Commit 1ed39bac authored by Ingo Molnar's avatar Ingo Molnar

Merge tag 'perf-core-for-mingo' of...

Merge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core

Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:

 User visible changes:

  - Show precise number of samples in at the end of a 'record' session, if
    processing build ids, since we will then traverse the whole perf.data file
    and see all the PERF_RECORD_SAMPLE records, otherwise stop showing the
    previous off-base heuristicly counted number of "samples"  (Namhyung Kim).

  - Support to read compressed module from build-id cache (Namhyung Kim)

 Infrastructure changes:

  - Cache eh/debug frame offset for dwarf unwind (Namhyung Kim)

  - Set header version correctly in all cases (Namhyung Kim)

  - Set attr.task bit for a tracking event, to be consistent (Namhyung Kim)
    perf tools: Use perf_data_file__fd() consistently
    perf symbols: Convert lseek + read to pread

  - Don't rely on malloc working for sz 0, fixing another problem when
    using uClibc (Vineet Gupta)

  - Provide stub for missing pthread_attr_setaffinity_np for libcs where this
    is not available, such as uClibc (Vineet Gupta)
Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: default avatarIngo Molnar <mingo@kernel.org>
parents b3890e47 c52686f9
...@@ -68,4 +68,17 @@ futex_cmp_requeue(u_int32_t *uaddr, u_int32_t val, u_int32_t *uaddr2, int nr_wak ...@@ -68,4 +68,17 @@ futex_cmp_requeue(u_int32_t *uaddr, u_int32_t val, u_int32_t *uaddr2, int nr_wak
val, opflags); val, opflags);
} }
#ifndef HAVE_PTHREAD_ATTR_SETAFFINITY_NP
#include <pthread.h>
static inline int pthread_attr_setaffinity_np(pthread_attr_t *attr,
size_t cpusetsize,
cpu_set_t *cpuset)
{
attr = attr;
cpusetsize = cpusetsize;
cpuset = cpuset;
return 0;
}
#endif
#endif /* _FUTEX_H */ #endif /* _FUTEX_H */
...@@ -343,6 +343,7 @@ static int __cmd_inject(struct perf_inject *inject) ...@@ -343,6 +343,7 @@ static int __cmd_inject(struct perf_inject *inject)
int ret = -EINVAL; int ret = -EINVAL;
struct perf_session *session = inject->session; struct perf_session *session = inject->session;
struct perf_data_file *file_out = &inject->output; struct perf_data_file *file_out = &inject->output;
int fd = perf_data_file__fd(file_out);
signal(SIGINT, sig_handler); signal(SIGINT, sig_handler);
...@@ -376,7 +377,7 @@ static int __cmd_inject(struct perf_inject *inject) ...@@ -376,7 +377,7 @@ static int __cmd_inject(struct perf_inject *inject)
} }
if (!file_out->is_pipe) if (!file_out->is_pipe)
lseek(file_out->fd, session->header.data_offset, SEEK_SET); lseek(fd, session->header.data_offset, SEEK_SET);
ret = perf_session__process_events(session, &inject->tool); ret = perf_session__process_events(session, &inject->tool);
...@@ -385,7 +386,7 @@ static int __cmd_inject(struct perf_inject *inject) ...@@ -385,7 +386,7 @@ static int __cmd_inject(struct perf_inject *inject)
perf_header__set_feat(&session->header, perf_header__set_feat(&session->header,
HEADER_BUILD_ID); HEADER_BUILD_ID);
session->header.data_size = inject->bytes_written; session->header.data_size = inject->bytes_written;
perf_session__write_header(session, session->evlist, file_out->fd, true); perf_session__write_header(session, session->evlist, fd, true);
} }
return ret; return ret;
......
...@@ -190,16 +190,30 @@ static int record__open(struct record *rec) ...@@ -190,16 +190,30 @@ static int record__open(struct record *rec)
return rc; return rc;
} }
static int process_sample_event(struct perf_tool *tool,
union perf_event *event,
struct perf_sample *sample,
struct perf_evsel *evsel,
struct machine *machine)
{
struct record *rec = container_of(tool, struct record, tool);
rec->samples++;
return build_id__mark_dso_hit(tool, event, sample, evsel, machine);
