* [PATCHv2 0/2] perf/libperf move
@ 2019-12-06 21:06 Jiri Olsa
2019-12-06 21:06 ` [PATCH 1/2] libperf: Move libperf under tools/lib/perf Jiri Olsa
2019-12-06 21:06 ` [PATCH 2/2] libperf: Add man pages Jiri Olsa
0 siblings, 2 replies; 5+ messages in thread
From: Jiri Olsa @ 2019-12-06 21:06 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo
Cc: lkml, Ingo Molnar, Namhyung Kim, Alexander Shishkin,
Peter Zijlstra, Michael Petlan
hi,
this patchset moves libperf under tools/lib.
It's now possible to build and install it with:
$ cd tools/lib/perf
$ make prefix=/tmp/libperf install
INSTALL libperf.a
INSTALL libperf.so
INSTALL libperf.so.0
INSTALL libperf.so.0.0.1
INSTALL headers
INSTALL libperf.pc
INSTALL man
INSTALL html
INSTALL examples
$ find /tmp/libperf/ -not -type d
/tmp/libperf/share/doc/libperf/examples/sampling.c
/tmp/libperf/share/doc/libperf/examples/counting.c
/tmp/libperf/share/doc/libperf/html/libperf-sampling.html
/tmp/libperf/share/doc/libperf/html/libperf-counting.html
/tmp/libperf/share/doc/libperf/html/libperf.html
/tmp/libperf/share/man/man7/libperf-sampling.7
/tmp/libperf/share/man/man7/libperf-counting.7
/tmp/libperf/share/man/man3/libperf.3
/tmp/libperf/include/perf/mmap.h
/tmp/libperf/include/perf/event.h
/tmp/libperf/include/perf/evsel.h
/tmp/libperf/include/perf/evlist.h
/tmp/libperf/include/perf/threadmap.h
/tmp/libperf/include/perf/cpumap.h
/tmp/libperf/include/perf/core.h
/tmp/libperf/lib64/pkgconfig/libperf.pc
/tmp/libperf/lib64/libperf.so.0.0.1
/tmp/libperf/lib64/libperf.so.0
/tmp/libperf/lib64/libperf.so
/tmp/libperf/lib64/libperf.a
There are few obvious fixes added like include path changes.
Plus adding Documentation changes to have the man page base
for more additions. I switched man pages to asciidoc and
added 3 man pages:
libperf.3 - overall description
libperf-counting.7 - counting basics explained on simple example
libperf-sampling.7 - sampling basics explained on simple example
The plan is to add more man pages to cover the basic API.
I put the html pages in here for you to check:
http://people.redhat.com/~jolsa/libperf/libperf.html
http://people.redhat.com/~jolsa/libperf/libperf-counting.html
http://people.redhat.com/~jolsa/libperf/libperf-sampling.html
It's also available in here:
git://git.kernel.org/pub/scm/linux/kernel/git/jolsa/perf.git
perf/lib
NOTE we are still able to change any libperf ABI,
nothing is committed by this move ;-)
NOTE 'make perf-targz-src-pkg' works nicely with this change,
but is currently failing because of recent bpf changes, I have
a fix for that and will send it shortly
v2 changes:
- squash 'fixes' patch with the move, so we don't break bisect
thanks,
jirka
---
Jiri Olsa (2):
libperf: Move libperf under tools/lib/perf
libperf: Add man pages
tools/{perf/lib => lib/perf}/Build | 0
tools/lib/perf/Documentation/Makefile | 156 +++++++++++++
tools/lib/perf/Documentation/asciidoc.conf | 120 ++++++++++
tools/lib/perf/Documentation/examples/sampling.c | 119 ++++++++++
tools/lib/perf/Documentation/libperf-counting.txt | 211 ++++++++++++++++++
tools/lib/perf/Documentation/libperf-sampling.txt | 243 ++++++++++++++++++++
tools/lib/perf/Documentation/libperf.txt | 246 +++++++++++++++++++++
tools/lib/perf/Documentation/manpage-1.72.xsl | 14 ++
tools/lib/perf/Documentation/manpage-base.xsl | 35 +++
.../perf/Documentation/manpage-bold-literal.xsl | 17 ++
tools/lib/perf/Documentation/manpage-normal.xsl | 13 ++
.../lib/perf/Documentation/manpage-suppress-sp.xsl | 21 ++
tools/{perf/lib => lib/perf}/Makefile | 7 +-
tools/{perf/lib => lib/perf}/core.c | 0
tools/{perf/lib => lib/perf}/cpumap.c | 0
tools/{perf/lib => lib/perf}/evlist.c | 0
tools/{perf/lib => lib/perf}/evsel.c | 0
.../lib => lib/perf}/include/internal/cpumap.h | 0
.../lib => lib/perf}/include/internal/evlist.h | 0
.../lib => lib/perf}/include/internal/evsel.h | 0
.../{perf/lib => lib/perf}/include/internal/lib.h | 0
.../{perf/lib => lib/perf}/include/internal/mmap.h | 0
.../lib => lib/perf}/include/internal/tests.h | 0
.../lib => lib/perf}/include/internal/threadmap.h | 0
.../lib => lib/perf}/include/internal/xyarray.h | 0
tools/{perf/lib => lib/perf}/include/perf/core.h | 0
tools/{perf/lib => lib/perf}/include/perf/cpumap.h | 0
tools/{perf/lib => lib/perf}/include/perf/event.h | 0
tools/{perf/lib => lib/perf}/include/perf/evlist.h | 0
tools/{perf/lib => lib/perf}/include/perf/evsel.h | 0
tools/{perf/lib => lib/perf}/include/perf/mmap.h | 0
.../lib => lib/perf}/include/perf/threadmap.h | 0
tools/{perf/lib => lib/perf}/internal.h | 0
tools/{perf/lib => lib/perf}/lib.c | 0
tools/{perf/lib => lib/perf}/libperf.map | 0
tools/{perf/lib => lib/perf}/libperf.pc.template | 0
tools/{perf/lib => lib/perf}/mmap.c | 0
tools/{perf/lib => lib/perf}/tests/Makefile | 2 +-
tools/{perf/lib => lib/perf}/tests/test-cpumap.c | 0
tools/{perf/lib => lib/perf}/tests/test-evlist.c | 0
tools/{perf/lib => lib/perf}/tests/test-evsel.c | 0
.../{perf/lib => lib/perf}/tests/test-threadmap.c | 0
tools/{perf/lib => lib/perf}/threadmap.c | 0
tools/{perf/lib => lib/perf}/xyarray.c | 0
tools/perf/MANIFEST | 1 +
tools/perf/Makefile.config | 2 +-
tools/perf/Makefile.perf | 2 +-
tools/perf/lib/Documentation/Makefile | 7 -
tools/perf/lib/Documentation/man/libperf.rst | 100 ---------
tools/perf/lib/Documentation/tutorial/tutorial.rst | 123 -----------
50 files changed, 1204 insertions(+), 235 deletions(-)
rename tools/{perf/lib => lib/perf}/Build (100%)
create mode 100644 tools/lib/perf/Documentation/Makefile
create mode 100644 tools/lib/perf/Documentation/asciidoc.conf
create mode 100644 tools/lib/perf/Documentation/examples/sampling.c
create mode 100644 tools/lib/perf/Documentation/libperf-counting.txt
create mode 100644 tools/lib/perf/Documentation/libperf-sampling.txt
create mode 100644 tools/lib/perf/Documentation/libperf.txt
create mode 100644 tools/lib/perf/Documentation/manpage-1.72.xsl
create mode 100644 tools/lib/perf/Documentation/manpage-base.xsl
create mode 100644 tools/lib/perf/Documentation/manpage-bold-literal.xsl
create mode 100644 tools/lib/perf/Documentation/manpage-normal.xsl
create mode 100644 tools/lib/perf/Documentation/manpage-suppress-sp.xsl
rename tools/{perf/lib => lib/perf}/Makefile (96%)
rename tools/{perf/lib => lib/perf}/core.c (100%)
rename tools/{perf/lib => lib/perf}/cpumap.c (100%)
rename tools/{perf/lib => lib/perf}/evlist.c (100%)
rename tools/{perf/lib => lib/perf}/evsel.c (100%)
rename tools/{perf/lib => lib/perf}/include/internal/cpumap.h (100%)
rename tools/{perf/lib => lib/perf}/include/internal/evlist.h (100%)
rename tools/{perf/lib => lib/perf}/include/internal/evsel.h (100%)
rename tools/{perf/lib => lib/perf}/include/internal/lib.h (100%)
rename tools/{perf/lib => lib/perf}/include/internal/mmap.h (100%)
rename tools/{perf/lib => lib/perf}/include/internal/tests.h (100%)
rename tools/{perf/lib => lib/perf}/include/internal/threadmap.h (100%)
rename tools/{perf/lib => lib/perf}/include/internal/xyarray.h (100%)
rename tools/{perf/lib => lib/perf}/include/perf/core.h (100%)
rename tools/{perf/lib => lib/perf}/include/perf/cpumap.h (100%)
rename tools/{perf/lib => lib/perf}/include/perf/event.h (100%)
rename tools/{perf/lib => lib/perf}/include/perf/evlist.h (100%)
rename tools/{perf/lib => lib/perf}/include/perf/evsel.h (100%)
rename tools/{perf/lib => lib/perf}/include/perf/mmap.h (100%)
rename tools/{perf/lib => lib/perf}/include/perf/threadmap.h (100%)
rename tools/{perf/lib => lib/perf}/internal.h (100%)
rename tools/{perf/lib => lib/perf}/lib.c (100%)
rename tools/{perf/lib => lib/perf}/libperf.map (100%)
rename tools/{perf/lib => lib/perf}/libperf.pc.template (100%)
rename tools/{perf/lib => lib/perf}/mmap.c (100%)
rename tools/{perf/lib => lib/perf}/tests/Makefile (93%)
rename tools/{perf/lib => lib/perf}/tests/test-cpumap.c (100%)
rename tools/{perf/lib => lib/perf}/tests/test-evlist.c (100%)
rename tools/{perf/lib => lib/perf}/tests/test-evsel.c (100%)
rename tools/{perf/lib => lib/perf}/tests/test-threadmap.c (100%)
rename tools/{perf/lib => lib/perf}/threadmap.c (100%)
rename tools/{perf/lib => lib/perf}/xyarray.c (100%)
delete mode 100644 tools/perf/lib/Documentation/Makefile
delete mode 100644 tools/perf/lib/Documentation/man/libperf.rst
delete mode 100644 tools/perf/lib/Documentation/tutorial/tutorial.rst
^ permalink raw reply [flat|nested] 5+ messages in thread* [PATCH 1/2] libperf: Move libperf under tools/lib/perf 2019-12-06 21:06 [PATCHv2 0/2] perf/libperf move Jiri Olsa @ 2019-12-06 21:06 ` Jiri Olsa 2020-01-10 17:53 ` [tip: perf/core] libperf: Move to tools/lib/perf tip-bot2 for Jiri Olsa 2019-12-06 21:06 ` [PATCH 2/2] libperf: Add man pages Jiri Olsa 1 sibling, 1 reply; 5+ messages in thread From: Jiri Olsa @ 2019-12-06 21:06 UTC (permalink / raw) To: Arnaldo Carvalho de Melo Cc: lkml, Ingo Molnar, Namhyung Kim, Alexander Shishkin, Peter Zijlstra, Michael Petlan Moving libperf from its current location under perf to separate directory under tools/lib. Also changing various paths (mainly include) to reflect the libperf move under separate directory and adding new directory under MANIFEST. Link: https://lkml.kernel.org/n/tip-78qoi1s5iitrfmsj67j2evzd@git.kernel.org Signed-off-by: Jiri Olsa <jolsa@kernel.org> --- tools/lib/perf/Build | 13 + tools/lib/perf/Documentation/Makefile | 7 + tools/lib/perf/Documentation/man/libperf.rst | 100 +++ .../perf/Documentation/tutorial/tutorial.rst | 123 ++++ tools/lib/perf/Makefile | 188 +++++ tools/lib/perf/core.c | 38 ++ tools/lib/perf/cpumap.c | 345 ++++++++++ tools/lib/perf/evlist.c | 641 ++++++++++++++++++ tools/lib/perf/evsel.c | 301 ++++++++ tools/lib/perf/include/internal/cpumap.h | 19 + tools/lib/perf/include/internal/evlist.h | 127 ++++ tools/lib/perf/include/internal/evsel.h | 63 ++ tools/lib/perf/include/internal/lib.h | 12 + tools/lib/perf/include/internal/mmap.h | 55 ++ tools/lib/perf/include/internal/tests.h | 33 + tools/lib/perf/include/internal/threadmap.h | 23 + tools/lib/perf/include/internal/xyarray.h | 36 + tools/lib/perf/include/perf/core.h | 25 + tools/lib/perf/include/perf/cpumap.h | 28 + tools/lib/perf/include/perf/event.h | 385 +++++++++++ tools/lib/perf/include/perf/evlist.h | 49 ++ tools/lib/perf/include/perf/evsel.h | 40 ++ tools/lib/perf/include/perf/mmap.h | 15 + tools/lib/perf/include/perf/threadmap.h | 20 + tools/lib/perf/internal.h | 23 + tools/lib/perf/lib.c | 48 ++ tools/lib/perf/libperf.map | 51 ++ tools/lib/perf/libperf.pc.template | 11 + tools/lib/perf/mmap.c | 275 ++++++++ tools/lib/perf/tests/Makefile | 38 ++ tools/lib/perf/tests/test-cpumap.c | 31 + tools/lib/perf/tests/test-evlist.c | 413 +++++++++++ tools/lib/perf/tests/test-evsel.c | 135 ++++ tools/lib/perf/tests/test-threadmap.c | 31 + tools/lib/perf/threadmap.c | 91 +++ tools/lib/perf/xyarray.c | 33 + tools/perf/MANIFEST | 1 + tools/perf/Makefile.config | 2 +- tools/perf/Makefile.perf | 2 +- tools/perf/lib/Build | 13 - tools/perf/lib/Documentation/Makefile | 7 - tools/perf/lib/Documentation/man/libperf.rst | 100 --- .../lib/Documentation/tutorial/tutorial.rst | 123 ---- tools/perf/lib/Makefile | 188 ----- tools/perf/lib/core.c | 38 -- tools/perf/lib/cpumap.c | 345 ---------- tools/perf/lib/evlist.c | 641 ------------------ tools/perf/lib/evsel.c | 301 -------- tools/perf/lib/include/internal/cpumap.h | 19 - tools/perf/lib/include/internal/evlist.h | 127 ---- tools/perf/lib/include/internal/evsel.h | 63 -- tools/perf/lib/include/internal/lib.h | 12 - tools/perf/lib/include/internal/mmap.h | 55 -- tools/perf/lib/include/internal/tests.h | 33 - tools/perf/lib/include/internal/threadmap.h | 23 - tools/perf/lib/include/internal/xyarray.h | 36 - tools/perf/lib/include/perf/core.h | 25 - tools/perf/lib/include/perf/cpumap.h | 28 - tools/perf/lib/include/perf/event.h | 385 ----------- tools/perf/lib/include/perf/evlist.h | 49 -- tools/perf/lib/include/perf/evsel.h | 40 -- tools/perf/lib/include/perf/mmap.h | 15 - tools/perf/lib/include/perf/threadmap.h | 20 - tools/perf/lib/internal.h | 23 - tools/perf/lib/lib.c | 48 -- tools/perf/lib/libperf.map | 51 -- tools/perf/lib/libperf.pc.template | 11 - tools/perf/lib/mmap.c | 275 -------- tools/perf/lib/tests/Makefile | 38 -- tools/perf/lib/tests/test-cpumap.c | 31 - tools/perf/lib/tests/test-evlist.c | 413 ----------- tools/perf/lib/tests/test-evsel.c | 135 ---- tools/perf/lib/tests/test-threadmap.c | 31 - tools/perf/lib/threadmap.c | 91 --- tools/perf/lib/xyarray.c | 33 - 75 files changed, 3869 insertions(+), 3868 deletions(-) create mode 100644 tools/lib/perf/Build create mode 100644 tools/lib/perf/Documentation/Makefile create mode 100644 tools/lib/perf/Documentation/man/libperf.rst create mode 100644 tools/lib/perf/Documentation/tutorial/tutorial.rst create mode 100644 tools/lib/perf/Makefile create mode 100644 tools/lib/perf/core.c create mode 100644 tools/lib/perf/cpumap.c create mode 100644 tools/lib/perf/evlist.c create mode 100644 tools/lib/perf/evsel.c create mode 100644 tools/lib/perf/include/internal/cpumap.h create mode 100644 tools/lib/perf/include/internal/evlist.h create mode 100644 tools/lib/perf/include/internal/evsel.h create mode 100644 tools/lib/perf/include/internal/lib.h create mode 100644 tools/lib/perf/include/internal/mmap.h create mode 100644 tools/lib/perf/include/internal/tests.h create mode 100644 tools/lib/perf/include/internal/threadmap.h create mode 100644 tools/lib/perf/include/internal/xyarray.h create mode 100644 tools/lib/perf/include/perf/core.h create mode 100644 tools/lib/perf/include/perf/cpumap.h create mode 100644 tools/lib/perf/include/perf/event.h create mode 100644 tools/lib/perf/include/perf/evlist.h create mode 100644 tools/lib/perf/include/perf/evsel.h create mode 100644 tools/lib/perf/include/perf/mmap.h create mode 100644 tools/lib/perf/include/perf/threadmap.h create mode 100644 tools/lib/perf/internal.h create mode 100644 tools/lib/perf/lib.c create mode 100644 tools/lib/perf/libperf.map create mode 100644 tools/lib/perf/libperf.pc.template create mode 100644 tools/lib/perf/mmap.c create mode 100644 tools/lib/perf/tests/Makefile create mode 100644 tools/lib/perf/tests/test-cpumap.c create mode 100644 tools/lib/perf/tests/test-evlist.c create mode 100644 tools/lib/perf/tests/test-evsel.c create mode 100644 tools/lib/perf/tests/test-threadmap.c create mode 100644 tools/lib/perf/threadmap.c create mode 100644 tools/lib/perf/xyarray.c delete mode 100644 tools/perf/lib/Build delete mode 100644 tools/perf/lib/Documentation/Makefile delete mode 100644 tools/perf/lib/Documentation/man/libperf.rst delete mode 100644 tools/perf/lib/Documentation/tutorial/tutorial.rst delete mode 100644 tools/perf/lib/Makefile delete mode 100644 tools/perf/lib/core.c delete mode 100644 tools/perf/lib/cpumap.c delete mode 100644 tools/perf/lib/evlist.c delete mode 100644 tools/perf/lib/evsel.c delete mode 100644 tools/perf/lib/include/internal/cpumap.h delete mode 100644 tools/perf/lib/include/internal/evlist.h delete mode 100644 tools/perf/lib/include/internal/evsel.h delete mode 100644 tools/perf/lib/include/internal/lib.h delete mode 100644 tools/perf/lib/include/internal/mmap.h delete mode 100644 tools/perf/lib/include/internal/tests.h delete mode 100644 tools/perf/lib/include/internal/threadmap.h delete mode 100644 tools/perf/lib/include/internal/xyarray.h delete mode 100644 tools/perf/lib/include/perf/core.h delete mode 100644 tools/perf/lib/include/perf/cpumap.h delete mode 100644 tools/perf/lib/include/perf/event.h delete mode 100644 tools/perf/lib/include/perf/evlist.h delete mode 100644 tools/perf/lib/include/perf/evsel.h delete mode 100644 tools/perf/lib/include/perf/mmap.h delete mode 100644 tools/perf/lib/include/perf/threadmap.h delete mode 100644 tools/perf/lib/internal.h delete mode 100644 tools/perf/lib/lib.c delete mode 100644 tools/perf/lib/libperf.map delete mode 100644 tools/perf/lib/libperf.pc.template delete mode 100644 tools/perf/lib/mmap.c delete mode 100644 tools/perf/lib/tests/Makefile delete mode 100644 tools/perf/lib/tests/test-cpumap.c delete mode 100644 tools/perf/lib/tests/test-evlist.c delete mode 100644 tools/perf/lib/tests/test-evsel.c delete mode 100644 tools/perf/lib/tests/test-threadmap.c delete mode 100644 tools/perf/lib/threadmap.c delete mode 100644 tools/perf/lib/xyarray.c diff --git a/tools/lib/perf/Build b/tools/lib/perf/Build new file mode 100644 index 000000000000..2ef9a4ec6d99 --- /dev/null +++ b/tools/lib/perf/Build @@ -0,0 +1,13 @@ +libperf-y += core.o +libperf-y += cpumap.o +libperf-y += threadmap.o +libperf-y += evsel.o +libperf-y += evlist.o +libperf-y += mmap.o +libperf-y += zalloc.o +libperf-y += xyarray.o +libperf-y += lib.o + +$(OUTPUT)zalloc.o: ../../lib/zalloc.c FORCE + $(call rule_mkdir) + $(call if_changed_dep,cc_o_c) diff --git a/tools/lib/perf/Documentation/Makefile b/tools/lib/perf/Documentation/Makefile new file mode 100644 index 000000000000..586425a88795 --- /dev/null +++ b/tools/lib/perf/Documentation/Makefile @@ -0,0 +1,7 @@ +all: + rst2man man/libperf.rst > man/libperf.7 + rst2pdf tutorial/tutorial.rst + +clean: + rm -f man/libperf.7 + rm -f tutorial/tutorial.pdf diff --git a/tools/lib/perf/Documentation/man/libperf.rst b/tools/lib/perf/Documentation/man/libperf.rst new file mode 100644 index 000000000000..09a270fccb9c --- /dev/null +++ b/tools/lib/perf/Documentation/man/libperf.rst @@ -0,0 +1,100 @@ +.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) + +libperf + +The libperf library provides an API to access the linux kernel perf +events subsystem. It provides the following high level objects: + + - struct perf_cpu_map + - struct perf_thread_map + - struct perf_evlist + - struct perf_evsel + +reference +========= +Function reference by header files: + +perf/core.h +----------- +.. code-block:: c + + typedef int (\*libperf_print_fn_t)(enum libperf_print_level level, + const char \*, va_list ap); + + void libperf_set_print(libperf_print_fn_t fn); + +perf/cpumap.h +------------- +.. code-block:: c + + struct perf_cpu_map \*perf_cpu_map__dummy_new(void); + struct perf_cpu_map \*perf_cpu_map__new(const char \*cpu_list); + struct perf_cpu_map \*perf_cpu_map__read(FILE \*file); + struct perf_cpu_map \*perf_cpu_map__get(struct perf_cpu_map \*map); + void perf_cpu_map__put(struct perf_cpu_map \*map); + int perf_cpu_map__cpu(const struct perf_cpu_map \*cpus, int idx); + int perf_cpu_map__nr(const struct perf_cpu_map \*cpus); + perf_cpu_map__for_each_cpu(cpu, idx, cpus) + +perf/threadmap.h +---------------- +.. code-block:: c + + struct perf_thread_map \*perf_thread_map__new_dummy(void); + void perf_thread_map__set_pid(struct perf_thread_map \*map, int thread, pid_t pid); + char \*perf_thread_map__comm(struct perf_thread_map \*map, int thread); + struct perf_thread_map \*perf_thread_map__get(struct perf_thread_map \*map); + void perf_thread_map__put(struct perf_thread_map \*map); + +perf/evlist.h +------------- +.. code-block:: + + void perf_evlist__init(struct perf_evlist \*evlist); + void perf_evlist__add(struct perf_evlist \*evlist, + struct perf_evsel \*evsel); + void perf_evlist__remove(struct perf_evlist \*evlist, + struct perf_evsel \*evsel); + struct perf_evlist \*perf_evlist__new(void); + void perf_evlist__delete(struct perf_evlist \*evlist); + struct perf_evsel\* perf_evlist__next(struct perf_evlist \*evlist, + struct perf_evsel \*evsel); + int perf_evlist__open(struct perf_evlist \*evlist); + void perf_evlist__close(struct perf_evlist \*evlist); + void perf_evlist__enable(struct perf_evlist \*evlist); + void perf_evlist__disable(struct perf_evlist \*evlist); + perf_evlist__for_each_evsel(evlist, pos) + void perf_evlist__set_maps(struct perf_evlist \*evlist, + struct perf_cpu_map \*cpus, + struct perf_thread_map \*threads); + +perf/evsel.h +------------ +.. code-block:: c + + struct perf_counts_values { + union { + struct { + uint64_t val; + uint64_t ena; + uint64_t run; + }; + uint64_t values[3]; + }; + }; + + void perf_evsel__init(struct perf_evsel \*evsel, + struct perf_event_attr \*attr); + struct perf_evsel \*perf_evsel__new(struct perf_event_attr \*attr); + void perf_evsel__delete(struct perf_evsel \*evsel); + int perf_evsel__open(struct perf_evsel \*evsel, struct perf_cpu_map \*cpus, + struct perf_thread_map \*threads); + void perf_evsel__close(struct perf_evsel \*evsel); + int perf_evsel__read(struct perf_evsel \*evsel, int cpu, int thread, + struct perf_counts_values \*count); + int perf_evsel__enable(struct perf_evsel \*evsel); + int perf_evsel__disable(struct perf_evsel \*evsel); + int perf_evsel__apply_filter(struct perf_evsel \*evsel, const char \*filter); + struct perf_cpu_map \*perf_evsel__cpus(struct perf_evsel \*evsel); + struct perf_thread_map \*perf_evsel__threads(struct perf_evsel \*evsel); + struct perf_event_attr \*perf_evsel__attr(struct perf_evsel \*evsel); diff --git a/tools/lib/perf/Documentation/tutorial/tutorial.rst b/tools/lib/perf/Documentation/tutorial/tutorial.rst new file mode 100644 index 000000000000..7be7bc27b385 --- /dev/null +++ b/tools/lib/perf/Documentation/tutorial/tutorial.rst @@ -0,0 +1,123 @@ +.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) + +libperf tutorial +================ + +Compile and install libperf from kernel sources +=============================================== +.. code-block:: bash + + git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git + cd linux/tools/perf/lib + make + sudo make install prefix=/usr + +Libperf object +============== +The libperf library provides several high level objects: + +struct perf_cpu_map + Provides a cpu list abstraction. + +struct perf_thread_map + Provides a thread list abstraction. + +struct perf_evsel + Provides an abstraction for single a perf event. + +struct perf_evlist + Gathers several struct perf_evsel object and performs functions on all of them. + +The exported API binds these objects together, +for full reference see the libperf.7 man page. + +Examples +======== +Examples aim to explain libperf functionality on simple use cases. +They are based in on a checked out linux kernel git tree: + +.. code-block:: bash + + $ cd tools/perf/lib/Documentation/tutorial/ + $ ls -d ex-* + ex-1-compile ex-2-evsel-stat ex-3-evlist-stat + +ex-1-compile example +==================== +This example shows the basic usage of *struct perf_cpu_map*, +how to create it and display its cpus: + +.. code-block:: bash + + $ cd ex-1-compile/ + $ make + gcc -o test test.c -lperf + $ ./test + 0 1 2 3 4 5 6 7 + + +The full code listing is here: + +.. code-block:: c + + 1 #include <perf/cpumap.h> + 2 + 3 int main(int argc, char **Argv) + 4 { + 5 struct perf_cpu_map *cpus; + 6 int cpu, tmp; + 7 + 8 cpus = perf_cpu_map__new(NULL); + 9 + 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) + 11 fprintf(stdout, "%d ", cpu); + 12 + 13 fprintf(stdout, "\n"); + 14 + 15 perf_cpu_map__put(cpus); + 16 return 0; + 17 } + + +First you need to include the proper header to have *struct perf_cpumap* +declaration and functions: + +.. code-block:: c + + 1 #include <perf/cpumap.h> + + +The *struct perf_cpumap* object is created by *perf_cpu_map__new* call. +The *NULL* argument asks it to populate the object with the current online CPUs list: + +.. code-block:: c + + 8 cpus = perf_cpu_map__new(NULL); + +This is paired with a *perf_cpu_map__put*, that drops its reference at the end, possibly deleting it. + +.. code-block:: c + + 15 perf_cpu_map__put(cpus); + +The iteration through the *struct perf_cpumap* CPUs is done using the *perf_cpu_map__for_each_cpu* +macro which requires 3 arguments: + +- cpu - the cpu numer +- tmp - iteration helper variable +- cpus - the *struct perf_cpumap* object + +.. code-block:: c + + 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) + 11 fprintf(stdout, "%d ", cpu); + +ex-2-evsel-stat example +======================= + +TBD + +ex-3-evlist-stat example +======================== + +TBD diff --git a/tools/lib/perf/Makefile b/tools/lib/perf/Makefile new file mode 100644 index 000000000000..768dd423730b --- /dev/null +++ b/tools/lib/perf/Makefile @@ -0,0 +1,188 @@ +# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) +# Most of this file is copied from tools/lib/bpf/Makefile + +LIBPERF_VERSION = 0 +LIBPERF_PATCHLEVEL = 0 +LIBPERF_EXTRAVERSION = 1 + +MAKEFLAGS += --no-print-directory + +ifeq ($(srctree),) +srctree := $(patsubst %/,%,$(dir $(CURDIR))) +srctree := $(patsubst %/,%,$(dir $(srctree))) +srctree := $(patsubst %/,%,$(dir $(srctree))) +#$(info Determined 'srctree' to be $(srctree)) +endif + +INSTALL = install + +# Use DESTDIR for installing into a different root directory. +# This is useful for building a package. The program will be +# installed in this directory as if it was the root directory. +# Then the build tool can move it later. +DESTDIR ?= +DESTDIR_SQ = '$(subst ','\'',$(DESTDIR))' + +include $(srctree)/tools/scripts/Makefile.include +include $(srctree)/tools/scripts/Makefile.arch + +ifeq ($(LP64), 1) + libdir_relative = lib64 +else + libdir_relative = lib +endif + +prefix ?= +libdir = $(prefix)/$(libdir_relative) + +# Shell quotes +libdir_SQ = $(subst ','\'',$(libdir)) +libdir_relative_SQ = $(subst ','\'',$(libdir_relative)) + +ifeq ("$(origin V)", "command line") + VERBOSE = $(V) +endif +ifndef VERBOSE + VERBOSE = 0 +endif + +ifeq ($(VERBOSE),1) + Q = +else + Q = @ +endif + +# Set compile option CFLAGS +ifdef EXTRA_CFLAGS + CFLAGS := $(EXTRA_CFLAGS) +else + CFLAGS := -g -Wall +endif + +INCLUDES = \ +-I$(srctree)/tools/lib/perf/include \ +-I$(srctree)/tools/lib/ \ +-I$(srctree)/tools/include \ +-I$(srctree)/tools/arch/$(SRCARCH)/include/ \ +-I$(srctree)/tools/arch/$(SRCARCH)/include/uapi \ +-I$(srctree)/tools/include/uapi + +# Append required CFLAGS +override CFLAGS += $(EXTRA_WARNINGS) +override CFLAGS += -Werror -Wall +override CFLAGS += -fPIC +override CFLAGS += $(INCLUDES) +override CFLAGS += -fvisibility=hidden + +all: + +export srctree OUTPUT CC LD CFLAGS V +export DESTDIR DESTDIR_SQ + +include $(srctree)/tools/build/Makefile.include + +VERSION_SCRIPT := libperf.map + +PATCHLEVEL = $(LIBPERF_PATCHLEVEL) +EXTRAVERSION = $(LIBPERF_EXTRAVERSION) +VERSION = $(LIBPERF_VERSION).$(LIBPERF_PATCHLEVEL).$(LIBPERF_EXTRAVERSION) + +LIBPERF_SO := $(OUTPUT)libperf.so.$(VERSION) +LIBPERF_A := $(OUTPUT)libperf.a +LIBPERF_IN := $(OUTPUT)libperf-in.o +LIBPERF_PC := $(OUTPUT)libperf.pc + +LIBPERF_ALL := $(LIBPERF_A) $(OUTPUT)libperf.so* + +LIB_DIR := $(srctree)/tools/lib/api/ + +ifneq ($(OUTPUT),) +ifneq ($(subdir),) + API_PATH=$(OUTPUT)/../lib/api/ +else + API_PATH=$(OUTPUT) +endif +else + API_PATH=$(LIB_DIR) +endif + +LIBAPI = $(API_PATH)libapi.a +export LIBAPI + +$(LIBAPI): FORCE + $(Q)$(MAKE) -C $(LIB_DIR) O=$(OUTPUT) $(OUTPUT)libapi.a + +$(LIBAPI)-clean: + $(call QUIET_CLEAN, libapi) + $(Q)$(MAKE) -C $(LIB_DIR) O=$(OUTPUT) clean >/dev/null + +$(LIBPERF_IN): FORCE + $(Q)$(MAKE) $(build)=libperf + +$(LIBPERF_A): $(LIBPERF_IN) + $(QUIET_AR)$(RM) $@ && $(AR) rcs $@ $(LIBPERF_IN) + +$(LIBPERF_SO): $(LIBPERF_IN) $(LIBAPI) + $(QUIET_LINK)$(CC) --shared -Wl,-soname,libperf.so \ + -Wl,--version-script=$(VERSION_SCRIPT) $^ -o $@ + @ln -sf $(@F) $(OUTPUT)libperf.so + @ln -sf $(@F) $(OUTPUT)libperf.so.$(LIBPERF_VERSION) + + +libs: $(LIBPERF_A) $(LIBPERF_SO) $(LIBPERF_PC) + +all: fixdep + $(Q)$(MAKE) libs + +clean: $(LIBAPI)-clean + $(call QUIET_CLEAN, libperf) $(RM) $(LIBPERF_A) \ + *.o *~ *.a *.so *.so.$(VERSION) *.so.$(LIBPERF_VERSION) .*.d .*.cmd LIBPERF-CFLAGS $(LIBPERF_PC) + $(Q)$(MAKE) -C tests clean + +tests: libs + $(Q)$(MAKE) -C tests + $(Q)$(MAKE) -C tests run + +$(LIBPERF_PC): + $(QUIET_GEN)sed -e "s|@PREFIX@|$(prefix)|" \ + -e "s|@LIBDIR@|$(libdir_SQ)|" \ + -e "s|@VERSION@|$(VERSION)|" \ + < libperf.pc.template > $@ + +define do_install_mkdir + if [ ! -d '$(DESTDIR_SQ)$1' ]; then \ + $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$1'; \ + fi +endef + +define do_install + if [ ! -d '$(DESTDIR_SQ)$2' ]; then \ + $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$2'; \ + fi; \ + $(INSTALL) $1 $(if $3,-m $3,) '$(DESTDIR_SQ)$2' +endef + +install_lib: libs + $(call QUIET_INSTALL, $(LIBPERF_ALL)) \ + $(call do_install_mkdir,$(libdir_SQ)); \ + cp -fpR $(LIBPERF_ALL) $(DESTDIR)$(libdir_SQ) + +install_headers: + $(call QUIET_INSTALL, headers) \ + $(call do_install,include/perf/core.h,$(prefix)/include/perf,644); \ + $(call do_install,include/perf/cpumap.h,$(prefix)/include/perf,644); \ + $(call do_install,include/perf/threadmap.h,$(prefix)/include/perf,644); \ + $(call do_install,include/perf/evlist.h,$(prefix)/include/perf,644); \ + $(call do_install,include/perf/evsel.h,$(prefix)/include/perf,644); \ + $(call do_install,include/perf/event.h,$(prefix)/include/perf,644); \ + $(call do_install,include/perf/mmap.h,$(prefix)/include/perf,644); + +install_pkgconfig: $(LIBPERF_PC) + $(call QUIET_INSTALL, $(LIBPERF_PC)) \ + $(call do_install,$(LIBPERF_PC),$(libdir_SQ)/pkgconfig,644) + +install: install_lib install_headers install_pkgconfig + +FORCE: + +.PHONY: all install clean tests FORCE diff --git a/tools/lib/perf/core.c b/tools/lib/perf/core.c new file mode 100644 index 000000000000..58fc894b76c5 --- /dev/null +++ b/tools/lib/perf/core.c @@ -0,0 +1,38 @@ +// SPDX-License-Identifier: GPL-2.0-only + +#define __printf(a, b) __attribute__((format(printf, a, b))) + +#include <stdio.h> +#include <stdarg.h> +#include <unistd.h> +#include <linux/compiler.h> +#include <perf/core.h> +#include <internal/lib.h> +#include "internal.h" + +static int __base_pr(enum libperf_print_level level __maybe_unused, const char *format, + va_list args) +{ + return vfprintf(stderr, format, args); +} + +static libperf_print_fn_t __libperf_pr = __base_pr; + +__printf(2, 3) +void libperf_print(enum libperf_print_level level, const char *format, ...) +{ + va_list args; + + if (!__libperf_pr) + return; + + va_start(args, format); + __libperf_pr(level, format, args); + va_end(args); +} + +void libperf_init(libperf_print_fn_t fn) +{ + page_size = sysconf(_SC_PAGE_SIZE); + __libperf_pr = fn; +} diff --git a/tools/lib/perf/cpumap.c b/tools/lib/perf/cpumap.c new file mode 100644 index 000000000000..f93f4e703e4c --- /dev/null +++ b/tools/lib/perf/cpumap.c @@ -0,0 +1,345 @@ +// SPDX-License-Identifier: GPL-2.0-only +#include <perf/cpumap.h> +#include <stdlib.h> +#include <linux/refcount.h> +#include <internal/cpumap.h> +#include <asm/bug.h> +#include <stdio.h> +#include <string.h> +#include <unistd.h> +#include <ctype.h> +#include <limits.h> + +struct perf_cpu_map *perf_cpu_map__dummy_new(void) +{ + struct perf_cpu_map *cpus = malloc(sizeof(*cpus) + sizeof(int)); + + if (cpus != NULL) { + cpus->nr = 1; + cpus->map[0] = -1; + refcount_set(&cpus->refcnt, 1); + } + + return cpus; +} + +static void cpu_map__delete(struct perf_cpu_map *map) +{ + if (map) { + WARN_ONCE(refcount_read(&map->refcnt) != 0, + "cpu_map refcnt unbalanced\n"); + free(map); + } +} + +struct perf_cpu_map *perf_cpu_map__get(struct perf_cpu_map *map) +{ + if (map) + refcount_inc(&map->refcnt); + return map; +} + +void perf_cpu_map__put(struct perf_cpu_map *map) +{ + if (map && refcount_dec_and_test(&map->refcnt)) + cpu_map__delete(map); +} + +static struct perf_cpu_map *cpu_map__default_new(void) +{ + struct perf_cpu_map *cpus; + int nr_cpus; + + nr_cpus = sysconf(_SC_NPROCESSORS_ONLN); + if (nr_cpus < 0) + return NULL; + + cpus = malloc(sizeof(*cpus) + nr_cpus * sizeof(int)); + if (cpus != NULL) { + int i; + + for (i = 0; i < nr_cpus; ++i) + cpus->map[i] = i; + + cpus->nr = nr_cpus; + refcount_set(&cpus->refcnt, 1); + } + + return cpus; +} + +static int cmp_int(const void *a, const void *b) +{ + return *(const int *)a - *(const int*)b; +} + +static struct perf_cpu_map *cpu_map__trim_new(int nr_cpus, int *tmp_cpus) +{ + size_t payload_size = nr_cpus * sizeof(int); + struct perf_cpu_map *cpus = malloc(sizeof(*cpus) + payload_size); + int i, j; + + if (cpus != NULL) { + memcpy(cpus->map, tmp_cpus, payload_size); + qsort(cpus->map, nr_cpus, sizeof(int), cmp_int); + /* Remove dups */ + j = 0; + for (i = 0; i < nr_cpus; i++) { + if (i == 0 || cpus->map[i] != cpus->map[i - 1]) + cpus->map[j++] = cpus->map[i]; + } + cpus->nr = j; + assert(j <= nr_cpus); + refcount_set(&cpus->refcnt, 1); + } + + return cpus; +} + +struct perf_cpu_map *perf_cpu_map__read(FILE *file) +{ + struct perf_cpu_map *cpus = NULL; + int nr_cpus = 0; + int *tmp_cpus = NULL, *tmp; + int max_entries = 0; + int n, cpu, prev; + char sep; + + sep = 0; + prev = -1; + for (;;) { + n = fscanf(file, "%u%c", &cpu, &sep); + if (n <= 0) + break; + if (prev >= 0) { + int new_max = nr_cpus + cpu - prev - 1; + + WARN_ONCE(new_max >= MAX_NR_CPUS, "Perf can support %d CPUs. " + "Consider raising MAX_NR_CPUS\n", MAX_NR_CPUS); + + if (new_max >= max_entries) { + max_entries = new_max + MAX_NR_CPUS / 2; + tmp = realloc(tmp_cpus, max_entries * sizeof(int)); + if (tmp == NULL) + goto out_free_tmp; + tmp_cpus = tmp; + } + + while (++prev < cpu) + tmp_cpus[nr_cpus++] = prev; + } + if (nr_cpus == max_entries) { + max_entries += MAX_NR_CPUS; + tmp = realloc(tmp_cpus, max_entries * sizeof(int)); + if (tmp == NULL) + goto out_free_tmp; + tmp_cpus = tmp; + } + + tmp_cpus[nr_cpus++] = cpu; + if (n == 2 && sep == '-') + prev = cpu; + else + prev = -1; + if (n == 1 || sep == '\n') + break; + } + + if (nr_cpus > 0) + cpus = cpu_map__trim_new(nr_cpus, tmp_cpus); + else + cpus = cpu_map__default_new(); +out_free_tmp: + free(tmp_cpus); + return cpus; +} + +static struct perf_cpu_map *cpu_map__read_all_cpu_map(void) +{ + struct perf_cpu_map *cpus = NULL; + FILE *onlnf; + + onlnf = fopen("/sys/devices/system/cpu/online", "r"); + if (!onlnf) + return cpu_map__default_new(); + + cpus = perf_cpu_map__read(onlnf); + fclose(onlnf); + return cpus; +} + +struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list) +{ + struct perf_cpu_map *cpus = NULL; + unsigned long start_cpu, end_cpu = 0; + char *p = NULL; + int i, nr_cpus = 0; + int *tmp_cpus = NULL, *tmp; + int max_entries = 0; + + if (!cpu_list) + return cpu_map__read_all_cpu_map(); + + /* + * must handle the case of empty cpumap to cover + * TOPOLOGY header for NUMA nodes with no CPU + * ( e.g., because of CPU hotplug) + */ + if (!isdigit(*cpu_list) && *cpu_list != '\0') + goto out; + + while (isdigit(*cpu_list)) { + p = NULL; + start_cpu = strtoul(cpu_list, &p, 0); + if (start_cpu >= INT_MAX + || (*p != '\0' && *p != ',' && *p != '-')) + goto invalid; + + if (*p == '-') { + cpu_list = ++p; + p = NULL; + end_cpu = strtoul(cpu_list, &p, 0); + + if (end_cpu >= INT_MAX || (*p != '\0' && *p != ',')) + goto invalid; + + if (end_cpu < start_cpu) + goto invalid; + } else { + end_cpu = start_cpu; + } + + WARN_ONCE(end_cpu >= MAX_NR_CPUS, "Perf can support %d CPUs. " + "Consider raising MAX_NR_CPUS\n", MAX_NR_CPUS); + + for (; start_cpu <= end_cpu; start_cpu++) { + /* check for duplicates */ + for (i = 0; i < nr_cpus; i++) + if (tmp_cpus[i] == (int)start_cpu) + goto invalid; + + if (nr_cpus == max_entries) { + max_entries += MAX_NR_CPUS; + tmp = realloc(tmp_cpus, max_entries * sizeof(int)); + if (tmp == NULL) + goto invalid; + tmp_cpus = tmp; + } + tmp_cpus[nr_cpus++] = (int)start_cpu; + } + if (*p) + ++p; + + cpu_list = p; + } + + if (nr_cpus > 0) + cpus = cpu_map__trim_new(nr_cpus, tmp_cpus); + else if (*cpu_list != '\0') + cpus = cpu_map__default_new(); + else + cpus = perf_cpu_map__dummy_new(); +invalid: + free(tmp_cpus); +out: + return cpus; +} + +int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx) +{ + if (idx < cpus->nr) + return cpus->map[idx]; + + return -1; +} + +int perf_cpu_map__nr(const struct perf_cpu_map *cpus) +{ + return cpus ? cpus->nr : 1; +} + +bool perf_cpu_map__empty(const struct perf_cpu_map *map) +{ + return map ? map->map[0] == -1 : true; +} + +int perf_cpu_map__idx(struct perf_cpu_map *cpus, int cpu) +{ + int i; + + for (i = 0; i < cpus->nr; ++i) { + if (cpus->map[i] == cpu) + return i; + } + + return -1; +} + +int perf_cpu_map__max(struct perf_cpu_map *map) +{ + int i, max = -1; + + for (i = 0; i < map->nr; i++) { + if (map->map[i] > max) + max = map->map[i]; + } + + return max; +} + +/* + * Merge two cpumaps + * + * orig either gets freed and replaced with a new map, or reused + * with no reference count change (similar to "realloc") + * other has its reference count increased. + */ + +struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig, + struct perf_cpu_map *other) +{ + int *tmp_cpus; + int tmp_len; + int i, j, k; + struct perf_cpu_map *merged; + + if (!orig && !other) + return NULL; + if (!orig) { + perf_cpu_map__get(other); + return other; + } + if (!other) + return orig; + if (orig->nr == other->nr && + !memcmp(orig->map, other->map, orig->nr * sizeof(int))) + return orig; + + tmp_len = orig->nr + other->nr; + tmp_cpus = malloc(tmp_len * sizeof(int)); + if (!tmp_cpus) + return NULL; + + /* Standard merge algorithm from wikipedia */ + i = j = k = 0; + while (i < orig->nr && j < other->nr) { + if (orig->map[i] <= other->map[j]) { + if (orig->map[i] == other->map[j]) + j++; + tmp_cpus[k++] = orig->map[i++]; + } else + tmp_cpus[k++] = other->map[j++]; + } + + while (i < orig->nr) + tmp_cpus[k++] = orig->map[i++]; + + while (j < other->nr) + tmp_cpus[k++] = other->map[j++]; + assert(k <= tmp_len); + + merged = cpu_map__trim_new(k, tmp_cpus); + free(tmp_cpus); + perf_cpu_map__put(orig); + return merged; +} diff --git a/tools/lib/perf/evlist.c b/tools/lib/perf/evlist.c new file mode 100644 index 000000000000..ae9e65aa2491 --- /dev/null +++ b/tools/lib/perf/evlist.c @@ -0,0 +1,641 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <perf/evlist.h> +#include <perf/evsel.h> +#include <linux/bitops.h> +#include <linux/list.h> +#include <linux/hash.h> +#include <sys/ioctl.h> +#include <internal/evlist.h> +#include <internal/evsel.h> +#include <internal/xyarray.h> +#include <internal/mmap.h> +#include <internal/cpumap.h> +#include <internal/threadmap.h> +#include <internal/xyarray.h> +#include <internal/lib.h> +#include <linux/zalloc.h> +#include <sys/ioctl.h> +#include <stdlib.h> +#include <errno.h> +#include <unistd.h> +#include <fcntl.h> +#include <signal.h> +#include <poll.h> +#include <sys/mman.h> +#include <perf/cpumap.h> +#include <perf/threadmap.h> +#include <api/fd/array.h> + +void perf_evlist__init(struct perf_evlist *evlist) +{ + int i; + + for (i = 0; i < PERF_EVLIST__HLIST_SIZE; ++i) + INIT_HLIST_HEAD(&evlist->heads[i]); + INIT_LIST_HEAD(&evlist->entries); + evlist->nr_entries = 0; + fdarray__init(&evlist->pollfd, 64); +} + +static void __perf_evlist__propagate_maps(struct perf_evlist *evlist, + struct perf_evsel *evsel) +{ + /* + * We already have cpus for evsel (via PMU sysfs) so + * keep it, if there's no target cpu list defined. + */ + if (!evsel->own_cpus || evlist->has_user_cpus) { + perf_cpu_map__put(evsel->cpus); + evsel->cpus = perf_cpu_map__get(evlist->cpus); + } else if (evsel->cpus != evsel->own_cpus) { + perf_cpu_map__put(evsel->cpus); + evsel->cpus = perf_cpu_map__get(evsel->own_cpus); + } + + perf_thread_map__put(evsel->threads); + evsel->threads = perf_thread_map__get(evlist->threads); + evlist->all_cpus = perf_cpu_map__merge(evlist->all_cpus, evsel->cpus); +} + +static void perf_evlist__propagate_maps(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + + perf_evlist__for_each_evsel(evlist, evsel) + __perf_evlist__propagate_maps(evlist, evsel); +} + +void perf_evlist__add(struct perf_evlist *evlist, + struct perf_evsel *evsel) +{ + list_add_tail(&evsel->node, &evlist->entries); + evlist->nr_entries += 1; + __perf_evlist__propagate_maps(evlist, evsel); +} + +void perf_evlist__remove(struct perf_evlist *evlist, + struct perf_evsel *evsel) +{ + list_del_init(&evsel->node); + evlist->nr_entries -= 1; +} + +struct perf_evlist *perf_evlist__new(void) +{ + struct perf_evlist *evlist = zalloc(sizeof(*evlist)); + + if (evlist != NULL) + perf_evlist__init(evlist); + + return evlist; +} + +struct perf_evsel * +perf_evlist__next(struct perf_evlist *evlist, struct perf_evsel *prev) +{ + struct perf_evsel *next; + + if (!prev) { + next = list_first_entry(&evlist->entries, + struct perf_evsel, + node); + } else { + next = list_next_entry(prev, node); + } + + /* Empty list is noticed here so don't need checking on entry. */ + if (&next->node == &evlist->entries) + return NULL; + + return next; +} + +static void perf_evlist__purge(struct perf_evlist *evlist) +{ + struct perf_evsel *pos, *n; + + perf_evlist__for_each_entry_safe(evlist, n, pos) { + list_del_init(&pos->node); + perf_evsel__delete(pos); + } + + evlist->nr_entries = 0; +} + +void perf_evlist__exit(struct perf_evlist *evlist) +{ + perf_cpu_map__put(evlist->cpus); + perf_thread_map__put(evlist->threads); + evlist->cpus = NULL; + evlist->threads = NULL; + fdarray__exit(&evlist->pollfd); +} + +void perf_evlist__delete(struct perf_evlist *evlist) +{ + if (evlist == NULL) + return; + + perf_evlist__munmap(evlist); + perf_evlist__close(evlist); + perf_evlist__purge(evlist); + perf_evlist__exit(evlist); + free(evlist); +} + +void perf_evlist__set_maps(struct perf_evlist *evlist, + struct perf_cpu_map *cpus, + struct perf_thread_map *threads) +{ + /* + * Allow for the possibility that one or another of the maps isn't being + * changed i.e. don't put it. Note we are assuming the maps that are + * being applied are brand new and evlist is taking ownership of the + * original reference count of 1. If that is not the case it is up to + * the caller to increase the reference count. + */ + if (cpus != evlist->cpus) { + perf_cpu_map__put(evlist->cpus); + evlist->cpus = perf_cpu_map__get(cpus); + } + + if (threads != evlist->threads) { + perf_thread_map__put(evlist->threads); + evlist->threads = perf_thread_map__get(threads); + } + + perf_evlist__propagate_maps(evlist); +} + +int perf_evlist__open(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + int err; + + perf_evlist__for_each_entry(evlist, evsel) { + err = perf_evsel__open(evsel, evsel->cpus, evsel->threads); + if (err < 0) + goto out_err; + } + + return 0; + +out_err: + perf_evlist__close(evlist); + return err; +} + +void perf_evlist__close(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + + perf_evlist__for_each_entry_reverse(evlist, evsel) + perf_evsel__close(evsel); +} + +void perf_evlist__enable(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + + perf_evlist__for_each_entry(evlist, evsel) + perf_evsel__enable(evsel); +} + +void perf_evlist__disable(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + + perf_evlist__for_each_entry(evlist, evsel) + perf_evsel__disable(evsel); +} + +u64 perf_evlist__read_format(struct perf_evlist *evlist) +{ + struct perf_evsel *first = perf_evlist__first(evlist); + + return first->attr.read_format; +} + +#define SID(e, x, y) xyarray__entry(e->sample_id, x, y) + +static void perf_evlist__id_hash(struct perf_evlist *evlist, + struct perf_evsel *evsel, + int cpu, int thread, u64 id) +{ + int hash; + struct perf_sample_id *sid = SID(evsel, cpu, thread); + + sid->id = id; + sid->evsel = evsel; + hash = hash_64(sid->id, PERF_EVLIST__HLIST_BITS); + hlist_add_head(&sid->node, &evlist->heads[hash]); +} + +void perf_evlist__id_add(struct perf_evlist *evlist, + struct perf_evsel *evsel, + int cpu, int thread, u64 id) +{ + perf_evlist__id_hash(evlist, evsel, cpu, thread, id); + evsel->id[evsel->ids++] = id; +} + +int perf_evlist__id_add_fd(struct perf_evlist *evlist, + struct perf_evsel *evsel, + int cpu, int thread, int fd) +{ + u64 read_data[4] = { 0, }; + int id_idx = 1; /* The first entry is the counter value */ + u64 id; + int ret; + + ret = ioctl(fd, PERF_EVENT_IOC_ID, &id); + if (!ret) + goto add; + + if (errno != ENOTTY) + return -1; + + /* Legacy way to get event id.. All hail to old kernels! */ + + /* + * This way does not work with group format read, so bail + * out in that case. + */ + if (perf_evlist__read_format(evlist) & PERF_FORMAT_GROUP) + return -1; + + if (!(evsel->attr.read_format & PERF_FORMAT_ID) || + read(fd, &read_data, sizeof(read_data)) == -1) + return -1; + + if (evsel->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) + ++id_idx; + if (evsel->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) + ++id_idx; + + id = read_data[id_idx]; + +add: + perf_evlist__id_add(evlist, evsel, cpu, thread, id); + return 0; +} + +int perf_evlist__alloc_pollfd(struct perf_evlist *evlist) +{ + int nr_cpus = perf_cpu_map__nr(evlist->cpus); + int nr_threads = perf_thread_map__nr(evlist->threads); + int nfds = 0; + struct perf_evsel *evsel; + + perf_evlist__for_each_entry(evlist, evsel) { + if (evsel->system_wide) + nfds += nr_cpus; + else + nfds += nr_cpus * nr_threads; + } + + if (fdarray__available_entries(&evlist->pollfd) < nfds && + fdarray__grow(&evlist->pollfd, nfds) < 0) + return -ENOMEM; + + return 0; +} + +int perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd, + void *ptr, short revent) +{ + int pos = fdarray__add(&evlist->pollfd, fd, revent | POLLERR | POLLHUP); + + if (pos >= 0) { + evlist->pollfd.priv[pos].ptr = ptr; + fcntl(fd, F_SETFL, O_NONBLOCK); + } + + return pos; +} + +static void perf_evlist__munmap_filtered(struct fdarray *fda, int fd, + void *arg __maybe_unused) +{ + struct perf_mmap *map = fda->priv[fd].ptr; + + if (map) + perf_mmap__put(map); +} + +int perf_evlist__filter_pollfd(struct perf_evlist *evlist, short revents_and_mask) +{ + return fdarray__filter(&evlist->pollfd, revents_and_mask, + perf_evlist__munmap_filtered, NULL); +} + +int perf_evlist__poll(struct perf_evlist *evlist, int timeout) +{ + return fdarray__poll(&evlist->pollfd, timeout); +} + +static struct perf_mmap* perf_evlist__alloc_mmap(struct perf_evlist *evlist, bool overwrite) +{ + int i; + struct perf_mmap *map; + + map = zalloc(evlist->nr_mmaps * sizeof(struct perf_mmap)); + if (!map) + return NULL; + + for (i = 0; i < evlist->nr_mmaps; i++) { + struct perf_mmap *prev = i ? &map[i - 1] : NULL; + + /* + * When the perf_mmap() call is made we grab one refcount, plus + * one extra to let perf_mmap__consume() get the last + * events after all real references (perf_mmap__get()) are + * dropped. + * + * Each PERF_EVENT_IOC_SET_OUTPUT points to this mmap and + * thus does perf_mmap__get() on it. + */ + perf_mmap__init(&map[i], prev, overwrite, NULL); + } + + return map; +} + +static void perf_evlist__set_sid_idx(struct perf_evlist *evlist, + struct perf_evsel *evsel, int idx, int cpu, + int thread) +{ + struct perf_sample_id *sid = SID(evsel, cpu, thread); + + sid->idx = idx; + if (evlist->cpus && cpu >= 0) + sid->cpu = evlist->cpus->map[cpu]; + else + sid->cpu = -1; + if (!evsel->system_wide && evlist->threads && thread >= 0) + sid->tid = perf_thread_map__pid(evlist->threads, thread); + else + sid->tid = -1; +} + +static struct perf_mmap* +perf_evlist__mmap_cb_get(struct perf_evlist *evlist, bool overwrite, int idx) +{ + struct perf_mmap *maps; + + maps = overwrite ? evlist->mmap_ovw : evlist->mmap; + + if (!maps) { + maps = perf_evlist__alloc_mmap(evlist, overwrite); + if (!maps) + return NULL; + + if (overwrite) + evlist->mmap_ovw = maps; + else + evlist->mmap = maps; + } + + return &maps[idx]; +} + +#define FD(e, x, y) (*(int *) xyarray__entry(e->fd, x, y)) + +static int +perf_evlist__mmap_cb_mmap(struct perf_mmap *map, struct perf_mmap_param *mp, + int output, int cpu) +{ + return perf_mmap__mmap(map, mp, output, cpu); +} + +static void perf_evlist__set_mmap_first(struct perf_evlist *evlist, struct perf_mmap *map, + bool overwrite) +{ + if (overwrite) + evlist->mmap_ovw_first = map; + else + evlist->mmap_first = map; +} + +static int +mmap_per_evsel(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, + int idx, struct perf_mmap_param *mp, int cpu_idx, + int thread, int *_output, int *_output_overwrite) +{ + int evlist_cpu = perf_cpu_map__cpu(evlist->cpus, cpu_idx); + struct perf_evsel *evsel; + int revent; + + perf_evlist__for_each_entry(evlist, evsel) { + bool overwrite = evsel->attr.write_backward; + struct perf_mmap *map; + int *output, fd, cpu; + + if (evsel->system_wide && thread) + continue; + + cpu = perf_cpu_map__idx(evsel->cpus, evlist_cpu); + if (cpu == -1) + continue; + + map = ops->get(evlist, overwrite, idx); + if (map == NULL) + return -ENOMEM; + + if (overwrite) { + mp->prot = PROT_READ; + output = _output_overwrite; + } else { + mp->prot = PROT_READ | PROT_WRITE; + output = _output; + } + + fd = FD(evsel, cpu, thread); + + if (*output == -1) { + *output = fd; + + /* + * The last one will be done at perf_mmap__consume(), so that we + * make sure we don't prevent tools from consuming every last event in + * the ring buffer. + * + * I.e. we can get the POLLHUP meaning that the fd doesn't exist + * anymore, but the last events for it are still in the ring buffer, + * waiting to be consumed. + * + * Tools can chose to ignore this at their own discretion, but the + * evlist layer can't just drop it when filtering events in + * perf_evlist__filter_pollfd(). + */ + refcount_set(&map->refcnt, 2); + + if (ops->mmap(map, mp, *output, evlist_cpu) < 0) + return -1; + + if (!idx) + perf_evlist__set_mmap_first(evlist, map, overwrite); + } else { + if (ioctl(fd, PERF_EVENT_IOC_SET_OUTPUT, *output) != 0) + return -1; + + perf_mmap__get(map); + } + + revent = !overwrite ? POLLIN : 0; + + if (!evsel->system_wide && + perf_evlist__add_pollfd(evlist, fd, map, revent) < 0) { + perf_mmap__put(map); + return -1; + } + + if (evsel->attr.read_format & PERF_FORMAT_ID) { + if (perf_evlist__id_add_fd(evlist, evsel, cpu, thread, + fd) < 0) + return -1; + perf_evlist__set_sid_idx(evlist, evsel, idx, cpu, + thread); + } + } + + return 0; +} + +static int +mmap_per_thread(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, + struct perf_mmap_param *mp) +{ + int thread; + int nr_threads = perf_thread_map__nr(evlist->threads); + + for (thread = 0; thread < nr_threads; thread++) { + int output = -1; + int output_overwrite = -1; + + if (ops->idx) + ops->idx(evlist, mp, thread, false); + + if (mmap_per_evsel(evlist, ops, thread, mp, 0, thread, + &output, &output_overwrite)) + goto out_unmap; + } + + return 0; + +out_unmap: + perf_evlist__munmap(evlist); + return -1; +} + +static int +mmap_per_cpu(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, + struct perf_mmap_param *mp) +{ + int nr_threads = perf_thread_map__nr(evlist->threads); + int nr_cpus = perf_cpu_map__nr(evlist->cpus); + int cpu, thread; + + for (cpu = 0; cpu < nr_cpus; cpu++) { + int output = -1; + int output_overwrite = -1; + + if (ops->idx) + ops->idx(evlist, mp, cpu, true); + + for (thread = 0; thread < nr_threads; thread++) { + if (mmap_per_evsel(evlist, ops, cpu, mp, cpu, + thread, &output, &output_overwrite)) + goto out_unmap; + } + } + + return 0; + +out_unmap: + perf_evlist__munmap(evlist); + return -1; +} + +static int perf_evlist__nr_mmaps(struct perf_evlist *evlist) +{ + int nr_mmaps; + + nr_mmaps = perf_cpu_map__nr(evlist->cpus); + if (perf_cpu_map__empty(evlist->cpus)) + nr_mmaps = perf_thread_map__nr(evlist->threads); + + return nr_mmaps; +} + +int perf_evlist__mmap_ops(struct perf_evlist *evlist, + struct perf_evlist_mmap_ops *ops, + struct perf_mmap_param *mp) +{ + struct perf_evsel *evsel; + const struct perf_cpu_map *cpus = evlist->cpus; + const struct perf_thread_map *threads = evlist->threads; + + if (!ops || !ops->get || !ops->mmap) + return -EINVAL; + + mp->mask = evlist->mmap_len - page_size - 1; + + evlist->nr_mmaps = perf_evlist__nr_mmaps(evlist); + + perf_evlist__for_each_entry(evlist, evsel) { + if ((evsel->attr.read_format & PERF_FORMAT_ID) && + evsel->sample_id == NULL && + perf_evsel__alloc_id(evsel, perf_cpu_map__nr(cpus), threads->nr) < 0) + return -ENOMEM; + } + + if (evlist->pollfd.entries == NULL && perf_evlist__alloc_pollfd(evlist) < 0) + return -ENOMEM; + + if (perf_cpu_map__empty(cpus)) + return mmap_per_thread(evlist, ops, mp); + + return mmap_per_cpu(evlist, ops, mp); +} + +int perf_evlist__mmap(struct perf_evlist *evlist, int pages) +{ + struct perf_mmap_param mp; + struct perf_evlist_mmap_ops ops = { + .get = perf_evlist__mmap_cb_get, + .mmap = perf_evlist__mmap_cb_mmap, + }; + + evlist->mmap_len = (pages + 1) * page_size; + + return perf_evlist__mmap_ops(evlist, &ops, &mp); +} + +void perf_evlist__munmap(struct perf_evlist *evlist) +{ + int i; + + if (evlist->mmap) { + for (i = 0; i < evlist->nr_mmaps; i++) + perf_mmap__munmap(&evlist->mmap[i]); + } + + if (evlist->mmap_ovw) { + for (i = 0; i < evlist->nr_mmaps; i++) + perf_mmap__munmap(&evlist->mmap_ovw[i]); + } + + zfree(&evlist->mmap); + zfree(&evlist->mmap_ovw); +} + +struct perf_mmap* +perf_evlist__next_mmap(struct perf_evlist *evlist, struct perf_mmap *map, + bool overwrite) +{ + if (map) + return map->next; + + return overwrite ? evlist->mmap_ovw_first : evlist->mmap_first; +} diff --git a/tools/lib/perf/evsel.c b/tools/lib/perf/evsel.c new file mode 100644 index 000000000000..4dc06289f4c7 --- /dev/null +++ b/tools/lib/perf/evsel.c @@ -0,0 +1,301 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <errno.h> +#include <unistd.h> +#include <sys/syscall.h> +#include <perf/evsel.h> +#include <perf/cpumap.h> +#include <perf/threadmap.h> +#include <linux/list.h> +#include <internal/evsel.h> +#include <linux/zalloc.h> +#include <stdlib.h> +#include <internal/xyarray.h> +#include <internal/cpumap.h> +#include <internal/threadmap.h> +#include <internal/lib.h> +#include <linux/string.h> +#include <sys/ioctl.h> + +void perf_evsel__init(struct perf_evsel *evsel, struct perf_event_attr *attr) +{ + INIT_LIST_HEAD(&evsel->node); + evsel->attr = *attr; +} + +struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr) +{ + struct perf_evsel *evsel = zalloc(sizeof(*evsel)); + + if (evsel != NULL) + perf_evsel__init(evsel, attr); + + return evsel; +} + +void perf_evsel__delete(struct perf_evsel *evsel) +{ + free(evsel); +} + +#define FD(e, x, y) (*(int *) xyarray__entry(e->fd, x, y)) + +int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads) +{ + evsel->fd = xyarray__new(ncpus, nthreads, sizeof(int)); + + if (evsel->fd) { + int cpu, thread; + for (cpu = 0; cpu < ncpus; cpu++) { + for (thread = 0; thread < nthreads; thread++) { + FD(evsel, cpu, thread) = -1; + } + } + } + + return evsel->fd != NULL ? 0 : -ENOMEM; +} + +static int +sys_perf_event_open(struct perf_event_attr *attr, + pid_t pid, int cpu, int group_fd, + unsigned long flags) +{ + return syscall(__NR_perf_event_open, attr, pid, cpu, group_fd, flags); +} + +int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus, + struct perf_thread_map *threads) +{ + int cpu, thread, err = 0; + + if (cpus == NULL) { + static struct perf_cpu_map *empty_cpu_map; + + if (empty_cpu_map == NULL) { + empty_cpu_map = perf_cpu_map__dummy_new(); + if (empty_cpu_map == NULL) + return -ENOMEM; + } + + cpus = empty_cpu_map; + } + + if (threads == NULL) { + static struct perf_thread_map *empty_thread_map; + + if (empty_thread_map == NULL) { + empty_thread_map = perf_thread_map__new_dummy(); + if (empty_thread_map == NULL) + return -ENOMEM; + } + + threads = empty_thread_map; + } + + if (evsel->fd == NULL && + perf_evsel__alloc_fd(evsel, cpus->nr, threads->nr) < 0) + return -ENOMEM; + + for (cpu = 0; cpu < cpus->nr; cpu++) { + for (thread = 0; thread < threads->nr; thread++) { + int fd; + + fd = sys_perf_event_open(&evsel->attr, + threads->map[thread].pid, + cpus->map[cpu], -1, 0); + + if (fd < 0) + return -errno; + + FD(evsel, cpu, thread) = fd; + } + } + + return err; +} + +static void perf_evsel__close_fd_cpu(struct perf_evsel *evsel, int cpu) +{ + int thread; + + for (thread = 0; thread < xyarray__max_y(evsel->fd); ++thread) { + if (FD(evsel, cpu, thread) >= 0) + close(FD(evsel, cpu, thread)); + FD(evsel, cpu, thread) = -1; + } +} + +void perf_evsel__close_fd(struct perf_evsel *evsel) +{ + int cpu; + + for (cpu = 0; cpu < xyarray__max_x(evsel->fd); cpu++) + perf_evsel__close_fd_cpu(evsel, cpu); +} + +void perf_evsel__free_fd(struct perf_evsel *evsel) +{ + xyarray__delete(evsel->fd); + evsel->fd = NULL; +} + +void perf_evsel__close(struct perf_evsel *evsel) +{ + if (evsel->fd == NULL) + return; + + perf_evsel__close_fd(evsel); + perf_evsel__free_fd(evsel); +} + +void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu) +{ + if (evsel->fd == NULL) + return; + + perf_evsel__close_fd_cpu(evsel, cpu); +} + +int perf_evsel__read_size(struct perf_evsel *evsel) +{ + u64 read_format = evsel->attr.read_format; + int entry = sizeof(u64); /* value */ + int size = 0; + int nr = 1; + + if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) + size += sizeof(u64); + + if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) + size += sizeof(u64); + + if (read_format & PERF_FORMAT_ID) + entry += sizeof(u64); + + if (read_format & PERF_FORMAT_GROUP) { + nr = evsel->nr_members; + size += sizeof(u64); + } + + size += entry * nr; + return size; +} + +int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, + struct perf_counts_values *count) +{ + size_t size = perf_evsel__read_size(evsel); + + memset(count, 0, sizeof(*count)); + + if (FD(evsel, cpu, thread) < 0) + return -EINVAL; + + if (readn(FD(evsel, cpu, thread), count->values, size) <= 0) + return -errno; + + return 0; +} + +static int perf_evsel__run_ioctl(struct perf_evsel *evsel, + int ioc, void *arg, + int cpu) +{ + int thread; + + for (thread = 0; thread < xyarray__max_y(evsel->fd); thread++) { + int fd = FD(evsel, cpu, thread), + err = ioctl(fd, ioc, arg); + + if (err) + return err; + } + + return 0; +} + +int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu) +{ + return perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_ENABLE, NULL, cpu); +} + +int perf_evsel__enable(struct perf_evsel *evsel) +{ + int i; + int err = 0; + + for (i = 0; i < xyarray__max_x(evsel->fd) && !err; i++) + err = perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_ENABLE, NULL, i); + return err; +} + +int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu) +{ + return perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_DISABLE, NULL, cpu); +} + +int perf_evsel__disable(struct perf_evsel *evsel) +{ + int i; + int err = 0; + + for (i = 0; i < xyarray__max_x(evsel->fd) && !err; i++) + err = perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_DISABLE, NULL, i); + return err; +} + +int perf_evsel__apply_filter(struct perf_evsel *evsel, const char *filter) +{ + int err = 0, i; + + for (i = 0; i < evsel->cpus->nr && !err; i++) + err = perf_evsel__run_ioctl(evsel, + PERF_EVENT_IOC_SET_FILTER, + (void *)filter, i); + return err; +} + +struct perf_cpu_map *perf_evsel__cpus(struct perf_evsel *evsel) +{ + return evsel->cpus; +} + +struct perf_thread_map *perf_evsel__threads(struct perf_evsel *evsel) +{ + return evsel->threads; +} + +struct perf_event_attr *perf_evsel__attr(struct perf_evsel *evsel) +{ + return &evsel->attr; +} + +int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads) +{ + if (ncpus == 0 || nthreads == 0) + return 0; + + if (evsel->system_wide) + nthreads = 1; + + evsel->sample_id = xyarray__new(ncpus, nthreads, sizeof(struct perf_sample_id)); + if (evsel->sample_id == NULL) + return -ENOMEM; + + evsel->id = zalloc(ncpus * nthreads * sizeof(u64)); + if (evsel->id == NULL) { + xyarray__delete(evsel->sample_id); + evsel->sample_id = NULL; + return -ENOMEM; + } + + return 0; +} + +void perf_evsel__free_id(struct perf_evsel *evsel) +{ + xyarray__delete(evsel->sample_id); + evsel->sample_id = NULL; + zfree(&evsel->id); + evsel->ids = 0; +} diff --git a/tools/lib/perf/include/internal/cpumap.h b/tools/lib/perf/include/internal/cpumap.h new file mode 100644 index 000000000000..840d4032587b --- /dev/null +++ b/tools/lib/perf/include/internal/cpumap.h @@ -0,0 +1,19 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_CPUMAP_H +#define __LIBPERF_INTERNAL_CPUMAP_H + +#include <linux/refcount.h> + +struct perf_cpu_map { + refcount_t refcnt; + int nr; + int map[]; +}; + +#ifndef MAX_NR_CPUS +#define MAX_NR_CPUS 2048 +#endif + +int perf_cpu_map__idx(struct perf_cpu_map *cpus, int cpu); + +#endif /* __LIBPERF_INTERNAL_CPUMAP_H */ diff --git a/tools/lib/perf/include/internal/evlist.h b/tools/lib/perf/include/internal/evlist.h new file mode 100644 index 000000000000..74dc8c3f0b66 --- /dev/null +++ b/tools/lib/perf/include/internal/evlist.h @@ -0,0 +1,127 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_EVLIST_H +#define __LIBPERF_INTERNAL_EVLIST_H + +#include <linux/list.h> +#include <api/fd/array.h> +#include <internal/evsel.h> + +#define PERF_EVLIST__HLIST_BITS 8 +#define PERF_EVLIST__HLIST_SIZE (1 << PERF_EVLIST__HLIST_BITS) + +struct perf_cpu_map; +struct perf_thread_map; +struct perf_mmap_param; + +struct perf_evlist { + struct list_head entries; + int nr_entries; + bool has_user_cpus; + struct perf_cpu_map *cpus; + struct perf_cpu_map *all_cpus; + struct perf_thread_map *threads; + int nr_mmaps; + size_t mmap_len; + struct fdarray pollfd; + struct hlist_head heads[PERF_EVLIST__HLIST_SIZE]; + struct perf_mmap *mmap; + struct perf_mmap *mmap_ovw; + struct perf_mmap *mmap_first; + struct perf_mmap *mmap_ovw_first; +}; + +typedef void +(*perf_evlist_mmap__cb_idx_t)(struct perf_evlist*, struct perf_mmap_param*, int, bool); +typedef struct perf_mmap* +(*perf_evlist_mmap__cb_get_t)(struct perf_evlist*, bool, int); +typedef int +(*perf_evlist_mmap__cb_mmap_t)(struct perf_mmap*, struct perf_mmap_param*, int, int); + +struct perf_evlist_mmap_ops { + perf_evlist_mmap__cb_idx_t idx; + perf_evlist_mmap__cb_get_t get; + perf_evlist_mmap__cb_mmap_t mmap; +}; + +int perf_evlist__alloc_pollfd(struct perf_evlist *evlist); +int perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd, + void *ptr, short revent); + +int perf_evlist__mmap_ops(struct perf_evlist *evlist, + struct perf_evlist_mmap_ops *ops, + struct perf_mmap_param *mp); + +void perf_evlist__init(struct perf_evlist *evlist); +void perf_evlist__exit(struct perf_evlist *evlist); + +/** + * __perf_evlist__for_each_entry - iterate thru all the evsels + * @list: list_head instance to iterate + * @evsel: struct perf_evsel iterator + */ +#define __perf_evlist__for_each_entry(list, evsel) \ + list_for_each_entry(evsel, list, node) + +/** + * evlist__for_each_entry - iterate thru all the evsels + * @evlist: perf_evlist instance to iterate + * @evsel: struct perf_evsel iterator + */ +#define perf_evlist__for_each_entry(evlist, evsel) \ + __perf_evlist__for_each_entry(&(evlist)->entries, evsel) + +/** + * __perf_evlist__for_each_entry_reverse - iterate thru all the evsels in reverse order + * @list: list_head instance to iterate + * @evsel: struct evsel iterator + */ +#define __perf_evlist__for_each_entry_reverse(list, evsel) \ + list_for_each_entry_reverse(evsel, list, node) + +/** + * perf_evlist__for_each_entry_reverse - iterate thru all the evsels in reverse order + * @evlist: evlist instance to iterate + * @evsel: struct evsel iterator + */ +#define perf_evlist__for_each_entry_reverse(evlist, evsel) \ + __perf_evlist__for_each_entry_reverse(&(evlist)->entries, evsel) + +/** + * __perf_evlist__for_each_entry_safe - safely iterate thru all the evsels + * @list: list_head instance to iterate + * @tmp: struct evsel temp iterator + * @evsel: struct evsel iterator + */ +#define __perf_evlist__for_each_entry_safe(list, tmp, evsel) \ + list_for_each_entry_safe(evsel, tmp, list, node) + +/** + * perf_evlist__for_each_entry_safe - safely iterate thru all the evsels + * @evlist: evlist instance to iterate + * @evsel: struct evsel iterator + * @tmp: struct evsel temp iterator + */ +#define perf_evlist__for_each_entry_safe(evlist, tmp, evsel) \ + __perf_evlist__for_each_entry_safe(&(evlist)->entries, tmp, evsel) + +static inline struct perf_evsel *perf_evlist__first(struct perf_evlist *evlist) +{ + return list_entry(evlist->entries.next, struct perf_evsel, node); +} + +static inline struct perf_evsel *perf_evlist__last(struct perf_evlist *evlist) +{ + return list_entry(evlist->entries.prev, struct perf_evsel, node); +} + +u64 perf_evlist__read_format(struct perf_evlist *evlist); + +void perf_evlist__id_add(struct perf_evlist *evlist, + struct perf_evsel *evsel, + int cpu, int thread, u64 id); + +int perf_evlist__id_add_fd(struct perf_evlist *evlist, + struct perf_evsel *evsel, + int cpu, int thread, int fd); + +#endif /* __LIBPERF_INTERNAL_EVLIST_H */ diff --git a/tools/lib/perf/include/internal/evsel.h b/tools/lib/perf/include/internal/evsel.h new file mode 100644 index 000000000000..1ffd083b235e --- /dev/null +++ b/tools/lib/perf/include/internal/evsel.h @@ -0,0 +1,63 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_EVSEL_H +#define __LIBPERF_INTERNAL_EVSEL_H + +#include <linux/types.h> +#include <linux/perf_event.h> +#include <stdbool.h> +#include <sys/types.h> + +struct perf_cpu_map; +struct perf_thread_map; +struct xyarray; + +/* + * Per fd, to map back from PERF_SAMPLE_ID to evsel, only used when there are + * more than one entry in the evlist. + */ +struct perf_sample_id { + struct hlist_node node; + u64 id; + struct perf_evsel *evsel; + /* + * 'idx' will be used for AUX area sampling. A sample will have AUX area + * data that will be queued for decoding, where there are separate + * queues for each CPU (per-cpu tracing) or task (per-thread tracing). + * The sample ID can be used to lookup 'idx' which is effectively the + * queue number. + */ + int idx; + int cpu; + pid_t tid; + + /* Holds total ID period value for PERF_SAMPLE_READ processing. */ + u64 period; +}; + +struct perf_evsel { + struct list_head node; + struct perf_event_attr attr; + struct perf_cpu_map *cpus; + struct perf_cpu_map *own_cpus; + struct perf_thread_map *threads; + struct xyarray *fd; + struct xyarray *sample_id; + u64 *id; + u32 ids; + + /* parse modifier helper */ + int nr_members; + bool system_wide; +}; + +void perf_evsel__init(struct perf_evsel *evsel, struct perf_event_attr *attr); +int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads); +void perf_evsel__close_fd(struct perf_evsel *evsel); +void perf_evsel__free_fd(struct perf_evsel *evsel); +int perf_evsel__read_size(struct perf_evsel *evsel); +int perf_evsel__apply_filter(struct perf_evsel *evsel, const char *filter); + +int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads); +void perf_evsel__free_id(struct perf_evsel *evsel); + +#endif /* __LIBPERF_INTERNAL_EVSEL_H */ diff --git a/tools/lib/perf/include/internal/lib.h b/tools/lib/perf/include/internal/lib.h new file mode 100644 index 000000000000..5175d491b2d4 --- /dev/null +++ b/tools/lib/perf/include/internal/lib.h @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_LIB_H +#define __LIBPERF_INTERNAL_LIB_H + +#include <sys/types.h> + +extern unsigned int page_size; + +ssize_t readn(int fd, void *buf, size_t n); +ssize_t writen(int fd, const void *buf, size_t n); + +#endif /* __LIBPERF_INTERNAL_CPUMAP_H */ diff --git a/tools/lib/perf/include/internal/mmap.h b/tools/lib/perf/include/internal/mmap.h new file mode 100644 index 000000000000..be7556e0a2b2 --- /dev/null +++ b/tools/lib/perf/include/internal/mmap.h @@ -0,0 +1,55 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_MMAP_H +#define __LIBPERF_INTERNAL_MMAP_H + +#include <linux/compiler.h> +#include <linux/refcount.h> +#include <linux/types.h> +#include <stdbool.h> + +/* perf sample has 16 bits size limit */ +#define PERF_SAMPLE_MAX_SIZE (1 << 16) + +struct perf_mmap; + +typedef void (*libperf_unmap_cb_t)(struct perf_mmap *map); + +/** + * struct perf_mmap - perf's ring buffer mmap details + * + * @refcnt - e.g. code using PERF_EVENT_IOC_SET_OUTPUT to share this + */ +struct perf_mmap { + void *base; + int mask; + int fd; + int cpu; + refcount_t refcnt; + u64 prev; + u64 start; + u64 end; + bool overwrite; + u64 flush; + libperf_unmap_cb_t unmap_cb; + char event_copy[PERF_SAMPLE_MAX_SIZE] __aligned(8); + struct perf_mmap *next; +}; + +struct perf_mmap_param { + int prot; + int mask; +}; + +size_t perf_mmap__mmap_len(struct perf_mmap *map); + +void perf_mmap__init(struct perf_mmap *map, struct perf_mmap *prev, + bool overwrite, libperf_unmap_cb_t unmap_cb); +int perf_mmap__mmap(struct perf_mmap *map, struct perf_mmap_param *mp, + int fd, int cpu); +void perf_mmap__munmap(struct perf_mmap *map); +void perf_mmap__get(struct perf_mmap *map); +void perf_mmap__put(struct perf_mmap *map); + +u64 perf_mmap__read_head(struct perf_mmap *map); + +#endif /* __LIBPERF_INTERNAL_MMAP_H */ diff --git a/tools/lib/perf/include/internal/tests.h b/tools/lib/perf/include/internal/tests.h new file mode 100644 index 000000000000..2093e8868a67 --- /dev/null +++ b/tools/lib/perf/include/internal/tests.h @@ -0,0 +1,33 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_TESTS_H +#define __LIBPERF_INTERNAL_TESTS_H + +#include <stdio.h> + +int tests_failed; + +#define __T_START \ +do { \ + fprintf(stdout, "- running %s...", __FILE__); \ + fflush(NULL); \ + tests_failed = 0; \ +} while (0) + +#define __T_END \ +do { \ + if (tests_failed) \ + fprintf(stdout, " FAILED (%d)\n", tests_failed); \ + else \ + fprintf(stdout, "OK\n"); \ +} while (0) + +#define __T(text, cond) \ +do { \ + if (!