struct numa_topology *numa_topology__new(void)
{
struct cpu_map *node_map = NULL;
struct numa_topology *tp = NULL;
char path[MAXPATHLEN];
char *buf = NULL;
size_t len = 0;
u32 nr, i;
FILE *fp;
char *c;
scnprintf(path, MAXPATHLEN, NODE_ONLINE_FMT,
sysfs__mountpoint());
fp = fopen(path, "r");
if (!fp)
return NULL;
if (getline(&buf, &len, fp) <= 0)
goto out;
c = strchr(buf, '\n');
if (c)
*c = '\0';
node_map = cpu_map__new(buf);
if (!node_map)
goto out;
nr = (u32) node_map->nr;
tp = zalloc(sizeof(*tp) + sizeof(tp->nodes[0])*nr);
if (!tp)
goto out;
tp->nr = nr;
for (i = 0; i < nr; i++) {
if (load_numa_node(&tp->nodes[i], node_map->map[i])) {
numa_topology__delete(tp);
tp = NULL;
break;
}
}
out:
free(buf);
fclose(fp);
cpu_map__put(node_map);
return tp;
}
struct cpu_topology *cpu_topology__new(void)
{
struct cpu_topology *tp = NULL;
void *addr;
u32 nr, i;
size_t sz;
long ncpus;
int ret = -1;
struct cpu_map *map;
ncpus = cpu__max_present_cpu();
/* build online CPU map */
map = cpu_map__new(NULL);
if (map == NULL) {
pr_debug("failed to get system cpumap\n");
return NULL;
}
nr = (u32)(ncpus & UINT_MAX);
sz = nr * sizeof(char *);
addr = calloc(1, sizeof(*tp) + 2 * sz);
if (!addr)
goto out_free;
tp = addr;
addr += sizeof(*tp);
tp->core_siblings = addr;
addr += sz;
tp->thread_siblings = addr;
for (i = 0; i < nr; i++) {
if (!cpu_map__has(map, i))
continue;
ret = build_cpu_topology(tp, i);
if (ret < 0)
break;
}
out_free:
cpu_map__put(map);
if (ret) {
cpu_topology__delete(tp);
tp = NULL;
}
return tp;
}
/**
* 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;
}
/*
* 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", "getpgid", };
pid_t (*syscalls[])(void) = { (void *)getsid, getppid, (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;
char sbuf[STRERR_BUFSIZE];
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_r(errno, sbuf, sizeof(sbuf)));
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 (IS_ERR(evsels[i])) {
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_r(errno, sbuf, sizeof(sbuf)));
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_r(errno, sbuf, sizeof(sbuf)));
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__put(cpus);
out_free_threads:
thread_map__put(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;
}
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;
};
}