}
static int process_buildids(struct record *rec) static int process_buildids(struct record *rec)
{ {
struct perf_data_file *file = &rec->file; struct perf_data_file *file = &rec->file;
struct perf_session *session = rec->session; struct perf_session *session = rec->session;
u64 start = session->header.data_offset;
u64 size = lseek(file->fd, 0, SEEK_CUR); u64 size = lseek(perf_data_file__fd(file), 0, SEEK_CUR);
if (size == 0) if (size == 0)
return 0; return 0;
file->size = size;
/* /*
* During this process, it'll load kernel map and replace the * During this process, it'll load kernel map and replace the
* dso->long_name to a real pathname it found. In this case * dso->long_name to a real pathname it found. In this case
...@@ -211,9 +225,7 @@ static int process_buildids(struct record *rec) ...@@ -211,9 +225,7 @@ static int process_buildids(struct record *rec)
*/ */
symbol_conf.ignore_vmlinux_buildid = true; symbol_conf.ignore_vmlinux_buildid = true;
return __perf_session__process_events(session, start, return perf_session__process_events(session, &rec->tool);
size - start,
size, &build_id__mark_dso_hit_ops);
} }
static void perf_event__synthesize_guest_os(struct machine *machine, void *data) static void perf_event__synthesize_guest_os(struct machine *machine, void *data)
...@@ -322,6 +334,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv) ...@@ -322,6 +334,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)
struct perf_data_file *file = &rec->file; struct perf_data_file *file = &rec->file;
struct perf_session *session; struct perf_session *session;
bool disabled = false, draining = false; bool disabled = false, draining = false;
int fd;
rec->progname = argv[0]; rec->progname = argv[0];
...@@ -336,6 +349,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv) ...@@ -336,6 +349,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)
return -1; return -1;
} }
fd = perf_data_file__fd(file);
rec->session = session; rec->session = session;
record__init_features(rec); record__init_features(rec);
...@@ -360,12 +374,11 @@ static int __cmd_record(struct record *rec, int argc, const char **argv) ...@@ -360,12 +374,11 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)
perf_header__clear_feat(&session->header, HEADER_GROUP_DESC); perf_header__clear_feat(&session->header, HEADER_GROUP_DESC);
if (file->is_pipe) { if (file->is_pipe) {
err = perf_header__write_pipe(file->fd); err = perf_header__write_pipe(fd);
if (err < 0) if (err < 0)
goto out_child; goto out_child;
} else { } else {
err = perf_session__write_header(session, rec->evlist, err = perf_session__write_header(session, rec->evlist, fd, false);
file->fd, false);
if (err < 0) if (err < 0)
goto out_child; goto out_child;
} }
...@@ -397,7 +410,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv) ...@@ -397,7 +410,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)
* return this more properly and also * return this more properly and also
* propagate errors that now are calling die() * propagate errors that now are calling die()
*/ */
err = perf_event__synthesize_tracing_data(tool, file->fd, rec->evlist, err = perf_event__synthesize_tracing_data(tool, fd, rec->evlist,
process_synthesized_event); process_synthesized_event);
if (err <= 0) { if (err <= 0) {
pr_err("Couldn't record tracing data.\n"); pr_err("Couldn't record tracing data.\n");
...@@ -504,19 +517,9 @@ static int __cmd_record(struct record *rec, int argc, const char **argv) ...@@ -504,19 +517,9 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)
goto out_child; goto out_child;
} }
if (!quiet) { if (!quiet)
fprintf(stderr, "[ perf record: Woken up %ld times to write data ]\n", waking); fprintf(stderr, "[ perf record: Woken up %ld times to write data ]\n", waking);