(cond)) { \ + fprintf(stderr, "FAILED %s:%d %s\n", __FILE__, __LINE__, text); \ + tests_failed++; \ + return -1; \ + } \ +} while (0) + +#endif /* __LIBPERF_INTERNAL_TESTS_H */ diff --git a/tools/lib/perf/include/internal/threadmap.h b/tools/lib/perf/include/internal/threadmap.h new file mode 100644 index 000000000000..df748baf9eda --- /dev/null +++ b/tools/lib/perf/include/internal/threadmap.h @@ -0,0 +1,23 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_THREADMAP_H +#define __LIBPERF_INTERNAL_THREADMAP_H + +#include <linux/refcount.h> +#include <sys/types.h> +#include <unistd.h> + +struct thread_map_data { + pid_t pid; + char *comm; +}; + +struct perf_thread_map { + refcount_t refcnt; + int nr; + int err_thread; + struct thread_map_data map[]; +}; + +struct perf_thread_map *perf_thread_map__realloc(struct perf_thread_map *map, int nr); + +#endif /* __LIBPERF_INTERNAL_THREADMAP_H */ diff --git a/tools/lib/perf/include/internal/xyarray.h b/tools/lib/perf/include/internal/xyarray.h new file mode 100644 index 000000000000..51e35d6c8ec4 --- /dev/null +++ b/tools/lib/perf/include/internal/xyarray.h @@ -0,0 +1,36 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_XYARRAY_H +#define __LIBPERF_INTERNAL_XYARRAY_H + +#include <linux/compiler.h> +#include <sys/types.h> + +struct xyarray { + size_t row_size; + size_t entry_size; + size_t entries; + size_t max_x; + size_t max_y; + char contents[] __aligned(8); +}; + +struct xyarray *xyarray__new(int xlen, int ylen, size_t entry_size); +void xyarray__delete(struct xyarray *xy); +void xyarray__reset(struct xyarray *xy); + +static inline void *xyarray__entry(struct xyarray *xy, int x, int y) +{ + return &xy->contents[x * xy->row_size + y * xy->entry_size]; +} + +static inline int xyarray__max_y(struct xyarray *xy) +{ + return xy->max_y; +} + +static inline int xyarray__max_x(struct xyarray *xy) +{ + return xy->max_x; +} + +#endif /* __LIBPERF_INTERNAL_XYARRAY_H */ diff --git a/tools/lib/perf/include/perf/core.h b/tools/lib/perf/include/perf/core.h new file mode 100644 index 000000000000..a3f6d68edad7 --- /dev/null +++ b/tools/lib/perf/include/perf/core.h @@ -0,0 +1,25 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_CORE_H +#define __LIBPERF_CORE_H + +#include <stdarg.h> + +#ifndef LIBPERF_API +#define LIBPERF_API __attribute__((visibility("default"))) +#endif + +enum libperf_print_level { + LIBPERF_ERR, + LIBPERF_WARN, + LIBPERF_INFO, + LIBPERF_DEBUG, + LIBPERF_DEBUG2, + LIBPERF_DEBUG3, +}; + +typedef int (*libperf_print_fn_t)(enum libperf_print_level level, + const char *, va_list ap); + +LIBPERF_API void libperf_init(libperf_print_fn_t fn); + +#endif /* __LIBPERF_CORE_H */ diff --git a/tools/lib/perf/include/perf/cpumap.h b/tools/lib/perf/include/perf/cpumap.h new file mode 100644 index 000000000000..6a17ad730cbc --- /dev/null +++ b/tools/lib/perf/include/perf/cpumap.h @@ -0,0 +1,28 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_CPUMAP_H +#define __LIBPERF_CPUMAP_H + +#include <perf/core.h> +#include <stdio.h> +#include <stdbool.h> + +struct perf_cpu_map; + +LIBPERF_API struct perf_cpu_map *perf_cpu_map__dummy_new(void); +LIBPERF_API struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list); +LIBPERF_API struct perf_cpu_map *perf_cpu_map__read(FILE *file); +LIBPERF_API struct perf_cpu_map *perf_cpu_map__get(struct perf_cpu_map *map); +LIBPERF_API struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig, + struct perf_cpu_map *other); +LIBPERF_API void perf_cpu_map__put(struct perf_cpu_map *map); +LIBPERF_API int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx); +LIBPERF_API int perf_cpu_map__nr(const struct perf_cpu_map *cpus); +LIBPERF_API bool perf_cpu_map__empty(const struct perf_cpu_map *map); +LIBPERF_API int perf_cpu_map__max(struct perf_cpu_map *map); + +#define perf_cpu_map__for_each_cpu(cpu, idx, cpus) \ + for ((idx) = 0, (cpu) = perf_cpu_map__cpu(cpus, idx); \ + (idx) < perf_cpu_map__nr(cpus); \ + (idx)++, (cpu) = perf_cpu_map__cpu(cpus, idx)) + +#endif /* __LIBPERF_CPUMAP_H */ diff --git a/tools/lib/perf/include/perf/event.h b/tools/lib/perf/include/perf/event.h new file mode 100644 index 000000000000..18106899cb4e --- /dev/null +++ b/tools/lib/perf/include/perf/event.h @@ -0,0 +1,385 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_EVENT_H +#define __LIBPERF_EVENT_H + +#include <linux/perf_event.h> +#include <linux/types.h> +#include <linux/limits.h> +#include <linux/bpf.h> +#include <sys/types.h> /* pid_t */ + +struct perf_record_mmap { + struct perf_event_header header; + __u32 pid, tid; + __u64 start; + __u64 len; + __u64 pgoff; + char filename[PATH_MAX]; +}; + +struct perf_record_mmap2 { + struct perf_event_header header; + __u32 pid, tid; + __u64 start; + __u64 len; + __u64 pgoff; + __u32 maj; + __u32 min; + __u64 ino; + __u64 ino_generation; + __u32 prot; + __u32 flags; + char filename[PATH_MAX]; +}; + +struct perf_record_comm { + struct perf_event_header header; + __u32 pid, tid; + char comm[16]; +}; + +struct perf_record_namespaces { + struct perf_event_header header; + __u32 pid, tid; + __u64 nr_namespaces; + struct perf_ns_link_info link_info[]; +}; + +struct perf_record_fork { + struct perf_event_header header; + __u32 pid, ppid; + __u32 tid, ptid; + __u64 time; +}; + +struct perf_record_lost { + struct perf_event_header header; + __u64 id; + __u64 lost; +}; + +struct perf_record_lost_samples { + struct perf_event_header header; + __u64 lost; +}; + +/* + * PERF_FORMAT_ENABLED | PERF_FORMAT_RUNNING | PERF_FORMAT_ID + */ +struct perf_record_read { + struct perf_event_header header; + __u32 pid, tid; + __u64 value; + __u64 time_enabled; + __u64 time_running; + __u64 id; +}; + +struct perf_record_throttle { + struct perf_event_header header; + __u64 time; + __u64 id; + __u64 stream_id; +}; + +#ifndef KSYM_NAME_LEN +#define KSYM_NAME_LEN 256 +#endif + +struct perf_record_ksymbol { + struct perf_event_header header; + __u64 addr; + __u32 len; + __u16 ksym_type; + __u16 flags; + char name[KSYM_NAME_LEN]; +}; + +struct perf_record_bpf_event { + struct perf_event_header header; + __u16 type; + __u16 flags; + __u32 id; + + /* for bpf_prog types */ + __u8 tag[BPF_TAG_SIZE]; // prog tag +}; + +struct perf_record_sample { + struct perf_event_header header; + __u64 array[]; +}; + +struct perf_record_switch { + struct perf_event_header header; + __u32 next_prev_pid; + __u32 next_prev_tid; +}; + +struct perf_record_header_attr { + struct perf_event_header header; + struct perf_event_attr attr; + __u64 id[]; +}; + +enum { + PERF_CPU_MAP__CPUS = 0, + PERF_CPU_MAP__MASK = 1, +}; + +struct cpu_map_entries { + __u16 nr; + __u16 cpu[]; +}; + +struct perf_record_record_cpu_map { + __u16 nr; + __u16 long_size; + unsigned long mask[]; +}; + +struct perf_record_cpu_map_data { + __u16 type; + char data[]; +}; + +struct perf_record_cpu_map { + struct perf_event_header header; + struct perf_record_cpu_map_data data; +}; + +enum { + PERF_EVENT_UPDATE__UNIT = 0, + PERF_EVENT_UPDATE__SCALE = 1, + PERF_EVENT_UPDATE__NAME = 2, + PERF_EVENT_UPDATE__CPUS = 3, +}; + +struct perf_record_event_update_cpus { + struct perf_record_cpu_map_data cpus; +}; + +struct perf_record_event_update_scale { + double scale; +}; + +struct perf_record_event_update { + struct perf_event_header header; + __u64 type; + __u64 id; + char data[]; +}; + +#define MAX_EVENT_NAME 64 + +struct perf_trace_event_type { + __u64 event_id; + char name[MAX_EVENT_NAME]; +}; + +struct perf_record_header_event_type { + struct perf_event_header header; + struct perf_trace_event_type event_type; +}; + +struct perf_record_header_tracing_data { + struct perf_event_header header; + __u32 size; +}; + +struct perf_record_header_build_id { + struct perf_event_header header; + pid_t pid; + __u8 build_id[24]; + char filename[]; +}; + +struct id_index_entry { + __u64 id; + __u64 idx; + __u64 cpu; + __u64 tid; +}; + +struct perf_record_id_index { + struct perf_event_header header; + __u64 nr; + struct id_index_entry entries[0]; +}; + +struct perf_record_auxtrace_info { + struct perf_event_header header; + __u32 type; + __u32 reserved__; /* For alignment */ + __u64 priv[]; +}; + +struct perf_record_auxtrace { + struct perf_event_header header; + __u64 size; + __u64 offset; + __u64 reference; + __u32 idx; + __u32 tid; + __u32 cpu; + __u32 reserved__; /* For alignment */ +}; + +#define MAX_AUXTRACE_ERROR_MSG 64 + +struct perf_record_auxtrace_error { + struct perf_event_header header; + __u32 type; + __u32 code; + __u32 cpu; + __u32 pid; + __u32 tid; + __u32 fmt; + __u64 ip; + __u64 time; + char msg[MAX_AUXTRACE_ERROR_MSG]; +}; + +struct perf_record_aux { + struct perf_event_header header; + __u64 aux_offset; + __u64 aux_size; + __u64 flags; +}; + +struct perf_record_itrace_start { + struct perf_event_header header; + __u32 pid; + __u32 tid; +}; + +struct perf_record_thread_map_entry { + __u64 pid; + char comm[16]; +}; + +struct perf_record_thread_map { + struct perf_event_header header; + __u64 nr; + struct perf_record_thread_map_entry entries[]; +}; + +enum { + PERF_STAT_CONFIG_TERM__AGGR_MODE = 0, + PERF_STAT_CONFIG_TERM__INTERVAL = 1, + PERF_STAT_CONFIG_TERM__SCALE = 2, + PERF_STAT_CONFIG_TERM__MAX = 3, +}; + +struct perf_record_stat_config_entry { + __u64 tag; + __u64 val; +}; + +struct perf_record_stat_config { + struct perf_event_header header; + __u64 nr; + struct perf_record_stat_config_entry data[]; +}; + +struct perf_record_stat { + struct perf_event_header header; + + __u64 id; + __u32 cpu; + __u32 thread; + + union { + struct { + __u64 val; + __u64 ena; + __u64 run; + }; + __u64 values[3]; + }; +}; + +struct perf_record_stat_round { + struct perf_event_header header; + __u64 type; + __u64 time; +}; + +struct perf_record_time_conv { + struct perf_event_header header; + __u64 time_shift; + __u64 time_mult; + __u64 time_zero; +}; + +struct perf_record_header_feature { + struct perf_event_header header; + __u64 feat_id; + char data[]; +}; + +struct perf_record_compressed { + struct perf_event_header header; + char data[]; +}; + +enum perf_user_event_type { /* above any possible kernel type */ + PERF_RECORD_USER_TYPE_START = 64, + PERF_RECORD_HEADER_ATTR = 64, + PERF_RECORD_HEADER_EVENT_TYPE = 65, /* deprecated */ + PERF_RECORD_HEADER_TRACING_DATA = 66, + PERF_RECORD_HEADER_BUILD_ID = 67, + PERF_RECORD_FINISHED_ROUND = 68, + PERF_RECORD_ID_INDEX = 69, + PERF_RECORD_AUXTRACE_INFO = 70, + PERF_RECORD_AUXTRACE = 71, + PERF_RECORD_AUXTRACE_ERROR = 72, + PERF_RECORD_THREAD_MAP = 73, + PERF_RECORD_CPU_MAP = 74, + PERF_RECORD_STAT_CONFIG = 75, + PERF_RECORD_STAT = 76, + PERF_RECORD_STAT_ROUND = 77, + PERF_RECORD_EVENT_UPDATE = 78, + PERF_RECORD_TIME_CONV = 79, + PERF_RECORD_HEADER_FEATURE = 80, + PERF_RECORD_COMPRESSED = 81, + PERF_RECORD_HEADER_MAX +}; + +union perf_event { + struct perf_event_header header; + struct perf_record_mmap mmap; + struct perf_record_mmap2 mmap2; + struct perf_record_comm comm; + struct perf_record_namespaces namespaces; + struct perf_record_fork fork; + struct perf_record_lost lost; + struct perf_record_lost_samples lost_samples; + struct perf_record_read read; + struct perf_record_throttle throttle; + struct perf_record_sample sample; + struct perf_record_bpf_event bpf; + struct perf_record_ksymbol ksymbol; + struct perf_record_header_attr attr; + struct perf_record_event_update event_update; + struct perf_record_header_event_type event_type; + struct perf_record_header_tracing_data tracing_data; + struct perf_record_header_build_id build_id; + struct perf_record_id_index id_index; + struct perf_record_auxtrace_info auxtrace_info; + struct perf_record_auxtrace auxtrace; + struct perf_record_auxtrace_error auxtrace_error; + struct perf_record_aux aux; + struct perf_record_itrace_start itrace_start; + struct perf_record_switch context_switch; + struct perf_record_thread_map thread_map; + struct perf_record_cpu_map cpu_map; + struct perf_record_stat_config stat_config; + struct perf_record_stat stat; + struct perf_record_stat_round stat_round; + struct perf_record_time_conv time_conv; + struct perf_record_header_feature feat; + struct perf_record_compressed pack; +}; + +#endif /* __LIBPERF_EVENT_H */ diff --git a/tools/lib/perf/include/perf/evlist.h b/tools/lib/perf/include/perf/evlist.h new file mode 100644 index 000000000000..0a7479dc13bf --- /dev/null +++ b/tools/lib/perf/include/perf/evlist.h @@ -0,0 +1,49 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_EVLIST_H +#define __LIBPERF_EVLIST_H + +#include <perf/core.h> +#include <stdbool.h> + +struct perf_evlist; +struct perf_evsel; +struct perf_cpu_map; +struct perf_thread_map; + +LIBPERF_API void perf_evlist__add(struct perf_evlist *evlist, + struct perf_evsel *evsel); +LIBPERF_API void perf_evlist__remove(struct perf_evlist *evlist, + struct perf_evsel *evsel); +LIBPERF_API struct perf_evlist *perf_evlist__new(void); +LIBPERF_API void perf_evlist__delete(struct perf_evlist *evlist); +LIBPERF_API struct perf_evsel* perf_evlist__next(struct perf_evlist *evlist, + struct perf_evsel *evsel); +LIBPERF_API int perf_evlist__open(struct perf_evlist *evlist); +LIBPERF_API void perf_evlist__close(struct perf_evlist *evlist); +LIBPERF_API void perf_evlist__enable(struct perf_evlist *evlist); +LIBPERF_API void perf_evlist__disable(struct perf_evlist *evlist); + +#define perf_evlist__for_each_evsel(evlist, pos) \ + for ((pos) = perf_evlist__next((evlist), NULL); \ + (pos) != NULL; \ + (pos) = perf_evlist__next((evlist), (pos))) + +LIBPERF_API void perf_evlist__set_maps(struct perf_evlist *evlist, + struct perf_cpu_map *cpus, + struct perf_thread_map *threads); +LIBPERF_API int perf_evlist__poll(struct perf_evlist *evlist, int timeout); +LIBPERF_API int perf_evlist__filter_pollfd(struct perf_evlist *evlist, + short revents_and_mask); + +LIBPERF_API int perf_evlist__mmap(struct perf_evlist *evlist, int pages); +LIBPERF_API void perf_evlist__munmap(struct perf_evlist *evlist); + +LIBPERF_API struct perf_mmap *perf_evlist__next_mmap(struct perf_evlist *evlist, + struct perf_mmap *map, + bool overwrite); +#define perf_evlist__for_each_mmap(evlist, pos, overwrite) \ + for ((pos) = perf_evlist__next_mmap((evlist), NULL, overwrite); \ + (pos) != NULL; \ + (pos) = perf_evlist__next_mmap((evlist), (pos), overwrite)) + +#endif /* __LIBPERF_EVLIST_H */ diff --git a/tools/lib/perf/include/perf/evsel.h b/tools/lib/perf/include/perf/evsel.h new file mode 100644 index 000000000000..c82ec39a4ad0 --- /dev/null +++ b/tools/lib/perf/include/perf/evsel.h @@ -0,0 +1,40 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_EVSEL_H +#define __LIBPERF_EVSEL_H + +#include <stdint.h> +#include <perf/core.h> + +struct perf_evsel; +struct perf_event_attr; +struct perf_cpu_map; +struct perf_thread_map; + +struct perf_counts_values { + union { + struct { + uint64_t val; + uint64_t ena; + uint64_t run; + }; + uint64_t values[3]; + }; +}; + +LIBPERF_API struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr); +LIBPERF_API void perf_evsel__delete(struct perf_evsel *evsel); +LIBPERF_API int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus, + struct perf_thread_map *threads); +LIBPERF_API void perf_evsel__close(struct perf_evsel *evsel); +LIBPERF_API void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu); +LIBPERF_API int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, + struct perf_counts_values *count); +LIBPERF_API int perf_evsel__enable(struct perf_evsel *evsel); +LIBPERF_API int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu); +LIBPERF_API int perf_evsel__disable(struct perf_evsel *evsel); +LIBPERF_API int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu); +LIBPERF_API struct perf_cpu_map *perf_evsel__cpus(struct perf_evsel *evsel); +LIBPERF_API struct perf_thread_map *perf_evsel__threads(struct perf_evsel *evsel); +LIBPERF_API struct perf_event_attr *perf_evsel__attr(struct perf_evsel *evsel); + +#endif /* __LIBPERF_EVSEL_H */ diff --git a/tools/lib/perf/include/perf/mmap.h b/tools/lib/perf/include/perf/mmap.h new file mode 100644 index 000000000000..9508ad90d8b9 --- /dev/null +++ b/tools/lib/perf/include/perf/mmap.h @@ -0,0 +1,15 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_MMAP_H +#define __LIBPERF_MMAP_H + +#include <perf/core.h> + +struct perf_mmap; +union perf_event; + +LIBPERF_API void perf_mmap__consume(struct perf_mmap *map); +LIBPERF_API int perf_mmap__read_init(struct perf_mmap *map); +LIBPERF_API void perf_mmap__read_done(struct perf_mmap *map); +LIBPERF_API union perf_event *perf_mmap__read_event(struct perf_mmap *map); + +#endif /* __LIBPERF_MMAP_H */ diff --git a/tools/lib/perf/include/perf/threadmap.h b/tools/lib/perf/include/perf/threadmap.h new file mode 100644 index 000000000000..a7c50de8d010 --- /dev/null +++ b/tools/lib/perf/include/perf/threadmap.h @@ -0,0 +1,20 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_THREADMAP_H +#define __LIBPERF_THREADMAP_H + +#include <perf/core.h> +#include <sys/types.h> + +struct perf_thread_map; + +LIBPERF_API struct perf_thread_map *perf_thread_map__new_dummy(void); + +LIBPERF_API void perf_thread_map__set_pid(struct perf_thread_map *map, int thread, pid_t pid); +LIBPERF_API char *perf_thread_map__comm(struct perf_thread_map *map, int thread); +LIBPERF_API int perf_thread_map__nr(struct perf_thread_map *threads); +LIBPERF_API pid_t perf_thread_map__pid(struct perf_thread_map *map, int thread); + +LIBPERF_API struct perf_thread_map *perf_thread_map__get(struct perf_thread_map *map); +LIBPERF_API void perf_thread_map__put(struct perf_thread_map *map); + +#endif /* __LIBPERF_THREADMAP_H */ diff --git a/tools/lib/perf/internal.h b/tools/lib/perf/internal.h new file mode 100644 index 000000000000..2c27e158de6b --- /dev/null +++ b/tools/lib/perf/internal.h @@ -0,0 +1,23 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_H +#define __LIBPERF_INTERNAL_H + +#include <perf/core.h> + +void libperf_print(enum libperf_print_level level, + const char *format, ...) + __attribute__((format(printf, 2, 3))); + +#define __pr(level, fmt, ...) \ +do { \ + libperf_print(level, "libperf: " fmt, ##__VA_ARGS__); \ +} while (0) + +#define pr_err(fmt, ...) __pr(LIBPERF_ERR, fmt, ##__VA_ARGS__) +#define pr_warning(fmt, ...) __pr(LIBPERF_WARN, fmt, ##__VA_ARGS__) +#define pr_info(fmt, ...) __pr(LIBPERF_INFO, fmt, ##__VA_ARGS__) +#define pr_debug(fmt, ...) __pr(LIBPERF_DEBUG, fmt, ##__VA_ARGS__) +#define pr_debug2(fmt, ...) __pr(LIBPERF_DEBUG2, fmt, ##__VA_ARGS__) +#define pr_debug3(fmt, ...) __pr(LIBPERF_DEBUG3, fmt, ##__VA_ARGS__) + +#endif /* __LIBPERF_INTERNAL_H */ diff --git a/tools/lib/perf/lib.c b/tools/lib/perf/lib.c new file mode 100644 index 000000000000..18658931fc71 --- /dev/null +++ b/tools/lib/perf/lib.c @@ -0,0 +1,48 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <unistd.h> +#include <stdbool.h> +#include <errno.h> +#include <linux/kernel.h> +#include <internal/lib.h> + +unsigned int page_size; + +static ssize_t ion(bool is_read, int fd, void *buf, size_t n) +{ + void *buf_start = buf; + size_t left = n; + + while (left) { + /* buf must be treated as const if !is_read. */ + ssize_t ret = is_read ? read(fd, buf, left) : + write(fd, buf, left); + + if (ret < 0 && errno == EINTR) + continue; + if (ret <= 0) + return ret; + + left -= ret; + buf += ret; + } + + BUG_ON((size_t)(buf - buf_start) != n); + return n; +} + +/* + * Read exactly 'n' bytes or return an error. + */ +ssize_t readn(int fd, void *buf, size_t n) +{ + return ion(true, fd, buf, n); +} + +/* + * Write exactly 'n' bytes or return an error. + */ +ssize_t writen(int fd, const void *buf, size_t n) +{ + /* ion does not modify buf. */ + return ion(false, fd, (void *)buf, n); +} diff --git a/tools/lib/perf/libperf.map b/tools/lib/perf/libperf.map new file mode 100644 index 000000000000..7be1af8a546c --- /dev/null +++ b/tools/lib/perf/libperf.map @@ -0,0 +1,51 @@ +LIBPERF_0.0.1 { + global: + libperf_init; + perf_cpu_map__dummy_new; + perf_cpu_map__get; + perf_cpu_map__put; + perf_cpu_map__new; + perf_cpu_map__read; + perf_cpu_map__nr; + perf_cpu_map__cpu; + perf_cpu_map__empty; + perf_cpu_map__max; + perf_thread_map__new_dummy; + perf_thread_map__set_pid; + perf_thread_map__comm; + perf_thread_map__nr; + perf_thread_map__pid; + perf_thread_map__get; + perf_thread_map__put; + perf_evsel__new; + perf_evsel__delete; + perf_evsel__enable; + perf_evsel__disable; + perf_evsel__open; + perf_evsel__close; + perf_evsel__read; + perf_evsel__cpus; + perf_evsel__threads; + perf_evsel__attr; + perf_evlist__new; + perf_evlist__delete; + perf_evlist__open; + perf_evlist__close; + perf_evlist__enable; + perf_evlist__disable; + perf_evlist__add; + perf_evlist__remove; + perf_evlist__next; + perf_evlist__set_maps; + perf_evlist__poll; + perf_evlist__mmap; + perf_evlist__munmap; + perf_evlist__filter_pollfd; + perf_evlist__next_mmap; + perf_mmap__consume; + perf_mmap__read_init; + perf_mmap__read_done; + perf_mmap__read_event; + local: + *; +}; diff --git a/tools/lib/perf/libperf.pc.template b/tools/lib/perf/libperf.pc.template new file mode 100644 index 000000000000..117e4a237b55 --- /dev/null +++ b/tools/lib/perf/libperf.pc.template @@ -0,0 +1,11 @@ +# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) + +prefix=@PREFIX@ +libdir=@LIBDIR@ +includedir=${prefix}/include + +Name: libperf +Description: perf library +Version: @VERSION@ +Libs: -L${libdir} -lperf +Cflags: -I${includedir} diff --git a/tools/lib/perf/mmap.c b/tools/lib/perf/mmap.c new file mode 100644 index 000000000000..79d5ed6c38cc --- /dev/null +++ b/tools/lib/perf/mmap.c @@ -0,0 +1,275 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <sys/mman.h> +#include <inttypes.h> +#include <asm/bug.h> +#include <errno.h> +#include <string.h> +#include <linux/ring_buffer.h> +#include <linux/perf_event.h> +#include <perf/mmap.h> +#include <perf/event.h> +#include <internal/mmap.h> +#include <internal/lib.h> +#include <linux/kernel.h> +#include "internal.h" + +void perf_mmap__init(struct perf_mmap *map, struct perf_mmap *prev, + bool overwrite, libperf_unmap_cb_t unmap_cb) +{ + map->fd = -1; + map->overwrite = overwrite; + map->unmap_cb = unmap_cb; + refcount_set(&map->refcnt, 0); + if (prev) + prev->next = map; +} + +size_t perf_mmap__mmap_len(struct perf_mmap *map) +{ + return map->mask + 1 + page_size; +} + +int perf_mmap__mmap(struct perf_mmap *map, struct perf_mmap_param *mp, + int fd, int cpu) +{ + map->prev = 0; + map->mask = mp->mask; + map->base = mmap(NULL, perf_mmap__mmap_len(map), mp->prot, + MAP_SHARED, fd, 0); + if (map->base == MAP_FAILED) { + map->base = NULL; + return -1; + } + + map->fd = fd; + map->cpu = cpu; + return 0; +} + +void perf_mmap__munmap(struct perf_mmap *map) +{ + if (map && map->base != NULL) { + munmap(map->base, perf_mmap__mmap_len(map)); + map->base = NULL; + map->fd = -1; + refcount_set(&map->refcnt, 0); + } + if (map && map->unmap_cb) + map->unmap_cb(map); +} + +void perf_mmap__get(struct perf_mmap *map) +{ + refcount_inc(&map->refcnt); +} + +void perf_mmap__put(struct perf_mmap *map) +{ + BUG_ON(map->base && refcount_read(&map->refcnt) == 0); + + if (refcount_dec_and_test(&map->refcnt)) + perf_mmap__munmap(map); +} + +static inline void perf_mmap__write_tail(struct perf_mmap *md, u64 tail) +{ + ring_buffer_write_tail(md->base, tail); +} + +u64 perf_mmap__read_head(struct perf_mmap *map) +{ + return ring_buffer_read_head(map->base); +} + +static bool perf_mmap__empty(struct perf_mmap *map) +{ + struct perf_event_mmap_page *pc = map->base; + + return perf_mmap__read_head(map) == map->prev && !pc->aux_size; +} + +void perf_mmap__consume(struct perf_mmap *map) +{ + if (!map->overwrite) { + u64 old = map->prev; + + perf_mmap__write_tail(map, old); + } + + if (refcount_read(&map->refcnt) == 1 && perf_mmap__empty(map)) + perf_mmap__put(map); +} + +static int overwrite_rb_find_range(void *buf, int mask, u64 *start, u64 *end) +{ + struct perf_event_header *pheader; + u64 evt_head = *start; + int size = mask + 1; + + pr_debug2("%s: buf=%p, start=%"PRIx64"\n", __func__, buf, *start); + pheader = (struct perf_event_header *)(buf + (*start & mask)); + while (true) { + if (evt_head - *start >= (unsigned int)size) { + pr_debug("Finished reading overwrite ring buffer: rewind\n"); + if (evt_head - *start > (unsigned int)size) + evt_head -= pheader->size; + *end = evt_head; + return 0; + } + + pheader = (struct perf_event_header *)(buf + (evt_head & mask)); + + if (pheader->size == 0) { + pr_debug("Finished reading overwrite ring buffer: get start\n"); + *end = evt_head; + return 0; + } + + evt_head += pheader->size; + pr_debug3("move evt_head: %"PRIx64"\n", evt_head); + } + WARN_ONCE(1, "Shouldn't get here\n"); + return -1; +} + +/* + * Report the start and end of the available data in ringbuffer + */ +static int __perf_mmap__read_init(struct perf_mmap *md) +{ + u64 head = perf_mmap__read_head(md); + u64 old = md->prev; + unsigned char *data = md->base + page_size; + unsigned long size; + + md->start = md->overwrite ? head : old; + md->end = md->overwrite ? old : head; + + if ((md->end - md->start) < md->flush) + return -EAGAIN; + + size = md->end - md->start; + if (size > (unsigned long)(md->mask) + 1) { + if (!md->overwrite) { + WARN_ONCE(1, "failed to keep up with mmap data. (warn only once)\n"); + + md->prev = head; + perf_mmap__consume(md); + return -EAGAIN; + } + + /* + * Backward ring buffer is full. We still have a chance to read + * most of data from it. + */ + if (overwrite_rb_find_range(data, md->mask, &md->start, &md->end)) + return -EINVAL; + } + + return 0; +} + +int perf_mmap__read_init(struct perf_mmap *map) +{ + /* + * Check if event was unmapped due to a POLLHUP/POLLERR. + */ + if (!refcount_read(&map->refcnt)) + return -ENOENT; + + return __perf_mmap__read_init(map); +} + +/* + * Mandatory for overwrite mode + * The direction of overwrite mode is backward. + * The last perf_mmap__read() will set tail to map->core.prev. + * Need to correct the map->core.prev to head which is the end of next read. + */ +void perf_mmap__read_done(struct perf_mmap *map) +{ + /* + * Check if event was unmapped due to a POLLHUP/POLLERR. + */ + if (!refcount_read(&map->refcnt)) + return; + + map->prev = perf_mmap__read_head(map); +} + +/* When check_messup is true, 'end' must points to a good entry */ +static union perf_event *perf_mmap__read(struct perf_mmap *map, + u64 *startp, u64 end) +{ + unsigned char *data = map->base + page_size; + union perf_event *event = NULL; + int diff = end - *startp; + + if (diff >= (int)sizeof(event->header)) { + size_t size; + + event = (union perf_event *)&data[*startp & map->mask]; + size = event->header.size; + + if (size < sizeof(event->header) || diff < (int)size) + return NULL; + + /* + * Event straddles the mmap boundary -- header should always + * be inside due to u64 alignment of output. + */ + if ((*startp & map->mask) + size != ((*startp + size) & map->mask)) { + unsigned int offset = *startp; + unsigned int len = min(sizeof(*event), size), cpy; + void *dst = map->event_copy; + + do { + cpy = min(map->mask + 1 - (offset & map->mask), len); + memcpy(dst, &data[offset & map->mask], cpy); + offset += cpy; + dst += cpy; + len -= cpy; + } while (len); + + event = (union perf_event *)map->event_copy; + } + + *startp += size; + } + + return event; +} + +/* + * Read event from ring buffer one by one. + * Return one event for each call. + * + * Usage: + * perf_mmap__read_init() + * while(event = perf_mmap__read_event()) { + * //process the event + * perf_mmap__consume() + * } + * perf_mmap__read_done() + */ +union perf_event *perf_mmap__read_event(struct perf_mmap *map) +{ + union perf_event *event; + + /* + * Check if event was unmapped due to a POLLHUP/POLLERR. + */ + if (!refcount_read(&map->refcnt)) + return NULL; + + /* non-overwirte doesn't pause the ringbuffer */ + if (!map->overwrite) + map->end = perf_mmap__read_head(map); + + event = perf_mmap__read(map, &map->start, map->end); + + if (!map->overwrite) + map->prev = map->start; + + return event; +} diff --git a/tools/lib/perf/tests/Makefile b/tools/lib/perf/tests/Makefile new file mode 100644 index 000000000000..96841775feaf --- /dev/null +++ b/tools/lib/perf/tests/Makefile @@ -0,0 +1,38 @@ +# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) + +TESTS = test-cpumap test-threadmap test-evlist test-evsel + +TESTS_SO := $(addsuffix -so,$(TESTS)) +TESTS_A := $(addsuffix -a,$(TESTS)) + +# Set compile option CFLAGS +ifdef EXTRA_CFLAGS + CFLAGS := $(EXTRA_CFLAGS) +else + CFLAGS := -g -Wall +endif + +all: + +include $(srctree)/tools/scripts/Makefile.include + +INCLUDE = -I$(srctree)/tools/lib/perf/include -I$(srctree)/tools/include -I$(srctree)/tools/lib + +$(TESTS_A): FORCE + $(QUIET_LINK)$(CC) $(INCLUDE) $(CFLAGS) -o $@ $(subst -a,.c,$@) ../libperf.a $(LIBAPI) + +$(TESTS_SO): FORCE + $(QUIET_LINK)$(CC) $(INCLUDE) $(CFLAGS) -L.. -o $@ $(subst -so,.c,$@) $(LIBAPI) -lperf + +all: $(TESTS_A) $(TESTS_SO) + +run: + @echo "running static:" + @for i in $(TESTS_A); do ./$$i; done + @echo "running dynamic:" + @for i in $(TESTS_SO); do LD_LIBRARY_PATH=../ ./$$i; done + +clean: + $(call QUIET_CLEAN, tests)$(RM) $(TESTS_A) $(TESTS_SO) + +.PHONY: all clean FORCE diff --git a/tools/lib/perf/tests/test-cpumap.c b/tools/lib/perf/tests/test-cpumap.c new file mode 100644 index 000000000000..c8d45091e7c2 --- /dev/null +++ b/tools/lib/perf/tests/test-cpumap.c @@ -0,0 +1,31 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <stdarg.h> +#include <stdio.h> +#include <perf/cpumap.h> +#include <internal/tests.h> + +static int libperf_print(enum libperf_print_level level, + const char *fmt, va_list ap) +{ + return vfprintf(stderr, fmt, ap); +} + +int main(int argc, char **argv) +{ + struct perf_cpu_map *cpus; + + __T_START; + + libperf_init(libperf_print); + + cpus = perf_cpu_map__dummy_new(); + if (!cpus) + return -1; + + perf_cpu_map__get(cpus); + perf_cpu_map__put(cpus); + perf_cpu_map__put(cpus); + + __T_END; + return 0; +} diff --git a/tools/lib/perf/tests/test-evlist.c b/tools/lib/perf/tests/test-evlist.c new file mode 100644 index 000000000000..6d8ebe0c2504 --- /dev/null +++ b/tools/lib/perf/tests/test-evlist.c @@ -0,0 +1,413 @@ +// SPDX-License-Identifier: GPL-2.0 +#define _GNU_SOURCE // needed for sched.h to get sched_[gs]etaffinity and CPU_(ZERO,SET) +#include <sched.h> +#include <stdio.h> +#include <stdarg.h> +#include <unistd.h> +#include <stdlib.h> +#include <linux/perf_event.h> +#include <linux/limits.h> +#include <sys/types.h> +#include <sys/wait.h> +#include <sys/prctl.h> +#include <perf/cpumap.h> +#include <perf/threadmap.h> +#include <perf/evlist.h> +#include <perf/evsel.h> +#include <perf/mmap.h> +#include <perf/event.h> +#include <internal/tests.h> +#include <api/fs/fs.h> + +static int libperf_print(enum libperf_print_level level, + const char *fmt, va_list ap) +{ + return vfprintf(stderr, fmt, ap); +} + +static int test_stat_cpu(void) +{ + struct perf_cpu_map *cpus; + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_event_attr attr1 = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_CPU_CLOCK, + }; + struct perf_event_attr attr2 = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_TASK_CLOCK, + }; + int err, cpu, tmp; + + cpus = perf_cpu_map__new(NULL); + __T("failed to create cpus", cpus); + + evlist = perf_evlist__new(); + __T("failed to create evlist", evlist); + + evsel = perf_evsel__new(&attr1); + __T("failed to create evsel1", evsel); + + perf_evlist__add(evlist, evsel); + + evsel = perf_evsel__new(&attr2); + __T("failed to create evsel2", evsel); + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, cpus, NULL); + + err = perf_evlist__open(evlist); + __T("failed to open evsel", err == 0); + + perf_evlist__for_each_evsel(evlist, evsel) { + cpus = perf_evsel__cpus(evsel); + + perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { + struct perf_counts_values counts = { .val = 0 }; + + perf_evsel__read(evsel, cpu, 0, &counts); + __T("failed to read value for evsel", counts.val != 0); + } + } + + perf_evlist__close(evlist); + perf_evlist__delete(evlist); + + perf_cpu_map__put(cpus); + return 0; +} + +static int test_stat_thread(void) +{ + struct perf_counts_values counts = { .val = 0 }; + struct perf_thread_map *threads; + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_event_attr attr1 = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_CPU_CLOCK, + }; + struct perf_event_attr attr2 = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_TASK_CLOCK, + }; + int err; + + threads = perf_thread_map__new_dummy(); + __T("failed to create threads", threads); + + perf_thread_map__set_pid(threads, 0, 0); + + evlist = perf_evlist__new(); + __T("failed to create evlist", evlist); + + evsel = perf_evsel__new(&attr1); + __T("failed to create evsel1", evsel); + + perf_evlist__add(evlist, evsel); + + evsel = perf_evsel__new(&attr2); + __T("failed to create evsel2", evsel); + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, NULL, threads); + + err = perf_evlist__open(evlist); + __T("failed to open evsel", err == 0); + + perf_evlist__for_each_evsel(evlist, evsel) { + perf_evsel__read(evsel, 0, 0, &counts); + __T("failed to read value for evsel", counts.val != 0); + } + + perf_evlist__close(evlist); + perf_evlist__delete(evlist); + + perf_thread_map__put(threads); + return 0; +} + +static int test_stat_thread_enable(void) +{ + struct perf_counts_values counts = { .val = 0 }; + struct perf_thread_map *threads; + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_event_attr attr1 = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_CPU_CLOCK, + .disabled = 1, + }; + struct perf_event_attr attr2 = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_TASK_CLOCK, + .disabled = 1, + }; + int err; + + threads = perf_thread_map__new_dummy(); + __T("failed to create threads", threads); + + perf_thread_map__set_pid(threads, 0, 0); + + evlist = perf_evlist__new(); + __T("failed to create evlist", evlist); + + evsel = perf_evsel__new(&attr1); + __T("failed to create evsel1", evsel); + + perf_evlist__add(evlist, evsel); + + evsel = perf_evsel__new(&attr2); + __T("failed to create evsel2", evsel); + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, NULL, threads); + + err = perf_evlist__open(evlist); + __T("failed to open evsel", err == 0); + + perf_evlist__for_each_evsel(evlist, evsel) { + perf_evsel__read(evsel, 0, 0, &counts); + __T("failed to read value for evsel", counts.val == 0); + } + + perf_evlist__enable(evlist); + + perf_evlist__for_each_evsel(evlist, evsel) { + perf_evsel__read(evsel, 0, 0, &counts); + __T("failed to read value for evsel", counts.val != 0); + } + + perf_evlist__disable(evlist); + + perf_evlist__close(evlist); + perf_evlist__delete(evlist); + + perf_thread_map__put(threads); + return 0; +} + +static int test_mmap_thread(void) +{ + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_mmap *map; + struct perf_cpu_map *cpus; + struct perf_thread_map *threads; + struct perf_event_attr attr = { + .type = PERF_TYPE_TRACEPOINT, + .sample_period = 1, + .wakeup_watermark = 1, + .disabled = 1, + }; + char path[PATH_MAX]; + int id, err, pid, go_pipe[2]; + union perf_event *event; + char bf; + int count = 0; + + snprintf(path, PATH_MAX, "%s/kernel/debug/tracing/events/syscalls/sys_enter_prctl/id", + sysfs__mountpoint()); + + if (filename__read_int(path, &id)) { + fprintf(stderr, "error: failed to get tracepoint id: %s\n", path); + return -1; + } + + attr.config = id; + + err = pipe(go_pipe); + __T("failed to create pipe", err == 0); + + fflush(NULL); + + pid = fork(); + if (!pid) { + int i; + + read(go_pipe[0], &bf, 1); + + /* Generate 100 prctl calls. */ + for (i = 0; i < 100; i++) + prctl(0, 0, 0, 0, 0); + + exit(0); + } + + threads = perf_thread_map__new_dummy(); + __T("failed to create threads", threads); + + cpus = perf_cpu_map__dummy_new(); + __T("failed to create cpus", cpus); + + perf_thread_map__set_pid(threads, 0, pid); + + evlist = perf_evlist__new(); + __T("failed to create evlist", evlist); + + evsel = perf_evsel__new(&attr); + __T("failed to create evsel1", evsel); + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, cpus, threads); + + err = perf_evlist__open(evlist); + __T("failed to open evlist", err == 0); + + err = perf_evlist__mmap(evlist, 4); + __T("failed to mmap evlist", err == 0); + + perf_evlist__enable(evlist); + + /* kick the child and wait for it to finish */ + write(go_pipe[1], &bf, 1); + waitpid(pid, NULL, 0); + + /* + * There's no need to call perf_evlist__disable, + * monitored process is dead now. + */ + + perf_evlist__for_each_mmap(evlist, map, false) { + if (perf_mmap__read_init(map) < 0) + continue; + + while ((event = perf_mmap__read_event(map)) != NULL) { + count++; + perf_mmap__consume(map); + } + + perf_mmap__read_done(map); + } + + /* calls perf_evlist__munmap/perf_evlist__close */ + perf_evlist__delete(evlist); + + perf_thread_map__put(threads); + perf_cpu_map__put(cpus); + + /* + * The generated prctl calls should match the + * number of events in the buffer. + */ + __T("failed count", count == 100); + + return 0; +} + +static int test_mmap_cpus(void) +{ + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_mmap *map; + struct perf_cpu_map *cpus; + struct perf_event_attr attr = { + .type = PERF_TYPE_TRACEPOINT, + .sample_period = 1, + .wakeup_watermark = 1, + .disabled = 1, + }; + cpu_set_t saved_mask; + char path[PATH_MAX]; + int id, err, cpu, tmp; + union perf_event *event; + int count = 0; + + snprintf(path, PATH_MAX, "%s/kernel/debug/tracing/events/syscalls/sys_enter_prctl/id", + sysfs__mountpoint()); + + if (filename__read_int(path, &id)) { + fprintf(stderr, "error: failed to get tracepoint id: %s\n", path); + return -1; + } + + attr.config = id; + + cpus = perf_cpu_map__new(NULL); + __T("failed to create cpus", cpus); + + evlist = perf_evlist__new(); + __T("failed to create evlist", evlist); + + evsel = perf_evsel__new(&attr); + __T("failed to create evsel1", evsel); + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, cpus, NULL); + + err = perf_evlist__open(evlist); + __T("failed to open evlist", err == 0); + + err = perf_evlist__mmap(evlist, 4); + __T("failed to mmap evlist", err == 0); + + perf_evlist__enable(evlist); + + err = sched_getaffinity(0, sizeof(saved_mask), &saved_mask); + __T("sched_getaffinity failed", err == 0); + + perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { + cpu_set_t mask; + + CPU_ZERO(&mask); + CPU_SET(cpu, &mask); + + err = sched_setaffinity(0, sizeof(mask), &mask); + __T("sched_setaffinity failed", err == 0); + + prctl(0, 0, 0, 0, 0); + } + + err = sched_setaffinity(0, sizeof(saved_mask), &saved_mask); + __T("sched_setaffinity failed", err == 0); + + perf_evlist__disable(evlist); + + perf_evlist__for_each_mmap(evlist, map, false) { + if (perf_mmap__read_init(map) < 0) + continue; + + while ((event = perf_mmap__read_event(map)) != NULL) { + count++; + perf_mmap__consume(map); + } + + perf_mmap__read_done(map); + } + + /* calls perf_evlist__munmap/perf_evlist__close */ + perf_evlist__delete(evlist); + + /* + * The generated prctl events should match the + * number of cpus or be bigger (we are system-wide). + */ + __T("failed count", count >= perf_cpu_map__nr(cpus)); + + perf_cpu_map__put(cpus); + + return 0; +} + +int main(int argc, char **argv) +{ + __T_START; + + libperf_init(libperf_print); + + test_stat_cpu(); + test_stat_thread(); + test_stat_thread_enable(); + test_mmap_thread(); + test_mmap_cpus(); + + __T_END; + return 0; +} diff --git a/tools/lib/perf/tests/test-evsel.c b/tools/lib/perf/tests/test-evsel.c new file mode 100644 index 000000000000..135722ac965b --- /dev/null +++ b/tools/lib/perf/tests/test-evsel.c @@ -0,0 +1,135 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <stdarg.h> +#include <stdio.h> +#include <linux/perf_event.h> +#include <perf/cpumap.h> +#include <perf/threadmap.h> +#include <perf/evsel.h> +#include <internal/tests.h> + +static int libperf_print(enum libperf_print_level level, + const char *fmt, va_list ap) +{ + return vfprintf(stderr, fmt, ap); +} + +static int test_stat_cpu(void) +{ + struct perf_cpu_map *cpus; + struct perf_evsel *evsel; + struct perf_event_attr attr = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_CPU_CLOCK, + }; + int err, cpu, tmp; + + cpus = perf_cpu_map__new(NULL); + __T("failed to create cpus", cpus); + + evsel = perf_evsel__new(&attr); + __T("failed to create evsel", evsel); + + err = perf_evsel__open(evsel, cpus, NULL); + __T("failed to open evsel", err == 0); + + perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { + struct perf_counts_values counts = { .val = 0 }; + + perf_evsel__read(evsel, cpu, 0, &counts); + __T("failed to read value for evsel", counts.val != 0); + } + + perf_evsel__close(evsel); + perf_evsel__delete(evsel); + + perf_cpu_map__put(cpus); + return 0; +} + +static int test_stat_thread(void) +{ + struct perf_counts_values counts = { .val = 0 }; + struct perf_thread_map *threads; + struct perf_evsel *evsel; + struct perf_event_attr attr = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_TASK_CLOCK, + }; + int err; + + threads = perf_thread_map__new_dummy(); + __T("failed to create threads", threads); + + perf_thread_map__set_pid(threads, 0, 0); + + evsel = perf_evsel__new(&attr); + __T("failed to create evsel", evsel); + + err = perf_evsel__open(evsel, NULL, threads); + __T("failed to open evsel", err == 0); + + perf_evsel__read(evsel, 0, 0, &counts); + __T("failed to read value for evsel", counts.val != 0); + + perf_evsel__close(evsel); + perf_evsel__delete(evsel); + + perf_thread_map__put(threads); + return 0; +} + +static int test_stat_thread_enable(void) +{ + struct perf_counts_values counts = { .val = 0 }; + struct perf_thread_map *threads; + struct perf_evsel *evsel; + struct perf_event_attr attr = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_TASK_CLOCK, + .disabled = 1, + }; + int err; + + threads = perf_thread_map__new_dummy(); + __T("failed to create threads", threads); + + perf_thread_map__set_pid(threads, 0, 0); + + evsel = perf_evsel__new(&attr); + __T("failed to create evsel", evsel); + + err = perf_evsel__open(evsel, NULL, threads); + __T("failed to open evsel", err == 0); + + perf_evsel__read(evsel, 0, 0, &counts); + __T("failed to read value for evsel", counts.val == 0); + + err = perf_evsel__enable(evsel); + __T("failed to enable evsel", err == 0); + + perf_evsel__read(evsel, 0, 0, &counts); + __T("failed to read value for evsel", counts.val != 0); + + err = perf_evsel__disable(evsel); + __T("failed to enable evsel", err == 0); + + perf_evsel__close(evsel); + perf_evsel__delete(evsel); + + perf_thread_map__put(threads); + return 0; +} + +int main(int argc, char **argv) +{ + __T_START; + + libperf_init(libperf_print); + + test_stat_cpu(); + test_stat_thread(); + test_stat_thread_enable(); + + __T_END; + return 0; +} diff --git a/tools/lib/perf/tests/test-threadmap.c b/tools/lib/perf/tests/test-threadmap.c new file mode 100644 index 000000000000..7dc4d6fbedde --- /dev/null +++ b/tools/lib/perf/tests/test-threadmap.c @@ -0,0 +1,31 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <stdarg.h> +#include <stdio.h> +#include <perf/threadmap.h> +#include <internal/tests.h> + +static int libperf_print(enum libperf_print_level level, + const char *fmt, va_list ap) +{ + return vfprintf(stderr, fmt, ap); +} + +int main(int argc, char **argv) +{ + struct perf_thread_map *threads; + + __T_START; + + libperf_init(libperf_print); + + threads = perf_thread_map__new_dummy(); + if (!threads) + return -1; + + perf_thread_map__get(threads); + perf_thread_map__put(threads); + perf_thread_map__put(threads); + + __T_END; + return 0; +} diff --git a/tools/lib/perf/threadmap.c b/tools/lib/perf/threadmap.c new file mode 100644 index 000000000000..e92c368b0a6c --- /dev/null +++ b/tools/lib/perf/threadmap.c @@ -0,0 +1,91 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <perf/threadmap.h> +#include <stdlib.h> +#include <linux/refcount.h> +#include <internal/threadmap.h> +#include <string.h> +#include <asm/bug.h> +#include <stdio.h> + +static void perf_thread_map__reset(struct perf_thread_map *map, int start, int nr) +{ + size_t size = (nr - start) * sizeof(map->map[0]); + + memset(&map->map[start], 0, size); + map->err_thread = -1; +} + +struct perf_thread_map *perf_thread_map__realloc(struct perf_thread_map *map, int nr) +{ + size_t size = sizeof(*map) + sizeof(map->map[0]) * nr; + int start = map ? map->nr : 0; + + map = realloc(map, size); + /* + * We only realloc to add more items, let's reset new items. + */ + if (map) + perf_thread_map__reset(map, start, nr); + + return map; +} + +#define thread_map__alloc(__nr) perf_thread_map__realloc(NULL, __nr) + +void perf_thread_map__set_pid(struct perf_thread_map *map, int thread, pid_t pid) +{ + map->map[thread].pid = pid; +} + +char *perf_thread_map__comm(struct perf_thread_map *map, int thread) +{ + return map->map[thread].comm; +} + +struct perf_thread_map *perf_thread_map__new_dummy(void) +{ + struct perf_thread_map *threads = thread_map__alloc(1); + + if (threads != NULL) { + perf_thread_map__set_pid(threads, 0, -1); + threads->nr = 1; + refcount_set(&threads->refcnt, 1); + } + return threads; +} + +static void perf_thread_map__delete(struct perf_thread_map *threads) +{ + if (threads) { + int i; + + WARN_ONCE(refcount_read(&threads->refcnt) != 0, + "thread map refcnt unbalanced\n"); + for (i = 0; i < threads->nr; i++) + free(perf_thread_map__comm(threads, i)); + free(threads); + } +} + +struct perf_thread_map *perf_thread_map__get(struct perf_thread_map *map) +{ + if (map) + refcount_inc(&map->refcnt); + return map; +} + +void perf_thread_map__put(struct perf_thread_map *map) +{ + if (map && refcount_dec_and_test(&map->refcnt)) + perf_thread_map__delete(map); +} + +int perf_thread_map__nr(struct perf_thread_map *threads) +{ + return threads ? threads->nr : 1; +} + +pid_t perf_thread_map__pid(struct perf_thread_map *map, int thread) +{ + return map->map[thread].pid; +} diff --git a/tools/lib/perf/xyarray.c b/tools/lib/perf/xyarray.c new file mode 100644 index 000000000000..dcd901d154bb --- /dev/null +++ b/tools/lib/perf/xyarray.c @@ -0,0 +1,33 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <internal/xyarray.h> +#include <linux/zalloc.h> +#include <stdlib.h> +#include <string.h> + +struct xyarray *xyarray__new(int xlen, int ylen, size_t entry_size) +{ + size_t row_size = ylen * entry_size; + struct xyarray *xy = zalloc(sizeof(*xy) + xlen * row_size); + + if (xy != NULL) { + xy->entry_size = entry_size; + xy->row_size = row_size; + xy->entries = xlen * ylen; + xy->max_x = xlen; + xy->max_y = ylen; + } + + return xy; +} + +void xyarray__reset(struct xyarray *xy) +{ + size_t n = xy->entries * xy->entry_size; + + memset(xy->contents, 0, n); +} + +void xyarray__delete(struct xyarray *xy) +{ + free(xy); +} diff --git a/tools/perf/MANIFEST b/tools/perf/MANIFEST index 70f1ff4e2eb4..ae2aab5b1be0 100644 --- a/tools/perf/MANIFEST +++ b/tools/perf/MANIFEST @@ -7,6 +7,7 @@ tools/lib/traceevent tools/lib/api tools/lib/bpf tools/lib/subcmd +tools/lib/perf tools/lib/argv_split.c tools/lib/ctype.c tools/lib/hweight.c diff --git a/tools/perf/Makefile.config b/tools/perf/Makefile.config index c90f4146e5a2..80e55e796be9 100644 --- a/tools/perf/Makefile.config +++ b/tools/perf/Makefile.config @@ -286,7 +286,7 @@ ifeq ($(DEBUG),0) endif endif -INC_FLAGS += -I$(src-perf)/lib/include +INC_FLAGS += -I$(srctree)/tools/lib/perf/include INC_FLAGS += -I$(src-perf)/util/include INC_FLAGS += -I$(src-perf)/arch/$(SRCARCH)/include INC_FLAGS += -I$(srctree)/tools/include/ diff --git a/tools/perf/Makefile.perf b/tools/perf/Makefile.perf index eae5d5e95952..3eda9d4b88e7 100644 --- a/tools/perf/Makefile.perf +++ b/tools/perf/Makefile.perf @@ -230,7 +230,7 @@ LIB_DIR = $(srctree)/tools/lib/api/ TRACE_EVENT_DIR = $(srctree)/tools/lib/traceevent/ BPF_DIR = $(srctree)/tools/lib/bpf/ SUBCMD_DIR = $(srctree)/tools/lib/subcmd/ -LIBPERF_DIR = $(srctree)/tools/perf/lib/ +LIBPERF_DIR = $(srctree)/tools/lib/perf/ # Set FEATURE_TESTS to 'all' so all possible feature checkers are executed. # Without this setting the output feature dump file misses some features, for diff --git a/tools/perf/lib/Build b/tools/perf/lib/Build deleted file mode 100644 index 2ef9a4ec6d99..000000000000 --- a/tools/perf/lib/Build +++ /dev/null @@ -1,13 +0,0 @@ -libperf-y += core.o -libperf-y += cpumap.o -libperf-y += threadmap.o -libperf-y += evsel.o -libperf-y += evlist.o -libperf-y += mmap.o -libperf-y += zalloc.o -libperf-y += xyarray.o -libperf-y += lib.o - -$(OUTPUT)zalloc.o: ../../lib/zalloc.c FORCE - $(call rule_mkdir) - $(call if_changed_dep,cc_o_c) diff --git a/tools/perf/lib/Documentation/Makefile b/tools/perf/lib/Documentation/Makefile deleted file mode 100644 index 586425a88795..000000000000 --- a/tools/perf/lib/Documentation/Makefile +++ /dev/null @@ -1,7 +0,0 @@ -all: - rst2man man/libperf.rst > man/libperf.7 - rst2pdf tutorial/tutorial.rst - -clean: - rm -f man/libperf.7 - rm -f tutorial/tutorial.pdf diff --git a/tools/perf/lib/Documentation/man/libperf.rst b/tools/perf/lib/Documentation/man/libperf.rst deleted file mode 100644 index 09a270fccb9c..000000000000 --- a/tools/perf/lib/Documentation/man/libperf.rst +++ /dev/null @@ -1,100 +0,0 @@ -.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -libperf - -The libperf library provides an API to access the linux kernel perf -events subsystem. It provides the following high level objects: - - - struct perf_cpu_map - - struct perf_thread_map - - struct perf_evlist - - struct perf_evsel - -reference -========= -Function reference by header files: - -perf/core.h ------------ -.. code-block:: c - - typedef int (\*libperf_print_fn_t)(enum libperf_print_level level, - const char \*, va_list ap); - - void libperf_set_print(libperf_print_fn_t fn); - -perf/cpumap.h -------------- -.. code-block:: c - - struct perf_cpu_map \*perf_cpu_map__dummy_new(void); - struct perf_cpu_map \*perf_cpu_map__new(const char \*cpu_list); - struct perf_cpu_map \*perf_cpu_map__read(FILE \*file); - struct perf_cpu_map \*perf_cpu_map__get(struct perf_cpu_map \*map); - void perf_cpu_map__put(struct perf_cpu_map \*map); - int perf_cpu_map__cpu(const struct perf_cpu_map \*cpus, int idx); - int perf_cpu_map__nr(const struct perf_cpu_map \*cpus); - perf_cpu_map__for_each_cpu(cpu, idx, cpus) - -perf/threadmap.h ----------------- -.. code-block:: c - - struct perf_thread_map \*perf_thread_map__new_dummy(void); - void perf_thread_map__set_pid(struct perf_thread_map \*map, int thread, pid_t pid); - char \*perf_thread_map__comm(struct perf_thread_map \*map, int thread); - struct perf_thread_map \*perf_thread_map__get(struct perf_thread_map \*map); - void perf_thread_map__put(struct perf_thread_map \*map); - -perf/evlist.h -------------- -.. code-block:: - - void perf_evlist__init(struct perf_evlist \*evlist); - void perf_evlist__add(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - void perf_evlist__remove(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - struct perf_evlist \*perf_evlist__new(void); - void perf_evlist__delete(struct perf_evlist \*evlist); - struct perf_evsel\* perf_evlist__next(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - int perf_evlist__open(struct perf_evlist \*evlist); - void perf_evlist__close(struct perf_evlist \*evlist); - void perf_evlist__enable(struct perf_evlist \*evlist); - void perf_evlist__disable(struct perf_evlist \*evlist); - perf_evlist__for_each_evsel(evlist, pos) - void perf_evlist__set_maps(struct perf_evlist \*evlist, - struct perf_cpu_map \*cpus, - struct perf_thread_map \*threads); - -perf/evsel.h ------------- -.. code-block:: c - - struct perf_counts_values { - union { - struct { - uint64_t val; - uint64_t ena; - uint64_t run; - }; - uint64_t values[3]; - }; - }; - - void perf_evsel__init(struct perf_evsel \*evsel, - struct perf_event_attr \*attr); - struct perf_evsel \*perf_evsel__new(struct perf_event_attr \*attr); - void perf_evsel__delete(struct perf_evsel \*evsel); - int perf_evsel__open(struct perf_evsel \*evsel, struct perf_cpu_map \*cpus, - struct perf_thread_map \*threads); - void perf_evsel__close(struct perf_evsel \*evsel); - int perf_evsel__read(struct perf_evsel \*evsel, int cpu, int thread, - struct perf_counts_values \*count); - int perf_evsel__enable(struct perf_evsel \*evsel); - int perf_evsel__disable(struct perf_evsel \*evsel); - int perf_evsel__apply_filter(struct perf_evsel \*evsel, const char \*filter); - struct perf_cpu_map \*perf_evsel__cpus(struct perf_evsel \*evsel); - struct perf_thread_map \*perf_evsel__threads(struct perf_evsel \*evsel); - struct perf_event_attr \*perf_evsel__attr(struct perf_evsel \*evsel); diff --git a/tools/perf/lib/Documentation/tutorial/tutorial.rst b/tools/perf/lib/Documentation/tutorial/tutorial.rst deleted file mode 100644 index 7be7bc27b385..000000000000 --- a/tools/perf/lib/Documentation/tutorial/tutorial.rst +++ /dev/null @@ -1,123 +0,0 @@ -.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -libperf tutorial -================ - -Compile and install libperf from kernel sources -=============================================== -.. code-block:: bash - - git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git - cd linux/tools/perf/lib - make - sudo make install prefix=/usr - -Libperf object -============== -The libperf library provides several high level objects: - -struct perf_cpu_map - Provides a cpu list abstraction. - -struct perf_thread_map - Provides a thread list abstraction. - -struct perf_evsel - Provides an abstraction for single a perf event. - -struct perf_evlist - Gathers several struct perf_evsel object and performs functions on all of them. - -The exported API binds these objects together, -for full reference see the libperf.7 man page. - -Examples -======== -Examples aim to explain libperf functionality on simple use cases. -They are based in on a checked out linux kernel git tree: - -.. code-block:: bash - - $ cd tools/perf/lib/Documentation/tutorial/ - $ ls -d ex-* - ex-1-compile ex-2-evsel-stat ex-3-evlist-stat - -ex-1-compile example -==================== -This example shows the basic usage of *struct perf_cpu_map*, -how to create it and display its cpus: - -.. code-block:: bash - - $ cd ex-1-compile/ - $ make - gcc -o test test.c -lperf - $ ./test - 0 1 2 3 4 5 6 7 - - -The full code listing is here: - -.. code-block:: c - - 1 #include <perf/cpumap.h> - 2 - 3 int main(int argc, char **Argv) - 4 { - 5 struct perf_cpu_map *cpus; - 6 int cpu, tmp; - 7 - 8 cpus = perf_cpu_map__new(NULL); - 9 - 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) - 11 fprintf(stdout, "%d ", cpu); - 12 - 13 fprintf(stdout, "\n"); - 14 - 15 perf_cpu_map__put(cpus); - 16 return 0; - 17 } - - -First you need to include the proper header to have *struct perf_cpumap* -declaration and functions: - -.. code-block:: c - - 1 #include <perf/cpumap.h> - - -The *struct perf_cpumap* object is created by *perf_cpu_map__new* call. -The *NULL* argument asks it to populate the object with the current online CPUs list: - -.. code-block:: c - - 8 cpus = perf_cpu_map__new(NULL); - -This is paired with a *perf_cpu_map__put*, that drops its reference at the end, possibly deleting it. - -.. code-block:: c - - 15 perf_cpu_map__put(cpus); - -The iteration through the *struct perf_cpumap* CPUs is done using the *perf_cpu_map__for_each_cpu* -macro which requires 3 arguments: - -- cpu - the cpu numer -- tmp - iteration helper variable -- cpus - the *struct perf_cpumap* object - -.. code-block:: c - - 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) - 11 fprintf(stdout, "%d ", cpu); - -ex-2-evsel-stat example -======================= - -TBD - -ex-3-evlist-stat example -======================== - -TBD diff --git a/tools/perf/lib/Makefile b/tools/perf/lib/Makefile deleted file mode 100644 index 0f233638ef1f..000000000000 --- a/tools/perf/lib/Makefile +++ /dev/null @@ -1,188 +0,0 @@ -# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) -# Most of this file is copied from tools/lib/bpf/Makefile - -LIBPERF_VERSION = 0 -LIBPERF_PATCHLEVEL = 0 -LIBPERF_EXTRAVERSION = 1 - -MAKEFLAGS += --no-print-directory - -ifeq ($(srctree),) -srctree := $(patsubst %/,%,$(dir $(CURDIR))) -srctree := $(patsubst %/,%,$(dir $(srctree))) -srctree := $(patsubst %/,%,$(dir $(srctree))) -#$(info Determined 'srctree' to be $(srctree)) -endif - -INSTALL = install - -# Use DESTDIR for installing into a different root directory. -# This is useful for building a package. The program will be -# installed in this directory as if it was the root directory. -# Then the build tool can move it later. -DESTDIR ?= -DESTDIR_SQ = '$(subst ','\'',$(DESTDIR))' - -include $(srctree)/tools/scripts/Makefile.include -include $(srctree)/tools/scripts/Makefile.arch - -ifeq ($(LP64), 1) - libdir_relative = lib64 -else - libdir_relative = lib -endif - -prefix ?= -libdir = $(prefix)/$(libdir_relative) - -# Shell quotes -libdir_SQ = $(subst ','\'',$(libdir)) -libdir_relative_SQ = $(subst ','\'',$(libdir_relative)) - -ifeq ("$(origin V)", "command line") - VERBOSE = $(V) -endif -ifndef VERBOSE - VERBOSE = 0 -endif - -ifeq ($(VERBOSE),1) - Q = -else - Q = @ -endif - -# Set compile option CFLAGS -ifdef EXTRA_CFLAGS - CFLAGS := $(EXTRA_CFLAGS) -else - CFLAGS := -g -Wall -endif - -INCLUDES = \ --I$(srctree)/tools/perf/lib/include \ --I$(srctree)/tools/lib/ \ --I$(srctree)/tools/include \ --I$(srctree)/tools/arch/$(SRCARCH)/include/ \ --I$(srctree)/tools/arch/$(SRCARCH)/include/uapi \ --I$(srctree)/tools/include/uapi - -# Append required CFLAGS -override CFLAGS += $(EXTRA_WARNINGS) -override CFLAGS += -Werror -Wall -override CFLAGS += -fPIC -override CFLAGS += $(INCLUDES) -override CFLAGS += -fvisibility=hidden - -all: - -export srctree OUTPUT CC LD CFLAGS V -export DESTDIR DESTDIR_SQ - -include $(srctree)/tools/build/Makefile.include - -VERSION_SCRIPT := libperf.map - -PATCHLEVEL = $(LIBPERF_PATCHLEVEL) -EXTRAVERSION = $(LIBPERF_EXTRAVERSION) -VERSION = $(LIBPERF_VERSION).$(LIBPERF_PATCHLEVEL).$(LIBPERF_EXTRAVERSION) - -LIBPERF_SO := $(OUTPUT)libperf.so.$(VERSION) -LIBPERF_A := $(OUTPUT)libperf.a -LIBPERF_IN := $(OUTPUT)libperf-in.o -LIBPERF_PC := $(OUTPUT)libperf.pc - -LIBPERF_ALL := $(LIBPERF_A) $(OUTPUT)libperf.so* - -LIB_DIR := $(srctree)/tools/lib/api/ - -ifneq ($(OUTPUT),) -ifneq ($(subdir),) - API_PATH=$(OUTPUT)/../lib/api/ -else - API_PATH=$(OUTPUT) -endif -else - API_PATH=$(LIB_DIR) -endif - -LIBAPI = $(API_PATH)libapi.a -export LIBAPI - -$(LIBAPI): FORCE - $(Q)$(MAKE) -C $(LIB_DIR) O=$(OUTPUT) $(OUTPUT)libapi.a - -$(LIBAPI)-clean: - $(call QUIET_CLEAN, libapi) - $(Q)$(MAKE) -C $(LIB_DIR) O=$(OUTPUT) clean >/dev/null - -$(LIBPERF_IN): FORCE - $(Q)$(MAKE) $(build)=libperf - -$(LIBPERF_A): $(LIBPERF_IN) - $(QUIET_AR)$(RM) $@ && $(AR) rcs $@ $(LIBPERF_IN) - -$(LIBPERF_SO): $(LIBPERF_IN) $(LIBAPI) - $(QUIET_LINK)$(CC) --shared -Wl,-soname,libperf.so \ - -Wl,--version-script=$(VERSION_SCRIPT) $^ -o $@ - @ln -sf $(@F) $(OUTPUT)libperf.so - @ln -sf $(@F) $(OUTPUT)libperf.so.$(LIBPERF_VERSION) - - -libs: $(LIBPERF_A) $(LIBPERF_SO) $(LIBPERF_PC) - -all: fixdep - $(Q)$(MAKE) libs - -clean: $(LIBAPI)-clean - $(call QUIET_CLEAN, libperf) $(RM) $(LIBPERF_A) \ - *.o *~ *.a *.so *.so.$(VERSION) *.so.$(LIBPERF_VERSION) .*.d .