static int do_test(struct perf_evlist *evlist, int mmap_pages,
int *sample_count, int *comm_count)
{
int err;
char sbuf[STRERR_BUFSIZE];
err = perf_evlist__mmap(evlist, mmap_pages, true);
if (err < 0) {
pr_debug("perf_evlist__mmap: %s\n",
strerror_r(errno, sbuf, sizeof(sbuf)));
return TEST_FAIL;
}
perf_evlist__enable(evlist);
testcase();
perf_evlist__disable(evlist);
err = count_samples(evlist, sample_count, comm_count);
perf_evlist__munmap(evlist);
return err;
}
/**
* test__keep_tracking - test using a dummy software event to keep tracking.
*
* This function implements a test that checks that tracking events continue
* when an event is disabled but a dummy software event is not disabled. If the
* test passes %0 is returned, otherwise %-1 is returned.
*/
int test__keep_tracking(void)
{
struct record_opts opts = {
.mmap_pages = UINT_MAX,
.user_freq = UINT_MAX,
.user_interval = ULLONG_MAX,
.freq = 4000,
.target = {
.uses_mmap = true,
},
};
struct thread_map *threads = NULL;
struct cpu_map *cpus = NULL;
struct perf_evlist *evlist = NULL;
struct perf_evsel *evsel = NULL;
int found, err = -1;
const char *comm;
threads = thread_map__new(-1, getpid(), UINT_MAX);
CHECK_NOT_NULL__(threads);
cpus = cpu_map__new(NULL);
CHECK_NOT_NULL__(cpus);
evlist = perf_evlist__new();
CHECK_NOT_NULL__(evlist);
perf_evlist__set_maps(evlist, cpus, threads);
CHECK__(parse_events(evlist, "dummy:u", NULL));
CHECK__(parse_events(evlist, "cycles:u", NULL));
perf_evlist__config(evlist, &opts);
evsel = perf_evlist__first(evlist);
evsel->attr.comm = 1;
evsel->attr.disabled = 1;
evsel->attr.enable_on_exec = 0;
if (perf_evlist__open(evlist) < 0) {
fprintf(stderr, " (not supported)");
err = 0;
goto out_err;
}
CHECK__(perf_evlist__mmap(evlist, UINT_MAX, false));
/*
* First, test that a 'comm' event can be found when the event is
* enabled.
*/
perf_evlist__enable(evlist);
comm = "Test COMM 1";
CHECK__(prctl(PR_SET_NAME, (unsigned long)comm, 0, 0, 0));
perf_evlist__disable(evlist);
found = find_comm(evlist, comm);
if (found != 1) {
pr_debug("First time, failed to find tracking event.\n");
goto out_err;
}
/*
* Secondly, test that a 'comm' event can be found when the event is
* disabled with the dummy event still enabled.
*/
perf_evlist__enable(evlist);
evsel = perf_evlist__last(evlist);
CHECK__(perf_evlist__disable_event(evlist, evsel));
comm = "Test COMM 2";
CHECK__(prctl(PR_SET_NAME, (unsigned long)comm, 0, 0, 0));
perf_evlist__disable(evlist);
found = find_comm(evlist, comm);
if (found != 1) {
pr_debug("Seconf time, failed to find tracking event.\n");
goto out_err;
}
err = 0;
out_err:
if (evlist) {
perf_evlist__disable(evlist);
perf_evlist__delete(evlist);
} else {
cpu_map__put(cpus);
thread_map__put(threads);
}
return err;
}
static int do_test(struct bpf_object *obj, int (*func)(void),
int expect)
{
struct record_opts opts = {
.target = {
.uid = UINT_MAX,
.uses_mmap = true,
},
.freq = 0,
.mmap_pages = 256,
.default_interval = 1,
};
char pid[16];
char sbuf[STRERR_BUFSIZE];
struct perf_evlist *evlist;
int i, ret = TEST_FAIL, err = 0, count = 0;
struct parse_events_evlist parse_evlist;
struct parse_events_error parse_error;
bzero(&parse_error, sizeof(parse_error));
bzero(&parse_evlist, sizeof(parse_evlist));
parse_evlist.error = &parse_error;
INIT_LIST_HEAD(&parse_evlist.list);
err = parse_events_load_bpf_obj(&parse_evlist, &parse_evlist.list, obj, NULL);
if (err || list_empty(&parse_evlist.list)) {
pr_debug("Failed to add events selected by BPF\n");
return TEST_FAIL;
}
snprintf(pid, sizeof(pid), "%d", getpid());
pid[sizeof(pid) - 1] = '\0';
opts.target.tid = opts.target.pid = pid;
/* Instead of perf_evlist__new_default, don't add default events */
evlist = perf_evlist__new();
if (!evlist) {
pr_debug("Not enough memory to create evlist\n");
return TEST_FAIL;
}
err = perf_evlist__create_maps(evlist, &opts.target);
if (err < 0) {
pr_debug("Not enough memory to create thread/cpu maps\n");
goto out_delete_evlist;
}
perf_evlist__splice_list_tail(evlist, &parse_evlist.list);