/*
* Approximate RIP event size: 24 bytes.
*/
fprintf(stderr,
"[ perf record: Captured and wrote %.3f MB %s (~%" PRIu64 " samples) ]\n",
(double)rec->bytes_written / 1024.0 / 1024.0,
file->path,
rec->bytes_written / 24);
}
out_child: out_child:
if (forks) { if (forks) {
int exit_status; int exit_status;
...@@ -535,13 +538,29 @@ static int __cmd_record(struct record *rec, int argc, const char **argv) ...@@ -535,13 +538,29 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)
} else } else
status = err; status = err;
/* this will be recalculated during process_buildids() */
rec->samples = 0;
if (!err && !file->is_pipe) { if (!err && !file->is_pipe) {
rec->session->header.data_size += rec->bytes_written; rec->session->header.data_size += rec->bytes_written;
if (!rec->no_buildid) if (!rec->no_buildid)
process_buildids(rec); process_buildids(rec);
perf_session__write_header(rec->session, rec->evlist, perf_session__write_header(rec->session, rec->evlist, fd, true);
file->fd, true); }
if (!err && !quiet) {
char samples[128];
if (rec->samples)
scnprintf(samples, sizeof(samples),
" (%" PRIu64 " samples)", rec->samples);
else
samples[0] = '\0';
fprintf(stderr, "[ perf record: Captured and wrote %.3f MB %s%s ]\n",
perf_data_file__size(file) / 1024.0 / 1024.0,
file->path, samples);
} }
out_delete_session: out_delete_session:
...@@ -720,6 +739,13 @@ static struct record record = { ...@@ -720,6 +739,13 @@ static struct record record = {
.default_per_cpu = true, .default_per_cpu = true,
}, },
}, },
.tool = {
.sample = process_sample_event,
.fork = perf_event__process_fork,
.comm = perf_event__process_comm,
.mmap = perf_event__process_mmap,
.mmap2 = perf_event__process_mmap2,
},
}; };
#define CALLCHAIN_HELP "setup and enables call-graph (stack chain/backtrace) recording: " #define CALLCHAIN_HELP "setup and enables call-graph (stack chain/backtrace) recording: "
......
...@@ -198,6 +198,7 @@ CORE_FEATURE_TESTS = \ ...@@ -198,6 +198,7 @@ CORE_FEATURE_TESTS = \
libpython-version \ libpython-version \
libslang \ libslang \
libunwind \ libunwind \
pthread-attr-setaffinity-np \
stackprotector-all \ stackprotector-all \
timerfd \ timerfd \
libdw-dwarf-unwind \ libdw-dwarf-unwind \
...@@ -226,6 +227,7 @@ VF_FEATURE_TESTS = \ ...@@ -226,6 +227,7 @@ VF_FEATURE_TESTS = \
libelf-getphdrnum \ libelf-getphdrnum \
libelf-mmap \ libelf-mmap \
libpython-version \ libpython-version \
pthread-attr-setaffinity-np \
stackprotector-all \ stackprotector-all \
timerfd \ timerfd \
libunwind-debug-frame \ libunwind-debug-frame \
...@@ -301,6 +303,10 @@ ifeq ($(feature-sync-compare-and-swap), 1) ...@@ -301,6 +303,10 @@ ifeq ($(feature-sync-compare-and-swap), 1)
CFLAGS += -DHAVE_SYNC_COMPARE_AND_SWAP_SUPPORT CFLAGS += -DHAVE_SYNC_COMPARE_AND_SWAP_SUPPORT
endif endif
ifeq ($(feature-pthread-attr-setaffinity-np), 1)
CFLAGS += -DHAVE_PTHREAD_ATTR_SETAFFINITY_NP
endif
ifndef NO_BIONIC ifndef NO_BIONIC
$(call feature_check,bionic) $(call feature_check,bionic)
ifeq ($(feature-bionic), 1) ifeq ($(feature-bionic), 1)
......
...@@ -25,6 +25,7 @@ FILES= \ ...@@ -25,6 +25,7 @@ FILES= \
test-libslang.bin \ test-libslang.bin \
test-libunwind.bin \ test-libunwind.bin \
test-libunwind-debug-frame.bin \ test-libunwind-debug-frame.bin \
test-pthread-attr-setaffinity-np.bin \
test-stackprotector-all.bin \ test-stackprotector-all.bin \
test-timerfd.bin \ test-timerfd.bin \
test-libdw-dwarf-unwind.bin \ test-libdw-dwarf-unwind.bin \
...@@ -47,6 +48,9 @@ test-all.bin: ...@@ -47,6 +48,9 @@ test-all.bin:
test-hello.bin: test-hello.bin:
$(BUILD) $(BUILD)
test-pthread-attr-setaffinity-np.bin:
$(BUILD) -Werror -lpthread
test-stackprotector-all.bin: test-stackprotector-all.bin:
$(BUILD) -Werror -fstack-protector-all $(BUILD) -Werror -fstack-protector-all