*.cmd LIBPERF-CFLAGS $(LIBPERF_PC) - $(Q)$(MAKE) -C tests clean - -tests: libs - $(Q)$(MAKE) -C tests - $(Q)$(MAKE) -C tests run - -$(LIBPERF_PC): - $(QUIET_GEN)sed -e "s|@PREFIX@|$(prefix)|" \ - -e "s|@LIBDIR@|$(libdir_SQ)|" \ - -e "s|@VERSION@|$(VERSION)|" \ - < libperf.pc.template > $@ - -define do_install_mkdir - if [ ! -d '$(DESTDIR_SQ)$1' ]; then \ - $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$1'; \ - fi -endef - -define do_install - if [ ! -d '$(DESTDIR_SQ)$2' ]; then \ - $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$2'; \ - fi; \ - $(INSTALL) $1 $(if $3,-m $3,) '$(DESTDIR_SQ)$2' -endef - -install_lib: libs - $(call QUIET_INSTALL, $(LIBPERF_ALL)) \ - $(call do_install_mkdir,$(libdir_SQ)); \ - cp -fpR $(LIBPERF_ALL) $(DESTDIR)$(libdir_SQ) - -install_headers: - $(call QUIET_INSTALL, headers) \ - $(call do_install,include/perf/core.h,$(prefix)/include/perf,644); \ - $(call do_install,include/perf/cpumap.h,$(prefix)/include/perf,644); \ - $(call do_install,include/perf/threadmap.h,$(prefix)/include/perf,644); \ - $(call do_install,include/perf/evlist.h,$(prefix)/include/perf,644); \ - $(call do_install,include/perf/evsel.h,$(prefix)/include/perf,644); \ - $(call do_install,include/perf/event.h,$(prefix)/include/perf,644); \ - $(call do_install,include/perf/mmap.h,$(prefix)/include/perf,644); - -install_pkgconfig: $(LIBPERF_PC) - $(call QUIET_INSTALL, $(LIBPERF_PC)) \ - $(call do_install,$(LIBPERF_PC),$(libdir_SQ)/pkgconfig,644) - -install: install_lib install_headers install_pkgconfig - -FORCE: - -.PHONY: all install clean tests FORCE diff --git a/tools/perf/lib/core.c b/tools/perf/lib/core.c deleted file mode 100644 index 58fc894b76c5..000000000000 --- a/tools/perf/lib/core.c +++ /dev/null @@ -1,38 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only - -#define __printf(a, b) __attribute__((format(printf, a, b))) - -#include <stdio.h> -#include <stdarg.h> -#include <unistd.h> -#include <linux/compiler.h> -#include <perf/core.h> -#include <internal/lib.h> -#include "internal.h" - -static int __base_pr(enum libperf_print_level level __maybe_unused, const char *format, - va_list args) -{ - return vfprintf(stderr, format, args); -} - -static libperf_print_fn_t __libperf_pr = __base_pr; - -__printf(2, 3) -void libperf_print(enum libperf_print_level level, const char *format, ...) -{ - va_list args; - - if (!__libperf_pr) - return; - - va_start(args, format); - __libperf_pr(level, format, args); - va_end(args); -} - -void libperf_init(libperf_print_fn_t fn) -{ - page_size = sysconf(_SC_PAGE_SIZE); - __libperf_pr = fn; -} diff --git a/tools/perf/lib/cpumap.c b/tools/perf/lib/cpumap.c deleted file mode 100644 index f93f4e703e4c..000000000000 --- a/tools/perf/lib/cpumap.c +++ /dev/null @@ -1,345 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -#include <perf/cpumap.h> -#include <stdlib.h> -#include <linux/refcount.h> -#include <internal/cpumap.h> -#include <asm/bug.h> -#include <stdio.h> -#include <string.h> -#include <unistd.h> -#include <ctype.h> -#include <limits.h> - -struct perf_cpu_map *perf_cpu_map__dummy_new(void) -{ - struct perf_cpu_map *cpus = malloc(sizeof(*cpus) + sizeof(int)); - - if (cpus != NULL) { - cpus->nr = 1; - cpus->map[0] = -1; - refcount_set(&cpus->refcnt, 1); - } - - return cpus; -} - -static void cpu_map__delete(struct perf_cpu_map *map) -{ - if (map) { - WARN_ONCE(refcount_read(&map->refcnt) != 0, - "cpu_map refcnt unbalanced\n"); - free(map); - } -} - -struct perf_cpu_map *perf_cpu_map__get(struct perf_cpu_map *map) -{ - if (map) - refcount_inc(&map->refcnt); - return map; -} - -void perf_cpu_map__put(struct perf_cpu_map *map) -{ - if (map && refcount_dec_and_test(&map->refcnt)) - cpu_map__delete(map); -} - -static struct perf_cpu_map *cpu_map__default_new(void) -{ - struct perf_cpu_map *cpus; - int nr_cpus; - - nr_cpus = sysconf(_SC_NPROCESSORS_ONLN); - if (nr_cpus < 0) - return NULL; - - cpus = malloc(sizeof(*cpus) + nr_cpus * sizeof(int)); - if (cpus != NULL) { - int i; - - for (i = 0; i < nr_cpus; ++i) - cpus->map[i] = i; - - cpus->nr = nr_cpus; - refcount_set(&cpus->refcnt, 1); - } - - return cpus; -} - -static int cmp_int(const void *a, const void *b) -{ - return *(const int *)a - *(const int*)b; -} - -static struct perf_cpu_map *cpu_map__trim_new(int nr_cpus, int *tmp_cpus) -{ - size_t payload_size = nr_cpus * sizeof(int); - struct perf_cpu_map *cpus = malloc(sizeof(*cpus) + payload_size); - int i, j; - - if (cpus != NULL) { - memcpy(cpus->map, tmp_cpus, payload_size); - qsort(cpus->map, nr_cpus, sizeof(int), cmp_int); - /* Remove dups */ - j = 0; - for (i = 0; i < nr_cpus; i++) { - if (i == 0 || cpus->map[i] != cpus->map[i - 1]) - cpus->map[j++] = cpus->map[i]; - } - cpus->nr = j; - assert(j <= nr_cpus); - refcount_set(&cpus->refcnt, 1); - } - - return cpus; -} - -struct perf_cpu_map *perf_cpu_map__read(FILE *file) -{ - struct perf_cpu_map *cpus = NULL; - int nr_cpus = 0; - int *tmp_cpus = NULL, *tmp; - int max_entries = 0; - int n, cpu, prev; - char sep; - - sep = 0; - prev = -1; - for (;;) { - n = fscanf(file, "%u%c", &cpu, &sep); - if (n <= 0) - break; - if (prev >= 0) { - int new_max = nr_cpus + cpu - prev - 1; - - WARN_ONCE(new_max >= MAX_NR_CPUS, "Perf can support %d CPUs. " - "Consider raising MAX_NR_CPUS\n", MAX_NR_CPUS); - - if (new_max >= max_entries) { - max_entries = new_max + MAX_NR_CPUS / 2; - tmp = realloc(tmp_cpus, max_entries * sizeof(int)); - if (tmp == NULL) - goto out_free_tmp; - tmp_cpus = tmp; - } - - while (++prev < cpu) - tmp_cpus[nr_cpus++] = prev; - } - if (nr_cpus == max_entries) { - max_entries += MAX_NR_CPUS; - tmp = realloc(tmp_cpus, max_entries * sizeof(int)); - if (tmp == NULL) - goto out_free_tmp; - tmp_cpus = tmp; - } - - tmp_cpus[nr_cpus++] = cpu; - if (n == 2 && sep == '-') - prev = cpu; - else - prev = -1; - if (n == 1 || sep == '\n') - break; - } - - if (nr_cpus > 0) - cpus = cpu_map__trim_new(nr_cpus, tmp_cpus); - else - cpus = cpu_map__default_new(); -out_free_tmp: - free(tmp_cpus); - return cpus; -} - -static struct perf_cpu_map *cpu_map__read_all_cpu_map(void) -{ - struct perf_cpu_map *cpus = NULL; - FILE *onlnf; - - onlnf = fopen("/sys/devices/system/cpu/online", "r"); - if (!onlnf) - return cpu_map__default_new(); - - cpus = perf_cpu_map__read(onlnf); - fclose(onlnf); - return cpus; -} - -struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list) -{ - struct perf_cpu_map *cpus = NULL; - unsigned long start_cpu, end_cpu = 0; - char *p = NULL; - int i, nr_cpus = 0; - int *tmp_cpus = NULL, *tmp; - int max_entries = 0; - - if (!cpu_list) - return cpu_map__read_all_cpu_map(); - - /* - * must handle the case of empty cpumap to cover - * TOPOLOGY header for NUMA nodes with no CPU - * ( e.g., because of CPU hotplug) - */ - if (!isdigit(*cpu_list) && *cpu_list != '\0') - goto out; - - while (isdigit(*cpu_list)) { - p = NULL; - start_cpu = strtoul(cpu_list, &p, 0); - if (start_cpu >= INT_MAX - || (*p != '\0' && *p != ',' && *p != '-')) - goto invalid; - - if (*p == '-') { - cpu_list = ++p; - p = NULL; - end_cpu = strtoul(cpu_list, &p, 0); - - if (end_cpu >= INT_MAX || (*p != '\0' && *p != ',')) - goto invalid; - - if (end_cpu < start_cpu) - goto invalid; - } else { - end_cpu = start_cpu; - } - - WARN_ONCE(end_cpu >= MAX_NR_CPUS, "Perf can support %d CPUs. " - "Consider raising MAX_NR_CPUS\n", MAX_NR_CPUS); - - for (; start_cpu <= end_cpu; start_cpu++) { - /* check for duplicates */ - for (i = 0; i < nr_cpus; i++) - if (tmp_cpus[i] == (int)start_cpu) - goto invalid; - - if (nr_cpus == max_entries) { - max_entries += MAX_NR_CPUS; - tmp = realloc(tmp_cpus, max_entries * sizeof(int)); - if (tmp == NULL) - goto invalid; - tmp_cpus = tmp; - } - tmp_cpus[nr_cpus++] = (int)start_cpu; - } - if (*p) - ++p; - - cpu_list = p; - } - - if (nr_cpus > 0) - cpus = cpu_map__trim_new(nr_cpus, tmp_cpus); - else if (*cpu_list != '\0') - cpus = cpu_map__default_new(); - else - cpus = perf_cpu_map__dummy_new(); -invalid: - free(tmp_cpus); -out: - return cpus; -} - -int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx) -{ - if (idx < cpus->nr) - return cpus->map[idx]; - - return -1; -} - -int perf_cpu_map__nr(const struct perf_cpu_map *cpus) -{ - return cpus ? cpus->nr : 1; -} - -bool perf_cpu_map__empty(const struct perf_cpu_map *map) -{ - return map ? map->map[0] == -1 : true; -} - -int perf_cpu_map__idx(struct perf_cpu_map *cpus, int cpu) -{ - int i; - - for (i = 0; i < cpus->nr; ++i) { - if (cpus->map[i] == cpu) - return i; - } - - return -1; -} - -int perf_cpu_map__max(struct perf_cpu_map *map) -{ - int i, max = -1; - - for (i = 0; i < map->nr; i++) { - if (map->map[i] > max) - max = map->map[i]; - } - - return max; -} - -/* - * Merge two cpumaps - * - * orig either gets freed and replaced with a new map, or reused - * with no reference count change (similar to "realloc") - * other has its reference count increased. - */ - -struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig, - struct perf_cpu_map *other) -{ - int *tmp_cpus; - int tmp_len; - int i, j, k; - struct perf_cpu_map *merged; - - if (!orig && !other) - return NULL; - if (!orig) { - perf_cpu_map__get(other); - return other; - } - if (!other) - return orig; - if (orig->nr == other->nr && - !memcmp(orig->map, other->map, orig->nr * sizeof(int))) - return orig; - - tmp_len = orig->nr + other->nr; - tmp_cpus = malloc(tmp_len * sizeof(int)); - if (!tmp_cpus) - return NULL; - - /* Standard merge algorithm from wikipedia */ - i = j = k = 0; - while (i < orig->nr && j < other->nr) { - if (orig->map[i] <= other->map[j]) { - if (orig->map[i] == other->map[j]) - j++; - tmp_cpus[k++] = orig->map[i++]; - } else - tmp_cpus[k++] = other->map[j++]; - } - - while (i < orig->nr) - tmp_cpus[k++] = orig->map[i++]; - - while (j < other->nr) - tmp_cpus[k++] = other->map[j++]; - assert(k <= tmp_len); - - merged = cpu_map__trim_new(k, tmp_cpus); - free(tmp_cpus); - perf_cpu_map__put(orig); - return merged; -} diff --git a/tools/perf/lib/evlist.c b/tools/perf/lib/evlist.c deleted file mode 100644 index ae9e65aa2491..000000000000 --- a/tools/perf/lib/evlist.c +++ /dev/null @@ -1,641 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <perf/evlist.h> -#include <perf/evsel.h> -#include <linux/bitops.h> -#include <linux/list.h> -#include <linux/hash.h> -#include <sys/ioctl.h> -#include <internal/evlist.h> -#include <internal/evsel.h> -#include <internal/xyarray.h> -#include <internal/mmap.h> -#include <internal/cpumap.h> -#include <internal/threadmap.h> -#include <internal/xyarray.h> -#include <internal/lib.h> -#include <linux/zalloc.h> -#include <sys/ioctl.h> -#include <stdlib.h> -#include <errno.h> -#include <unistd.h> -#include <fcntl.h> -#include <signal.h> -#include <poll.h> -#include <sys/mman.h> -#include <perf/cpumap.h> -#include <perf/threadmap.h> -#include <api/fd/array.h> - -void perf_evlist__init(struct perf_evlist *evlist) -{ - int i; - - for (i = 0; i < PERF_EVLIST__HLIST_SIZE; ++i) - INIT_HLIST_HEAD(&evlist->heads[i]); - INIT_LIST_HEAD(&evlist->entries); - evlist->nr_entries = 0; - fdarray__init(&evlist->pollfd, 64); -} - -static void __perf_evlist__propagate_maps(struct perf_evlist *evlist, - struct perf_evsel *evsel) -{ - /* - * We already have cpus for evsel (via PMU sysfs) so - * keep it, if there's no target cpu list defined. - */ - if (!evsel->own_cpus || evlist->has_user_cpus) { - perf_cpu_map__put(evsel->cpus); - evsel->cpus = perf_cpu_map__get(evlist->cpus); - } else if (evsel->cpus != evsel->own_cpus) { - perf_cpu_map__put(evsel->cpus); - evsel->cpus = perf_cpu_map__get(evsel->own_cpus); - } - - perf_thread_map__put(evsel->threads); - evsel->threads = perf_thread_map__get(evlist->threads); - evlist->all_cpus = perf_cpu_map__merge(evlist->all_cpus, evsel->cpus); -} - -static void perf_evlist__propagate_maps(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel; - - perf_evlist__for_each_evsel(evlist, evsel) - __perf_evlist__propagate_maps(evlist, evsel); -} - -void perf_evlist__add(struct perf_evlist *evlist, - struct perf_evsel *evsel) -{ - list_add_tail(&evsel->node, &evlist->entries); - evlist->nr_entries += 1; - __perf_evlist__propagate_maps(evlist, evsel); -} - -void perf_evlist__remove(struct perf_evlist *evlist, - struct perf_evsel *evsel) -{ - list_del_init(&evsel->node); - evlist->nr_entries -= 1; -} - -struct perf_evlist *perf_evlist__new(void) -{ - struct perf_evlist *evlist = zalloc(sizeof(*evlist)); - - if (evlist != NULL) - perf_evlist__init(evlist); - - return evlist; -} - -struct perf_evsel * -perf_evlist__next(struct perf_evlist *evlist, struct perf_evsel *prev) -{ - struct perf_evsel *next; - - if (!prev) { - next = list_first_entry(&evlist->entries, - struct perf_evsel, - node); - } else { - next = list_next_entry(prev, node); - } - - /* Empty list is noticed here so don't need checking on entry. */ - if (&next->node == &evlist->entries) - return NULL; - - return next; -} - -static void perf_evlist__purge(struct perf_evlist *evlist) -{ - struct perf_evsel *pos, *n; - - perf_evlist__for_each_entry_safe(evlist, n, pos) { - list_del_init(&pos->node); - perf_evsel__delete(pos); - } - - evlist->nr_entries = 0; -} - -void perf_evlist__exit(struct perf_evlist *evlist) -{ - perf_cpu_map__put(evlist->cpus); - perf_thread_map__put(evlist->threads); - evlist->cpus = NULL; - evlist->threads = NULL; - fdarray__exit(&evlist->pollfd); -} - -void perf_evlist__delete(struct perf_evlist *evlist) -{ - if (evlist == NULL) - return; - - perf_evlist__munmap(evlist); - perf_evlist__close(evlist); - perf_evlist__purge(evlist); - perf_evlist__exit(evlist); - free(evlist); -} - -void perf_evlist__set_maps(struct perf_evlist *evlist, - struct perf_cpu_map *cpus, - struct perf_thread_map *threads) -{ - /* - * Allow for the possibility that one or another of the maps isn't being - * changed i.e. don't put it. Note we are assuming the maps that are - * being applied are brand new and evlist is taking ownership of the - * original reference count of 1. If that is not the case it is up to - * the caller to increase the reference count. - */ - if (cpus != evlist->cpus) { - perf_cpu_map__put(evlist->cpus); - evlist->cpus = perf_cpu_map__get(cpus); - } - - if (threads != evlist->threads) { - perf_thread_map__put(evlist->threads); - evlist->threads = perf_thread_map__get(threads); - } - - perf_evlist__propagate_maps(evlist); -} - -int perf_evlist__open(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel; - int err; - - perf_evlist__for_each_entry(evlist, evsel) { - err = perf_evsel__open(evsel, evsel->cpus, evsel->threads); - if (err < 0) - goto out_err; - } - - return 0; - -out_err: - perf_evlist__close(evlist); - return err; -} - -void perf_evlist__close(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel; - - perf_evlist__for_each_entry_reverse(evlist, evsel) - perf_evsel__close(evsel); -} - -void perf_evlist__enable(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel; - - perf_evlist__for_each_entry(evlist, evsel) - perf_evsel__enable(evsel); -} - -void perf_evlist__disable(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel; - - perf_evlist__for_each_entry(evlist, evsel) - perf_evsel__disable(evsel); -} - -u64 perf_evlist__read_format(struct perf_evlist *evlist) -{ - struct perf_evsel *first = perf_evlist__first(evlist); - - return first->attr.read_format; -} - -#define SID(e, x, y) xyarray__entry(e->sample_id, x, y) - -static void perf_evlist__id_hash(struct perf_evlist *evlist, - struct perf_evsel *evsel, - int cpu, int thread, u64 id) -{ - int hash; - struct perf_sample_id *sid = SID(evsel, cpu, thread); - - sid->id = id; - sid->evsel = evsel; - hash = hash_64(sid->id, PERF_EVLIST__HLIST_BITS); - hlist_add_head(&sid->node, &evlist->heads[hash]); -} - -void perf_evlist__id_add(struct perf_evlist *evlist, - struct perf_evsel *evsel, - int cpu, int thread, u64 id) -{ - perf_evlist__id_hash(evlist, evsel, cpu, thread, id); - evsel->id[evsel->ids++] = id; -} - -int perf_evlist__id_add_fd(struct perf_evlist *evlist, - struct perf_evsel *evsel, - int cpu, int thread, int fd) -{ - u64 read_data[4] = { 0, }; - int id_idx = 1; /* The first entry is the counter value */ - u64 id; - int ret; - - ret = ioctl(fd, PERF_EVENT_IOC_ID, &id); - if (!ret) - goto add; - - if (errno != ENOTTY) - return -1; - - /* Legacy way to get event id.. All hail to old kernels! */ - - /* - * This way does not work with group format read, so bail - * out in that case. - */ - if (perf_evlist__read_format(evlist) & PERF_FORMAT_GROUP) - return -1; - - if (!(evsel->attr.read_format & PERF_FORMAT_ID) || - read(fd, &read_data, sizeof(read_data)) == -1) - return -1; - - if (evsel->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) - ++id_idx; - if (evsel->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) - ++id_idx; - - id = read_data[id_idx]; - -add: - perf_evlist__id_add(evlist, evsel, cpu, thread, id); - return 0; -} - -int perf_evlist__alloc_pollfd(struct perf_evlist *evlist) -{ - int nr_cpus = perf_cpu_map__nr(evlist->cpus); - int nr_threads = perf_thread_map__nr(evlist->threads); - int nfds = 0; - struct perf_evsel *evsel; - - perf_evlist__for_each_entry(evlist, evsel) { - if (evsel->system_wide) - nfds += nr_cpus; - else - nfds += nr_cpus * nr_threads; - } - - if (fdarray__available_entries(&evlist->pollfd) < nfds && - fdarray__grow(&evlist->pollfd, nfds) < 0) - return -ENOMEM; - - return 0; -} - -int perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd, - void *ptr, short revent) -{ - int pos = fdarray__add(&evlist->pollfd, fd, revent | POLLERR | POLLHUP); - - if (pos >= 0) { - evlist->pollfd.priv[pos].ptr = ptr; - fcntl(fd, F_SETFL, O_NONBLOCK); - } - - return pos; -} - -static void perf_evlist__munmap_filtered(struct fdarray *fda, int fd, - void *arg __maybe_unused) -{ - struct perf_mmap *map = fda->priv[fd].ptr; - - if (map) - perf_mmap__put(map); -} - -int perf_evlist__filter_pollfd(struct perf_evlist *evlist, short revents_and_mask) -{ - return fdarray__filter(&evlist->pollfd, revents_and_mask, - perf_evlist__munmap_filtered, NULL); -} - -int perf_evlist__poll(struct perf_evlist *evlist, int timeout) -{ - return fdarray__poll(&evlist->pollfd, timeout); -} - -static struct perf_mmap* perf_evlist__alloc_mmap(struct perf_evlist *evlist, bool overwrite) -{ - int i; - struct perf_mmap *map; - - map = zalloc(evlist->nr_mmaps * sizeof(struct perf_mmap)); - if (!map) - return NULL; - - for (i = 0; i < evlist->nr_mmaps; i++) { - struct perf_mmap *prev = i ? &map[i - 1] : NULL; - - /* - * When the perf_mmap() call is made we grab one refcount, plus - * one extra to let perf_mmap__consume() get the last - * events after all real references (perf_mmap__get()) are - * dropped. - * - * Each PERF_EVENT_IOC_SET_OUTPUT points to this mmap and - * thus does perf_mmap__get() on it. - */ - perf_mmap__init(&map[i], prev, overwrite, NULL); - } - - return map; -} - -static void perf_evlist__set_sid_idx(struct perf_evlist *evlist, - struct perf_evsel *evsel, int idx, int cpu, - int thread) -{ - struct perf_sample_id *sid = SID(evsel, cpu, thread); - - sid->idx = idx; - if (evlist->cpus && cpu >= 0) - sid->cpu = evlist->cpus->map[cpu]; - else - sid->cpu = -1; - if (!evsel->system_wide && evlist->threads && thread >= 0) - sid->tid = perf_thread_map__pid(evlist->threads, thread); - else - sid->tid = -1; -} - -static struct perf_mmap* -perf_evlist__mmap_cb_get(struct perf_evlist *evlist, bool overwrite, int idx) -{ - struct perf_mmap *maps; - - maps = overwrite ? evlist->mmap_ovw : evlist->mmap; - - if (!maps) { - maps = perf_evlist__alloc_mmap(evlist, overwrite); - if (!maps) - return NULL; - - if (overwrite) - evlist->mmap_ovw = maps; - else - evlist->mmap = maps; - } - - return &maps[idx]; -} - -#define FD(e, x, y) (*(int *) xyarray__entry(e->fd, x, y)) - -static int -perf_evlist__mmap_cb_mmap(struct perf_mmap *map, struct perf_mmap_param *mp, - int output, int cpu) -{ - return perf_mmap__mmap(map, mp, output, cpu); -} - -static void perf_evlist__set_mmap_first(struct perf_evlist *evlist, struct perf_mmap *map, - bool overwrite) -{ - if (overwrite) - evlist->mmap_ovw_first = map; - else - evlist->mmap_first = map; -} - -static int -mmap_per_evsel(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, - int idx, struct perf_mmap_param *mp, int cpu_idx, - int thread, int *_output, int *_output_overwrite) -{ - int evlist_cpu = perf_cpu_map__cpu(evlist->cpus, cpu_idx); - struct perf_evsel *evsel; - int revent; - - perf_evlist__for_each_entry(evlist, evsel) { - bool overwrite = evsel->attr.write_backward; - struct perf_mmap *map; - int *output, fd, cpu; - - if (evsel->system_wide && thread) - continue; - - cpu = perf_cpu_map__idx(evsel->cpus, evlist_cpu); - if (cpu == -1) - continue; - - map = ops->get(evlist, overwrite, idx); - if (map == NULL) - return -ENOMEM; - - if (overwrite) { - mp->prot = PROT_READ; - output = _output_overwrite; - } else { - mp->prot = PROT_READ | PROT_WRITE; - output = _output; - } - - fd = FD(evsel, cpu, thread); - - if (*output == -1) { - *output = fd; - - /* - * The last one will be done at perf_mmap__consume(), so that we - * make sure we don't prevent tools from consuming every last event in - * the ring buffer. - * - * I.e. we can get the POLLHUP meaning that the fd doesn't exist - * anymore, but the last events for it are still in the ring buffer, - * waiting to be consumed. - * - * Tools can chose to ignore this at their own discretion, but the - * evlist layer can't just drop it when filtering events in - * perf_evlist__filter_pollfd(). - */ - refcount_set(&map->refcnt, 2); - - if (ops->mmap(map, mp, *output, evlist_cpu) < 0) - return -1; - - if (!idx) - perf_evlist__set_mmap_first(evlist, map, overwrite); - } else { - if (ioctl(fd, PERF_EVENT_IOC_SET_OUTPUT, *output) != 0) - return -1; - - perf_mmap__get(map); - } - - revent = !overwrite ? POLLIN : 0; - - if (!evsel->system_wide && - perf_evlist__add_pollfd(evlist, fd, map, revent) < 0) { - perf_mmap__put(map); - return -1; - } - - if (evsel->attr.read_format & PERF_FORMAT_ID) { - if (perf_evlist__id_add_fd(evlist, evsel, cpu, thread, - fd) < 0) - return -1; - perf_evlist__set_sid_idx(evlist, evsel, idx, cpu, - thread); - } - } - - return 0; -} - -static int -mmap_per_thread(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, - struct perf_mmap_param *mp) -{ - int thread; - int nr_threads = perf_thread_map__nr(evlist->threads); - - for (thread = 0; thread < nr_threads; thread++) { - int output = -1; - int output_overwrite = -1; - - if (ops->idx) - ops->idx(evlist, mp, thread, false); - - if (mmap_per_evsel(evlist, ops, thread, mp, 0, thread, - &output, &output_overwrite)) - goto out_unmap; - } - - return 0; - -out_unmap: - perf_evlist__munmap(evlist); - return -1; -} - -static int -mmap_per_cpu(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, - struct perf_mmap_param *mp) -{ - int nr_threads = perf_thread_map__nr(evlist->threads); - int nr_cpus = perf_cpu_map__nr(evlist->cpus); - int cpu, thread; - - for (cpu = 0; cpu < nr_cpus; cpu++) { - int output = -1; - int output_overwrite = -1; - - if (ops->idx) - ops->idx(evlist, mp, cpu, true); - - for (thread = 0; thread < nr_threads; thread++) { - if (mmap_per_evsel(evlist, ops, cpu, mp, cpu, - thread, &output, &output_overwrite)) - goto out_unmap; - } - } - - return 0; - -out_unmap: - perf_evlist__munmap(evlist); - return -1; -} - -static int perf_evlist__nr_mmaps(struct perf_evlist *evlist) -{ - int nr_mmaps; - - nr_mmaps = perf_cpu_map__nr(evlist->cpus); - if (perf_cpu_map__empty(evlist->cpus)) - nr_mmaps = perf_thread_map__nr(evlist->threads); - - return nr_mmaps; -} - -int perf_evlist__mmap_ops(struct perf_evlist *evlist, - struct perf_evlist_mmap_ops *ops, - struct perf_mmap_param *mp) -{ - struct perf_evsel *evsel; - const struct perf_cpu_map *cpus = evlist->cpus; - const struct perf_thread_map *threads = evlist->threads; - - if (!ops || !ops->get || !ops->mmap) - return -EINVAL; - - mp->mask = evlist->mmap_len - page_size - 1; - - evlist->nr_mmaps = perf_evlist__nr_mmaps(evlist); - - perf_evlist__for_each_entry(evlist, evsel) { - if ((evsel->attr.read_format & PERF_FORMAT_ID) && - evsel->sample_id == NULL && - perf_evsel__alloc_id(evsel, perf_cpu_map__nr(cpus), threads->nr) < 0) - return -ENOMEM; - } - - if (evlist->pollfd.entries == NULL && perf_evlist__alloc_pollfd(evlist) < 0) - return -ENOMEM; - - if (perf_cpu_map__empty(cpus)) - return mmap_per_thread(evlist, ops, mp); - - return mmap_per_cpu(evlist, ops, mp); -} - -int perf_evlist__mmap(struct perf_evlist *evlist, int pages) -{ - struct perf_mmap_param mp; - struct perf_evlist_mmap_ops ops = { - .get = perf_evlist__mmap_cb_get, - .mmap = perf_evlist__mmap_cb_mmap, - }; - - evlist->mmap_len = (pages + 1) * page_size; - - return perf_evlist__mmap_ops(evlist, &ops, &mp); -} - -void perf_evlist__munmap(struct perf_evlist *evlist) -{ - int i; - - if (evlist->mmap) { - for (i = 0; i < evlist->nr_mmaps; i++) - perf_mmap__munmap(&evlist->mmap[i]); - } - - if (evlist->mmap_ovw) { - for (i = 0; i < evlist->nr_mmaps; i++) - perf_mmap__munmap(&evlist->mmap_ovw[i]); - } - - zfree(&evlist->mmap); - zfree(&evlist->mmap_ovw); -} - -struct perf_mmap* -perf_evlist__next_mmap(struct perf_evlist *evlist, struct perf_mmap *map, - bool overwrite) -{ - if (map) - return map->next; - - return overwrite ? evlist->mmap_ovw_first : evlist->mmap_first; -} diff --git a/tools/perf/lib/evsel.c b/tools/perf/lib/evsel.c deleted file mode 100644 index 4dc06289f4c7..000000000000 --- a/tools/perf/lib/evsel.c +++ /dev/null @@ -1,301 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <errno.h> -#include <unistd.h> -#include <sys/syscall.h> -#include <perf/evsel.h> -#include <perf/cpumap.h> -#include <perf/threadmap.h> -#include <linux/list.h> -#include <internal/evsel.h> -#include <linux/zalloc.h> -#include <stdlib.h> -#include <internal/xyarray.h> -#include <internal/cpumap.h> -#include <internal/threadmap.h> -#include <internal/lib.h> -#include <linux/string.h> -#include <sys/ioctl.h> - -void perf_evsel__init(struct perf_evsel *evsel, struct perf_event_attr *attr) -{ - INIT_LIST_HEAD(&evsel->node); - evsel->attr = *attr; -} - -struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr) -{ - struct perf_evsel *evsel = zalloc(sizeof(*evsel)); - - if (evsel != NULL) - perf_evsel__init(evsel, attr); - - return evsel; -} - -void perf_evsel__delete(struct perf_evsel *evsel) -{ - free(evsel); -} - -#define FD(e, x, y) (*(int *) xyarray__entry(e->fd, x, y)) - -int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads) -{ - evsel->fd = xyarray__new(ncpus, nthreads, sizeof(int)); - - if (evsel->fd) { - int cpu, thread; - for (cpu = 0; cpu < ncpus; cpu++) { - for (thread = 0; thread < nthreads; thread++) { - FD(evsel, cpu, thread) = -1; - } - } - } - - return evsel->fd != NULL ? 0 : -ENOMEM; -} - -static int -sys_perf_event_open(struct perf_event_attr *attr, - pid_t pid, int cpu, int group_fd, - unsigned long flags) -{ - return syscall(__NR_perf_event_open, attr, pid, cpu, group_fd, flags); -} - -int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus, - struct perf_thread_map *threads) -{ - int cpu, thread, err = 0; - - if (cpus == NULL) { - static struct perf_cpu_map *empty_cpu_map; - - if (empty_cpu_map == NULL) { - empty_cpu_map = perf_cpu_map__dummy_new(); - if (empty_cpu_map == NULL) - return -ENOMEM; - } - - cpus = empty_cpu_map; - } - - if (threads == NULL) { - static struct perf_thread_map *empty_thread_map; - - if (empty_thread_map == NULL) { - empty_thread_map = perf_thread_map__new_dummy(); - if (empty_thread_map == NULL) - return -ENOMEM; - } - - threads = empty_thread_map; - } - - if (evsel->fd == NULL && - perf_evsel__alloc_fd(evsel, cpus->nr, threads->nr) < 0) - return -ENOMEM; - - for (cpu = 0; cpu < cpus->nr; cpu++) { - for (thread = 0; thread < threads->nr; thread++) { - int fd; - - fd = sys_perf_event_open(&evsel->attr, - threads->map[thread].pid, - cpus->map[cpu], -1, 0); - - if (fd < 0) - return -errno; - - FD(evsel, cpu, thread) = fd; - } - } - - return err; -} - -static void perf_evsel__close_fd_cpu(struct perf_evsel *evsel, int cpu) -{ - int thread; - - for (thread = 0; thread < xyarray__max_y(evsel->fd); ++thread) { - if (FD(evsel, cpu, thread) >= 0) - close(FD(evsel, cpu, thread)); - FD(evsel, cpu, thread) = -1; - } -} - -void perf_evsel__close_fd(struct perf_evsel *evsel) -{ - int cpu; - - for (cpu = 0; cpu < xyarray__max_x(evsel->fd); cpu++) - perf_evsel__close_fd_cpu(evsel, cpu); -} - -void perf_evsel__free_fd(struct perf_evsel *evsel) -{ - xyarray__delete(evsel->fd); - evsel->fd = NULL; -} - -void perf_evsel__close(struct perf_evsel *evsel) -{ - if (evsel->fd == NULL) - return; - - perf_evsel__close_fd(evsel); - perf_evsel__free_fd(evsel); -} - -void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu) -{ - if (evsel->fd == NULL) - return; - - perf_evsel__close_fd_cpu(evsel, cpu); -} - -int perf_evsel__read_size(struct perf_evsel *evsel) -{ - u64 read_format = evsel->attr.read_format; - int entry = sizeof(u64); /* value */ - int size = 0; - int nr = 1; - - if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) - size += sizeof(u64); - - if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) - size += sizeof(u64); - - if (read_format & PERF_FORMAT_ID) - entry += sizeof(u64); - - if (read_format & PERF_FORMAT_GROUP) { - nr = evsel->nr_members; - size += sizeof(u64); - } - - size += entry * nr; - return size; -} - -int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, - struct perf_counts_values *count) -{ - size_t size = perf_evsel__read_size(evsel); - - memset(count, 0, sizeof(*count)); - - if (FD(evsel, cpu, thread) < 0) - return -EINVAL; - - if (readn(FD(evsel, cpu, thread), count->values, size) <= 0) - return -errno; - - return 0; -} - -static int perf_evsel__run_ioctl(struct perf_evsel *evsel, - int ioc, void *arg, - int cpu) -{ - int thread; - - for (thread = 0; thread < xyarray__max_y(evsel->fd); thread++) { - int fd = FD(evsel, cpu, thread), - err = ioctl(fd, ioc, arg); - - if (err) - return err; - } - - return 0; -} - -int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu) -{ - return perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_ENABLE, NULL, cpu); -} - -int perf_evsel__enable(struct perf_evsel *evsel) -{ - int i; - int err = 0; - - for (i = 0; i < xyarray__max_x(evsel->fd) && !err; i++) - err = perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_ENABLE, NULL, i); - return err; -} - -int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu) -{ - return perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_DISABLE, NULL, cpu); -} - -int perf_evsel__disable(struct perf_evsel *evsel) -{ - int i; - int err = 0; - - for (i = 0; i < xyarray__max_x(evsel->fd) && !err; i++) - err = perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_DISABLE, NULL, i); - return err; -} - -int perf_evsel__apply_filter(struct perf_evsel *evsel, const char *filter) -{ - int err = 0, i; - - for (i = 0; i < evsel->cpus->nr && !err; i++) - err = perf_evsel__run_ioctl(evsel, - PERF_EVENT_IOC_SET_FILTER, - (void *)filter, i); - return err; -} - -struct perf_cpu_map *perf_evsel__cpus(struct perf_evsel *evsel) -{ - return evsel->cpus; -} - -struct perf_thread_map *perf_evsel__threads(struct perf_evsel *evsel) -{ - return evsel->threads; -} - -struct perf_event_attr *perf_evsel__attr(struct perf_evsel *evsel) -{ - return &evsel->attr; -} - -int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads) -{ - if (ncpus == 0 || nthreads == 0) - return 0; - - if (evsel->system_wide) - nthreads = 1; - - evsel->sample_id = xyarray__new(ncpus, nthreads, sizeof(struct perf_sample_id)); - if (evsel->sample_id == NULL) - return -ENOMEM; - - evsel->id = zalloc(ncpus * nthreads * sizeof(u64)); - if (evsel->id == NULL) { - xyarray__delete(evsel->sample_id); - evsel->sample_id = NULL; - return -ENOMEM; - } - - return 0; -} - -void perf_evsel__free_id(struct perf_evsel *evsel) -{ - xyarray__delete(evsel->sample_id); - evsel->sample_id = NULL; - zfree(&evsel->id); - evsel->ids = 0; -} diff --git a/tools/perf/lib/include/internal/cpumap.h b/tools/perf/lib/include/internal/cpumap.h deleted file mode 100644 index 840d4032587b..000000000000 --- a/tools/perf/lib/include/internal/cpumap.h +++ /dev/null @@ -1,19 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_CPUMAP_H -#define __LIBPERF_INTERNAL_CPUMAP_H - -#include <linux/refcount.h> - -struct perf_cpu_map { - refcount_t refcnt; - int nr; - int map[]; -}; - -#ifndef MAX_NR_CPUS -#define MAX_NR_CPUS 2048 -#endif - -int perf_cpu_map__idx(struct perf_cpu_map *cpus, int cpu); - -#endif /* __LIBPERF_INTERNAL_CPUMAP_H */ diff --git a/tools/perf/lib/include/internal/evlist.h b/tools/perf/lib/include/internal/evlist.h deleted file mode 100644 index 74dc8c3f0b66..000000000000 --- a/tools/perf/lib/include/internal/evlist.h +++ /dev/null @@ -1,127 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_EVLIST_H -#define __LIBPERF_INTERNAL_EVLIST_H - -#include <linux/list.h> -#include <api/fd/array.h> -#include <internal/evsel.h> - -#define PERF_EVLIST__HLIST_BITS 8 -#define PERF_EVLIST__HLIST_SIZE (1 << PERF_EVLIST__HLIST_BITS) - -struct perf_cpu_map; -struct perf_thread_map; -struct perf_mmap_param; - -struct perf_evlist { - struct list_head entries; - int nr_entries; - bool has_user_cpus; - struct perf_cpu_map *cpus; - struct perf_cpu_map *all_cpus; - struct perf_thread_map *threads; - int nr_mmaps; - size_t mmap_len; - struct fdarray pollfd; - struct hlist_head heads[PERF_EVLIST__HLIST_SIZE]; - struct perf_mmap *mmap; - struct perf_mmap *mmap_ovw; - struct perf_mmap *mmap_first; - struct perf_mmap *mmap_ovw_first; -}; - -typedef void -(*perf_evlist_mmap__cb_idx_t)(struct perf_evlist*, struct perf_mmap_param*, int, bool); -typedef struct perf_mmap* -(*perf_evlist_mmap__cb_get_t)(struct perf_evlist*, bool, int); -typedef int -(*perf_evlist_mmap__cb_mmap_t)(struct perf_mmap*, struct perf_mmap_param*, int, int); - -struct perf_evlist_mmap_ops { - perf_evlist_mmap__cb_idx_t idx; - perf_evlist_mmap__cb_get_t get; - perf_evlist_mmap__cb_mmap_t mmap; -}; - -int perf_evlist__alloc_pollfd(struct perf_evlist *evlist); -int perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd, - void *ptr, short revent); - -int perf_evlist__mmap_ops(struct perf_evlist *evlist, - struct perf_evlist_mmap_ops *ops, - struct perf_mmap_param *mp); - -void perf_evlist__init(struct perf_evlist *evlist); -void perf_evlist__exit(struct perf_evlist *evlist); - -/** - * __perf_evlist__for_each_entry - iterate thru all the evsels - * @list: list_head instance to iterate - * @evsel: struct perf_evsel iterator - */ -#define __perf_evlist__for_each_entry(list, evsel) \ - list_for_each_entry(evsel, list, node) - -/** - * evlist__for_each_entry - iterate thru all the evsels - * @evlist: perf_evlist instance to iterate - * @evsel: struct perf_evsel iterator - */ -#define perf_evlist__for_each_entry(evlist, evsel) \ - __perf_evlist__for_each_entry(&(evlist)->entries, evsel) - -/** - * __perf_evlist__for_each_entry_reverse - iterate thru all the evsels in reverse order - * @list: list_head instance to iterate - * @evsel: struct evsel iterator - */ -#define __perf_evlist__for_each_entry_reverse(list, evsel) \ - list_for_each_entry_reverse(evsel, list, node) - -/** - * perf_evlist__for_each_entry_reverse - iterate thru all the evsels in reverse order - * @evlist: evlist instance to iterate - * @evsel: struct evsel iterator - */ -#define perf_evlist__for_each_entry_reverse(evlist, evsel) \ - __perf_evlist__for_each_entry_reverse(&(evlist)->entries, evsel) - -/** - * __perf_evlist__for_each_entry_safe - safely iterate thru all the evsels - * @list: list_head instance to iterate - * @tmp: struct evsel temp iterator - * @evsel: struct evsel iterator - */ -#define __perf_evlist__for_each_entry_safe(list, tmp, evsel) \ - list_for_each_entry_safe(evsel, tmp, list, node) - -/** - * perf_evlist__for_each_entry_safe - safely iterate thru all the evsels - * @evlist: evlist instance to iterate - * @evsel: struct evsel iterator - * @tmp: struct evsel temp iterator - */ -#define perf_evlist__for_each_entry_safe(evlist, tmp, evsel) \ - __perf_evlist__for_each_entry_safe(&(evlist)->entries, tmp, evsel) - -static inline struct perf_evsel *perf_evlist__first(struct perf_evlist *evlist) -{ - return list_entry(evlist->entries.next, struct perf_evsel, node); -} - -static inline struct perf_evsel *perf_evlist__last(struct perf_evlist *evlist) -{ - return list_entry(evlist->entries.prev, struct perf_evsel, node); -} - -u64 perf_evlist__read_format(struct perf_evlist *evlist); - -void perf_evlist__id_add(struct perf_evlist *evlist, - struct perf_evsel *evsel, - int cpu, int thread, u64 id); - -int perf_evlist__id_add_fd(struct perf_evlist *evlist, - struct perf_evsel *evsel, - int cpu, int thread, int fd); - -#endif /* __LIBPERF_INTERNAL_EVLIST_H */ diff --git a/tools/perf/lib/include/internal/evsel.h b/tools/perf/lib/include/internal/evsel.h deleted file mode 100644 index 1ffd083b235e..000000000000 --- a/tools/perf/lib/include/internal/evsel.h +++ /dev/null @@ -1,63 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_EVSEL_H -#define __LIBPERF_INTERNAL_EVSEL_H - -#include <linux/types.h> -#include <linux/perf_event.h> -#include <stdbool.h> -#include <sys/types.h> - -struct perf_cpu_map; -struct perf_thread_map; -struct xyarray; - -/* - * Per fd, to map back from PERF_SAMPLE_ID to evsel, only used when there are - * more than one entry in the evlist. - */ -struct perf_sample_id { - struct hlist_node node; - u64 id; - struct perf_evsel *evsel; - /* - * 'idx' will be used for AUX area sampling. A sample will have AUX area - * data that will be queued for decoding, where there are separate - * queues for each CPU (per-cpu tracing) or task (per-thread tracing). - * The sample ID can be used to lookup 'idx' which is effectively the - * queue number. - */ - int idx; - int cpu; - pid_t tid; - - /* Holds total ID period value for PERF_SAMPLE_READ processing. */ - u64 period; -}; - -struct perf_evsel { - struct list_head node; - struct perf_event_attr attr; - struct perf_cpu_map *cpus; - struct perf_cpu_map *own_cpus; - struct perf_thread_map *threads; - struct xyarray *fd; - struct xyarray *sample_id; - u64 *id; - u32 ids; - - /* parse modifier helper */ - int nr_members; - bool system_wide; -}; - -void perf_evsel__init(struct perf_evsel *evsel, struct perf_event_attr *attr); -int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads); -void perf_evsel__close_fd(struct perf_evsel *evsel); -void perf_evsel__free_fd(struct perf_evsel *evsel); -int perf_evsel__read_size(struct perf_evsel *evsel); -int perf_evsel__apply_filter(struct perf_evsel *evsel, const char *filter); - -int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads); -void perf_evsel__free_id(struct perf_evsel *evsel); - -#endif /* __LIBPERF_INTERNAL_EVSEL_H */ diff --git a/tools/perf/lib/include/internal/lib.h b/tools/perf/lib/include/internal/lib.h deleted file mode 100644 index 5175d491b2d4..000000000000 --- a/tools/perf/lib/include/internal/lib.h +++ /dev/null @@ -1,12 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_LIB_H -#define __LIBPERF_INTERNAL_LIB_H - -#include <sys/types.h> - -extern unsigned int page_size; - -ssize_t readn(int fd, void *buf, size_t n); -ssize_t writen(int fd, const void *buf, size_t n); - -#endif /* __LIBPERF_INTERNAL_CPUMAP_H */ diff --git a/tools/perf/lib/include/internal/mmap.h b/tools/perf/lib/include/internal/mmap.h deleted file mode 100644 index be7556e0a2b2..000000000000 --- a/tools/perf/lib/include/internal/mmap.h +++ /dev/null @@ -1,55 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_MMAP_H -#define __LIBPERF_INTERNAL_MMAP_H - -#include <linux/compiler.h> -#include <linux/refcount.h> -#include <linux/types.h> -#include <stdbool.h> - -/* perf sample has 16 bits size limit */ -#define PERF_SAMPLE_MAX_SIZE (1 << 16) - -struct perf_mmap; - -typedef void (*libperf_unmap_cb_t)(struct perf_mmap *map); - -/** - * struct perf_mmap - perf's ring buffer mmap details - * - * @refcnt - e.g. code using PERF_EVENT_IOC_SET_OUTPUT to share this - */ -struct perf_mmap { - void *base; - int mask; - int fd; - int cpu; - refcount_t refcnt; - u64 prev; - u64 start; - u64 end; - bool overwrite; - u64 flush; - libperf_unmap_cb_t unmap_cb; - char event_copy[PERF_SAMPLE_MAX_SIZE] __aligned(8); - struct perf_mmap *next; -}; - -struct perf_mmap_param { - int prot; - int mask; -}; - -size_t perf_mmap__mmap_len(struct perf_mmap *map); - -void perf_mmap__init(struct perf_mmap *map, struct perf_mmap *prev, - bool overwrite, libperf_unmap_cb_t unmap_cb); -int perf_mmap__mmap(struct perf_mmap *map, struct perf_mmap_param *mp, - int fd, int cpu); -void perf_mmap__munmap(struct perf_mmap *map); -void perf_mmap__get(struct perf_mmap *map); -void perf_mmap__put(struct perf_mmap *map); - -u64 perf_mmap__read_head(struct perf_mmap *map); - -#endif /* __LIBPERF_INTERNAL_MMAP_H */ diff --git a/tools/perf/lib/include/internal/tests.h b/tools/perf/lib/include/internal/tests.h deleted file mode 100644 index 2093e8868a67..000000000000 --- a/tools/perf/lib/include/internal/tests.h +++ /dev/null @@ -1,33 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_TESTS_H -#define __LIBPERF_INTERNAL_TESTS_H - -#include <stdio.h> - -int tests_failed; - -#define __T_START \ -do { \ - fprintf(stdout, "- running %s...", __FILE__); \ - fflush(NULL); \ - tests_failed = 0; \ -} while (0) - -#define __T_END \ -do { \ - if (tests_failed) \ - fprintf(stdout, " FAILED (%d)\n", tests_failed); \ - else \ - fprintf(stdout, "OK\n"); \ -} while (0) - -#define __T(text, cond) \ -do { \ - if (!(cond)) { \ - fprintf(stderr, "FAILED %s:%d %s\n", __FILE__, __LINE__, text); \ - tests_failed++; \ - return -1; \ - } \ -} while (0) - -#endif /* __LIBPERF_INTERNAL_TESTS_H */ diff --git a/tools/perf/lib/include/internal/threadmap.h b/tools/perf/lib/include/internal/threadmap.h deleted file mode 100644 index df748baf9eda..000000000000 --- a/tools/perf/lib/include/internal/threadmap.h +++ /dev/null @@ -1,23 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_THREADMAP_H -#define __LIBPERF_INTERNAL_THREADMAP_H - -#include <linux/refcount.h> -#include <sys/types.h> -#include <unistd.h> - -struct thread_map_data { - pid_t pid; - char *comm; -}; - -struct perf_thread_map { - refcount_t refcnt; - int nr; - int err_thread; - struct thread_map_data map[]; -}; - -struct perf_thread_map *perf_thread_map__realloc(struct perf_thread_map *map, int nr); - -#endif /* __LIBPERF_INTERNAL_THREADMAP_H */ diff --git a/tools/perf/lib/include/internal/xyarray.h b/tools/perf/lib/include/internal/xyarray.h deleted file mode 100644 index 51e35d6c8ec4..000000000000 --- a/tools/perf/lib/include/internal/xyarray.h +++ /dev/null @@ -1,36 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_XYARRAY_H -#define __LIBPERF_INTERNAL_XYARRAY_H - -#include <linux/compiler.h> -#include <sys/types.h> - -struct xyarray { - size_t row_size; - size_t entry_size; - size_t entries; - size_t max_x; - size_t max_y; - char contents[] __aligned(8); -}; - -struct xyarray *xyarray__new(int xlen, int ylen, size_t entry_size); -void xyarray__delete(struct xyarray *xy); -void xyarray__reset(struct xyarray *xy); - -static inline void *xyarray__entry(struct xyarray *xy, int x, int y) -{ - return &xy->contents[x * xy->row_size + y * xy->entry_size]; -} - -static inline int xyarray__max_y(struct xyarray *xy) -{ - return xy->max_y; -} - -static inline int xyarray__max_x(struct xyarray *xy) -{ - return xy->max_x; -} - -#endif /* __LIBPERF_INTERNAL_XYARRAY_H */ diff --git a/tools/perf/lib/include/perf/core.h b/tools/perf/lib/include/perf/core.h deleted file mode 100644 index a3f6d68edad7..000000000000 --- a/tools/perf/lib/include/perf/core.h +++ /dev/null @@ -1,25 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_CORE_H -#define __LIBPERF_CORE_H - -#include <stdarg.h> - -#ifndef LIBPERF_API -#define LIBPERF_API __attribute__((visibility("default"))) -#endif - -enum libperf_print_level { - LIBPERF_ERR, - LIBPERF_WARN, - LIBPERF_INFO, - LIBPERF_DEBUG, - LIBPERF_DEBUG2, - LIBPERF_DEBUG3, -}; - -typedef int (*libperf_print_fn_t)(enum libperf_print_level level, - const char *, va_list ap); - -LIBPERF_API void libperf_init(libperf_print_fn_t fn); - -#endif /* __LIBPERF_CORE_H */ diff --git a/tools/perf/lib/include/perf/cpumap.h b/tools/perf/lib/include/perf/cpumap.h deleted file mode 100644 index 6a17ad730cbc..000000000000 --- a/tools/perf/lib/include/perf/cpumap.h +++ /dev/null @@ -1,28 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_CPUMAP_H -#define __LIBPERF_CPUMAP_H - -#include <perf/core.h> -#include <stdio.h> -#include <stdbool.h> - -struct perf_cpu_map; - -LIBPERF_API struct perf_cpu_map *perf_cpu_map__dummy_new(void); -LIBPERF_API struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list); -LIBPERF_API struct perf_cpu_map *perf_cpu_map__read(FILE *file); -LIBPERF_API struct perf_cpu_map *perf_cpu_map__get(struct perf_cpu_map *map); -LIBPERF_API struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig, - struct perf_cpu_map *other); -LIBPERF_API void perf_cpu_map__put(struct perf_cpu_map *map); -LIBPERF_API int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx); -LIBPERF_API int perf_cpu_map__nr(const struct perf_cpu_map *cpus); -LIBPERF_API bool perf_cpu_map__empty(const struct perf_cpu_map *map); -LIBPERF_API int perf_cpu_map__max(struct perf_cpu_map *map); - -#define perf_cpu_map__for_each_cpu(cpu, idx, cpus) \ - for ((idx) = 0, (cpu) = perf_cpu_map__cpu(cpus, idx); \ - (idx) < perf_cpu_map__nr(cpus); \ - (idx)++, (cpu) = perf_cpu_map__cpu(cpus, idx)) - -#endif /* __LIBPERF_CPUMAP_H */ diff --git a/tools/perf/lib/include/perf/event.h b/tools/perf/lib/include/perf/event.h deleted file mode 100644 index 18106899cb4e..000000000000 --- a/tools/perf/lib/include/perf/event.h +++ /dev/null @@ -1,385 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_EVENT_H -#define __LIBPERF_EVENT_H - -#include <linux/perf_event.h> -#include <linux/types.h> -#include <linux/limits.h> -#include <linux/bpf.h> -#include <sys/types.h> /* pid_t */ - -struct perf_record_mmap { - struct perf_event_header header; - __u32 pid, tid; - __u64 start; - __u64 len; - __u64 pgoff; - char filename[PATH_MAX]; -}; - -struct perf_record_mmap2 { - struct perf_event_header header; - __u32 pid, tid; - __u64 start; - __u64 len; - __u64 pgoff; - __u32 maj; - __u32 min; - __u64 ino; - __u64 ino_generation; - __u32 prot; - __u32 flags; - char filename[PATH_MAX]; -}; - -struct perf_record_comm { - struct perf_event_header header; - __u32 pid, tid; - char comm[16]; -}; - -struct perf_record_namespaces { - struct perf_event_header header; - __u32 pid, tid; - __u64 nr_namespaces; - struct perf_ns_link_info link_info[]; -}; - -struct perf_record_fork { - struct perf_event_header header; - __u32 pid, ppid; - __u32 tid, ptid; - __u64 time; -}; - -struct perf_record_lost { - struct perf_event_header header; - __u64 id; - __u64 lost; -}; - -struct perf_record_lost_samples { - struct perf_event_header header; - __u64 lost; -}; - -/* - * PERF_FORMAT_ENABLED | PERF_FORMAT_RUNNING | PERF_FORMAT_ID - */ -struct perf_record_read { - struct perf_event_header header; - __u32 pid, tid; - __u64 value; - __u64 time_enabled; - __u64 time_running; - __u64 id; -}; - -struct perf_record_throttle { - struct perf_event_header header; - __u64 time; - __u64 id; - __u64 stream_id; -}; - -#ifndef KSYM_NAME_LEN -#define KSYM_NAME_LEN 256 -#endif - -struct perf_record_ksymbol { - struct perf_event_header header; - __u64 addr; - __u32 len; - __u16 ksym_type; - __u16 flags; - char name[KSYM_NAME_LEN]; -}; - -struct perf_record_bpf_event { - struct perf_event_header header; - __u16 type; - __u16 flags; - __u32 id; - - /* for bpf_prog types */ - __u8 tag[BPF_TAG_SIZE]; // prog tag -}; - -struct perf_record_sample { - struct perf_event_header header; - __u64 array[]; -}; - -struct perf_record_switch { - struct perf_event_header header; - __u32 next_prev_pid; - __u32 next_prev_tid; -}; - -struct perf_record_header_attr { - struct perf_event_header header; - struct perf_event_attr attr; - __u64 id[]; -}; - -enum { - PERF_CPU_MAP__CPUS = 0, - PERF_CPU_MAP__MASK = 1, -}; - -struct cpu_map_entries { - __u16 nr; - __u16 cpu[]; -}; - -struct perf_record_record_cpu_map { - __u16 nr; - __u16 long_size; - unsigned long mask[]; -}; - -struct perf_record_cpu_map_data { - __u16 type; - char data[]; -}; - -struct perf_record_cpu_map { - struct perf_event_header header; - struct perf_record_cpu_map_data data; -}; - -enum { - PERF_EVENT_UPDATE__UNIT = 0, - PERF_EVENT_UPDATE__SCALE = 1, - PERF_EVENT_UPDATE__NAME = 2, - PERF_EVENT_UPDATE__CPUS = 3, -}; - -struct perf_record_event_update_cpus { - struct perf_record_cpu_map_data cpus; -}; - -struct perf_record_event_update_scale { - double scale; -}; - -struct perf_record_event_update { - struct perf_event_header header; - __u64 type; - __u64 id; - char data[]; -}; - -#define MAX_EVENT_NAME 64 - -struct perf_trace_event_type { - __u64 event_id; - char name[MAX_EVENT_NAME]; -}; - -struct perf_record_header_event_type { - struct perf_event_header header; - struct perf_trace_event_type event_type; -}; - -struct perf_record_header_tracing_data { - struct perf_event_header header; - __u32 size; -}; - -struct perf_record_header_build_id { - struct perf_event_header header; - pid_t pid; - __u8 build_id[24]; - char filename[]; -}; - -struct id_index_entry { - __u64 id; - __u64 idx; - __u64 cpu; - __u64 tid; -}; - -struct perf_record_id_index { - struct perf_event_header header; - __u64 nr; - struct id_index_entry entries[0]; -}; - -struct perf_record_auxtrace_info { - struct perf_event_header header; - __u32 type; - __u32 reserved__; /* For alignment */ - __u64 priv[]; -}; - -struct perf_record_auxtrace { - struct perf_event_header header; - __u64 size; - __u64 offset; - __u64 reference; - __u32 idx; - __u32 tid; - __u32 cpu; - __u32 reserved__; /* For alignment */ -}; - -#define MAX_AUXTRACE_ERROR_MSG 64 - -struct perf_record_auxtrace_error { - struct perf_event_header header; - __u32 type; - __u32 code; - __u32 cpu; - __u32 pid; - __u32 tid; - __u32 fmt; - __u64 ip; - __u64 time; - char msg[MAX_AUXTRACE_ERROR_MSG]; -}; - -struct perf_record_aux { - struct perf_event_header header; - __u64 aux_offset; - __u64 aux_size; - __u64 flags; -}; - -struct perf_record_itrace_start { - struct perf_event_header header; - __u32 pid; - __u32 tid; -}; - -struct perf_record_thread_map_entry { - __u64 pid; - char comm[16]; -}; - -struct perf_record_thread_map { - struct perf_event_header header; - __u64 nr; - struct perf_record_thread_map_entry entries[]; -}; - -enum { - PERF_STAT_CONFIG_TERM__AGGR_MODE = 0, - PERF_STAT_CONFIG_TERM__INTERVAL = 1, - PERF_STAT_CONFIG_TERM__SCALE = 2, - PERF_STAT_CONFIG_TERM__MAX = 3, -}; - -struct perf_record_stat_config_entry { - __u64 tag; - __u64 val; -}; - -struct perf_record_stat_config { - struct perf_event_header header; - __u64 nr; - struct perf_record_stat_config_entry data[]; -}; - -struct perf_record_stat { - struct perf_event_header header; - - __u64 id; - __u32 cpu; - __u32 thread; - - union { - struct { - __u64 val; - __u64 ena; - __u64 run; - }; - __u64 values[3]; - }; -}; - -struct perf_record_stat_round { - struct perf_event_header header; - __u64 type; - __u64 time; -}; - -struct perf_record_time_conv { - struct perf_event_header header; - __u64 time_shift; - __u64 time_mult; - __u64 time_zero; -}; - -struct perf_record_header_feature { - struct perf_event_header header; - __u64 feat_id; - char data[]; -}; - -struct perf_record_compressed { - struct perf_event_header header; - char data[]; -}; - -enum perf_user_event_type { /* above any possible kernel type */ - PERF_RECORD_USER_TYPE_START = 64, - PERF_RECORD_HEADER_ATTR = 64, - PERF_RECORD_HEADER_EVENT_TYPE = 65, /* deprecated */ - PERF_RECORD_HEADER_TRACING_DATA = 66, - PERF_RECORD_HEADER_BUILD_ID = 67, - PERF_RECORD_FINISHED_ROUND = 68, - PERF_RECORD_ID_INDEX = 69, - PERF_RECORD_AUXTRACE_INFO = 70, - PERF_RECORD_AUXTRACE = 71, - PERF_RECORD_AUXTRACE_ERROR = 72, - PERF_RECORD_THREAD_MAP = 73, - PERF_RECORD_CPU_MAP = 74, - PERF_RECORD_STAT_CONFIG = 75, - PERF_RECORD_STAT = 76, - PERF_RECORD_STAT_ROUND = 77, - PERF_RECORD_EVENT_UPDATE = 78, - PERF_RECORD_TIME_CONV = 79, - PERF_RECORD_HEADER_FEATURE = 80, - PERF_RECORD_COMPRESSED = 81, - PERF_RECORD_HEADER_MAX -}; - -union perf_event { - struct perf_event_header header; - struct perf_record_mmap mmap; - struct perf_record_mmap2 mmap2; - struct perf_record_comm comm; - struct perf_record_namespaces namespaces; - struct perf_record_fork fork; - struct perf_record_lost lost; - struct perf_record_lost_samples lost_samples; - struct perf_record_read read; - struct perf_record_throttle throttle; - struct perf_record_sample sample; - struct perf_record_bpf_event bpf; - struct perf_record_ksymbol ksymbol; - struct perf_record_header_attr attr; - struct perf_record_event_update event_update; - struct perf_record_header_event_type event_type; - struct perf_record_header_tracing_data tracing_data; - struct perf_record_header_build_id build_id; - struct perf_record_id_index id_index; - struct perf_record_auxtrace_info auxtrace_info; - struct perf_record_auxtrace auxtrace; - struct perf_record_auxtrace_error auxtrace_error; - struct perf_record_aux aux; - struct perf_record_itrace_start itrace_start; - struct perf_record_switch context_switch; - struct perf_record_thread_map thread_map; - struct perf_record_cpu_map cpu_map; - struct perf_record_stat_config stat_config; - struct perf_record_stat stat; - struct perf_record_stat_round stat_round; - struct perf_record_time_conv time_conv; - struct perf_record_header_feature feat; - struct perf_record_compressed pack; -}; - -#endif /* __LIBPERF_EVENT_H */ diff --git a/tools/perf/lib/include/perf/evlist.h b/tools/perf/lib/include/perf/evlist.h deleted file mode 100644 index 0a7479dc13bf..000000000000 --- a/tools/perf/lib/include/perf/evlist.h +++ /dev/null @@ -1,49 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_EVLIST_H -#define __LIBPERF_EVLIST_H - -#include <perf/core.h> -#include <stdbool.h> - -struct perf_evlist; -struct perf_evsel; -struct perf_cpu_map; -struct perf_thread_map; - -LIBPERF_API void perf_evlist__add(struct perf_evlist *evlist, - struct perf_evsel *evsel); -LIBPERF_API void perf_evlist__remove(struct perf_evlist *evlist, - struct perf_evsel *evsel); -LIBPERF_API struct perf_evlist *perf_evlist__new(void); -LIBPERF_API void perf_evlist__delete(struct perf_evlist *evlist); -LIBPERF_API struct perf_evsel* perf_evlist__next(struct perf_evlist *evlist, - struct perf_evsel *evsel); -LIBPERF_API int perf_evlist__open(struct perf_evlist *evlist); -LIBPERF_API void perf_evlist__close(struct perf_evlist *evlist); -LIBPERF_API void perf_evlist__enable(struct perf_evlist *evlist); -LIBPERF_API void perf_evlist__disable(struct perf_evlist *evlist); - -#define perf_evlist__for_each_evsel(evlist, pos) \ - for ((pos) = perf_evlist__next((evlist), NULL); \ - (pos) != NULL; \ - (pos) = perf_evlist__next((evlist), (pos))) - -LIBPERF_API void perf_evlist__set_maps(struct perf_evlist *evlist, - struct perf_cpu_map *cpus, - struct perf_thread_map *threads); -LIBPERF_API int perf_evlist__poll(struct perf_evlist *evlist, int timeout); -LIBPERF_API int perf_evlist__filter_pollfd(struct perf_evlist *evlist, - short revents_and_mask); - -LIBPERF_API int perf_evlist__mmap(struct perf_evlist *evlist, int pages); -LIBPERF_API void perf_evlist__munmap(struct perf_evlist *evlist); - -LIBPERF_API struct perf_mmap *perf_evlist__next_mmap(struct perf_evlist *evlist, - struct perf_mmap *map, - bool overwrite); -#define perf_evlist__for_each_mmap(evlist, pos, overwrite) \ - for ((pos) = perf_evlist__next_mmap((evlist), NULL, overwrite); \ - (pos) != NULL; \ - (pos) = perf_evlist__next_mmap((evlist), (pos), overwrite)) - -#endif /* __LIBPERF_EVLIST_H */ diff --git a/tools/perf/lib/include/perf/evsel.h b/tools/perf/lib/include/perf/evsel.h deleted file mode 100644 index c82ec39a4ad0..000000000000 --- a/tools/perf/lib/include/perf/evsel.h +++ /dev/null @@ -1,40 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_EVSEL_H -#define __LIBPERF_EVSEL_H - -#include <stdint.h> -#include <perf/core.h> - -struct perf_evsel; -struct perf_event_attr; -struct perf_cpu_map; -struct perf_thread_map; - -struct perf_counts_values { - union { - struct { - uint64_t val; - uint64_t ena; - uint64_t run; - }; - uint64_t values[3]; - }; -}; - -LIBPERF_API struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr); -LIBPERF_API void perf_evsel__delete(struct perf_evsel *evsel); -LIBPERF_API int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus, - struct perf_thread_map *threads); -LIBPERF_API void perf_evsel__close(struct perf_evsel *evsel); -LIBPERF_API void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu); -LIBPERF_API int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, - struct perf_counts_values *count); -LIBPERF_API int perf_evsel__enable(struct perf_evsel *evsel); -LIBPERF_API int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu); -LIBPERF_API int perf_evsel__disable(struct perf_evsel *evsel); -LIBPERF_API int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu); -LIBPERF_API struct perf_cpu_map *perf_evsel__cpus(struct perf_evsel *evsel); -LIBPERF_API struct perf_thread_map *perf_evsel__threads(struct perf_evsel *evsel); -LIBPERF_API struct perf_event_attr *perf_evsel__attr(struct perf_evsel *evsel); - -#endif /* __LIBPERF_EVSEL_H */ diff --git a/tools/perf/lib/include/perf/mmap.h b/tools/perf/lib/include/perf/mmap.h deleted file mode 100644 index 9508ad90d8b9..000000000000 --- a/tools/perf/lib/include/perf/mmap.h +++ /dev/null @@ -1,15 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_MMAP_H -#define __LIBPERF_MMAP_H - -#include <perf/core.h> - -struct perf_mmap; -union perf_event; - -LIBPERF_API void perf_mmap__consume(struct perf_mmap *map); -LIBPERF_API int perf_mmap__read_init(struct perf_mmap *map); -LIBPERF_API void perf_mmap__read_done(struct perf_mmap *map); -LIBPERF_API union perf_event *perf_mmap__read_event(struct perf_mmap *map); - -#endif /* __LIBPERF_MMAP_H */ diff --git a/tools/perf/lib/include/perf/threadmap.h b/tools/perf/lib/include/perf/threadmap.h deleted file mode 100644 index a7c50de8d010..000000000000 --- a/tools/perf/lib/include/perf/threadmap.h +++ /dev/null @@ -1,20 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_THREADMAP_H -#define __LIBPERF_THREADMAP_H - -#include <perf/core.h> -#include <sys/types.h> - -struct perf_thread_map; - -LIBPERF_API struct perf_thread_map *perf_thread_map__new_dummy(void); - -LIBPERF_API void perf_thread_map__set_pid(struct perf_thread_map *map, int thread, pid_t pid); -LIBPERF_API char *perf_thread_map__comm(struct perf_thread_map *map, int thread); -LIBPERF_API int perf_thread_map__nr(struct perf_thread_map *threads); -LIBPERF_API pid_t perf_thread_map__pid(struct perf_thread_map *map, int thread); - -LIBPERF_API struct perf_thread_map *perf_thread_map__get(struct perf_thread_map *map); -LIBPERF_API void perf_thread_map__put(struct perf_thread_map *map); - -#endif /* __LIBPERF_THREADMAP_H */ diff --git a/tools/perf/lib/internal.h b/tools/perf/lib/internal.h deleted file mode 100644 index 2c27e158de6b..000000000000 --- a/tools/perf/lib/internal.h +++ /dev/null @@ -1,23 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_H -#define __LIBPERF_INTERNAL_H - -#include <perf/core.h> - -void libperf_print(enum libperf_print_level level, - const char *format, ...) - __attribute__((format(printf, 2, 3))); - -#define __pr(level, fmt, ...) \ -do { \ - libperf_print(level, "libperf: " fmt, ##__VA_ARGS__); \ -} while (0) - -#define pr_err(fmt, ...) __pr(LIBPERF_ERR, fmt, ##__VA_ARGS__) -#define pr_warning(fmt, ...) __pr(LIBPERF_WARN, fmt, ##__VA_ARGS__) -#define pr_info(fmt, ...) __pr(LIBPERF_INFO, fmt, ##__VA_ARGS__) -#define pr_debug(fmt, ...) __pr(LIBPERF_DEBUG, fmt, ##__VA_ARGS__) -#define pr_debug2(fmt, ...) __pr(LIBPERF_DEBUG2, fmt, ##__VA_ARGS__) -#define pr_debug3(fmt, ...) __pr(LIBPERF_DEBUG3, fmt, ##__VA_ARGS__) - -#endif /* __LIBPERF_INTERNAL_H */ diff --git a/tools/perf/lib/lib.c b/tools/perf/lib/lib.c deleted file mode 100644 index 18658931fc71..000000000000 --- a/tools/perf/lib/lib.c +++ /dev/null @@ -1,48 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <unistd.h> -#include <stdbool.h> -#include <errno.h> -#include <linux/kernel.h> -#include <internal/lib.h> - -unsigned int page_size; - -static ssize_t ion(bool is_read, int fd, void *buf, size_t n) -{ - void *buf_start = buf; - size_t left = n; - - while (left) { - /* buf must be treated as const if !is_read. */ - ssize_t ret = is_read ? read(fd, buf, left) : - write(fd, buf, left); - - if (ret < 0 && errno == EINTR) - continue; - if (ret <= 0) - return ret; - - left -= ret; - buf += ret; - } - - BUG_ON((size_t)(buf - buf_start) != n); - return n; -} - -/* - * Read exactly 'n' bytes or return an error. - */ -ssize_t readn(int fd, void *buf, size_t n) -{ - return ion(true, fd, buf, n); -} - -/* - * Write exactly 'n' bytes or return an error. - */ -ssize_t writen(int fd, const void *buf, size_t n) -{ - /* ion does not modify buf. */ - return ion(false, fd, (void *)buf, n); -} diff --git a/tools/perf/lib/libperf.map b/tools/perf/lib/libperf.map deleted file mode 100644 index 7be1af8a546c..000000000000 --- a/tools/perf/lib/libperf.map +++ /dev/null @@ -1,51 +0,0 @@ -LIBPERF_0.0.1 { - global: - libperf_init; - perf_cpu_map__dummy_new; - perf_cpu_map__get; - perf_cpu_map__put; - perf_cpu_map__new; - perf_cpu_map__read; - perf_cpu_map__nr; - perf_cpu_map__cpu; - perf_cpu_map__empty; - perf_cpu_map__max; - perf_thread_map__new_dummy; - perf_thread_map__set_pid; - perf_thread_map__comm; - perf_thread_map__nr; - perf_thread_map__pid; - perf_thread_map__get; - perf_thread_map__put; - perf_evsel__new; - perf_evsel__delete; - perf_evsel__enable; - perf_evsel__disable; - perf_evsel__open; - perf_evsel__close; - perf_evsel__read; - perf_evsel__cpus; - perf_evsel__threads; - perf_evsel__attr; - perf_evlist__new; - perf_evlist__delete; - perf_evlist__open; - perf_evlist__close; - perf_evlist__enable; - perf_evlist__disable; - perf_evlist__add; - perf_evlist__remove; - perf_evlist__next; - perf_evlist__set_maps; - perf_evlist__poll; - perf_evlist__mmap; - perf_evlist__munmap; - perf_evlist__filter_pollfd; - perf_evlist__next_mmap; - perf_mmap__consume; - perf_mmap__read_init; - perf_mmap__read_done; - perf_mmap__read_event; - local: - *; -}; diff --git a/tools/perf/lib/libperf.pc.template b/tools/perf/lib/libperf.pc.template deleted file mode 100644 index 117e4a237b55..000000000000 --- a/tools/perf/lib/libperf.pc.template +++ /dev/null @@ -1,11 +0,0 @@ -# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -prefix=@PREFIX@ -libdir=@LIBDIR@ -includedir=${prefix}/include - -Name: libperf -Description: perf library -Version: @VERSION@ -Libs: -L${libdir} -lperf -Cflags: -I${includedir} diff --git a/tools/perf/lib/mmap.c b/tools/perf/lib/mmap.c deleted file mode 100644 index 79d5ed6c38cc..000000000000 --- a/tools/perf/lib/mmap.c +++ /dev/null @@ -1,275 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <sys/mman.h> -#include <inttypes.h> -#include <asm/bug.h> -#include <errno.h> -#include <string.h> -#include <linux/ring_buffer.h> -#include <linux/perf_event.h> -#include <perf/mmap.h> -#include <perf/event.h> -#include <internal/mmap.h> -#include <internal/lib.h> -#include <linux/kernel.h> -#include "internal.h" - -void perf_mmap__init(struct perf_mmap *map, struct perf_mmap *prev, - bool overwrite, libperf_unmap_cb_t unmap_cb) -{ - map->fd = -1; - map->overwrite = overwrite; - map->unmap_cb = unmap_cb; - refcount_set(&map->refcnt, 0); - if (prev) - prev->next = map; -} - -size_t perf_mmap__mmap_len(struct perf_mmap *map) -{ - return map->mask + 1 + page_size; -} - -int perf_mmap__mmap(struct perf_mmap *map, struct perf_mmap_param *mp, - int fd, int cpu) -{ - map->prev = 0; - map->mask = mp->mask; - map->base = mmap(NULL, perf_mmap__mmap_len(map), mp->prot, - MAP_SHARED, fd, 0); - if (map->base == MAP_FAILED) { - map->base = NULL; - return -1; - } - - map->fd = fd; - map->cpu = cpu; - return 0; -} - -void perf_mmap__munmap(struct perf_mmap *map) -{ - if (map && map->base != NULL) { - munmap(map->base, perf_mmap__mmap_len(map)); - map->base = NULL; - map->fd = -1; - refcount_set(&map->refcnt, 0); - } - if (map && map->unmap_cb) - map->unmap_cb(map); -} - -void perf_mmap__get(struct perf_mmap *map) -{ - refcount_inc(&map->refcnt); -} - -void perf_mmap__put(struct perf_mmap *map) -{ - BUG_ON(map->base && refcount_read(&map->refcnt) == 