evlist->nr_groups = parse_evlist.nr_groups;
perf_evlist__config(evlist, &opts, NULL);
err = perf_evlist__open(evlist);
if (err < 0) {
pr_debug("perf_evlist__open: %s\n",
str_error_r(errno, sbuf, sizeof(sbuf)));
goto out_delete_evlist;
}
err = perf_evlist__mmap(evlist, opts.mmap_pages, false);
if (err < 0) {
pr_debug("perf_evlist__mmap: %s\n",
str_error_r(errno, sbuf, sizeof(sbuf)));
goto out_delete_evlist;
}
perf_evlist__enable(evlist);
(*func)();
perf_evlist__disable(evlist);
for (i = 0; i < evlist->nr_mmaps; i++) {
union perf_event *event;
while ((event = perf_evlist__mmap_read(evlist, i)) != NULL) {
const u32 type = event->header.type;
if (type == PERF_RECORD_SAMPLE)
count ++;
}
}
if (count != expect) {
pr_debug("BPF filter result incorrect\n");
goto out_delete_evlist;
}
ret = TEST_OK;
out_delete_evlist:
perf_evlist__delete(evlist);
return ret;
}
static int do_test_code_reading(bool try_kcore)
{
struct machines machines;
struct machine *machine;
struct thread *thread;
struct record_opts opts = {
.mmap_pages = UINT_MAX,
.user_freq = UINT_MAX,
.user_interval = ULLONG_MAX,
.freq = 4000,
.target = {
.uses_mmap = true,
},
};
struct state state = {
.done_cnt = 0,
};
struct thread_map *threads = NULL;
struct cpu_map *cpus = NULL;
struct perf_evlist *evlist = NULL;
struct perf_evsel *evsel = NULL;
int err = -1, ret;
pid_t pid;
struct map *map;
bool have_vmlinux, have_kcore, excl_kernel = false;
pid = getpid();
machines__init(&machines);
machine = &machines.host;
ret = machine__create_kernel_maps(machine);
if (ret < 0) {
pr_debug("machine__create_kernel_maps failed\n");
goto out_err;
}
/* Force the use of kallsyms instead of vmlinux to try kcore */
if (try_kcore)
symbol_conf.kallsyms_name = "/proc/kallsyms";
/* Load kernel map */
map = machine->vmlinux_maps[MAP__FUNCTION];
ret = map__load(map, NULL);
if (ret < 0) {
pr_debug("map__load failed\n");
goto out_err;
}
have_vmlinux = dso__is_vmlinux(map->dso);
have_kcore = dso__is_kcore(map->dso);
/* 2nd time through we just try kcore */
if (try_kcore && !have_kcore)
return TEST_CODE_READING_NO_KCORE;
/* No point getting kernel events if there is no kernel object */
if (!have_vmlinux && !have_kcore)
excl_kernel = true;
threads = thread_map__new_by_tid(pid);
if (!threads) {
pr_debug("thread_map__new_by_tid failed\n");
goto out_err;
}
ret = perf_event__synthesize_thread_map(NULL, threads,
perf_event__process, machine, false);
if (ret < 0) {
pr_debug("perf_event__synthesize_thread_map failed\n");
goto out_err;
}
thread = machine__findnew_thread(machine, pid, pid);
if (!thread) {
pr_debug("machine__findnew_thread failed\n");
goto out_err;
}
cpus = cpu_map__new(NULL);
if (!cpus) {
pr_debug("cpu_map__new failed\n");
goto out_err;
}
while (1) {
const char *str;
evlist = perf_evlist__new();
if (!evlist) {
pr_debug("perf_evlist__new failed\n");
goto out_err;
}
perf_evlist__set_maps(evlist, cpus, threads);
if (excl_kernel)
str = "cycles:u";
else
str = "cycles";
pr_debug("Parsing event '%s'\n", str);
ret = parse_events(evlist, str);
if (ret < 0) {
pr_debug("parse_events failed\n");
goto out_err;
}
perf_evlist__config(evlist, &opts);
evsel = perf_evlist__first(evlist);
evsel->attr.comm = 1;
evsel->attr.disabled = 1;
evsel->attr.enable_on_exec = 0;
ret = perf_evlist__open(evlist);
if (ret < 0) {
if (!excl_kernel) {
excl_kernel = true;
perf_evlist__set_maps(evlist, NULL, NULL);
perf_evlist__delete(evlist);
evlist = NULL;
continue;
}
pr_debug("perf_evlist__open failed\n");
goto out_err;
}
break;
}
ret = perf_evlist__mmap(evlist, UINT_MAX, false);
if (ret < 0) {
pr_debug("perf_evlist__mmap failed\n");
goto out_err;
}
perf_evlist__enable(evlist);
do_something();
perf_evlist__disable(evlist);
ret = process_events(machine, evlist, &state);
if (ret < 0)
goto out_err;
if (!have_vmlinux && !have_kcore && !try_kcore)
err = TEST_CODE_READING_NO_KERNEL_OBJ;
else if (!have_vmlinux && !try_kcore)
err = TEST_CODE_READING_NO_VMLINUX;
else if (excl_kernel)
err = TEST_CODE_READING_NO_ACCESS;
