- 29 Jul, 2019 40 commits
-
-
Jiri Olsa authored
Add the perf_evsel__init() function to initialize perf_evsel struct. Signed-off-by:
Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Michael Petlan <mpetlan@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20190721112506.12306-38-jolsa@kernel.orgSigned-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com>
-
Jiri Olsa authored
Include perf_evlist in the evlist object, will continue to move other generic things into libperf's perf_evlist. Signed-off-by:
-
Jiri Olsa authored
Including perf_evsel in evsel object, will continue to move other generic things into libperf's perf_evsel struct. Signed-off-by:
-
Jiri Olsa authored
Add the perf_evlist and perf_evsel structs to libperf. It's added as a declarations into: include/perf/evlist.h include/perf/evsel.h which will be included by users. The definitions are added into: include/internal/evlist.h include/internal/evsel.h which is not to be included by users, but shared within perf and libperf. Signed-off-by:
-
Jiri Olsa authored
Move the following functions: thread_map__get() thread_map__put() thread_map__comm() to libperf with the following names: perf_thread_map__get() perf_thread_map__put() perf_thread_map__comm() Add the perf_thread_map__comm() function for it to work/compile. Signed-off-by:
-
Jiri Olsa authored
Moving the following functions: thread_map__new_dummy() thread_map__realloc() thread_map__set_pid() to libperf with the following names: perf_thread_map__new_dummy() perf_thread_map__realloc() perf_thread_map__set_pid() the other 2 functions are dependencies of the perf_thread_map__new_dummy() function. The perf_thread_map__realloc() function is not exported. Signed-off-by:
-
Jiri Olsa authored
Add perf_thread_map struct to libperf. It's added as a declaration into into: include/perf/threadmap.h which will be included by users. The perf_thread_map struct definition is added into: include/internal/threadmap.h which is not to be included by users, but shared within perf and libperf. We tried the total separation of the perf_thread_map struct in libperf, but it lead to complications and much bigger changes in perf code, so we decided to share the declaration. Signed-off-by:
-
Jiri Olsa authored
Moving the following functions: cpu_map__get() cpu_map__put() to libperf with following names: perf_cpu_map__get() perf_cpu_map__put() Committer notes: Added fixes for arm/arm64 Signed-off-by:
-
Jiri Olsa authored
Move cpu_map__dummy_new() to libperf as perf_cpu_map__dummy_new() function. Signed-off-by:
-
Jiri Olsa authored
Add perf_cpu_map struct to libperf. It's added as a declaration into: include/perf/cpumap.h which will be included by users. The perf_cpu_map struct definition is added into: include/internal/cpumap.h which is not to be included by users, but shared within perf and libperf. We tried the total separation of the perf_cpu_map struct in libperf, but it lead to complications and much bigger changes in perf code, so we decided to share the declaration. Signed-off-by:
-
Jiri Olsa authored
Add the perf_set_print() function to allow setting an output function for warn/info/debug messages. Signed-off-by:
-
Jiri Olsa authored
Add perf/core.h header to be used in header files coming in the following patches. Signed-off-by:
-
Jiri Olsa authored
Link libperf.a with python.so. Committer testing: Continues to work: # perf test python 18: 'import perf' in python : Ok # Signed-off-by:
-
Jiri Olsa authored
Add a shared library version, generating the following files: $ ll tools/perf/lib/libperf.so* libperf.so -> libperf.so.0.0.1 libperf.so.0 -> libperf.so.0.0.1 libperf.so.0.0.1 Committer testing: One has to build just libbperf to get this, building perf so far doesn't trigger this, i.e. I tried: $ make O=/tmp/build/perf -C tools/perf And the files above were not created, so one has to do: $ make O=/tmp/build/perf -C tools/perf/lib/ make: Entering directory '/home/acme/git/perf/tools/perf/lib' LINK /tmp/build/perf/libperf.so.0.0.1 make: Leaving directory '/home/acme/git/perf/tools/perf/lib' $ ls -la /tmp/build/perf/*.so.* lrwxrwxrwx. 1 acme acme 16 Jul 22 15:37 /tmp/build/perf/libperf.so.0 -> libperf.so.0.0.1 -rwxrwxr-x. 1 acme acme 16368 Jul 22 15:37 /tmp/build/perf/libperf.so.0.0.1 $ Signed-off-by: -
Jiri Olsa authored