......
...@@ -97,6 +97,10 @@ ...@@ -97,6 +97,10 @@
# include "test-zlib.c" # include "test-zlib.c"
#undef main #undef main
#define main main_test_pthread_attr_setaffinity_np
# include "test-pthread_attr_setaffinity_np.c"
#undef main
int main(int argc, char *argv[]) int main(int argc, char *argv[])
{ {
main_test_libpython(); main_test_libpython();
...@@ -121,6 +125,7 @@ int main(int argc, char *argv[]) ...@@ -121,6 +125,7 @@ int main(int argc, char *argv[])
main_test_libdw_dwarf_unwind(); main_test_libdw_dwarf_unwind();
main_test_sync_compare_and_swap(argc, argv); main_test_sync_compare_and_swap(argc, argv);
main_test_zlib(); main_test_zlib();
main_test_pthread_attr_setaffinity_np();
return 0; return 0;
} }
#include <stdint.h>
#include <pthread.h>
int main(void)
{
int ret = 0;
pthread_attr_t thread_attr;
pthread_attr_init(&thread_attr);
/* don't care abt exact args, just the API itself in libpthread */
ret = pthread_attr_setaffinity_np(&thread_attr, 0, NULL);
return ret;
}
...@@ -532,12 +532,8 @@ dso_cache__read(struct dso *dso, u64 offset, u8 *data, ssize_t size) ...@@ -532,12 +532,8 @@ dso_cache__read(struct dso *dso, u64 offset, u8 *data, ssize_t size)
break; break;
cache_offset = offset & DSO__DATA_CACHE_MASK; cache_offset = offset & DSO__DATA_CACHE_MASK;
ret = -EINVAL;
if (-1 == lseek(dso->data.fd, cache_offset, SEEK_SET)) ret = pread(dso->data.fd, cache->data, DSO__DATA_CACHE_SIZE, cache_offset);
break;
ret = read(dso->data.fd, cache->data, DSO__DATA_CACHE_SIZE);
if (ret <= 0) if (ret <= 0)
break; break;
......
...@@ -139,6 +139,7 @@ struct dso { ...@@ -139,6 +139,7 @@ struct dso {
u32 status_seen; u32 status_seen;
size_t file_size; size_t file_size;
struct list_head open_entry; struct list_head open_entry;
u64 frame_offset;
} data; } data;
union { /* Tool specific area */ union { /* Tool specific area */
......
...@@ -709,6 +709,7 @@ void perf_evsel__config(struct perf_evsel *evsel, struct record_opts *opts) ...@@ -709,6 +709,7 @@ void perf_evsel__config(struct perf_evsel *evsel, struct record_opts *opts)
if (opts->sample_weight) if (opts->sample_weight)
perf_evsel__set_sample_bit(evsel, WEIGHT); perf_evsel__set_sample_bit(evsel, WEIGHT);
attr->task = track;
attr->mmap = track; attr->mmap = track;
attr->mmap2 = track && !perf_missing_features.mmap2; attr->mmap2 = track && !perf_missing_features.mmap2;
attr->comm = track; attr->comm = track;
...@@ -797,6 +798,9 @@ int perf_evsel__enable(struct perf_evsel *evsel, int ncpus, int nthreads) ...@@ -797,6 +798,9 @@ int perf_evsel__enable(struct perf_evsel *evsel, int ncpus, int nthreads)
int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads) int perf_evsel__alloc_id(struct perf_evsel *evsel, int ncpus, int nthreads)
{ {
if (ncpus == 0 || nthreads == 0)
return 0;
if (evsel->system_wide) if (evsel->system_wide)
nthreads = 1; nthreads = 1;
......
...@@ -2237,6 +2237,7 @@ static int check_magic_endian(u64 magic, uint64_t hdr_sz, ...@@ -2237,6 +2237,7 @@ static int check_magic_endian(u64 magic, uint64_t hdr_sz,
* - unique number to identify actual perf.data files * - unique number to identify actual perf.data files
* - encode endianness of file * - encode endianness of file
*/ */
ph->version = PERF_HEADER_VERSION_2;
/* check magic number with one endianness */ /* check magic number with one endianness */
if (magic == __perf_magic2) if (magic == __perf_magic2)
...@@ -2247,7 +2248,6 @@ static int check_magic_endian(u64 magic, uint64_t hdr_sz, ...@@ -2247,7 +2248,6 @@ static int check_magic_endian(u64 magic, uint64_t hdr_sz,
return -1; return -1;
ph->needs_swap = true; ph->needs_swap = true;
ph->version = PERF_HEADER_VERSION_2;
return 0; return 0;
} }
......