else
err = TEST_CODE_READING_OK;
out_err:
if (evlist) {
perf_evlist__delete(evlist);
} else {
cpu_map__delete(cpus);
thread_map__delete(threads);
}
machines__destroy_kernel_maps(&machines);
machine__delete_threads(machine);
machines__exit(&machines);
return err;
}
int test__code_reading(void)
{
int ret;
ret = do_test_code_reading(false);
if (!ret)
ret = do_test_code_reading(true);
switch (ret) {
case TEST_CODE_READING_OK:
return 0;
case TEST_CODE_READING_NO_VMLINUX:
fprintf(stderr, " (no vmlinux)");
return 0;
case TEST_CODE_READING_NO_KCORE:
fprintf(stderr, " (no kcore)");
return 0;
case TEST_CODE_READING_NO_ACCESS:
fprintf(stderr, " (no access)");
return 0;
case TEST_CODE_READING_NO_KERNEL_OBJ:
fprintf(stderr, " (no kernel obj)");
return 0;
default:
return -1;
};
}
int test__PERF_RECORD(int subtest __maybe_unused)
{
struct record_opts opts = {
.target = {
.uid = UINT_MAX,
.uses_mmap = true,
},
.no_buffering = true,
.mmap_pages = 256,
};
cpu_set_t cpu_mask;
size_t cpu_mask_size = sizeof(cpu_mask);
struct perf_evlist *evlist = perf_evlist__new_dummy();
struct perf_evsel *evsel;
struct perf_sample sample;
const char *cmd = "sleep";
const char *argv[] = { cmd, "1", NULL, };
char *bname, *mmap_filename;
u64 prev_time = 0;
bool found_cmd_mmap = false,
found_libc_mmap = false,
found_vdso_mmap = false,
found_ld_mmap = false;
int err = -1, errs = 0, i, wakeups = 0;
u32 cpu;
int total_events = 0, nr_events[PERF_RECORD_MAX] = { 0, };
char sbuf[STRERR_BUFSIZE];
if (evlist == NULL) /* Fallback for kernels lacking PERF_COUNT_SW_DUMMY */
evlist = perf_evlist__new_default();
if (evlist == NULL || argv == NULL) {
pr_debug("Not enough memory to create evlist\n");
goto out;
}
/*
* Create maps of threads and cpus to monitor. In this case
* we start with all threads and cpus (-1, -1) but then in
* perf_evlist__prepare_workload we'll fill in the only thread
* we're monitoring, the one forked there.
*/
err = perf_evlist__create_maps(evlist, &opts.target);
if (err < 0) {
pr_debug("Not enough memory to create thread/cpu maps\n");
goto out_delete_evlist;
}
/*
* Prepare the workload in argv[] to run, it'll fork it, and then wait
* for perf_evlist__start_workload() to exec it. This is done this way
* so that we have time to open the evlist (calling sys_perf_event_open
* on all the fds) and then mmap them.
*/
err = perf_evlist__prepare_workload(evlist, &opts.target, argv, false, NULL);
if (err < 0) {
pr_debug("Couldn't run the workload!\n");
goto out_delete_evlist;
}
/*
* Config the evsels, setting attr->comm on the first one, etc.
*/
evsel = perf_evlist__first(evlist);
perf_evsel__set_sample_bit(evsel, CPU);
perf_evsel__set_sample_bit(evsel, TID);
perf_evsel__set_sample_bit(evsel, TIME);
perf_evlist__config(evlist, &opts);
err = sched__get_first_possible_cpu(evlist->workload.pid, &cpu_mask);
if (err < 0) {
pr_debug("sched__get_first_possible_cpu: %s\n",
strerror_r(errno, sbuf, sizeof(sbuf)));
goto out_delete_evlist;
}
cpu = err;
/*
* So that we can check perf_sample.cpu on all the samples.
*/
if (sched_setaffinity(evlist->workload.pid, cpu_mask_size, &cpu_mask) < 0) {
pr_debug("sched_setaffinity: %s\n",
strerror_r(errno, sbuf, sizeof(sbuf)));
goto out_delete_evlist;
}
/*
* Call sys_perf_event_open on all the fds on all the evsels,
* grouping them if asked to.
*/
err = perf_evlist__open(evlist);
if (err < 0) {
pr_debug("perf_evlist__open: %s\n",
strerror_r(errno, sbuf, sizeof(sbuf)));
goto out_delete_evlist;
}
/*
* mmap the first fd on a given CPU and ask for events for the other
* fds in the same CPU to be injected in the same mmap ring buffer
* (using ioctl(PERF_EVENT_IOC_SET_OUTPUT)).
*/
err = perf_evlist__mmap(evlist, opts.mmap_pages, false);
if (err < 0) {
pr_debug("perf_evlist__mmap: %s\n",
strerror_r(errno, sbuf, sizeof(sbuf)));
goto out_delete_evlist;
}
/*
* Now that all is properly set up, enable the events, they will
* count just on workload.pid, which will start...
*/
perf_evlist__enable(evlist);
/*
* Now!