0); - - if (refcount_dec_and_test(&map->refcnt)) - perf_mmap__munmap(map); -} - -static inline void perf_mmap__write_tail(struct perf_mmap *md, u64 tail) -{ - ring_buffer_write_tail(md->base, tail); -} - -u64 perf_mmap__read_head(struct perf_mmap *map) -{ - return ring_buffer_read_head(map->base); -} - -static bool perf_mmap__empty(struct perf_mmap *map) -{ - struct perf_event_mmap_page *pc = map->base; - - return perf_mmap__read_head(map) == map->prev && !pc->aux_size; -} - -void perf_mmap__consume(struct perf_mmap *map) -{ - if (!map->overwrite) { - u64 old = map->prev; - - perf_mmap__write_tail(map, old); - } - - if (refcount_read(&map->refcnt) == 1 && perf_mmap__empty(map)) - perf_mmap__put(map); -} - -static int overwrite_rb_find_range(void *buf, int mask, u64 *start, u64 *end) -{ - struct perf_event_header *pheader; - u64 evt_head = *start; - int size = mask + 1; - - pr_debug2("%s: buf=%p, start=%"PRIx64"\n", __func__, buf, *start); - pheader = (struct perf_event_header *)(buf + (*start & mask)); - while (true) { - if (evt_head - *start >= (unsigned int)size) { - pr_debug("Finished reading overwrite ring buffer: rewind\n"); - if (evt_head - *start > (unsigned int)size) - evt_head -= pheader->size; - *end = evt_head; - return 0; - } - - pheader = (struct perf_event_header *)(buf + (evt_head & mask)); - - if (pheader->size == 0) { - pr_debug("Finished reading overwrite ring buffer: get start\n"); - *end = evt_head; - return 0; - } - - evt_head += pheader->size; - pr_debug3("move evt_head: %"PRIx64"\n", evt_head); - } - WARN_ONCE(1, "Shouldn't get here\n"); - return -1; -} - -/* - * Report the start and end of the available data in ringbuffer - */ -static int __perf_mmap__read_init(struct perf_mmap *md) -{ - u64 head = perf_mmap__read_head(md); - u64 old = md->prev; - unsigned char *data = md->base + page_size; - unsigned long size; - - md->start = md->overwrite ? head : old; - md->end = md->overwrite ? old : head; - - if ((md->end - md->start) < md->flush) - return -EAGAIN; - - size = md->end - md->start; - if (size > (unsigned long)(md->mask) + 1) { - if (!md->overwrite) { - WARN_ONCE(1, "failed to keep up with mmap data. (warn only once)\n"); - - md->prev = head; - perf_mmap__consume(md); - return -EAGAIN; - } - - /* - * Backward ring buffer is full. We still have a chance to read - * most of data from it. - */ - if (overwrite_rb_find_range(data, md->mask, &md->start, &md->end)) - return -EINVAL; - } - - return 0; -} - -int perf_mmap__read_init(struct perf_mmap *map) -{ - /* - * Check if event was unmapped due to a POLLHUP/POLLERR. - */ - if (!refcount_read(&map->refcnt)) - return -ENOENT; - - return __perf_mmap__read_init(map); -} - -/* - * Mandatory for overwrite mode - * The direction of overwrite mode is backward. - * The last perf_mmap__read() will set tail to map->core.prev. - * Need to correct the map->core.prev to head which is the end of next read. - */ -void perf_mmap__read_done(struct perf_mmap *map) -{ - /* - * Check if event was unmapped due to a POLLHUP/POLLERR. - */ - if (!refcount_read(&map->refcnt)) - return; - - map->prev = perf_mmap__read_head(map); -} - -/* When check_messup is true, 'end' must points to a good entry */ -static union perf_event *perf_mmap__read(struct perf_mmap *map, - u64 *startp, u64 end) -{ - unsigned char *data = map->base + page_size; - union perf_event *event = NULL; - int diff = end - *startp; - - if (diff >= (int)sizeof(event->header)) { - size_t size; - - event = (union perf_event *)&data[*startp & map->mask]; - size = event->header.size; - - if (size < sizeof(event->header) || diff < (int)size) - return NULL; - - /* - * Event straddles the mmap boundary -- header should always - * be inside due to u64 alignment of output. - */ - if ((*startp & map->mask) + size != ((*startp + size) & map->mask)) { - unsigned int offset = *startp; - unsigned int len = min(sizeof(*event), size), cpy; - void *dst = map->event_copy; - - do { - cpy = min(map->mask + 1 - (offset & map->mask), len); - memcpy(dst, &data[offset & map->mask], cpy); - offset += cpy; - dst += cpy; - len -= cpy; - } while (len); - - event = (union perf_event *)map->event_copy; - } - - *startp += size; - } - - return event; -} - -/* - * Read event from ring buffer one by one. - * Return one event for each call. - * - * Usage: - * perf_mmap__read_init() - * while(event = perf_mmap__read_event()) { - * //process the event - * perf_mmap__consume() - * } - * perf_mmap__read_done() - */ -union perf_event *perf_mmap__read_event(struct perf_mmap *map) -{ - union perf_event *event; - - /* - * Check if event was unmapped due to a POLLHUP/POLLERR. - */ - if (!refcount_read(&map->refcnt)) - return NULL; - - /* non-overwirte doesn't pause the ringbuffer */ - if (!map->overwrite) - map->end = perf_mmap__read_head(map); - - event = perf_mmap__read(map, &map->start, map->end); - - if (!map->overwrite) - map->prev = map->start; - - return event; -} diff --git a/tools/perf/lib/tests/Makefile b/tools/perf/lib/tests/Makefile deleted file mode 100644 index a43cd08c5c03..000000000000 --- a/tools/perf/lib/tests/Makefile +++ /dev/null @@ -1,38 +0,0 @@ -# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -TESTS = test-cpumap test-threadmap test-evlist test-evsel - -TESTS_SO := $(addsuffix -so,$(TESTS)) -TESTS_A := $(addsuffix -a,$(TESTS)) - -# Set compile option CFLAGS -ifdef EXTRA_CFLAGS - CFLAGS := $(EXTRA_CFLAGS) -else - CFLAGS := -g -Wall -endif - -all: - -include $(srctree)/tools/scripts/Makefile.include - -INCLUDE = -I$(srctree)/tools/perf/lib/include -I$(srctree)/tools/include -I$(srctree)/tools/lib - -$(TESTS_A): FORCE - $(QUIET_LINK)$(CC) $(INCLUDE) $(CFLAGS) -o $@ $(subst -a,.c,$@) ../libperf.a $(LIBAPI) - -$(TESTS_SO): FORCE - $(QUIET_LINK)$(CC) $(INCLUDE) $(CFLAGS) -L.. -o $@ $(subst -so,.c,$@) $(LIBAPI) -lperf - -all: $(TESTS_A) $(TESTS_SO) - -run: - @echo "running static:" - @for i in $(TESTS_A); do ./$$i; done - @echo "running dynamic:" - @for i in $(TESTS_SO); do LD_LIBRARY_PATH=../ ./$$i; done - -clean: - $(call QUIET_CLEAN, tests)$(RM) $(TESTS_A) $(TESTS_SO) - -.PHONY: all clean FORCE diff --git a/tools/perf/lib/tests/test-cpumap.c b/tools/perf/lib/tests/test-cpumap.c deleted file mode 100644 index c8d45091e7c2..000000000000 --- a/tools/perf/lib/tests/test-cpumap.c +++ /dev/null @@ -1,31 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <stdarg.h> -#include <stdio.h> -#include <perf/cpumap.h> -#include <internal/tests.h> - -static int libperf_print(enum libperf_print_level level, - const char *fmt, va_list ap) -{ - return vfprintf(stderr, fmt, ap); -} - -int main(int argc, char **argv) -{ - struct perf_cpu_map *cpus; - - __T_START; - - libperf_init(libperf_print); - - cpus = perf_cpu_map__dummy_new(); - if (!cpus) - return -1; - - perf_cpu_map__get(cpus); - perf_cpu_map__put(cpus); - perf_cpu_map__put(cpus); - - __T_END; - return 0; -} diff --git a/tools/perf/lib/tests/test-evlist.c b/tools/perf/lib/tests/test-evlist.c deleted file mode 100644 index 6d8ebe0c2504..000000000000 --- a/tools/perf/lib/tests/test-evlist.c +++ /dev/null @@ -1,413 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#define _GNU_SOURCE // needed for sched.h to get sched_[gs]etaffinity and CPU_(ZERO,SET) -#include <sched.h> -#include <stdio.h> -#include <stdarg.h> -#include <unistd.h> -#include <stdlib.h> -#include <linux/perf_event.h> -#include <linux/limits.h> -#include <sys/types.h> -#include <sys/wait.h> -#include <sys/prctl.h> -#include <perf/cpumap.h> -#include <perf/threadmap.h> -#include <perf/evlist.h> -#include <perf/evsel.h> -#include <perf/mmap.h> -#include <perf/event.h> -#include <internal/tests.h> -#include <api/fs/fs.h> - -static int libperf_print(enum libperf_print_level level, - const char *fmt, va_list ap) -{ - return vfprintf(stderr, fmt, ap); -} - -static int test_stat_cpu(void) -{ - struct perf_cpu_map *cpus; - struct perf_evlist *evlist; - struct perf_evsel *evsel; - struct perf_event_attr attr1 = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_CPU_CLOCK, - }; - struct perf_event_attr attr2 = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_TASK_CLOCK, - }; - int err, cpu, tmp; - - cpus = perf_cpu_map__new(NULL); - __T("failed to create cpus", cpus); - - evlist = perf_evlist__new(); - __T("failed to create evlist", evlist); - - evsel = perf_evsel__new(&attr1); - __T("failed to create evsel1", evsel); - - perf_evlist__add(evlist, evsel); - - evsel = perf_evsel__new(&attr2); - __T("failed to create evsel2", evsel); - - perf_evlist__add(evlist, evsel); - - perf_evlist__set_maps(evlist, cpus, NULL); - - err = perf_evlist__open(evlist); - __T("failed to open evsel", err == 0); - - perf_evlist__for_each_evsel(evlist, evsel) { - cpus = perf_evsel__cpus(evsel); - - perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { - struct perf_counts_values counts = { .val = 0 }; - - perf_evsel__read(evsel, cpu, 0, &counts); - __T("failed to read value for evsel", counts.val != 0); - } - } - - perf_evlist__close(evlist); - perf_evlist__delete(evlist); - - perf_cpu_map__put(cpus); - return 0; -} - -static int test_stat_thread(void) -{ - struct perf_counts_values counts = { .val = 0 }; - struct perf_thread_map *threads; - struct perf_evlist *evlist; - struct perf_evsel *evsel; - struct perf_event_attr attr1 = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_CPU_CLOCK, - }; - struct perf_event_attr attr2 = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_TASK_CLOCK, - }; - int err; - - threads = perf_thread_map__new_dummy(); - __T("failed to create threads", threads); - - perf_thread_map__set_pid(threads, 0, 0); - - evlist = perf_evlist__new(); - __T("failed to create evlist", evlist); - - evsel = perf_evsel__new(&attr1); - __T("failed to create evsel1", evsel); - - perf_evlist__add(evlist, evsel); - - evsel = perf_evsel__new(&attr2); - __T("failed to create evsel2", evsel); - - perf_evlist__add(evlist, evsel); - - perf_evlist__set_maps(evlist, NULL, threads); - - err = perf_evlist__open(evlist); - __T("failed to open evsel", err == 0); - - perf_evlist__for_each_evsel(evlist, evsel) { - perf_evsel__read(evsel, 0, 0, &counts); - __T("failed to read value for evsel", counts.val != 0); - } - - perf_evlist__close(evlist); - perf_evlist__delete(evlist); - - perf_thread_map__put(threads); - return 0; -} - -static int test_stat_thread_enable(void) -{ - struct perf_counts_values counts = { .val = 0 }; - struct perf_thread_map *threads; - struct perf_evlist *evlist; - struct perf_evsel *evsel; - struct perf_event_attr attr1 = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_CPU_CLOCK, - .disabled = 1, - }; - struct perf_event_attr attr2 = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_TASK_CLOCK, - .disabled = 1, - }; - int err; - - threads = perf_thread_map__new_dummy(); - __T("failed to create threads", threads); - - perf_thread_map__set_pid(threads, 0, 0); - - evlist = perf_evlist__new(); - __T("failed to create evlist", evlist); - - evsel = perf_evsel__new(&attr1); - __T("failed to create evsel1", evsel); - - perf_evlist__add(evlist, evsel); - - evsel = perf_evsel__new(&attr2); - __T("failed to create evsel2", evsel); - - perf_evlist__add(evlist, evsel); - - perf_evlist__set_maps(evlist, NULL, threads); - - err = perf_evlist__open(evlist); - __T("failed to open evsel", err == 0); - - perf_evlist__for_each_evsel(evlist, evsel) { - perf_evsel__read(evsel, 0, 0, &counts); - __T("failed to read value for evsel", counts.val == 0); - } - - perf_evlist__enable(evlist); - - perf_evlist__for_each_evsel(evlist, evsel) { - perf_evsel__read(evsel, 0, 0, &counts); - __T("failed to read value for evsel", counts.val != 0); - } - - perf_evlist__disable(evlist); - - perf_evlist__close(evlist); - perf_evlist__delete(evlist); - - perf_thread_map__put(threads); - return 0; -} - -static int test_mmap_thread(void) -{ - struct perf_evlist *evlist; - struct perf_evsel *evsel; - struct perf_mmap *map; - struct perf_cpu_map *cpus; - struct perf_thread_map *threads; - struct perf_event_attr attr = { - .type = PERF_TYPE_TRACEPOINT, - .sample_period = 1, - .wakeup_watermark = 1, - .disabled = 1, - }; - char path[PATH_MAX]; - int id, err, pid, go_pipe[2]; - union perf_event *event; - char bf; - int count = 0; - - snprintf(path, PATH_MAX, "%s/kernel/debug/tracing/events/syscalls/sys_enter_prctl/id", - sysfs__mountpoint()); - - if (filename__read_int(path, &id)) { - fprintf(stderr, "error: failed to get tracepoint id: %s\n", path); - return -1; - } - - attr.config = id; - - err = pipe(go_pipe); - __T("failed to create pipe", err == 0); - - fflush(NULL); - - pid = fork(); - if (!pid) { - int i; - - read(go_pipe[0], &bf, 1); - - /* Generate 100 prctl calls. */ - for (i = 0; i < 100; i++) - prctl(0, 0, 0, 0, 0); - - exit(0); - } - - threads = perf_thread_map__new_dummy(); - __T("failed to create threads", threads); - - cpus = perf_cpu_map__dummy_new(); - __T("failed to create cpus", cpus); - - perf_thread_map__set_pid(threads, 0, pid); - - evlist = perf_evlist__new(); - __T("failed to create evlist", evlist); - - evsel = perf_evsel__new(&attr); - __T("failed to create evsel1", evsel); - - perf_evlist__add(evlist, evsel); - - perf_evlist__set_maps(evlist, cpus, threads); - - err = perf_evlist__open(evlist); - __T("failed to open evlist", err == 0); - - err = perf_evlist__mmap(evlist, 4); - __T("failed to mmap evlist", err == 0); - - perf_evlist__enable(evlist); - - /* kick the child and wait for it to finish */ - write(go_pipe[1], &bf, 1); - waitpid(pid, NULL, 0); - - /* - * There's no need to call perf_evlist__disable, - * monitored process is dead now. - */ - - perf_evlist__for_each_mmap(evlist, map, false) { - if (perf_mmap__read_init(map) < 0) - continue; - - while ((event = perf_mmap__read_event(map)) != NULL) { - count++; - perf_mmap__consume(map); - } - - perf_mmap__read_done(map); - } - - /* calls perf_evlist__munmap/perf_evlist__close */ - perf_evlist__delete(evlist); - - perf_thread_map__put(threads); - perf_cpu_map__put(cpus); - - /* - * The generated prctl calls should match the - * number of events in the buffer. - */ - __T("failed count", count == 100); - - return 0; -} - -static int test_mmap_cpus(void) -{ - struct perf_evlist *evlist; - struct perf_evsel *evsel; - struct perf_mmap *map; - struct perf_cpu_map *cpus; - struct perf_event_attr attr = { - .type = PERF_TYPE_TRACEPOINT, - .sample_period = 1, - .wakeup_watermark = 1, - .disabled = 1, - }; - cpu_set_t saved_mask; - char path[PATH_MAX]; - int id, err, cpu, tmp; - union perf_event *event; - int count = 0; - - snprintf(path, PATH_MAX, "%s/kernel/debug/tracing/events/syscalls/sys_enter_prctl/id", - sysfs__mountpoint()); - - if (filename__read_int(path, &id)) { - fprintf(stderr, "error: failed to get tracepoint id: %s\n", path); - return -1; - } - - attr.config = id; - - cpus = perf_cpu_map__new(NULL); - __T("failed to create cpus", cpus); - - evlist = perf_evlist__new(); - __T("failed to create evlist", evlist); - - evsel = perf_evsel__new(&attr); - __T("failed to create evsel1", evsel); - - perf_evlist__add(evlist, evsel); - - perf_evlist__set_maps(evlist, cpus, NULL); - - err = perf_evlist__open(evlist); - __T("failed to open evlist", err == 0); - - err = perf_evlist__mmap(evlist, 4); - __T("failed to mmap evlist", err == 0); - - perf_evlist__enable(evlist); - - err = sched_getaffinity(0, sizeof(saved_mask), &saved_mask); - __T("sched_getaffinity failed", err == 0); - - perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { - cpu_set_t mask; - - CPU_ZERO(&mask); - CPU_SET(cpu, &mask); - - err = sched_setaffinity(0, sizeof(mask), &mask); - __T("sched_setaffinity failed", err == 0); - - prctl(0, 0, 0, 0, 0); - } - - err = sched_setaffinity(0, sizeof(saved_mask), &saved_mask); - __T("sched_setaffinity failed", err == 0); - - perf_evlist__disable(evlist); - - perf_evlist__for_each_mmap(evlist, map, false) { - if (perf_mmap__read_init(map) < 0) - continue; - - while ((event = perf_mmap__read_event(map)) != NULL) { - count++; - perf_mmap__consume(map); - } - - perf_mmap__read_done(map); - } - - /* calls perf_evlist__munmap/perf_evlist__close */ - perf_evlist__delete(evlist); - - /* - * The generated prctl events should match the - * number of cpus or be bigger (we are system-wide). - */ - __T("failed count", count >= perf_cpu_map__nr(cpus)); - - perf_cpu_map__put(cpus); - - return 0; -} - -int main(int argc, char **argv) -{ - __T_START; - - libperf_init(libperf_print); - - test_stat_cpu(); - test_stat_thread(); - test_stat_thread_enable(); - test_mmap_thread(); - test_mmap_cpus(); - - __T_END; - return 0; -} diff --git a/tools/perf/lib/tests/test-evsel.c b/tools/perf/lib/tests/test-evsel.c deleted file mode 100644 index 135722ac965b..000000000000 --- a/tools/perf/lib/tests/test-evsel.c +++ /dev/null @@ -1,135 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <stdarg.h> -#include <stdio.h> -#include <linux/perf_event.h> -#include <perf/cpumap.h> -#include <perf/threadmap.h> -#include <perf/evsel.h> -#include <internal/tests.h> - -static int libperf_print(enum libperf_print_level level, - const char *fmt, va_list ap) -{ - return vfprintf(stderr, fmt, ap); -} - -static int test_stat_cpu(void) -{ - struct perf_cpu_map *cpus; - struct perf_evsel *evsel; - struct perf_event_attr attr = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_CPU_CLOCK, - }; - int err, cpu, tmp; - - cpus = perf_cpu_map__new(NULL); - __T("failed to create cpus", cpus); - - evsel = perf_evsel__new(&attr); - __T("failed to create evsel", evsel); - - err = perf_evsel__open(evsel, cpus, NULL); - __T("failed to open evsel", err == 0); - - perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { - struct perf_counts_values counts = { .val = 0 }; - - perf_evsel__read(evsel, cpu, 0, &counts); - __T("failed to read value for evsel", counts.val != 0); - } - - perf_evsel__close(evsel); - perf_evsel__delete(evsel); - - perf_cpu_map__put(cpus); - return 0; -} - -static int test_stat_thread(void) -{ - struct perf_counts_values counts = { .val = 0 }; - struct perf_thread_map *threads; - struct perf_evsel *evsel; - struct perf_event_attr attr = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_TASK_CLOCK, - }; - int err; - - threads = perf_thread_map__new_dummy(); - __T("failed to create threads", threads); - - perf_thread_map__set_pid(threads, 0, 0); - - evsel = perf_evsel__new(&attr); - __T("failed to create evsel", evsel); - - err = perf_evsel__open(evsel, NULL, threads); - __T("failed to open evsel", err == 0); - - perf_evsel__read(evsel, 0, 0, &counts); - __T("failed to read value for evsel", counts.val != 0); - - perf_evsel__close(evsel); - perf_evsel__delete(evsel); - - perf_thread_map__put(threads); - return 0; -} - -static int test_stat_thread_enable(void) -{ - struct perf_counts_values counts = { .val = 0 }; - struct perf_thread_map *threads; - struct perf_evsel *evsel; - struct perf_event_attr attr = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_TASK_CLOCK, - .disabled = 1, - }; - int err; - - threads = perf_thread_map__new_dummy(); - __T("failed to create threads", threads); - - perf_thread_map__set_pid(threads, 0, 0); - - evsel = perf_evsel__new(&attr); - __T("failed to create evsel", evsel); - - err = perf_evsel__open(evsel, NULL, threads); - __T("failed to open evsel", err == 0); - - perf_evsel__read(evsel, 0, 0, &counts); - __T("failed to read value for evsel", counts.val == 0); - - err = perf_evsel__enable(evsel); - __T("failed to enable evsel", err == 0); - - perf_evsel__read(evsel, 0, 0, &counts); - __T("failed to read value for evsel", counts.val != 0); - - err = perf_evsel__disable(evsel); - __T("failed to enable evsel", err == 0); - - perf_evsel__close(evsel); - perf_evsel__delete(evsel); - - perf_thread_map__put(threads); - return 0; -} - -int main(int argc, char **argv) -{ - __T_START; - - libperf_init(libperf_print); - - test_stat_cpu(); - test_stat_thread(); - test_stat_thread_enable(); - - __T_END; - return 0; -} diff --git a/tools/perf/lib/tests/test-threadmap.c b/tools/perf/lib/tests/test-threadmap.c deleted file mode 100644 index 7dc4d6fbedde..000000000000 --- a/tools/perf/lib/tests/test-threadmap.c +++ /dev/null @@ -1,31 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <stdarg.h> -#include <stdio.h> -#include <perf/threadmap.h> -#include <internal/tests.h> - -static int libperf_print(enum libperf_print_level level, - const char *fmt, va_list ap) -{ - return vfprintf(stderr, fmt, ap); -} - -int main(int argc, char **argv) -{ - struct perf_thread_map *threads; - - __T_START; - - libperf_init(libperf_print); - - threads = perf_thread_map__new_dummy(); - if (!threads) - return -1; - - perf_thread_map__get(threads); - perf_thread_map__put(threads); - perf_thread_map__put(threads); - - __T_END; - return 0; -} diff --git a/tools/perf/lib/threadmap.c b/tools/perf/lib/threadmap.c deleted file mode 100644 index e92c368b0a6c..000000000000 --- a/tools/perf/lib/threadmap.c +++ /dev/null @@ -1,91 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <perf/threadmap.h> -#include <stdlib.h> -#include <linux/refcount.h> -#include <internal/threadmap.h> -#include <string.h> -#include <asm/bug.h> -#include <stdio.h> - -static void perf_thread_map__reset(struct perf_thread_map *map, int start, int nr) -{ - size_t size = (nr - start) * sizeof(map->map[0]); - - memset(&map->map[start], 0, size); - map->err_thread = -1; -} - -struct perf_thread_map *perf_thread_map__realloc(struct perf_thread_map *map, int nr) -{ - size_t size = sizeof(*map) + sizeof(map->map[0]) * nr; - int start = map ? map->nr : 0; - - map = realloc(map, size); - /* - * We only realloc to add more items, let's reset new items. - */ - if (map) - perf_thread_map__reset(map, start, nr); - - return map; -} - -#define thread_map__alloc(__nr) perf_thread_map__realloc(NULL, __nr) - -void perf_thread_map__set_pid(struct perf_thread_map *map, int thread, pid_t pid) -{ - map->map[thread].pid = pid; -} - -char *perf_thread_map__comm(struct perf_thread_map *map, int thread) -{ - return map->map[thread].comm; -} - -struct perf_thread_map *perf_thread_map__new_dummy(void) -{ - struct perf_thread_map *threads = thread_map__alloc(1); - - if (threads != NULL) { - perf_thread_map__set_pid(threads, 0, -1); - threads->nr = 1; - refcount_set(&threads->refcnt, 1); - } - return threads; -} - -static void perf_thread_map__delete(struct perf_thread_map *threads) -{ - if (threads) { - int i; - - WARN_ONCE(refcount_read(&threads->refcnt) != 0, - "thread map refcnt unbalanced\n"); - for (i = 0; i < threads->nr; i++) - free(perf_thread_map__comm(threads, i)); - free(threads); - } -} - -struct perf_thread_map *perf_thread_map__get(struct perf_thread_map *map) -{ - if (map) - refcount_inc(&map->refcnt); - return map; -} - -void perf_thread_map__put(struct perf_thread_map *map) -{ - if (map && refcount_dec_and_test(&map->refcnt)) - perf_thread_map__delete(map); -} - -int perf_thread_map__nr(struct perf_thread_map *threads) -{ - return threads ? threads->nr : 1; -} - -pid_t perf_thread_map__pid(struct perf_thread_map *map, int thread) -{ - return map->map[thread].pid; -} diff --git a/tools/perf/lib/xyarray.c b/tools/perf/lib/xyarray.c deleted file mode 100644 index dcd901d154bb..000000000000 --- a/tools/perf/lib/xyarray.c +++ /dev/null @@ -1,33 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <internal/xyarray.h> -#include <linux/zalloc.h> -#include <stdlib.h> -#include <string.h> - -struct xyarray *xyarray__new(int xlen, int ylen, size_t entry_size) -{ - size_t row_size = ylen * entry_size; - struct xyarray *xy = zalloc(sizeof(*xy) + xlen * row_size); - - if (xy != NULL) { - xy->entry_size = entry_size; - xy->row_size = row_size; - xy->entries = xlen * ylen; - xy->max_x = xlen; - xy->max_y = ylen; - } - - return xy; -} - -void xyarray__reset(struct xyarray *xy) -{ - size_t n = xy->entries * xy->entry_size; - - memset(xy->contents, 0, n); -} - -void xyarray__delete(struct xyarray *xy) -{ - free(xy); -} -- 2.21.0 ^ permalink raw reply related [flat|nested] 5+ messages in thread
* [tip: perf/core] libperf: Move to tools/lib/perf 2019-12-06 21:06 ` [PATCH 1/2] libperf: Move libperf under tools/lib/perf Jiri Olsa @ 2020-01-10 17:53 ` tip-bot2 for Jiri Olsa 0 siblings, 0 replies; 5+ messages in thread From: tip-bot2 for Jiri Olsa @ 2020-01-10 17:53 UTC (permalink / raw) To: linux-tip-commits Cc: Jiri Olsa, Arnaldo Carvalho de Melo, Alexander Shishkin, Michael Petlan, Namhyung Kim, Peter Zijlstra, x86, LKML The following commit has been merged into the perf/core branch of tip: Commit-ID: 3ce311afb5583cf3d3b7f54ab088949da28aea05 Gitweb: https://git.kernel.org/tip/3ce311afb5583cf3d3b7f54ab088949da28aea05 Author: Jiri Olsa <jolsa@kernel.org> AuthorDate: Fri, 06 Dec 2019 22:06:11 +01:00 Committer: Arnaldo Carvalho de Melo <acme@redhat.com> CommitterDate: Mon, 06 Jan 2020 11:46:09 -03:00 libperf: Move to tools/lib/perf Move libperf from its current location under tools/perf to a separate directory under tools/lib/. Also change various paths (mainly includes) to reflect the libperf move to a separate directory and add a new directory under MANIFEST. Signed-off-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Michael Petlan <mpetlan@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lore.kernel.org/lkml/20191206210612.8676-2-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> --- tools/lib/perf/Build | 13 +- tools/lib/perf/Documentation/Makefile | 7 +- tools/lib/perf/Documentation/man/libperf.rst | 100 ++- tools/lib/perf/Documentation/tutorial/tutorial.rst | 123 ++- tools/lib/perf/Makefile | 188 ++++- tools/lib/perf/core.c | 38 +- tools/lib/perf/cpumap.c | 345 ++++++- tools/lib/perf/evlist.c | 641 ++++++++++++- tools/lib/perf/evsel.c | 301 ++++++- tools/lib/perf/include/internal/cpumap.h | 19 +- tools/lib/perf/include/internal/evlist.h | 127 ++- tools/lib/perf/include/internal/evsel.h | 63 +- tools/lib/perf/include/internal/lib.h | 12 +- tools/lib/perf/include/internal/mmap.h | 55 +- tools/lib/perf/include/internal/tests.h | 33 +- tools/lib/perf/include/internal/threadmap.h | 23 +- tools/lib/perf/include/internal/xyarray.h | 36 +- tools/lib/perf/include/perf/core.h | 25 +- tools/lib/perf/include/perf/cpumap.h | 28 +- tools/lib/perf/include/perf/event.h | 385 +++++++- tools/lib/perf/include/perf/evlist.h | 49 +- tools/lib/perf/include/perf/evsel.h | 40 +- tools/lib/perf/include/perf/mmap.h | 15 +- tools/lib/perf/include/perf/threadmap.h | 20 +- tools/lib/perf/internal.h | 23 +- tools/lib/perf/lib.c | 48 +- tools/lib/perf/libperf.map | 51 +- tools/lib/perf/libperf.pc.template | 11 +- tools/lib/perf/mmap.c | 275 +++++- tools/lib/perf/tests/Makefile | 38 +- tools/lib/perf/tests/test-cpumap.c | 31 +- tools/lib/perf/tests/test-evlist.c | 413 ++++++++- tools/lib/perf/tests/test-evsel.c | 135 +++- tools/lib/perf/tests/test-threadmap.c | 31 +- tools/lib/perf/threadmap.c | 91 ++- tools/lib/perf/xyarray.c | 33 +- tools/perf/MANIFEST | 1 +- tools/perf/Makefile.config | 2 +- tools/perf/Makefile.perf | 2 +- tools/perf/lib/Build | 13 +- tools/perf/lib/Documentation/Makefile | 7 +- tools/perf/lib/Documentation/man/libperf.rst | 100 +-- tools/perf/lib/Documentation/tutorial/tutorial.rst | 123 +-- tools/perf/lib/Makefile | 188 +---- tools/perf/lib/core.c | 38 +- tools/perf/lib/cpumap.c | 345 +------ tools/perf/lib/evlist.c | 641 +------------ tools/perf/lib/evsel.c | 301 +------ tools/perf/lib/include/internal/cpumap.h | 19 +- tools/perf/lib/include/internal/evlist.h | 127 +-- tools/perf/lib/include/internal/evsel.h | 63 +- tools/perf/lib/include/internal/lib.h | 12 +- tools/perf/lib/include/internal/mmap.h | 55 +- tools/perf/lib/include/internal/tests.h | 33 +- tools/perf/lib/include/internal/threadmap.h | 23 +- tools/perf/lib/include/internal/xyarray.h | 36 +- tools/perf/lib/include/perf/core.h | 25 +- tools/perf/lib/include/perf/cpumap.h | 28 +- tools/perf/lib/include/perf/event.h | 385 +------- tools/perf/lib/include/perf/evlist.h | 49 +- tools/perf/lib/include/perf/evsel.h | 40 +- tools/perf/lib/include/perf/mmap.h | 15 +- tools/perf/lib/include/perf/threadmap.h | 20 +- tools/perf/lib/internal.h | 23 +- tools/perf/lib/lib.c | 48 +- tools/perf/lib/libperf.map | 51 +- tools/perf/lib/libperf.pc.template | 11 +- tools/perf/lib/mmap.c | 275 +----- tools/perf/lib/tests/Makefile | 38 +- tools/perf/lib/tests/test-cpumap.c | 31 +- tools/perf/lib/tests/test-evlist.c | 413 +-------- tools/perf/lib/tests/test-evsel.c | 135 +--- tools/perf/lib/tests/test-threadmap.c | 31 +- tools/perf/lib/threadmap.c | 91 +-- tools/perf/lib/xyarray.c | 33 +- 75 files changed, 3869 insertions(+), 3868 deletions(-) create mode 100644 tools/lib/perf/Build create mode 100644 tools/lib/perf/Documentation/Makefile create mode 100644 tools/lib/perf/Documentation/man/libperf.rst create mode 100644 tools/lib/perf/Documentation/tutorial/tutorial.rst create mode 100644 tools/lib/perf/Makefile create mode 100644 tools/lib/perf/core.c create mode 100644 tools/lib/perf/cpumap.c create mode 100644 tools/lib/perf/evlist.c create mode 100644 tools/lib/perf/evsel.c create mode 100644 tools/lib/perf/include/internal/cpumap.h create mode 100644 tools/lib/perf/include/internal/evlist.h create mode 100644 tools/lib/perf/include/internal/evsel.h create mode 100644 tools/lib/perf/include/internal/lib.h create mode 100644 tools/lib/perf/include/internal/mmap.h create mode 100644 tools/lib/perf/include/internal/tests.h create mode 100644 tools/lib/perf/include/internal/threadmap.h create mode 100644 tools/lib/perf/include/internal/xyarray.h create mode 100644 tools/lib/perf/include/perf/core.h create mode 100644 tools/lib/perf/include/perf/cpumap.h create mode 100644 tools/lib/perf/include/perf/event.h create mode 100644 tools/lib/perf/include/perf/evlist.h create mode 100644 tools/lib/perf/include/perf/evsel.h create mode 100644 tools/lib/perf/include/perf/mmap.h create mode 100644 tools/lib/perf/include/perf/threadmap.h create mode 100644 tools/lib/perf/internal.h create mode 100644 tools/lib/perf/lib.c create mode 100644 tools/lib/perf/libperf.map create mode 100644 tools/lib/perf/libperf.pc.template create mode 100644 tools/lib/perf/mmap.c create mode 100644 tools/lib/perf/tests/Makefile create mode 100644 tools/lib/perf/tests/test-cpumap.c create mode 100644 tools/lib/perf/tests/test-evlist.c create mode 100644 tools/lib/perf/tests/test-evsel.c create mode 100644 tools/lib/perf/tests/test-threadmap.c create mode 100644 tools/lib/perf/threadmap.c create mode 100644 tools/lib/perf/xyarray.c delete mode 100644 tools/perf/lib/Build delete mode 100644 tools/perf/lib/Documentation/Makefile delete mode 100644 tools/perf/lib/Documentation/man/libperf.rst delete mode 100644 tools/perf/lib/Documentation/tutorial/tutorial.rst delete mode 100644 tools/perf/lib/Makefile delete mode 100644 tools/perf/lib/core.c delete mode 100644 tools/perf/lib/cpumap.c delete mode 100644 tools/perf/lib/evlist.c delete mode 100644 tools/perf/lib/evsel.c delete mode 100644 tools/perf/lib/include/internal/cpumap.h delete mode 100644 tools/perf/lib/include/internal/evlist.h delete mode 100644 tools/perf/lib/include/internal/evsel.h delete mode 100644 tools/perf/lib/include/internal/lib.h delete mode 100644 tools/perf/lib/include/internal/mmap.h delete mode 100644 tools/perf/lib/include/internal/tests.h delete mode 100644 tools/perf/lib/include/internal/threadmap.h delete mode 100644 tools/perf/lib/include/internal/xyarray.h delete mode 100644 tools/perf/lib/include/perf/core.h delete mode 100644 tools/perf/lib/include/perf/cpumap.h delete mode 100644 tools/perf/lib/include/perf/event.h delete mode 100644 tools/perf/lib/include/perf/evlist.h delete mode 100644 tools/perf/lib/include/perf/evsel.h delete mode 100644 tools/perf/lib/include/perf/mmap.h delete mode 100644 tools/perf/lib/include/perf/threadmap.h delete mode 100644 tools/perf/lib/internal.h delete mode 100644 tools/perf/lib/lib.c delete mode 100644 tools/perf/lib/libperf.map delete mode 100644 tools/perf/lib/libperf.pc.template delete mode 100644 tools/perf/lib/mmap.c delete mode 100644 tools/perf/lib/tests/Makefile delete mode 100644 tools/perf/lib/tests/test-cpumap.c delete mode 100644 tools/perf/lib/tests/test-evlist.c delete mode 100644 tools/perf/lib/tests/test-evsel.c delete mode 100644 tools/perf/lib/tests/test-threadmap.c delete mode 100644 tools/perf/lib/threadmap.c delete mode 100644 tools/perf/lib/xyarray.c diff --git a/tools/lib/perf/Build b/tools/lib/perf/Build new file mode 100644 index 0000000..2ef9a4e --- /dev/null +++ b/tools/lib/perf/Build @@ -0,0 +1,13 @@ +libperf-y += core.o +libperf-y += cpumap.o +libperf-y += threadmap.o +libperf-y += evsel.o +libperf-y += evlist.o +libperf-y += mmap.o +libperf-y += zalloc.o +libperf-y += xyarray.o +libperf-y += lib.o + +$(OUTPUT)zalloc.o: ../../lib/zalloc.c FORCE + $(call rule_mkdir) + $(call if_changed_dep,cc_o_c) diff --git a/tools/lib/perf/Documentation/Makefile b/tools/lib/perf/Documentation/Makefile new file mode 100644 index 0000000..586425a --- /dev/null +++ b/tools/lib/perf/Documentation/Makefile @@ -0,0 +1,7 @@ +all: + rst2man man/libperf.rst > man/libperf.7 + rst2pdf tutorial/tutorial.rst + +clean: + rm -f man/libperf.7 + rm -f tutorial/tutorial.pdf diff --git a/tools/lib/perf/Documentation/man/libperf.rst b/tools/lib/perf/Documentation/man/libperf.rst new file mode 100644 index 0000000..09a270f --- /dev/null +++ b/tools/lib/perf/Documentation/man/libperf.rst @@ -0,0 +1,100 @@ +.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) + +libperf + +The libperf library provides an API to access the linux kernel perf +events subsystem. It provides the following high level objects: + + - struct perf_cpu_map + - struct perf_thread_map + - struct perf_evlist + - struct perf_evsel + +reference +========= +Function reference by header files: + +perf/core.h +----------- +.. code-block:: c + + typedef int (\*libperf_print_fn_t)(enum libperf_print_level level, + const char \*, va_list ap); + + void libperf_set_print(libperf_print_fn_t fn); + +perf/cpumap.h +------------- +.. code-block:: c + + struct perf_cpu_map \*perf_cpu_map__dummy_new(void); + struct perf_cpu_map \*perf_cpu_map__new(const char \*cpu_list); + struct perf_cpu_map \*perf_cpu_map__read(FILE \*file); + struct perf_cpu_map \*perf_cpu_map__get(struct perf_cpu_map \*map); + void perf_cpu_map__put(struct perf_cpu_map \*map); + int perf_cpu_map__cpu(const struct perf_cpu_map \*cpus, int idx); + int perf_cpu_map__nr(const struct perf_cpu_map \*cpus); + perf_cpu_map__for_each_cpu(cpu, idx, cpus) + +perf/threadmap.h +---------------- +.. code-block:: c + + struct perf_thread_map \*perf_thread_map__new_dummy(void); + void perf_thread_map__set_pid(struct perf_thread_map \*map, int thread, pid_t pid); + char \*perf_thread_map__comm(struct perf_thread_map \*map, int thread); + struct perf_thread_map \*perf_thread_map__get(struct perf_thread_map \*map); + void perf_thread_map__put(struct perf_thread_map \*map); + +perf/evlist.h +------------- +.. code-block:: + + void perf_evlist__init(struct perf_evlist \*evlist); + void perf_evlist__add(struct perf_evlist \*evlist, + struct perf_evsel \*evsel); + void perf_evlist__remove(struct perf_evlist \*evlist, + struct perf_evsel \*evsel); + struct perf_evlist \*perf_evlist__new(void); + void perf_evlist__delete(struct perf_evlist \*evlist); + struct perf_evsel\* perf_evlist__next(struct perf_evlist \*evlist, + struct perf_evsel \*evsel); + int perf_evlist__open(struct perf_evlist \*evlist); + void perf_evlist__close(struct perf_evlist \*evlist); + void perf_evlist__enable(struct perf_evlist \*evlist); + void perf_evlist__disable(struct perf_evlist \*evlist); + perf_evlist__for_each_evsel(evlist, pos) + void perf_evlist__set_maps(struct perf_evlist \*evlist, + struct perf_cpu_map \*cpus, + struct perf_thread_map \*threads); + +perf/evsel.h +------------ +.. code-block:: c + + struct perf_counts_values { + union { + struct { + uint64_t val; + uint64_t ena; + uint64_t run; + }; + uint64_t values[3]; + }; + }; + + void perf_evsel__init(struct perf_evsel \*evsel, + struct perf_event_attr \*attr); + struct perf_evsel \*perf_evsel__new(struct perf_event_attr \*attr); + void perf_evsel__delete(struct perf_evsel \*evsel); + int perf_evsel__open(struct perf_evsel \*evsel, struct perf_cpu_map \*cpus, + struct perf_thread_map \*threads); + void perf_evsel__close(struct perf_evsel \*evsel); + int perf_evsel__read(struct perf_evsel \*evsel, int cpu, int thread, + struct perf_counts_values \*count); + int perf_evsel__enable(struct perf_evsel \*evsel); + int perf_evsel__disable(struct perf_evsel \*evsel); + int perf_evsel__apply_filter(struct perf_evsel \*evsel, const char \*filter); + struct perf_cpu_map \*perf_evsel__cpus(struct perf_evsel \*evsel); + struct perf_thread_map \*perf_evsel__threads(struct perf_evsel \*evsel); + struct perf_event_attr \*perf_evsel__attr(struct perf_evsel \*evsel); diff --git a/tools/lib/perf/Documentation/tutorial/tutorial.rst b/tools/lib/perf/Documentation/tutorial/tutorial.rst new file mode 100644 index 0000000..7be7bc2 --- /dev/null +++ b/tools/lib/perf/Documentation/tutorial/tutorial.rst @@ -0,0 +1,123 @@ +.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) + +libperf tutorial +================ + +Compile and install libperf from kernel sources +=============================================== +.. code-block:: bash + + git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git + cd linux/tools/perf/lib + make + sudo make install prefix=/usr + +Libperf object +============== +The libperf library provides several high level objects: + +struct perf_cpu_map + Provides a cpu list abstraction. + +struct perf_thread_map + Provides a thread list abstraction. + +struct perf_evsel + Provides an abstraction for single a perf event. + +struct perf_evlist + Gathers several struct perf_evsel object and performs functions on all of them. + +The exported API binds these objects together, +for full reference see the libperf.7 man page. + +Examples +======== +Examples aim to explain libperf functionality on simple use cases. +They are based in on a checked out linux kernel git tree: + +.. code-block:: bash + + $ cd tools/perf/lib/Documentation/tutorial/ + $ ls -d ex-* + ex-1-compile ex-2-evsel-stat ex-3-evlist-stat + +ex-1-compile example +==================== +This example shows the basic usage of *struct perf_cpu_map*, +how to create it and display its cpus: + +.. code-block:: bash + + $ cd ex-1-compile/ + $ make + gcc -o test test.c -lperf + $ ./test + 0 1 2 3 4 5 6 7 + + +The full code listing is here: + +.. code-block:: c + + 1 #include <perf/cpumap.h> + 2 + 3 int main(int argc, char **Argv) + 4 { + 5 struct perf_cpu_map *cpus; + 6 int cpu, tmp; + 7 + 8 cpus = perf_cpu_map__new(NULL); + 9 + 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) + 11 fprintf(stdout, "%d ", cpu); + 12 + 13 fprintf(stdout, "\n"); + 14 + 15 perf_cpu_map__put(cpus); + 16 return 0; + 17 } + + +First you need to include the proper header to have *struct perf_cpumap* +declaration and functions: + +.. code-block:: c + + 1 #include <perf/cpumap.h> + + +The *struct perf_cpumap* object is created by *perf_cpu_map__new* call. +The *NULL* argument asks it to populate the object with the current online CPUs list: + +.. code-block:: c + + 8 cpus = perf_cpu_map__new(NULL); + +This is paired with a *perf_cpu_map__put*, that drops its reference at the end, possibly deleting it. + +.. code-block:: c + + 15 perf_cpu_map__put(cpus); + +The iteration through the *struct perf_cpumap* CPUs is done using the *perf_cpu_map__for_each_cpu* +macro which requires 3 arguments: + +- cpu - the cpu numer +- tmp - iteration helper variable +- cpus - the *struct perf_cpumap* object + +.. code-block:: c + + 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) + 11 fprintf(stdout, "%d ", cpu); + +ex-2-evsel-stat example +======================= + +TBD + +ex-3-evlist-stat example +======================== + +TBD diff --git a/tools/lib/perf/Makefile b/tools/lib/perf/Makefile new file mode 100644 index 0000000..768dd42 --- /dev/null +++ b/tools/lib/perf/Makefile @@ -0,0 +1,188 @@ +# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) +# Most of this file is copied from tools/lib/bpf/Makefile + +LIBPERF_VERSION = 0 +LIBPERF_PATCHLEVEL = 0 +LIBPERF_EXTRAVERSION = 1 + +MAKEFLAGS += --no-print-directory + +ifeq ($(srctree),) +srctree := $(patsubst %/,%,$(dir $(CURDIR))) +srctree := $(patsubst %/,%,$(dir $(srctree))) +srctree := $(patsubst %/,%,$(dir $(srctree))) +#$(info Determined 'srctree' to be $(srctree)) +endif + +INSTALL = install + +# Use DESTDIR for installing into a different root directory. +# This is useful for building a package. The program will be +# installed in this directory as if it was the root directory. +# Then the build tool can move it later. +DESTDIR ?= +DESTDIR_SQ = '$(subst ','\'',$(DESTDIR))' + +include $(srctree)/tools/scripts/Makefile.include +include $(srctree)/tools/scripts/Makefile.arch + +ifeq ($(LP64), 1) + libdir_relative = lib64 +else + libdir_relative = lib +endif + +prefix ?= +libdir = $(prefix)/$(libdir_relative) + +# Shell quotes +libdir_SQ = $(subst ','\'',$(libdir)) +libdir_relative_SQ = $(subst ','\'',$(libdir_relative)) + +ifeq ("$(origin V)", "command line") + VERBOSE = $(V) +endif +ifndef VERBOSE + VERBOSE = 0 +endif + +ifeq ($(VERBOSE),1) + Q = +else + Q = @ +endif + +# Set compile option CFLAGS +ifdef EXTRA_CFLAGS + CFLAGS := $(EXTRA_CFLAGS) +else + CFLAGS := -g -Wall +endif + +INCLUDES = \ +-I$(srctree)/tools/lib/perf/include \ +-I$(srctree)/tools/lib/ \ +-I$(srctree)/tools/include \ +-I$(srctree)/tools/arch/$(SRCARCH)/include/ \ +-I$(srctree)/tools/arch/$(SRCARCH)/include/uapi \ +-I$(srctree)/tools/include/uapi + +# Append required CFLAGS +override CFLAGS += $(EXTRA_WARNINGS) +override CFLAGS += -Werror -Wall +override CFLAGS += -fPIC +override CFLAGS += $(INCLUDES) +override CFLAGS += -fvisibility=hidden + +all: + +export srctree OUTPUT CC LD CFLAGS V +export DESTDIR DESTDIR_SQ + +include $(srctree)/tools/build/Makefile.include + +VERSION_SCRIPT := libperf.map + +PATCHLEVEL = $(LIBPERF_PATCHLEVEL) +EXTRAVERSION = $(LIBPERF_EXTRAVERSION) +VERSION = $(LIBPERF_VERSION).$(LIBPERF_PATCHLEVEL).$(LIBPERF_EXTRAVERSION) + +LIBPERF_SO := $(OUTPUT)libperf.so.$(VERSION) +LIBPERF_A := $(OUTPUT)libperf.a +LIBPERF_IN := $(OUTPUT)libperf-in.o +LIBPERF_PC := $(OUTPUT)libperf.pc + +LIBPERF_ALL := $(LIBPERF_A) $(OUTPUT)libperf.so* + +LIB_DIR := $(srctree)/tools/lib/api/ + +ifneq ($(OUTPUT),) +ifneq ($(subdir),) + API_PATH=$(OUTPUT)/../lib/api/ +else + API_PATH=$(OUTPUT) +endif +else + API_PATH=$(LIB_DIR) +endif + +LIBAPI = $(API_PATH)libapi.a +export LIBAPI + +$(LIBAPI): FORCE + $(Q)$(MAKE) -C $(LIB_DIR) O=$(OUTPUT) $(OUTPUT)libapi.a + +$(LIBAPI)-clean: + $(call QUIET_CLEAN, libapi) + $(Q)$(MAKE) -C $(LIB_DIR) O=$(OUTPUT) clean >/dev/null + +$(LIBPERF_IN): FORCE + $(Q)$(MAKE) $(build)=libperf + +$(LIBPERF_A): $(LIBPERF_IN) + $(QUIET_AR)$(RM) $@ && $(AR) rcs $@ $(LIBPERF_IN) + +$(LIBPERF_SO): $(LIBPERF_IN) $(LIBAPI) + $(QUIET_LINK)$(CC) --shared -Wl,-soname,libperf.so \ + -Wl,--version-script=$(VERSION_SCRIPT) $^ -o $@ + @ln -sf $(@F) $(OUTPUT)libperf.so + @ln -sf $(@F) $(OUTPUT)libperf.so.$(LIBPERF_VERSION) + + +libs: $(LIBPERF_A) $(LIBPERF_SO) $(LIBPERF_PC) + +all: fixdep + $(Q)$(MAKE) libs + +clean: $(LIBAPI)-clean + $(call QUIET_CLEAN, libperf) $(RM) $(LIBPERF_A) \ + *.o *~ *.a *.so *.so.$(VERSION) *.so.$(LIBPERF_VERSION) .*.d .*.cmd LIBPERF-CFLAGS $(LIBPERF_PC) + $(Q)$(MAKE) -C tests clean + +tests: libs + $(Q)$(MAKE) -C tests + $(Q)$(MAKE) -C tests run + +$(LIBPERF_PC): + $(QUIET_GEN)sed -e "s|@PREFIX@|$(prefix)|" \ + -e "s|@LIBDIR@|$(libdir_SQ)|" \ + -e "s|@VERSION@|$(VERSION)|" \ + < libperf.pc.template > $@ + +define do_install_mkdir + if [ ! -d '$(DESTDIR_SQ)$1' ]; then \ + $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$1'; \ + fi +endef + +define do_install + if [ ! -d '$(DESTDIR_SQ)$2' ]; then \ + $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$2'; \ + fi; \ + $(INSTALL) $1 $(if $3,-m $3,) '$(DESTDIR_SQ)$2' +endef + +install_lib: libs + $(call QUIET_INSTALL, $(LIBPERF_ALL)) \ + $(call do_install_mkdir,$(libdir_SQ)); \ + cp -fpR $(LIBPERF_ALL) $(DESTDIR)$(libdir_SQ) + +install_headers: + $(call QUIET_INSTALL, headers) \ + $(call do_install,include/perf/core.h,$(prefix)/include/perf,644); \ + $(call do_install,include/perf/cpumap.h,$(prefix)/include/perf,644); \ + $(call do_install,include/perf/threadmap.h,$(prefix)/include/perf,644); \ + $(call do_install,include/perf/evlist.h,$(prefix)/include/perf,644); \ + $(call do_install,include/perf/evsel.h,$(prefix)/include/perf,644); \ + $(call do_install,include/perf/event.h,$(prefix)/include/perf,644); \ + $(call do_install,include/perf/mmap.h,$(prefix)/include/perf,644); + +install_pkgconfig: $(LIBPERF_PC) + $(call QUIET_INSTALL, $(LIBPERF_PC)) \ + $(call do_install,$(LIBPERF_PC),$(libdir_SQ)/pkgconfig,644) + +install: install_lib install_headers install_pkgconfig + +FORCE: + +.PHONY: all install clean tests FORCE diff --git a/tools/lib/perf/core.c b/tools/lib/perf/core.c new file mode 100644 index 0000000..58fc894 --- /dev/null +++ b/tools/lib/perf/core.c @@ -0,0 +1,38 @@ +// SPDX-License-Identifier: GPL-2.0-only + +#define __printf(a, b) __attribute__((format(printf, a, b))) + +#include <stdio.h> +#include <stdarg.h> +#include <unistd.h> +#include <linux/compiler.h> +#include <perf/core.h> +#include <internal/lib.h> +#include "internal.h" + +static int __base_pr(enum libperf_print_level level __maybe_unused, const char *format, + va_list args) +{ + return vfprintf(stderr, format, args); +} + +static libperf_print_fn_t __libperf_pr = __base_pr; + +__printf(2, 3) +void libperf_print(enum libperf_print_level level, const char *format, ...) +{ + va_list args; + + if (!__libperf_pr) + return; + + va_start(args, format); + __libperf_pr(level, format, args); + va_end(args); +} + +void libperf_init(libperf_print_fn_t fn) +{ + page_size = sysconf(_SC_PAGE_SIZE); + __libperf_pr = fn; +} diff --git a/tools/lib/perf/cpumap.c b/tools/lib/perf/cpumap.c new file mode 100644 index 0000000..f93f4e7 --- /dev/null +++ b/tools/lib/perf/cpumap.c @@ -0,0 +1,345 @@ +// SPDX-License-Identifier: GPL-2.0-only +#include <perf/cpumap.h> +#include <stdlib.h> +#include <linux/refcount.h> +#include <internal/cpumap.h> +#include <asm/bug.h> +#include <stdio.h> +#include <string.h> +#include <unistd.h> +#include <ctype.h> +#include <limits.h> + +struct perf_cpu_map *perf_cpu_map__dummy_new(void) +{ + struct perf_cpu_map *cpus = malloc(sizeof(*cpus) + sizeof(int)); + + if (cpus != NULL) { + cpus->nr = 1; + cpus->map[0] = -1; + refcount_set(&cpus->refcnt, 1); + } + + return cpus; +} + +static void cpu_map__delete(struct perf_cpu_map *map) +{ + if (map) { + WARN_ONCE(refcount_read(&map->refcnt) != 0, + "cpu_map refcnt unbalanced\n"); + free(map); + } +} + +struct perf_cpu_map *perf_cpu_map__get(struct perf_cpu_map *map) +{ + if (map) + refcount_inc(&map->refcnt); + return map; +} + +void perf_cpu_map__put(struct perf_cpu_map *map) +{ + if (map && refcount_dec_and_test(&map->refcnt)) + cpu_map__delete(map); +} + +static struct perf_cpu_map *cpu_map__default_new(void) +{ + struct perf_cpu_map *cpus; + int nr_cpus; + + nr_cpus = sysconf(_SC_NPROCESSORS_ONLN); + if (nr_cpus < 0) + return NULL; + + cpus = malloc(sizeof(*cpus) + nr_cpus * sizeof(int)); + if (cpus != NULL) { + int i; + + for (i = 0; i < nr_cpus; ++i) + cpus->map[i] = i; + + cpus->nr = nr_cpus; + refcount_set(&cpus->refcnt, 1); + } + + return cpus; +} + +static int cmp_int(const void *a, const void *b) +{ + return *(const int *)a - *(const int*)b; +} + +static struct perf_cpu_map *cpu_map__trim_new(int nr_cpus, int *tmp_cpus) +{ + size_t payload_size = nr_cpus * sizeof(int); + struct perf_cpu_map *cpus = malloc(sizeof(*cpus) + payload_size); + int i, j; + + if (cpus != NULL) { + memcpy(cpus->map, tmp_cpus, payload_size); + qsort(cpus->map, nr_cpus, sizeof(int), cmp_int); + /* Remove dups */ + j = 0; + for (i = 0; i < nr_cpus; i++) { + if (i == 0 || cpus->map[i] != cpus->map[i - 1]) + cpus->map[j++] = cpus->map[i]; + } + cpus->nr = j; + assert(j <= nr_cpus); + refcount_set(&cpus->refcnt, 1); + } + + return cpus; +} + +struct perf_cpu_map *perf_cpu_map__read(FILE *file) +{ + struct perf_cpu_map *cpus = NULL; + int nr_cpus = 0; + int *tmp_cpus = NULL, *tmp; + int max_entries = 0; + int n, cpu, prev; + char sep; + + sep = 0; + prev = -1; + for (;;) { + n = fscanf(file, "%u%c", &cpu, &sep); + if (n <= 0) + break; + if (prev >= 0) { + int new_max = nr_cpus + cpu - prev - 1; + + WARN_ONCE(new_max >= MAX_NR_CPUS, "Perf can support %d CPUs. " + "Consider raising MAX_NR_CPUS\n", MAX_NR_CPUS); + + if (new_max >= max_entries) { + max_entries = new_max + MAX_NR_CPUS / 2; + tmp = realloc(tmp_cpus, max_entries * sizeof(int)); + if (tmp == NULL) + goto out_free_tmp; + tmp_cpus = tmp; + } + + while (++prev < cpu) + tmp_cpus[nr_cpus++] = prev; + } + if (nr_cpus == max_entries) { + max_entries += MAX_NR_CPUS; + tmp = realloc(tmp_cpus, max_entries * sizeof(int)); + if (tmp == NULL) + goto out_free_tmp; + tmp_cpus = tmp; + } + + tmp_cpus[nr_cpus++] = cpu; + if (n == 2 && sep == '-') + prev = cpu; + else + prev = -1; + if (n == 1 || sep == '\n') + break; + } + + if (nr_cpus > 0) + cpus = cpu_map__trim_new(nr_cpus, tmp_cpus); + else + cpus = cpu_map__default_new(); +out_free_tmp: + free(tmp_cpus); + return cpus; +} + +static struct perf_cpu_map *cpu_map__read_all_cpu_map(void) +{ + struct perf_cpu_map *cpus = NULL; + FILE *onlnf; + + onlnf = fopen("/sys/devices/system/cpu/online", "r"); + if (!onlnf) + return cpu_map__default_new(); + + cpus = perf_cpu_map__read(onlnf); + fclose(onlnf); + return cpus; +} + +struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list) +{ + struct perf_cpu_map *cpus = NULL; + unsigned long start_cpu, end_cpu = 0; + char *p = NULL; + int i, nr_cpus = 0; + int *tmp_cpus = NULL, *tmp; + int max_entries = 0; + + if (!cpu_list) + return cpu_map__read_all_cpu_map(); + + /* + * must handle the case of empty cpumap to cover + * TOPOLOGY header for NUMA nodes with no CPU + * ( e.g., because of CPU hotplug) + */ + if (!isdigit(*cpu_list) && *cpu_list != '\0') + goto out; + + while (isdigit(*cpu_list)) { + p = NULL; + start_cpu = strtoul(cpu_list, &p, 0); + if (start_cpu >= INT_MAX + || (*p != '\0' && *p != ',' && *p != '-')) + goto invalid; + + if (*p == '-') { + cpu_list = ++p; + p = NULL; + end_cpu = strtoul(cpu_list, &p, 0); + + if (end_cpu >= INT_MAX || (*p != '\0' && *p != ',')) + goto invalid; + + if (end_cpu < start_cpu) + goto invalid; + } else { + end_cpu = start_cpu; + } + + WARN_ONCE(end_cpu >= MAX_NR_CPUS, "Perf can support %d CPUs. " + "Consider raising MAX_NR_CPUS\n", MAX_NR_CPUS); + + for (; start_cpu <= end_cpu; start_cpu++) { + /* check for duplicates */ + for (i = 0; i < nr_cpus; i++) + if (tmp_cpus[i] == (int)start_cpu) + goto invalid; + + if (nr_cpus == max_entries) { + max_entries += MAX_NR_CPUS; + tmp = realloc(tmp_cpus, max_entries * sizeof(int)); + if (tmp == NULL) + goto invalid; + tmp_cpus = tmp; + } + tmp_cpus[nr_cpus++] = (int)start_cpu; + } + if (*p) + ++p; + + cpu_list = p; + } + + if (nr_cpus > 0) + cpus = cpu_map__trim_new(nr_cpus, tmp_cpus); + else if (*cpu_list != '\0') + cpus = cpu_map__default_new(); + else + cpus = perf_cpu_map__dummy_new(); +invalid: + free(tmp_cpus); +out: + return cpus; +} + +int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx) +{ + if (idx < cpus->nr) + return cpus->map[idx]; + + return -1; +} + +int perf_cpu_map__nr(const struct perf_cpu_map *cpus) +{ + return cpus ? cpus->nr : 1; +} + +bool perf_cpu_map__empty(const struct perf_cpu_map *map) +{ + return map ? map->map[0] == -1 : true; +} + +int perf_cpu_map__idx(struct perf_cpu_map *cpus, int cpu) +{ + int i; + + for (i = 0; i < cpus->nr; ++i) { + if (cpus->map[i] == cpu) + return i; + } + + return -1; +} + +int perf_cpu_map__max(struct perf_cpu_map *map) +{ + int i, max = -1; + + for (i = 0; i < map->nr; i++) { + if (map->map[i] > max) + max = map->map[i]; + } + + return max; +} + +/* + * Merge two cpumaps + * + * orig either gets freed and replaced with a new map, or reused + * with no reference count change (similar to "realloc") + * other has its reference count increased. + */ + +struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig, + struct perf_cpu_map *other) +{ + int *tmp_cpus; + int tmp_len; + int i, j, k; + struct perf_cpu_map *merged; + + if (!orig && !other) + return NULL; + if (!orig) { + perf_cpu_map__get(other); + return other; + } + if (!other) + return orig; + if (orig->nr == other->nr && + !memcmp(orig->map, other->map, orig->nr * sizeof(int))) + return orig; + + tmp_len = orig->nr + other->nr; + tmp_cpus = malloc(tmp_len * sizeof(int)); + if (!tmp_cpus) + return NULL; + + /* Standard merge algorithm from wikipedia */ + i = j = k = 0; + while (i < orig->nr && j < other->nr) { + if (orig->map[i] <= other->map[j]) { + if (orig->map[i] == other->map[j]) + j++; + tmp_cpus[k++] = orig->map[i++]; + } else + tmp_cpus[k++] = other->map[j++]; + } + + while (i < orig->nr) + tmp_cpus[k++] = orig->map[i++]; + + while (j < other->nr) + tmp_cpus[k++] = other->map[j++]; + assert(k <= tmp_len); + + merged = cpu_map__trim_new(k, tmp_cpus); + free(tmp_cpus); + perf_cpu_map__put(orig); + return merged; +} diff --git a/tools/lib/perf/evlist.c b/tools/lib/perf/evlist.c new file mode 100644 index 0000000..ae9e65a --- /dev/null +++ b/tools/lib/perf/evlist.c @@ -0,0 +1,641 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <perf/evlist.h> +#include <perf/evsel.h> +#include <linux/bitops.h> +#include <linux/list.h> +#include <linux/hash.h> +#include <sys/ioctl.h> +#include <internal/evlist.h> +#include <internal/evsel.h> +#include <internal/xyarray.h> +#include <internal/mmap.h> +#include <internal/cpumap.h> +#include <internal/threadmap.h> +#include <internal/xyarray.h> +#include <internal/lib.h> +#include <linux/zalloc.h> +#include <sys/ioctl.h> +#include <stdlib.h> +#include <errno.h> +#include <unistd.h> +#include <fcntl.h> +#include <signal.h> +#include <poll.h> +#include <sys/mman.h> +#include <perf/cpumap.h> +#include <perf/threadmap.h> +#include <api/fd/array.h> + +void perf_evlist__init(struct perf_evlist *evlist) +{ + int i; + + for (i = 0; i < PERF_EVLIST__HLIST_SIZE; ++i) + INIT_HLIST_HEAD(&evlist->heads[i]); + INIT_LIST_HEAD(&evlist->entries); + evlist->nr_entries = 0; + fdarray__init(&evlist->pollfd, 64); +} + +static void __perf_evlist__propagate_maps(struct perf_evlist *evlist, + struct perf_evsel *evsel) +{ + /* + * We already have cpus for evsel (via PMU sysfs) so + * keep it, if there's no target cpu list defined. + */ + if (!evsel->own_cpus || evlist->has_user_cpus) { + perf_cpu_map__put(evsel->cpus); + evsel->cpus = perf_cpu_map__get(evlist->cpus); + } else if (evsel->cpus != evsel->own_cpus) { + perf_cpu_map__put(evsel->cpus); + evsel->cpus = perf_cpu_map__get(evsel->own_cpus); + } + + perf_thread_map__put(evsel->threads); + evsel->threads = perf_thread_map__get(evlist->threads); + evlist->all_cpus = perf_cpu_map__merge(evlist->all_cpus, evsel->cpus); +} + +static void perf_evlist__propagate_maps(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + + perf_evlist__for_each_evsel(evlist, evsel) + __perf_evlist__propagate_maps(evlist, evsel); +} + +void perf_evlist__add(struct perf_evlist *evlist, + struct perf_evsel *evsel) +{ + list_add_tail(&evsel->node, &evlist->entries); + evlist->nr_entries += 1; + __perf_evlist__propagate_maps(evlist, evsel); +} + +void perf_evlist__remove(struct perf_evlist *evlist, + struct perf_evsel *evsel) +{ + list_del_init(&evsel->node); + evlist->nr_entries -= 1; +} + +struct perf_evlist *perf_evlist__new(void) +{ + struct perf_evlist *evlist = zalloc(sizeof(*evlist)); + + if (evlist != NULL) + perf_evlist__init(evlist); + + return evlist; +} + +struct perf_evsel * +perf_evlist__next(struct perf_evlist *evlist, struct perf_evsel *prev) +{ + struct perf_evsel *next; + + if (!prev) { + next = list_first_entry(&evlist->entries, + struct perf_evsel, + node); + } else { + next = list_next_entry(prev, node); + } + + /* Empty list is noticed here so don't need checking on entry. */ + if (&next->node == &evlist->entries) + return NULL; + + return next; +} + +static void perf_evlist__purge(struct perf_evlist *evlist) +{ + struct perf_evsel *pos, *n; + + perf_evlist__for_each_entry_safe(evlist, n, pos) { + list_del_init(&pos->node); + perf_evsel__delete(pos); + } + + evlist->nr_entries = 0; +} + +void perf_evlist__exit(struct perf_evlist *evlist) +{ + perf_cpu_map__put(evlist->cpus); + perf_thread_map__put(evlist->threads); + evlist->cpus = NULL; + evlist->threads = NULL; + fdarray__exit(&evlist->pollfd); +} + +void perf_evlist__delete(struct perf_evlist *evlist) +{ + if (evlist == NULL) + return; + + perf_evlist__munmap(evlist); + perf_evlist__close(evlist); + perf_evlist__purge(evlist); + perf_evlist__exit(evlist); + free(evlist); +} + +void perf_evlist__set_maps(struct perf_evlist *evlist, + struct perf_cpu_map *cpus, + struct perf_thread_map *threads) +{ + /* + * Allow for the possibility that one or another of the maps isn't being + * changed i.e. don't put it. Note we are assuming the maps that are + * being applied are brand new and evlist is taking ownership of the + * original reference count of 1. If that is not the case it is up to + * the caller to increase the reference count. + */ + if (cpus != evlist->cpus) { + perf_cpu_map__put(evlist->cpus); + evlist->cpus = perf_cpu_map__get(cpus); + } + + if (threads != evlist->threads) { + perf_thread_map__put(evlist->threads); + evlist->threads = perf_thread_map__get(threads); + } + + perf_evlist__propagate_maps(evlist); +} + +int perf_evlist__open(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + int err; + + perf_evlist__for_each_entry(evlist, evsel) { + err = perf_evsel__open(evsel, evsel->cpus, evsel->threads); + if (err < 0) + goto out_err; + } + + return 0; + +out_err: + perf_evlist__close(evlist); + return err; +} + +void perf_evlist__close(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + + perf_evlist__for_each_entry_reverse(evlist, evsel) + perf_evsel__close(evsel); +} + +void perf_evlist__enable(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + + perf_evlist__for_each_entry(evlist, evsel) + perf_evsel__enable(evsel); +} + +void perf_evlist__disable(struct perf_evlist *evlist) +{ + struct perf_evsel *evsel; + + perf_evlist__for_each_entry(evlist, evsel) + perf_evsel__disable(evsel); +} + +u64 perf_evlist__read_format(struct perf_evlist *evlist) +{ + struct perf_evsel *first = perf_evlist__first(evlist); + + return first->attr.read_format; +} + +#define SID(e, x, y) xyarray__entry(e->sample_id, x, y) + +static void perf_evlist__id_hash(struct perf_evlist *evlist, + struct perf_evsel *evsel, + int cpu, int thread, u64 id) +{ + int hash; + struct perf_sample_id *sid = SID(evsel, cpu, thread); + + sid->id = id; + sid->evsel = evsel; + hash = hash_64(sid->id, PERF_EVLIST__HLIST_BITS); + hlist_add_head(&sid->node, &evlist->heads[hash]); +} + +void perf_evlist__id_add(struct perf_evlist *evlist, + struct perf_evsel *evsel, + int cpu, int thread, u64 id) +{ + perf_evlist__id_hash(evlist, evsel, cpu, thread, id); + evsel->id[evsel->ids++] = id; +} + +int perf_evlist__id_add_fd(struct perf_evlist *evlist, + struct perf_evsel *evsel, + int cpu, int thread, int fd) +{ + u64 read_data[4] = { 0, }; + int id_idx = 1; /* The first entry is the counter value */ + u64 id; + int ret; + + ret = ioctl(fd, PERF_EVENT_IOC_ID, &id); + if (!ret) + goto add; + + if (errno != ENOTTY) + return -1; + + /* Legacy way to get event id.. All hail to old kernels! */ + + /* + * This way does not work with group format read, so bail + * out in that case. + */ + if (perf_evlist__read_format(evlist) & PERF_FORMAT_GROUP) + return -1; + + if (!(evsel->attr.read_format & PERF_FORMAT_ID) || + read(fd, &read_data, sizeof(read_data)) == -1) + return -1; + + if (evsel->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) + ++id_idx; + if (evsel->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) + ++id_idx; + + id = read_data[id_idx]; + +add: + perf_evlist__id_add(evlist, evsel, cpu, thread, id); + return 0; +} + +int perf_evlist__alloc_pollfd(struct perf_evlist *evlist) +{ + int nr_cpus = perf_cpu_map__nr(evlist->cpus); + int nr_threads = perf_thread_map__nr(evlist->threads); + int nfds = 0; + struct perf_evsel *evsel; + + perf_evlist__for_each_entry(evlist, evsel) { + if (evsel->system_wide) + nfds += nr_cpus; + else + nfds += nr_cpus * nr_threads; + } + + if (fdarray__available_entries(&evlist->pollfd) < nfds && + fdarray__grow(&evlist->pollfd, nfds) < 0) + return -ENOMEM; + + return 0; +} + +int perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd, + void *ptr, short revent) +{ + int pos = fdarray__add(&evlist->pollfd, fd, revent | POLLERR | POLLHUP); + + if (pos >= 0) { + evlist->pollfd.priv[pos].ptr = ptr; + fcntl(fd, F_SETFL, O_NONBLOCK); + } + + return pos; +} + +static void perf_evlist__munmap_filtered(struct fdarray *fda, int fd, + void *arg __maybe_unused) +{ + struct perf_mmap *map = fda->priv[fd].ptr; + + if (map) + perf_mmap__put(map); +} + +int perf_evlist__filter_pollfd(struct perf_evlist *evlist, short revents_and_mask) +{ + return fdarray__filter(&evlist->pollfd, revents_and_mask, + perf_evlist__munmap_filtered, NULL); +} + +int perf_evlist__poll(struct perf_evlist *evlist, int timeout) +{ + return fdarray__poll(&evlist->pollfd, timeout); +} + +static struct perf_mmap* perf_evlist__alloc_mmap(struct perf_evlist *evlist, bool overwrite) +{ + int i; + struct perf_mmap *map; + + map = zalloc(evlist->nr_mmaps * sizeof(struct perf_mmap)); + if (!map) + return NULL; + + for (i = 0; i < evlist->nr_mmaps; i++) { + struct perf_mmap *prev = i ? &map[i - 1] : NULL; + + /* + * When the perf_mmap() call is made we grab one refcount, plus + * one extra to let perf_mmap__consume() get the last + * events after all real references (perf_mmap__get()) are + * dropped. + * + * Each PERF_EVENT_IOC_SET_OUTPUT points to this mmap and + * thus does perf_mmap__get() on it. + */ + perf_mmap__init(&map[i], prev, overwrite, NULL); + } + + return map; +} + +static void perf_evlist__set_sid_idx(struct perf_evlist *evlist, + struct perf_evsel *evsel, int idx, int cpu, + int thread) +{ + struct perf_sample_id *sid = SID(evsel, cpu, thread); + + sid->idx = idx; + if (evlist->cpus && cpu >= 0) + sid->cpu = evlist->cpus->map[cpu]; + else + sid->cpu = -1; + if (!evsel->system_wide && evlist->threads && thread >= 0) + sid->tid = perf_thread_map__pid(evlist->threads, thread); + else + sid->tid = -1; +} + +static struct perf_mmap* +perf_evlist__mmap_cb_get(struct perf_evlist *evlist, bool overwrite, int idx) +{ + struct perf_mmap *maps; + + maps = overwrite ? evlist->mmap_ovw : evlist->mmap; + + if (!maps) { + maps = perf_evlist__alloc_mmap(evlist, overwrite); + if (!maps) + return NULL; + + if (overwrite) + evlist->mmap_ovw = maps; + else + evlist->mmap = maps; + } + + return &maps[idx]; +} + +#define FD(e, x, y) (*(int *) xyarray__entry(e->fd, x, y)) + +static int +perf_evlist__mmap_cb_mmap(struct perf_mmap *map, struct perf_mmap_param *mp, + int output, int cpu) +{ + return perf_mmap__mmap(map, mp, output, cpu); +} + +static void perf_evlist__set_mmap_first(struct perf_evlist *evlist, struct perf_mmap *map, + bool overwrite) +{ + if (overwrite) + evlist->mmap_ovw_first = map; + else + evlist->mmap_first = map; +} + +static int +mmap_per_evsel(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, + int idx, struct perf_mmap_param *mp, int cpu_idx, + int thread, int *_output, int *_output_overwrite) +{ + int evlist_cpu = perf_cpu_map__cpu(evlist->cpus, cpu_idx); + struct perf_evsel *evsel; + int revent; + + perf_evlist__for_each_entry(evlist, evsel) { + bool overwrite = evsel->attr.write_backward; + struct perf_mmap *map; + int *output, fd, cpu; + + if (evsel->system_wide && thread) + continue; + + cpu = perf_cpu_map__idx(evsel->cpus, evlist_cpu); + if (cpu == -1) + continue; + + map = ops->get(evlist, overwrite, idx); + if (map == NULL) + return -ENOMEM; + + if (overwrite) { + mp->prot = PROT_READ; + output = _output_overwrite; + } else { + mp->prot = PROT_READ | PROT_WRITE; + output = _output; + } + + fd = FD(evsel, cpu, thread); + + if (*output == -1) { + *output = fd; + + /* + * The last one will be done at perf_mmap__consume(), so that we + * make sure we don't prevent tools from consuming every last event in + * the ring buffer. + * + * I.e. we can get the POLLHUP meaning that the fd doesn't exist + * anymore, but the last events for it are still in the ring buffer, + * waiting to be consumed. + * + * Tools can chose to ignore this at their own discretion, but the + * evlist layer can't just drop it when filtering events in + * perf_evlist__filter_pollfd(). + */ + refcount_set(&map->refcnt, 2); + + if (ops->mmap(map, mp, *output, evlist_cpu) < 0) + return -1; + + if (!idx) + perf_evlist__set_mmap_first(evlist, map, overwrite); + } else { + if (ioctl(fd, PERF_EVENT_IOC_SET_OUTPUT, *output) != 0) + return -1; + + perf_mmap__get(map); + } + + revent = !overwrite ? POLLIN : 0; + + if (!evsel->system_wide && + perf_evlist__add_pollfd(evlist, fd, map, revent) < 0) { + perf_mmap__put(map); + return -1; + } + + if (evsel->attr.read_format & PERF_FORMAT_ID) { + if (perf_evlist__id_add_fd(evlist, evsel, cpu, thread, + fd) < 0) + return -1; + perf_evlist__set_sid_idx(evlist, evsel, idx, cpu, + thread); + } + } + + return 0; +} + +static int +mmap_per_thread(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, + struct perf_mmap_param *mp) +{ + int thread; + int nr_threads = perf_thread_map__nr(evlist->threads); + + for (thread = 0; thread < nr_threads; thread++) { + int output = -1; + int output_overwrite = -1; + + if (ops->idx) + ops->idx(evlist, mp, thread, false); + + if (mmap_per_evsel(evlist, ops, thread, mp, 0, thread, + &output, &output_overwrite)) + goto out_unmap; + } + + return 0; + +out_unmap: + perf_evlist__munmap(evlist); + return -1; +} + +static int +mmap_per_cpu(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, + struct perf_mmap_param *mp) +{ + int nr_threads = perf_thread_map__nr(evlist->threads); + int nr_cpus = perf_cpu_map__nr(evlist->cpus); + int cpu, thread; + + for (cpu = 0; cpu < nr_cpus; cpu++) { + int output = -1; + int output_overwrite = -1; + + if (ops->idx) + ops->idx(evlist, mp, cpu, true); + + for (thread = 0; thread < nr_threads; thread++) { + if (mmap_per_evsel(evlist, ops, cpu, mp, cpu, + thread, &output, &output_overwrite)) + goto out_unmap; + } + } + + return 0; + +out_unmap: + perf_evlist__munmap(evlist); + return -1; +} + +static int perf_evlist__nr_mmaps(struct perf_evlist *evlist) +{ + int nr_mmaps; + + nr_mmaps = perf_cpu_map__nr(evlist->cpus); + if (perf_cpu_map__empty(evlist->cpus)) + nr_mmaps = perf_thread_map__nr(evlist->threads); + + return nr_mmaps; +} + +int perf_evlist__mmap_ops(struct perf_evlist *evlist, + struct perf_evlist_mmap_ops *ops, + struct perf_mmap_param *mp) +{ + struct perf_evsel *evsel; + const struct perf_cpu_map *cpus = evlist->cpus; + const struct perf_thread_map *threads = evlist->threads; + + if (!ops || !ops->get || !ops->mmap) + return -EINVAL; + + mp->mask = evlist->mmap_len - page_size - 1; + + evlist->nr_mmaps = perf_evlist__nr_mmaps(evlist); + + perf_evlist__for_each_entry(evlist, evsel) { + if ((evsel->attr.read_format & PERF_FORMAT_ID) && + evsel->sample_id == NULL && + perf_evsel__alloc_id(evsel, perf_cpu_map__nr(cpus), threads->nr) < 0) + return -ENOMEM; + } + + if (evlist->pollfd.entries == NULL && perf_evlist__alloc_pollfd(evlist) < 0) + return -ENOMEM; + + if (perf_cpu_map__empty(cpus)) + return mmap_per_thread(evlist, ops, mp); + + return mmap_per_cpu(evlist, ops, mp); +} + +int perf_evlist__mmap(struct perf_evlist *evlist, int pages) +{ + struct perf_mmap_param mp; + struct perf_evlist_mmap_ops ops = { + .get = perf_evlist__mmap_cb_get, + .mmap = perf_evlist__mmap_cb_mmap, + }; + + evlist->mmap_len = (pages + 1) * page_size; + + return perf_evlist__mmap_ops(evlist, &ops, &mp); +} + +void perf_evlist__munmap(struct perf_evlist *evlist) +{ + int i; + + if (evlist->mmap) { + for (i = 0; i < evlist->nr_mmaps; i++) + perf_mmap__munmap(&evlist->mmap[i]); + } + + if (evlist->mmap_ovw) { + for (i = 0; i < evlist->nr_mmaps; i++) + perf_mmap__munmap(&evlist->mmap_ovw[i]); + } + + zfree(&evlist->mmap); + zfree(&evlist->mmap_ovw); +} + +struct perf_mmap* +perf_evlist__next_mmap(struct perf_evlist *evlist, struct perf_mmap *map, + bool overwrite) +{ + if (map) + return map->next; + + return overwrite ? evlist->mmap_ovw_first : evlist->mmap_first; +} diff --git a/tools/lib/perf/evsel.c b/tools/lib/perf/evsel.c new file mode 100644 index 0000000..4dc0628 --- /dev/null +++ b/tools/lib/perf/evsel.c @@ -0,0 +1,301 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <errno.h> +#include <unistd.h> +#include <sys/syscall.h> +#include <perf/evsel.h> +#include <perf/cpumap.h> +#include <perf/threadmap.h> +#include <linux/list.h> +#include <internal/evsel.h> +#include <linux/zalloc.h> +#include <stdlib.h> +#include <internal/xyarray.h> +#include <internal/cpumap.h> +#include <internal/threadmap.h> +#include <internal/lib.h> +#include <linux/string.h> +#include <sys/ioctl.h> + +void perf_evsel__init(struct perf_evsel *evsel, struct perf_event_attr *attr) +{ + INIT_LIST_HEAD(&evsel->node); + evsel->attr = *attr; +} + +struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr) +{ + struct perf_evsel *evsel = zalloc(sizeof(*evsel)); + + if (evsel != NULL) + perf_evsel__init(evsel, attr); + + return evsel; +} + +void perf_evsel__delete(struct perf_evsel *evsel) +{ + free(evsel); +} + +#define FD(e, x, y) (*(int *) xyarray__entry(e->fd, x, y)) + +int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads) +{ + evsel->fd = xyarray__new(ncpus, nthreads, sizeof(int)); + + if (evsel->fd) { + int cpu, thread; + for (cpu = 0; cpu < ncpus; cpu++) { + for (thread = 0; thread < nthreads; thread++) { + FD(evsel, cpu, thread) = -1; + } + } + } + + return evsel->fd != NULL ? 0 : -ENOMEM; +} + +static int +sys_perf_event_open(struct perf_event_attr *attr, + pid_t pid, int cpu, int group_fd, + unsigned long flags) +{ + return syscall(__NR_perf_event_open, attr, pid, cpu, group_fd, flags); +} + +int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus, + struct perf_thread_map *threads) +{ + int cpu, thread, err = 0; + + if (cpus == NULL) { + static struct perf_cpu_map *empty_cpu_map; + + if (empty_cpu_map == NULL) { + empty_cpu_map = perf_cpu_map__dummy_new(); + if (empty_cpu_map == NULL) + return -ENOMEM; + } + + cpus = empty_cpu_map; + } + + if (threads == NULL) { + static struct perf_thread_map *empty_thread_map; + + if (empty_thread_map == NULL) { + empty_thread_map = perf_thread_map__new_dummy(); + if (empty_thread_map == NULL) + return -ENOMEM; + } + + threads = empty_thread_map; + } + + if (evsel->fd == NULL && + perf_evsel__alloc_fd(evsel, cpus->nr, threads->nr) < 0) + return -ENOMEM; + + for (cpu = 0; cpu < cpus->nr; cpu++) { + for (thread = 0; thread < threads->nr; thread++) { + int fd; + + fd = sys_perf_event_open(&evsel->attr, + threads->map[thread].pid, + cpus->map[cpu], -1, 0); + + if (fd < 0) + return -errno; + + FD(evsel, cpu, thread) = fd; + } + } + + return err; +} + +static void perf_evsel__close_fd_cpu(struct perf_evsel *evsel, int cpu) +{ + int thread; + + for (thread = 0; thread < xyarray__max_y(evsel->fd); ++thread) { + if (FD(evsel, cpu, thread) >= 0) + close(FD(evsel, cpu, thread)); + FD(evsel, cpu, thread) = -1; + } +} + +void perf_evsel__close_fd(struct perf_evsel *evsel) +{ + int cpu; + + for (cpu = 0; cpu < xyarray__max_x(evsel->fd); cpu++) + perf_evsel__close_fd_cpu(evsel, cpu); +} + +void perf_evsel__free_fd(struct perf_evsel *evsel) +{ + xyarray__delete(evsel->fd); + evsel->fd = NULL; +} + +void perf_evsel__close(struct perf_evsel *evsel) +{ + if (evsel->fd == NULL) + return; + + perf_evsel__close_fd(evsel); + perf_evsel__free_fd(evsel); +} + +void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu) +{ + if (evsel->fd == NULL) + return; + + perf_evsel__close_fd_cpu(evsel, cpu); +} + +int perf_evsel__read_size(struct perf_evsel *evsel) +{ + u64 read_format = evsel->attr.read_format; + int entry = sizeof(u64); /* value */ + int size = 0; + int nr = 1; + + if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) + size += sizeof(u64); + + if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) + size += sizeof(u64); + + if (read_format & PERF_FORMAT_ID) + entry += sizeof(u64); + + if (read_format & PERF_FORMAT_GROUP) { + nr = evsel->nr_members; + size += sizeof(u64); + } + + size += entry * nr; + return size; +} + +int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, + struct perf_counts_values *count) +{ + size_t size = perf_evsel__read_size(evsel); + + memset(count, 0, sizeof(*count)); + + if (FD(evsel, cpu, thread) < 0) + return -EINVAL; + + if (readn(FD(evsel, cpu, thread), count->values, size) <= 0) + return -errno; + + return 0; +} + +static int perf_evsel__run_ioctl(struct perf_evsel *evsel, + int ioc, void *arg, + int cpu) +{ + int thread; + + for (thread = 0; thread < xyarray__max_y(evsel->fd); thread++) { + int fd = FD(evsel, cpu, thread), + err = ioctl(fd, ioc, arg); + + if (err) + return err; + } + + return 0; +} + +int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu) +{ + return perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_ENABLE, NULL, cpu); +} + +int perf_evsel__enable(struct perf_evsel *evsel) +{ + int i; + int err = 0; + + for (i = 0; i < xyarray__max_x(evsel->fd) && !err; i++) + err = perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_ENABLE, NULL, i); + return err; +} + +int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu) +{ + return perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_DISABLE, NULL, cpu); +} + +int perf_evsel__disable(struct perf_evsel *evsel) +{ + int i; + int err = 0; + + for (i = 0; i < xyarray__max_x(evsel->fd) && !err; i++) + err = perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_DISABLE, NULL, i); + return err; +} + +int perf_evsel__apply_filter(struct perf_evsel *evsel, const char *filter) +{ + int err = 0, i; + + for (i = 0; i < evsel->cpus->nr && !err; i++) + err = perf_evsel__run_ioctl(evsel, + PERF_EVENT_IOC_SET_FILTER, + (void *)filter, i); + return err; +} + +struct perf_cpu_map *perf_evsel__cpus(struct perf_evsel *evsel) +{ + return evsel->cpus; +} + +struct perf_thread_map *perf_evsel__threads(struct perf_evsel *evsel) +{ + return evsel->threads; +} + +struct perf_event_attr *perf_evsel__attr(struct perf_evsel *evsel) +{ + return &evsel->attr; +} + +int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads) +{ + if (ncpus == 0 || nthreads == 0) + return 0; + + if (evsel->system_wide) + nthreads = 1; + + evsel->sample_id = xyarray__new(ncpus, nthreads, sizeof(struct perf_sample_id)); + if (evsel->sample_id == NULL) + return -ENOMEM; + + evsel->id = zalloc(ncpus * nthreads * sizeof(u64)); + if (evsel->id == NULL) { + xyarray__delete(evsel->sample_id); + evsel->sample_id = NULL; + return -ENOMEM; + } + + return 0; +} + +void perf_evsel__free_id(struct perf_evsel *evsel) +{ + xyarray__delete(evsel->sample_id); + evsel->sample_id = NULL; + zfree(&evsel->id); + evsel->ids = 0; +} diff --git a/tools/lib/perf/include/internal/cpumap.h b/tools/lib/perf/include/internal/cpumap.h new file mode 100644 index 0000000..840d403 --- /dev/null +++ b/tools/lib/perf/include/internal/cpumap.h @@ -0,0 +1,19 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_CPUMAP_H +#define __LIBPERF_INTERNAL_CPUMAP_H + +#include <linux/refcount.h> + +struct perf_cpu_map { + refcount_t refcnt; + int nr; + int map[]; +}; + +#ifndef MAX_NR_CPUS +#define MAX_NR_CPUS 2048 +#endif + +int perf_cpu_map__idx(struct perf_cpu_map *cpus, int cpu); + +#endif /* __LIBPERF_INTERNAL_CPUMAP_H */ diff --git a/tools/lib/perf/include/internal/evlist.h b/tools/lib/perf/include/internal/evlist.h new file mode 100644 index 0000000..74dc8c3 --- /dev/null +++ b/tools/lib/perf/include/internal/evlist.h @@ -0,0 +1,127 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_EVLIST_H +#define __LIBPERF_INTERNAL_EVLIST_H + +#include <linux/list.h> +#include <api/fd/array.h> +#include <internal/evsel.h> + +#define PERF_EVLIST__HLIST_BITS 8 +#define PERF_EVLIST__HLIST_SIZE (1 << PERF_EVLIST__HLIST_BITS) + +struct perf_cpu_map; +struct perf_thread_map; +struct perf_mmap_param; + +struct perf_evlist { + struct list_head entries; + int nr_entries; + bool has_user_cpus; + struct perf_cpu_map *cpus; + struct perf_cpu_map *all_cpus; + struct perf_thread_map *threads; + int nr_mmaps; + size_t mmap_len; + struct fdarray pollfd; + struct hlist_head heads[PERF_EVLIST__HLIST_SIZE]; + struct perf_mmap *mmap; + struct perf_mmap *mmap_ovw; + struct perf_mmap *mmap_first; + struct perf_mmap *mmap_ovw_first; +}; + +typedef void +(*perf_evlist_mmap__cb_idx_t)(struct perf_evlist*, struct perf_mmap_param*, int, bool); +typedef struct perf_mmap* +(*perf_evlist_mmap__cb_get_t)(struct perf_evlist*, bool, int); +typedef int +(*perf_evlist_mmap__cb_mmap_t)(struct perf_mmap*, struct perf_mmap_param*, int, int); + +struct perf_evlist_mmap_ops { + perf_evlist_mmap__cb_idx_t idx; + perf_evlist_mmap__cb_get_t get; + perf_evlist_mmap__cb_mmap_t mmap; +}; + +int perf_evlist__alloc_pollfd(struct perf_evlist *evlist); +int perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd, + void *ptr, short revent); + +int perf_evlist__mmap_ops(struct perf_evlist *evlist, + struct perf_evlist_mmap_ops *ops, + struct perf_mmap_param *mp); + +void perf_evlist__init(struct perf_evlist *evlist); +void perf_evlist__exit(struct perf_evlist *evlist); + +/** + * __perf_evlist__for_each_entry - iterate thru all the evsels + * @list: list_head instance to iterate + * @evsel: struct perf_evsel iterator + */ +#define __perf_evlist__for_each_entry(list, evsel) \ + list_for_each_entry(evsel, list, node) + +/** + * evlist__for_each_entry - iterate thru all the evsels + * @evlist: perf_evlist instance to iterate + * @evsel: struct perf_evsel iterator + */ +#define perf_evlist__for_each_entry(evlist, evsel) \ + __perf_evlist__for_each_entry(&(evlist)->entries, evsel) + +/** + * __perf_evlist__for_each_entry_reverse - iterate thru all the evsels in reverse order + * @list: list_head instance to iterate + * @evsel: struct evsel iterator + */ +#define __perf_evlist__for_each_entry_reverse(list, evsel) \ + list_for_each_entry_reverse(evsel, list, node) + +/** + * perf_evlist__for_each_entry_reverse - iterate thru all the evsels in reverse order + * @evlist: evlist instance to iterate + * @evsel: struct evsel iterator + */ +#define perf_evlist__for_each_entry_reverse(evlist, evsel) \ + __perf_evlist__for_each_entry_reverse(&(evlist)->entries, evsel) + +/** + * __perf_evlist__for_each_entry_safe - safely iterate thru all the evsels + * @list: list_head instance to iterate + * @tmp: struct evsel temp iterator + * @evsel: struct evsel iterator + */ +#define __perf_evlist__for_each_entry_safe(list, tmp, evsel) \ + list_for_each_entry_safe(evsel, tmp, list, node) + +/** + * perf_evlist__for_each_entry_safe - safely iterate thru all the evsels + * @evlist: evlist instance to iterate + * @evsel: struct evsel iterator + * @tmp: struct evsel temp iterator + */ +#define perf_evlist__for_each_entry_safe(evlist, tmp, evsel) \ + __perf_evlist__for_each_entry_safe(&(evlist)->entries, tmp, evsel) + +static inline struct perf_evsel *perf_evlist__first(struct perf_evlist *evlist) +{ + return list_entry(evlist->entries.next, struct perf_evsel, node); +} + +static inline struct perf_evsel *perf_evlist__last(struct perf_evlist *evlist) +{ + return list_entry(evlist->entries.prev, struct perf_evsel, node); +} + +u64 perf_evlist__read_format(struct perf_evlist *evlist); + +void perf_evlist__id_add(struct perf_evlist *evlist, + struct perf_evsel *evsel, + int cpu, int thread, u64 id); + +int perf_evlist__id_add_fd(struct perf_evlist *evlist, + struct perf_evsel *evsel, + int cpu, int thread, int fd); + +#endif /* __LIBPERF_INTERNAL_EVLIST_H */ diff --git a/tools/lib/perf/include/internal/evsel.h b/tools/lib/perf/include/internal/evsel.h new file mode 100644 index 0000000..1ffd083 --- /dev/null +++ b/tools/lib/perf/include/internal/evsel.h @@ -0,0 +1,63 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_EVSEL_H +#define __LIBPERF_INTERNAL_EVSEL_H + +#include <linux/types.h> +#include <linux/perf_event.h> +#include <stdbool.h> +#include <sys/types.h> + +struct perf_cpu_map; +struct perf_thread_map; +struct xyarray; + +/* + * Per fd, to map back from PERF_SAMPLE_ID to evsel, only used when there are + * more than one entry in the evlist. + */ +struct perf_sample_id { + struct hlist_node node; + u64 id; + struct perf_evsel *evsel; + /* + * 'idx' will be used for AUX area sampling. A sample will have AUX area + * data that will be queued for decoding, where there are separate + * queues for each CPU (per-cpu tracing) or task (per-thread tracing). + * The sample ID can be used to lookup 'idx' which is effectively the + * queue number. + */ + int idx; + int cpu; + pid_t tid; + + /* Holds total ID period value for PERF_SAMPLE_READ processing. */ + u64 period; +}; + +struct perf_evsel { + struct list_head node; + struct perf_event_attr attr; + struct perf_cpu_map *cpus; + struct perf_cpu_map *own_cpus; + struct perf_thread_map *threads; + struct xyarray *fd; + struct xyarray *sample_id; + u64 *id; + u32 ids; + + /* parse modifier helper */ + int nr_members; + bool system_wide; +}; + +void perf_evsel__init(struct perf_evsel *evsel, struct perf_event_attr *attr); +int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads); +void perf_evsel__close_fd(struct perf_evsel *evsel); +void perf_evsel__free_fd(struct perf_evsel *evsel); +int perf_evsel__read_size(struct perf_evsel *evsel); +int perf_evsel__apply_filter(struct perf_evsel *evsel, const char *filter); + +int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads); +void perf_evsel__free_id(struct perf_evsel *evsel); + +#endif /* __LIBPERF_INTERNAL_EVSEL_H */ diff --git a/tools/lib/perf/include/internal/lib.h b/tools/lib/perf/include/internal/lib.h new file mode 100644 index 0000000..5175d49 --- /dev/null +++ b/tools/lib/perf/include/internal/lib.h @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_LIB_H +#define __LIBPERF_INTERNAL_LIB_H + +#include <sys/types.h> + +extern unsigned int page_size; + +ssize_t readn(int fd, void *buf, size_t n); +ssize_t writen(int fd, const void *buf, size_t n); + +#endif /* __LIBPERF_INTERNAL_CPUMAP_H */ diff --git a/tools/lib/perf/include/internal/mmap.h b/tools/lib/perf/include/internal/mmap.h new file mode 100644 index 0000000..be7556e --- /dev/null +++ b/tools/lib/perf/include/internal/mmap.h @@ -0,0 +1,55 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_MMAP_H +#define __LIBPERF_INTERNAL_MMAP_H + +#include <linux/compiler.h> +#include <linux/refcount.h> +#include <linux/types.h> +#include <stdbool.h> + +/* perf sample has 16 bits size limit */ +#define PERF_SAMPLE_MAX_SIZE (1 << 16) + +struct perf_mmap; + +typedef void (*libperf_unmap_cb_t)(struct perf_mmap *map); + +/** + * struct perf_mmap - perf's ring buffer mmap details + * + * @refcnt - e.g. code using PERF_EVENT_IOC_SET_OUTPUT to share this + */ +struct perf_mmap { + void *base; + int mask; + int fd; + int cpu; + refcount_t refcnt; + u64 prev; + u64 start; + u64 end; + bool overwrite; + u64 flush; + libperf_unmap_cb_t unmap_cb; + char event_copy[PERF_SAMPLE_MAX_SIZE] __aligned(8); + struct perf_mmap *next; +}; + +struct perf_mmap_param { + int prot; + int mask; +}; + +size_t perf_mmap__mmap_len(struct perf_mmap *map); + +void perf_mmap__init(struct perf_mmap *map, struct perf_mmap *prev, + bool overwrite, libperf_unmap_cb_t unmap_cb); +int perf_mmap__mmap(struct perf_mmap *map, struct perf_mmap_param *mp, + int fd, int cpu); +void perf_mmap__munmap(struct perf_mmap *map); +void perf_mmap__get(struct perf_mmap *map); +void perf_mmap__put(struct perf_mmap *map); + +u64 perf_mmap__read_head(struct perf_mmap *map); + +#endif /* __LIBPERF_INTERNAL_MMAP_H */ diff --git a/tools/lib/perf/include/internal/tests.h b/tools/lib/perf/include/internal/tests.h new file mode 100644 index 0000000..2093e88 --- /dev/null +++ b/tools/lib/perf/include/internal/tests.h @@ -0,0 +1,33 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_TESTS_H +#define __LIBPERF_INTERNAL_TESTS_H + +#include <stdio.h> + +int tests_failed; + +#define __T_START \ +do { \ + fprintf(stdout, "- running %s...", __FILE__); \ + fflush(NULL); \ + tests_failed = 0; \ +} while (0) + +#define __T_END \ +do { \ + if (tests_failed) \ + fprintf(stdout, " FAILED (%d)\n", tests_failed); \ + else \ + fprintf(stdout, "OK\n"); \ +} while (0) + +#define __T(text, cond) \ +do { \ + if (!(cond)) { \ + fprintf(stderr, "FAILED %s:%d %s\n", __FILE__, __LINE__, text); \ + tests_failed++; \ + return -1; \ + } \ +} while (0) + +#endif /* __LIBPERF_INTERNAL_TESTS_H */ diff --git a/tools/lib/perf/include/internal/threadmap.h b/tools/lib/perf/include/internal/threadmap.h new file mode 100644 index 0000000..df748ba --- /dev/null +++ b/tools/lib/perf/include/internal/threadmap.h @@ -0,0 +1,23 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_THREADMAP_H +#define __LIBPERF_INTERNAL_THREADMAP_H + +#include <linux/refcount.h> +#include <sys/types.h> +#include <unistd.h> + +struct thread_map_data { + pid_t pid; + char *comm; +}; + +struct perf_thread_map { + refcount_t refcnt; + int nr; + int err_thread; + struct thread_map_data map[]; +}; + +struct perf_thread_map *perf_thread_map__realloc(struct perf_thread_map *map, int nr); + +#endif /* __LIBPERF_INTERNAL_THREADMAP_H */ diff --git a/tools/lib/perf/include/internal/xyarray.h b/tools/lib/perf/include/internal/xyarray.h new file mode 100644 index 0000000..51e35d6 --- /dev/null +++ b/tools/lib/perf/include/internal/xyarray.h @@ -0,0 +1,36 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_XYARRAY_H +#define __LIBPERF_INTERNAL_XYARRAY_H + +#include <linux/compiler.h> +#include <sys/types.h> + +struct xyarray { + size_t row_size; + size_t entry_size; + size_t entries; + size_t max_x; + size_t max_y; + char contents[] __aligned(8); +}; + +struct xyarray *xyarray__new(int xlen, int ylen, size_t entry_size); +void xyarray__delete(struct xyarray *xy); +void xyarray__reset(struct xyarray *xy); + +static inline void *xyarray__entry(struct xyarray *xy, int x, int y) +{ + return &xy->contents[x * xy->row_size + y * xy->entry_size]; +} + +static inline int xyarray__max_y(struct xyarray *xy) +{ + return xy->max_y; +} + +static inline int xyarray__max_x(struct xyarray *xy) +{ + return xy->max_x; +} + +#endif /* __LIBPERF_INTERNAL_XYARRAY_H */ diff --git a/tools/lib/perf/include/perf/core.h b/tools/lib/perf/include/perf/core.h new file mode 100644 index 0000000..a3f6d68 --- /dev/null +++ b/tools/lib/perf/include/perf/core.h @@ -0,0 +1,25 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_CORE_H +#define __LIBPERF_CORE_H + +#include <stdarg.h> + +#ifndef LIBPERF_API +#define LIBPERF_API __attribute__((visibility("default"))) +#endif + +enum libperf_print_level { + LIBPERF_ERR, + LIBPERF_WARN, + LIBPERF_INFO, + LIBPERF_DEBUG, + LIBPERF_DEBUG2, + LIBPERF_DEBUG3, +}; + +typedef int (*libperf_print_fn_t)(enum libperf_print_level level, + const char *, va_list ap); + +LIBPERF_API void libperf_init(libperf_print_fn_t fn); + +#endif /* __LIBPERF_CORE_H */ diff --git a/tools/lib/perf/include/perf/cpumap.h b/tools/lib/perf/include/perf/cpumap.h new file mode 100644 index 0000000..6a17ad7 --- /dev/null +++ b/tools/lib/perf/include/perf/cpumap.h @@ -0,0 +1,28 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_CPUMAP_H +#define __LIBPERF_CPUMAP_H + +#include <perf/core.h> +#include <stdio.h> +#include <stdbool.h> + +struct perf_cpu_map; + +LIBPERF_API struct perf_cpu_map *perf_cpu_map__dummy_new(void); +LIBPERF_API struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list); +LIBPERF_API struct perf_cpu_map *perf_cpu_map__read(FILE *file); +LIBPERF_API struct perf_cpu_map *perf_cpu_map__get(struct perf_cpu_map *map); +LIBPERF_API struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig, + struct perf_cpu_map *other); +LIBPERF_API void perf_cpu_map__put(struct perf_cpu_map *map); +LIBPERF_API int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx); +LIBPERF_API int perf_cpu_map__nr(const struct perf_cpu_map *cpus); +LIBPERF_API bool perf_cpu_map__empty(const struct perf_cpu_map *map); +LIBPERF_API int perf_cpu_map__max(struct perf_cpu_map *map); + +#define perf_cpu_map__for_each_cpu(cpu, idx, cpus) \ + for ((idx) = 0, (cpu) = perf_cpu_map__cpu(cpus, idx); \ + (idx) < perf_cpu_map__nr(cpus); \ + (idx)++, (cpu) = perf_cpu_map__cpu(cpus, idx)) + +#endif /* __LIBPERF_CPUMAP_H */ diff --git a/tools/lib/perf/include/perf/event.h b/tools/lib/perf/include/perf/event.h new file mode 100644 index 0000000..1810689 --- /dev/null +++ b/tools/lib/perf/include/perf/event.h @@ -0,0 +1,385 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_EVENT_H +#define __LIBPERF_EVENT_H + +#include <linux/perf_event.h> +#include <linux/types.h> +#include <linux/limits.h> +#include <linux/bpf.h> +#include <sys/types.h> /* pid_t */ + +struct perf_record_mmap { + struct perf_event_header header; + __u32 pid, tid; + __u64 start; + __u64 len; + __u64 pgoff; + char filename[PATH_MAX]; +}; + +struct perf_record_mmap2 { + struct perf_event_header header; + __u32 pid, tid; + __u64 start; + __u64 len; + __u64 pgoff; + __u32 maj; + __u32 min; + __u64 ino; + __u64 ino_generation; + __u32 prot; + __u32 flags; + char filename[PATH_MAX]; +}; + +struct perf_record_comm { + struct perf_event_header header; + __u32 pid, tid; + char comm[16]; +}; + +struct perf_record_namespaces { + struct perf_event_header header; + __u32 pid, tid; + __u64 nr_namespaces; + struct perf_ns_link_info link_info[]; +}; + +struct perf_record_fork { + struct perf_event_header header; + __u32 pid, ppid; + __u32 tid, ptid; + __u64 time; +}; + +struct perf_record_lost { + struct perf_event_header header; + __u64 id; + __u64 lost; +}; + +struct perf_record_lost_samples { + struct perf_event_header header; + __u64 lost; +}; + +/* + * PERF_FORMAT_ENABLED | PERF_FORMAT_RUNNING | PERF_FORMAT_ID + */ +struct perf_record_read { + struct perf_event_header header; + __u32 pid, tid; + __u64 value; + __u64 time_enabled; + __u64 time_running; + __u64 id; +}; + +struct perf_record_throttle { + struct perf_event_header header; + __u64 time; + __u64 id; + __u64 stream_id; +}; + +#ifndef KSYM_NAME_LEN +#define KSYM_NAME_LEN 256 +#endif + +struct perf_record_ksymbol { + struct perf_event_header header; + __u64 addr; + __u32 len; + __u16 ksym_type; + __u16 flags; + char name[KSYM_NAME_LEN]; +}; + +struct perf_record_bpf_event { + struct perf_event_header header; + __u16 type; + __u16 flags; + __u32 id; + + /* for bpf_prog types */ + __u8 tag[BPF_TAG_SIZE]; // prog tag +}; + +struct perf_record_sample { + struct perf_event_header header; + __u64 array[]; +}; + +struct perf_record_switch { + struct perf_event_header header; + __u32 next_prev_pid; + __u32 next_prev_tid; +}; + +struct perf_record_header_attr { + struct perf_event_header header; + struct perf_event_attr attr; + __u64 id[]; +}; + +enum { + PERF_CPU_MAP__CPUS = 0, + PERF_CPU_MAP__MASK = 1, +}; + +struct cpu_map_entries { + __u16 nr; + __u16 cpu[]; +}; + +struct perf_record_record_cpu_map { + __u16 nr; + __u16 long_size; + unsigned long mask[]; +}; + +struct perf_record_cpu_map_data { + __u16 type; + char data[]; +}; + +struct perf_record_cpu_map { + struct perf_event_header header; + struct perf_record_cpu_map_data data; +}; + +enum { + PERF_EVENT_UPDATE__UNIT = 0, + PERF_EVENT_UPDATE__SCALE = 1, + PERF_EVENT_UPDATE__NAME = 2, + PERF_EVENT_UPDATE__CPUS = 3, +}; + +struct perf_record_event_update_cpus { + struct perf_record_cpu_map_data cpus; +}; + +struct perf_record_event_update_scale { + double scale; +}; + +struct perf_record_event_update { + struct perf_event_header header; + __u64 type; + __u64 id; + char data[]; +}; + +#define MAX_EVENT_NAME 64 + +struct perf_trace_event_type { + __u64 event_id; + char name[MAX_EVENT_NAME]; +}; + +struct perf_record_header_event_type { + struct perf_event_header header; + struct perf_trace_event_type event_type; +}; + +struct perf_record_header_tracing_data { + struct perf_event_header header; + __u32 size; +}; + +struct perf_record_header_build_id { + struct perf_event_header header; + pid_t pid; + __u8 build_id[24]; + char filename[]; +}; + +struct id_index_entry { + __u64 id; + __u64 idx; + __u64 cpu; + __u64 tid; +}; + +struct perf_record_id_index { + struct perf_event_header header; + __u64 nr; + struct id_index_entry entries[0]; +}; + +struct perf_record_auxtrace_info { + struct perf_event_header header; + __u32 type; + __u32 reserved__; /* For alignment */ + __u64 priv[]; +}; + +struct perf_record_auxtrace { + struct perf_event_header header; + __u64 size; + __u64 offset; + __u64 reference; + __u32 idx; + __u32 tid; + __u32 cpu; + __u32 reserved__; /* For alignment */ +}; + +#define MAX_AUXTRACE_ERROR_MSG 64 + +struct perf_record_auxtrace_error { + struct perf_event_header header; + __u32 type; + __u32 code; + __u32 cpu; + __u32 pid; + __u32 tid; + __u32 fmt; + __u64 ip; + __u64 time; + char msg[MAX_AUXTRACE_ERROR_MSG]; +}; + +struct perf_record_aux { + struct perf_event_header header; + __u64 aux_offset; + __u64 aux_size; + __u64 flags; +}; + +struct perf_record_itrace_start { + struct perf_event_header header; + __u32 pid; + __u32 tid; +}; + +struct perf_record_thread_map_entry { + __u64 pid; + char comm[16]; +}; + +struct perf_record_thread_map { + struct perf_event_header header; + __u64 nr; + struct perf_record_thread_map_entry entries[]; +}; + +enum { + PERF_STAT_CONFIG_TERM__AGGR_MODE = 0, + PERF_STAT_CONFIG_TERM__INTERVAL = 1, + PERF_STAT_CONFIG_TERM__SCALE = 2, + PERF_STAT_CONFIG_TERM__MAX = 3, +}; + +struct perf_record_stat_config_entry { + __u64 tag; + __u64 val; +}; + +struct perf_record_stat_config { + struct perf_event_header header; + __u64 nr; + struct perf_record_stat_config_entry data[]; +}; + +struct perf_record_stat { + struct perf_event_header header; + + __u64 id; + __u32 cpu; + __u32 thread; + + union { + struct { + __u64 val; + __u64 ena; + __u64 run; + }; + __u64 values[3]; + }; +}; + +struct perf_record_stat_round { + struct perf_event_header header; + __u64 type; + __u64 time; +}; + +struct perf_record_time_conv { + struct perf_event_header header; + __u64 time_shift; + __u64 time_mult; + __u64 time_zero; +}; + +struct perf_record_header_feature { + struct perf_event_header header; + __u64 feat_id; + char data[]; +}; + +struct perf_record_compressed { + struct perf_event_header header; + char data[]; +}; + +enum perf_user_event_type { /* above any possible kernel type */ + PERF_RECORD_USER_TYPE_START = 64, + PERF_RECORD_HEADER_ATTR = 64, + PERF_RECORD_HEADER_EVENT_TYPE = 65, /* deprecated */ + PERF_RECORD_HEADER_TRACING_DATA = 66, + PERF_RECORD_HEADER_BUILD_ID = 67, + PERF_RECORD_FINISHED_ROUND = 68, + PERF_RECORD_ID_INDEX = 69, + PERF_RECORD_AUXTRACE_INFO = 70, + PERF_RECORD_AUXTRACE = 71, + PERF_RECORD_AUXTRACE_ERROR = 72, + PERF_RECORD_THREAD_MAP = 73, + PERF_RECORD_CPU_MAP = 74, + PERF_RECORD_STAT_CONFIG = 75, + PERF_RECORD_STAT = 76, + PERF_RECORD_STAT_ROUND = 77, + PERF_RECORD_EVENT_UPDATE = 78, + PERF_RECORD_TIME_CONV = 79, + PERF_RECORD_HEADER_FEATURE = 80, + PERF_RECORD_COMPRESSED = 81, + PERF_RECORD_HEADER_MAX +}; + +union perf_event { + struct perf_event_header header; + struct perf_record_mmap mmap; + struct perf_record_mmap2 mmap2; + struct perf_record_comm comm; + struct perf_record_namespaces namespaces; + struct perf_record_fork fork; + struct perf_record_lost lost; + struct perf_record_lost_samples lost_samples; + struct perf_record_read read; + struct perf_record_throttle throttle; + struct perf_record_sample sample; + struct perf_record_bpf_event bpf; + struct perf_record_ksymbol ksymbol; + struct perf_record_header_attr attr; + struct perf_record_event_update event_update; + struct perf_record_header_event_type event_type; + struct perf_record_header_tracing_data tracing_data; + struct perf_record_header_build_id build_id; + struct perf_record_id_index id_index; + struct perf_record_auxtrace_info auxtrace_info; + struct perf_record_auxtrace auxtrace; + struct perf_record_auxtrace_error auxtrace_error; + struct perf_record_aux aux; + struct perf_record_itrace_start itrace_start; + struct perf_record_switch context_switch; + struct perf_record_thread_map thread_map; + struct perf_record_cpu_map cpu_map; + struct perf_record_stat_config stat_config; + struct perf_record_stat stat; + struct perf_record_stat_round stat_round; + struct perf_record_time_conv time_conv; + struct perf_record_header_feature feat; + struct perf_record_compressed pack; +}; + +#endif /* __LIBPERF_EVENT_H */ diff --git a/tools/lib/perf/include/perf/evlist.h b/tools/lib/perf/include/perf/evlist.h new file mode 100644 index 0000000..0a7479d --- /dev/null +++ b/tools/lib/perf/include/perf/evlist.h @@ -0,0 +1,49 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_EVLIST_H +#define __LIBPERF_EVLIST_H + +#include <perf/core.h> +#include <stdbool.h> + +struct perf_evlist; +struct perf_evsel; +struct perf_cpu_map; +struct perf_thread_map; + +LIBPERF_API void perf_evlist__add(struct perf_evlist *evlist, + struct perf_evsel *evsel); +LIBPERF_API void perf_evlist__remove(struct perf_evlist *evlist, + struct perf_evsel *evsel); +LIBPERF_API struct perf_evlist *perf_evlist__new(void); +LIBPERF_API void perf_evlist__delete(struct perf_evlist *evlist); +LIBPERF_API struct perf_evsel* perf_evlist__next(struct perf_evlist *evlist, + struct perf_evsel *evsel); +LIBPERF_API int perf_evlist__open(struct perf_evlist *evlist); +LIBPERF_API void perf_evlist__close(struct perf_evlist *evlist); +LIBPERF_API void perf_evlist__enable(struct perf_evlist *evlist); +LIBPERF_API void perf_evlist__disable(struct perf_evlist *evlist); + +#define perf_evlist__for_each_evsel(evlist, pos) \ + for ((pos) = perf_evlist__next((evlist), NULL); \ + (pos) != NULL; \ + (pos) = perf_evlist__next((evlist), (pos))) + +LIBPERF_API void perf_evlist__set_maps(struct perf_evlist *evlist, + struct perf_cpu_map *cpus, + struct perf_thread_map *threads); +LIBPERF_API int perf_evlist__poll(struct perf_evlist *evlist, int timeout); +LIBPERF_API int perf_evlist__filter_pollfd(struct perf_evlist *evlist, + short revents_and_mask); + +LIBPERF_API int perf_evlist__mmap(struct perf_evlist *evlist, int pages); +LIBPERF_API void perf_evlist__munmap(struct perf_evlist *evlist); + +LIBPERF_API struct perf_mmap *perf_evlist__next_mmap(struct perf_evlist *evlist, + struct perf_mmap *map, + bool overwrite); +#define perf_evlist__for_each_mmap(evlist, pos, overwrite) \ + for ((pos) = perf_evlist__next_mmap((evlist), NULL, overwrite); \ + (pos) != NULL; \ + (pos) = perf_evlist__next_mmap((evlist), (pos), overwrite)) + +#endif /* __LIBPERF_EVLIST_H */ diff --git a/tools/lib/perf/include/perf/evsel.h b/tools/lib/perf/include/perf/evsel.h new file mode 100644 index 0000000..c82ec39 --- /dev/null +++ b/tools/lib/perf/include/perf/evsel.h @@ -0,0 +1,40 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_EVSEL_H +#define __LIBPERF_EVSEL_H + +#include <stdint.h> +#include <perf/core.h> + +struct perf_evsel; +struct perf_event_attr; +struct perf_cpu_map; +struct perf_thread_map; + +struct perf_counts_values { + union { + struct { + uint64_t val; + uint64_t ena; + uint64_t run; + }; + uint64_t values[3]; + }; +}; + +LIBPERF_API struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr); +LIBPERF_API void perf_evsel__delete(struct perf_evsel *evsel); +LIBPERF_API int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus, + struct perf_thread_map *threads); +LIBPERF_API void perf_evsel__close(struct perf_evsel *evsel); +LIBPERF_API void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu); +LIBPERF_API int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, + struct perf_counts_values *count); +LIBPERF_API int perf_evsel__enable(struct perf_evsel *evsel); +LIBPERF_API int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu); +LIBPERF_API int perf_evsel__disable(struct perf_evsel *evsel); +LIBPERF_API int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu); +LIBPERF_API struct perf_cpu_map *perf_evsel__cpus(struct perf_evsel *evsel); +LIBPERF_API struct perf_thread_map *perf_evsel__threads(struct perf_evsel *evsel); +LIBPERF_API struct perf_event_attr *perf_evsel__attr(struct perf_evsel *evsel); + +#endif /* __LIBPERF_EVSEL_H */ diff --git a/tools/lib/perf/include/perf/mmap.h b/tools/lib/perf/include/perf/mmap.h new file mode 100644 index 0000000..9508ad9 --- /dev/null +++ b/tools/lib/perf/include/perf/mmap.h @@ -0,0 +1,15 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_MMAP_H +#define __LIBPERF_MMAP_H + +#include <perf/core.h> + +struct perf_mmap; +union perf_event; + +LIBPERF_API void perf_mmap__consume(struct perf_mmap *map); +LIBPERF_API int perf_mmap__read_init(struct perf_mmap *map); +LIBPERF_API void perf_mmap__read_done(struct perf_mmap *map); +LIBPERF_API union perf_event *perf_mmap__read_event(struct perf_mmap *map); + +#endif /* __LIBPERF_MMAP_H */ diff --git a/tools/lib/perf/include/perf/threadmap.h b/tools/lib/perf/include/perf/threadmap.h new file mode 100644 index 0000000..a7c50de --- /dev/null +++ b/tools/lib/perf/include/perf/threadmap.h @@ -0,0 +1,20 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_THREADMAP_H +#define __LIBPERF_THREADMAP_H + +#include <perf/core.h> +#include <sys/types.h> + +struct perf_thread_map; + +LIBPERF_API struct perf_thread_map *perf_thread_map__new_dummy(void); + +LIBPERF_API void perf_thread_map__set_pid(struct perf_thread_map *map, int thread, pid_t pid); +LIBPERF_API char *perf_thread_map__comm(struct perf_thread_map *map, int thread); +LIBPERF_API int perf_thread_map__nr(struct perf_thread_map *threads); +LIBPERF_API pid_t perf_thread_map__pid(struct perf_thread_map *map, int thread); + +LIBPERF_API struct perf_thread_map *perf_thread_map__get(struct perf_thread_map *map); +LIBPERF_API void perf_thread_map__put(struct perf_thread_map *map); + +#endif /* __LIBPERF_THREADMAP_H */ diff --git a/tools/lib/perf/internal.h b/tools/lib/perf/internal.h new file mode 100644 index 0000000..2c27e15 --- /dev/null +++ b/tools/lib/perf/internal.h @@ -0,0 +1,23 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __LIBPERF_INTERNAL_H +#define __LIBPERF_INTERNAL_H + +#include <perf/core.h> + +void libperf_print(enum libperf_print_level level, + const char *format, ...) + __attribute__((format(printf, 2, 3))); + +#define __pr(level, fmt, ...) \ +do { \ + libperf_print(level, "libperf: " fmt, ##__VA_ARGS__); \ +} while (0) + +#define pr_err(fmt, ...) __pr(LIBPERF_ERR, fmt, ##__VA_ARGS__) +#define pr_warning(fmt, ...) __pr(LIBPERF_WARN, fmt, ##__VA_ARGS__) +#define pr_info(fmt, ...) __pr(LIBPERF_INFO, fmt, ##__VA_ARGS__) +#define pr_debug(fmt, ...) __pr(LIBPERF_DEBUG, fmt, ##__VA_ARGS__) +#define pr_debug2(fmt, ...) __pr(LIBPERF_DEBUG2, fmt, ##__VA_ARGS__) +#define pr_debug3(fmt, ...) __pr(LIBPERF_DEBUG3, fmt, ##__VA_ARGS__) + +#endif /* __LIBPERF_INTERNAL_H */ diff --git a/tools/lib/perf/lib.c b/tools/lib/perf/lib.c new file mode 100644 index 0000000..1865893 --- /dev/null +++ b/tools/lib/perf/lib.c @@ -0,0 +1,48 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <unistd.h> +#include <stdbool.h> +#include <errno.h> +#include <linux/kernel.h> +#include <internal/lib.h> + +unsigned int page_size; + +static ssize_t ion(bool is_read, int fd, void *buf, size_t n) +{ + void *buf_start = buf; + size_t left = n; + + while (left) { + /* buf must be treated as const if !is_read. */ + ssize_t ret = is_read ? read(fd, buf, left) : + write(fd, buf, left); + + if (ret < 0 && errno == EINTR) + continue; + if (ret <= 0) + return ret; + + left -= ret; + buf += ret; + } + + BUG_ON((size_t)(buf - buf_start) != n); + return n; +} + +/* + * Read exactly 'n' bytes or return an error. + */ +ssize_t readn(int fd, void *buf, size_t n) +{ + return ion(true, fd, buf, n); +} + +/* + * Write exactly 'n' bytes or return an error. + */ +ssize_t writen(int fd, const void *buf, size_t n) +{ + /* ion does not modify buf. */ + return ion(false, fd, (void *)buf, n); +} diff --git a/tools/lib/perf/libperf.map b/tools/lib/perf/libperf.map new file mode 100644 index 0000000..7be1af8 --- /dev/null +++ b/tools/lib/perf/libperf.map @@ -0,0 +1,51 @@ +LIBPERF_0.0.1 { + global: + libperf_init; + perf_cpu_map__dummy_new; + perf_cpu_map__get; + perf_cpu_map__put; + perf_cpu_map__new; + perf_cpu_map__read; + perf_cpu_map__nr; + perf_cpu_map__cpu; + perf_cpu_map__empty; + perf_cpu_map__max; + perf_thread_map__new_dummy; + perf_thread_map__set_pid; + perf_thread_map__comm; + perf_thread_map__nr; + perf_thread_map__pid; + perf_thread_map__get; + perf_thread_map__put; + perf_evsel__new; + perf_evsel__delete; + perf_evsel__enable; + perf_evsel__disable; + perf_evsel__open; + perf_evsel__close; + perf_evsel__read; + perf_evsel__cpus; + perf_evsel__threads; + perf_evsel__attr; + perf_evlist__new; + perf_evlist__delete; + perf_evlist__open; + perf_evlist__close; + perf_evlist__enable; + perf_evlist__disable; + perf_evlist__add; + perf_evlist__remove; + perf_evlist__next; + perf_evlist__set_maps; + perf_evlist__poll; + perf_evlist__mmap; + perf_evlist__munmap; + perf_evlist__filter_pollfd; + perf_evlist__next_mmap; + perf_mmap__consume; + perf_mmap__read_init; + perf_mmap__read_done; + perf_mmap__read_event; + local: + *; +}; diff --git a/tools/lib/perf/libperf.pc.template b/tools/lib/perf/libperf.pc.template new file mode 100644 index 0000000..117e4a2 --- /dev/null +++ b/tools/lib/perf/libperf.pc.template @@ -0,0 +1,11 @@ +# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) + +prefix=@PREFIX@ +libdir=@LIBDIR@ +includedir=${prefix}/include + +Name: libperf +Description: perf library +Version: @VERSION@ +Libs: -L${libdir} -lperf +Cflags: -I${includedir} diff --git a/tools/lib/perf/mmap.c b/tools/lib/perf/mmap.c new file mode 100644 index 0000000..79d5ed6 --- /dev/null +++ b/tools/lib/perf/mmap.c @@ -0,0 +1,275 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <sys/mman.h> +#include <inttypes.h> +#include <asm/bug.h> +#include <errno.h> +#include <string.h> +#include <linux/ring_buffer.h> +#include <linux/perf_event.h> +#include <perf/mmap.h> +#include <perf/event.h> +#include <internal/mmap.h> +#include <internal/lib.h> +#include <linux/kernel.h> +#include "internal.h" + +void perf_mmap__init(struct perf_mmap *map, struct perf_mmap *prev, + bool overwrite, libperf_unmap_cb_t unmap_cb) +{ + map->fd = -1; + map->overwrite = overwrite; + map->unmap_cb = unmap_cb; + refcount_set(&map->refcnt, 0); + if (prev) + prev->next = map; +} + +size_t perf_mmap__mmap_len(struct perf_mmap *map) +{ + return map->mask + 1 + page_size; +} + +int perf_mmap__mmap(struct perf_mmap *map, struct perf_mmap_param *mp, + int fd, int cpu) +{ + map->prev = 0; + map->mask = mp->mask; + map->base = mmap(NULL, perf_mmap__mmap_len(map), mp->prot, + MAP_SHARED, fd, 0); + if (map->base == MAP_FAILED) { + map->base = NULL; + return -1; + } + + map->fd = fd; + map->cpu = cpu; + return 0; +} + +void perf_mmap__munmap(struct perf_mmap *map) +{ + if (map && map->base != NULL) { + munmap(map->base, perf_mmap__mmap_len(map)); + map->base = NULL; + map->fd = -1; + refcount_set(&map->refcnt, 0); + } + if (map && map->unmap_cb) + map->unmap_cb(map); +} + +void perf_mmap__get(struct perf_mmap *map) +{ + refcount_inc(&map->refcnt); +} + +void perf_mmap__put(struct perf_mmap *map) +{ + BUG_ON(map->base && refcount_read(&map->refcnt) == 0); + + if (refcount_dec_and_test(&map->refcnt)) + perf_mmap__munmap(map); +} + +static inline void perf_mmap__write_tail(struct perf_mmap *md, u64 tail) +{ + ring_buffer_write_tail(md->base, tail); +} + +u64 perf_mmap__read_head(struct perf_mmap *map) +{ + return ring_buffer_read_head(map->base); +} + +static bool perf_mmap__empty(struct perf_mmap *map) +{ + struct perf_event_mmap_page *pc = map->base; + + return perf_mmap__read_head(map) == map->prev && !pc->aux_size; +} + +void perf_mmap__consume(struct perf_mmap *map) +{ + if (!map->overwrite) { + u64 old = map->prev; + + perf_mmap__write_tail(map, old); + } + + if (refcount_read(&map->refcnt) == 1 && perf_mmap__empty(map)) + perf_mmap__put(map); +} + +static int overwrite_rb_find_range(void *buf, int mask, u64 *start, u64 *end) +{ + struct perf_event_header *pheader; + u64 evt_head = *start; + int size = mask + 1; + + pr_debug2("%s: buf=%p, start=%"PRIx64"\n", __func__, buf, *start); + pheader = (struct perf_event_header *)(buf + (*start & mask)); + while (true) { + if (evt_head - *start >= (unsigned int)size) { + pr_debug("Finished reading overwrite ring buffer: rewind\n"); + if (evt_head - *start > (unsigned int)size) + evt_head -= pheader->size; + *end = evt_head; + return 0; + } + + pheader = (struct perf_event_header *)(buf + (evt_head & mask)); + + if (pheader->size == 0) { + pr_debug("Finished reading overwrite ring buffer: get start\n"); + *end = evt_head; + return 0; + } + + evt_head += pheader->size; + pr_debug3("move evt_head: %"PRIx64"\n", evt_head); + } + WARN_ONCE(1, "Shouldn't get here\n"); + return -1; +} + +/* + * Report the start and end of the available data in ringbuffer + */ +static int __perf_mmap__read_init(struct perf_mmap *md) +{ + u64 head = perf_mmap__read_head(md); + u64 old = md->prev; + unsigned char *data = md->base + page_size; + unsigned long size; + + md->start = md->overwrite ? head : old; + md->end = md->overwrite ? old : head; + + if ((md->end - md->start) < md->flush) + return -EAGAIN; + + size = md->end - md->start; + if (size > (unsigned long)(md->mask) + 1) { + if (!md->overwrite) { + WARN_ONCE(1, "failed to keep up with mmap data. (warn only once)\n"); + + md->prev = head; + perf_mmap__consume(md); + return -EAGAIN; + } + + /* + * Backward ring buffer is full. We still have a chance to read + * most of data from it. + */ + if (overwrite_rb_find_range(data, md->mask, &md->start, &md->end)) + return -EINVAL; + } + + return 0; +} + +int perf_mmap__read_init(struct perf_mmap *map) +{ + /* + * Check if event was unmapped due to a POLLHUP/POLLERR. + */ + if (!refcount_read(&map->refcnt)) + return -ENOENT; + + return __perf_mmap__read_init(map); +} + +/* + * Mandatory for overwrite mode + * The direction of overwrite mode is backward. + * The last perf_mmap__read() will set tail to map->core.prev. + * Need to correct the map->core.prev to head which is the end of next read. + */ +void perf_mmap__read_done(struct perf_mmap *map) +{ + /* + * Check if event was unmapped due to a POLLHUP/POLLERR. + */ + if (!refcount_read(&map->refcnt)) + return; + + map->prev = perf_mmap__read_head(map); +} + +/* When check_messup is true, 'end' must points to a good entry */ +static union perf_event *perf_mmap__read(struct perf_mmap *map, + u64 *startp, u64 end) +{ + unsigned char *data = map->base + page_size; + union perf_event *event = NULL; + int diff = end - *startp; + + if (diff >= (int)sizeof(event->header)) { + size_t size; + + event = (union perf_event *)&data[*startp & map->mask]; + size = event->header.size; + + if (size < sizeof(event->header) || diff < (int)size) + return NULL; + + /* + * Event straddles the mmap boundary -- header should always + * be inside due to u64 alignment of output. + */ + if ((*startp & map->mask) + size != ((*startp + size) & map->mask)) { + unsigned int offset = *startp; + unsigned int len = min(sizeof(*event), size), cpy; + void *dst = map->event_copy; + + do { + cpy = min(map->mask + 1 - (offset & map->mask), len); + memcpy(dst, &data[offset & map->mask], cpy); + offset += cpy; + dst += cpy; + len -= cpy; + } while (len); + + event = (union perf_event *)map->event_copy; + } + + *startp += size; + } + + return event; +} + +/* + * Read event from ring buffer one by one. + * Return one event for each call. + * + * Usage: + * perf_mmap__read_init() + * while(event = perf_mmap__read_event()) { + * //process the event + * perf_mmap__consume() + * } + * perf_mmap__read_done() + */ +union perf_event *perf_mmap__read_event(struct perf_mmap *map) +{ + union perf_event *event; + + /* + * Check if event was unmapped due to a POLLHUP/POLLERR. + */ + if (!refcount_read(&map->refcnt)) + return NULL; + + /* non-overwirte doesn't pause the ringbuffer */ + if (!map->overwrite) + map->end = perf_mmap__read_head(map); + + event = perf_mmap__read(map, &map->start, map->end); + + if (!map->overwrite) + map->prev = map->start; + + return event; +} diff --git a/tools/lib/perf/tests/Makefile b/tools/lib/perf/tests/Makefile new file mode 100644 index 0000000..9684177 --- /dev/null +++ b/tools/lib/perf/tests/Makefile @@ -0,0 +1,38 @@ +# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) + +TESTS = test-cpumap test-threadmap test-evlist test-evsel + +TESTS_SO := $(addsuffix -so,$(TESTS)) +TESTS_A := $(addsuffix -a,$(TESTS)) + +# Set compile option CFLAGS +ifdef EXTRA_CFLAGS + CFLAGS := $(EXTRA_CFLAGS) +else + CFLAGS := -g -Wall +endif + +all: + +include $(srctree)/tools/scripts/Makefile.include + +INCLUDE = -I$(srctree)/tools/lib/perf/include -I$(srctree)/tools/include -I$(srctree)/tools/lib + +$(TESTS_A): FORCE + $(QUIET_LINK)$(CC) $(INCLUDE) $(CFLAGS) -o $@ $(subst -a,.c,$@) ../libperf.a $(LIBAPI) + +$(TESTS_SO): FORCE + $(QUIET_LINK)$(CC) $(INCLUDE) $(CFLAGS) -L.. -o $@ $(subst -so,.c,$@) $(LIBAPI) -lperf + +all: $(TESTS_A) $(TESTS_SO) + +run: + @echo "running static:" + @for i in $(TESTS_A); do ./$$i; done + @echo "running dynamic:" + @for i in $(TESTS_SO); do LD_LIBRARY_PATH=../ ./$$i; done + +clean: + $(call QUIET_CLEAN, tests)$(RM) $(TESTS_A) $(TESTS_SO) + +.PHONY: all clean FORCE diff --git a/tools/lib/perf/tests/test-cpumap.c b/tools/lib/perf/tests/test-cpumap.c new file mode 100644 index 0000000..c8d4509 --- /dev/null +++ b/tools/lib/perf/tests/test-cpumap.c @@ -0,0 +1,31 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <stdarg.h> +#include <stdio.h> +#include <perf/cpumap.h> +#include <internal/tests.h> + +static int libperf_print(enum libperf_print_level level, + const char *fmt, va_list ap) +{ + return vfprintf(stderr, fmt, ap); +} + +int main(int argc, char **argv) +{ + struct perf_cpu_map *cpus; + + __T_START; + + libperf_init(libperf_print); + + cpus = perf_cpu_map__dummy_new(); + if (!cpus) + return -1; + + perf_cpu_map__get(cpus); + perf_cpu_map__put(cpus); + perf_cpu_map__put(cpus); + + __T_END; + return 0; +} diff --git a/tools/lib/perf/tests/test-evlist.c b/tools/lib/perf/tests/test-evlist.c new file mode 100644 index 0000000..6d8ebe0 --- /dev/null +++ b/tools/lib/perf/tests/test-evlist.c @@ -0,0 +1,413 @@ +// SPDX-License-Identifier: GPL-2.0 +#define _GNU_SOURCE // needed for sched.h to get sched_[gs]etaffinity and CPU_(ZERO,SET) +#include <sched.h> +#include <stdio.h> +#include <stdarg.h> +#include <unistd.h> +#include <stdlib.h> +#include <linux/perf_event.h> +#include <linux/limits.h> +#include <sys/types.h> +#include <sys/wait.h> +#include <sys/prctl.h> +#include <perf/cpumap.h> +#include <perf/threadmap.h> +#include <perf/evlist.h> +#include <perf/evsel.h> +#include <perf/mmap.h> +#include <perf/event.h> +#include <internal/tests.h> +#include <api/fs/fs.h> + +static int libperf_print(enum libperf_print_level level, + const char *fmt, va_list ap) +{ + return vfprintf(stderr, fmt, ap); +} + +static int test_stat_cpu(void) +{ + struct perf_cpu_map *cpus; + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_event_attr attr1 = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_CPU_CLOCK, + }; + struct perf_event_attr attr2 = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_TASK_CLOCK, + }; + int err, cpu, tmp; + + cpus = perf_cpu_map__new(NULL); + __T("failed to create cpus", cpus); + + evlist = perf_evlist__new(); + __T("failed to create evlist", evlist); + + evsel = perf_evsel__new(&attr1); + __T("failed to create evsel1", evsel); + + perf_evlist__add(evlist, evsel); + + evsel = perf_evsel__new(&attr2); + __T("failed to create evsel2", evsel); + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, cpus, NULL); + + err = perf_evlist__open(evlist); + __T("failed to open evsel", err == 0); + + perf_evlist__for_each_evsel(evlist, evsel) { + cpus = perf_evsel__cpus(evsel); + + perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { + struct perf_counts_values counts = { .val = 0 }; + + perf_evsel__read(evsel, cpu, 0, &counts); + __T("failed to read value for evsel", counts.val != 0); + } + } + + perf_evlist__close(evlist); + perf_evlist__delete(evlist); + + perf_cpu_map__put(cpus); + return 0; +} + +static int test_stat_thread(void) +{ + struct perf_counts_values counts = { .val = 0 }; + struct perf_thread_map *threads; + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_event_attr attr1 = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_CPU_CLOCK, + }; + struct perf_event_attr attr2 = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_TASK_CLOCK, + }; + int err; + + threads = perf_thread_map__new_dummy(); + __T("failed to create threads", threads); + + perf_thread_map__set_pid(threads, 0, 0); + + evlist = perf_evlist__new(); + __T("failed to create evlist", evlist); + + evsel = perf_evsel__new(&attr1); + __T("failed to create evsel1", evsel); + + perf_evlist__add(evlist, evsel); + + evsel = perf_evsel__new(&attr2); + __T("failed to create evsel2", evsel); + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, NULL, threads); + + err = perf_evlist__open(evlist); + __T("failed to open evsel", err == 0); + + perf_evlist__for_each_evsel(evlist, evsel) { + perf_evsel__read(evsel, 0, 0, &counts); + __T("failed to read value for evsel", counts.val != 0); + } + + perf_evlist__close(evlist); + perf_evlist__delete(evlist); + + perf_thread_map__put(threads); + return 0; +} + +static int test_stat_thread_enable(void) +{ + struct perf_counts_values counts = { .val = 0 }; + struct perf_thread_map *threads; + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_event_attr attr1 = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_CPU_CLOCK, + .disabled = 1, + }; + struct perf_event_attr attr2 = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_TASK_CLOCK, + .disabled = 1, + }; + int err; + + threads = perf_thread_map__new_dummy(); + __T("failed to create threads", threads); + + perf_thread_map__set_pid(threads, 0, 0); + + evlist = perf_evlist__new(); + __T("failed to create evlist", evlist); + + evsel = perf_evsel__new(&attr1); + __T("failed to create evsel1", evsel); + + perf_evlist__add(evlist, evsel); + + evsel = perf_evsel__new(&attr2); + __T("failed to create evsel2", evsel); + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, NULL, threads); + + err = perf_evlist__open(evlist); + __T("failed to open evsel", err == 0); + + perf_evlist__for_each_evsel(evlist, evsel) { + perf_evsel__read(evsel, 0, 0, &counts); + __T("failed to read value for evsel", counts.val == 0); + } + + perf_evlist__enable(evlist); + + perf_evlist__for_each_evsel(evlist, evsel) { + perf_evsel__read(evsel, 0, 0, &counts); + __T("failed to read value for evsel", counts.val != 0); + } + + perf_evlist__disable(evlist); + + perf_evlist__close(evlist); + perf_evlist__delete(evlist); + + perf_thread_map__put(threads); + return 0; +} + +static int test_mmap_thread(void) +{ + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_mmap *map; + struct perf_cpu_map *cpus; + struct perf_thread_map *threads; + struct perf_event_attr attr = { + .type = PERF_TYPE_TRACEPOINT, + .sample_period = 1, + .wakeup_watermark = 1, + .disabled = 1, + }; + char path[PATH_MAX]; + int id, err, pid, go_pipe[2]; + union perf_event *event; + char bf; + int count = 0; + + snprintf(path, PATH_MAX, "%s/kernel/debug/tracing/events/syscalls/sys_enter_prctl/id", + sysfs__mountpoint()); + + if (filename__read_int(path, &id)) { + fprintf(stderr, "error: failed to get tracepoint id: %s\n", path); + return -1; + } + + attr.config = id; + + err = pipe(go_pipe); + __T("failed to create pipe", err == 0); + + fflush(NULL); + + pid = fork(); + if (!pid) { + int i; + + read(go_pipe[0], &bf, 1); + + /* Generate 100 prctl calls. */ + for (i = 0; i < 100; i++) + prctl(0, 0, 0, 0, 0); + + exit(0); + } + + threads = perf_thread_map__new_dummy(); + __T("failed to create threads", threads); + + cpus = perf_cpu_map__dummy_new(); + __T("failed to create cpus", cpus); + + perf_thread_map__set_pid(threads, 0, pid); + + evlist = perf_evlist__new(); + __T("failed to create evlist", evlist); + + evsel = perf_evsel__new(&attr); + __T("failed to create evsel1", evsel); + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, cpus, threads); + + err = perf_evlist__open(evlist); + __T("failed to open evlist", err == 0); + + err = perf_evlist__mmap(evlist, 4); + __T("failed to mmap evlist", err == 0); + + perf_evlist__enable(evlist); + + /* kick the child and wait for it to finish */ + write(go_pipe[1], &bf, 1); + waitpid(pid, NULL, 0); + + /* + * There's no need to call perf_evlist__disable, + * monitored process is dead now. + */ + + perf_evlist__for_each_mmap(evlist, map, false) { + if (perf_mmap__read_init(map) < 0) + continue; + + while ((event = perf_mmap__read_event(map)) != NULL) { + count++; + perf_mmap__consume(map); + } + + perf_mmap__read_done(map); + } + + /* calls perf_evlist__munmap/perf_evlist__close */ + perf_evlist__delete(evlist); + + perf_thread_map__put(threads); + perf_cpu_map__put(cpus); + + /* + * The generated prctl calls should match the + * number of events in the buffer. + */ + __T("failed count", count == 100); + + return 0; +} + +static int test_mmap_cpus(void) +{ + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_mmap *map; + struct perf_cpu_map *cpus; + struct perf_event_attr attr = { + .type = PERF_TYPE_TRACEPOINT, + .sample_period = 1, + .wakeup_watermark = 1, + .disabled = 1, + }; + cpu_set_t saved_mask; + char path[PATH_MAX]; + int id, err, cpu, tmp; + union perf_event *event; + int count = 0; + + snprintf(path, PATH_MAX, "%s/kernel/debug/tracing/events/syscalls/sys_enter_prctl/id", + sysfs__mountpoint()); + + if (filename__read_int(path, &id)) { + fprintf(stderr, "error: failed to get tracepoint id: %s\n", path); + return -1; + } + + attr.config = id; + + cpus = perf_cpu_map__new(NULL); + __T("failed to create cpus", cpus); + + evlist = perf_evlist__new(); + __T("failed to create evlist", evlist); + + evsel = perf_evsel__new(&attr); + __T("failed to create evsel1", evsel); + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, cpus, NULL); + + err = perf_evlist__open(evlist); + __T("failed to open evlist", err == 0); + + err = perf_evlist__mmap(evlist, 4); + __T("failed to mmap evlist", err == 0); + + perf_evlist__enable(evlist); + + err = sched_getaffinity(0, sizeof(saved_mask), &saved_mask); + __T("sched_getaffinity failed", err == 0); + + perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { + cpu_set_t mask; + + CPU_ZERO(&mask); + CPU_SET(cpu, &mask); + + err = sched_setaffinity(0, sizeof(mask), &mask); + __T("sched_setaffinity failed", err == 0); + + prctl(0, 0, 0, 0, 0); + } + + err = sched_setaffinity(0, sizeof(saved_mask), &saved_mask); + __T("sched_setaffinity failed", err == 0); + + perf_evlist__disable(evlist); + + perf_evlist__for_each_mmap(evlist, map, false) { + if (perf_mmap__read_init(map) < 0) + continue; + + while ((event = perf_mmap__read_event(map)) != NULL) { + count++; + perf_mmap__consume(map); + } + + perf_mmap__read_done(map); + } + + /* calls perf_evlist__munmap/perf_evlist__close */ + perf_evlist__delete(evlist); + + /* + * The generated prctl events should match the + * number of cpus or be bigger (we are system-wide). + */ + __T("failed count", count >= perf_cpu_map__nr(cpus)); + + perf_cpu_map__put(cpus); + + return 0; +} + +int main(int argc, char **argv) +{ + __T_START; + + libperf_init(libperf_print); + + test_stat_cpu(); + test_stat_thread(); + test_stat_thread_enable(); + test_mmap_thread(); + test_mmap_cpus(); + + __T_END; + return 0; +} diff --git a/tools/lib/perf/tests/test-evsel.c b/tools/lib/perf/tests/test-evsel.c new file mode 100644 index 0000000..135722a --- /dev/null +++ b/tools/lib/perf/tests/test-evsel.c @@ -0,0 +1,135 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <stdarg.h> +#include <stdio.h> +#include <linux/perf_event.h> +#include <perf/cpumap.h> +#include <perf/threadmap.h> +#include <perf/evsel.h> +#include <internal/tests.h> + +static int libperf_print(enum libperf_print_level level, + const char *fmt, va_list ap) +{ + return vfprintf(stderr, fmt, ap); +} + +static int test_stat_cpu(void) +{ + struct perf_cpu_map *cpus; + struct perf_evsel *evsel; + struct perf_event_attr attr = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_CPU_CLOCK, + }; + int err, cpu, tmp; + + cpus = perf_cpu_map__new(NULL); + __T("failed to create cpus", cpus); + + evsel = perf_evsel__new(&attr); + __T("failed to create evsel", evsel); + + err = perf_evsel__open(evsel, cpus, NULL); + __T("failed to open evsel", err == 0); + + perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { + struct perf_counts_values counts = { .val = 0 }; + + perf_evsel__read(evsel, cpu, 0, &counts); + __T("failed to read value for evsel", counts.val != 0); + } + + perf_evsel__close(evsel); + perf_evsel__delete(evsel); + + perf_cpu_map__put(cpus); + return 0; +} + +static int test_stat_thread(void) +{ + struct perf_counts_values counts = { .val = 0 }; + struct perf_thread_map *threads; + struct perf_evsel *evsel; + struct perf_event_attr attr = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_TASK_CLOCK, + }; + int err; + + threads = perf_thread_map__new_dummy(); + __T("failed to create threads", threads); + + perf_thread_map__set_pid(threads, 0, 0); + + evsel = perf_evsel__new(&attr); + __T("failed to create evsel", evsel); + + err = perf_evsel__open(evsel, NULL, threads); + __T("failed to open evsel", err == 0); + + perf_evsel__read(evsel, 0, 0, &counts); + __T("failed to read value for evsel", counts.val != 0); + + perf_evsel__close(evsel); + perf_evsel__delete(evsel); + + perf_thread_map__put(threads); + return 0; +} + +static int test_stat_thread_enable(void) +{ + struct perf_counts_values counts = { .val = 0 }; + struct perf_thread_map *threads; + struct perf_evsel *evsel; + struct perf_event_attr attr = { + .type = PERF_TYPE_SOFTWARE, + .config = PERF_COUNT_SW_TASK_CLOCK, + .disabled = 1, + }; + int err; + + threads = perf_thread_map__new_dummy(); + __T("failed to create threads", threads); + + perf_thread_map__set_pid(threads, 0, 0); + + evsel = perf_evsel__new(&attr); + __T("failed to create evsel", evsel); + + err = perf_evsel__open(evsel, NULL, threads); + __T("failed to open evsel", err == 0); + + perf_evsel__read(evsel, 0, 0, &counts); + __T("failed to read value for evsel", counts.val == 0); + + err = perf_evsel__enable(evsel); + __T("failed to enable evsel", err == 0); + + perf_evsel__read(evsel, 0, 0, &counts); + __T("failed to read value for evsel", counts.val != 0); + + err = perf_evsel__disable(evsel); + __T("failed to enable evsel", err == 0); + + perf_evsel__close(evsel); + perf_evsel__delete(evsel); + + perf_thread_map__put(threads); + return 0; +} + +int main(int argc, char **argv) +{ + __T_START; + + libperf_init(libperf_print); + + test_stat_cpu(); + test_stat_thread(); + test_stat_thread_enable(); + + __T_END; + return 0; +} diff --git a/tools/lib/perf/tests/test-threadmap.c b/tools/lib/perf/tests/test-threadmap.c new file mode 100644 index 0000000..7dc4d6f --- /dev/null +++ b/tools/lib/perf/tests/test-threadmap.c @@ -0,0 +1,31 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <stdarg.h> +#include <stdio.h> +#include <perf/threadmap.h> +#include <internal/tests.h> + +static int libperf_print(enum libperf_print_level level, + const char *fmt, va_list ap) +{ + return vfprintf(stderr, fmt, ap); +} + +int main(int argc, char **argv) +{ + struct perf_thread_map *threads; + + __T_START; + + libperf_init(libperf_print); + + threads = perf_thread_map__new_dummy(); + if (!threads) + return -1; + + perf_thread_map__get(threads); + perf_thread_map__put(threads); + perf_thread_map__put(threads); + + __T_END; + return 0; +} diff --git a/tools/lib/perf/threadmap.c b/tools/lib/perf/threadmap.c new file mode 100644 index 0000000..e92c368 --- /dev/null +++ b/tools/lib/perf/threadmap.c @@ -0,0 +1,91 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <perf/threadmap.h> +#include <stdlib.h> +#include <linux/refcount.h> +#include <internal/threadmap.h> +#include <string.h> +#include <asm/bug.h> +#include <stdio.h> + +static void perf_thread_map__reset(struct perf_thread_map *map, int start, int nr) +{ + size_t size = (nr - start) * sizeof(map->map[0]); + + memset(&map->map[start], 0, size); + map->err_thread = -1; +} + +struct perf_thread_map *perf_thread_map__realloc(struct perf_thread_map *map, int nr) +{ + size_t size = sizeof(*map) + sizeof(map->map[0]) * nr; + int start = map ? map->nr : 0; + + map = realloc(map, size); + /* + * We only realloc to add more items, let's reset new items. + */ + if (map) + perf_thread_map__reset(map, start, nr); + + return map; +} + +#define thread_map__alloc(__nr) perf_thread_map__realloc(NULL, __nr) + +void perf_thread_map__set_pid(struct perf_thread_map *map, int thread, pid_t pid) +{ + map->map[thread].pid = pid; +} + +char *perf_thread_map__comm(struct perf_thread_map *map, int thread) +{ + return map->map[thread].comm; +} + +struct perf_thread_map *perf_thread_map__new_dummy(void) +{ + struct perf_thread_map *threads = thread_map__alloc(1); + + if (threads != NULL) { + perf_thread_map__set_pid(threads, 0, -1); + threads->nr = 1; + refcount_set(&threads->refcnt, 1); + } + return threads; +} + +static void perf_thread_map__delete(struct perf_thread_map *threads) +{ + if (threads) { + int i; + + WARN_ONCE(refcount_read(&threads->refcnt) != 0, + "thread map refcnt unbalanced\n"); + for (i = 0; i < threads->nr; i++) + free(perf_thread_map__comm(threads, i)); + free(threads); + } +} + +struct perf_thread_map *perf_thread_map__get(struct perf_thread_map *map) +{ + if (map) + refcount_inc(&map->refcnt); + return map; +} + +void perf_thread_map__put(struct perf_thread_map *map) +{ + if (map && refcount_dec_and_test(&map->refcnt)) + perf_thread_map__delete(map); +} + +int perf_thread_map__nr(struct perf_thread_map *threads) +{ + return threads ? threads->nr : 1; +} + +pid_t perf_thread_map__pid(struct perf_thread_map *map, int thread) +{ + return map->map[thread].pid; +} diff --git a/tools/lib/perf/xyarray.c b/tools/lib/perf/xyarray.c new file mode 100644 index 0000000..dcd901d --- /dev/null +++ b/tools/lib/perf/xyarray.c @@ -0,0 +1,33 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <internal/xyarray.h> +#include <linux/zalloc.h> +#include <stdlib.h> +#include <string.h> + +struct xyarray *xyarray__new(int xlen, int ylen, size_t entry_size) +{ + size_t row_size = ylen * entry_size; + struct xyarray *xy = zalloc(sizeof(*xy) + xlen * row_size); + + if (xy != NULL) { + xy->entry_size = entry_size; + xy->row_size = row_size; + xy->entries = xlen * ylen; + xy->max_x = xlen; + xy->max_y = ylen; + } + + return xy; +} + +void xyarray__reset(struct xyarray *xy) +{ + size_t n = xy->entries * xy->entry_size; + + memset(xy->contents, 0, n); +} + +void xyarray__delete(struct xyarray *xy) +{ + free(xy); +} diff --git a/tools/perf/MANIFEST b/tools/perf/MANIFEST index 4934edb..5d7b947 100644 --- a/tools/perf/MANIFEST +++ b/tools/perf/MANIFEST @@ -7,6 +7,7 @@ tools/lib/traceevent tools/lib/api tools/lib/bpf tools/lib/subcmd +tools/lib/perf tools/lib/argv_split.c tools/lib/ctype.c tools/lib/hweight.c diff --git a/tools/perf/Makefile.config b/tools/perf/Makefile.config index c90f414..80e55e7 100644 --- a/tools/perf/Makefile.config +++ b/tools/perf/Makefile.config @@ -286,7 +286,7 @@ ifeq ($(DEBUG),0) endif endif -INC_FLAGS += -I$(src-perf)/lib/include +INC_FLAGS += -I$(srctree)/tools/lib/perf/include INC_FLAGS += -I$(src-perf)/util/include INC_FLAGS += -I$(src-perf)/arch/$(SRCARCH)/include INC_FLAGS += -I$(srctree)/tools/include/ diff --git a/tools/perf/Makefile.perf b/tools/perf/Makefile.perf index eae5d5e..3eda9d4 100644 --- a/tools/perf/Makefile.perf +++ b/tools/perf/Makefile.perf @@ -230,7 +230,7 @@ LIB_DIR = $(srctree)/tools/lib/api/ TRACE_EVENT_DIR = $(srctree)/tools/lib/traceevent/ BPF_DIR = $(srctree)/tools/lib/bpf/ SUBCMD_DIR = $(srctree)/tools/lib/subcmd/ -LIBPERF_DIR = $(srctree)/tools/perf/lib/ +LIBPERF_DIR = $(srctree)/tools/lib/perf/ # Set FEATURE_TESTS to 'all' so all possible feature checkers are executed. # Without this setting the output feature dump file misses some features, for diff --git a/tools/perf/lib/Build b/tools/perf/lib/Build deleted file mode 100644 index 2ef9a4e..0000000 --- a/tools/perf/lib/Build +++ /dev/null @@ -1,13 +0,0 @@ -libperf-y += core.o -libperf-y += cpumap.o -libperf-y += threadmap.o -libperf-y += evsel.o -libperf-y += evlist.o -libperf-y += mmap.o -libperf-y += zalloc.o -libperf-y += xyarray.o -libperf-y += lib.o - -$(OUTPUT)zalloc.o: ../../lib/zalloc.c FORCE - $(call rule_mkdir) - $(call if_changed_dep,cc_o_c) diff --git a/tools/perf/lib/Documentation/Makefile b/tools/perf/lib/Documentation/Makefile deleted file mode 100644 index 586425a..0000000 --- a/tools/perf/lib/Documentation/Makefile +++ /dev/null @@ -1,7 +0,0 @@ -all: - rst2man man/libperf.rst > man/libperf.7 - rst2pdf tutorial/tutorial.rst - -clean: - rm -f man/libperf.7 - rm -f tutorial/tutorial.pdf diff --git a/tools/perf/lib/Documentation/man/libperf.rst b/tools/perf/lib/Documentation/man/libperf.rst deleted file mode 100644 index 09a270f..0000000 --- a/tools/perf/lib/Documentation/man/libperf.rst +++ /dev/null @@ -1,100 +0,0 @@ -.