Add an empty libperf.a under tools/perf/lib and link it with perf. It can also be built separately with: $ cd tools/perf/lib && make CC core.o LD libperf-in.o AR libperf.a LINK libperf.so Committer testing: $ make O=/tmp/build/perf -C tools/perf/lib/ make: Entering directory '/home/acme/git/perf/tools/perf/lib' LINK /tmp/build/perf/libperf.so make: Leaving directory '/home/acme/git/perf/tools/perf/lib' $ ls -la /tmp/build/perf/libperf.so -rwxrwxr-x. 1 acme acme 16232 Jul 22 15:30 /tmp/build/perf/libperf.so $ file /tmp/build/perf/libperf.so /tmp/build/perf/libperf.so: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked, BuildID[sha1]=7a51d227d871b381ddb686dcf94145c4dd908221, not stripped $ git status tools/perf On branch perf/core nothing to commit, working tree clean $ $ ls -lart tools/perf/lib/ total 16 drwxrwxr-x. 16 acme acme 4096 Jul 22 15:29 .. -rw-rw-r--. 1 acme acme 1633 Jul 22 15:29 Makefile -rw-rw-r--. 1 acme acme 0 Jul 22 15:29 core.c -rw-rw-r--. 1 acme acme 20 Jul 22 15:29 Build drwxrwxr-x. 2 acme acme 4096 Jul 22 15:29 . $ Committer notes: Need to add -I$(srctree)/tools/arch/$(ARCH)/include/uapi -I$(srctree)/tools/include/uapi to tools/perf/lib/Makefile's INCLUDE variable to pick up the latest versions of kernel headers, even in older systems, this is in line with what is in tools/lib/bpf/Makefile. Signed-off-by: -
Jiri Olsa authored
Rename perf_evlist__disable() to evlist__disable(), so we don't have a name clash when we add perf_evlist__disable() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evlist__enable() to evlist__enable(), so we don't have a name clash when we add perf_evlist__enable() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evlist__close() to evlist__close(), so we don't have a name clash when we add perf_evlist__close() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evlist__open() to evlist__open(), so we don't have a name clash when we add perf_evlist__open() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evsel__cpus() to evsel__cpus(), so we don't have a name clash when we add perf_evsel__cpus() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evsel__apply_filter() to evsel__apply_filter(), so we don't have a name clash when we add perf_evsel__apply_filter() in libperf. Signed-off-by:
-
Jiri Olsa authored
Renaming perf_evsel__disable() to evsel__disable(), so we don't have a name clash when we add perf_evsel__disable() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evsel__enable() to evsel__enable(), so we don't have a name clash when we add perf_evsel__enable() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evsel__open() to evsel__open(), so we don't have a name clash when we add perf_evsel__open() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evlist__remove() to evlist__remove(), so we don't have a name clash when we add perf_evlist__remove() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evlist__add() to evlist__add(), so we don't have a name clash when we add perf_evlist__add() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evsel__new() to evsel__new(), so we don't have a name clash when we add perf_evsel__new() in libperf. Signed-off-by:
-
Jiri Olsa authored
Remame perf_evsel__delete() to evsel__delete(), so we don't have a name clash when we add perf_evsel__delete() in libperf. Also renaming perf_evsel__delete_priv() to evsel__delete_priv(). Signed-off-by:
-
Jiri Olsa authored
Rename perf_evlist__delete() to evlist__delete(), so we don't have a name clash when we add perf_evlist__delete() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evlist__new() to evlist__new(), so we don't have a name clash when we add perf_evlist__new() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evlist__init() to evlist__init(), so we don't have a name clash when we add perf_evlist__init() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename perf_evsel__init() to evsel__init(), so we don't have a name clash when we add perf_evsel__init() in libperf. Signed-off-by:
-
Jiri Olsa authored
Rename struct perf_evlist to struct evlist, so we don't have a name clash when we add struct perf_evlist in libperf. Committer notes: Added fixes to build on arm64, from Jiri and from me (tools/perf/util/cs-etm.c) Signed-off-by:
-
Jiri Olsa authored
Rename struct perf_evsel to struct evsel, so we don't have a name clash when we add struct perf_evsel in libperf. Committer notes: Added fixes for arm64, provided by Jiri. Signed-off-by:
-
Jiri Olsa authored
Rename struct thread_map to struct perf_thread_map, so it could be part of libperf. Signed-off-by:
-
Jiri Olsa authored