*/
perf_evlist__start_workload(evlist);
while (1) {
int before = total_events;
for (i = 0; i < evlist->nr_mmaps; i++) {
union perf_event *event;
while ((event = perf_evlist__mmap_read(evlist, i)) != NULL) {
const u32 type = event->header.type;
const char *name = perf_event__name(type);
++total_events;
if (type < PERF_RECORD_MAX)
nr_events[type]++;
err = perf_evlist__parse_sample(evlist, event, &sample);
if (err < 0) {
if (verbose)
perf_event__fprintf(event, stderr);
pr_debug("Couldn't parse sample\n");
goto out_delete_evlist;
}
if (verbose) {
pr_info("%" PRIu64" %d ", sample.time, sample.cpu);
perf_event__fprintf(event, stderr);
}
if (prev_time > sample.time) {
pr_debug("%s going backwards in time, prev=%" PRIu64 ", curr=%" PRIu64 "\n",
name, prev_time, sample.time);
++errs;
}
prev_time = sample.time;
if (sample.cpu != cpu) {
pr_debug("%s with unexpected cpu, expected %d, got %d\n",
name, cpu, sample.cpu);
++errs;
}
if ((pid_t)sample.pid != evlist->workload.pid) {
pr_debug("%s with unexpected pid, expected %d, got %d\n",
name, evlist->workload.pid, sample.pid);
++errs;
}
if ((pid_t)sample.tid != evlist->workload.pid) {
pr_debug("%s with unexpected tid, expected %d, got %d\n",
name, evlist->workload.pid, sample.tid);
++errs;
}
if ((type == PERF_RECORD_COMM ||
type == PERF_RECORD_MMAP ||
type == PERF_RECORD_MMAP2 ||
type == PERF_RECORD_FORK ||
type == PERF_RECORD_EXIT) &&
(pid_t)event->comm.pid != evlist->workload.pid) {
pr_debug("%s with unexpected pid/tid\n", name);
++errs;
}
if ((type == PERF_RECORD_COMM ||
type == PERF_RECORD_MMAP ||
type == PERF_RECORD_MMAP2) &&
event->comm.pid != event->comm.tid) {
pr_debug("%s with different pid/tid!\n", name);
++errs;
}
switch (type) {
case PERF_RECORD_COMM:
if (strcmp(event->comm.comm, cmd)) {
pr_debug("%s with unexpected comm!\n", name);
++errs;
}
break;
case PERF_RECORD_EXIT:
goto found_exit;
case PERF_RECORD_MMAP:
mmap_filename = event->mmap.filename;
goto check_bname;
case PERF_RECORD_MMAP2:
mmap_filename = event->mmap2.filename;
check_bname:
bname = strrchr(mmap_filename, '/');
if (bname != NULL) {
if (!found_cmd_mmap)
found_cmd_mmap = !strcmp(bname + 1, cmd);
if (!found_libc_mmap)
found_libc_mmap = !strncmp(bname + 1, "libc", 4);
if (!found_ld_mmap)
found_ld_mmap = !strncmp(bname + 1, "ld", 2);
} else if (!found_vdso_mmap)
found_vdso_mmap = !strcmp(mmap_filename, "[vdso]");
break;
case PERF_RECORD_SAMPLE:
/* Just ignore samples for now */
break;
default:
pr_debug("Unexpected perf_event->header.type %d!\n",
type);
++errs;
}
perf_evlist__mmap_consume(evlist, i);
}
}
/*
* We don't use poll here because at least at 3.1 times the
* PERF_RECORD_{!SAMPLE} events don't honour
* perf_event_attr.wakeup_events, just PERF_EVENT_SAMPLE does.
*/
if (total_events == before && false)
perf_evlist__poll(evlist, -1);
sleep(1);
if (++wakeups > 5) {
pr_debug("No PERF_RECORD_EXIT event!\n");
break;
}
}
found_exit:
if (nr_events[PERF_RECORD_COMM] > 1) {
pr_debug("Excessive number of PERF_RECORD_COMM events!\n");
++errs;
}
if (nr_events[PERF_RECORD_COMM] == 0) {
pr_debug("Missing PERF_RECORD_COMM for %s!\n", cmd);
++errs;
}
if (!found_cmd_mmap) {
pr_debug("PERF_RECORD_MMAP for %s missing!\n", cmd);
++errs;
}
if (!found_libc_mmap) {
pr_debug("PERF_RECORD_MMAP for %s missing!\n", "libc");
++errs;
}
if (!found_ld_mmap) {
pr_debug("PERF_RECORD_MMAP for %s missing!\n", "ld");
++errs;
}
if (!found_vdso_mmap) {
pr_debug("PERF_RECORD_MMAP for %s missing!\n", "[vdso]");
++errs;
}
out_delete_evlist:
perf_evlist__delete(evlist);
out:
return (err < 0 || errs > 0) ? -1 : 0;
}
/*
* This test will start a workload that does nothing then it checks
* if the number of exit event reported by the kernel is 1 or not
* in order to check the kernel returns correct number of event.
*/
int test__task_exit(void)
{
int err = -1;
union perf_event *event;
struct perf_evsel *evsel;
struct perf_evlist *evlist;
struct perf_target target = {
.uid = UINT_MAX,
.uses_mmap = true,
};
const char *argv[] = { "true", NULL };
signal(SIGCHLD, sig_handler);
signal(SIGUSR1, sig_handler);
evlist = perf_evlist__new();
if (evlist == NULL) {
pr_debug("perf_evlist__new\n");
return -1;
}
/*
* We need at least one evsel in the evlist, use the default
* one: "cycles".
*/
err = perf_evlist__add_default(evlist);
if (err < 0) {
pr_debug("Not enough memory to create evsel\n");
goto out_free_evlist;
}
/*
* Create maps of threads and cpus to monitor. In this case
* we start with all threads and cpus (-1, -1) but then in
* perf_evlist__prepare_workload we'll fill in the only thread
* we're monitoring, the one forked there.
