linux/tools/perf/tests/mmap-thread-lookup.c
Kan Liang 340b47f510 perf top: Implement multithreading for perf_event__synthesize_threads
The proc files which is sorted with alphabetical order are evenly
assigned to several synthesize threads to be processed in parallel.

For 'perf top', the threads number hard code to online CPU number. The
following patch will introduce an option to set it.

For other perf tools, the thread number is 1. Because the process
function is not ready for multithreading, e.g.
process_synthesized_event.

This patch series only support event synthesize multithreading for 'perf
top'. For other tools, it can be done separately later.

With multithread applied, the total processing time can get up to 1.56x
speedup on Knights Mill for 'perf top'.

For specific single event processing, the processing time could increase
because of the lock contention. So proc_map_timeout may need to be
increased. Otherwise some proc maps will be truncated.

Based on my test, increasing the proc_map_timeout has small impact
on the total processing time. The total processing time still get 1.49x
speedup on Knights Mill after increasing the proc_map_timeout.
The patch itself doesn't increase the proc_map_timeout.

Doesn't need to implement multithreading for per task monitoring,
perf_event__synthesize_thread_map. It doesn't have performance issue.

Committer testing:

  # getconf _NPROCESSORS_ONLN
  4
  # perf trace --no-inherit -e clone -o /tmp/output perf top
  # tail -4 /tmp/bla
     0.124 ( 0.041 ms): clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7fc3eb3a8f30, parent_tidptr: 0x7fc3eb3a99d0, child_tidptr: 0x7fc3eb3a99d0, tls: 0x7fc3eb3a9700) = 9548 (perf)
     0.246 ( 0.023 ms): clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7fc3eaba7f30, parent_tidptr: 0x7fc3eaba89d0, child_tidptr: 0x7fc3eaba89d0, tls: 0x7fc3eaba8700) = 9549 (perf)
     0.286 ( 0.019 ms): clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7fc3ea3a6f30, parent_tidptr: 0x7fc3ea3a79d0, child_tidptr: 0x7fc3ea3a79d0, tls: 0x7fc3ea3a7700) = 9550 (perf)
   246.540 ( 0.047 ms): clone(flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, child_stack: 0x7fc3ea3a6f30, parent_tidptr: 0x7fc3ea3a79d0, child_tidptr: 0x7fc3ea3a79d0, tls: 0x7fc3ea3a7700) = 9551 (perf)
  #

Signed-off-by: Kan Liang <kan.liang@intel.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Lukasz Odzioba <lukasz.odzioba@intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1506696477-146932-4-git-send-email-kan.liang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-10-03 09:27:46 -03:00