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -libperf - -The libperf library provides an API to access the linux kernel perf -events subsystem. It provides the following high level objects: - - - struct perf_cpu_map - - struct perf_thread_map - - struct perf_evlist - - struct perf_evsel - -reference -========= -Function reference by header files: - -perf/core.h ------------ -.. code-block:: c - - typedef int (\*libperf_print_fn_t)(enum libperf_print_level level, - const char \*, va_list ap); - - void libperf_set_print(libperf_print_fn_t fn); - -perf/cpumap.h -------------- -.. code-block:: c - - struct perf_cpu_map \*perf_cpu_map__dummy_new(void); - struct perf_cpu_map \*perf_cpu_map__new(const char \*cpu_list); - struct perf_cpu_map \*perf_cpu_map__read(FILE \*file); - struct perf_cpu_map \*perf_cpu_map__get(struct perf_cpu_map \*map); - void perf_cpu_map__put(struct perf_cpu_map \*map); - int perf_cpu_map__cpu(const struct perf_cpu_map \*cpus, int idx); - int perf_cpu_map__nr(const struct perf_cpu_map \*cpus); - perf_cpu_map__for_each_cpu(cpu, idx, cpus) - -perf/threadmap.h ----------------- -.. code-block:: c - - struct perf_thread_map \*perf_thread_map__new_dummy(void); - void perf_thread_map__set_pid(struct perf_thread_map \*map, int thread, pid_t pid); - char \*perf_thread_map__comm(struct perf_thread_map \*map, int thread); - struct perf_thread_map \*perf_thread_map__get(struct perf_thread_map \*map); - void perf_thread_map__put(struct perf_thread_map \*map); - -perf/evlist.h -------------- -.. code-block:: - - void perf_evlist__init(struct perf_evlist \*evlist); - void perf_evlist__add(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - void perf_evlist__remove(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - struct perf_evlist \*perf_evlist__new(void); - void perf_evlist__delete(struct perf_evlist \*evlist); - struct perf_evsel\* perf_evlist__next(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - int perf_evlist__open(struct perf_evlist \*evlist); - void perf_evlist__close(struct perf_evlist \*evlist); - void perf_evlist__enable(struct perf_evlist \*evlist); - void perf_evlist__disable(struct perf_evlist \*evlist); - perf_evlist__for_each_evsel(evlist, pos) - void perf_evlist__set_maps(struct perf_evlist \*evlist, - struct perf_cpu_map \*cpus, - struct perf_thread_map \*threads); - -perf/evsel.h ------------- -.. code-block:: c - - struct perf_counts_values { - union { - struct { - uint64_t val; - uint64_t ena; - uint64_t run; - }; - uint64_t values[3]; - }; - }; - - void perf_evsel__init(struct perf_evsel \*evsel, - struct perf_event_attr \*attr); - struct perf_evsel \*perf_evsel__new(struct perf_event_attr \*attr); - void perf_evsel__delete(struct perf_evsel \*evsel); - int perf_evsel__open(struct perf_evsel \*evsel, struct perf_cpu_map \*cpus, - struct perf_thread_map \*threads); - void perf_evsel__close(struct perf_evsel \*evsel); - int perf_evsel__read(struct perf_evsel \*evsel, int cpu, int thread, - struct perf_counts_values \*count); - int perf_evsel__enable(struct perf_evsel \*evsel); - int perf_evsel__disable(struct perf_evsel \*evsel); - int perf_evsel__apply_filter(struct perf_evsel \*evsel, const char \*filter); - struct perf_cpu_map \*perf_evsel__cpus(struct perf_evsel \*evsel); - struct perf_thread_map \*perf_evsel__threads(struct perf_evsel \*evsel); - struct perf_event_attr \*perf_evsel__attr(struct perf_evsel \*evsel); diff --git a/tools/perf/lib/Documentation/tutorial/tutorial.rst b/tools/perf/lib/Documentation/tutorial/tutorial.rst deleted file mode 100644 index 7be7bc2..0000000 --- a/tools/perf/lib/Documentation/tutorial/tutorial.rst +++ /dev/null @@ -1,123 +0,0 @@ -.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -libperf tutorial -================ - -Compile and install libperf from kernel sources -=============================================== -.. code-block:: bash - - git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git - cd linux/tools/perf/lib - make - sudo make install prefix=/usr - -Libperf object -============== -The libperf library provides several high level objects: - -struct perf_cpu_map - Provides a cpu list abstraction. - -struct perf_thread_map - Provides a thread list abstraction. - -struct perf_evsel - Provides an abstraction for single a perf event. - -struct perf_evlist - Gathers several struct perf_evsel object and performs functions on all of them. - -The exported API binds these objects together, -for full reference see the libperf.7 man page. - -Examples -======== -Examples aim to explain libperf functionality on simple use cases. -They are based in on a checked out linux kernel git tree: - -.. code-block:: bash - - $ cd tools/perf/lib/Documentation/tutorial/ - $ ls -d ex-* - ex-1-compile ex-2-evsel-stat ex-3-evlist-stat - -ex-1-compile example -==================== -This example shows the basic usage of *struct perf_cpu_map*, -how to create it and display its cpus: - -.. code-block:: bash - - $ cd ex-1-compile/ - $ make - gcc -o test test.c -lperf - $ ./test - 0 1 2 3 4 5 6 7 - - -The full code listing is here: - -.. code-block:: c - - 1 #include <perf/cpumap.h> - 2 - 3 int main(int argc, char **Argv) - 4 { - 5 struct perf_cpu_map *cpus; - 6 int cpu, tmp; - 7 - 8 cpus = perf_cpu_map__new(NULL); - 9 - 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) - 11 fprintf(stdout, "%d ", cpu); - 12 - 13 fprintf(stdout, "\n"); - 14 - 15 perf_cpu_map__put(cpus); - 16 return 0; - 17 } - - -First you need to include the proper header to have *struct perf_cpumap* -declaration and functions: - -.. code-block:: c - - 1 #include <perf/cpumap.h> - - -The *struct perf_cpumap* object is created by *perf_cpu_map__new* call. -The *NULL* argument asks it to populate the object with the current online CPUs list: - -.. code-block:: c - - 8 cpus = perf_cpu_map__new(NULL); - -This is paired with a *perf_cpu_map__put*, that drops its reference at the end, possibly deleting it. - -.. code-block:: c - - 15 perf_cpu_map__put(cpus); - -The iteration through the *struct perf_cpumap* CPUs is done using the *perf_cpu_map__for_each_cpu* -macro which requires 3 arguments: - -- cpu - the cpu numer -- tmp - iteration helper variable -- cpus - the *struct perf_cpumap* object - -.. code-block:: c - - 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) - 11 fprintf(stdout, "%d ", cpu); - -ex-2-evsel-stat example -======================= - -TBD - -ex-3-evlist-stat example -======================== - -TBD diff --git a/tools/perf/lib/Makefile b/tools/perf/lib/Makefile deleted file mode 100644 index 0f23363..0000000 --- a/tools/perf/lib/Makefile +++ /dev/null @@ -1,188 +0,0 @@ -# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) -# Most of this file is copied from tools/lib/bpf/Makefile - -LIBPERF_VERSION = 0 -LIBPERF_PATCHLEVEL = 0 -LIBPERF_EXTRAVERSION = 1 - -MAKEFLAGS += --no-print-directory - -ifeq ($(srctree),) -srctree := $(patsubst %/,%,$(dir $(CURDIR))) -srctree := $(patsubst %/,%,$(dir $(srctree))) -srctree := $(patsubst %/,%,$(dir $(srctree))) -#$(info Determined 'srctree' to be $(srctree)) -endif - -INSTALL = install - -# Use DESTDIR for installing into a different root directory. -# This is useful for building a package. The program will be -# installed in this directory as if it was the root directory. -# Then the build tool can move it later. -DESTDIR ?= -DESTDIR_SQ = '$(subst ','\'',$(DESTDIR))' - -include $(srctree)/tools/scripts/Makefile.include -include $(srctree)/tools/scripts/Makefile.arch - -ifeq ($(LP64), 1) - libdir_relative = lib64 -else - libdir_relative = lib -endif - -prefix ?= -libdir = $(prefix)/$(libdir_relative) - -# Shell quotes -libdir_SQ = $(subst ','\'',$(libdir)) -libdir_relative_SQ = $(subst ','\'',$(libdir_relative)) - -ifeq ("$(origin V)", "command line") - VERBOSE = $(V) -endif -ifndef VERBOSE - VERBOSE = 0 -endif - -ifeq ($(VERBOSE),1) - Q = -else - Q = @ -endif - -# Set compile option CFLAGS -ifdef EXTRA_CFLAGS - CFLAGS := $(EXTRA_CFLAGS) -else - CFLAGS := -g -Wall -endif - -INCLUDES = \ --I$(srctree)/tools/perf/lib/include \ --I$(srctree)/tools/lib/ \ --I$(srctree)/tools/include \ --I$(srctree)/tools/arch/$(SRCARCH)/include/ \ --I$(srctree)/tools/arch/$(SRCARCH)/include/uapi \ --I$(srctree)/tools/include/uapi - -# Append required CFLAGS -override CFLAGS += $(EXTRA_WARNINGS) -override CFLAGS += -Werror -Wall -override CFLAGS += -fPIC -override CFLAGS += $(INCLUDES) -override CFLAGS += -fvisibility=hidden - -all: - -export srctree OUTPUT CC LD CFLAGS V -export DESTDIR DESTDIR_SQ - -include $(srctree)/tools/build/Makefile.include - -VERSION_SCRIPT := libperf.map - -PATCHLEVEL = $(LIBPERF_PATCHLEVEL) -EXTRAVERSION = $(LIBPERF_EXTRAVERSION) -VERSION = $(LIBPERF_VERSION).$(LIBPERF_PATCHLEVEL).$(LIBPERF_EXTRAVERSION) - -LIBPERF_SO := $(OUTPUT)libperf.so.$(VERSION) -LIBPERF_A := $(OUTPUT)libperf.a -LIBPERF_IN := $(OUTPUT)libperf-in.o -LIBPERF_PC := $(OUTPUT)libperf.pc - -LIBPERF_ALL := $(LIBPERF_A) $(OUTPUT)libperf.so* - -LIB_DIR := $(srctree)/tools/lib/api/ - -ifneq ($(OUTPUT),) -ifneq ($(subdir),) - API_PATH=$(OUTPUT)/../lib/api/ -else - API_PATH=$(OUTPUT) -endif -else - API_PATH=$(LIB_DIR) -endif - -LIBAPI = $(API_PATH)libapi.a -export LIBAPI - -$(LIBAPI): FORCE - $(Q)$(MAKE) -C $(LIB_DIR) O=$(OUTPUT) $(OUTPUT)libapi.a - -$(LIBAPI)-clean: - $(call QUIET_CLEAN, libapi) - $(Q)$(MAKE) -C $(LIB_DIR) O=$(OUTPUT) clean >/dev/null - -$(LIBPERF_IN): FORCE - $(Q)$(MAKE) $(build)=libperf - -$(LIBPERF_A): $(LIBPERF_IN) - $(QUIET_AR)$(RM) $@ && $(AR) rcs $@ $(LIBPERF_IN) - -$(LIBPERF_SO): $(LIBPERF_IN) $(LIBAPI) - $(QUIET_LINK)$(CC) --shared -Wl,-soname,libperf.so \ - -Wl,--version-script=$(VERSION_SCRIPT) $^ -o $@ - @ln -sf $(@F) $(OUTPUT)libperf.so - @ln -sf $(@F) $(OUTPUT)libperf.so.$(LIBPERF_VERSION) - - -libs: $(LIBPERF_A) $(LIBPERF_SO) $(LIBPERF_PC) - -all: fixdep - $(Q)$(MAKE) libs - -clean: $(LIBAPI)-clean - $(call QUIET_CLEAN, libperf) $(RM) $(LIBPERF_A) \ - *.o *~ *.a *.so *.so.$(VERSION) *.so.$(LIBPERF_VERSION) .*.d .*.cmd LIBPERF-CFLAGS $(LIBPERF_PC) - $(Q)$(MAKE) -C tests clean - -tests: libs - $(Q)$(MAKE) -C tests - $(Q)$(MAKE) -C tests run - -$(LIBPERF_PC): - $(QUIET_GEN)sed -e "s|@PREFIX@|$(prefix)|" \ - -e "s|@LIBDIR@|$(libdir_SQ)|" \ - -e "s|@VERSION@|$(VERSION)|" \ - < libperf.pc.template > $@ - -define do_install_mkdir - if [ ! -d '$(DESTDIR_SQ)$1' ]; then \ - $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$1'; \ - fi -endef - -define do_install - if [ ! -d '$(DESTDIR_SQ)$2' ]; then \ - $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$2'; \ - fi; \ - $(INSTALL) $1 $(if $3,-m $3,) '$(DESTDIR_SQ)$2' -endef - -install_lib: libs - $(call QUIET_INSTALL, $(LIBPERF_ALL)) \ - $(call do_install_mkdir,$(libdir_SQ)); \ - cp -fpR $(LIBPERF_ALL) $(DESTDIR)$(libdir_SQ) - -install_headers: - $(call QUIET_INSTALL, headers) \ - $(call do_install,include/perf/core.h,$(prefix)/include/perf,644); \ - $(call do_install,include/perf/cpumap.h,$(prefix)/include/perf,644); \ - $(call do_install,include/perf/threadmap.h,$(prefix)/include/perf,644); \ - $(call do_install,include/perf/evlist.h,$(prefix)/include/perf,644); \ - $(call do_install,include/perf/evsel.h,$(prefix)/include/perf,644); \ - $(call do_install,include/perf/event.h,$(prefix)/include/perf,644); \ - $(call do_install,include/perf/mmap.h,$(prefix)/include/perf,644); - -install_pkgconfig: $(LIBPERF_PC) - $(call QUIET_INSTALL, $(LIBPERF_PC)) \ - $(call do_install,$(LIBPERF_PC),$(libdir_SQ)/pkgconfig,644) - -install: install_lib install_headers install_pkgconfig - -FORCE: - -.PHONY: all install clean tests FORCE diff --git a/tools/perf/lib/core.c b/tools/perf/lib/core.c deleted file mode 100644 index 58fc894..0000000 --- a/tools/perf/lib/core.c +++ /dev/null @@ -1,38 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only - -#define __printf(a, b) __attribute__((format(printf, a, b))) - -#include <stdio.h> -#include <stdarg.h> -#include <unistd.h> -#include <linux/compiler.h> -#include <perf/core.h> -#include <internal/lib.h> -#include "internal.h" - -static int __base_pr(enum libperf_print_level level __maybe_unused, const char *format, - va_list args) -{ - return vfprintf(stderr, format, args); -} - -static libperf_print_fn_t __libperf_pr = __base_pr; - -__printf(2, 3) -void libperf_print(enum libperf_print_level level, const char *format, ...) -{ - va_list args; - - if (!__libperf_pr) - return; - - va_start(args, format); - __libperf_pr(level, format, args); - va_end(args); -} - -void libperf_init(libperf_print_fn_t fn) -{ - page_size = sysconf(_SC_PAGE_SIZE); - __libperf_pr = fn; -} diff --git a/tools/perf/lib/cpumap.c b/tools/perf/lib/cpumap.c deleted file mode 100644 index f93f4e7..0000000 --- a/tools/perf/lib/cpumap.c +++ /dev/null @@ -1,345 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-only -#include <perf/cpumap.h> -#include <stdlib.h> -#include <linux/refcount.h> -#include <internal/cpumap.h> -#include <asm/bug.h> -#include <stdio.h> -#include <string.h> -#include <unistd.h> -#include <ctype.h> -#include <limits.h> - -struct perf_cpu_map *perf_cpu_map__dummy_new(void) -{ - struct perf_cpu_map *cpus = malloc(sizeof(*cpus) + sizeof(int)); - - if (cpus != NULL) { - cpus->nr = 1; - cpus->map[0] = -1; - refcount_set(&cpus->refcnt, 1); - } - - return cpus; -} - -static void cpu_map__delete(struct perf_cpu_map *map) -{ - if (map) { - WARN_ONCE(refcount_read(&map->refcnt) != 0, - "cpu_map refcnt unbalanced\n"); - free(map); - } -} - -struct perf_cpu_map *perf_cpu_map__get(struct perf_cpu_map *map) -{ - if (map) - refcount_inc(&map->refcnt); - return map; -} - -void perf_cpu_map__put(struct perf_cpu_map *map) -{ - if (map && refcount_dec_and_test(&map->refcnt)) - cpu_map__delete(map); -} - -static struct perf_cpu_map *cpu_map__default_new(void) -{ - struct perf_cpu_map *cpus; - int nr_cpus; - - nr_cpus = sysconf(_SC_NPROCESSORS_ONLN); - if (nr_cpus < 0) - return NULL; - - cpus = malloc(sizeof(*cpus) + nr_cpus * sizeof(int)); - if (cpus != NULL) { - int i; - - for (i = 0; i < nr_cpus; ++i) - cpus->map[i] = i; - - cpus->nr = nr_cpus; - refcount_set(&cpus->refcnt, 1); - } - - return cpus; -} - -static int cmp_int(const void *a, const void *b) -{ - return *(const int *)a - *(const int*)b; -} - -static struct perf_cpu_map *cpu_map__trim_new(int nr_cpus, int *tmp_cpus) -{ - size_t payload_size = nr_cpus * sizeof(int); - struct perf_cpu_map *cpus = malloc(sizeof(*cpus) + payload_size); - int i, j; - - if (cpus != NULL) { - memcpy(cpus->map, tmp_cpus, payload_size); - qsort(cpus->map, nr_cpus, sizeof(int), cmp_int); - /* Remove dups */ - j = 0; - for (i = 0; i < nr_cpus; i++) { - if (i == 0 || cpus->map[i] != cpus->map[i - 1]) - cpus->map[j++] = cpus->map[i]; - } - cpus->nr = j; - assert(j <= nr_cpus); - refcount_set(&cpus->refcnt, 1); - } - - return cpus; -} - -struct perf_cpu_map *perf_cpu_map__read(FILE *file) -{ - struct perf_cpu_map *cpus = NULL; - int nr_cpus = 0; - int *tmp_cpus = NULL, *tmp; - int max_entries = 0; - int n, cpu, prev; - char sep; - - sep = 0; - prev = -1; - for (;;) { - n = fscanf(file, "%u%c", &cpu, &sep); - if (n <= 0) - break; - if (prev >= 0) { - int new_max = nr_cpus + cpu - prev - 1; - - WARN_ONCE(new_max >= MAX_NR_CPUS, "Perf can support %d CPUs. " - "Consider raising MAX_NR_CPUS\n", MAX_NR_CPUS); - - if (new_max >= max_entries) { - max_entries = new_max + MAX_NR_CPUS / 2; - tmp = realloc(tmp_cpus, max_entries * sizeof(int)); - if (tmp == NULL) - goto out_free_tmp; - tmp_cpus = tmp; - } - - while (++prev < cpu) - tmp_cpus[nr_cpus++] = prev; - } - if (nr_cpus == max_entries) { - max_entries += MAX_NR_CPUS; - tmp = realloc(tmp_cpus, max_entries * sizeof(int)); - if (tmp == NULL) - goto out_free_tmp; - tmp_cpus = tmp; - } - - tmp_cpus[nr_cpus++] = cpu; - if (n == 2 && sep == '-') - prev = cpu; - else - prev = -1; - if (n == 1 || sep == '\n') - break; - } - - if (nr_cpus > 0) - cpus = cpu_map__trim_new(nr_cpus, tmp_cpus); - else - cpus = cpu_map__default_new(); -out_free_tmp: - free(tmp_cpus); - return cpus; -} - -static struct perf_cpu_map *cpu_map__read_all_cpu_map(void) -{ - struct perf_cpu_map *cpus = NULL; - FILE *onlnf; - - onlnf = fopen("/sys/devices/system/cpu/online", "r"); - if (!onlnf) - return cpu_map__default_new(); - - cpus = perf_cpu_map__read(onlnf); - fclose(onlnf); - return cpus; -} - -struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list) -{ - struct perf_cpu_map *cpus = NULL; - unsigned long start_cpu, end_cpu = 0; - char *p = NULL; - int i, nr_cpus = 0; - int *tmp_cpus = NULL, *tmp; - int max_entries = 0; - - if (!cpu_list) - return cpu_map__read_all_cpu_map(); - - /* - * must handle the case of empty cpumap to cover - * TOPOLOGY header for NUMA nodes with no CPU - * ( e.g., because of CPU hotplug) - */ - if (!isdigit(*cpu_list) && *cpu_list != '\0') - goto out; - - while (isdigit(*cpu_list)) { - p = NULL; - start_cpu = strtoul(cpu_list, &p, 0); - if (start_cpu >= INT_MAX - || (*p != '\0' && *p != ',' && *p != '-')) - goto invalid; - - if (*p == '-') { - cpu_list = ++p; - p = NULL; - end_cpu = strtoul(cpu_list, &p, 0); - - if (end_cpu >= INT_MAX || (*p != '\0' && *p != ',')) - goto invalid; - - if (end_cpu < start_cpu) - goto invalid; - } else { - end_cpu = start_cpu; - } - - WARN_ONCE(end_cpu >= MAX_NR_CPUS, "Perf can support %d CPUs. " - "Consider raising MAX_NR_CPUS\n", MAX_NR_CPUS); - - for (; start_cpu <= end_cpu; start_cpu++) { - /* check for duplicates */ - for (i = 0; i < nr_cpus; i++) - if (tmp_cpus[i] == (int)start_cpu) - goto invalid; - - if (nr_cpus == max_entries) { - max_entries += MAX_NR_CPUS; - tmp = realloc(tmp_cpus, max_entries * sizeof(int)); - if (tmp == NULL) - goto invalid; - tmp_cpus = tmp; - } - tmp_cpus[nr_cpus++] = (int)start_cpu; - } - if (*p) - ++p; - - cpu_list = p; - } - - if (nr_cpus > 0) - cpus = cpu_map__trim_new(nr_cpus, tmp_cpus); - else if (*cpu_list != '\0') - cpus = cpu_map__default_new(); - else - cpus = perf_cpu_map__dummy_new(); -invalid: - free(tmp_cpus); -out: - return cpus; -} - -int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx) -{ - if (idx < cpus->nr) - return cpus->map[idx]; - - return -1; -} - -int perf_cpu_map__nr(const struct perf_cpu_map *cpus) -{ - return cpus ? cpus->nr : 1; -} - -bool perf_cpu_map__empty(const struct perf_cpu_map *map) -{ - return map ? map->map[0] == -1 : true; -} - -int perf_cpu_map__idx(struct perf_cpu_map *cpus, int cpu) -{ - int i; - - for (i = 0; i < cpus->nr; ++i) { - if (cpus->map[i] == cpu) - return i; - } - - return -1; -} - -int perf_cpu_map__max(struct perf_cpu_map *map) -{ - int i, max = -1; - - for (i = 0; i < map->nr; i++) { - if (map->map[i] > max) - max = map->map[i]; - } - - return max; -} - -/* - * Merge two cpumaps - * - * orig either gets freed and replaced with a new map, or reused - * with no reference count change (similar to "realloc") - * other has its reference count increased. - */ - -struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig, - struct perf_cpu_map *other) -{ - int *tmp_cpus; - int tmp_len; - int i, j, k; - struct perf_cpu_map *merged; - - if (!orig && !other) - return NULL; - if (!orig) { - perf_cpu_map__get(other); - return other; - } - if (!other) - return orig; - if (orig->nr == other->nr && - !memcmp(orig->map, other->map, orig->nr * sizeof(int))) - return orig; - - tmp_len = orig->nr + other->nr; - tmp_cpus = malloc(tmp_len * sizeof(int)); - if (!tmp_cpus) - return NULL; - - /* Standard merge algorithm from wikipedia */ - i = j = k = 0; - while (i < orig->nr && j < other->nr) { - if (orig->map[i] <= other->map[j]) { - if (orig->map[i] == other->map[j]) - j++; - tmp_cpus[k++] = orig->map[i++]; - } else - tmp_cpus[k++] = other->map[j++]; - } - - while (i < orig->nr) - tmp_cpus[k++] = orig->map[i++]; - - while (j < other->nr) - tmp_cpus[k++] = other->map[j++]; - assert(k <= tmp_len); - - merged = cpu_map__trim_new(k, tmp_cpus); - free(tmp_cpus); - perf_cpu_map__put(orig); - return merged; -} diff --git a/tools/perf/lib/evlist.c b/tools/perf/lib/evlist.c deleted file mode 100644 index ae9e65a..0000000 --- a/tools/perf/lib/evlist.c +++ /dev/null @@ -1,641 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <perf/evlist.h> -#include <perf/evsel.h> -#include <linux/bitops.h> -#include <linux/list.h> -#include <linux/hash.h> -#include <sys/ioctl.h> -#include <internal/evlist.h> -#include <internal/evsel.h> -#include <internal/xyarray.h> -#include <internal/mmap.h> -#include <internal/cpumap.h> -#include <internal/threadmap.h> -#include <internal/xyarray.h> -#include <internal/lib.h> -#include <linux/zalloc.h> -#include <sys/ioctl.h> -#include <stdlib.h> -#include <errno.h> -#include <unistd.h> -#include <fcntl.h> -#include <signal.h> -#include <poll.h> -#include <sys/mman.h> -#include <perf/cpumap.h> -#include <perf/threadmap.h> -#include <api/fd/array.h> - -void perf_evlist__init(struct perf_evlist *evlist) -{ - int i; - - for (i = 0; i < PERF_EVLIST__HLIST_SIZE; ++i) - INIT_HLIST_HEAD(&evlist->heads[i]); - INIT_LIST_HEAD(&evlist->entries); - evlist->nr_entries = 0; - fdarray__init(&evlist->pollfd, 64); -} - -static void __perf_evlist__propagate_maps(struct perf_evlist *evlist, - struct perf_evsel *evsel) -{ - /* - * We already have cpus for evsel (via PMU sysfs) so - * keep it, if there's no target cpu list defined. - */ - if (!evsel->own_cpus || evlist->has_user_cpus) { - perf_cpu_map__put(evsel->cpus); - evsel->cpus = perf_cpu_map__get(evlist->cpus); - } else if (evsel->cpus != evsel->own_cpus) { - perf_cpu_map__put(evsel->cpus); - evsel->cpus = perf_cpu_map__get(evsel->own_cpus); - } - - perf_thread_map__put(evsel->threads); - evsel->threads = perf_thread_map__get(evlist->threads); - evlist->all_cpus = perf_cpu_map__merge(evlist->all_cpus, evsel->cpus); -} - -static void perf_evlist__propagate_maps(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel; - - perf_evlist__for_each_evsel(evlist, evsel) - __perf_evlist__propagate_maps(evlist, evsel); -} - -void perf_evlist__add(struct perf_evlist *evlist, - struct perf_evsel *evsel) -{ - list_add_tail(&evsel->node, &evlist->entries); - evlist->nr_entries += 1; - __perf_evlist__propagate_maps(evlist, evsel); -} - -void perf_evlist__remove(struct perf_evlist *evlist, - struct perf_evsel *evsel) -{ - list_del_init(&evsel->node); - evlist->nr_entries -= 1; -} - -struct perf_evlist *perf_evlist__new(void) -{ - struct perf_evlist *evlist = zalloc(sizeof(*evlist)); - - if (evlist != NULL) - perf_evlist__init(evlist); - - return evlist; -} - -struct perf_evsel * -perf_evlist__next(struct perf_evlist *evlist, struct perf_evsel *prev) -{ - struct perf_evsel *next; - - if (!prev) { - next = list_first_entry(&evlist->entries, - struct perf_evsel, - node); - } else { - next = list_next_entry(prev, node); - } - - /* Empty list is noticed here so don't need checking on entry. */ - if (&next->node == &evlist->entries) - return NULL; - - return next; -} - -static void perf_evlist__purge(struct perf_evlist *evlist) -{ - struct perf_evsel *pos, *n; - - perf_evlist__for_each_entry_safe(evlist, n, pos) { - list_del_init(&pos->node); - perf_evsel__delete(pos); - } - - evlist->nr_entries = 0; -} - -void perf_evlist__exit(struct perf_evlist *evlist) -{ - perf_cpu_map__put(evlist->cpus); - perf_thread_map__put(evlist->threads); - evlist->cpus = NULL; - evlist->threads = NULL; - fdarray__exit(&evlist->pollfd); -} - -void perf_evlist__delete(struct perf_evlist *evlist) -{ - if (evlist == NULL) - return; - - perf_evlist__munmap(evlist); - perf_evlist__close(evlist); - perf_evlist__purge(evlist); - perf_evlist__exit(evlist); - free(evlist); -} - -void perf_evlist__set_maps(struct perf_evlist *evlist, - struct perf_cpu_map *cpus, - struct perf_thread_map *threads) -{ - /* - * Allow for the possibility that one or another of the maps isn't being - * changed i.e. don't put it. Note we are assuming the maps that are - * being applied are brand new and evlist is taking ownership of the - * original reference count of 1. If that is not the case it is up to - * the caller to increase the reference count. - */ - if (cpus != evlist->cpus) { - perf_cpu_map__put(evlist->cpus); - evlist->cpus = perf_cpu_map__get(cpus); - } - - if (threads != evlist->threads) { - perf_thread_map__put(evlist->threads); - evlist->threads = perf_thread_map__get(threads); - } - - perf_evlist__propagate_maps(evlist); -} - -int perf_evlist__open(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel; - int err; - - perf_evlist__for_each_entry(evlist, evsel) { - err = perf_evsel__open(evsel, evsel->cpus, evsel->threads); - if (err < 0) - goto out_err; - } - - return 0; - -out_err: - perf_evlist__close(evlist); - return err; -} - -void perf_evlist__close(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel; - - perf_evlist__for_each_entry_reverse(evlist, evsel) - perf_evsel__close(evsel); -} - -void perf_evlist__enable(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel; - - perf_evlist__for_each_entry(evlist, evsel) - perf_evsel__enable(evsel); -} - -void perf_evlist__disable(struct perf_evlist *evlist) -{ - struct perf_evsel *evsel; - - perf_evlist__for_each_entry(evlist, evsel) - perf_evsel__disable(evsel); -} - -u64 perf_evlist__read_format(struct perf_evlist *evlist) -{ - struct perf_evsel *first = perf_evlist__first(evlist); - - return first->attr.read_format; -} - -#define SID(e, x, y) xyarray__entry(e->sample_id, x, y) - -static void perf_evlist__id_hash(struct perf_evlist *evlist, - struct perf_evsel *evsel, - int cpu, int thread, u64 id) -{ - int hash; - struct perf_sample_id *sid = SID(evsel, cpu, thread); - - sid->id = id; - sid->evsel = evsel; - hash = hash_64(sid->id, PERF_EVLIST__HLIST_BITS); - hlist_add_head(&sid->node, &evlist->heads[hash]); -} - -void perf_evlist__id_add(struct perf_evlist *evlist, - struct perf_evsel *evsel, - int cpu, int thread, u64 id) -{ - perf_evlist__id_hash(evlist, evsel, cpu, thread, id); - evsel->id[evsel->ids++] = id; -} - -int perf_evlist__id_add_fd(struct perf_evlist *evlist, - struct perf_evsel *evsel, - int cpu, int thread, int fd) -{ - u64 read_data[4] = { 0, }; - int id_idx = 1; /* The first entry is the counter value */ - u64 id; - int ret; - - ret = ioctl(fd, PERF_EVENT_IOC_ID, &id); - if (!ret) - goto add; - - if (errno != ENOTTY) - return -1; - - /* Legacy way to get event id.. All hail to old kernels! */ - - /* - * This way does not work with group format read, so bail - * out in that case. - */ - if (perf_evlist__read_format(evlist) & PERF_FORMAT_GROUP) - return -1; - - if (!(evsel->attr.read_format & PERF_FORMAT_ID) || - read(fd, &read_data, sizeof(read_data)) == -1) - return -1; - - if (evsel->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) - ++id_idx; - if (evsel->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) - ++id_idx; - - id = read_data[id_idx]; - -add: - perf_evlist__id_add(evlist, evsel, cpu, thread, id); - return 0; -} - -int perf_evlist__alloc_pollfd(struct perf_evlist *evlist) -{ - int nr_cpus = perf_cpu_map__nr(evlist->cpus); - int nr_threads = perf_thread_map__nr(evlist->threads); - int nfds = 0; - struct perf_evsel *evsel; - - perf_evlist__for_each_entry(evlist, evsel) { - if (evsel->system_wide) - nfds += nr_cpus; - else - nfds += nr_cpus * nr_threads; - } - - if (fdarray__available_entries(&evlist->pollfd) < nfds && - fdarray__grow(&evlist->pollfd, nfds) < 0) - return -ENOMEM; - - return 0; -} - -int perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd, - void *ptr, short revent) -{ - int pos = fdarray__add(&evlist->pollfd, fd, revent | POLLERR | POLLHUP); - - if (pos >= 0) { - evlist->pollfd.priv[pos].ptr = ptr; - fcntl(fd, F_SETFL, O_NONBLOCK); - } - - return pos; -} - -static void perf_evlist__munmap_filtered(struct fdarray *fda, int fd, - void *arg __maybe_unused) -{ - struct perf_mmap *map = fda->priv[fd].ptr; - - if (map) - perf_mmap__put(map); -} - -int perf_evlist__filter_pollfd(struct perf_evlist *evlist, short revents_and_mask) -{ - return fdarray__filter(&evlist->pollfd, revents_and_mask, - perf_evlist__munmap_filtered, NULL); -} - -int perf_evlist__poll(struct perf_evlist *evlist, int timeout) -{ - return fdarray__poll(&evlist->pollfd, timeout); -} - -static struct perf_mmap* perf_evlist__alloc_mmap(struct perf_evlist *evlist, bool overwrite) -{ - int i; - struct perf_mmap *map; - - map = zalloc(evlist->nr_mmaps * sizeof(struct perf_mmap)); - if (!map) - return NULL; - - for (i = 0; i < evlist->nr_mmaps; i++) { - struct perf_mmap *prev = i ? &map[i - 1] : NULL; - - /* - * When the perf_mmap() call is made we grab one refcount, plus - * one extra to let perf_mmap__consume() get the last - * events after all real references (perf_mmap__get()) are - * dropped. - * - * Each PERF_EVENT_IOC_SET_OUTPUT points to this mmap and - * thus does perf_mmap__get() on it. - */ - perf_mmap__init(&map[i], prev, overwrite, NULL); - } - - return map; -} - -static void perf_evlist__set_sid_idx(struct perf_evlist *evlist, - struct perf_evsel *evsel, int idx, int cpu, - int thread) -{ - struct perf_sample_id *sid = SID(evsel, cpu, thread); - - sid->idx = idx; - if (evlist->cpus && cpu >= 0) - sid->cpu = evlist->cpus->map[cpu]; - else - sid->cpu = -1; - if (!evsel->system_wide && evlist->threads && thread >= 0) - sid->tid = perf_thread_map__pid(evlist->threads, thread); - else - sid->tid = -1; -} - -static struct perf_mmap* -perf_evlist__mmap_cb_get(struct perf_evlist *evlist, bool overwrite, int idx) -{ - struct perf_mmap *maps; - - maps = overwrite ? evlist->mmap_ovw : evlist->mmap; - - if (!maps) { - maps = perf_evlist__alloc_mmap(evlist, overwrite); - if (!maps) - return NULL; - - if (overwrite) - evlist->mmap_ovw = maps; - else - evlist->mmap = maps; - } - - return &maps[idx]; -} - -#define FD(e, x, y) (*(int *) xyarray__entry(e->fd, x, y)) - -static int -perf_evlist__mmap_cb_mmap(struct perf_mmap *map, struct perf_mmap_param *mp, - int output, int cpu) -{ - return perf_mmap__mmap(map, mp, output, cpu); -} - -static void perf_evlist__set_mmap_first(struct perf_evlist *evlist, struct perf_mmap *map, - bool overwrite) -{ - if (overwrite) - evlist->mmap_ovw_first = map; - else - evlist->mmap_first = map; -} - -static int -mmap_per_evsel(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, - int idx, struct perf_mmap_param *mp, int cpu_idx, - int thread, int *_output, int *_output_overwrite) -{ - int evlist_cpu = perf_cpu_map__cpu(evlist->cpus, cpu_idx); - struct perf_evsel *evsel; - int revent; - - perf_evlist__for_each_entry(evlist, evsel) { - bool overwrite = evsel->attr.write_backward; - struct perf_mmap *map; - int *output, fd, cpu; - - if (evsel->system_wide && thread) - continue; - - cpu = perf_cpu_map__idx(evsel->cpus, evlist_cpu); - if (cpu == -1) - continue; - - map = ops->get(evlist, overwrite, idx); - if (map == NULL) - return -ENOMEM; - - if (overwrite) { - mp->prot = PROT_READ; - output = _output_overwrite; - } else { - mp->prot = PROT_READ | PROT_WRITE; - output = _output; - } - - fd = FD(evsel, cpu, thread); - - if (*output == -1) { - *output = fd; - - /* - * The last one will be done at perf_mmap__consume(), so that we - * make sure we don't prevent tools from consuming every last event in - * the ring buffer. - * - * I.e. we can get the POLLHUP meaning that the fd doesn't exist - * anymore, but the last events for it are still in the ring buffer, - * waiting to be consumed. - * - * Tools can chose to ignore this at their own discretion, but the - * evlist layer can't just drop it when filtering events in - * perf_evlist__filter_pollfd(). - */ - refcount_set(&map->refcnt, 2); - - if (ops->mmap(map, mp, *output, evlist_cpu) < 0) - return -1; - - if (!idx) - perf_evlist__set_mmap_first(evlist, map, overwrite); - } else { - if (ioctl(fd, PERF_EVENT_IOC_SET_OUTPUT, *output) != 0) - return -1; - - perf_mmap__get(map); - } - - revent = !overwrite ? POLLIN : 0; - - if (!evsel->system_wide && - perf_evlist__add_pollfd(evlist, fd, map, revent) < 0) { - perf_mmap__put(map); - return -1; - } - - if (evsel->attr.read_format & PERF_FORMAT_ID) { - if (perf_evlist__id_add_fd(evlist, evsel, cpu, thread, - fd) < 0) - return -1; - perf_evlist__set_sid_idx(evlist, evsel, idx, cpu, - thread); - } - } - - return 0; -} - -static int -mmap_per_thread(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, - struct perf_mmap_param *mp) -{ - int thread; - int nr_threads = perf_thread_map__nr(evlist->threads); - - for (thread = 0; thread < nr_threads; thread++) { - int output = -1; - int output_overwrite = -1; - - if (ops->idx) - ops->idx(evlist, mp, thread, false); - - if (mmap_per_evsel(evlist, ops, thread, mp, 0, thread, - &output, &output_overwrite)) - goto out_unmap; - } - - return 0; - -out_unmap: - perf_evlist__munmap(evlist); - return -1; -} - -static int -mmap_per_cpu(struct perf_evlist *evlist, struct perf_evlist_mmap_ops *ops, - struct perf_mmap_param *mp) -{ - int nr_threads = perf_thread_map__nr(evlist->threads); - int nr_cpus = perf_cpu_map__nr(evlist->cpus); - int cpu, thread; - - for (cpu = 0; cpu < nr_cpus; cpu++) { - int output = -1; - int output_overwrite = -1; - - if (ops->idx) - ops->idx(evlist, mp, cpu, true); - - for (thread = 0; thread < nr_threads; thread++) { - if (mmap_per_evsel(evlist, ops, cpu, mp, cpu, - thread, &output, &output_overwrite)) - goto out_unmap; - } - } - - return 0; - -out_unmap: - perf_evlist__munmap(evlist); - return -1; -} - -static int perf_evlist__nr_mmaps(struct perf_evlist *evlist) -{ - int nr_mmaps; - - nr_mmaps = perf_cpu_map__nr(evlist->cpus); - if (perf_cpu_map__empty(evlist->cpus)) - nr_mmaps = perf_thread_map__nr(evlist->threads); - - return nr_mmaps; -} - -int perf_evlist__mmap_ops(struct perf_evlist *evlist, - struct perf_evlist_mmap_ops *ops, - struct perf_mmap_param *mp) -{ - struct perf_evsel *evsel; - const struct perf_cpu_map *cpus = evlist->cpus; - const struct perf_thread_map *threads = evlist->threads; - - if (!ops || !ops->get || !ops->mmap) - return -EINVAL; - - mp->mask = evlist->mmap_len - page_size - 1; - - evlist->nr_mmaps = perf_evlist__nr_mmaps(evlist); - - perf_evlist__for_each_entry(evlist, evsel) { - if ((evsel->attr.read_format & PERF_FORMAT_ID) && - evsel->sample_id == NULL && - perf_evsel__alloc_id(evsel, perf_cpu_map__nr(cpus), threads->nr) < 0) - return -ENOMEM; - } - - if (evlist->pollfd.entries == NULL && perf_evlist__alloc_pollfd(evlist) < 0) - return -ENOMEM; - - if (perf_cpu_map__empty(cpus)) - return mmap_per_thread(evlist, ops, mp); - - return mmap_per_cpu(evlist, ops, mp); -} - -int perf_evlist__mmap(struct perf_evlist *evlist, int pages) -{ - struct perf_mmap_param mp; - struct perf_evlist_mmap_ops ops = { - .get = perf_evlist__mmap_cb_get, - .mmap = perf_evlist__mmap_cb_mmap, - }; - - evlist->mmap_len = (pages + 1) * page_size; - - return perf_evlist__mmap_ops(evlist, &ops, &mp); -} - -void perf_evlist__munmap(struct perf_evlist *evlist) -{ - int i; - - if (evlist->mmap) { - for (i = 0; i < evlist->nr_mmaps; i++) - perf_mmap__munmap(&evlist->mmap[i]); - } - - if (evlist->mmap_ovw) { - for (i = 0; i < evlist->nr_mmaps; i++) - perf_mmap__munmap(&evlist->mmap_ovw[i]); - } - - zfree(&evlist->mmap); - zfree(&evlist->mmap_ovw); -} - -struct perf_mmap* -perf_evlist__next_mmap(struct perf_evlist *evlist, struct perf_mmap *map, - bool overwrite) -{ - if (map) - return map->next; - - return overwrite ? evlist->mmap_ovw_first : evlist->mmap_first; -} diff --git a/tools/perf/lib/evsel.c b/tools/perf/lib/evsel.c deleted file mode 100644 index 4dc0628..0000000 --- a/tools/perf/lib/evsel.c +++ /dev/null @@ -1,301 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <errno.h> -#include <unistd.h> -#include <sys/syscall.h> -#include <perf/evsel.h> -#include <perf/cpumap.h> -#include <perf/threadmap.h> -#include <linux/list.h> -#include <internal/evsel.h> -#include <linux/zalloc.h> -#include <stdlib.h> -#include <internal/xyarray.h> -#include <internal/cpumap.h> -#include <internal/threadmap.h> -#include <internal/lib.h> -#include <linux/string.h> -#include <sys/ioctl.h> - -void perf_evsel__init(struct perf_evsel *evsel, struct perf_event_attr *attr) -{ - INIT_LIST_HEAD(&evsel->node); - evsel->attr = *attr; -} - -struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr) -{ - struct perf_evsel *evsel = zalloc(sizeof(*evsel)); - - if (evsel != NULL) - perf_evsel__init(evsel, attr); - - return evsel; -} - -void perf_evsel__delete(struct perf_evsel *evsel) -{ - free(evsel); -} - -#define FD(e, x, y) (*(int *) xyarray__entry(e->fd, x, y)) - -int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads) -{ - evsel->fd = xyarray__new(ncpus, nthreads, sizeof(int)); - - if (evsel->fd) { - int cpu, thread; - for (cpu = 0; cpu < ncpus; cpu++) { - for (thread = 0; thread < nthreads; thread++) { - FD(evsel, cpu, thread) = -1; - } - } - } - - return evsel->fd != NULL ? 0 : -ENOMEM; -} - -static int -sys_perf_event_open(struct perf_event_attr *attr, - pid_t pid, int cpu, int group_fd, - unsigned long flags) -{ - return syscall(__NR_perf_event_open, attr, pid, cpu, group_fd, flags); -} - -int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus, - struct perf_thread_map *threads) -{ - int cpu, thread, err = 0; - - if (cpus == NULL) { - static struct perf_cpu_map *empty_cpu_map; - - if (empty_cpu_map == NULL) { - empty_cpu_map = perf_cpu_map__dummy_new(); - if (empty_cpu_map == NULL) - return -ENOMEM; - } - - cpus = empty_cpu_map; - } - - if (threads == NULL) { - static struct perf_thread_map *empty_thread_map; - - if (empty_thread_map == NULL) { - empty_thread_map = perf_thread_map__new_dummy(); - if (empty_thread_map == NULL) - return -ENOMEM; - } - - threads = empty_thread_map; - } - - if (evsel->fd == NULL && - perf_evsel__alloc_fd(evsel, cpus->nr, threads->nr) < 0) - return -ENOMEM; - - for (cpu = 0; cpu < cpus->nr; cpu++) { - for (thread = 0; thread < threads->nr; thread++) { - int fd; - - fd = sys_perf_event_open(&evsel->attr, - threads->map[thread].pid, - cpus->map[cpu], -1, 0); - - if (fd < 0) - return -errno; - - FD(evsel, cpu, thread) = fd; - } - } - - return err; -} - -static void perf_evsel__close_fd_cpu(struct perf_evsel *evsel, int cpu) -{ - int thread; - - for (thread = 0; thread < xyarray__max_y(evsel->fd); ++thread) { - if (FD(evsel, cpu, thread) >= 0) - close(FD(evsel, cpu, thread)); - FD(evsel, cpu, thread) = -1; - } -} - -void perf_evsel__close_fd(struct perf_evsel *evsel) -{ - int cpu; - - for (cpu = 0; cpu < xyarray__max_x(evsel->fd); cpu++) - perf_evsel__close_fd_cpu(evsel, cpu); -} - -void perf_evsel__free_fd(struct perf_evsel *evsel) -{ - xyarray__delete(evsel->fd); - evsel->fd = NULL; -} - -void perf_evsel__close(struct perf_evsel *evsel) -{ - if (evsel->fd == NULL) - return; - - perf_evsel__close_fd(evsel); - perf_evsel__free_fd(evsel); -} - -void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu) -{ - if (evsel->fd == NULL) - return; - - perf_evsel__close_fd_cpu(evsel, cpu); -} - -int perf_evsel__read_size(struct perf_evsel *evsel) -{ - u64 read_format = evsel->attr.read_format; - int entry = sizeof(u64); /* value */ - int size = 0; - int nr = 1; - - if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) - size += sizeof(u64); - - if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) - size += sizeof(u64); - - if (read_format & PERF_FORMAT_ID) - entry += sizeof(u64); - - if (read_format & PERF_FORMAT_GROUP) { - nr = evsel->nr_members; - size += sizeof(u64); - } - - size += entry * nr; - return size; -} - -int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, - struct perf_counts_values *count) -{ - size_t size = perf_evsel__read_size(evsel); - - memset(count, 0, sizeof(*count)); - - if (FD(evsel, cpu, thread) < 0) - return -EINVAL; - - if (readn(FD(evsel, cpu, thread), count->values, size) <= 0) - return -errno; - - return 0; -} - -static int perf_evsel__run_ioctl(struct perf_evsel *evsel, - int ioc, void *arg, - int cpu) -{ - int thread; - - for (thread = 0; thread < xyarray__max_y(evsel->fd); thread++) { - int fd = FD(evsel, cpu, thread), - err = ioctl(fd, ioc, arg); - - if (err) - return err; - } - - return 0; -} - -int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu) -{ - return perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_ENABLE, NULL, cpu); -} - -int perf_evsel__enable(struct perf_evsel *evsel) -{ - int i; - int err = 0; - - for (i = 0; i < xyarray__max_x(evsel->fd) && !err; i++) - err = perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_ENABLE, NULL, i); - return err; -} - -int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu) -{ - return perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_DISABLE, NULL, cpu); -} - -int perf_evsel__disable(struct perf_evsel *evsel) -{ - int i; - int err = 0; - - for (i = 0; i < xyarray__max_x(evsel->fd) && !err; i++) - err = perf_evsel__run_ioctl(evsel, PERF_EVENT_IOC_DISABLE, NULL, i); - return err; -} - -int perf_evsel__apply_filter(struct perf_evsel *evsel, const char *filter) -{ - int err = 0, i; - - for (i = 0; i < evsel->cpus->nr && !err; i++) - err = perf_evsel__run_ioctl(evsel, - PERF_EVENT_IOC_SET_FILTER, - (void *)filter, i); - return err; -} - -struct perf_cpu_map *perf_evsel__cpus(struct perf_evsel *evsel) -{ - return evsel->cpus; -} - -struct perf_thread_map *perf_evsel__threads(struct perf_evsel *evsel) -{ - return evsel->threads; -} - -struct perf_event_attr *perf_evsel__attr(struct perf_evsel *evsel) -{ - return &evsel->attr; -} - -int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads) -{ - if (ncpus == 0 || nthreads == 0) - return 0; - - if (evsel->system_wide) - nthreads = 1; - - evsel->sample_id = xyarray__new(ncpus, nthreads, sizeof(struct perf_sample_id)); - if (evsel->sample_id == NULL) - return -ENOMEM; - - evsel->id = zalloc(ncpus * nthreads * sizeof(u64)); - if (evsel->id == NULL) { - xyarray__delete(evsel->sample_id); - evsel->sample_id = NULL; - return -ENOMEM; - } - - return 0; -} - -void perf_evsel__free_id(struct perf_evsel *evsel) -{ - xyarray__delete(evsel->sample_id); - evsel->sample_id = NULL; - zfree(&evsel->id); - evsel->ids = 0; -} diff --git a/tools/perf/lib/include/internal/cpumap.h b/tools/perf/lib/include/internal/cpumap.h deleted file mode 100644 index 840d403..0000000 --- a/tools/perf/lib/include/internal/cpumap.h +++ /dev/null @@ -1,19 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_CPUMAP_H -#define __LIBPERF_INTERNAL_CPUMAP_H - -#include <linux/refcount.h> - -struct perf_cpu_map { - refcount_t refcnt; - int nr; - int map[]; -}; - -#ifndef MAX_NR_CPUS -#define MAX_NR_CPUS 2048 -#endif - -int perf_cpu_map__idx(struct perf_cpu_map *cpus, int cpu); - -#endif /* __LIBPERF_INTERNAL_CPUMAP_H */ diff --git a/tools/perf/lib/include/internal/evlist.h b/tools/perf/lib/include/internal/evlist.h deleted file mode 100644 index 74dc8c3..0000000 --- a/tools/perf/lib/include/internal/evlist.h +++ /dev/null @@ -1,127 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_EVLIST_H -#define __LIBPERF_INTERNAL_EVLIST_H - -#include <linux/list.h> -#include <api/fd/array.h> -#include <internal/evsel.h> - -#define PERF_EVLIST__HLIST_BITS 8 -#define PERF_EVLIST__HLIST_SIZE (1 << PERF_EVLIST__HLIST_BITS) - -struct perf_cpu_map; -struct perf_thread_map; -struct perf_mmap_param; - -struct perf_evlist { - struct list_head entries; - int nr_entries; - bool has_user_cpus; - struct perf_cpu_map *cpus; - struct perf_cpu_map *all_cpus; - struct perf_thread_map *threads; - int nr_mmaps; - size_t mmap_len; - struct fdarray pollfd; - struct hlist_head heads[PERF_EVLIST__HLIST_SIZE]; - struct perf_mmap *mmap; - struct perf_mmap *mmap_ovw; - struct perf_mmap *mmap_first; - struct perf_mmap *mmap_ovw_first; -}; - -typedef void -(*perf_evlist_mmap__cb_idx_t)(struct perf_evlist*, struct perf_mmap_param*, int, bool); -typedef struct perf_mmap* -(*perf_evlist_mmap__cb_get_t)(struct perf_evlist*, bool, int); -typedef int -(*perf_evlist_mmap__cb_mmap_t)(struct perf_mmap*, struct perf_mmap_param*, int, int); - -struct perf_evlist_mmap_ops { - perf_evlist_mmap__cb_idx_t idx; - perf_evlist_mmap__cb_get_t get; - perf_evlist_mmap__cb_mmap_t mmap; -}; - -int perf_evlist__alloc_pollfd(struct perf_evlist *evlist); -int perf_evlist__add_pollfd(struct perf_evlist *evlist, int fd, - void *ptr, short revent); - -int perf_evlist__mmap_ops(struct perf_evlist *evlist, - struct perf_evlist_mmap_ops *ops, - struct perf_mmap_param *mp); - -void perf_evlist__init(struct perf_evlist *evlist); -void perf_evlist__exit(struct perf_evlist *evlist); - -/** - * __perf_evlist__for_each_entry - iterate thru all the evsels - * @list: list_head instance to iterate - * @evsel: struct perf_evsel iterator - */ -#define __perf_evlist__for_each_entry(list, evsel) \ - list_for_each_entry(evsel, list, node) - -/** - * evlist__for_each_entry - iterate thru all the evsels - * @evlist: perf_evlist instance to iterate - * @evsel: struct perf_evsel iterator - */ -#define perf_evlist__for_each_entry(evlist, evsel) \ - __perf_evlist__for_each_entry(&(evlist)->entries, evsel) - -/** - * __perf_evlist__for_each_entry_reverse - iterate thru all the evsels in reverse order - * @list: list_head instance to iterate - * @evsel: struct evsel iterator - */ -#define __perf_evlist__for_each_entry_reverse(list, evsel) \ - list_for_each_entry_reverse(evsel, list, node) - -/** - * perf_evlist__for_each_entry_reverse - iterate thru all the evsels in reverse order - * @evlist: evlist instance to iterate - * @evsel: struct evsel iterator - */ -#define perf_evlist__for_each_entry_reverse(evlist, evsel) \ - __perf_evlist__for_each_entry_reverse(&(evlist)->entries, evsel) - -/** - * __perf_evlist__for_each_entry_safe - safely iterate thru all the evsels - * @list: list_head instance to iterate - * @tmp: struct evsel temp iterator - * @evsel: struct evsel iterator - */ -#define __perf_evlist__for_each_entry_safe(list, tmp, evsel) \ - list_for_each_entry_safe(evsel, tmp, list, node) - -/** - * perf_evlist__for_each_entry_safe - safely iterate thru all the evsels - * @evlist: evlist instance to iterate - * @evsel: struct evsel iterator - * @tmp: struct evsel temp iterator - */ -#define perf_evlist__for_each_entry_safe(evlist, tmp, evsel) \ - __perf_evlist__for_each_entry_safe(&(evlist)->entries, tmp, evsel) - -static inline struct perf_evsel *perf_evlist__first(struct perf_evlist *evlist) -{ - return list_entry(evlist->entries.next, struct perf_evsel, node); -} - -static inline struct perf_evsel *perf_evlist__last(struct perf_evlist *evlist) -{ - return list_entry(evlist->entries.prev, struct perf_evsel, node); -} - -u64 perf_evlist__read_format(struct perf_evlist *evlist); - -void perf_evlist__id_add(struct perf_evlist *evlist, - struct perf_evsel *evsel, - int cpu, int thread, u64 id); - -int perf_evlist__id_add_fd(struct perf_evlist *evlist, - struct perf_evsel *evsel, - int cpu, int thread, int fd); - -#endif /* __LIBPERF_INTERNAL_EVLIST_H */ diff --git a/tools/perf/lib/include/internal/evsel.h b/tools/perf/lib/include/internal/evsel.h deleted file mode 100644 index 1ffd083..0000000 --- a/tools/perf/lib/include/internal/evsel.h +++ /dev/null @@ -1,63 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_EVSEL_H -#define __LIBPERF_INTERNAL_EVSEL_H - -#include <linux/types.h> -#include <linux/perf_event.h> -#include <stdbool.h> -#include <sys/types.h> - -struct perf_cpu_map; -struct perf_thread_map; -struct xyarray; - -/* - * Per fd, to map back from PERF_SAMPLE_ID to evsel, only used when there are - * more than one entry in the evlist. - */ -struct perf_sample_id { - struct hlist_node node; - u64 id; - struct perf_evsel *evsel; - /* - * 'idx' will be used for AUX area sampling. A sample will have AUX area - * data that will be queued for decoding, where there are separate - * queues for each CPU (per-cpu tracing) or task (per-thread tracing). - * The sample ID can be used to lookup 'idx' which is effectively the - * queue number. - */ - int idx; - int cpu; - pid_t tid; - - /* Holds total ID period value for PERF_SAMPLE_READ processing. */ - u64 period; -}; - -struct perf_evsel { - struct list_head node; - struct perf_event_attr attr; - struct perf_cpu_map *cpus; - struct perf_cpu_map *own_cpus; - struct perf_thread_map *threads; - struct xyarray *fd; - struct xyarray *sample_id; - u64 *id; - u32 ids; - - /* parse modifier helper */ - int nr_members; - bool system_wide; -}; - -void perf_evsel__init(struct perf_evsel *evsel, struct perf_event_attr *attr); -int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads); -void perf_evsel__close_fd(struct perf_evsel *evsel); -void perf_evsel__free_fd(struct perf_evsel *evsel); -int perf_evsel__read_size(struct perf_evsel *evsel); -int perf_evsel__apply_filter(struct perf_evsel *evsel, const char *filter); - -int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads); -void perf_evsel__free_id(struct perf_evsel *evsel); - -#endif /* __LIBPERF_INTERNAL_EVSEL_H */ diff --git a/tools/perf/lib/include/internal/lib.h b/tools/perf/lib/include/internal/lib.h deleted file mode 100644 index 5175d49..0000000 --- a/tools/perf/lib/include/internal/lib.h +++ /dev/null @@ -1,12 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_LIB_H -#define __LIBPERF_INTERNAL_LIB_H - -#include <sys/types.h> - -extern unsigned int page_size; - -ssize_t readn(int fd, void *buf, size_t n); -ssize_t writen(int fd, const void *buf, size_t n); - -#endif /* __LIBPERF_INTERNAL_CPUMAP_H */ diff --git a/tools/perf/lib/include/internal/mmap.h b/tools/perf/lib/include/internal/mmap.h deleted file mode 100644 index be7556e..0000000 --- a/tools/perf/lib/include/internal/mmap.h +++ /dev/null @@ -1,55 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_MMAP_H -#define __LIBPERF_INTERNAL_MMAP_H - -#include <linux/compiler.h> -#include <linux/refcount.h> -#include <linux/types.h> -#include <stdbool.h> - -/* perf sample has 16 bits size limit */ -#define PERF_SAMPLE_MAX_SIZE (1 << 16) - -struct perf_mmap; - -typedef void (*libperf_unmap_cb_t)(struct perf_mmap *map); - -/** - * struct perf_mmap - perf's ring buffer mmap details - * - * @refcnt - e.g. code using PERF_EVENT_IOC_SET_OUTPUT to share this - */ -struct perf_mmap { - void *base; - int mask; - int fd; - int cpu; - refcount_t refcnt; - u64 prev; - u64 start; - u64 end; - bool overwrite; - u64 flush; - libperf_unmap_cb_t unmap_cb; - char event_copy[PERF_SAMPLE_MAX_SIZE] __aligned(8); - struct perf_mmap *next; -}; - -struct perf_mmap_param { - int prot; - int mask; -}; - -size_t perf_mmap__mmap_len(struct perf_mmap *map); - -void perf_mmap__init(struct perf_mmap *map, struct perf_mmap *prev, - bool overwrite, libperf_unmap_cb_t unmap_cb); -int perf_mmap__mmap(struct perf_mmap *map, struct perf_mmap_param *mp, - int fd, int cpu); -void perf_mmap__munmap(struct perf_mmap *map); -void perf_mmap__get(struct perf_mmap *map); -void perf_mmap__put(struct perf_mmap *map); - -u64 perf_mmap__read_head(struct perf_mmap *map); - -#endif /* __LIBPERF_INTERNAL_MMAP_H */ diff --git a/tools/perf/lib/include/internal/tests.h b/tools/perf/lib/include/internal/tests.h deleted file mode 100644 index 2093e88..0000000 --- a/tools/perf/lib/include/internal/tests.h +++ /dev/null @@ -1,33 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_TESTS_H -#define __LIBPERF_INTERNAL_TESTS_H - -#include <stdio.h> - -int tests_failed; - -#define __T_START \ -do { \ - fprintf(stdout, "- running %s...", __FILE__); \ - fflush(NULL); \ - tests_failed = 0; \ -} while (0) - -#define __T_END \ -do { \ - if (tests_failed) \ - fprintf(stdout, " FAILED (%d)\n", tests_failed); \ - else \ - fprintf(stdout, "OK\n"); \ -} while (0) - -#define __T(text, cond) \ -do { \ - if (!(cond)) { \ - fprintf(stderr, "FAILED %s:%d %s\n", __FILE__, __LINE__, text); \ - tests_failed++; \ - return -1; \ - } \ -} while (0) - -#endif /* __LIBPERF_INTERNAL_TESTS_H */ diff --git a/tools/perf/lib/include/internal/threadmap.h b/tools/perf/lib/include/internal/threadmap.h deleted file mode 100644 index df748ba..0000000 --- a/tools/perf/lib/include/internal/threadmap.h +++ /dev/null @@ -1,23 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_THREADMAP_H -#define __LIBPERF_INTERNAL_THREADMAP_H - -#include <linux/refcount.h> -#include <sys/types.h> -#include <unistd.h> - -struct thread_map_data { - pid_t pid; - char *comm; -}; - -struct perf_thread_map { - refcount_t refcnt; - int nr; - int err_thread; - struct thread_map_data map[]; -}; - -struct perf_thread_map *perf_thread_map__realloc(struct perf_thread_map *map, int nr); - -#endif /* __LIBPERF_INTERNAL_THREADMAP_H */ diff --git a/tools/perf/lib/include/internal/xyarray.h b/tools/perf/lib/include/internal/xyarray.h deleted file mode 100644 index 51e35d6..0000000 --- a/tools/perf/lib/include/internal/xyarray.h +++ /dev/null @@ -1,36 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_XYARRAY_H -#define __LIBPERF_INTERNAL_XYARRAY_H - -#include <linux/compiler.h> -#include <sys/types.h> - -struct xyarray { - size_t row_size; - size_t entry_size; - size_t entries; - size_t max_x; - size_t max_y; - char contents[] __aligned(8); -}; - -struct xyarray *xyarray__new(int xlen, int ylen, size_t entry_size); -void xyarray__delete(struct xyarray *xy); -void xyarray__reset(struct xyarray *xy); - -static inline void *xyarray__entry(struct xyarray *xy, int x, int y) -{ - return &xy->contents[x * xy->row_size + y * xy->entry_size]; -} - -static inline int xyarray__max_y(struct xyarray *xy) -{ - return xy->max_y; -} - -static inline int xyarray__max_x(struct xyarray *xy) -{ - return xy->max_x; -} - -#endif /* __LIBPERF_INTERNAL_XYARRAY_H */ diff --git a/tools/perf/lib/include/perf/core.h b/tools/perf/lib/include/perf/core.h deleted file mode 100644 index a3f6d68..0000000 --- a/tools/perf/lib/include/perf/core.h +++ /dev/null @@ -1,25 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_CORE_H -#define __LIBPERF_CORE_H - -#include <stdarg.h> - -#ifndef LIBPERF_API -#define LIBPERF_API __attribute__((visibility("default"))) -#endif - -enum libperf_print_level { - LIBPERF_ERR, - LIBPERF_WARN, - LIBPERF_INFO, - LIBPERF_DEBUG, - LIBPERF_DEBUG2, - LIBPERF_DEBUG3, -}; - -typedef int (*libperf_print_fn_t)(enum libperf_print_level level, - const char *, va_list ap); - -LIBPERF_API void libperf_init(libperf_print_fn_t fn); - -#endif /* __LIBPERF_CORE_H */ diff --git a/tools/perf/lib/include/perf/cpumap.h b/tools/perf/lib/include/perf/cpumap.h deleted file mode 100644 index 6a17ad7..0000000 --- a/tools/perf/lib/include/perf/cpumap.h +++ /dev/null @@ -1,28 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_CPUMAP_H -#define __LIBPERF_CPUMAP_H - -#include <perf/core.h> -#include <stdio.h> -#include <stdbool.h> - -struct perf_cpu_map; - -LIBPERF_API struct perf_cpu_map *perf_cpu_map__dummy_new(void); -LIBPERF_API struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list); -LIBPERF_API struct perf_cpu_map *perf_cpu_map__read(FILE *file); -LIBPERF_API struct perf_cpu_map *perf_cpu_map__get(struct perf_cpu_map *map); -LIBPERF_API struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig, - struct perf_cpu_map *other); -LIBPERF_API void perf_cpu_map__put(struct perf_cpu_map *map); -LIBPERF_API int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx); -LIBPERF_API int perf_cpu_map__nr(const struct perf_cpu_map *cpus); -LIBPERF_API bool perf_cpu_map__empty(const struct perf_cpu_map *map); -LIBPERF_API int perf_cpu_map__max(struct perf_cpu_map *map); - -#define perf_cpu_map__for_each_cpu(cpu, idx, cpus) \ - for ((idx) = 0, (cpu) = perf_cpu_map__cpu(cpus, idx); \ - (idx) < perf_cpu_map__nr(cpus); \ - (idx)++, (cpu) = perf_cpu_map__cpu(cpus, idx)) - -#endif /* __LIBPERF_CPUMAP_H */ diff --git a/tools/perf/lib/include/perf/event.h b/tools/perf/lib/include/perf/event.h deleted file mode 100644 index 1810689..0000000 --- a/tools/perf/lib/include/perf/event.h +++ /dev/null @@ -1,385 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_EVENT_H -#define __LIBPERF_EVENT_H - -#include <linux/perf_event.h> -#include <linux/types.h> -#include <linux/limits.h> -#include <linux/bpf.h> -#include <sys/types.h> /* pid_t */ - -struct perf_record_mmap { - struct perf_event_header header; - __u32 pid, tid; - __u64 start; - __u64 len; - __u64 pgoff; - char filename[PATH_MAX]; -}; - -struct perf_record_mmap2 { - struct perf_event_header header; - __u32 pid, tid; - __u64 start; - __u64 len; - __u64 pgoff; - __u32 maj; - __u32 min; - __u64 ino; - __u64 ino_generation; - __u32 prot; - __u32 flags; - char filename[PATH_MAX]; -}; - -struct perf_record_comm { - struct perf_event_header header; - __u32 pid, tid; - char comm[16]; -}; - -struct perf_record_namespaces { - struct perf_event_header header; - __u32 pid, tid; - __u64 nr_namespaces; - struct perf_ns_link_info link_info[]; -}; - -struct perf_record_fork { - struct perf_event_header header; - __u32 pid, ppid; - __u32 tid, ptid; - __u64 time; -}; - -struct perf_record_lost { - struct perf_event_header header; - __u64 id; - __u64 lost; -}; - -struct perf_record_lost_samples { - struct perf_event_header header; - __u64 lost; -}; - -/* - * PERF_FORMAT_ENABLED | PERF_FORMAT_RUNNING | PERF_FORMAT_ID - */ -struct perf_record_read { - struct perf_event_header header; - __u32 pid, tid; - __u64 value; - __u64 time_enabled; - __u64 time_running; - __u64 id; -}; - -struct perf_record_throttle { - struct perf_event_header header; - __u64 time; - __u64 id; - __u64 stream_id; -}; - -#ifndef KSYM_NAME_LEN -#define KSYM_NAME_LEN 256 -#endif - -struct perf_record_ksymbol { - struct perf_event_header header; - __u64 addr; - __u32 len; - __u16 ksym_type; - __u16 flags; - char name[KSYM_NAME_LEN]; -}; - -struct perf_record_bpf_event { - struct perf_event_header header; - __u16 type; - __u16 flags; - __u32 id; - - /* for bpf_prog types */ - __u8 tag[BPF_TAG_SIZE]; // prog tag -}; - -struct perf_record_sample { - struct perf_event_header header; - __u64 array[]; -}; - -struct perf_record_switch { - struct perf_event_header header; - __u32 next_prev_pid; - __u32 next_prev_tid; -}; - -struct perf_record_header_attr { - struct perf_event_header header; - struct perf_event_attr attr; - __u64 id[]; -}; - -enum { - PERF_CPU_MAP__CPUS = 0, - PERF_CPU_MAP__MASK = 1, -}; - -struct cpu_map_entries { - __u16 nr; - __u16 cpu[]; -}; - -struct perf_record_record_cpu_map { - __u16 nr; - __u16 long_size; - unsigned long mask[]; -}; - -struct perf_record_cpu_map_data { - __u16 type; - char data[]; -}; - -struct perf_record_cpu_map { - struct perf_event_header header; - struct perf_record_cpu_map_data data; -}; - -enum { - PERF_EVENT_UPDATE__UNIT = 0, - PERF_EVENT_UPDATE__SCALE = 1, - PERF_EVENT_UPDATE__NAME = 2, - PERF_EVENT_UPDATE__CPUS = 3, -}; - -struct perf_record_event_update_cpus { - struct perf_record_cpu_map_data cpus; -}; - -struct perf_record_event_update_scale { - double scale; -}; - -struct perf_record_event_update { - struct perf_event_header header; - __u64 type; - __u64 id; - char data[]; -}; - -#define MAX_EVENT_NAME 64 - -struct perf_trace_event_type { - __u64 event_id; - char name[MAX_EVENT_NAME]; -}; - -struct perf_record_header_event_type { - struct perf_event_header header; - struct perf_trace_event_type event_type; -}; - -struct perf_record_header_tracing_data { - struct perf_event_header header; - __u32 size; -}; - -struct perf_record_header_build_id { - struct perf_event_header header; - pid_t pid; - __u8 build_id[24]; - char filename[]; -}; - -struct id_index_entry { - __u64 id; - __u64 idx; - __u64 cpu; - __u64 tid; -}; - -struct perf_record_id_index { - struct perf_event_header header; - __u64 nr; - struct id_index_entry entries[0]; -}; - -struct perf_record_auxtrace_info { - struct perf_event_header header; - __u32 type; - __u32 reserved__; /* For alignment */ - __u64 priv[]; -}; - -struct perf_record_auxtrace { - struct perf_event_header header; - __u64 size; - __u64 offset; - __u64 reference; - __u32 idx; - __u32 tid; - __u32 cpu; - __u32 reserved__; /* For alignment */ -}; - -#define MAX_AUXTRACE_ERROR_MSG 64 - -struct perf_record_auxtrace_error { - struct perf_event_header header; - __u32 type; - __u32 code; - __u32 cpu; - __u32 pid; - __u32 tid; - __u32 fmt; - __u64 ip; - __u64 time; - char msg[MAX_AUXTRACE_ERROR_MSG]; -}; - -struct perf_record_aux { - struct perf_event_header header; - __u64 aux_offset; - __u64 aux_size; - __u64 flags; -}; - -struct perf_record_itrace_start { - struct perf_event_header header; - __u32 pid; - __u32 tid; -}; - -struct perf_record_thread_map_entry { - __u64 pid; - char comm[16]; -}; - -struct perf_record_thread_map { - struct perf_event_header header; - __u64 nr; - struct perf_record_thread_map_entry entries[]; -}; - -enum { - PERF_STAT_CONFIG_TERM__AGGR_MODE = 0, - PERF_STAT_CONFIG_TERM__INTERVAL = 1, - PERF_STAT_CONFIG_TERM__SCALE = 2, - PERF_STAT_CONFIG_TERM__MAX = 3, -}; - -struct perf_record_stat_config_entry { - __u64 tag; - __u64 val; -}; - -struct perf_record_stat_config { - struct perf_event_header header; - __u64 nr; - struct perf_record_stat_config_entry data[]; -}; - -struct perf_record_stat { - struct perf_event_header header; - - __u64 id; - __u32 cpu; - __u32 thread; - - union { - struct { - __u64 val; - __u64 ena; - __u64 run; - }; - __u64 values[3]; - }; -}; - -struct perf_record_stat_round { - struct perf_event_header header; - __u64 type; - __u64 time; -}; - -struct perf_record_time_conv { - struct perf_event_header header; - __u64 time_shift; - __u64 time_mult; - __u64 time_zero; -}; - -struct perf_record_header_feature { - struct perf_event_header header; - __u64 feat_id; - char data[]; -}; - -struct perf_record_compressed { - struct perf_event_header header; - char data[]; -}; - -enum perf_user_event_type { /* above any possible kernel type */ - PERF_RECORD_USER_TYPE_START = 64, - PERF_RECORD_HEADER_ATTR = 64, - PERF_RECORD_HEADER_EVENT_TYPE = 65, /* deprecated */ - PERF_RECORD_HEADER_TRACING_DATA = 66, - PERF_RECORD_HEADER_BUILD_ID = 67, - PERF_RECORD_FINISHED_ROUND = 68, - PERF_RECORD_ID_INDEX = 69, - PERF_RECORD_AUXTRACE_INFO = 70, - PERF_RECORD_AUXTRACE = 71, - PERF_RECORD_AUXTRACE_ERROR = 72, - PERF_RECORD_THREAD_MAP = 73, - PERF_RECORD_CPU_MAP = 74, - PERF_RECORD_STAT_CONFIG = 75, - PERF_RECORD_STAT = 76, - PERF_RECORD_STAT_ROUND = 77, - PERF_RECORD_EVENT_UPDATE = 78, - PERF_RECORD_TIME_CONV = 79, - PERF_RECORD_HEADER_FEATURE = 80, - PERF_RECORD_COMPRESSED = 81, - PERF_RECORD_HEADER_MAX -}; - -union perf_event { - struct perf_event_header header; - struct perf_record_mmap mmap; - struct perf_record_mmap2 mmap2; - struct perf_record_comm comm; - struct perf_record_namespaces namespaces; - struct perf_record_fork fork; - struct perf_record_lost lost; - struct perf_record_lost_samples lost_samples; - struct perf_record_read read; - struct perf_record_throttle throttle; - struct perf_record_sample sample; - struct perf_record_bpf_event bpf; - struct perf_record_ksymbol ksymbol; - struct perf_record_header_attr attr; - struct perf_record_event_update event_update; - struct perf_record_header_event_type event_type; - struct perf_record_header_tracing_data tracing_data; - struct perf_record_header_build_id build_id; - struct perf_record_id_index id_index; - struct perf_record_auxtrace_info auxtrace_info; - struct perf_record_auxtrace auxtrace; - struct perf_record_auxtrace_error auxtrace_error; - struct perf_record_aux aux; - struct perf_record_itrace_start itrace_start; - struct perf_record_switch context_switch; - struct perf_record_thread_map thread_map; - struct perf_record_cpu_map cpu_map; - struct perf_record_stat_config stat_config; - struct perf_record_stat stat; - struct perf_record_stat_round stat_round; - struct perf_record_time_conv time_conv; - struct perf_record_header_feature feat; - struct perf_record_compressed pack; -}; - -#endif /* __LIBPERF_EVENT_H */ diff --git a/tools/perf/lib/include/perf/evlist.h b/tools/perf/lib/include/perf/evlist.h deleted file mode 100644 index 0a7479d..0000000 --- a/tools/perf/lib/include/perf/evlist.h +++ /dev/null @@ -1,49 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_EVLIST_H -#define __LIBPERF_EVLIST_H - -#include <perf/core.h> -#include <stdbool.h> - -struct perf_evlist; -struct perf_evsel; -struct perf_cpu_map; -struct perf_thread_map; - -LIBPERF_API void perf_evlist__add(struct perf_evlist *evlist, - struct perf_evsel *evsel); -LIBPERF_API void perf_evlist__remove(struct perf_evlist *evlist, - struct perf_evsel *evsel); -LIBPERF_API struct perf_evlist *perf_evlist__new(void); -LIBPERF_API void perf_evlist__delete(struct perf_evlist *evlist); -LIBPERF_API struct perf_evsel* perf_evlist__next(struct perf_evlist *evlist, - struct perf_evsel *evsel); -LIBPERF_API int perf_evlist__open(struct perf_evlist *evlist); -LIBPERF_API void perf_evlist__close(struct perf_evlist *evlist); -LIBPERF_API void perf_evlist__enable(struct perf_evlist *evlist); -LIBPERF_API void perf_evlist__disable(struct perf_evlist *evlist); - -#define perf_evlist__for_each_evsel(evlist, pos) \ - for ((pos) = perf_evlist__next((evlist), NULL); \ - (pos) != NULL; \ - (pos) = perf_evlist__next((evlist), (pos))) - -LIBPERF_API void perf_evlist__set_maps(struct perf_evlist *evlist, - struct perf_cpu_map *cpus, - struct perf_thread_map *threads); -LIBPERF_API int perf_evlist__poll(struct perf_evlist *evlist, int timeout); -LIBPERF_API int perf_evlist__filter_pollfd(struct perf_evlist *evlist, - short revents_and_mask); - -LIBPERF_API int perf_evlist__mmap(struct perf_evlist *evlist, int pages); -LIBPERF_API void perf_evlist__munmap(struct perf_evlist *evlist); - -LIBPERF_API struct perf_mmap *perf_evlist__next_mmap(struct perf_evlist *evlist, - struct perf_mmap *map, - bool overwrite); -#define perf_evlist__for_each_mmap(evlist, pos, overwrite) \ - for ((pos) = perf_evlist__next_mmap((evlist), NULL, overwrite); \ - (pos) != NULL; \ - (pos) = perf_evlist__next_mmap((evlist), (pos), overwrite)) - -#endif /* __LIBPERF_EVLIST_H */ diff --git a/tools/perf/lib/include/perf/evsel.h b/tools/perf/lib/include/perf/evsel.h deleted file mode 100644 index c82ec39..0000000 --- a/tools/perf/lib/include/perf/evsel.h +++ /dev/null @@ -1,40 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_EVSEL_H -#define __LIBPERF_EVSEL_H - -#include <stdint.h> -#include <perf/core.h> - -struct perf_evsel; -struct perf_event_attr; -struct perf_cpu_map; -struct perf_thread_map; - -struct perf_counts_values { - union { - struct { - uint64_t val; - uint64_t ena; - uint64_t run; - }; - uint64_t values[3]; - }; -}; - -LIBPERF_API struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr); -LIBPERF_API void perf_evsel__delete(struct perf_evsel *evsel); -LIBPERF_API int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus, - struct perf_thread_map *threads); -LIBPERF_API void perf_evsel__close(struct perf_evsel *evsel); -LIBPERF_API void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu); -LIBPERF_API int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, - struct perf_counts_values *count); -LIBPERF_API int perf_evsel__enable(struct perf_evsel *evsel); -LIBPERF_API int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu); -LIBPERF_API int perf_evsel__disable(struct perf_evsel *evsel); -LIBPERF_API int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu); -LIBPERF_API struct perf_cpu_map *perf_evsel__cpus(struct perf_evsel *evsel); -LIBPERF_API struct perf_thread_map *perf_evsel__threads(struct perf_evsel *evsel); -LIBPERF_API struct perf_event_attr *perf_evsel__attr(struct perf_evsel *evsel); - -#endif /* __LIBPERF_EVSEL_H */ diff --git a/tools/perf/lib/include/perf/mmap.h b/tools/perf/lib/include/perf/mmap.h deleted file mode 100644 index 9508ad9..0000000 --- a/tools/perf/lib/include/perf/mmap.h +++ /dev/null @@ -1,15 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_MMAP_H -#define __LIBPERF_MMAP_H - -#include <perf/core.h> - -struct perf_mmap; -union perf_event; - -LIBPERF_API void perf_mmap__consume(struct perf_mmap *map); -LIBPERF_API int perf_mmap__read_init(struct perf_mmap *map); -LIBPERF_API void perf_mmap__read_done(struct perf_mmap *map); -LIBPERF_API union perf_event *perf_mmap__read_event(struct perf_mmap *map); - -#endif /* __LIBPERF_MMAP_H */ diff --git a/tools/perf/lib/include/perf/threadmap.h b/tools/perf/lib/include/perf/threadmap.h deleted file mode 100644 index a7c50de..0000000 --- a/tools/perf/lib/include/perf/threadmap.h +++ /dev/null @@ -1,20 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_THREADMAP_H -#define __LIBPERF_THREADMAP_H - -#include <perf/core.h> -#include <sys/types.h> - -struct perf_thread_map; - -LIBPERF_API struct perf_thread_map *perf_thread_map__new_dummy(void); - -LIBPERF_API void perf_thread_map__set_pid(struct perf_thread_map *map, int thread, pid_t pid); -LIBPERF_API char *perf_thread_map__comm(struct perf_thread_map *map, int thread); -LIBPERF_API int perf_thread_map__nr(struct perf_thread_map *threads); -LIBPERF_API pid_t perf_thread_map__pid(struct perf_thread_map *map, int thread); - -LIBPERF_API struct perf_thread_map *perf_thread_map__get(struct perf_thread_map *map); -LIBPERF_API void perf_thread_map__put(struct perf_thread_map *map); - -#endif /* __LIBPERF_THREADMAP_H */ diff --git a/tools/perf/lib/internal.h b/tools/perf/lib/internal.h deleted file mode 100644 index 2c27e15..0000000 --- a/tools/perf/lib/internal.h +++ /dev/null @@ -1,23 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LIBPERF_INTERNAL_H -#define __LIBPERF_INTERNAL_H - -#include <perf/core.h> - -void libperf_print(enum libperf_print_level level, - const char *format, ...) - __attribute__((format(printf, 2, 3))); - -#define __pr(level, fmt, ...) \ -do { \ - libperf_print(level, "libperf: " fmt, ##__VA_ARGS__); \ -} while (0) - -#define pr_err(fmt, ...) __pr(LIBPERF_ERR, fmt, ##__VA_ARGS__) -#define pr_warning(fmt, ...) __pr(LIBPERF_WARN, fmt, ##__VA_ARGS__) -#define pr_info(fmt, ...) __pr(LIBPERF_INFO, fmt, ##__VA_ARGS__) -#define pr_debug(fmt, ...) __pr(LIBPERF_DEBUG, fmt, ##__VA_ARGS__) -#define pr_debug2(fmt, ...) __pr(LIBPERF_DEBUG2, fmt, ##__VA_ARGS__) -#define pr_debug3(fmt, ...) __pr(LIBPERF_DEBUG3, fmt, ##__VA_ARGS__) - -#endif /* __LIBPERF_INTERNAL_H */ diff --git a/tools/perf/lib/lib.c b/tools/perf/lib/lib.c deleted file mode 100644 index 1865893..0000000 --- a/tools/perf/lib/lib.c +++ /dev/null @@ -1,48 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <unistd.h> -#include <stdbool.h> -#include <errno.h> -#include <linux/kernel.h> -#include <internal/lib.h> - -unsigned int page_size; - -static ssize_t ion(bool is_read, int fd, void *buf, size_t n) -{ - void *buf_start = buf; - size_t left = n; - - while (left) { - /* buf must be treated as const if !is_read. */ - ssize_t ret = is_read ? read(fd, buf, left) : - write(fd, buf, left); - - if (ret < 0 && errno == EINTR) - continue; - if (ret <= 0) - return ret; - - left -= ret; - buf += ret; - } - - BUG_ON((size_t)(buf - buf_start) != n); - return n; -} - -/* - * Read exactly 'n' bytes or return an error. - */ -ssize_t readn(int fd, void *buf, size_t n) -{ - return ion(true, fd, buf, n); -} - -/* - * Write exactly 'n' bytes or return an error. - */ -ssize_t writen(int fd, const void *buf, size_t n) -{ - /* ion does not modify buf. */ - return ion(false, fd, (void *)buf, n); -} diff --git a/tools/perf/lib/libperf.map b/tools/perf/lib/libperf.map deleted file mode 100644 index 7be1af8..0000000 --- a/tools/perf/lib/libperf.map +++ /dev/null @@ -1,51 +0,0 @@ -LIBPERF_0.0.1 { - global: - libperf_init; - perf_cpu_map__dummy_new; - perf_cpu_map__get; - perf_cpu_map__put; - perf_cpu_map__new; - perf_cpu_map__read; - perf_cpu_map__nr; - perf_cpu_map__cpu; - perf_cpu_map__empty; - perf_cpu_map__max; - perf_thread_map__new_dummy; - perf_thread_map__set_pid; - perf_thread_map__comm; - perf_thread_map__nr; - perf_thread_map__pid; - perf_thread_map__get; - perf_thread_map__put; - perf_evsel__new; - perf_evsel__delete; - perf_evsel__enable; - perf_evsel__disable; - perf_evsel__open; - perf_evsel__close; - perf_evsel__read; - perf_evsel__cpus; - perf_evsel__threads; - perf_evsel__attr; - perf_evlist__new; - perf_evlist__delete; - perf_evlist__open; - perf_evlist__close; - perf_evlist__enable; - perf_evlist__disable; - perf_evlist__add; - perf_evlist__remove; - perf_evlist__next; - perf_evlist__set_maps; - perf_evlist__poll; - perf_evlist__mmap; - perf_evlist__munmap; - perf_evlist__filter_pollfd; - perf_evlist__next_mmap; - perf_mmap__consume; - perf_mmap__read_init; - perf_mmap__read_done; - perf_mmap__read_event; - local: - *; -}; diff --git a/tools/perf/lib/libperf.pc.template b/tools/perf/lib/libperf.pc.template deleted file mode 100644 index 117e4a2..0000000 --- a/tools/perf/lib/libperf.pc.template +++ /dev/null @@ -1,11 +0,0 @@ -# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -prefix=@PREFIX@ -libdir=@LIBDIR@ -includedir=${prefix}/include - -Name: libperf -Description: perf library -Version: @VERSION@ -Libs: -L${libdir} -lperf -Cflags: -I${includedir} diff --git a/tools/perf/lib/mmap.c b/tools/perf/lib/mmap.c deleted file mode 100644 index 79d5ed6..0000000 --- a/tools/perf/lib/mmap.c +++ /dev/null @@ -1,275 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <sys/mman.h> -#include <inttypes.h> -#include <asm/bug.h> -#include <errno.h> -#include <string.h> -#include <linux/ring_buffer.h> -#include <linux/perf_event.h> -#include <perf/mmap.h> -#include <perf/event.h> -#include <internal/mmap.h> -#include <internal/lib.h> -#include <linux/kernel.h> -#include "internal.h" - -void perf_mmap__init(struct perf_mmap *map, struct perf_mmap *prev, - bool overwrite, libperf_unmap_cb_t unmap_cb) -{ - map->fd = -1; - map->overwrite = overwrite; - map->unmap_cb = unmap_cb; - refcount_set(&map->refcnt, 0); - if (prev) - prev->next = map; -} - -size_t perf_mmap__mmap_len(struct perf_mmap *map) -{ - return map->mask + 1 + page_size; -} - -int perf_mmap__mmap(struct perf_mmap *map, struct perf_mmap_param *mp, - int fd, int cpu) -{ - map->prev = 0; - map->mask = mp->mask; - map->base = mmap(NULL, perf_mmap__mmap_len(map), mp->prot, - MAP_SHARED, fd, 0); - if (map->base == MAP_FAILED) { - map->base = NULL; - return -1; - } - - map->fd = fd; - map->cpu = cpu; - return 0; -} - -void perf_mmap__munmap(struct perf_mmap *map) -{ - if (map && map->base != NULL) { - munmap(map->base, perf_mmap__mmap_len(map)); - map->base = NULL; - map->fd = -1; - refcount_set(&map->refcnt, 0); - } - if (map && map->unmap_cb) - map->unmap_cb(map); -} - -void perf_mmap__get(struct perf_mmap *map) -{ - refcount_inc(&map->refcnt); -} - -void perf_mmap__put(struct perf_mmap *map) -{ - BUG_ON(map->base && refcount_read(&map->refcnt) == 0); - - if (refcount_dec_and_test(&map->refcnt)) - perf_mmap__munmap(map); -} - -static inline void perf_mmap__write_tail(struct perf_mmap *md, u64 tail) -{ - ring_buffer_write_tail(md->base, tail); -} - -u64 perf_mmap__read_head(struct perf_mmap *map) -{ - return ring_buffer_read_head(map->base); -} - -static bool perf_mmap__empty(struct perf_mmap *map) -{ - struct perf_event_mmap_page *pc = map->base; - - return perf_mmap__read_head(map) == map->prev && !pc->aux_size; -} - -void perf_mmap__consume(struct perf_mmap *map) -{ - if (!map->overwrite) { - u64 old = map->prev; - - perf_mmap__write_tail(map, old); - } - - if (refcount_read(&map->refcnt) == 1 && perf_mmap__empty(map)) - perf_mmap__put(map); -} - -static int overwrite_rb_find_range(void *buf, int mask, u64 *start, u64 *end) -{ - struct perf_event_header *pheader; - u64 evt_head = *start; - int size = mask + 1; - - pr_debug2("%s: buf=%p, start=%"PRIx64"\n", __func__, buf, *start); - pheader = (struct perf_event_header *)(buf + (*start & mask)); - while (true) { - if (evt_head - *start >= (unsigned int)size) { - pr_debug("Finished reading overwrite ring buffer: rewind\n"); - if (evt_head - *start > (unsigned int)size) - evt_head -= pheader->size; - *end = evt_head; - return 0; - } - - pheader = (struct perf_event_header *)(buf + (evt_head & mask)); - - if (pheader->size == 0) { - pr_debug("Finished reading overwrite ring buffer: get start\n"); - *end = evt_head; - return 0; - } - - evt_head += pheader->size; - pr_debug3("move evt_head: %"PRIx64"\n", evt_head); - } - WARN_ONCE(1, "Shouldn't get here\n"); - return -1; -} - -/* - * Report the start and end of the available data in ringbuffer - */ -static int __perf_mmap__read_init(struct perf_mmap *md) -{ - u64 head = perf_mmap__read_head(md); - u64 old = md->prev; - unsigned char *data = md->base + page_size; - unsigned long size; - - md->start = md->overwrite ? head : old; - md->end = md->overwrite ? old : head; - - if ((md->end - md->start) < md->flush) - return -EAGAIN; - - size = md->end - md->start; - if (size > (unsigned long)(md->mask) + 1) { - if (!md->overwrite) { - WARN_ONCE(1, "failed to keep up with mmap data. (warn only once)\n"); - - md->prev = head; - perf_mmap__consume(md); - return -EAGAIN; - } - - /* - * Backward ring buffer is full. We still have a chance to read - * most of data from it. - */ - if (overwrite_rb_find_range(data, md->mask, &md->start, &md->end)) - return -EINVAL; - } - - return 0; -} - -int perf_mmap__read_init(struct perf_mmap *map) -{ - /* - * Check if event was unmapped due to a POLLHUP/POLLERR. - */ - if (!refcount_read(&map->refcnt)) - return -ENOENT; - - return __perf_mmap__read_init(map); -} - -/* - * Mandatory for overwrite mode - * The direction of overwrite mode is backward. - * The last perf_mmap__read() will set tail to map->core.prev. - * Need to correct the map->core.prev to head which is the end of next read. - */ -void perf_mmap__read_done(struct perf_mmap *map) -{ - /* - * Check if event was unmapped due to a POLLHUP/POLLERR. - */ - if (!refcount_read(&map->refcnt)) - return; - - map->prev = perf_mmap__read_head(map); -} - -/* When check_messup is true, 'end' must points to a good entry */ -static union perf_event *perf_mmap__read(struct perf_mmap *map, - u64 *startp, u64 end) -{ - unsigned char *data = map->base + page_size; - union perf_event *event = NULL; - int diff = end - *startp; - - if (diff >= (int)sizeof(event->header)) { - size_t size; - - event = (union perf_event *)&data[*startp & map->mask]; - size = event->header.size; - - if (size < sizeof(event->header) || diff < (int)size) - return NULL; - - /* - * Event straddles the mmap boundary -- header should always - * be inside due to u64 alignment of output. - */ - if ((*startp & map->mask) + size != ((*startp + size) & map->mask)) { - unsigned int offset = *startp; - unsigned int len = min(sizeof(*event), size), cpy; - void *dst = map->event_copy; - - do { - cpy = min(map->mask + 1 - (offset & map->mask), len); - memcpy(dst, &data[offset & map->mask], cpy); - offset += cpy; - dst += cpy; - len -= cpy; - } while (len); - - event = (union perf_event *)map->event_copy; - } - - *startp += size; - } - - return event; -} - -/* - * Read event from ring buffer one by one. - * Return one event for each call. - * - * Usage: - * perf_mmap__read_init() - * while(event = perf_mmap__read_event()) { - * //process the event - * perf_mmap__consume() - * } - * perf_mmap__read_done() - */ -union perf_event *perf_mmap__read_event(struct perf_mmap *map) -{ - union perf_event *event; - - /* - * Check if event was unmapped due to a POLLHUP/POLLERR. - */ - if (!refcount_read(&map->refcnt)) - return NULL; - - /* non-overwirte doesn't pause the ringbuffer */ - if (!map->overwrite) - map->end = perf_mmap__read_head(map); - - event = perf_mmap__read(map, &map->start, map->end); - - if (!map->overwrite) - map->prev = map->start; - - return event; -} diff --git a/tools/perf/lib/tests/Makefile b/tools/perf/lib/tests/Makefile deleted file mode 100644 index a43cd08..0000000 --- a/tools/perf/lib/tests/Makefile +++ /dev/null @@ -1,38 +0,0 @@ -# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -TESTS = test-cpumap test-threadmap test-evlist test-evsel - -TESTS_SO := $(addsuffix -so,$(TESTS)) -TESTS_A := $(addsuffix -a,$(TESTS)) - -# Set compile option CFLAGS -ifdef EXTRA_CFLAGS - CFLAGS := $(EXTRA_CFLAGS) -else - CFLAGS := -g -Wall -endif - -all: - -include $(srctree)/tools/scripts/Makefile.include - -INCLUDE = -I$(srctree)/tools/perf/lib/include -I$(srctree)/tools/include -I$(srctree)/tools/lib - -$(TESTS_A): FORCE - $(QUIET_LINK)$(CC) $(INCLUDE) $(CFLAGS) -o $@ $(subst -a,.c,$@) ../libperf.a $(LIBAPI) - -$(TESTS_SO): FORCE - $(QUIET_LINK)$(CC) $(INCLUDE) $(CFLAGS) -L.. -o $@ $(subst -so,.c,$@) $(LIBAPI) -lperf - -all: $(TESTS_A) $(TESTS_SO) - -run: - @echo "running static:" - @for i in $(TESTS_A); do ./$$i; done - @echo "running dynamic:" - @for i in $(TESTS_SO); do LD_LIBRARY_PATH=../ ./$$i; done - -clean: - $(call QUIET_CLEAN, tests)$(RM) $(TESTS_A) $(TESTS_SO) - -.PHONY: all clean FORCE diff --git a/tools/perf/lib/tests/test-cpumap.c b/tools/perf/lib/tests/test-cpumap.c deleted file mode 100644 index c8d4509..0000000 --- a/tools/perf/lib/tests/test-cpumap.c +++ /dev/null @@ -1,31 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <stdarg.h> -#include <stdio.h> -#include <perf/cpumap.h> -#include <internal/tests.h> - -static int libperf_print(enum libperf_print_level level, - const char *fmt, va_list ap) -{ - return vfprintf(stderr, fmt, ap); -} - -int main(int argc, char **argv) -{ - struct perf_cpu_map *cpus; - - __T_START; - - libperf_init(libperf_print); - - cpus = perf_cpu_map__dummy_new(); - if (!cpus) - return -1; - - perf_cpu_map__get(cpus); - perf_cpu_map__put(cpus); - perf_cpu_map__put(cpus); - - __T_END; - return 0; -} diff --git a/tools/perf/lib/tests/test-evlist.c b/tools/perf/lib/tests/test-evlist.c deleted file mode 100644 index 6d8ebe0..0000000 --- a/tools/perf/lib/tests/test-evlist.c +++ /dev/null @@ -1,413 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#define _GNU_SOURCE // needed for sched.h to get sched_[gs]etaffinity and CPU_(ZERO,SET) -#include <sched.h> -#include <stdio.h> -#include <stdarg.h> -#include <unistd.h> -#include <stdlib.h> -#include <linux/perf_event.h> -#include <linux/limits.h> -#include <sys/types.h> -#include <sys/wait.h> -#include <sys/prctl.h> -#include <perf/cpumap.h> -#include <perf/threadmap.h> -#include <perf/evlist.h> -#include <perf/evsel.h> -#include <perf/mmap.h> -#include <perf/event.h> -#include <internal/tests.h> -#include <api/fs/fs.h> - -static int libperf_print(enum libperf_print_level level, - const char *fmt, va_list ap) -{ - return vfprintf(stderr, fmt, ap); -} - -static int test_stat_cpu(void) -{ - struct perf_cpu_map *cpus; - struct perf_evlist *evlist; - struct perf_evsel *evsel; - struct perf_event_attr attr1 = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_CPU_CLOCK, - }; - struct perf_event_attr attr2 = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_TASK_CLOCK, - }; - int err, cpu, tmp; - - cpus = perf_cpu_map__new(NULL); - __T("failed to create cpus", cpus); - - evlist = perf_evlist__new(); - __T("failed to create evlist", evlist); - - evsel = perf_evsel__new(&attr1); - __T("failed to create evsel1", evsel); - - perf_evlist__add(evlist, evsel); - - evsel = perf_evsel__new(&attr2); - __T("failed to create evsel2", evsel); - - perf_evlist__add(evlist, evsel); - - perf_evlist__set_maps(evlist, cpus, NULL); - - err = perf_evlist__open(evlist); - __T("failed to open evsel", err == 0); - - perf_evlist__for_each_evsel(evlist, evsel) { - cpus = perf_evsel__cpus(evsel); - - perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { - struct perf_counts_values counts = { .val = 0 }; - - perf_evsel__read(evsel, cpu, 0, &counts); - __T("failed to read value for evsel", counts.val != 0); - } - } - - perf_evlist__close(evlist); - perf_evlist__delete(evlist); - - perf_cpu_map__put(cpus); - return 0; -} - -static int test_stat_thread(void) -{ - struct perf_counts_values counts = { .val = 0 }; - struct perf_thread_map *threads; - struct perf_evlist *evlist; - struct perf_evsel *evsel; - struct perf_event_attr attr1 = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_CPU_CLOCK, - }; - struct perf_event_attr attr2 = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_TASK_CLOCK, - }; - int err; - - threads = perf_thread_map__new_dummy(); - __T("failed to create threads", threads); - - perf_thread_map__set_pid(threads, 0, 0); - - evlist = perf_evlist__new(); - __T("failed to create evlist", evlist); - - evsel = perf_evsel__new(&attr1); - __T("failed to create evsel1", evsel); - - perf_evlist__add(evlist, evsel); - - evsel = perf_evsel__new(&attr2); - __T("failed to create evsel2", evsel); - - perf_evlist__add(evlist, evsel); - - perf_evlist__set_maps(evlist, NULL, threads); - - err = perf_evlist__open(evlist); - __T("failed to open evsel", err == 0); - - perf_evlist__for_each_evsel(evlist, evsel) { - perf_evsel__read(evsel, 0, 0, &counts); - __T("failed to read value for evsel", counts.val != 0); - } - - perf_evlist__close(evlist); - perf_evlist__delete(evlist); - - perf_thread_map__put(threads); - return 0; -} - -static int test_stat_thread_enable(void) -{ - struct perf_counts_values counts = { .val = 0 }; - struct perf_thread_map *threads; - struct perf_evlist *evlist; - struct perf_evsel *evsel; - struct perf_event_attr attr1 = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_CPU_CLOCK, - .disabled = 1, - }; - struct perf_event_attr attr2 = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_TASK_CLOCK, - .disabled = 1, - }; - int err; - - threads = perf_thread_map__new_dummy(); - __T("failed to create threads", threads); - - perf_thread_map__set_pid(threads, 0, 0); - - evlist = perf_evlist__new(); - __T("failed to create evlist", evlist); - - evsel = perf_evsel__new(&attr1); - __T("failed to create evsel1", evsel); - - perf_evlist__add(evlist, evsel); - - evsel = perf_evsel__new(&attr2); - __T("failed to create evsel2", evsel); - - perf_evlist__add(evlist, evsel); - - perf_evlist__set_maps(evlist, NULL, threads); - - err = perf_evlist__open(evlist); - __T("failed to open evsel", err == 0); - - perf_evlist__for_each_evsel(evlist, evsel) { - perf_evsel__read(evsel, 0, 0, &counts); - __T("failed to read value for evsel", counts.val == 0); - } - - perf_evlist__enable(evlist); - - perf_evlist__for_each_evsel(evlist, evsel) { - perf_evsel__read(evsel, 0, 0, &counts); - __T("failed to read value for evsel", counts.val != 0); - } - - perf_evlist__disable(evlist); - - perf_evlist__close(evlist); - perf_evlist__delete(evlist); - - perf_thread_map__put(threads); - return 0; -} - -static int test_mmap_thread(void) -{ - struct perf_evlist *evlist; - struct perf_evsel *evsel; - struct perf_mmap *map; - struct perf_cpu_map *cpus; - struct perf_thread_map *threads; - struct perf_event_attr attr = { - .type = PERF_TYPE_TRACEPOINT, - .sample_period = 1, - .wakeup_watermark = 1, - .disabled = 1, - }; - char path[PATH_MAX]; - int id, err, pid, go_pipe[2]; - union perf_event *event; - char bf; - int count = 0; - - snprintf(path, PATH_MAX, "%s/kernel/debug/tracing/events/syscalls/sys_enter_prctl/id", - sysfs__mountpoint()); - - if (filename__read_int(path, &id)) { - fprintf(stderr, "error: failed to get tracepoint id: %s\n", path); - return -1; - } - - attr.config = id; - - err = pipe(go_pipe); - __T("failed to create pipe", err == 0); - - fflush(NULL); - - pid = fork(); - if (!pid) { - int i; - - read(go_pipe[0], &bf, 1); - - /* Generate 100 prctl calls. */ - for (i = 0; i < 100; i++) - prctl(0, 0, 0, 0, 0); - - exit(0); - } - - threads = perf_thread_map__new_dummy(); - __T("failed to create threads", threads); - - cpus = perf_cpu_map__dummy_new(); - __T("failed to create cpus", cpus); - - perf_thread_map__set_pid(threads, 0, pid); - - evlist = perf_evlist__new(); - __T("failed to create evlist", evlist); - - evsel = perf_evsel__new(&attr); - __T("failed to create evsel1", evsel); - - perf_evlist__add(evlist, evsel); - - perf_evlist__set_maps(evlist, cpus, threads); - - err = perf_evlist__open(evlist); - __T("failed to open evlist", err == 0); - - err = perf_evlist__mmap(evlist, 4); - __T("failed to mmap evlist", err == 0); - - perf_evlist__enable(evlist); - - /* kick the child and wait for it to finish */ - write(go_pipe[1], &bf, 1); - waitpid(pid, NULL, 0); - - /* - * There's no need to call perf_evlist__disable, - * monitored process is dead now. - */ - - perf_evlist__for_each_mmap(evlist, map, false) { - if (perf_mmap__read_init(map) < 0) - continue; - - while ((event = perf_mmap__read_event(map)) != NULL) { - count++; - perf_mmap__consume(map); - } - - perf_mmap__read_done(map); - } - - /* calls perf_evlist__munmap/perf_evlist__close */ - perf_evlist__delete(evlist); - - perf_thread_map__put(threads); - perf_cpu_map__put(cpus); - - /* - * The generated prctl calls should match the - * number of events in the buffer. - */ - __T("failed count", count == 100); - - return 0; -} - -static int test_mmap_cpus(void) -{ - struct perf_evlist *evlist; - struct perf_evsel *evsel; - struct perf_mmap *map; - struct perf_cpu_map *cpus; - struct perf_event_attr attr = { - .type = PERF_TYPE_TRACEPOINT, - .sample_period = 1, - .wakeup_watermark = 1, - .disabled = 1, - }; - cpu_set_t saved_mask; - char path[PATH_MAX]; - int id, err, cpu, tmp; - union perf_event *event; - int count = 0; - - snprintf(path, PATH_MAX, "%s/kernel/debug/tracing/events/syscalls/sys_enter_prctl/id", - sysfs__mountpoint()); - - if (filename__read_int(path, &id)) { - fprintf(stderr, "error: failed to get tracepoint id: %s\n", path); - return -1; - } - - attr.config = id; - - cpus = perf_cpu_map__new(NULL); - __T("failed to create cpus", cpus); - - evlist = perf_evlist__new(); - __T("failed to create evlist", evlist); - - evsel = perf_evsel__new(&attr); - __T("failed to create evsel1", evsel); - - perf_evlist__add(evlist, evsel); - - perf_evlist__set_maps(evlist, cpus, NULL); - - err = perf_evlist__open(evlist); - __T("failed to open evlist", err == 0); - - err = perf_evlist__mmap(evlist, 4); - __T("failed to mmap evlist", err == 0); - - perf_evlist__enable(evlist); - - err = sched_getaffinity(0, sizeof(saved_mask), &saved_mask); - __T("sched_getaffinity failed", err == 0); - - perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { - cpu_set_t mask; - - CPU_ZERO(&mask); - CPU_SET(cpu, &mask); - - err = sched_setaffinity(0, sizeof(mask), &mask); - __T("sched_setaffinity failed", err == 0); - - prctl(0, 0, 0, 0, 0); - } - - err = sched_setaffinity(0, sizeof(saved_mask), &saved_mask); - __T("sched_setaffinity failed", err == 0); - - perf_evlist__disable(evlist); - - perf_evlist__for_each_mmap(evlist, map, false) { - if (perf_mmap__read_init(map) < 0) - continue; - - while ((event = perf_mmap__read_event(map)) != NULL) { - count++; - perf_mmap__consume(map); - } - - perf_mmap__read_done(map); - } - - /* calls perf_evlist__munmap/perf_evlist__close */ - perf_evlist__delete(evlist); - - /* - * The generated prctl events should match the - * number of cpus or be bigger (we are system-wide). - */ - __T("failed count", count >= perf_cpu_map__nr(cpus)); - - perf_cpu_map__put(cpus); - - return 0; -} - -int main(int argc, char **argv) -{ - __T_START; - - libperf_init(libperf_print); - - test_stat_cpu(); - test_stat_thread(); - test_stat_thread_enable(); - test_mmap_thread(); - test_mmap_cpus(); - - __T_END; - return 0; -} diff --git a/tools/perf/lib/tests/test-evsel.c b/tools/perf/lib/tests/test-evsel.c deleted file mode 100644 index 135722a..0000000 --- a/tools/perf/lib/tests/test-evsel.c +++ /dev/null @@ -1,135 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <stdarg.h> -#include <stdio.h> -#include <linux/perf_event.h> -#include <perf/cpumap.h> -#include <perf/threadmap.h> -#include <perf/evsel.h> -#include <internal/tests.h> - -static int libperf_print(enum libperf_print_level level, - const char *fmt, va_list ap) -{ - return vfprintf(stderr, fmt, ap); -} - -static int test_stat_cpu(void) -{ - struct perf_cpu_map *cpus; - struct perf_evsel *evsel; - struct perf_event_attr attr = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_CPU_CLOCK, - }; - int err, cpu, tmp; - - cpus = perf_cpu_map__new(NULL); - __T("failed to create cpus", cpus); - - evsel = perf_evsel__new(&attr); - __T("failed to create evsel", evsel); - - err = perf_evsel__open(evsel, cpus, NULL); - __T("failed to open evsel", err == 0); - - perf_cpu_map__for_each_cpu(cpu, tmp, cpus) { - struct perf_counts_values counts = { .val = 0 }; - - perf_evsel__read(evsel, cpu, 0, &counts); - __T("failed to read value for evsel", counts.val != 0); - } - - perf_evsel__close(evsel); - perf_evsel__delete(evsel); - - perf_cpu_map__put(cpus); - return 0; -} - -static int test_stat_thread(void) -{ - struct perf_counts_values counts = { .val = 0 }; - struct perf_thread_map *threads; - struct perf_evsel *evsel; - struct perf_event_attr attr = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_TASK_CLOCK, - }; - int err; - - threads = perf_thread_map__new_dummy(); - __T("failed to create threads", threads); - - perf_thread_map__set_pid(threads, 0, 0); - - evsel = perf_evsel__new(&attr); - __T("failed to create evsel", evsel); - - err = perf_evsel__open(evsel, NULL, threads); - __T("failed to open evsel", err == 0); - - perf_evsel__read(evsel, 0, 0, &counts); - __T("failed to read value for evsel", counts.val != 0); - - perf_evsel__close(evsel); - perf_evsel__delete(evsel); - - perf_thread_map__put(threads); - return 0; -} - -static int test_stat_thread_enable(void) -{ - struct perf_counts_values counts = { .val = 0 }; - struct perf_thread_map *threads; - struct perf_evsel *evsel; - struct perf_event_attr attr = { - .type = PERF_TYPE_SOFTWARE, - .config = PERF_COUNT_SW_TASK_CLOCK, - .disabled = 1, - }; - int err; - - threads = perf_thread_map__new_dummy(); - __T("failed to create threads", threads); - - perf_thread_map__set_pid(threads, 0, 0); - - evsel = perf_evsel__new(&attr); - __T("failed to create evsel", evsel); - - err = perf_evsel__open(evsel, NULL, threads); - __T("failed to open evsel", err == 0); - - perf_evsel__read(evsel, 0, 0, &counts); - __T("failed to read value for evsel", counts.val == 0); - - err = perf_evsel__enable(evsel); - __T("failed to enable evsel", err == 0); - - perf_evsel__read(evsel, 0, 0, &counts); - __T("failed to read value for evsel", counts.val != 0); - - err = perf_evsel__disable(evsel); - __T("failed to enable evsel", err == 0); - - perf_evsel__close(evsel); - perf_evsel__delete(evsel); - - perf_thread_map__put(threads); - return 0; -} - -int main(int argc, char **argv) -{ - __T_START; - - libperf_init(libperf_print); - - test_stat_cpu(); - test_stat_thread(); - test_stat_thread_enable(); - - __T_END; - return 0; -} diff --git a/tools/perf/lib/tests/test-threadmap.c b/tools/perf/lib/tests/test-threadmap.c deleted file mode 100644 index 7dc4d6f..0000000 --- a/tools/perf/lib/tests/test-threadmap.c +++ /dev/null @@ -1,31 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <stdarg.h> -#include <stdio.h> -#include <perf/threadmap.h> -#include <internal/tests.h> - -static int libperf_print(enum libperf_print_level level, - const char *fmt, va_list ap) -{ - return vfprintf(stderr, fmt, ap); -} - -int main(int argc, char **argv) -{ - struct perf_thread_map *threads; - - __T_START; - - libperf_init(libperf_print); - - threads = perf_thread_map__new_dummy(); - if (!threads) - return -1; - - perf_thread_map__get(threads); - perf_thread_map__put(threads); - perf_thread_map__put(threads); - - __T_END; - return 0; -} diff --git a/tools/perf/lib/threadmap.c b/tools/perf/lib/threadmap.c deleted file mode 100644 index e92c368..0000000 --- a/tools/perf/lib/threadmap.c +++ /dev/null @@ -1,91 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <perf/threadmap.h> -#include <stdlib.h> -#include <linux/refcount.h> -#include <internal/threadmap.h> -#include <string.h> -#include <asm/bug.h> -#include <stdio.h> - -static void perf_thread_map__reset(struct perf_thread_map *map, int start, int nr) -{ - size_t size = (nr - start) * sizeof(map->map[0]); - - memset(&map->map[start], 0, size); - map->err_thread = -1; -} - -struct perf_thread_map *perf_thread_map__realloc(struct perf_thread_map *map, int nr) -{ - size_t size = sizeof(*map) + sizeof(map->map[0]) * nr; - int start = map ? map->nr : 0; - - map = realloc(map, size); - /* - * We only realloc to add more items, let's reset new items. - */ - if (map) - perf_thread_map__reset(map, start, nr); - - return map; -} - -#define thread_map__alloc(__nr) perf_thread_map__realloc(NULL, __nr) - -void perf_thread_map__set_pid(struct perf_thread_map *map, int thread, pid_t pid) -{ - map->map[thread].pid = pid; -} - -char *perf_thread_map__comm(struct perf_thread_map *map, int thread) -{ - return map->map[thread].comm; -} - -struct perf_thread_map *perf_thread_map__new_dummy(void) -{ - struct perf_thread_map *threads = thread_map__alloc(1); - - if (threads != NULL) { - perf_thread_map__set_pid(threads, 0, -1); - threads->nr = 1; - refcount_set(&threads->refcnt, 1); - } - return threads; -} - -static void perf_thread_map__delete(struct perf_thread_map *threads) -{ - if (threads) { - int i; - - WARN_ONCE(refcount_read(&threads->refcnt) != 0, - "thread map refcnt unbalanced\n"); - for (i = 0; i < threads->nr; i++) - free(perf_thread_map__comm(threads, i)); - free(threads); - } -} - -struct perf_thread_map *perf_thread_map__get(struct perf_thread_map *map) -{ - if (map) - refcount_inc(&map->refcnt); - return map; -} - -void perf_thread_map__put(struct perf_thread_map *map) -{ - if (map && refcount_dec_and_test(&map->refcnt)) - perf_thread_map__delete(map); -} - -int perf_thread_map__nr(struct perf_thread_map *threads) -{ - return threads ? threads->nr : 1; -} - -pid_t perf_thread_map__pid(struct perf_thread_map *map, int thread) -{ - return map->map[thread].pid; -} diff --git a/tools/perf/lib/xyarray.c b/tools/perf/lib/xyarray.c deleted file mode 100644 index dcd901d..0000000 --- a/tools/perf/lib/xyarray.c +++ /dev/null @@ -1,33 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -#include <internal/xyarray.h> -#include <linux/zalloc.h> -#include <stdlib.h> -#include <string.h> - -struct xyarray *xyarray__new(int xlen, int ylen, size_t entry_size) -{ - size_t row_size = ylen * entry_size; - struct xyarray *xy = zalloc(sizeof(*xy) + xlen * row_size); - - if (xy != NULL) { - xy->entry_size = entry_size; - xy->row_size = row_size; - xy->entries = xlen * ylen; - xy->max_x = xlen; - xy->max_y = ylen; - } - - return xy; -} - -void xyarray__reset(struct xyarray *xy) -{ - size_t n = xy->entries * xy->entry_size; - - memset(xy->contents, 0, n); -} - -void xyarray__delete(struct xyarray *xy) -{ - free(xy); -} ^ permalink raw reply related [flat|nested] 5+ messages in thread
* [PATCH 2/2] libperf: Add man pages 2019-12-06 21:06 [PATCHv2 0/2] perf/libperf move Jiri Olsa 2019-12-06 21:06 ` [PATCH 1/2] libperf: Move libperf under tools/lib/perf Jiri Olsa @ 2019-12-06 21:06 ` Jiri Olsa 2020-01-10 17:53 ` [tip: perf/core] " tip-bot2 for Jiri Olsa 1 sibling, 1 reply; 5+ messages in thread From: Jiri Olsa @ 2019-12-06 21:06 UTC (permalink / raw) To: Arnaldo Carvalho de Melo Cc: lkml, Ingo Molnar, Namhyung Kim, Alexander Shishkin, Peter Zijlstra, Michael Petlan Changing the man page generation to, because it's easier to use and been more common in related projects. Removing the current rst pages. Adding 3 man pages to have the man page base for more additions: libperf.3 - overall description libperf-counting.7 - counting basics explained on simple example libperf-sampling.7 - sampling basics explained on simple example The plan is to add more man pages to cover the basic API. The build generates html and man pages: $ cd tools/lib/perf/Documentation $ make ASCIIDOC libperf.xml XMLTO libperf.3 ASCIIDOC libperf-counting.xml XMLTO libperf-counting.7 ASCIIDOC libperf-sampling.xml XMLTO libperf-sampling.7 ASCIIDOC libperf.html ASCIIDOC libperf-counting.html ASCIIDOC libperf-sampling.html Adding following install targets: install-man - man pages install-html - html version of man pages install-examples - examples mentioned in the man pages Link: https://lkml.kernel.org/n/tip-nr27ee1oxca14n491pkud7tu@git.kernel.org Signed-off-by: Jiri Olsa <jolsa@kernel.org> --- tools/lib/perf/Documentation/Makefile | 159 ++++++++++- tools/lib/perf/Documentation/asciidoc.conf | 120 +++++++++ .../perf/Documentation/examples/sampling.c | 119 +++++++++ .../perf/Documentation/libperf-counting.txt | 211 +++++++++++++++ .../perf/Documentation/libperf-sampling.txt | 243 +++++++++++++++++ tools/lib/perf/Documentation/libperf.txt | 246 ++++++++++++++++++ tools/lib/perf/Documentation/man/libperf.rst | 100 ------- tools/lib/perf/Documentation/manpage-1.72.xsl | 14 + tools/lib/perf/Documentation/manpage-base.xsl | 35 +++ .../Documentation/manpage-bold-literal.xsl | 17 ++ .../lib/perf/Documentation/manpage-normal.xsl | 13 + .../Documentation/manpage-suppress-sp.xsl | 21 ++ .../perf/Documentation/tutorial/tutorial.rst | 123 --------- tools/lib/perf/Makefile | 5 +- 14 files changed, 1197 insertions(+), 229 deletions(-) create mode 100644 tools/lib/perf/Documentation/asciidoc.conf create mode 100644 tools/lib/perf/Documentation/examples/sampling.c create mode 100644 tools/lib/perf/Documentation/libperf-counting.txt create mode 100644 tools/lib/perf/Documentation/libperf-sampling.txt create mode 100644 tools/lib/perf/Documentation/libperf.txt delete mode 100644 tools/lib/perf/Documentation/man/libperf.rst create mode 100644 tools/lib/perf/Documentation/manpage-1.72.xsl create mode 100644 tools/lib/perf/Documentation/manpage-base.xsl create mode 100644 tools/lib/perf/Documentation/manpage-bold-literal.xsl create mode 100644 tools/lib/perf/Documentation/manpage-normal.xsl create mode 100644 tools/lib/perf/Documentation/manpage-suppress-sp.xsl delete mode 100644 tools/lib/perf/Documentation/tutorial/tutorial.rst diff --git a/tools/lib/perf/Documentation/Makefile b/tools/lib/perf/Documentation/Makefile index 586425a88795..972754082a85 100644 --- a/tools/lib/perf/Documentation/Makefile +++ b/tools/lib/perf/Documentation/Makefile @@ -1,7 +1,156 @@ -all: - rst2man man/libperf.rst > man/libperf.7 - rst2pdf tutorial/tutorial.rst +# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) +# Most of this file is copied from tools/perf/Documentation/Makefile + +include ../../../scripts/Makefile.include +include ../../../scripts/utilities.mak + +MAN3_TXT = libperf.txt +MAN7_TXT = libperf-counting.txt libperf-sampling.txt +MAN_EX = examples/*.c + +MAN_TXT = $(MAN3_TXT) $(MAN7_TXT) + +_MAN_XML = $(patsubst %.txt,%.xml,$(MAN_TXT)) +_MAN_HTML = $(patsubst %.txt,%.html,$(MAN_TXT)) +_MAN_3 = $(patsubst %.txt,%.3,$(MAN3_TXT)) +_MAN_7 = $(patsubst %.txt,%.7,$(MAN7_TXT)) + +MAN_XML = $(addprefix $(OUTPUT),$(_MAN_XML)) +MAN_HTML = $(addprefix $(OUTPUT),$(_MAN_HTML)) +MAN_3 = $(addprefix $(OUTPUT),$(_MAN_3)) +MAN_7 = $(addprefix $(OUTPUT),$(_MAN_7)) +MAN_X = $(MAN_3) $(MAN_7) + +# Make the path relative to DESTDIR, not prefix +ifndef DESTDIR + prefix ?=$(HOME) +endif + +mandir ?= $(prefix)/share/man +man3dir = $(mandir)/man3 +man7dir = $(mandir)/man7 + +docdir ?= $(prefix)/share/doc/libperf +htmldir = $(docdir)/html +exdir = $(docdir)/examples + +ASCIIDOC = asciidoc +ASCIIDOC_EXTRA = --unsafe -f asciidoc.conf +ASCIIDOC_HTML = xhtml11 +MANPAGE_XSL = manpage-normal.xsl +XMLTO_EXTRA = +XMLTO =xmlto + +INSTALL ?= install +RM ?= rm -f + +# For asciidoc ... +# -7.1.2, no extra settings are needed. +# 8.0-, set ASCIIDOC8. +# + +# For docbook-xsl ... +# -1.68.1, set ASCIIDOC_NO_ROFF? (based on changelog from 1.73.0) +# 1.69.0, no extra settings are needed? +# 1.69.1-1.71.0, set DOCBOOK_SUPPRESS_SP? +# 1.71.1, no extra settings are needed? +# 1.72.0, set DOCBOOK_XSL_172. +# 1.73.0-, set ASCIIDOC_NO_ROFF + +# If you had been using DOCBOOK_XSL_172 in an attempt to get rid +# of 'the ".ft C" problem' in your generated manpages, and you +# instead ended up with weird characters around callouts, try +# using ASCIIDOC_NO_ROFF instead (it works fine with ASCIIDOC8). + +ifdef ASCIIDOC8 + ASCIIDOC_EXTRA += -a asciidoc7compatible +endif +ifdef DOCBOOK_XSL_172 + ASCIIDOC_EXTRA += -a libperf-asciidoc-no-roff + MANPAGE_XSL = manpage-1.72.xsl +else + ifdef ASCIIDOC_NO_ROFF + # docbook-xsl after 1.72 needs the regular XSL, but will not + # pass-thru raw roff codes from asciidoc.conf, so turn them off. + ASCIIDOC_EXTRA += -a libperf-asciidoc-no-roff + endif +endif +ifdef MAN_BOLD_LITERAL + XMLTO_EXTRA += -m manpage-bold-literal.xsl +endif +ifdef DOCBOOK_SUPPRESS_SP + XMLTO_EXTRA += -m manpage-suppress-sp.xsl +endif + +DESTDIR ?= +DESTDIR_SQ = '$(subst ','\'',$(DESTDIR))' + +export DESTDIR DESTDIR_SQ + +# Please note that there is a minor bug in asciidoc. +# The version after 6.0.3 _will_ include the patch found here: +# http://marc.theaimsgroup.com/?l=libtraceevent&m=111558757202243&w=2 +# +# Until that version is released you may have to apply the patch +# yourself - yes, all 6 characters of it! + +QUIET_SUBDIR0 = +$(MAKE) -C # space to separate -C and subdir +QUIET_SUBDIR1 = + +ifneq ($(findstring $(MAKEFLAGS),w),w) + PRINT_DIR = --no-print-directory +else # "make -w" + NO_SUBDIR = : +endif + +ifneq ($(findstring $(MAKEFLAGS),s),s) + ifneq ($(V),1) + QUIET_ASCIIDOC = @echo ' ASCIIDOC '$@; + QUIET_XMLTO = @echo ' XMLTO '$@; + endif +endif + +all: $(MAN_X) $(MAN_HTML) + +$(MAN_HTML) $(MAN_X): asciidoc.conf + +install-man: all + $(call QUIET_INSTALL, man) \ + $(INSTALL) -d -m 755 $(DESTDIR)$(man3dir); \ + $(INSTALL) -m 644 $(MAN_3) $(DESTDIR)$(man3dir); \ + $(INSTALL) -d -m 755 $(DESTDIR)$(man7dir); \ + $(INSTALL) -m 644 $(MAN_7) $(DESTDIR)$(man7dir); + +install-html: + $(call QUIET_INSTALL, html) \ + $(INSTALL) -d -m 755 $(DESTDIR)$(htmldir); \ + $(INSTALL) -m 644 $(MAN_HTML) $(DESTDIR)$(htmldir); \ + +install-examples: + $(call QUIET_INSTALL, examples) \ + $(INSTALL) -d -m 755 $(DESTDIR)$(exdir); \ + $(INSTALL) -m 644 $(MAN_EX) $(DESTDIR)$(exdir); \ + +CLEAN_FILES = \ + $(MAN_XML) $(addsuffix +,$(MAN_XML)) \ + $(MAN_HTML) $(addsuffix +,$(MAN_HTML)) \ + $(MAN_X) clean: - rm -f man/libperf.7 - rm -f tutorial/tutorial.pdf + $(call QUIET_CLEAN, Documentation) $(RM) $(CLEAN_FILES) + +$(MAN_3): $(OUTPUT)%.3: %.xml + $(QUIET_XMLTO)$(XMLTO) -o $(OUTPUT). -m $(MANPAGE_XSL) $(XMLTO_EXTRA) man $< + +$(MAN_7): $(OUTPUT)%.7: %.xml + $(QUIET_XMLTO)$(XMLTO) -o $(OUTPUT). -m $(MANPAGE_XSL) $(XMLTO_EXTRA) man $< + +$(MAN_XML): $(OUTPUT)%.xml: %.txt + $(QUIET_ASCIIDOC)$(ASCIIDOC) -b docbook -d manpage \ + $(ASCIIDOC_EXTRA) -alibperf_version=$(EVENT_PARSE_VERSION) -o $@+ $< && \ + mv $@+ $@ + +$(MAN_HTML): $(OUTPUT)%.html: %.txt + $(QUIET_ASCIIDOC)$(ASCIIDOC) -b $(ASCIIDOC_HTML) -d manpage \ + $(ASCIIDOC_EXTRA) -aperf_version=$(EVENT_PARSE_VERSION) -o $@+ $< && \ + mv $@+ $@ diff --git a/tools/lib/perf/Documentation/asciidoc.conf b/tools/lib/perf/Documentation/asciidoc.conf new file mode 100644 index 000000000000..9d5a5a5ee091 --- /dev/null +++ b/tools/lib/perf/Documentation/asciidoc.conf @@ -0,0 +1,120 @@ +## linktep: macro +# +# Usage: linktep:command[manpage-section] +# +# Note, {0} is the manpage section, while {target} is the command. +# +# Show TEP link as: <command>(<section>); if section is defined, else just show +# the command. + +[macros] +(?su)[\\]?(?P<name>linktep):(?P<target>\S*?)\[(?P<attrlist>.*?)\]= + +[attributes] +asterisk=* +plus=+ +caret=^ +startsb=[ +endsb=] +tilde=~ + +ifdef::backend-docbook[] +[linktep-inlinemacro] +{0%{target}} +{0#<citerefentry>} +{0#<refentrytitle>{target}</refentrytitle><manvolnum>{0}</manvolnum>} +{0#</citerefentry>} +endif::backend-docbook[] + +ifdef::backend-docbook[] +ifndef::tep-asciidoc-no-roff[] +# "unbreak" docbook-xsl v1.68 for manpages. v1.69 works with or without this. +# v1.72 breaks with this because it replaces dots not in roff requests. +[listingblock] +<example><title>{title}</title> +<literallayout> +ifdef::doctype-manpage[] + .ft C +endif::doctype-manpage[] +| +ifdef::doctype-manpage[] + .ft +endif::doctype-manpage[] +</literallayout> +{title#}</example> +endif::tep-asciidoc-no-roff[] + +ifdef::tep-asciidoc-no-roff[] +ifdef::doctype-manpage[] +# The following two small workarounds insert a simple paragraph after screen +[listingblock] +<example><title>{title}</title> +<literallayout> +| +</literallayout><simpara></simpara> +{title#}</example> + +[verseblock] +<formalpara{id? id="{id}"}><title>{title}</title><para> +{title%}<literallayout{id? id="{id}"}> +{title#}<literallayout> +| +</literallayout> +{title#}</para></formalpara> +{title%}<simpara></simpara> +endif::doctype-manpage[] +endif::tep-asciidoc-no-roff[] +endif::backend-docbook[] + +ifdef::doctype-manpage[] +ifdef::backend-docbook[] +[header] +template::[header-declarations] +<refentry> +<refmeta> +<refentrytitle>{mantitle}</refentrytitle> +<manvolnum>{manvolnum}</manvolnum> +<refmiscinfo class="source">libperf</refmiscinfo> +<refmiscinfo class="version">{libperf_version}</refmiscinfo> +<refmiscinfo class="manual">libperf Manual</refmiscinfo> +</refmeta> +<refnamediv> + <refname>{manname1}</refname> + <refname>{manname2}</refname> + <refname>{manname3}</refname> + <refname>{manname4}</refname> + <refname>{manname5}</refname> + <refname>{manname6}</refname> + <refname>{manname7}</refname> + <refname>{manname8}</refname> + <refname>{manname9}</refname> + <refname>{manname10}</refname> + <refname>{manname11}</refname> + <refname>{manname12}</refname> + <refname>{manname13}</refname> + <refname>{manname14}</refname> + <refname>{manname15}</refname> + <refname>{manname16}</refname> + <refname>{manname17}</refname> + <refname>{manname18}</refname> + <refname>{manname19}</refname> + <refname>{manname20}</refname> + <refname>{manname21}</refname> + <refname>{manname22}</refname> + <refname>{manname23}</refname> + <refname>{manname24}</refname> + <refname>{manname25}</refname> + <refname>{manname26}</refname> + <refname>{manname27}</refname> + <refname>{manname28}</refname> + <refname>{manname29}</refname> + <refname>{manname30}</refname> + <refpurpose>{manpurpose}</refpurpose> +</refnamediv> +endif::backend-docbook[] +endif::doctype-manpage[] + +ifdef::backend-xhtml11[] +[linktep-inlinemacro] +<a href="{target}.html">{target}{0?({0})}</a> +endif::backend-xhtml11[] diff --git a/tools/lib/perf/Documentation/examples/sampling.c b/tools/lib/perf/Documentation/examples/sampling.c new file mode 100644 index 000000000000..8e1a926a9cfe --- /dev/null +++ b/tools/lib/perf/Documentation/examples/sampling.c @@ -0,0 +1,119 @@ +#include <linux/perf_event.h> +#include <perf/evlist.h> +#include <perf/evsel.h> +#include <perf/cpumap.h> +#include <perf/threadmap.h> +#include <perf/mmap.h> +#include <perf/core.h> +#include <perf/event.h> +#include <stdio.h> +#include <unistd.h> + +static int libperf_print(enum libperf_print_level level, + const char *fmt, va_list ap) +{ + return vfprintf(stderr, fmt, ap); +} + +union u64_swap { + __u64 val64; + __u32 val32[2]; +}; + +int main(int argc, char **argv) +{ + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_mmap *map; + struct perf_cpu_map *cpus; + struct perf_event_attr attr = { + .type = PERF_TYPE_HARDWARE, + .config = PERF_COUNT_HW_CPU_CYCLES, + .disabled = 1, + .freq = 1, + .sample_freq = 10, + .sample_type = PERF_SAMPLE_IP|PERF_SAMPLE_TID|PERF_SAMPLE_CPU|PERF_SAMPLE_PERIOD, + }; + int err = -1; + union perf_event *event; + + libperf_init(libperf_print); + + cpus = perf_cpu_map__new(NULL); + if (!cpus) { + fprintf(stderr, "failed to create cpus\n"); + return -1; + } + + evlist = perf_evlist__new(); + if (!evlist) { + fprintf(stderr, "failed to create evlist\n"); + goto out_cpus; + } + + evsel = perf_evsel__new(&attr); + if (!evsel) { + fprintf(stderr, "failed to create cycles\n"); + goto out_cpus; + } + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, cpus, NULL); + + err = perf_evlist__open(evlist); + if (err) { + fprintf(stderr, "failed to open evlist\n"); + goto out_evlist; + } + + err = perf_evlist__mmap(evlist, 4); + if (err) { + fprintf(stderr, "failed to mmap evlist\n"); + goto out_evlist; + } + + perf_evlist__enable(evlist); + sleep(3); + perf_evlist__disable(evlist); + + perf_evlist__for_each_mmap(evlist, map, false) { + if (perf_mmap__read_init(map) < 0) + continue; + + while ((event = perf_mmap__read_event(map)) != NULL) { + int cpu, pid, tid; + __u64 ip, period, *array; + union u64_swap u; + + array = event->sample.array; + + ip = *array; + array++; + + u.val64 = *array; + pid = u.val32[0]; + tid = u.val32[1]; + array++; + + u.val64 = *array; + cpu = u.val32[0]; + array++; + + period = *array; + + fprintf(stdout, "cpu %3d, pid %6d, tid %6d, ip %20llx, period %20llu\n", + cpu, pid, tid, ip, period); + + perf_mmap__consume(map); + } + + perf_mmap__read_done(map); + } + +out_evlist: + perf_evlist__delete(evlist); +out_cpus: + perf_cpu_map__put(cpus); + return err; +} diff --git a/tools/lib/perf/Documentation/libperf-counting.txt b/tools/lib/perf/Documentation/libperf-counting.txt new file mode 100644 index 000000000000..cae9757f49c1 --- /dev/null +++ b/tools/lib/perf/Documentation/libperf-counting.txt @@ -0,0 +1,211 @@ +libperf-counting(7) +=================== + +NAME +---- +libperf-counting - counting interface + +DESCRIPTION +----------- +The counting interface provides API to meassure and get count for specific perf events. + +The following test tries to explain count on `counting.c` example. + +It is by no means complete guide to counting, but shows libperf basic API for counting. + +The `counting.c` comes with libbperf package and can be compiled and run like: + +[source,bash] +-- +$ gcc -o counting counting.c -lperf +$ sudo ./counting +count 176792, enabled 176944, run 176944 +count 176242, enabled 176242, run 176242 +-- + +It requires root access, because of the `PERF_COUNT_SW_CPU_CLOCK` event, +which is available only for root. + +The `counting.c` example monitors two events on the current process and displays their count, in a nutshel it: + +* creates events +* adds them to the event list +* opens and enables events through the event list +* does some workload +* disables events +* reads and displays event counts +* destroys the event list + +The first thing you need to do before using libperf is to call init function: + +[source,c] +-- + 8 static int libperf_print(enum libperf_print_level level, + 9 const char *fmt, va_list ap) + 10 { + 11 return vfprintf(stderr, fmt, ap); + 12 } + + 14 int main(int argc, char **argv) + 15 { + ... + 35 libperf_init(libperf_print); +-- + +It will setup the library and sets function for debug output from library. + +The `libperf_print` callback will receive any message with its debug level, +defined as: + +[source,c] +-- +enum libperf_print_level { + LIBPERF_ERR, + LIBPERF_WARN, + LIBPERF_INFO, + LIBPERF_DEBUG, + LIBPERF_DEBUG2, + LIBPERF_DEBUG3, +}; +-- + +Once the setup is complete we start by defining specific events using the `struct perf_event_attr`. + +We create software events for cpu and task: + +[source,c] +-- + 20 struct perf_event_attr attr1 = { + 21 .type = PERF_TYPE_SOFTWARE, + 22 .config = PERF_COUNT_SW_CPU_CLOCK, + 23 .read_format = PERF_FORMAT_TOTAL_TIME_ENABLED|PERF_FORMAT_TOTAL_TIME_RUNNING, + 24 .disabled = 1, + 25 }; + 26 struct perf_event_attr attr2 = { + 27 .type = PERF_TYPE_SOFTWARE, + 28 .config = PERF_COUNT_SW_TASK_CLOCK, + 29 .read_format = PERF_FORMAT_TOTAL_TIME_ENABLED|PERF_FORMAT_TOTAL_TIME_RUNNING, + 30 .disabled = 1, + 31 }; +-- + +The `read_format` setup tells perf to include timing details together with each count. + +Next step is to prepare threads map. + +In this case we will monitor current process, so we create threads map with single pid (0): + +[source,c] +-- + 37 threads = perf_thread_map__new_dummy(); + 38 if (!threads) { + 39 fprintf(stderr, "failed to create threads\n"); + 40 return -1; + 41 } + 42 + 43 perf_thread_map__set_pid(threads, 0, 0); +-- + +Now we create libperf's event list, which will serve as holder for the events we want: + +[source,c] +-- + 45 evlist = perf_evlist__new(); + 46 if (!evlist) { + 47 fprintf(stderr, "failed to create evlist\n"); + 48 goto out_threads; + 49 } +-- + +We create libperf's events for the attributes we defined earlier and add them to the list: + +[source,c] +-- + 51 evsel = perf_evsel__new(&attr1); + 52 if (!evsel) { + 53 fprintf(stderr, "failed to create evsel1\n"); + 54 goto out_evlist; + 55 } + 56 + 57 perf_evlist__add(evlist, evsel); + 58 + 59 evsel = perf_evsel__new(&attr2); + 60 if (!evsel) { + 61 fprintf(stderr, "failed to create evsel2\n"); + 62 goto out_evlist; + 63 } + 64 + 65 perf_evlist__add(evlist, evsel); +-- + +Configure event list with the thread map and open events: + +[source,c] +-- + 67 perf_evlist__set_maps(evlist, NULL, threads); + 68 + 69 err = perf_evlist__open(evlist); + 70 if (err) { + 71 fprintf(stderr, "failed to open evsel\n"); + 72 goto out_evlist; + 73 } +-- + +Both events are created as disabled (note the `disabled = 1` assignment above), +so we need to enable the whole list explicitely (both events). + +From this moment events are counting and we can do our workload. + +When we are done we disable the events list. + +[source,c] +-- + 75 perf_evlist__enable(evlist); + 76 + 77 while (count--); + 78 + 79 perf_evlist__disable(evlist); +-- + +Now we need to get the counts from events, following code iterates throught the events list and read counts: + +[source,c] +-- + 81 perf_evlist__for_each_evsel(evlist, evsel) { + 82 perf_evsel__read(evsel, 0, 0, &counts); + 83 fprintf(stdout, "count %llu, enabled %llu, run %llu\n", + 84 counts.val, counts.ena, counts.run); + 85 } +-- + +And finaly cleanup. + +We close the whole events list (both events) and remove it together with the threads map: + +[source,c] +-- + 87 perf_evlist__close(evlist); + 88 + 89 out_evlist: + 90 perf_evlist__delete(evlist); + 91 out_threads: + 92 perf_thread_map__put(threads); + 93 return err; + 94 } +-- + +REPORTING BUGS +-------------- +Report bugs to <linux-perf-users@vger.kernel.org>. + +LICENSE +------- +libperf is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git + +SEE ALSO +-------- +libperf(3), libperf-sampling(7) diff --git a/tools/lib/perf/Documentation/libperf-sampling.txt b/tools/lib/perf/Documentation/libperf-sampling.txt new file mode 100644 index 000000000000..d71a7b4fcf5f --- /dev/null +++ b/tools/lib/perf/Documentation/libperf-sampling.txt @@ -0,0 +1,243 @@ +libperf-sampling(7) +=================== + +NAME +---- +libperf-sampling - sampling interface + + +DESCRIPTION +----------- +The sampling interface provides API to meassure and get count for specific perf events. + +The following test tries to explain count on `sampling.c` example. + +It is by no means complete guide to sampling, but shows libperf basic API for sampling. + +The `sampling.c` comes with libbperf package and can be compiled and run like: + +[source,bash] +-- +$ gcc -o sampling sampling.c -lperf +$ sudo ./sampling +cpu 0, pid 0, tid 0, ip ffffffffad06c4e6, period 1 +cpu 0, pid 4465, tid 4469, ip ffffffffad118748, period 18322959 +cpu 0, pid 0, tid 0, ip ffffffffad115722, period 33544846 +cpu 0, pid 4465, tid 4470, ip 7f84fe0cdad6, period 23687474 +cpu 0, pid 0, tid 0, ip ffffffffad9e0349, period 34255790 +cpu 0, pid 4465, tid 4469, ip ffffffffad136581, period 38664069 +cpu 0, pid 0, tid 0, ip ffffffffad9e55e2, period 21922384 +cpu 0, pid 4465, tid 4470, ip 7f84fe0ebebf, period 17655175 +... +-- + +It requires root access, because it uses hardware cycles event. + +The `sampling.c` example profiles/samples all CPUs with hardware cycles, in a nutshel it: + +- creates events +- adds them to the event list +- opens and enables events through the event list +- sleeps for 3 seconds +- disables events +- reads and displays recorded samples +- destroys the event list + +The first thing you need to do before using libperf is to call init function: + +[source,c] +-- + 12 static int libperf_print(enum libperf_print_level level, + 13 const char *fmt, va_list ap) + 14 { + 15 return vfprintf(stderr, fmt, ap); + 16 } + + 23 int main(int argc, char **argv) + 24 { + ... + 40 libperf_init(libperf_print); +-- + +It will setup the library and sets function for debug output from library. + +The `libperf_print` callback will receive any message with its debug level, +defined as: + +[source,c] +-- +enum libperf_print_level { + LIBPERF_ERR, + LIBPERF_WARN, + LIBPERF_INFO, + LIBPERF_DEBUG, + LIBPERF_DEBUG2, + LIBPERF_DEBUG3, +}; +-- + +Once the setup is complete we start by defining cycles event using the `struct perf_event_attr`: + +[source,c] +-- + 29 struct perf_event_attr attr = { + 30 .type = PERF_TYPE_HARDWARE, + 31 .config = PERF_COUNT_HW_CPU_CYCLES, + 32 .disabled = 1, + 33 .freq = 1, + 34 .sample_freq = 10, + 35 .sample_type = PERF_SAMPLE_IP|PERF_SAMPLE_TID|PERF_SAMPLE_CPU|PERF_SAMPLE_PERIOD, + 36 }; +-- + +Next step is to prepare cpus map. + +In this case we will monitor all the available CPUs: + +[source,c] +-- + 42 cpus = perf_cpu_map__new(NULL); + 43 if (!cpus) { + 44 fprintf(stderr, "failed to create cpus\n"); + 45 return -1; + 46 } +-- + +Now we create libperf's event list, which will serve as holder for the cycles event: + +[source,c] +-- + 48 evlist = perf_evlist__new(); + 49 if (!evlist) { + 50 fprintf(stderr, "failed to create evlist\n"); + 51 goto out_cpus; + 52 } +-- + +We create libperf's event for the cycles attribute we defined earlier and add it to the list: + +[source,c] +-- + 54 evsel = perf_evsel__new(&attr); + 55 if (!evsel) { + 56 fprintf(stderr, "failed to create cycles\n"); + 57 goto out_cpus; + 58 } + 59 + 60 perf_evlist__add(evlist, evsel); +-- + +Configure event list with the cpus map and open event: + +[source,c] +-- + 62 perf_evlist__set_maps(evlist, cpus, NULL); + 63 + 64 err = perf_evlist__open(evlist); + 65 if (err) { + 66 fprintf(stderr, "failed to open evlist\n"); + 67 goto out_evlist; + 68 } +-- + +Once the events list is open, we can create memory maps AKA perf ring buffers: + +[source,c] +-- + 70 err = perf_evlist__mmap(evlist, 4); + 71 if (err) { + 72 fprintf(stderr, "failed to mmap evlist\n"); + 73 goto out_evlist; + 74 } +-- + +The event is created as disabled (note the `disabled = 1` assignment above), +so we need to enable the events list explicitely. + +From this moment the cycles event is sampling. + +We will sleep for 3 seconds while the ring buffers get data from all CPUs, then we disable the events list. + +[source,c] +-- + 76 perf_evlist__enable(evlist); + 77 sleep(3); + 78 perf_evlist__disable(evlist); +-- + +Following code walks through the ring buffers and reads stored events/samples: + +[source,c] +-- + 80 perf_evlist__for_each_mmap(evlist, map, false) { + 81 if (perf_mmap__read_init(map) < 0) + 82 continue; + 83 + 84 while ((event = perf_mmap__read_event(map)) != NULL) { + + /* process event */ + +108 perf_mmap__consume(map); +109 } +110 perf_mmap__read_done(map); +111 } + +-- + +Each sample needs to get parsed: + +[source,c] +-- + 85 int cpu, pid, tid; + 86 __u64 ip, period, *array; + 87 union u64_swap u; + 88 + 89 array = event->sample.array; + 90 + 91 ip = *array; + 92 array++; + 93 + 94 u.val64 = *array; + 95 pid = u.val32[0]; + 96 tid = u.val32[1]; + 97 array++; + 98 + 99 u.val64 = *array; +100 cpu = u.val32[0]; +101 array++; +102 +103 period = *array; +104 +105 fprintf(stdout, "cpu %3d, pid %6d, tid %6d, ip %20llx, period %20llu\n", +106 cpu, pid, tid, ip, period); +-- + +And finaly cleanup. + +We close the whole events list (both events) and remove it together with the threads map: + +[source,c] +-- +113 out_evlist: +114 perf_evlist__delete(evlist); +115 out_cpus: +116 perf_cpu_map__put(cpus); +117 return err; +118 } +-- + +REPORTING BUGS +-------------- +Report bugs to <linux-perf-users@vger.kernel.org>. + +LICENSE +------- +libperf is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git + +SEE ALSO +-------- +libperf(3), libperf-counting(7) diff --git a/tools/lib/perf/Documentation/libperf.txt b/tools/lib/perf/Documentation/libperf.txt new file mode 100644 index 000000000000..5a6bb512789d --- /dev/null +++ b/tools/lib/perf/Documentation/libperf.txt @@ -0,0 +1,246 @@ +libperf(3) +========== + +NAME +---- +libperf - Linux kernel perf event library + + +SYNOPSIS +-------- +*Generic API:* + +[source,c] +-- + #include <perf/core.h> + + enum libperf_print_level { + LIBPERF_ERR, + LIBPERF_WARN, + LIBPERF_INFO, + LIBPERF_DEBUG, + LIBPERF_DEBUG2, + LIBPERF_DEBUG3, + }; + + typedef int (*libperf_print_fn_t)(enum libperf_print_level level, + const char *, va_list ap); + + void libperf_init(libperf_print_fn_t fn); +-- + +*API to handle cpu maps:* + +[source,c] +-- + #include <perf/cpumap.h> + + struct perf_cpu_map; + + struct perf_cpu_map *perf_cpu_map__dummy_new(void); + struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list); + struct perf_cpu_map *perf_cpu_map__read(FILE *file); + struct perf_cpu_map *perf_cpu_map__get(struct perf_cpu_map *map); + struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig, + struct perf_cpu_map *other); + void perf_cpu_map__put(struct perf_cpu_map *map); + int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx); + int perf_cpu_map__nr(const struct perf_cpu_map *cpus); + bool perf_cpu_map__empty(const struct perf_cpu_map *map); + int perf_cpu_map__max(struct perf_cpu_map *map); + + #define perf_cpu_map__for_each_cpu(cpu, idx, cpus) +-- + +*API to handle thread maps:* + +[source,c] +-- + #include <perf/threadmap.h> + + struct perf_thread_map; + + struct perf_thread_map *perf_thread_map__new_dummy(void); + + void perf_thread_map__set_pid(struct perf_thread_map *map, int thread, pid_t pid); + char *perf_thread_map__comm(struct perf_thread_map *map, int thread); + int perf_thread_map__nr(struct perf_thread_map *threads); + pid_t perf_thread_map__pid(struct perf_thread_map *map, int thread); + + struct perf_thread_map *perf_thread_map__get(struct perf_thread_map *map); + void perf_thread_map__put(struct perf_thread_map *map); +-- + +*API to handle event lists:* + +[source,c] +-- + #include <perf/evlist.h> + + struct perf_evlist; + + void perf_evlist__add(struct perf_evlist *evlist, + struct perf_evsel *evsel); + void perf_evlist__remove(struct perf_evlist *evlist, + struct perf_evsel *evsel); + struct perf_evlist *perf_evlist__new(void); + void perf_evlist__delete(struct perf_evlist *evlist); + struct perf_evsel* perf_evlist__next(struct perf_evlist *evlist, + struct perf_evsel *evsel); + int perf_evlist__open(struct perf_evlist *evlist); + void perf_evlist__close(struct perf_evlist *evlist); + void perf_evlist__enable(struct perf_evlist *evlist); + void perf_evlist__disable(struct perf_evlist *evlist); + + #define perf_evlist__for_each_evsel(evlist, pos) + + void perf_evlist__set_maps(struct perf_evlist *evlist, + struct perf_cpu_map *cpus, + struct perf_thread_map *threads); + int perf_evlist__poll(struct perf_evlist *evlist, int timeout); + int perf_evlist__filter_pollfd(struct perf_evlist *evlist, + short revents_and_mask); + + int perf_evlist__mmap(struct perf_evlist *evlist, int pages); + void perf_evlist__munmap(struct perf_evlist *evlist); + + struct perf_mmap *perf_evlist__next_mmap(struct perf_evlist *evlist, + struct perf_mmap *map, + bool overwrite); + + #define perf_evlist__for_each_mmap(evlist, pos, overwrite) +-- + +*API to handle events:* + +[source,c] +-- + #include <perf/evsel.h>* + + struct perf_evsel; + + struct perf_counts_values { + union { + struct { + uint64_t val; + uint64_t ena; + uint64_t run; + }; + uint64_t values[3]; + }; + }; + + struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr); + void perf_evsel__delete(struct perf_evsel *evsel); + int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus, + struct perf_thread_map *threads); + void perf_evsel__close(struct perf_evsel *evsel); + void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu); + int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, + struct perf_counts_values *count); + int perf_evsel__enable(struct perf_evsel *evsel); + int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu); + int perf_evsel__disable(struct perf_evsel *evsel); + int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu); + struct perf_cpu_map *perf_evsel__cpus(struct perf_evsel *evsel); + struct perf_thread_map *perf_evsel__threads(struct perf_evsel *evsel); + struct perf_event_attr *perf_evsel__attr(struct perf_evsel *evsel); +-- + +*API to handle maps (perf ring buffers):* + +[source,c] +-- + #include <perf/mmap.h> + + struct perf_mmap; + + void perf_mmap__consume(struct perf_mmap *map); + int perf_mmap__read_init(struct perf_mmap *map); + void perf_mmap__read_done(struct perf_mmap *map); + union perf_event *perf_mmap__read_event(struct perf_mmap *map); +-- + +*Structures to access perf API events:* + +[source,c] +-- + #include <perf/event.h> + + struct perf_record_mmap; + struct perf_record_mmap2; + struct perf_record_comm; + struct perf_record_namespaces; + struct perf_record_fork; + struct perf_record_lost; + struct perf_record_lost_samples; + struct perf_record_read; + struct perf_record_throttle; + struct perf_record_ksymbol; + struct perf_record_bpf_event; + struct perf_record_sample; + struct perf_record_switch; + struct perf_record_header_attr; + struct perf_record_record_cpu_map; + struct perf_record_cpu_map_data; + struct perf_record_cpu_map; + struct perf_record_event_update_cpus; + struct perf_record_event_update_scale; + struct perf_record_event_update; + struct perf_trace_event_type; + struct perf_record_header_event_type; + struct perf_record_header_tracing_data; + struct perf_record_header_build_id; + struct perf_record_id_index; + struct perf_record_auxtrace_info; + struct perf_record_auxtrace; + struct perf_record_auxtrace_error; + struct perf_record_aux; + struct perf_record_itrace_start; + struct perf_record_thread_map_entry; + struct perf_record_thread_map; + struct perf_record_stat_config_entry; + struct perf_record_stat_config; + struct perf_record_stat; + struct perf_record_stat_round; + struct perf_record_time_conv; + struct perf_record_header_feature; + struct perf_record_compressed; +-- + +DESCRIPTION +----------- +The libperf library provides an API to access the linux kernel perf +events subsystem. + +Following objects are key to the libperf interface: + +[horizontal] + +struct perf_cpu_map:: Provides a cpu list abstraction. + +struct perf_thread_map:: Provides a thread list abstraction. + +struct perf_evsel:: Provides an abstraction for single a perf event. + +struct perf_evlist:: Gathers several struct perf_evsel object and performs functions on all of them. + +struct perf_mmap:: Provides an abstraction for accessing perf ring buffer. + +The exported API functions bind these objects together. + +REPORTING BUGS +-------------- +Report bugs to <linux-perf-users@vger.kernel.org>. + +LICENSE +------- +libperf is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git + +SEE ALSO +-------- +libperf-sampling(7), libperf-counting(7) diff --git a/tools/lib/perf/Documentation/man/libperf.rst b/tools/lib/perf/Documentation/man/libperf.rst deleted file mode 100644 index 09a270fccb9c..000000000000 --- a/tools/lib/perf/Documentation/man/libperf.rst +++ /dev/null @@ -1,100 +0,0 @@ -.