*/
evlist->cpus = cpu_map__dummy_new();
evlist->threads = thread_map__new_by_tid(-1);
if (!evlist->cpus || !evlist->threads) {
err = -ENOMEM;
pr_debug("Not enough memory to create thread/cpu maps\n");
goto out_delete_maps;
}
err = perf_evlist__prepare_workload(evlist, &target, argv, false, true);
if (err < 0) {
pr_debug("Couldn't run the workload!\n");
goto out_delete_maps;
}
evsel = perf_evlist__first(evlist);
evsel->attr.task = 1;
evsel->attr.sample_freq = 0;
evsel->attr.inherit = 0;
evsel->attr.watermark = 0;
evsel->attr.wakeup_events = 1;
evsel->attr.exclude_kernel = 1;
err = perf_evlist__open(evlist);
if (err < 0) {
pr_debug("Couldn't open the evlist: %s\n", strerror(-err));
goto out_delete_maps;
}
if (perf_evlist__mmap(evlist, 128, true) < 0) {
pr_debug("failed to mmap events: %d (%s)\n", errno,
strerror(errno));
goto out_close_evlist;
}
perf_evlist__start_workload(evlist);
retry:
while ((event = perf_evlist__mmap_read(evlist, 0)) != NULL) {
if (event->header.type != PERF_RECORD_EXIT)
continue;
nr_exit++;
}
if (!exited || !nr_exit) {
poll(evlist->pollfd, evlist->nr_fds, -1);
goto retry;
}
if (nr_exit != 1) {
pr_debug("received %d EXIT records\n", nr_exit);
err = -1;
}
perf_evlist__munmap(evlist);
out_close_evlist:
perf_evlist__close(evlist);
out_delete_maps:
perf_evlist__delete_maps(evlist);
out_free_evlist:
perf_evlist__delete(evlist);
return err;
}
int test__syscall_open_tp_fields(void)
{
struct perf_record_opts opts = {
.target = {
.uid = UINT_MAX,
.uses_mmap = true,
},
.no_delay = true,
.freq = 1,
.mmap_pages = 256,
.raw_samples = true,
};
const char *filename = "/etc/passwd";
int flags = O_RDONLY | O_DIRECTORY;
struct perf_evlist *evlist = perf_evlist__new();
struct perf_evsel *evsel;
int err = -1, i, nr_events = 0, nr_polls = 0;
if (evlist == NULL) {
pr_debug("%s: perf_evlist__new\n", __func__);
goto out;
}
evsel = perf_evsel__newtp("syscalls", "sys_enter_open");
if (evsel == NULL) {
pr_debug("%s: perf_evsel__newtp\n", __func__);
goto out_delete_evlist;
}
perf_evlist__add(evlist, evsel);
err = perf_evlist__create_maps(evlist, &opts.target);
if (err < 0) {
pr_debug("%s: perf_evlist__create_maps\n", __func__);
goto out_delete_evlist;
}
perf_evsel__config(evsel, &opts);
evlist->threads->map[0] = getpid();
err = perf_evlist__open(evlist);
if (err < 0) {
pr_debug("perf_evlist__open: %s\n", strerror(errno));
goto out_delete_maps;
}
err = perf_evlist__mmap(evlist, UINT_MAX, false);
if (err < 0) {
pr_debug("perf_evlist__mmap: %s\n", strerror(errno));
goto out_close_evlist;
}
perf_evlist__enable(evlist);
/*
* Generate the event:
*/
open(filename, flags);
while (1) {
int before = nr_events;
for (i = 0; i < evlist->nr_mmaps; i++) {
union perf_event *event;
while ((event = perf_evlist__mmap_read(evlist, i)) != NULL) {
const u32 type = event->header.type;
int tp_flags;
struct perf_sample sample;
++nr_events;
if (type != PERF_RECORD_SAMPLE) {
perf_evlist__mmap_consume(evlist, i);
continue;
}
err = perf_evsel__parse_sample(evsel, event, &sample);
if (err) {
pr_err("Can't parse sample, err = %d\n", err);
goto out_munmap;
}
tp_flags = perf_evsel__intval(evsel, &sample, "flags");
if (flags != tp_flags) {
pr_debug("%s: Expected flags=%#x, got %#x\n",
__func__, flags, tp_flags);
goto out_munmap;
}
goto out_ok;
}
}
if (nr_events == before)
poll(evlist->pollfd, evlist->nr_fds, 10);
if (++nr_polls > 5) {
pr_debug("%s: no events!\n", __func__);
goto out_munmap;
}
}
out_ok:
err = 0;
out_munmap:
perf_evlist__munmap(evlist);
out_close_evlist:
perf_evlist__close(evlist);
out_delete_maps:
perf_evlist__delete_maps(evlist);
out_delete_evlist:
perf_evlist__delete(evlist);
out:
return err;
}
/*
* This test will generate random numbers of calls to some getpid syscalls,
* then establish an mmap for a group of events that are created to monitor
* the syscalls.
*
* It will receive the events, using mmap, use its PERF_SAMPLE_ID generated
* sample.id field to map back to its respective perf_evsel instance.
*
* Then it checks if the number of syscalls reported as perf events by
* the kernel corresponds to the number of syscalls made.
*/
int test__basic_mmap(void)
{
int err = -1;
union perf_event *event;
struct thread_map *threads;
struct cpu_map *cpus;
struct perf_evlist *evlist;
cpu_set_t cpu_set;
const char *syscall_names[] = { "getsid", "getppid", "getpgrp",
"getpgid", };
pid_t (*syscalls[])(void) = { (void *)getsid, getppid, getpgrp,
(void*)getpgid };
#define nsyscalls ARRAY_SIZE(syscall_names)
unsigned int nr_events[nsyscalls],
expected_nr_events[nsyscalls], i, j;
struct perf_evsel *evsels[nsyscalls], *evsel;
threads = thread_map__new(-1, getpid(), UINT_MAX);
if (threads == NULL) {
pr_debug("thread_map__new\n");
return -1;
}
cpus = cpu_map__new(NULL);
if (cpus == NULL) {
pr_debug("cpu_map__new\n");
goto out_free_threads;
}
CPU_ZERO(&cpu_set);
CPU_SET(cpus->map[0], &cpu_set);
sched_setaffinity(0, sizeof(cpu_set), &cpu_set);
if (sched_setaffinity(0, sizeof(cpu_set), &cpu_set) < 0) {
pr_debug("sched_setaffinity() failed on CPU %d: %s ",
cpus->map[0], strerror(errno));
goto out_free_cpus;
}
evlist = perf_evlist__new();
if (evlist == NULL) {
pr_debug("perf_evlist__new\n");
goto out_free_cpus;
}
perf_evlist__set_maps(evlist, cpus, threads);
for (i = 0; i < nsyscalls; ++i) {
char name[64];
snprintf(name, sizeof(name), "sys_enter_%s", syscall_names[i]);
evsels[i] = perf_evsel__newtp("syscalls", name);
if (evsels[i] == NULL) {
pr_debug("perf_evsel__new\n");
goto out_delete_evlist;
}
evsels[i]->attr.wakeup_events = 1;
perf_evsel__set_sample_id(evsels[i], false);
perf_evlist__add(evlist, evsels[i]);
if (perf_evsel__open(evsels[i], cpus, threads) < 0) {
pr_debug("failed to open counter: %s, "
"tweak /proc/sys/kernel/perf_event_paranoid?\n",
strerror(errno));
goto out_delete_evlist;
}
nr_events[i] = 0;
expected_nr_events[i] = 1 + rand() % 127;
}
if (perf_evlist__mmap(evlist, 128, true) < 0) {
pr_debug("failed to mmap events: %d (%s)\n", errno,
strerror(errno));
goto out_delete_evlist;
}
for (i = 0; i < nsyscalls; ++i)
for (j = 0; j < expected_nr_events[i]; ++j) {
int foo = syscalls[i]();
++foo;
}
while ((event = perf_evlist__mmap_read(evlist, 0)) != NULL) {
struct perf_sample sample;
if (event->header.type != PERF_RECORD_SAMPLE) {
pr_debug("unexpected %s event\n",
perf_event__name(event->header.type));
goto out_delete_evlist;
}
err = perf_evlist__parse_sample(evlist, event, &sample);
if (err) {
pr_err("Can't parse sample, err = %d\n", err);
goto out_delete_evlist;
}
err = -1;
evsel = perf_evlist__id2evsel(evlist, sample.id);
if (evsel == NULL) {
pr_debug("event with id %" PRIu64
" doesn't map to an evsel\n", sample.id);
goto out_delete_evlist;
}
nr_events[evsel->idx]++;
perf_evlist__mmap_consume(evlist, 0);
}
err = 0;
evlist__for_each(evlist, evsel) {
if (nr_events[evsel->idx] != expected_nr_events[evsel->idx]) {
pr_debug("expected %d %s events, got %d\n",
expected_nr_events[evsel->idx],
perf_evsel__name(evsel), nr_events[evsel->idx]);
err = -1;
goto out_delete_evlist;
}
}
out_delete_evlist:
perf_evlist__delete(evlist);
cpus = NULL;
threads = NULL;
out_free_cpus:
cpu_map__delete(cpus);
out_free_threads:
thread_map__delete(threads);
return err;
}
/*
* This test will open software clock events (cpu-clock, task-clock)
* then check their frequency -> period conversion has no artifact of
* setting period to 1 forcefully.
*/
static int __test__sw_clock_freq(enum perf_sw_ids clock_id)
{
int i, err = -1;
volatile int tmp = 0;
u64 total_periods = 0;
int nr_samples = 0;
char sbuf[STRERR_BUFSIZE];
union perf_event *event;
struct perf_evsel *evsel;
struct perf_evlist *evlist;
struct perf_event_attr attr = {
.type = PERF_TYPE_SOFTWARE,
.config = clock_id,
.sample_type = PERF_SAMPLE_PERIOD,
.exclude_kernel = 1,
.disabled = 1,
.freq = 1,
};
struct cpu_map *cpus;
struct thread_map *threads;
attr.sample_freq = 500;
evlist = perf_evlist__new();
if (evlist == NULL) {
pr_debug("perf_evlist__new\n");
return -1;
}
evsel = perf_evsel__new(&attr);
if (evsel == NULL) {
pr_debug("perf_evsel__new\n");
goto out_delete_evlist;
}
perf_evlist__add(evlist, evsel);
cpus = cpu_map__dummy_new();
threads = thread_map__new_by_tid(getpid());
if (!cpus || !threads) {
err = -ENOMEM;
pr_debug("Not enough memory to create thread/cpu maps\n");
goto out_free_maps;
}
perf_evlist__set_maps(evlist, cpus, threads);
cpus = NULL;
threads = NULL;
if (perf_evlist__open(evlist)) {
const char *knob = "/proc/sys/kernel/perf_event_max_sample_rate";
err = -errno;
pr_debug("Couldn't open evlist: %s\nHint: check %s, using %" PRIu64 " in this test.\n",
str_error_r(errno, sbuf, sizeof(sbuf)),
knob, (u64)attr.sample_freq);
goto out_delete_evlist;
}
err = perf_evlist__mmap(evlist, 128);
if (err < 0) {
pr_debug("failed to mmap event: %d (%s)\n", errno,
str_error_r(errno, sbuf, sizeof(sbuf)));
goto out_delete_evlist;
}
perf_evlist__enable(evlist);
/* collect samples */
for (i = 0; i < NR_LOOPS; i++)
tmp++;
perf_evlist__disable(evlist);
while ((event = perf_evlist__mmap_read(evlist, 0)) != NULL) {
struct perf_sample sample;
if (event->header.type != PERF_RECORD_SAMPLE)
goto next_event;
err = perf_evlist__parse_sample(evlist, event, &sample);
if (err < 0) {
pr_debug("Error during parse sample\n");
goto out_delete_evlist;
}
total_periods += sample.period;
nr_samples++;
next_event:
perf_evlist__mmap_consume(evlist, 0);
}
if ((u64) nr_samples == total_periods) {
pr_debug("All (%d) samples have period value of 1!\n",
nr_samples);
err = -1;
}
out_free_maps:
cpu_map__put(cpus);
thread_map__put(threads);
out_delete_evlist:
perf_evlist__delete(evlist);
return err;
}
int test__sw_clock_freq(struct test *test __maybe_unused, int subtest __maybe_unused)
{
int ret;
ret = __test__sw_clock_freq(PERF_COUNT_SW_CPU_CLOCK);
if (!ret)
ret = __test__sw_clock_freq(PERF_COUNT_SW_TASK_CLOCK);
return ret;
}
/**
* test__perf_time_to_tsc - test converting perf time to TSC.
*
* This function implements a test that checks that the conversion of perf time
* to and from TSC is consistent with the order of events. If the test passes
* %0 is returned, otherwise %-1 is returned. If TSC conversion is not
* supported then then the test passes but " (not supported)" is printed.
*/
int test__perf_time_to_tsc(void)
{
struct record_opts opts = {
.mmap_pages = UINT_MAX,
.user_freq = UINT_MAX,
.user_interval = ULLONG_MAX,
.freq = 4000,
.target = {
.uses_mmap = true,
},
.sample_time = true,
};
struct thread_map *threads = NULL;
struct cpu_map *cpus = NULL;
struct perf_evlist *evlist = NULL;
struct perf_evsel *evsel = NULL;
int err = -1, ret, i;
const char *comm1, *comm2;
struct perf_tsc_conversion tc;
struct perf_event_mmap_page *pc;
union perf_event *event;
u64 test_tsc, comm1_tsc, comm2_tsc;
u64 test_time, comm1_time = 0, comm2_time = 0;
threads = thread_map__new(-1, getpid(), UINT_MAX);
CHECK_NOT_NULL__(threads);
cpus = cpu_map__new(NULL);
CHECK_NOT_NULL__(cpus);
evlist = perf_evlist__new();
CHECK_NOT_NULL__(evlist);
perf_evlist__set_maps(evlist, cpus, threads);
CHECK__(parse_events(evlist, "cycles:u", NULL));
perf_evlist__config(evlist, &opts);
evsel = perf_evlist__first(evlist);
evsel->attr.comm = 1;
evsel->attr.disabled = 1;
evsel->attr.enable_on_exec = 0;
CHECK__(perf_evlist__open(evlist));
CHECK__(perf_evlist__mmap(evlist, UINT_MAX, false));
pc = evlist->mmap[0].base;
ret = perf_read_tsc_conversion(pc, &tc);
if (ret) {
if (ret == -EOPNOTSUPP) {
fprintf(stderr, " (not supported)");
return 0;
}
goto out_err;
}
perf_evlist__enable(evlist);
comm1 = "Test COMM 1";
CHECK__(prctl(PR_SET_NAME, (unsigned long)comm1, 0, 0, 0));
test_tsc = rdtsc();
comm2 = "Test COMM 2";
CHECK__(prctl(PR_SET_NAME, (unsigned long)comm2, 0, 0, 0));
perf_evlist__disable(evlist);
for (i = 0; i < evlist->nr_mmaps; i++) {
while ((event = perf_evlist__mmap_read(evlist, i)) != NULL) {
struct perf_sample sample;
if (event->header.type != PERF_RECORD_COMM ||
(pid_t)event->comm.pid != getpid() ||
(pid_t)event->comm.tid != getpid())
goto next_event;
if (strcmp(event->comm.comm, comm1) == 0) {
CHECK__(perf_evsel__parse_sample(evsel, event,
&sample));
comm1_time = sample.time;
}
if (strcmp(event->comm.comm, comm2) == 0) {
CHECK__(perf_evsel__parse_sample(evsel, event,
&sample));
comm2_time = sample.time;
}
next_event:
perf_evlist__mmap_consume(evlist, i);
}
}
if (!comm1_time || !comm2_time)
goto out_err;
test_time = tsc_to_perf_time(test_tsc, &tc);
comm1_tsc = perf_time_to_tsc(comm1_time, &tc);
comm2_tsc = perf_time_to_tsc(comm2_time, &tc);
pr_debug("1st event perf time %"PRIu64" tsc %"PRIu64"\n",
comm1_time, comm1_tsc);
pr_debug("rdtsc time %"PRIu64" tsc %"PRIu64"\n",
test_time, test_tsc);
pr_debug("2nd event perf time %"PRIu64" tsc %"PRIu64"\n",
comm2_time, comm2_tsc);
if (test_time <= comm1_time ||
test_time >= comm2_time)
goto out_err;
if (test_tsc <= comm1_tsc ||
test_tsc >= comm2_tsc)
goto out_err;
err = 0;
out_err:
if (evlist) {
perf_evlist__disable(evlist);
perf_evlist__delete(evlist);
}
return err;
}
static int do_test_code_reading(bool try_kcore)
{
struct machine *machine;
struct thread *thread;
struct record_opts opts = {
.mmap_pages = UINT_MAX,
.user_freq = UINT_MAX,
.user_interval = ULLONG_MAX,
.freq = 500,
.target = {
.uses_mmap = true,
},
};
struct state state = {
.done_cnt = 0,
};
struct thread_map *threads = NULL;
struct cpu_map *cpus = NULL;
struct perf_evlist *evlist = NULL;
struct perf_evsel *evsel = NULL;
int err = -1, ret;
pid_t pid;
struct map *map;
bool have_vmlinux, have_kcore, excl_kernel = false;
pid = getpid();