...@@ -1251,7 +1251,7 @@ fetch_mmaped_event(struct perf_session *session, ...@@ -1251,7 +1251,7 @@ fetch_mmaped_event(struct perf_session *session,
#define NUM_MMAPS 128 #define NUM_MMAPS 128
#endif #endif
int __perf_session__process_events(struct perf_session *session, static int __perf_session__process_events(struct perf_session *session,
u64 data_offset, u64 data_size, u64 data_offset, u64 data_size,
u64 file_size, struct perf_tool *tool) u64 file_size, struct perf_tool *tool)
{ {
......
...@@ -49,9 +49,6 @@ int perf_session__peek_event(struct perf_session *session, off_t file_offset, ...@@ -49,9 +49,6 @@ int perf_session__peek_event(struct perf_session *session, off_t file_offset,
union perf_event **event_ptr, union perf_event **event_ptr,
struct perf_sample *sample); struct perf_sample *sample);
int __perf_session__process_events(struct perf_session *session,
u64 data_offset, u64 data_size, u64 size,
struct perf_tool *tool);
int perf_session__process_events(struct perf_session *session, int perf_session__process_events(struct perf_session *session,
struct perf_tool *tool); struct perf_tool *tool);
......
...@@ -574,13 +574,16 @@ static int decompress_kmodule(struct dso *dso, const char *name, ...@@ -574,13 +574,16 @@ static int decompress_kmodule(struct dso *dso, const char *name,
const char *ext = strrchr(name, '.'); const char *ext = strrchr(name, '.');
char tmpbuf[] = "/tmp/perf-kmod-XXXXXX"; char tmpbuf[] = "/tmp/perf-kmod-XXXXXX";
if ((type != DSO_BINARY_TYPE__SYSTEM_PATH_KMODULE_COMP && if (type != DSO_BINARY_TYPE__SYSTEM_PATH_KMODULE_COMP &&
type != DSO_BINARY_TYPE__GUEST_KMODULE_COMP) || type != DSO_BINARY_TYPE__GUEST_KMODULE_COMP &&
type != dso->symtab_type) type != DSO_BINARY_TYPE__BUILD_ID_CACHE)
return -1; return -1;
if (!ext || !is_supported_compression(ext + 1)) {
ext = strrchr(dso->name, '.');
if (!ext || !is_supported_compression(ext + 1)) if (!ext || !is_supported_compression(ext + 1))
return -1; return -1;
}
fd = mkstemp(tmpbuf); fd = mkstemp(tmpbuf);
if (fd < 0) if (fd < 0)
......
...@@ -266,14 +266,17 @@ static int read_unwind_spec_eh_frame(struct dso *dso, struct machine *machine, ...@@ -266,14 +266,17 @@ static int read_unwind_spec_eh_frame(struct dso *dso, struct machine *machine,
u64 *fde_count) u64 *fde_count)
{ {
int ret = -EINVAL, fd; int ret = -EINVAL, fd;
u64 offset; u64 offset = dso->data.frame_offset;
if (offset == 0) {
fd = dso__data_fd(dso, machine); fd = dso__data_fd(dso, machine);
if (fd < 0) if (fd < 0)
return -EINVAL; return -EINVAL;
/* Check the .eh_frame section for unwinding info */ /* Check the .eh_frame section for unwinding info */
offset = elf_section_offset(fd, ".eh_frame_hdr"); offset = elf_section_offset(fd, ".eh_frame_hdr");
dso->data.frame_offset = offset;
}
if (offset) if (offset)
ret = unwind_spec_ehframe(dso, machine, offset, ret = unwind_spec_ehframe(dso, machine, offset,
...@@ -287,14 +290,20 @@ static int read_unwind_spec_eh_frame(struct dso *dso, struct machine *machine, ...@@ -287,14 +290,20 @@ static int read_unwind_spec_eh_frame(struct dso *dso, struct machine *machine,
static int read_unwind_spec_debug_frame(struct dso *dso, static int read_unwind_spec_debug_frame(struct dso *dso,
struct machine *machine, u64 *offset) struct machine *machine, u64 *offset)
{ {
int fd = dso__data_fd(dso, machine); int fd;
u64 ofs = dso->data.frame_offset;
if (ofs == 0) {
fd = dso__data_fd(dso, machine);
if (fd < 0) if (fd < 0)
return -EINVAL; return -EINVAL;
/* Check the .debug_frame section for unwinding info */ /* Check the .debug_frame section for unwinding info */
*offset = elf_section_offset(fd, ".debug_frame"); ofs = elf_section_offset(fd, ".debug_frame");
dso->data.frame_offset = ofs;
}
*offset = ofs;
if (*offset) if (*offset)
return 0; return 0;
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment