static timing_t
do_benchmark (size_t num_threads, size_t *iters)
{
timing_t elapsed = 0;
if (num_threads == 1)
{
timing_t start, stop;
void *working_set[WORKING_SET_SIZE];
memset (working_set, 0, sizeof (working_set));
TIMING_NOW (start);
*iters = malloc_benchmark_loop (working_set);
TIMING_NOW (stop);
TIMING_DIFF (elapsed, start, stop);
}
else
{
struct thread_args args[num_threads];
void *working_set[num_threads][WORKING_SET_SIZE];
pthread_t threads[num_threads];
memset (working_set, 0, sizeof (working_set));
*iters = 0;
for (size_t i = 0; i < num_threads; i++)
{
args[i].working_set = working_set[i];
pthread_create(&threads[i], NULL, benchmark_thread, &args[i]);
}
for (size_t i = 0; i < num_threads; i++)
{
pthread_join(threads[i], NULL);
TIMING_ACCUM (elapsed, args[i].elapsed);
*iters += args[i].iters;
}
}
return elapsed;
}
int
main (int argc, char **argv)
{
unsigned long i, k;
struct timespec runtime;
timing_t start, end;
startup();
memset (&runtime, 0, sizeof (runtime));
unsigned long iters, res;
TIMING_INIT (res);
iters = 1000 * res;
for (int v = 0; v < NUM_VARIANTS; v++)
{
/* Run for approximately DURATION seconds. */
clock_gettime (CLOCK_MONOTONIC_RAW, &runtime);
runtime.tv_sec += DURATION;
double d_total_i = 0;
timing_t total = 0, max = 0, min = 0x7fffffffffffffff;
while (1)
{
for (i = 0; i < NUM_SAMPLES (v); i++)
{
uint64_t cur;
TIMING_NOW (start);
for (k = 0; k < iters; k++)
BENCH_FUNC (v, i);
TIMING_NOW (end);
TIMING_DIFF (cur, start, end);
if (cur > max)
max = cur;
if (cur < min)
min = cur;
TIMING_ACCUM (total, cur);
d_total_i += iters;
}
struct timespec curtime;
memset (&curtime, 0, sizeof (curtime));
clock_gettime (CLOCK_MONOTONIC_RAW, &curtime);
if (TIMESPEC_AFTER (curtime, runtime))
goto done;
}
double d_total_s;
double d_iters;
done:
d_total_s = total;
d_iters = iters;
TIMING_PRINT_STATS (VARIANT (v), d_total_s, d_iters, d_total_i, max,
min);
}
return 0;
}
int
main (int argc, char **argv)
{
unsigned long i, k;
struct timespec runtime;
timing_t start, end;
bool detailed = false;
json_ctx_t json_ctx;
if (argc == 2 && !strcmp (argv[1], "-d"))
detailed = true;
bench_start ();
memset (&runtime, 0, sizeof (runtime));
unsigned long iters, res;
#ifdef BENCH_INIT
BENCH_INIT ();
#endif
TIMING_INIT (res);
iters = 1000 * res;
json_init (&json_ctx, 2, stdout);
/* Begin function. */
json_attr_object_begin (&json_ctx, FUNCNAME);
for (int v = 0; v < NUM_VARIANTS; v++)
{
/* Run for approximately DURATION seconds. */
clock_gettime (CLOCK_MONOTONIC_RAW, &runtime);
runtime.tv_sec += DURATION;
double d_total_i = 0;
timing_t total = 0, max = 0, min = 0x7fffffffffffffff;
int64_t c = 0;
while (1)
{
for (i = 0; i < NUM_SAMPLES (v); i++)
{
uint64_t cur;
TIMING_NOW (start);
for (k = 0; k < iters; k++)
BENCH_FUNC (v, i);
TIMING_NOW (end);
TIMING_DIFF (cur, start, end);
if (cur > max)
max = cur;
if (cur < min)
min = cur;
TIMING_ACCUM (total, cur);
/* Accumulate timings for the value. In the end we will divide
by the total iterations. */
RESULT_ACCUM (cur, v, i, c * iters, (c + 1) * iters);
d_total_i += iters;
}
c++;
struct timespec curtime;
memset (&curtime, 0, sizeof (curtime));
clock_gettime (CLOCK_MONOTONIC_RAW, &curtime);
if (TIMESPEC_AFTER (curtime, runtime))
goto done;
}
double d_total_s;
double d_iters;
done:
d_total_s = total;
d_iters = iters;
/* Begin variant. */
json_attr_object_begin (&json_ctx, VARIANT (v));
json_attr_double (&json_ctx, "duration", d_total_s);
json_attr_double (&json_ctx, "iterations", d_total_i);
json_attr_double (&json_ctx, "max", max / d_iters);
json_attr_double (&json_ctx, "min", min / d_iters);
json_attr_double (&json_ctx, "mean", d_total_s / d_total_i);
if (detailed)
{
json_array_begin (&json_ctx, "timings");
for (int i = 0; i < NUM_SAMPLES (v); i++)
json_element_double (&json_ctx, RESULT (v, i));
json_array_end (&json_ctx);
}
/* End variant. */
json_attr_object_end (&json_ctx);
}
/* End function. */
json_attr_object_end (&json_ctx);
return 0;
}