machine = machine__new_host();
ret = machine__create_kernel_maps(machine);
if (ret < 0) {
pr_debug("machine__create_kernel_maps failed\n");
goto out_err;
}
/* Force the use of kallsyms instead of vmlinux to try kcore */
if (try_kcore)
symbol_conf.kallsyms_name = "/proc/kallsyms";
/* Load kernel map */
map = machine__kernel_map(machine);
ret = map__load(map, NULL);
if (ret < 0) {
pr_debug("map__load failed\n");
goto out_err;
}
have_vmlinux = dso__is_vmlinux(map->dso);
have_kcore = dso__is_kcore(map->dso);
/* 2nd time through we just try kcore */
if (try_kcore && !have_kcore)
return TEST_CODE_READING_NO_KCORE;
/* No point getting kernel events if there is no kernel object */
if (!have_vmlinux && !have_kcore)
excl_kernel = true;
threads = thread_map__new_by_tid(pid);
if (!threads) {
pr_debug("thread_map__new_by_tid failed\n");
goto out_err;
}
ret = perf_event__synthesize_thread_map(NULL, threads,
perf_event__process, machine, false, 500);
if (ret < 0) {
pr_debug("perf_event__synthesize_thread_map failed\n");
goto out_err;
}
thread = machine__findnew_thread(machine, pid, pid);
if (!thread) {
pr_debug("machine__findnew_thread failed\n");
goto out_put;
}
cpus = cpu_map__new(NULL);
if (!cpus) {
pr_debug("cpu_map__new failed\n");
goto out_put;
}
while (1) {
const char *str;
evlist = perf_evlist__new();
if (!evlist) {
pr_debug("perf_evlist__new failed\n");
goto out_put;
}
perf_evlist__set_maps(evlist, cpus, threads);
if (excl_kernel)
str = "cycles:u";
else
str = "cycles";
pr_debug("Parsing event '%s'\n", str);
ret = parse_events(evlist, str, NULL);
if (ret < 0) {
pr_debug("parse_events failed\n");
goto out_put;
}
perf_evlist__config(evlist, &opts);
evsel = perf_evlist__first(evlist);
evsel->attr.comm = 1;
evsel->attr.disabled = 1;
evsel->attr.enable_on_exec = 0;
ret = perf_evlist__open(evlist);
if (ret < 0) {
if (!excl_kernel) {
excl_kernel = true;
/*
* Both cpus and threads are now owned by evlist
* and will be freed by following perf_evlist__set_maps
* call. Getting refference to keep them alive.
*/
cpu_map__get(cpus);
thread_map__get(threads);
perf_evlist__set_maps(evlist, NULL, NULL);
perf_evlist__delete(evlist);
evlist = NULL;
continue;
}
if (verbose) {
char errbuf[512];
perf_evlist__strerror_open(evlist, errno, errbuf, sizeof(errbuf));
pr_debug("perf_evlist__open() failed!\n%s\n", errbuf);
}
goto out_put;
}
break;
}
ret = perf_evlist__mmap(evlist, UINT_MAX, false);
if (ret < 0) {
pr_debug("perf_evlist__mmap failed\n");
goto out_put;
}
perf_evlist__enable(evlist);
do_something();
perf_evlist__disable(evlist);
ret = process_events(machine, evlist, &state);
if (ret < 0)
goto out_put;
if (!have_vmlinux && !have_kcore && !try_kcore)
err = TEST_CODE_READING_NO_KERNEL_OBJ;
else if (!have_vmlinux && !try_kcore)
err = TEST_CODE_READING_NO_VMLINUX;
else if (excl_kernel)
err = TEST_CODE_READING_NO_ACCESS;
else
err = TEST_CODE_READING_OK;
out_put:
thread__put(thread);
out_err:
if (evlist) {
perf_evlist__delete(evlist);
} else {
cpu_map__put(cpus);
thread_map__put(threads);
}
machine__delete_threads(machine);
machine__delete(machine);
return err;
}
int test__code_reading(int subtest __maybe_unused)
{
int ret;
ret = do_test_code_reading(false);
if (!ret)
ret = do_test_code_reading(true);
switch (ret) {
case TEST_CODE_READING_OK:
return 0;
case TEST_CODE_READING_NO_VMLINUX:
pr_debug("no vmlinux\n");
return 0;
case TEST_CODE_READING_NO_KCORE:
pr_debug("no kcore\n");
return 0;
case TEST_CODE_READING_NO_ACCESS:
pr_debug("no access\n");
return 0;
case TEST_CODE_READING_NO_KERNEL_OBJ:
pr_debug("no kernel obj\n");
return 0;
default:
return -1;
};
}
