on ]
... libelf: [ on ]
... libnuma: [ on ]
... numa_num_possible_cpus: [ on ]
... libperl: [ on ]
... libpython: [ on ]
... libcrypto: [ on ]
... libunwind: [ on ]
... libdw-dwarf-unwind: [ on ]
... zlib: [ on ]
... lzma: [ on ]
... get_cpuid: [ on ]
... bpf: [ on ]
... libaio: [ on ]
... libzstd: [ on ]
... disassembler-four-args: [ on ]
Makefile.config:665: No libaudit.h found, disables 'trace' tool, please install audit-libs-devel or libaudit-dev
GEN /tmp/build/perf/common-cmds.h
MKDIR /tmp/build/perf/fd/
MKDIR /tmp/build/perf/fs/
<SNIP>
$
The libaudit test is forced and it fails when audit-libs-devel isn't available:
$ cat /tmp/build/perf/feature/test-libaudit.make.output
test-libaudit.c:2:10: fatal error: libaudit.h: No such file or directory
2 | #include <libaudit.h>
| ^~~~~~~~~~~~
compilation terminated.
$
If we install audit-libs-devel and rebuild it continues not to be shown as OFF
in the main auto-detection summary, but again gets tested and this time:
$ rpm -q audit-libs-devel
audit-libs-devel-3.0-0.15.20191104git1c2f876.fc31.x86_64
$
The make output for the feature detection comes clean:
$ cat /tmp/build/perf/feature/test-libaudit.make.output
And the feature detection binary is successfully built and is dynamicly linked
with libaudit:
$ ldd /tmp/build/perf/feature/test-libaudit.bin | grep audit
libaudit.so.1 => /lib64/libaudit.so.1 (0x00007f5bf5177000)
$
As well as the resulting perf binary:
$ ldd /tmp/build/perf/perf | grep audit
libaudit.so.1 => /lib64/libaudit.so.1 (0x00007fad511c7000)
$
And 'perf trace' works using the libaudit method:
$ sudo /tmp/build/perf/perf trace -e nanosleep sleep 1
0.000 (1000.067 ms): sleep/281872 nanosleep(rqtp: 0x7ffedbbe69d0) = 0
$
If we leave audit-libs-devel installed but don't disable the use of the best
method, the one using SYSCALL_TABLE, the default for architectures that provide
the script to build the syscall id/name mapping using the .tbl files copied
from the kernel sources, we get:
$ rm -rf /tmp/build/perf ; mkdir -p /tmp/build/perf
$ make -C tools/perf O=/tmp/build/perf
Auto-detecting system features:
... dwarf: [ on ]
... dwarf_getlocations: [ on ]
... glibc: [ on ]
... gtk2: [ on ]
... libbfd: [ on ]
... libcap: [ on ]
... libelf: [ on ]
... libnuma: [ on ]
... numa_num_possible_cpus: [ on ]
... libperl: [ on ]
... libpython: [ on ]
... libcrypto: [ on ]
... libunwind: [ on ]
... libdw-dwarf-unwind: [ on ]
... zlib: [ on ]
... lzma: [ on ]
... get_cpuid: [ on ]
... bpf: [ on ]
... libaio: [ on ]
... libzstd: [ on ]
... disassembler-four-args: [ on ]
GEN /tmp/build/perf/common-cmds.h
<SNIP>
$
Again, no mention of libaudit being on or OFF and:
$ cat /tmp/build/perf/feature/test-libaudit.make.output
cat: /tmp/build/perf/feature/test-libaudit.make.output: No such file or directory
$
We didn't even bother checking for its availability, slightly speeding up the
build process and:
$ ldd /tmp/build/perf/perf | grep libaudit
$
We don't link with it, also:
$ sudo /tmp/build/perf/perf trace -e nanosleep sleep 1
0.000 (1000.053 ms): sleep/299125 nanosleep(rqtp: 0x7ffc24611b50) = 0
$
And globs become available:
$ sudo /tmp/build/perf/perf trace -e *sleep sleep 1
0.000 (1000.072 ms): sleep/299136 nanosleep(rqtp: 0x7ffe7a3c4ff0) = 0
$
Reported-by: Ingo Molnar <mingo@kernel.org>
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>
The audit-libs API doesn't provide a way to figure out what is the
syscall with the greatest number/id, take that into account when using
that method to go on growing the syscall table as we the syscalls go on
appearing on the radar.