236 lines
4.6 KiB
C

#include <inttypes.h>
#include <unistd.h>
#include <sys/syscall.h>
#include <sys/types.h>
#include <sys/mman.h>
#include <pthread.h>
#include <stdlib.h>
#include <stdio.h>
#include "debug.h"
#include "tests.h"
#include "machine.h"
#include "thread_map.h"
#include "symbol.h"
#include "thread.h"
#include "util.h"
#define THREADS 4
static int go_away;
struct thread_data {
pthread_t pt;
pid_t tid;
void *map;
int ready[2];
};
static struct thread_data threads[THREADS];
static int thread_init(struct thread_data *td)
{
void *map;
map = mmap(NULL, page_size,
PROT_READ|PROT_WRITE|PROT_EXEC,
MAP_SHARED|MAP_ANONYMOUS, -1, 0);
if (map == MAP_FAILED) {
perror("mmap failed");
return -1;
}
td->map = map;
td->tid = syscall(SYS_gettid);
pr_debug("tid = %d, map = %p\n", td->tid, map);
return 0;
}
static void *thread_fn(void *arg)
{
struct thread_data *td = arg;
ssize_t ret;
int go;
if (thread_init(td))
return NULL;
/* Signal thread_create thread is initialized. */
ret = write(td->ready[1], &go, sizeof(int));
if (ret != sizeof(int)) {
pr_err("failed to notify\n");
return NULL;
}
while (!go_away) {
/* Waiting for main thread to kill us. */
usleep(100);
}
munmap(td->map, page_size);
return NULL;
}
static int thread_create(int i)
{
struct thread_data *td = &threads[i];
int err, go;
if (pipe(td->ready))
return -1;
err = pthread_create(&td->pt, NULL, thread_fn, td);
if (!err) {
/* Wait for thread initialization. */
ssize_t ret = read(td->ready[0], &go, sizeof(int));
err = ret != sizeof(int);
}
close(td->ready[0]);
close(td->ready[1]);
return err;
}
static int threads_create(void)
{
struct thread_data *td0 = &threads[0];
int i, err = 0;
go_away = 0;
/* 0 is main thread */
if (thread_init(td0))
return -1;
for (i = 1; !err && i < THREADS; i++)
err = thread_create(i);
return err;
}
static int threads_destroy(void)
{
struct thread_data *td0 = &threads[0];
int i, err = 0;
/* cleanup the main thread */
munmap(td0->map, page_size);
go_away = 1;
for (i = 1; !err && i < THREADS; i++)
err = pthread_join(threads[i].pt, NULL);
return err;
}
typedef int (*synth_cb)(struct machine *machine);
static int synth_all(struct machine *machine)
{
return perf_event__synthesize_threads(NULL,
perf_event__process,
machine, 0, 500, 1);
}
static int synth_process(struct machine *machine)
{
struct thread_map *map;
int err;
map = thread_map__new_by_pid(getpid());
err = perf_event__synthesize_thread_map(NULL, map,
perf_event__process,
machine, 0, 500);
thread_map__put(map);
return err;
}
static int mmap_events(synth_cb synth)
{
struct machine *machine;
int err, i;
/*
* The threads_create will not return before all threads
* are spawned and all created memory map.
*
* They will loop until threads_destroy is called, so we
* can safely run synthesizing function.
*/
TEST_ASSERT_VAL("failed to create threads", !threads_create());
machine = machine__new_host();
dump_trace = verbose > 1 ? 1 : 0;
err = synth(machine);
dump_trace = 0;
TEST_ASSERT_VAL("failed to destroy threads", !threads_destroy());
TEST_ASSERT_VAL("failed to synthesize maps", !err);
/*
* All data is synthesized, try to find map for each
* thread object.
*/
for (i = 0; i < THREADS; i++) {
struct thread_data *td = &threads[i];
struct addr_location al;
struct thread *thread;
thread = machine__findnew_thread(machine, getpid(), td->tid);
pr_debug("looking for map %p\n", td->map);
thread__find_addr_map(thread,
PERF_RECORD_MISC_USER, MAP__FUNCTION,
(unsigned long) (td->map + 1), &al);
thread__put(thread);
if (!al.map) {
pr_debug("failed, couldn't find map\n");
err = -1;
break;
}
pr_debug("map %p, addr %" PRIx64 "\n", al.map, al.map->start);
}
machine__delete_threads(machine);
machine__delete(machine);
return err;
}
/*
* This test creates 'THREADS' number of threads (including
* main thread) and each thread creates memory map.
*
* When threads are created, we synthesize them with both
* (separate tests):
* perf_event__synthesize_thread_map (process based)
* perf_event__synthesize_threads (global)
*
* We test we can find all memory maps via:
* thread__find_addr_map
*
* by using all thread objects.
*/
int test__mmap_thread_lookup(struct test *test __maybe_unused, int subtest __maybe_unused)
{
/* perf_event__synthesize_threads synthesize */
TEST_ASSERT_VAL("failed with sythesizing all",
!mmap_events(synth_all));
/* perf_event__synthesize_thread_map synthesize */
TEST_ASSERT_VAL("failed with sythesizing process",
!mmap_events(synth_process));
return 0;
}