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -libperf - -The libperf library provides an API to access the linux kernel perf -events subsystem. It provides the following high level objects: - - - struct perf_cpu_map - - struct perf_thread_map - - struct perf_evlist - - struct perf_evsel - -reference -========= -Function reference by header files: - -perf/core.h ------------ -.. code-block:: c - - typedef int (\*libperf_print_fn_t)(enum libperf_print_level level, - const char \*, va_list ap); - - void libperf_set_print(libperf_print_fn_t fn); - -perf/cpumap.h -------------- -.. code-block:: c - - struct perf_cpu_map \*perf_cpu_map__dummy_new(void); - struct perf_cpu_map \*perf_cpu_map__new(const char \*cpu_list); - struct perf_cpu_map \*perf_cpu_map__read(FILE \*file); - struct perf_cpu_map \*perf_cpu_map__get(struct perf_cpu_map \*map); - void perf_cpu_map__put(struct perf_cpu_map \*map); - int perf_cpu_map__cpu(const struct perf_cpu_map \*cpus, int idx); - int perf_cpu_map__nr(const struct perf_cpu_map \*cpus); - perf_cpu_map__for_each_cpu(cpu, idx, cpus) - -perf/threadmap.h ----------------- -.. code-block:: c - - struct perf_thread_map \*perf_thread_map__new_dummy(void); - void perf_thread_map__set_pid(struct perf_thread_map \*map, int thread, pid_t pid); - char \*perf_thread_map__comm(struct perf_thread_map \*map, int thread); - struct perf_thread_map \*perf_thread_map__get(struct perf_thread_map \*map); - void perf_thread_map__put(struct perf_thread_map \*map); - -perf/evlist.h -------------- -.. code-block:: - - void perf_evlist__init(struct perf_evlist \*evlist); - void perf_evlist__add(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - void perf_evlist__remove(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - struct perf_evlist \*perf_evlist__new(void); - void perf_evlist__delete(struct perf_evlist \*evlist); - struct perf_evsel\* perf_evlist__next(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - int perf_evlist__open(struct perf_evlist \*evlist); - void perf_evlist__close(struct perf_evlist \*evlist); - void perf_evlist__enable(struct perf_evlist \*evlist); - void perf_evlist__disable(struct perf_evlist \*evlist); - perf_evlist__for_each_evsel(evlist, pos) - void perf_evlist__set_maps(struct perf_evlist \*evlist, - struct perf_cpu_map \*cpus, - struct perf_thread_map \*threads); - -perf/evsel.h ------------- -.. code-block:: c - - struct perf_counts_values { - union { - struct { - uint64_t val; - uint64_t ena; - uint64_t run; - }; - uint64_t values[3]; - }; - }; - - void perf_evsel__init(struct perf_evsel \*evsel, - struct perf_event_attr \*attr); - struct perf_evsel \*perf_evsel__new(struct perf_event_attr \*attr); - void perf_evsel__delete(struct perf_evsel \*evsel); - int perf_evsel__open(struct perf_evsel \*evsel, struct perf_cpu_map \*cpus, - struct perf_thread_map \*threads); - void perf_evsel__close(struct perf_evsel \*evsel); - int perf_evsel__read(struct perf_evsel \*evsel, int cpu, int thread, - struct perf_counts_values \*count); - int perf_evsel__enable(struct perf_evsel \*evsel); - int perf_evsel__disable(struct perf_evsel \*evsel); - int perf_evsel__apply_filter(struct perf_evsel \*evsel, const char \*filter); - struct perf_cpu_map \*perf_evsel__cpus(struct perf_evsel \*evsel); - struct perf_thread_map \*perf_evsel__threads(struct perf_evsel \*evsel); - struct perf_event_attr \*perf_evsel__attr(struct perf_evsel \*evsel); diff --git a/tools/lib/perf/Documentation/manpage-1.72.xsl b/tools/lib/perf/Documentation/manpage-1.72.xsl new file mode 100644 index 000000000000..b4d315cb8c47 --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-1.72.xsl @@ -0,0 +1,14 @@ +<!-- manpage-1.72.xsl: + special settings for manpages rendered from asciidoc+docbook + handles peculiarities in docbook-xsl 1.72.0 --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<xsl:import href="manpage-base.xsl"/> + +<!-- these are the special values for the roff control characters + needed for docbook-xsl 1.72.0 --> +<xsl:param name="git.docbook.backslash">▓</xsl:param> +<xsl:param name="git.docbook.dot" >⌂</xsl:param> + +</xsl:stylesheet> diff --git a/tools/lib/perf/Documentation/manpage-base.xsl b/tools/lib/perf/Documentation/manpage-base.xsl new file mode 100644 index 000000000000..a264fa616093 --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-base.xsl @@ -0,0 +1,35 @@ +<!-- manpage-base.xsl: + special formatting for manpages rendered from asciidoc+docbook --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<!-- these params silence some output from xmlto --> +<xsl:param name="man.output.quietly" select="1"/> +<xsl:param name="refentry.meta.get.quietly" select="1"/> + +<!-- convert asciidoc callouts to man page format; + git.docbook.backslash and git.docbook.dot params + must be supplied by another XSL file or other means --> +<xsl:template match="co"> + <xsl:value-of select="concat( + $git.docbook.backslash,'fB(', + substring-after(@id,'-'),')', + $git.docbook.backslash,'fR')"/> +</xsl:template> +<xsl:template match="calloutlist"> + <xsl:value-of select="$git.docbook.dot"/> + <xsl:text>sp </xsl:text> + <xsl:apply-templates/> + <xsl:text> </xsl:text> +</xsl:template> +<xsl:template match="callout"> + <xsl:value-of select="concat( + $git.docbook.backslash,'fB', + substring-after(@arearefs,'-'), + '. ',$git.docbook.backslash,'fR')"/> + <xsl:apply-templates/> + <xsl:value-of select="$git.docbook.dot"/> + <xsl:text>br </xsl:text> +</xsl:template> + +</xsl:stylesheet> diff --git a/tools/lib/perf/Documentation/manpage-bold-literal.xsl b/tools/lib/perf/Documentation/manpage-bold-literal.xsl new file mode 100644 index 000000000000..608eb5df6281 --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-bold-literal.xsl @@ -0,0 +1,17 @@ +<!-- manpage-bold-literal.xsl: + special formatting for manpages rendered from asciidoc+docbook --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<!-- render literal text as bold (instead of plain or monospace); + this makes literal text easier to distinguish in manpages + viewed on a tty --> +<xsl:template match="literal"> + <xsl:value-of select="$git.docbook.backslash"/> + <xsl:text>fB</xsl:text> + <xsl:apply-templates/> + <xsl:value-of select="$git.docbook.backslash"/> + <xsl:text>fR</xsl:text> +</xsl:template> + +</xsl:stylesheet> diff --git a/tools/lib/perf/Documentation/manpage-normal.xsl b/tools/lib/perf/Documentation/manpage-normal.xsl new file mode 100644 index 000000000000..a48f5b11f3dc --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-normal.xsl @@ -0,0 +1,13 @@ +<!-- manpage-normal.xsl: + special settings for manpages rendered from asciidoc+docbook + handles anything we want to keep away from docbook-xsl 1.72.0 --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<xsl:import href="manpage-base.xsl"/> + +<!-- these are the normal values for the roff control characters --> +<xsl:param name="git.docbook.backslash">\</xsl:param> +<xsl:param name="git.docbook.dot" >.</xsl:param> + +</xsl:stylesheet> diff --git a/tools/lib/perf/Documentation/manpage-suppress-sp.xsl b/tools/lib/perf/Documentation/manpage-suppress-sp.xsl new file mode 100644 index 000000000000..a63c7632a87d --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-suppress-sp.xsl @@ -0,0 +1,21 @@ +<!-- manpage-suppress-sp.xsl: + special settings for manpages rendered from asciidoc+docbook + handles erroneous, inline .sp in manpage output of some + versions of docbook-xsl --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<!-- attempt to work around spurious .sp at the tail of the line + that some versions of docbook stylesheets seem to add --> +<xsl:template match="simpara"> + <xsl:variable name="content"> + <xsl:apply-templates/> + </xsl:variable> + <xsl:value-of select="normalize-space($content)"/> + <xsl:if test="not(ancestor::authorblurb) and + not(ancestor::personblurb)"> + <xsl:text> </xsl:text> + </xsl:if> +</xsl:template> + +</xsl:stylesheet> diff --git a/tools/lib/perf/Documentation/tutorial/tutorial.rst b/tools/lib/perf/Documentation/tutorial/tutorial.rst deleted file mode 100644 index 7be7bc27b385..000000000000 --- a/tools/lib/perf/Documentation/tutorial/tutorial.rst +++ /dev/null @@ -1,123 +0,0 @@ -.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -libperf tutorial -================ - -Compile and install libperf from kernel sources -=============================================== -.. code-block:: bash - - git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git - cd linux/tools/perf/lib - make - sudo make install prefix=/usr - -Libperf object -============== -The libperf library provides several high level objects: - -struct perf_cpu_map - Provides a cpu list abstraction. - -struct perf_thread_map - Provides a thread list abstraction. - -struct perf_evsel - Provides an abstraction for single a perf event. - -struct perf_evlist - Gathers several struct perf_evsel object and performs functions on all of them. - -The exported API binds these objects together, -for full reference see the libperf.7 man page. - -Examples -======== -Examples aim to explain libperf functionality on simple use cases. -They are based in on a checked out linux kernel git tree: - -.. code-block:: bash - - $ cd tools/perf/lib/Documentation/tutorial/ - $ ls -d ex-* - ex-1-compile ex-2-evsel-stat ex-3-evlist-stat - -ex-1-compile example -==================== -This example shows the basic usage of *struct perf_cpu_map*, -how to create it and display its cpus: - -.. code-block:: bash - - $ cd ex-1-compile/ - $ make - gcc -o test test.c -lperf - $ ./test - 0 1 2 3 4 5 6 7 - - -The full code listing is here: - -.. code-block:: c - - 1 #include <perf/cpumap.h> - 2 - 3 int main(int argc, char **Argv) - 4 { - 5 struct perf_cpu_map *cpus; - 6 int cpu, tmp; - 7 - 8 cpus = perf_cpu_map__new(NULL); - 9 - 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) - 11 fprintf(stdout, "%d ", cpu); - 12 - 13 fprintf(stdout, "\n"); - 14 - 15 perf_cpu_map__put(cpus); - 16 return 0; - 17 } - - -First you need to include the proper header to have *struct perf_cpumap* -declaration and functions: - -.. code-block:: c - - 1 #include <perf/cpumap.h> - - -The *struct perf_cpumap* object is created by *perf_cpu_map__new* call. -The *NULL* argument asks it to populate the object with the current online CPUs list: - -.. code-block:: c - - 8 cpus = perf_cpu_map__new(NULL); - -This is paired with a *perf_cpu_map__put*, that drops its reference at the end, possibly deleting it. - -.. code-block:: c - - 15 perf_cpu_map__put(cpus); - -The iteration through the *struct perf_cpumap* CPUs is done using the *perf_cpu_map__for_each_cpu* -macro which requires 3 arguments: - -- cpu - the cpu numer -- tmp - iteration helper variable -- cpus - the *struct perf_cpumap* object - -.. code-block:: c - - 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) - 11 fprintf(stdout, "%d ", cpu); - -ex-2-evsel-stat example -======================= - -TBD - -ex-3-evlist-stat example -======================== - -TBD diff --git a/tools/lib/perf/Makefile b/tools/lib/perf/Makefile index 768dd423730b..3718d65cffac 100644 --- a/tools/lib/perf/Makefile +++ b/tools/lib/perf/Makefile @@ -181,7 +181,10 @@ install_pkgconfig: $(LIBPERF_PC) $(call QUIET_INSTALL, $(LIBPERF_PC)) \ $(call do_install,$(LIBPERF_PC),$(libdir_SQ)/pkgconfig,644) -install: install_lib install_headers install_pkgconfig +install_doc: + $(Q)$(MAKE) -C Documentation install-man install-html install-examples + +install: install_lib install_headers install_pkgconfig install_doc FORCE: -- 2.21.0 ^ permalink raw reply related [flat|nested] 5+ messages in thread
* [tip: perf/core] libperf: Add man pages 2019-12-06 21:06 ` [PATCH 2/2] libperf: Add man pages Jiri Olsa @ 2020-01-10 17:53 ` tip-bot2 for Jiri Olsa 0 siblings, 0 replies; 5+ messages in thread From: tip-bot2 for Jiri Olsa @ 2020-01-10 17:53 UTC (permalink / raw) To: linux-tip-commits Cc: Jiri Olsa, Alexander Shishkin, Michael Petlan, Namhyung Kim, Peter Zijlstra, Arnaldo Carvalho de Melo, x86, LKML The following commit has been merged into the perf/core branch of tip: Commit-ID: 81de3bf37a8bf58ecdbef608d16ddb0f4bbb71ca Gitweb: https://git.kernel.org/tip/81de3bf37a8bf58ecdbef608d16ddb0f4bbb71ca Author: Jiri Olsa <jolsa@kernel.org> AuthorDate: Fri, 06 Dec 2019 22:06:12 +01:00 Committer: Arnaldo Carvalho de Melo <acme@redhat.com> CommitterDate: Mon, 06 Jan 2020 11:46:09 -03:00 libperf: Add man pages Change the man page generation to asciidoc, because it's easier to use and has been more commonly used in related projects. Remove the current rst pages. Add 3 man pages to have a base for more additions: libperf.3 - overall description libperf-counting.7 - counting basics explained on simple example libperf-sampling.7 - sampling basics explained on simple example The plan is to add more man pages to cover the basic API. The build generates html and man pages: $ cd tools/lib/perf/Documentation $ make ASCIIDOC libperf.xml XMLTO libperf.3 ASCIIDOC libperf-counting.xml XMLTO libperf-counting.7 ASCIIDOC libperf-sampling.xml XMLTO libperf-sampling.7 ASCIIDOC libperf.html ASCIIDOC libperf-counting.html ASCIIDOC libperf-sampling.html Add the following install targets: install-man - man pages install-html - html version of man pages install-examples - examples mentioned in the man pages Signed-off-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Michael Petlan <mpetlan@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lore.kernel.org/lkml/20191206210612.8676-3-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> --- tools/lib/perf/Documentation/Makefile | 159 +++++- tools/lib/perf/Documentation/asciidoc.conf | 120 ++++- tools/lib/perf/Documentation/examples/sampling.c | 119 ++++- tools/lib/perf/Documentation/libperf-counting.txt | 211 ++++++++- tools/lib/perf/Documentation/libperf-sampling.txt | 243 +++++++++- tools/lib/perf/Documentation/libperf.txt | 246 +++++++++- tools/lib/perf/Documentation/man/libperf.rst | 100 +---- tools/lib/perf/Documentation/manpage-1.72.xsl | 14 +- tools/lib/perf/Documentation/manpage-base.xsl | 35 +- tools/lib/perf/Documentation/manpage-bold-literal.xsl | 17 +- tools/lib/perf/Documentation/manpage-normal.xsl | 13 +- tools/lib/perf/Documentation/manpage-suppress-sp.xsl | 21 +- tools/lib/perf/Documentation/tutorial/tutorial.rst | 123 +----- tools/lib/perf/Makefile | 5 +- 14 files changed, 1197 insertions(+), 229 deletions(-) create mode 100644 tools/lib/perf/Documentation/asciidoc.conf create mode 100644 tools/lib/perf/Documentation/examples/sampling.c create mode 100644 tools/lib/perf/Documentation/libperf-counting.txt create mode 100644 tools/lib/perf/Documentation/libperf-sampling.txt create mode 100644 tools/lib/perf/Documentation/libperf.txt delete mode 100644 tools/lib/perf/Documentation/man/libperf.rst create mode 100644 tools/lib/perf/Documentation/manpage-1.72.xsl create mode 100644 tools/lib/perf/Documentation/manpage-base.xsl create mode 100644 tools/lib/perf/Documentation/manpage-bold-literal.xsl create mode 100644 tools/lib/perf/Documentation/manpage-normal.xsl create mode 100644 tools/lib/perf/Documentation/manpage-suppress-sp.xsl delete mode 100644 tools/lib/perf/Documentation/tutorial/tutorial.rst diff --git a/tools/lib/perf/Documentation/Makefile b/tools/lib/perf/Documentation/Makefile index 586425a..9727540 100644 --- a/tools/lib/perf/Documentation/Makefile +++ b/tools/lib/perf/Documentation/Makefile @@ -1,7 +1,156 @@ -all: - rst2man man/libperf.rst > man/libperf.7 - rst2pdf tutorial/tutorial.rst +# SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) +# Most of this file is copied from tools/perf/Documentation/Makefile + +include ../../../scripts/Makefile.include +include ../../../scripts/utilities.mak + +MAN3_TXT = libperf.txt +MAN7_TXT = libperf-counting.txt libperf-sampling.txt +MAN_EX = examples/*.c + +MAN_TXT = $(MAN3_TXT) $(MAN7_TXT) + +_MAN_XML = $(patsubst %.txt,%.xml,$(MAN_TXT)) +_MAN_HTML = $(patsubst %.txt,%.html,$(MAN_TXT)) +_MAN_3 = $(patsubst %.txt,%.3,$(MAN3_TXT)) +_MAN_7 = $(patsubst %.txt,%.7,$(MAN7_TXT)) + +MAN_XML = $(addprefix $(OUTPUT),$(_MAN_XML)) +MAN_HTML = $(addprefix $(OUTPUT),$(_MAN_HTML)) +MAN_3 = $(addprefix $(OUTPUT),$(_MAN_3)) +MAN_7 = $(addprefix $(OUTPUT),$(_MAN_7)) +MAN_X = $(MAN_3) $(MAN_7) + +# Make the path relative to DESTDIR, not prefix +ifndef DESTDIR + prefix ?=$(HOME) +endif + +mandir ?= $(prefix)/share/man +man3dir = $(mandir)/man3 +man7dir = $(mandir)/man7 + +docdir ?= $(prefix)/share/doc/libperf +htmldir = $(docdir)/html +exdir = $(docdir)/examples + +ASCIIDOC = asciidoc +ASCIIDOC_EXTRA = --unsafe -f asciidoc.conf +ASCIIDOC_HTML = xhtml11 +MANPAGE_XSL = manpage-normal.xsl +XMLTO_EXTRA = +XMLTO =xmlto + +INSTALL ?= install +RM ?= rm -f + +# For asciidoc ... +# -7.1.2, no extra settings are needed. +# 8.0-, set ASCIIDOC8. +# + +# For docbook-xsl ... +# -1.68.1, set ASCIIDOC_NO_ROFF? (based on changelog from 1.73.0) +# 1.69.0, no extra settings are needed? +# 1.69.1-1.71.0, set DOCBOOK_SUPPRESS_SP? +# 1.71.1, no extra settings are needed? +# 1.72.0, set DOCBOOK_XSL_172. +# 1.73.0-, set ASCIIDOC_NO_ROFF + +# If you had been using DOCBOOK_XSL_172 in an attempt to get rid +# of 'the ".ft C" problem' in your generated manpages, and you +# instead ended up with weird characters around callouts, try +# using ASCIIDOC_NO_ROFF instead (it works fine with ASCIIDOC8). + +ifdef ASCIIDOC8 + ASCIIDOC_EXTRA += -a asciidoc7compatible +endif +ifdef DOCBOOK_XSL_172 + ASCIIDOC_EXTRA += -a libperf-asciidoc-no-roff + MANPAGE_XSL = manpage-1.72.xsl +else + ifdef ASCIIDOC_NO_ROFF + # docbook-xsl after 1.72 needs the regular XSL, but will not + # pass-thru raw roff codes from asciidoc.conf, so turn them off. + ASCIIDOC_EXTRA += -a libperf-asciidoc-no-roff + endif +endif +ifdef MAN_BOLD_LITERAL + XMLTO_EXTRA += -m manpage-bold-literal.xsl +endif +ifdef DOCBOOK_SUPPRESS_SP + XMLTO_EXTRA += -m manpage-suppress-sp.xsl +endif + +DESTDIR ?= +DESTDIR_SQ = '$(subst ','\'',$(DESTDIR))' + +export DESTDIR DESTDIR_SQ + +# Please note that there is a minor bug in asciidoc. +# The version after 6.0.3 _will_ include the patch found here: +# http://marc.theaimsgroup.com/?l=libtraceevent&m=111558757202243&w=2 +# +# Until that version is released you may have to apply the patch +# yourself - yes, all 6 characters of it! + +QUIET_SUBDIR0 = +$(MAKE) -C # space to separate -C and subdir +QUIET_SUBDIR1 = + +ifneq ($(findstring $(MAKEFLAGS),w),w) + PRINT_DIR = --no-print-directory +else # "make -w" + NO_SUBDIR = : +endif + +ifneq ($(findstring $(MAKEFLAGS),s),s) + ifneq ($(V),1) + QUIET_ASCIIDOC = @echo ' ASCIIDOC '$@; + QUIET_XMLTO = @echo ' XMLTO '$@; + endif +endif + +all: $(MAN_X) $(MAN_HTML) + +$(MAN_HTML) $(MAN_X): asciidoc.conf + +install-man: all + $(call QUIET_INSTALL, man) \ + $(INSTALL) -d -m 755 $(DESTDIR)$(man3dir); \ + $(INSTALL) -m 644 $(MAN_3) $(DESTDIR)$(man3dir); \ + $(INSTALL) -d -m 755 $(DESTDIR)$(man7dir); \ + $(INSTALL) -m 644 $(MAN_7) $(DESTDIR)$(man7dir); + +install-html: + $(call QUIET_INSTALL, html) \ + $(INSTALL) -d -m 755 $(DESTDIR)$(htmldir); \ + $(INSTALL) -m 644 $(MAN_HTML) $(DESTDIR)$(htmldir); \ + +install-examples: + $(call QUIET_INSTALL, examples) \ + $(INSTALL) -d -m 755 $(DESTDIR)$(exdir); \ + $(INSTALL) -m 644 $(MAN_EX) $(DESTDIR)$(exdir); \ + +CLEAN_FILES = \ + $(MAN_XML) $(addsuffix +,$(MAN_XML)) \ + $(MAN_HTML) $(addsuffix +,$(MAN_HTML)) \ + $(MAN_X) clean: - rm -f man/libperf.7 - rm -f tutorial/tutorial.pdf + $(call QUIET_CLEAN, Documentation) $(RM) $(CLEAN_FILES) + +$(MAN_3): $(OUTPUT)%.3: %.xml + $(QUIET_XMLTO)$(XMLTO) -o $(OUTPUT). -m $(MANPAGE_XSL) $(XMLTO_EXTRA) man $< + +$(MAN_7): $(OUTPUT)%.7: %.xml + $(QUIET_XMLTO)$(XMLTO) -o $(OUTPUT). -m $(MANPAGE_XSL) $(XMLTO_EXTRA) man $< + +$(MAN_XML): $(OUTPUT)%.xml: %.txt + $(QUIET_ASCIIDOC)$(ASCIIDOC) -b docbook -d manpage \ + $(ASCIIDOC_EXTRA) -alibperf_version=$(EVENT_PARSE_VERSION) -o $@+ $< && \ + mv $@+ $@ + +$(MAN_HTML): $(OUTPUT)%.html: %.txt + $(QUIET_ASCIIDOC)$(ASCIIDOC) -b $(ASCIIDOC_HTML) -d manpage \ + $(ASCIIDOC_EXTRA) -aperf_version=$(EVENT_PARSE_VERSION) -o $@+ $< && \ + mv $@+ $@ diff --git a/tools/lib/perf/Documentation/asciidoc.conf b/tools/lib/perf/Documentation/asciidoc.conf new file mode 100644 index 0000000..9d5a5a5 --- /dev/null +++ b/tools/lib/perf/Documentation/asciidoc.conf @@ -0,0 +1,120 @@ +## linktep: macro +# +# Usage: linktep:command[manpage-section] +# +# Note, {0} is the manpage section, while {target} is the command. +# +# Show TEP link as: <command>(<section>); if section is defined, else just show +# the command. + +[macros] +(?su)[\\]?(?P<name>linktep):(?P<target>\S*?)\[(?P<attrlist>.*?)\]= + +[attributes] +asterisk=* +plus=+ +caret=^ +startsb=[ +endsb=] +tilde=~ + +ifdef::backend-docbook[] +[linktep-inlinemacro] +{0%{target}} +{0#<citerefentry>} +{0#<refentrytitle>{target}</refentrytitle><manvolnum>{0}</manvolnum>} +{0#</citerefentry>} +endif::backend-docbook[] + +ifdef::backend-docbook[] +ifndef::tep-asciidoc-no-roff[] +# "unbreak" docbook-xsl v1.68 for manpages. v1.69 works with or without this. +# v1.72 breaks with this because it replaces dots not in roff requests. +[listingblock] +<example><title>{title}</title> +<literallayout> +ifdef::doctype-manpage[] + .ft C +endif::doctype-manpage[] +| +ifdef::doctype-manpage[] + .ft +endif::doctype-manpage[] +</literallayout> +{title#}</example> +endif::tep-asciidoc-no-roff[] + +ifdef::tep-asciidoc-no-roff[] +ifdef::doctype-manpage[] +# The following two small workarounds insert a simple paragraph after screen +[listingblock] +<example><title>{title}</title> +<literallayout> +| +</literallayout><simpara></simpara> +{title#}</example> + +[verseblock] +<formalpara{id? id="{id}"}><title>{title}</title><para> +{title%}<literallayout{id? id="{id}"}> +{title#}<literallayout> +| +</literallayout> +{title#}</para></formalpara> +{title%}<simpara></simpara> +endif::doctype-manpage[] +endif::tep-asciidoc-no-roff[] +endif::backend-docbook[] + +ifdef::doctype-manpage[] +ifdef::backend-docbook[] +[header] +template::[header-declarations] +<refentry> +<refmeta> +<refentrytitle>{mantitle}</refentrytitle> +<manvolnum>{manvolnum}</manvolnum> +<refmiscinfo class="source">libperf</refmiscinfo> +<refmiscinfo class="version">{libperf_version}</refmiscinfo> +<refmiscinfo class="manual">libperf Manual</refmiscinfo> +</refmeta> +<refnamediv> + <refname>{manname1}</refname> + <refname>{manname2}</refname> + <refname>{manname3}</refname> + <refname>{manname4}</refname> + <refname>{manname5}</refname> + <refname>{manname6}</refname> + <refname>{manname7}</refname> + <refname>{manname8}</refname> + <refname>{manname9}</refname> + <refname>{manname10}</refname> + <refname>{manname11}</refname> + <refname>{manname12}</refname> + <refname>{manname13}</refname> + <refname>{manname14}</refname> + <refname>{manname15}</refname> + <refname>{manname16}</refname> + <refname>{manname17}</refname> + <refname>{manname18}</refname> + <refname>{manname19}</refname> + <refname>{manname20}</refname> + <refname>{manname21}</refname> + <refname>{manname22}</refname> + <refname>{manname23}</refname> + <refname>{manname24}</refname> + <refname>{manname25}</refname> + <refname>{manname26}</refname> + <refname>{manname27}</refname> + <refname>{manname28}</refname> + <refname>{manname29}</refname> + <refname>{manname30}</refname> + <refpurpose>{manpurpose}</refpurpose> +</refnamediv> +endif::backend-docbook[] +endif::doctype-manpage[] + +ifdef::backend-xhtml11[] +[linktep-inlinemacro] +<a href="{target}.html">{target}{0?({0})}</a> +endif::backend-xhtml11[] diff --git a/tools/lib/perf/Documentation/examples/sampling.c b/tools/lib/perf/Documentation/examples/sampling.c new file mode 100644 index 0000000..8e1a926 --- /dev/null +++ b/tools/lib/perf/Documentation/examples/sampling.c @@ -0,0 +1,119 @@ +#include <linux/perf_event.h> +#include <perf/evlist.h> +#include <perf/evsel.h> +#include <perf/cpumap.h> +#include <perf/threadmap.h> +#include <perf/mmap.h> +#include <perf/core.h> +#include <perf/event.h> +#include <stdio.h> +#include <unistd.h> + +static int libperf_print(enum libperf_print_level level, + const char *fmt, va_list ap) +{ + return vfprintf(stderr, fmt, ap); +} + +union u64_swap { + __u64 val64; + __u32 val32[2]; +}; + +int main(int argc, char **argv) +{ + struct perf_evlist *evlist; + struct perf_evsel *evsel; + struct perf_mmap *map; + struct perf_cpu_map *cpus; + struct perf_event_attr attr = { + .type = PERF_TYPE_HARDWARE, + .config = PERF_COUNT_HW_CPU_CYCLES, + .disabled = 1, + .freq = 1, + .sample_freq = 10, + .sample_type = PERF_SAMPLE_IP|PERF_SAMPLE_TID|PERF_SAMPLE_CPU|PERF_SAMPLE_PERIOD, + }; + int err = -1; + union perf_event *event; + + libperf_init(libperf_print); + + cpus = perf_cpu_map__new(NULL); + if (!cpus) { + fprintf(stderr, "failed to create cpus\n"); + return -1; + } + + evlist = perf_evlist__new(); + if (!evlist) { + fprintf(stderr, "failed to create evlist\n"); + goto out_cpus; + } + + evsel = perf_evsel__new(&attr); + if (!evsel) { + fprintf(stderr, "failed to create cycles\n"); + goto out_cpus; + } + + perf_evlist__add(evlist, evsel); + + perf_evlist__set_maps(evlist, cpus, NULL); + + err = perf_evlist__open(evlist); + if (err) { + fprintf(stderr, "failed to open evlist\n"); + goto out_evlist; + } + + err = perf_evlist__mmap(evlist, 4); + if (err) { + fprintf(stderr, "failed to mmap evlist\n"); + goto out_evlist; + } + + perf_evlist__enable(evlist); + sleep(3); + perf_evlist__disable(evlist); + + perf_evlist__for_each_mmap(evlist, map, false) { + if (perf_mmap__read_init(map) < 0) + continue; + + while ((event = perf_mmap__read_event(map)) != NULL) { + int cpu, pid, tid; + __u64 ip, period, *array; + union u64_swap u; + + array = event->sample.array; + + ip = *array; + array++; + + u.val64 = *array; + pid = u.val32[0]; + tid = u.val32[1]; + array++; + + u.val64 = *array; + cpu = u.val32[0]; + array++; + + period = *array; + + fprintf(stdout, "cpu %3d, pid %6d, tid %6d, ip %20llx, period %20llu\n", + cpu, pid, tid, ip, period); + + perf_mmap__consume(map); + } + + perf_mmap__read_done(map); + } + +out_evlist: + perf_evlist__delete(evlist); +out_cpus: + perf_cpu_map__put(cpus); + return err; +} diff --git a/tools/lib/perf/Documentation/libperf-counting.txt b/tools/lib/perf/Documentation/libperf-counting.txt new file mode 100644 index 0000000..cae9757 --- /dev/null +++ b/tools/lib/perf/Documentation/libperf-counting.txt @@ -0,0 +1,211 @@ +libperf-counting(7) +=================== + +NAME +---- +libperf-counting - counting interface + +DESCRIPTION +----------- +The counting interface provides API to meassure and get count for specific perf events. + +The following test tries to explain count on `counting.c` example. + +It is by no means complete guide to counting, but shows libperf basic API for counting. + +The `counting.c` comes with libbperf package and can be compiled and run like: + +[source,bash] +-- +$ gcc -o counting counting.c -lperf +$ sudo ./counting +count 176792, enabled 176944, run 176944 +count 176242, enabled 176242, run 176242 +-- + +It requires root access, because of the `PERF_COUNT_SW_CPU_CLOCK` event, +which is available only for root. + +The `counting.c` example monitors two events on the current process and displays their count, in a nutshel it: + +* creates events +* adds them to the event list +* opens and enables events through the event list +* does some workload +* disables events +* reads and displays event counts +* destroys the event list + +The first thing you need to do before using libperf is to call init function: + +[source,c] +-- + 8 static int libperf_print(enum libperf_print_level level, + 9 const char *fmt, va_list ap) + 10 { + 11 return vfprintf(stderr, fmt, ap); + 12 } + + 14 int main(int argc, char **argv) + 15 { + ... + 35 libperf_init(libperf_print); +-- + +It will setup the library and sets function for debug output from library. + +The `libperf_print` callback will receive any message with its debug level, +defined as: + +[source,c] +-- +enum libperf_print_level { + LIBPERF_ERR, + LIBPERF_WARN, + LIBPERF_INFO, + LIBPERF_DEBUG, + LIBPERF_DEBUG2, + LIBPERF_DEBUG3, +}; +-- + +Once the setup is complete we start by defining specific events using the `struct perf_event_attr`. + +We create software events for cpu and task: + +[source,c] +-- + 20 struct perf_event_attr attr1 = { + 21 .type = PERF_TYPE_SOFTWARE, + 22 .config = PERF_COUNT_SW_CPU_CLOCK, + 23 .read_format = PERF_FORMAT_TOTAL_TIME_ENABLED|PERF_FORMAT_TOTAL_TIME_RUNNING, + 24 .disabled = 1, + 25 }; + 26 struct perf_event_attr attr2 = { + 27 .type = PERF_TYPE_SOFTWARE, + 28 .config = PERF_COUNT_SW_TASK_CLOCK, + 29 .read_format = PERF_FORMAT_TOTAL_TIME_ENABLED|PERF_FORMAT_TOTAL_TIME_RUNNING, + 30 .disabled = 1, + 31 }; +-- + +The `read_format` setup tells perf to include timing details together with each count. + +Next step is to prepare threads map. + +In this case we will monitor current process, so we create threads map with single pid (0): + +[source,c] +-- + 37 threads = perf_thread_map__new_dummy(); + 38 if (!threads) { + 39 fprintf(stderr, "failed to create threads\n"); + 40 return -1; + 41 } + 42 + 43 perf_thread_map__set_pid(threads, 0, 0); +-- + +Now we create libperf's event list, which will serve as holder for the events we want: + +[source,c] +-- + 45 evlist = perf_evlist__new(); + 46 if (!evlist) { + 47 fprintf(stderr, "failed to create evlist\n"); + 48 goto out_threads; + 49 } +-- + +We create libperf's events for the attributes we defined earlier and add them to the list: + +[source,c] +-- + 51 evsel = perf_evsel__new(&attr1); + 52 if (!evsel) { + 53 fprintf(stderr, "failed to create evsel1\n"); + 54 goto out_evlist; + 55 } + 56 + 57 perf_evlist__add(evlist, evsel); + 58 + 59 evsel = perf_evsel__new(&attr2); + 60 if (!evsel) { + 61 fprintf(stderr, "failed to create evsel2\n"); + 62 goto out_evlist; + 63 } + 64 + 65 perf_evlist__add(evlist, evsel); +-- + +Configure event list with the thread map and open events: + +[source,c] +-- + 67 perf_evlist__set_maps(evlist, NULL, threads); + 68 + 69 err = perf_evlist__open(evlist); + 70 if (err) { + 71 fprintf(stderr, "failed to open evsel\n"); + 72 goto out_evlist; + 73 } +-- + +Both events are created as disabled (note the `disabled = 1` assignment above), +so we need to enable the whole list explicitely (both events). + +From this moment events are counting and we can do our workload. + +When we are done we disable the events list. + +[source,c] +-- + 75 perf_evlist__enable(evlist); + 76 + 77 while (count--); + 78 + 79 perf_evlist__disable(evlist); +-- + +Now we need to get the counts from events, following code iterates throught the events list and read counts: + +[source,c] +-- + 81 perf_evlist__for_each_evsel(evlist, evsel) { + 82 perf_evsel__read(evsel, 0, 0, &counts); + 83 fprintf(stdout, "count %llu, enabled %llu, run %llu\n", + 84 counts.val, counts.ena, counts.run); + 85 } +-- + +And finaly cleanup. + +We close the whole events list (both events) and remove it together with the threads map: + +[source,c] +-- + 87 perf_evlist__close(evlist); + 88 + 89 out_evlist: + 90 perf_evlist__delete(evlist); + 91 out_threads: + 92 perf_thread_map__put(threads); + 93 return err; + 94 } +-- + +REPORTING BUGS +-------------- +Report bugs to <linux-perf-users@vger.kernel.org>. + +LICENSE +------- +libperf is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git + +SEE ALSO +-------- +libperf(3), libperf-sampling(7) diff --git a/tools/lib/perf/Documentation/libperf-sampling.txt b/tools/lib/perf/Documentation/libperf-sampling.txt new file mode 100644 index 0000000..d71a7b4 --- /dev/null +++ b/tools/lib/perf/Documentation/libperf-sampling.txt @@ -0,0 +1,243 @@ +libperf-sampling(7) +=================== + +NAME +---- +libperf-sampling - sampling interface + + +DESCRIPTION +----------- +The sampling interface provides API to meassure and get count for specific perf events. + +The following test tries to explain count on `sampling.c` example. + +It is by no means complete guide to sampling, but shows libperf basic API for sampling. + +The `sampling.c` comes with libbperf package and can be compiled and run like: + +[source,bash] +-- +$ gcc -o sampling sampling.c -lperf +$ sudo ./sampling +cpu 0, pid 0, tid 0, ip ffffffffad06c4e6, period 1 +cpu 0, pid 4465, tid 4469, ip ffffffffad118748, period 18322959 +cpu 0, pid 0, tid 0, ip ffffffffad115722, period 33544846 +cpu 0, pid 4465, tid 4470, ip 7f84fe0cdad6, period 23687474 +cpu 0, pid 0, tid 0, ip ffffffffad9e0349, period 34255790 +cpu 0, pid 4465, tid 4469, ip ffffffffad136581, period 38664069 +cpu 0, pid 0, tid 0, ip ffffffffad9e55e2, period 21922384 +cpu 0, pid 4465, tid 4470, ip 7f84fe0ebebf, period 17655175 +... +-- + +It requires root access, because it uses hardware cycles event. + +The `sampling.c` example profiles/samples all CPUs with hardware cycles, in a nutshel it: + +- creates events +- adds them to the event list +- opens and enables events through the event list +- sleeps for 3 seconds +- disables events +- reads and displays recorded samples +- destroys the event list + +The first thing you need to do before using libperf is to call init function: + +[source,c] +-- + 12 static int libperf_print(enum libperf_print_level level, + 13 const char *fmt, va_list ap) + 14 { + 15 return vfprintf(stderr, fmt, ap); + 16 } + + 23 int main(int argc, char **argv) + 24 { + ... + 40 libperf_init(libperf_print); +-- + +It will setup the library and sets function for debug output from library. + +The `libperf_print` callback will receive any message with its debug level, +defined as: + +[source,c] +-- +enum libperf_print_level { + LIBPERF_ERR, + LIBPERF_WARN, + LIBPERF_INFO, + LIBPERF_DEBUG, + LIBPERF_DEBUG2, + LIBPERF_DEBUG3, +}; +-- + +Once the setup is complete we start by defining cycles event using the `struct perf_event_attr`: + +[source,c] +-- + 29 struct perf_event_attr attr = { + 30 .type = PERF_TYPE_HARDWARE, + 31 .config = PERF_COUNT_HW_CPU_CYCLES, + 32 .disabled = 1, + 33 .freq = 1, + 34 .sample_freq = 10, + 35 .sample_type = PERF_SAMPLE_IP|PERF_SAMPLE_TID|PERF_SAMPLE_CPU|PERF_SAMPLE_PERIOD, + 36 }; +-- + +Next step is to prepare cpus map. + +In this case we will monitor all the available CPUs: + +[source,c] +-- + 42 cpus = perf_cpu_map__new(NULL); + 43 if (!cpus) { + 44 fprintf(stderr, "failed to create cpus\n"); + 45 return -1; + 46 } +-- + +Now we create libperf's event list, which will serve as holder for the cycles event: + +[source,c] +-- + 48 evlist = perf_evlist__new(); + 49 if (!evlist) { + 50 fprintf(stderr, "failed to create evlist\n"); + 51 goto out_cpus; + 52 } +-- + +We create libperf's event for the cycles attribute we defined earlier and add it to the list: + +[source,c] +-- + 54 evsel = perf_evsel__new(&attr); + 55 if (!evsel) { + 56 fprintf(stderr, "failed to create cycles\n"); + 57 goto out_cpus; + 58 } + 59 + 60 perf_evlist__add(evlist, evsel); +-- + +Configure event list with the cpus map and open event: + +[source,c] +-- + 62 perf_evlist__set_maps(evlist, cpus, NULL); + 63 + 64 err = perf_evlist__open(evlist); + 65 if (err) { + 66 fprintf(stderr, "failed to open evlist\n"); + 67 goto out_evlist; + 68 } +-- + +Once the events list is open, we can create memory maps AKA perf ring buffers: + +[source,c] +-- + 70 err = perf_evlist__mmap(evlist, 4); + 71 if (err) { + 72 fprintf(stderr, "failed to mmap evlist\n"); + 73 goto out_evlist; + 74 } +-- + +The event is created as disabled (note the `disabled = 1` assignment above), +so we need to enable the events list explicitely. + +From this moment the cycles event is sampling. + +We will sleep for 3 seconds while the ring buffers get data from all CPUs, then we disable the events list. + +[source,c] +-- + 76 perf_evlist__enable(evlist); + 77 sleep(3); + 78 perf_evlist__disable(evlist); +-- + +Following code walks through the ring buffers and reads stored events/samples: + +[source,c] +-- + 80 perf_evlist__for_each_mmap(evlist, map, false) { + 81 if (perf_mmap__read_init(map) < 0) + 82 continue; + 83 + 84 while ((event = perf_mmap__read_event(map)) != NULL) { + + /* process event */ + +108 perf_mmap__consume(map); +109 } +110 perf_mmap__read_done(map); +111 } + +-- + +Each sample needs to get parsed: + +[source,c] +-- + 85 int cpu, pid, tid; + 86 __u64 ip, period, *array; + 87 union u64_swap u; + 88 + 89 array = event->sample.array; + 90 + 91 ip = *array; + 92 array++; + 93 + 94 u.val64 = *array; + 95 pid = u.val32[0]; + 96 tid = u.val32[1]; + 97 array++; + 98 + 99 u.val64 = *array; +100 cpu = u.val32[0]; +101 array++; +102 +103 period = *array; +104 +105 fprintf(stdout, "cpu %3d, pid %6d, tid %6d, ip %20llx, period %20llu\n", +106 cpu, pid, tid, ip, period); +-- + +And finaly cleanup. + +We close the whole events list (both events) and remove it together with the threads map: + +[source,c] +-- +113 out_evlist: +114 perf_evlist__delete(evlist); +115 out_cpus: +116 perf_cpu_map__put(cpus); +117 return err; +118 } +-- + +REPORTING BUGS +-------------- +Report bugs to <linux-perf-users@vger.kernel.org>. + +LICENSE +------- +libperf is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git + +SEE ALSO +-------- +libperf(3), libperf-counting(7) diff --git a/tools/lib/perf/Documentation/libperf.txt b/tools/lib/perf/Documentation/libperf.txt new file mode 100644 index 0000000..5a6bb51 --- /dev/null +++ b/tools/lib/perf/Documentation/libperf.txt @@ -0,0 +1,246 @@ +libperf(3) +========== + +NAME +---- +libperf - Linux kernel perf event library + + +SYNOPSIS +-------- +*Generic API:* + +[source,c] +-- + #include <perf/core.h> + + enum libperf_print_level { + LIBPERF_ERR, + LIBPERF_WARN, + LIBPERF_INFO, + LIBPERF_DEBUG, + LIBPERF_DEBUG2, + LIBPERF_DEBUG3, + }; + + typedef int (*libperf_print_fn_t)(enum libperf_print_level level, + const char *, va_list ap); + + void libperf_init(libperf_print_fn_t fn); +-- + +*API to handle cpu maps:* + +[source,c] +-- + #include <perf/cpumap.h> + + struct perf_cpu_map; + + struct perf_cpu_map *perf_cpu_map__dummy_new(void); + struct perf_cpu_map *perf_cpu_map__new(const char *cpu_list); + struct perf_cpu_map *perf_cpu_map__read(FILE *file); + struct perf_cpu_map *perf_cpu_map__get(struct perf_cpu_map *map); + struct perf_cpu_map *perf_cpu_map__merge(struct perf_cpu_map *orig, + struct perf_cpu_map *other); + void perf_cpu_map__put(struct perf_cpu_map *map); + int perf_cpu_map__cpu(const struct perf_cpu_map *cpus, int idx); + int perf_cpu_map__nr(const struct perf_cpu_map *cpus); + bool perf_cpu_map__empty(const struct perf_cpu_map *map); + int perf_cpu_map__max(struct perf_cpu_map *map); + + #define perf_cpu_map__for_each_cpu(cpu, idx, cpus) +-- + +*API to handle thread maps:* + +[source,c] +-- + #include <perf/threadmap.h> + + struct perf_thread_map; + + struct perf_thread_map *perf_thread_map__new_dummy(void); + + void perf_thread_map__set_pid(struct perf_thread_map *map, int thread, pid_t pid); + char *perf_thread_map__comm(struct perf_thread_map *map, int thread); + int perf_thread_map__nr(struct perf_thread_map *threads); + pid_t perf_thread_map__pid(struct perf_thread_map *map, int thread); + + struct perf_thread_map *perf_thread_map__get(struct perf_thread_map *map); + void perf_thread_map__put(struct perf_thread_map *map); +-- + +*API to handle event lists:* + +[source,c] +-- + #include <perf/evlist.h> + + struct perf_evlist; + + void perf_evlist__add(struct perf_evlist *evlist, + struct perf_evsel *evsel); + void perf_evlist__remove(struct perf_evlist *evlist, + struct perf_evsel *evsel); + struct perf_evlist *perf_evlist__new(void); + void perf_evlist__delete(struct perf_evlist *evlist); + struct perf_evsel* perf_evlist__next(struct perf_evlist *evlist, + struct perf_evsel *evsel); + int perf_evlist__open(struct perf_evlist *evlist); + void perf_evlist__close(struct perf_evlist *evlist); + void perf_evlist__enable(struct perf_evlist *evlist); + void perf_evlist__disable(struct perf_evlist *evlist); + + #define perf_evlist__for_each_evsel(evlist, pos) + + void perf_evlist__set_maps(struct perf_evlist *evlist, + struct perf_cpu_map *cpus, + struct perf_thread_map *threads); + int perf_evlist__poll(struct perf_evlist *evlist, int timeout); + int perf_evlist__filter_pollfd(struct perf_evlist *evlist, + short revents_and_mask); + + int perf_evlist__mmap(struct perf_evlist *evlist, int pages); + void perf_evlist__munmap(struct perf_evlist *evlist); + + struct perf_mmap *perf_evlist__next_mmap(struct perf_evlist *evlist, + struct perf_mmap *map, + bool overwrite); + + #define perf_evlist__for_each_mmap(evlist, pos, overwrite) +-- + +*API to handle events:* + +[source,c] +-- + #include <perf/evsel.h>* + + struct perf_evsel; + + struct perf_counts_values { + union { + struct { + uint64_t val; + uint64_t ena; + uint64_t run; + }; + uint64_t values[3]; + }; + }; + + struct perf_evsel *perf_evsel__new(struct perf_event_attr *attr); + void perf_evsel__delete(struct perf_evsel *evsel); + int perf_evsel__open(struct perf_evsel *evsel, struct perf_cpu_map *cpus, + struct perf_thread_map *threads); + void perf_evsel__close(struct perf_evsel *evsel); + void perf_evsel__close_cpu(struct perf_evsel *evsel, int cpu); + int perf_evsel__read(struct perf_evsel *evsel, int cpu, int thread, + struct perf_counts_values *count); + int perf_evsel__enable(struct perf_evsel *evsel); + int perf_evsel__enable_cpu(struct perf_evsel *evsel, int cpu); + int perf_evsel__disable(struct perf_evsel *evsel); + int perf_evsel__disable_cpu(struct perf_evsel *evsel, int cpu); + struct perf_cpu_map *perf_evsel__cpus(struct perf_evsel *evsel); + struct perf_thread_map *perf_evsel__threads(struct perf_evsel *evsel); + struct perf_event_attr *perf_evsel__attr(struct perf_evsel *evsel); +-- + +*API to handle maps (perf ring buffers):* + +[source,c] +-- + #include <perf/mmap.h> + + struct perf_mmap; + + void perf_mmap__consume(struct perf_mmap *map); + int perf_mmap__read_init(struct perf_mmap *map); + void perf_mmap__read_done(struct perf_mmap *map); + union perf_event *perf_mmap__read_event(struct perf_mmap *map); +-- + +*Structures to access perf API events:* + +[source,c] +-- + #include <perf/event.h> + + struct perf_record_mmap; + struct perf_record_mmap2; + struct perf_record_comm; + struct perf_record_namespaces; + struct perf_record_fork; + struct perf_record_lost; + struct perf_record_lost_samples; + struct perf_record_read; + struct perf_record_throttle; + struct perf_record_ksymbol; + struct perf_record_bpf_event; + struct perf_record_sample; + struct perf_record_switch; + struct perf_record_header_attr; + struct perf_record_record_cpu_map; + struct perf_record_cpu_map_data; + struct perf_record_cpu_map; + struct perf_record_event_update_cpus; + struct perf_record_event_update_scale; + struct perf_record_event_update; + struct perf_trace_event_type; + struct perf_record_header_event_type; + struct perf_record_header_tracing_data; + struct perf_record_header_build_id; + struct perf_record_id_index; + struct perf_record_auxtrace_info; + struct perf_record_auxtrace; + struct perf_record_auxtrace_error; + struct perf_record_aux; + struct perf_record_itrace_start; + struct perf_record_thread_map_entry; + struct perf_record_thread_map; + struct perf_record_stat_config_entry; + struct perf_record_stat_config; + struct perf_record_stat; + struct perf_record_stat_round; + struct perf_record_time_conv; + struct perf_record_header_feature; + struct perf_record_compressed; +-- + +DESCRIPTION +----------- +The libperf library provides an API to access the linux kernel perf +events subsystem. + +Following objects are key to the libperf interface: + +[horizontal] + +struct perf_cpu_map:: Provides a cpu list abstraction. + +struct perf_thread_map:: Provides a thread list abstraction. + +struct perf_evsel:: Provides an abstraction for single a perf event. + +struct perf_evlist:: Gathers several struct perf_evsel object and performs functions on all of them. + +struct perf_mmap:: Provides an abstraction for accessing perf ring buffer. + +The exported API functions bind these objects together. + +REPORTING BUGS +-------------- +Report bugs to <linux-perf-users@vger.kernel.org>. + +LICENSE +------- +libperf is Free Software licensed under the GNU LGPL 2.1 + +RESOURCES +--------- +https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git + +SEE ALSO +-------- +libperf-sampling(7), libperf-counting(7) diff --git a/tools/lib/perf/Documentation/man/libperf.rst b/tools/lib/perf/Documentation/man/libperf.rst deleted file mode 100644 index 09a270f..0000000 --- a/tools/lib/perf/Documentation/man/libperf.rst +++ /dev/null @@ -1,100 +0,0 @@ -.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -libperf - -The libperf library provides an API to access the linux kernel perf -events subsystem. It provides the following high level objects: - - - struct perf_cpu_map - - struct perf_thread_map - - struct perf_evlist - - struct perf_evsel - -reference -========= -Function reference by header files: - -perf/core.h ------------ -.. code-block:: c - - typedef int (\*libperf_print_fn_t)(enum libperf_print_level level, - const char \*, va_list ap); - - void libperf_set_print(libperf_print_fn_t fn); - -perf/cpumap.h -------------- -.. code-block:: c - - struct perf_cpu_map \*perf_cpu_map__dummy_new(void); - struct perf_cpu_map \*perf_cpu_map__new(const char \*cpu_list); - struct perf_cpu_map \*perf_cpu_map__read(FILE \*file); - struct perf_cpu_map \*perf_cpu_map__get(struct perf_cpu_map \*map); - void perf_cpu_map__put(struct perf_cpu_map \*map); - int perf_cpu_map__cpu(const struct perf_cpu_map \*cpus, int idx); - int perf_cpu_map__nr(const struct perf_cpu_map \*cpus); - perf_cpu_map__for_each_cpu(cpu, idx, cpus) - -perf/threadmap.h ----------------- -.. code-block:: c - - struct perf_thread_map \*perf_thread_map__new_dummy(void); - void perf_thread_map__set_pid(struct perf_thread_map \*map, int thread, pid_t pid); - char \*perf_thread_map__comm(struct perf_thread_map \*map, int thread); - struct perf_thread_map \*perf_thread_map__get(struct perf_thread_map \*map); - void perf_thread_map__put(struct perf_thread_map \*map); - -perf/evlist.h -------------- -.. code-block:: - - void perf_evlist__init(struct perf_evlist \*evlist); - void perf_evlist__add(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - void perf_evlist__remove(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - struct perf_evlist \*perf_evlist__new(void); - void perf_evlist__delete(struct perf_evlist \*evlist); - struct perf_evsel\* perf_evlist__next(struct perf_evlist \*evlist, - struct perf_evsel \*evsel); - int perf_evlist__open(struct perf_evlist \*evlist); - void perf_evlist__close(struct perf_evlist \*evlist); - void perf_evlist__enable(struct perf_evlist \*evlist); - void perf_evlist__disable(struct perf_evlist \*evlist); - perf_evlist__for_each_evsel(evlist, pos) - void perf_evlist__set_maps(struct perf_evlist \*evlist, - struct perf_cpu_map \*cpus, - struct perf_thread_map \*threads); - -perf/evsel.h ------------- -.. code-block:: c - - struct perf_counts_values { - union { - struct { - uint64_t val; - uint64_t ena; - uint64_t run; - }; - uint64_t values[3]; - }; - }; - - void perf_evsel__init(struct perf_evsel \*evsel, - struct perf_event_attr \*attr); - struct perf_evsel \*perf_evsel__new(struct perf_event_attr \*attr); - void perf_evsel__delete(struct perf_evsel \*evsel); - int perf_evsel__open(struct perf_evsel \*evsel, struct perf_cpu_map \*cpus, - struct perf_thread_map \*threads); - void perf_evsel__close(struct perf_evsel \*evsel); - int perf_evsel__read(struct perf_evsel \*evsel, int cpu, int thread, - struct perf_counts_values \*count); - int perf_evsel__enable(struct perf_evsel \*evsel); - int perf_evsel__disable(struct perf_evsel \*evsel); - int perf_evsel__apply_filter(struct perf_evsel \*evsel, const char \*filter); - struct perf_cpu_map \*perf_evsel__cpus(struct perf_evsel \*evsel); - struct perf_thread_map \*perf_evsel__threads(struct perf_evsel \*evsel); - struct perf_event_attr \*perf_evsel__attr(struct perf_evsel \*evsel); diff --git a/tools/lib/perf/Documentation/manpage-1.72.xsl b/tools/lib/perf/Documentation/manpage-1.72.xsl new file mode 100644 index 0000000..b4d315c --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-1.72.xsl @@ -0,0 +1,14 @@ +<!-- manpage-1.72.xsl: + special settings for manpages rendered from asciidoc+docbook + handles peculiarities in docbook-xsl 1.72.0 --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<xsl:import href="manpage-base.xsl"/> + +<!-- these are the special values for the roff control characters + needed for docbook-xsl 1.72.0 --> +<xsl:param name="git.docbook.backslash">▓</xsl:param> +<xsl:param name="git.docbook.dot" >⌂</xsl:param> + +</xsl:stylesheet> diff --git a/tools/lib/perf/Documentation/manpage-base.xsl b/tools/lib/perf/Documentation/manpage-base.xsl new file mode 100644 index 0000000..a264fa6 --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-base.xsl @@ -0,0 +1,35 @@ +<!-- manpage-base.xsl: + special formatting for manpages rendered from asciidoc+docbook --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<!-- these params silence some output from xmlto --> +<xsl:param name="man.output.quietly" select="1"/> +<xsl:param name="refentry.meta.get.quietly" select="1"/> + +<!-- convert asciidoc callouts to man page format; + git.docbook.backslash and git.docbook.dot params + must be supplied by another XSL file or other means --> +<xsl:template match="co"> + <xsl:value-of select="concat( + $git.docbook.backslash,'fB(', + substring-after(@id,'-'),')', + $git.docbook.backslash,'fR')"/> +</xsl:template> +<xsl:template match="calloutlist"> + <xsl:value-of select="$git.docbook.dot"/> + <xsl:text>sp </xsl:text> + <xsl:apply-templates/> + <xsl:text> </xsl:text> +</xsl:template> +<xsl:template match="callout"> + <xsl:value-of select="concat( + $git.docbook.backslash,'fB', + substring-after(@arearefs,'-'), + '. ',$git.docbook.backslash,'fR')"/> + <xsl:apply-templates/> + <xsl:value-of select="$git.docbook.dot"/> + <xsl:text>br </xsl:text> +</xsl:template> + +</xsl:stylesheet> diff --git a/tools/lib/perf/Documentation/manpage-bold-literal.xsl b/tools/lib/perf/Documentation/manpage-bold-literal.xsl new file mode 100644 index 0000000..608eb5d --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-bold-literal.xsl @@ -0,0 +1,17 @@ +<!-- manpage-bold-literal.xsl: + special formatting for manpages rendered from asciidoc+docbook --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<!-- render literal text as bold (instead of plain or monospace); + this makes literal text easier to distinguish in manpages + viewed on a tty --> +<xsl:template match="literal"> + <xsl:value-of select="$git.docbook.backslash"/> + <xsl:text>fB</xsl:text> + <xsl:apply-templates/> + <xsl:value-of select="$git.docbook.backslash"/> + <xsl:text>fR</xsl:text> +</xsl:template> + +</xsl:stylesheet> diff --git a/tools/lib/perf/Documentation/manpage-normal.xsl b/tools/lib/perf/Documentation/manpage-normal.xsl new file mode 100644 index 0000000..a48f5b1 --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-normal.xsl @@ -0,0 +1,13 @@ +<!-- manpage-normal.xsl: + special settings for manpages rendered from asciidoc+docbook + handles anything we want to keep away from docbook-xsl 1.72.0 --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<xsl:import href="manpage-base.xsl"/> + +<!-- these are the normal values for the roff control characters --> +<xsl:param name="git.docbook.backslash">\</xsl:param> +<xsl:param name="git.docbook.dot" >.</xsl:param> + +</xsl:stylesheet> diff --git a/tools/lib/perf/Documentation/manpage-suppress-sp.xsl b/tools/lib/perf/Documentation/manpage-suppress-sp.xsl new file mode 100644 index 0000000..a63c763 --- /dev/null +++ b/tools/lib/perf/Documentation/manpage-suppress-sp.xsl @@ -0,0 +1,21 @@ +<!-- manpage-suppress-sp.xsl: + special settings for manpages rendered from asciidoc+docbook + handles erroneous, inline .sp in manpage output of some + versions of docbook-xsl --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<!-- attempt to work around spurious .sp at the tail of the line + that some versions of docbook stylesheets seem to add --> +<xsl:template match="simpara"> + <xsl:variable name="content"> + <xsl:apply-templates/> + </xsl:variable> + <xsl:value-of select="normalize-space($content)"/> + <xsl:if test="not(ancestor::authorblurb) and + not(ancestor::personblurb)"> + <xsl:text> </xsl:text> + </xsl:if> +</xsl:template> + +</xsl:stylesheet> diff --git a/tools/lib/perf/Documentation/tutorial/tutorial.rst b/tools/lib/perf/Documentation/tutorial/tutorial.rst deleted file mode 100644 index 7be7bc2..0000000 --- a/tools/lib/perf/Documentation/tutorial/tutorial.rst +++ /dev/null @@ -1,123 +0,0 @@ -.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) - -libperf tutorial -================ - -Compile and install libperf from kernel sources -=============================================== -.. code-block:: bash - - git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git - cd linux/tools/perf/lib - make - sudo make install prefix=/usr - -Libperf object -============== -The libperf library provides several high level objects: - -struct perf_cpu_map - Provides a cpu list abstraction. - -struct perf_thread_map - Provides a thread list abstraction. - -struct perf_evsel - Provides an abstraction for single a perf event. - -struct perf_evlist - Gathers several struct perf_evsel object and performs functions on all of them. - -The exported API binds these objects together, -for full reference see the libperf.7 man page. - -Examples -======== -Examples aim to explain libperf functionality on simple use cases. -They are based in on a checked out linux kernel git tree: - -.. code-block:: bash - - $ cd tools/perf/lib/Documentation/tutorial/ - $ ls -d ex-* - ex-1-compile ex-2-evsel-stat ex-3-evlist-stat - -ex-1-compile example -==================== -This example shows the basic usage of *struct perf_cpu_map*, -how to create it and display its cpus: - -.. code-block:: bash - - $ cd ex-1-compile/ - $ make - gcc -o test test.c -lperf - $ ./test - 0 1 2 3 4 5 6 7 - - -The full code listing is here: - -.. code-block:: c - - 1 #include <perf/cpumap.h> - 2 - 3 int main(int argc, char **Argv) - 4 { - 5 struct perf_cpu_map *cpus; - 6 int cpu, tmp; - 7 - 8 cpus = perf_cpu_map__new(NULL); - 9 - 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) - 11 fprintf(stdout, "%d ", cpu); - 12 - 13 fprintf(stdout, "\n"); - 14 - 15 perf_cpu_map__put(cpus); - 16 return 0; - 17 } - - -First you need to include the proper header to have *struct perf_cpumap* -declaration and functions: - -.. code-block:: c - - 1 #include <perf/cpumap.h> - - -The *struct perf_cpumap* object is created by *perf_cpu_map__new* call. -The *NULL* argument asks it to populate the object with the current online CPUs list: - -.. code-block:: c - - 8 cpus = perf_cpu_map__new(NULL); - -This is paired with a *perf_cpu_map__put*, that drops its reference at the end, possibly deleting it. - -.. code-block:: c - - 15 perf_cpu_map__put(cpus); - -The iteration through the *struct perf_cpumap* CPUs is done using the *perf_cpu_map__for_each_cpu* -macro which requires 3 arguments: - -- cpu - the cpu numer -- tmp - iteration helper variable -- cpus - the *struct perf_cpumap* object - -.. code-block:: c - - 10 perf_cpu_map__for_each_cpu(cpu, tmp, cpus) - 11 fprintf(stdout, "%d ", cpu); - -ex-2-evsel-stat example -======================= - -TBD - -ex-3-evlist-stat example -======================== - -TBD diff --git a/tools/lib/perf/Makefile b/tools/lib/perf/Makefile index 768dd42..3718d65 100644 --- a/tools/lib/perf/Makefile +++ b/tools/lib/perf/Makefile @@ -181,7 +181,10 @@ install_pkgconfig: $(LIBPERF_PC) $(call QUIET_INSTALL, $(LIBPERF_PC)) \ $(call do_install,$(LIBPERF_PC),$(libdir_SQ)/pkgconfig,644) -install: install_lib install_headers install_pkgconfig +install_doc: + $(Q)$(MAKE) -C Documentation install-man install-html install-examples + +install: install_lib install_headers install_pkgconfig install_doc FORCE: ^ permalink raw reply related [flat|nested] 5+ messages in thread
end of thread, other threads:[~2020-01-10 17:54 UTC | newest] Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2019-12-06 21:06 [PATCHv2 0/2] perf/libperf move Jiri Olsa 2019-12-06 21:06 ` [PATCH 1/2] libperf: Move libperf under tools/lib/perf Jiri Olsa 2020-01-10 17:53 ` [tip: perf/core] libperf: Move to tools/lib/perf tip-bot2 for Jiri Olsa 2019-12-06 21:06 ` [PATCH 2/2] libperf: Add man pages Jiri Olsa 2020-01-10 17:53 ` [tip: perf/core] " tip-bot2 for Jiri Olsa
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).