With this the libaudit based method is back working, i.e. when building
with:
$ make NO_SYSCALL_TABLE=1 O=/tmp/build/perf -C tools/perf install-bin
<SNIP>
Auto-detecting system features:
<SNIP>
... libaudit: [ on ]
... libbfd: [ on ]
... libcap: [ on ]
<SNIP>
$ ldd ~/bin/perf | grep audit
libaudit.so.1 => /lib64/libaudit.so.1 (0x00007faef22df000)
$
perf trace is back working, which makes it functional in arches other
than x86_64, powerpc, arm64 and s390, that provides these generators:
$ find tools/perf/arch/ -name "*syscalltbl*"
tools/perf/arch/x86/entry/syscalls/syscalltbl.sh
tools/perf/arch/arm64/entry/syscalls/mksyscalltbl
tools/perf/arch/s390/entry/syscalls/mksyscalltbl
tools/perf/arch/powerpc/entry/syscalls/mksyscalltbl
$
Example output forcing the libaudit method on x86_64:
# perf trace -e file,nanosleep sleep 0.001
? ( ): sleep/859090 ... [continued]: execve()) = 0
0.045 ( 0.005 ms): sleep/859090 access(filename: 0x8733e850, mode: R) = -1 ENOENT (No such file or directory)
0.055 ( 0.005 ms): sleep/859090 openat(dfd: CWD, filename: 0x8733ba29, flags: RDONLY|CLOEXEC) = 3
0.079 ( 0.005 ms): sleep/859090 openat(dfd: CWD, filename: 0x87345d20, flags: RDONLY|CLOEXEC) = 3
0.085 ( 0.002 ms): sleep/859090 read(fd: 3, buf: 0x7ffd9d483f58, count: 832) = 832
0.090 ( 0.002 ms): sleep/859090 read(fd: 3, buf: 0x7ffd9d483b50, count: 784) = 784
0.094 ( 0.002 ms): sleep/859090 read(fd: 3, buf: 0x7ffd9d483b20, count: 32) = 32
0.098 ( 0.002 ms): sleep/859090 read(fd: 3, buf: 0x7ffd9d483ad0, count: 68) = 68
0.109 ( 0.002 ms): sleep/859090 read(fd: 3, buf: 0x7ffd9d483a50, count: 784) = 784
0.113 ( 0.002 ms): sleep/859090 read(fd: 3, buf: 0x7ffd9d483730, count: 32) = 32
0.117 ( 0.002 ms): sleep/859090 read(fd: 3, buf: 0x7ffd9d483710, count: 68) = 68
0.320 ( 0.008 ms): sleep/859090 openat(dfd: CWD, filename: 0x872c3660, flags: RDONLY|CLOEXEC) = 3
0.372 ( 1.057 ms): sleep/859090 nanosleep(rqtp: 0x7ffd9d484ac0) = 0
#
There are still some limitations when using the libaudit method, that
will be fixed at some point, i.e., this works with the mksyscalltbl
method but not with libaudit's:
# perf trace -e file,*sleep sleep 0.001
event syntax error: '*sleep'
\___ parser error
Run 'perf list' for a list of valid events
Usage: perf trace [<options>] [<command>]
or: perf trace [<options>] -- <command> [<options>]
or: perf trace record [<options>] [<command>]
or: perf trace record [<options>] -- <command> [<options>]
-e, --event <event> event/syscall selector. use 'perf list' to list available events
#
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
In the past this wasn't needed as the libaudit based code would use just
one field, and the alternative constructor would fill in all the fields,
but now that even when using the libaudit based method we need the other
fields, switch to zalloc() to make sure the other fields are zeroed at
instantiation time.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When we moved to a syscalltbl generated from the kernel syscall tables
(arch/..../syscall*.tbl) the idea was to either use it, when having the
generator (e.g. tools/perf/arch/x86/entry/syscalls/syscalltbl.sh), or
falling back to the previous audit-libs based way of mapping syscall ids
to strings and the other way around.
At first we just needed the audit_detect_machine() return to then use it
to the str->id/id->str, or the other fields for the now used by default
in the most well developed arches method of using the syscall table
generator.
The problem is that then the libaudit code fell into disrepair, and
architectures where it is the method used are not working.
Now, with NO_SYSCALL_TABLE=1 being possible to pass on the make command
line we can automate the testing of that method even on x86-64, arm64,
etc.
And doing it I noted that we actually use fields in both entries in the
union, oops, so ditch the union, as we need all those fields at the same
time.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This is useful to see if, on x86, the legacy libaudit still works, as it
is used in architectures that don't have the SYSCALL_TABLE logic and we
want to have it tested in 'make -C tools/perf/ build-test'.
E.g.:
Without having audit-libs-devel installed:
$ make NO_SYSCALL_TABLE=1 O=/tmp/build/perf -C tools/perf install-bin
make: Entering directory '/home/acme/git/perf/tools/perf'
BUILD: Doing 'make -j12' parallel build
<SNIP>
Auto-detecting system features:
<SNIP>
... libaudit: [ OFF ]
... libbfd: [ on ]
... libcap: [ on ]
<SNIP>
Makefile.config:664: No libaudit.h found, disables 'trace' tool, please install audit-libs-devel or libaudit-dev
<SNIP>
After installing it:
$ rm -rf /tmp/build/perf ; mkdir -p /tmp/build/perf
$ time make NO_SYSCALL_TABLE=1 O=/tmp/build/perf -C tools/perf install-bin ; perf test python
make: Entering directory '/home/acme/git/perf/tools/perf'
BUILD: Doing 'make -j12' parallel build
HOSTCC /tmp/build/perf/fixdep.o
HOSTLD /tmp/build/perf/fixdep-in.o
LINK /tmp/build/perf/fixdep
Warning: Kernel ABI header at 'tools/arch/x86/include/asm/msr-index.h' differs from latest version at 'arch/x86/include/asm/msr-index.h'
diff -u tools/arch/x86/include/asm/msr-index.h arch/x86/include/asm/msr-index.h
Warning: Kernel ABI header at 'tools/perf/util/hashmap.h' differs from latest version at 'tools/lib/bpf/hashmap.h'
diff -u tools/perf/util/hashmap.h tools/lib/bpf/hashmap.h
Warning: Kernel ABI header at 'tools/perf/util/hashmap.c' differs from latest version at 'tools/lib/bpf/hashmap.c'
diff -u tools/perf/util/hashmap.c tools/lib/bpf/hashmap.c
Auto-detecting system features:
<SNIP>
... libaudit: [ on ]
... libbfd: [ on ]
... libcap: [ on ]
<SNIP>
$ ldd ~/bin/perf | grep audit
libaudit.so.1 => /lib64/libaudit.so.1 (0x00007fc18978e000)
$
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: http://lore.kernel.org/lkml/20200529155552.463-3-acme@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
To help in allowing to disable it from the make command line.
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: http://lore.kernel.org/lkml/20200529155552.463-2-acme@kernel.org
[ Fixed the logic for the filter part, it should be ifeq, not ifneq ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Stop the message displaying when user space is not being traced.
Example:
Prerequisites:
sudo setcap "cap_sys_rawio,cap_sys_admin,cap_sys_ptrace,cap_syslog,cap_ipc_lock=ep" ~/bin/perf
sudo chmod +r /proc/kcore
Before:
$ perf record --no-switch-events --kcore -a -e intel_pt//k -- sleep 0.001
Warning:
Intel Processor Trace decoding will not be possible except for kernel tracing!
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.838 MB perf.data ]
After:
$ perf record --no-switch-events --kcore -a -e intel_pt//k -- sleep 0.001
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 1.068 MB perf.data ]
$ sudo chmod go-r /proc/kcore
$ sudo setcap -r ~/bin/perf
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: http://lore.kernel.org/lkml/20200528120859.21604-2-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Context switch events are added automatically by Intel PT and Coresight.
Make it possible to suppress them. That is useful for tracing the
scheduler without the disturbance that the switch event processing
creates.
Example:
Prerequisites:
$ which perf
~/bin/perf
$ sudo setcap "cap_sys_rawio,cap_sys_admin,cap_sys_ptrace,cap_syslog,cap_ipc_lock=ep" ~/bin/perf
$ sudo chmod +r /proc/kcore
Before:
$ perf record --no-switch-events --kcore -a -e intel_pt//k -- sleep 0.001
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.938 MB perf.data ]
$ perf script -D | grep PERF_RECORD_SWITCH | wc -l
572
After:
$ perf record --no-switch-events --kcore -a -e intel_pt//k -- sleep 0.001
Warning:
Intel Processor Trace decoding will not be possible except for kernel tracing!
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.838 MB perf.data ]
$ perf script -D | grep PERF_RECORD_SWITCH | wc -l
0
$ sudo chmod go-r /proc/kcore
$ sudo setcap -r ~/bin/perf
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: http://lore.kernel.org/lkml/20200528120859.21604-1-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Events marked as 'immediate' are started before other events to ensure
that there is context at the start of the main tracing events. The same
is true at the end of tracing, so disable 'immediate' events after other
events.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt (VMware) <rostedt@goodmis.org>
Cc: x86@kernel.org
Link: http://lore.kernel.org/lkml/20200512121922.8997-11-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
In the absence of any modules, no "modules" map is created, but there
are other executable pages to map, due to eBPF JIT, kprobe or ftrace.
Map them by recognizing that the first "module" symbol is not
necessarily from a module, and adjust the map accordingly.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt (VMware) <rostedt@goodmis.org>
Cc: x86@kernel.org
Link: http://lore.kernel.org/lkml/20200512121922.8997-10-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
For a Java method signature like:
Ljava/lang/AbstractStringBuilder;appendChars(Ljava/lang/String;II)V
The demangler produces:
void class java.lang.AbstractStringBuilder.appendChars(class java.lang., shorttring., int, int)
The arguments should be (java.lang.String, int, int) but the demangler
interprets the "S" in String as the type code for "short". Correct this
and two other minor things:
- There is no "bool" type in Java, should be "boolean".
- The demangler prepends "class" to every Java class name. This is not
standard Java syntax and it wastes a lot of horizontal space if the
signature is long. Remove this as there isn't any ambiguity between
class names and primitives.
Committer notes:
This was split from a larger patch that also added a java demangler
'perf test' entry, that, before this patch shows the error being fixed
by it:
$ perf test java
65: Demangle Java : FAILED!
$ perf test -v java
Couldn't bump rlimit(MEMLOCK), failures may take place when creating BPF maps, etc
65: Demangle Java :
--- start ---
test child forked, pid 307264
FAILED: Ljava/lang/StringLatin1;equals([B[B)Z: bool class java.lang.StringLatin1.equals(byte[], byte[]) != boolean java.lang.StringLatin1.equals(byte[], byte[])
FAILED: Ljava/util/zip/ZipUtils;CENSIZ([BI)J: long class java.util.zip.ZipUtils.CENSIZ(byte[], int) != long java.util.zip.ZipUtils.CENSIZ(byte[], int)
FAILED: Ljava/util/regex/Pattern$BmpCharProperty;match(Ljava/util/regex/Matcher;ILjava/lang/CharSequence;)Z: bool class java.util.regex.Pattern$BmpCharProperty.match(class java.util.regex.Matcher., int, class java.lang., charhar, shortequence) != boolean java.util.regex.Pattern$BmpCharProperty.match(java.util.regex.Matcher, int, java.lang.CharSequence)
FAILED: Ljava/lang/AbstractStringBuilder;appendChars(Ljava/lang/String;II)V: void class java.lang.AbstractStringBuilder.appendChars(class java.lang., shorttring., int, int) != void java.lang.AbstractStringBuilder.appendChars(java.lang.String, int, int)
FAILED: Ljava/lang/Object;<init>()V: void class java.lang.Object<init>() != void java.lang.Object<init>()
test child finished with -1
---- end ----
Demangle Java: FAILED!
$
After applying this patch:
$ perf test java
65: Demangle Java : Ok
$
Signed-off-by: Nick Gasson <nick.gasson@arm.com>
Reviewed-by: Ian Rogers <irogers@google.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Tested-by: Ian Rogers <irogers@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20200427061520.24905-4-nick.gasson@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Split from a larger patch that was also fixing a problem with the java
demangler, so, before applying that patch we see:
$ perf test java
65: Demangle Java : FAILED!
$ perf test -v java
65: Demangle Java :
--- start ---
test child forked, pid 307264
FAILED: Ljava/lang/StringLatin1;equals([B[B)Z: bool class java.lang.StringLatin1.equals(byte[], byte[]) != boolean java.lang.StringLatin1.equals(byte[], byte[])
FAILED: Ljava/util/zip/ZipUtils;CENSIZ([BI)J: long class java.util.zip.ZipUtils.CENSIZ(byte[], int) != long java.util.zip.ZipUtils.CENSIZ(byte[], int)
FAILED: Ljava/util/regex/Pattern$BmpCharProperty;match(Ljava/util/regex/Matcher;ILjava/lang/CharSequence;)Z: bool class java.util.regex.Pattern$BmpCharProperty.match(class java.util.regex.Matcher., int, class java.lang., charhar, shortequence) != boolean java.util.regex.Pattern$BmpCharProperty.match(java.util.regex.Matcher, int, java.lang.CharSequence)
FAILED: Ljava/lang/AbstractStringBuilder;appendChars(Ljava/lang/String;II)V: void class java.lang.AbstractStringBuilder.appendChars(class java.lang., shorttring., int, int) != void java.lang.AbstractStringBuilder.appendChars(java.lang.String, int, int)
FAILED: Ljava/lang/Object;<init>()V: void class java.lang.Object<init>() != void java.lang.Object<init>()
test child finished with -1
---- end ----
Demangle Java: FAILED!
$
Next patch should fix this.
Signed-off-by: Nick Gasson <nick.gasson@arm.com>
Reviewed-by: Ian Rogers <irogers@google.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Tested-by: Ian Rogers <irogers@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20200427061520.24905-4-nick.gasson@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
If the Java sources are compiled with -g:none to disable debug
information the perf JVMTI plugin reports a lot of errors like:
java: GetLineNumberTable failed with JVMTI_ERROR_ABSENT_INFORMATION
java: GetLineNumberTable failed with JVMTI_ERROR_ABSENT_INFORMATION
java: GetLineNumberTable failed with JVMTI_ERROR_ABSENT_INFORMATION
java: GetLineNumberTable failed with JVMTI_ERROR_ABSENT_INFORMATION
java: GetLineNumberTable failed with JVMTI_ERROR_ABSENT_INFORMATION
Instead if GetLineNumberTable returns JVMTI_ERROR_ABSENT_INFORMATION
simply skip emitting line number information for that method. Unlike the
previous patch these errors don't affect the jitdump generation, they
just generate a lot of noise.
Similarly for native methods which also don't have line tables.
Signed-off-by: Nick Gasson <nick.gasson@arm.com>
Reviewed-by: Ian Rogers <irogers@google.com>
Tested-by: Ian Rogers <irogers@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20200427061520.24905-3-nick.gasson@arm.com
[ Moved || operator to the end of the line, not at the start of 2nd if condition ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
If a Java class is compiled with -g:none to omit debug information, the
JVMTI plugin won't write jitdump entries for any method in this class
and prints a lot of errors like:
java: GetSourceFileName failed with JVMTI_ERROR_ABSENT_INFORMATION
The call to GetSourceFileName is used to derive the file name `fn`, but
this value is not actually used since commit ca58d7e64b ("perf jvmti:
Generate correct debug information for inlined code") which moved the
file name lookup into fill_source_filenames(). So the call to
GetSourceFileName and related code can be safely removed.
Signed-off-by: Nick Gasson <nick.gasson@arm.com>
Reviewed-by: Ian Rogers <irogers@google.com>
Tested-by: Ian Rogers <irogers@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20200427061520.24905-2-nick.gasson@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We need to pass more data to the scanner so let's start with having it
to take pointer to 'struct parse_events_state' object instead of just
start token.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Michael Petlan <mpetlan@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lore.kernel.org/lkml/20200524224219.234847-4-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
There's no need to pass the given evsel's count to metric data, because
it will be pushed again within the following metric_events loop.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Michael Petlan <mpetlan@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lore.kernel.org/lkml/20200524224219.234847-3-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Before:
Usage: perf list [<options>] [hw|sw|cache|tracepoint|pmu|sdt|event_glob]
After:
Usage: perf list [<options>] [hw|sw|cache|tracepoint|pmu|sdt|metric|metricgroup|event_glob]
Committer testing:
Before and after we get these outputs on a Lenovo t480s (i7-8650U):
# perf list metricgroup
List of pre-defined events (to be used in -e):
Metric Groups:
BrMispredicts
BrMispredicts_SMT
Branches
Cache_Misses
DSB
FLOPS
FLOPS_SMT
Fetch_BW
IcMiss
Instruction_Type
Memory_BW
Memory_Bound
Memory_Lat
No_group
PGO
Pipeline
Power
Retire
SMT
Summary
TLB
TLB_SMT
TopDownL1
TopDownL1_SMT
TopdownL1
TopdownL1_SMT
#
# perf list metric | head -11
Metrics:
Backend_Bound
[This category represents fraction of slots where no uops are being delivered due to a lack of required resources for accepting new uops in the Backend]
Backend_Bound_SMT
[This category represents fraction of slots where no uops are being delivered due to a lack of required resources for accepting new uops in the Backend. SMT version; use when SMT is enabled and measuring per logical CPU]
Bad_Speculation
[This category represents fraction of slots wasted due to incorrect speculations]
Bad_Speculation_SMT
[This category represents fraction of slots wasted due to incorrect speculations. SMT version; use when SMT is enabled and measuring per logical CPU]
#
Signed-off-by: Ian Rogers <irogers@google.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lore.kernel.org/lkml/20200522064546.164259-1-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
--xed currently forces less. When piping the output to other scripts
this can waste a lot of CPU time because less is rather slow.
I've seen it using up a full core on its own in a pipeline.
Only force less when the output is actually a terminal.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Link: http://lore.kernel.org/lkml/20200522020914.527564-1-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Remove unnecessary commas from events before they are parsed. This
avoids ',' being echoed by parse-events.l.
Signed-off-by: Ian Rogers <irogers@google.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Andrii Nakryiko <andriin@fb.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Kim Phillips <kim.phillips@amd.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: bpf@vger.kernel.org
Cc: netdev@vger.kernel.org
Link: http://lore.kernel.org/lkml/20200520182011.32236-8-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add --metric-no-group that causes all events within metrics to not be
grouped. This can allow the event to get more time when multiplexed, but
may also lower accuracy.
Add --metric-no-merge option. By default events in different metrics may
be shared if the group of events for one metric is the same or larger
than that of the second. Sharing may increase or lower accuracy and so
is now configurable.
Signed-off-by: Ian Rogers <irogers@google.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Andrii Nakryiko <andriin@fb.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Kim Phillips <kim.phillips@amd.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: bpf@vger.kernel.org
Cc: netdev@vger.kernel.org
Link: http://lore.kernel.org/lkml/20200520182011.32236-7-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
A metric group contains multiple metrics. These metrics may use the same
events. If metrics use separate events then it leads to more
multiplexing and overall metric counts fail to sum to 100%.
Modify how metrics are associated with events so that if the events in
an earlier group satisfy the current metric, the same events are used.
A record of used events is kept and at the end of processing unnecessary
events are eliminated.
Before:
$ perf stat -a -M TopDownL1 sleep 1
Performance counter stats for 'system wide':
920,211,343 uops_issued.any # 0.5 Backend_Bound (16.56%)
1,977,733,128 idq_uops_not_delivered.core (16.56%)
51,668,510 int_misc.recovery_cycles (16.56%)
732,305,692 uops_retired.retire_slots (16.56%)
1,497,621,849 cycles (16.56%)
721,098,274 uops_issued.any # 0.1 Bad_Speculation (16.79%)
1,332,681,791 cycles (16.79%)
552,475,482 uops_retired.retire_slots (16.79%)
47,708,340 int_misc.recovery_cycles (16.79%)
1,383,713,292 cycles
# 0.4 Frontend_Bound (16.76%)
2,013,757,701 idq_uops_not_delivered.core (16.76%)
1,373,363,790 cycles
# 0.1 Retiring (33.54%)
577,302,589 uops_retired.retire_slots (33.54%)
392,766,987 inst_retired.any # 0.3 IPC (50.24%)
1,351,873,350 cpu_clk_unhalted.thread (50.24%)
1,332,510,318 cycles
# 5330041272.0 SLOTS (49.90%)
1.006336145 seconds time elapsed
After:
$ perf stat -a -M TopDownL1 sleep 1
Performance counter stats for 'system wide':
765,949,145 uops_issued.any # 0.1 Bad_Speculation
# 0.5 Backend_Bound (50.09%)
1,883,830,591 idq_uops_not_delivered.core # 0.3 Frontend_Bound (50.09%)
48,237,080 int_misc.recovery_cycles (50.09%)
581,798,385 uops_retired.retire_slots # 0.1 Retiring (50.09%)
1,361,628,527 cycles
# 5446514108.0 SLOTS (50.09%)
391,415,714 inst_retired.any # 0.3 IPC (49.91%)
1,336,486,781 cpu_clk_unhalted.thread (49.91%)
1.005469298 seconds time elapsed
Note: Bad_Speculation + Backend_Bound + Frontend_Bound + Retiring = 100%
after, where as before it is 110%. After there are 2 groups, whereas
before there are 6. After the cycles event appears once, before it
appeared 5 times.
Signed-off-by: Ian Rogers <irogers@google.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Andrii Nakryiko <andriin@fb.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Kim Phillips <kim.phillips@amd.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: bpf@vger.kernel.org
Cc: netdev@vger.kernel.org
Link: http://lore.kernel.org/lkml/20200520182011.32236-6-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When adding event groups to the group list, insert them in size order.
This performs an insertion sort on the group list. By placing the
largest groups at the front of the group list it is possible to see if a
larger group contains the same events as a later group. This can make
the later group redundant - it can reuse the events from the large
group. A later patch will add this sharing.
Signed-off-by: Ian Rogers <irogers@google.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Andrii Nakryiko <andriin@fb.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Kim Phillips <kim.phillips@amd.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: bpf@vger.kernel.org
Cc: netdev@vger.kernel.org
Link: http://lore.kernel.org/lkml/20200520182011.32236-5-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Currently event groups are placed into groups_list at the same time as
the events string containing the events is built. Separate these two
operations and build the groups_list first, then the event string from
the groups_list. This adds an ability to reorder the groups_list that
will be used in a later patch.
Signed-off-by: Ian Rogers <irogers@google.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Andrii Nakryiko <andriin@fb.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Kim Phillips <kim.phillips@amd.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: bpf@vger.kernel.org
Cc: netdev@vger.kernel.org
Link: http://lore.kernel.org/lkml/20200520182011.32236-4-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Use early return in metricgroup__add_metric and try to make the intent
of the returns more intention revealing.
Suggested-by: Jiri Olsa <jolsa@redhat.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Andrii Nakryiko <andriin@fb.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Kim Phillips <kim.phillips@amd.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: bpf@vger.kernel.org
Cc: netdev@vger.kernel.org
Link: http://lore.kernel.org/lkml/20200520182011.32236-3-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
If a metric contains the duration_time event then the event is placed
outside of the metric's group of events. Rather than split the group,
make it so the duration_time is immediately after the group.
Signed-off-by: Ian Rogers <irogers@google.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Andrii Nakryiko <andriin@fb.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Kim Phillips <kim.phillips@amd.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: bpf@vger.kernel.org
Cc: netdev@vger.kernel.org
Link: http://lore.kernel.org/lkml/20200520182011.32236-2-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This patch fix potential segment fault triggered in
put_tracepoints_path() when the address of the local variable 'path' be
freed in error path of record_saved_cmdline.
Signed-off-by: Li Bin <huawei.libin@huawei.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Hongbo Yao <yaohongbo@huawei.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Xie XiuQi <xiexiuqi@huawei.com>
Link: http://lore.kernel.org/lkml/20200521133218.30150-5-liwei391@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Need to free "str" before return when asprintf() failed to avoid memory
leak.
Signed-off-by: Xie XiuQi <xiexiuqi@huawei.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Hongbo Yao <yaohongbo@huawei.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Li Bin <huawei.libin@huawei.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: http://lore.kernel.org/lkml/20200521133218.30150-4-liwei391@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Currently there's no error message prompted if we failed to start
workload. And we still get some trace which is confusing. Let's tell
users what happened.
Committer testing:
Before:
# perf ftrace nonsense |& head
5) | switch_mm_irqs_off() {
5) 0.400 us | load_new_mm_cr3();
5) 3.261 us | }
------------------------------------------
5) <idle>-0 => <...>-3494
------------------------------------------
5) | finish_task_switch() {
5) ==========> |
5) | smp_irq_work_interrupt() {
# type nonsense
-bash: type: nonsense: not found
#
After:
# perf ftrace nonsense |& head
workload failed: No such file or directory
# type nonsense
-bash: type: nonsense: not found
#
Signed-off-by: Changbin Du <changbin.du@gmail.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt (VMware) <rostedt@goodmis.org>
Link: http://lore.kernel.org/lkml/20200510150628.16610-3-changbin.du@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This align ftrace to other perf sub-commands that if no target specified
then we trace all functions.
Signed-off-by: Changbin Du <changbin.du@gmail.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt (VMware) <rostedt@goodmis.org>
Link: http://lore.kernel.org/lkml/20200510150628.16610-2-changbin.du@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array
member[1][2], introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning in
case the flexible array does not occur last in the structure, which will
help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by this
change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
sizeof(flexible-array-member) triggers a warning because flexible array
members have incomplete type[1]. There are some instances of code in
which the sizeof operator is being incorrectly/erroneously applied to
zero-length arrays and the result is zero. Such instances may be hiding
some bugs. So, this work (flexible-array member conversions) will also
help to get completely rid of those sorts of issues.
This issue was found with the help of Coccinelle and audited _manually_.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 7649773293 ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Gustavo A. R. Silva <gustavo@embeddedor.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20200520191613.GA26869@embeddedor
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add support for new "stat.big-num" boolean option.