Rename struct cpu_map to struct perf_cpu_map, so it could be part of libperf. Committer notes: Added fixes for arm64, provided by Jiri. Signed-off-by:
-
Jiri Olsa authored
Because we will make struct perf_counts_values public in following patches and 'loaded' is implementation related. No functional change is expected. Signed-off-by:
-
Arnaldo Carvalho de Melo authored
We were looking in tracefs for: /sys/kernel/debug/tracing/events/syscalls/sys_enter_sendfile/format when what is there is just /sys/kernel/debug/tracing/events/syscalls/sys_enter_sendfile/format Its the same id, 40 in x86_64, so just add an alias and let the existing logic take care of that. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-km2hmg7hru6u4pawi5fi903q@git.kernel.orgSigned-off-by:
-
Arnaldo Carvalho de Melo authored
We have an augmenter for the "open" syscall, which has just one pointer, in the first argument, a "const char *", so any other syscall that has just one pointer and that is the first can reuse the "open" BPF augmenter program. Even more, syscalls that get two pointers with the first being a string can reuse "open"'s BPF augmenter till we have an augmenter that better matches that syscall with two pointers. With this the few augmenters we have, for open (first arg is a string), openat (2nd arg is a string), renameat (2nd and 4th are strings) can be reused by a lot of syscalls, ditto for "bind" reusing "connect" because both have the 2nd argument as a sockaddr and the 3rd as its len. Lets see how this makes the "bind" syscall reuse the "connect" BPF prog augmenter found in tools/perf/examples/bpf/augmented_raw_syscalls.c: # perf trace -e bind,connect systemctl restart sshd connect(3, { .family: PF_LOCAL, path: /run/systemd/private }, 23) = 0 # Oh, it just connects to some daemon, so we better do it system wide and then stop/start sshd: # perf trace -e bind,connect systemctl/10124 connect(3, { .family: PF_LOCAL, path: /run/systemd/private }, 23) = 0 sshd/10102 connect(7, { .family: PF_LOCAL, path: /dev/log }, 110) = 0 systemctl/10126 connect(3, { .family: PF_LOCAL, path: /run/systemd/private }, 23) = 0 systemd/10128 ... [continued]: connect()) = 0 (sshd)/10128 connect(3, { .family: PF_LOCAL, path: /run/systemd/journal/stdout }, 30) ... sshd/10128 bind(3, { .family: PF_NETLINK }, 12) = 0 sshd/10128 connect(4, { .family: PF_LOCAL, path: /var/run/nscd/socket }, 110) = -1 ENOENT (No such file or directory) sshd/10128 connect(3, { .family: PF_INET6, port: 22, addr: :: }, 28) = 0 sshd/10128 connect(3, { .family: PF_UNSPEC }, 16) = 0 sshd/10128 connect(3, { .family: PF_INET, port: 22, addr: 0.0.0.0 }, 16) = 0 sshd/10128 connect(3, { .family: PF_LOCAL, path: /var/run/nscd/socket }, 110) = -1 ENOENT (No such file or directory) sshd/10128 connect(3, { .family: PF_LOCAL, path: /var/run/nscd/socket }, 110) = -1 ENOENT (No such file or directory) sshd/10128 connect(5, { .family: PF_LOCAL, path: /var/run/nscd/socket }, 110) = -1 ENOENT (No such file or directory) sshd/10128 connect(5, { .family: PF_LOCAL, path: /var/run/nscd/socket }, 110) = -1 ENOENT (No such file or directory) sshd/10128 bind(4, { .family: PF_INET, port: 22, addr: 0.0.0.0 }, 16) = 0 sshd/10128 connect(6, { .family: PF_LOCAL, path: /dev/log }, 110) = 0 sshd/10128 bind(6, { .family: PF_INET6, port: 22, addr: :: }, 28) = 0 sshd/10128 connect(7, { .family: PF_LOCAL, path: /dev/log }, 110) = 0 ^C# Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-zfley2ghs4nim1uq4nu6ed3l@git.kernel.orgSigned-off-by: -
Arnaldo Carvalho de Melo authored
We'll continue reading its details from tracefs as we need it, but preallocate the whole thing otherwise we may realloc and end up with pointers to the previous buffer. I.e. in an upcoming algorithm we'll look for syscalls that have function signatures that are similar to a given syscall to see if we can reuse its BPF augmenter, so we may be at syscall 42, having a 'struct syscall' pointing to that slot in trace->syscalls.table[] and try to read the slot for an yet unread syscall, which would realloc that table to read the info for syscall 43, say, which would trigger a realoc of trace->syscalls.table[], and then the pointer we had for syscall 42 would be pointing to the previous block of memory. b00m. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Brendan Gregg <brendan.d.gregg@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Luis Cláudio Gonçalves <lclaudio@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-m3cjzzifibs13imafhkk77a0@git.kernel.orgSigned-off-by:
-