This allows a user to set a default for "--no-big-num" for "perf stat"
commands.
--
$ perf config stat.big-num
$ perf stat --event cycles /bin/true
Performance counter stats for '/bin/true':
778,849 cycles
[...]
$ perf config stat.big-num=false
$ perf config stat.big-num
stat.big-num=false
$ perf stat --event cycles /bin/true
Performance counter stats for '/bin/true':
769622 cycles
[...]
--
There is an interaction with "--field-separator" that must be
accommodated, such that specifying "--big-num --field-separator={x}"
still reports an invalid combination of options.
Documentation for perf-config and perf-stat updated.
Signed-off-by: Paul Clarke <pc@us.ibm.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Link: http://lore.kernel.org/lkml/1589991815-17951-1-git-send-email-pc@us.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
key_scan_pos is a pointer for getting scan position in
bpf__obj_config_map() for each BPF map configuration term,
but it's misused when error not happened.
Committer notes:
The point is that the only user of this is:
tools/perf/util/parse-events.c
err = bpf__config_obj(obj, term, parse_state->evlist, &error_pos);
if (err) bpf__strerror_config_obj(obj, term, parse_state->evlist, &error_pos, err, errbuf, sizeof(errbuf));
And then:
int bpf__strerror_config_obj(struct bpf_object *obj __maybe_unused,
struct parse_events_term *term __maybe_unused,
struct evlist *evlist __maybe_unused,
int *error_pos __maybe_unused, int err,
char *buf, size_t size)
{
bpf__strerror_head(err, buf, size);
bpf__strerror_entry(BPF_LOADER_ERRNO__OBJCONF_MAP_TYPE,
"Can't use this config term with this map type");
bpf__strerror_end(buf, size);
return 0;
}
So this is infrastructure that Wang Nan put in place for providing
better error messages but that he ended up not using, so I'll apply the
fix, its correct even not fixing any real problem at this time.
Fixes: 066dacbf2a ("perf bpf: Add API to set values to map entries in a bpf object")
Signed-off-by: Wang ShaoBo <bobo.shaobowang@huawei.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Cheng Jian <cj.chengjian@huawei.com>
Cc: Hanjun Guo <guohanjun@huawei.com>
Cc: Li Bin <huawei.libin@huawei.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: Xie XiuQi <xiexiuqi@huawei.com>
Cc: bpf@vger.kernel.org
Link: http://lore.kernel.org/lkml/20200520033216.48310-1-bobo.shaobowang@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
To collect the overall statistics for interval mode, we copy the counts
from evsel->prev_raw_counts to evsel->counts.
For AGGR_GLOBAL mode, because the perf_stat_process_counter creates aggr
values from per cpu values, but the per cpu values are 0, so the
calculated aggr values will be always 0.
This patch uses a trick that saves the previous aggr value to the first
member of perf_counts, then aggr calculation in process_counter_values
can work correctly for AGGR_GLOBAL.
v6:
---
Add comments in perf_evlist__save_aggr_prev_raw_counts.
Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
Reviewed-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jin Yao <yao.jin@intel.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20200520042737.24160-5-yao.jin@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
It would be useful to support the overall statistics for perf-stat
interval mode. For example, report the summary at the end of "perf-stat
-I" output.
But since perf-stat can support many aggregation modes, such as
--per-thread, --per-socket, -M and etc, we need a solution which doesn't
bring much complexity.
The idea is to use 'evsel->prev_raw_counts' which is updated in each
interval and it's saved with the latest counts. Before reporting the
summary, we copy the counts from evsel->prev_raw_counts to
evsel->counts, and next we just follow non-interval processing.
v5:
---
Don't save the previous aggr value to the member of [cpu0,thread0]
in perf_counts. Originally that was a trick because the
perf_stat_process_counter would create aggr values from per cpu
values. But we don't need to do that all the time. We will
handle it in next patch.
Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
Reviewed-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jin Yao <yao.jin@intel.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20200520042737.24160-4-yao.jin@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When we want to reset the evsel->prev_raw_counts, zeroing the aggr is
not enough, we need to reset the perf_counts too.
The perf_counts__reset zeros the perf_counts, and it should zero the
aggr too. This patch changes perf_counts__reset to non-static, and calls
it in evsel__reset_prev_raw_counts to reset the prev_raw_counts.
Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
Reviewed-by: Jiri Olsa <jolsa@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jin Yao <yao.jin@intel.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lore.kernel.org/lkml/20200520042737.24160-3-yao.jin@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
