Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Running some tests in parallel by using pthread #26

Open
wants to merge 6 commits into
base: master
Choose a base branch
from
Open

Running some tests in parallel by using pthread #26

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
f370214
Add support for pthread
sun409-1 Dec 28, 2020
7269d31
Add test for reading only
sun409-1 Dec 28, 2020
7a7514d
Update readme.txt
sun409-1 Dec 28, 2020
7e5b555
Add tests of unaligned fetch and fill
tim609 Jul 4, 2023
b8a2264
Add tests of unaligned fetch and fill
tim609 Jul 4, 2023
f178ce7
Merge branch 'master' of https://github.com/sun409/tinymembench-pthread
tim609 Jul 4, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 4 additions & 1 deletion Makefile
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,11 +1,14 @@
LDFLAGS ?= -pthread
CFLAGS ?= -march=native

all: tinymembench

ifdef WINDIR
CC = gcc
endif

tinymembench: main.c util.o util.h asm-opt.h version.h asm-opt.o x86-sse2.o arm-neon.o mips-32.o aarch64-asm.o
${CC} -O2 ${CFLAGS} -o tinymembench main.c util.o asm-opt.o x86-sse2.o arm-neon.o mips-32.o aarch64-asm.o -lm
${CC} -O2 ${CFLAGS} ${LDFLAGS} -o tinymembench main.c util.o asm-opt.o x86-sse2.o arm-neon.o mips-32.o aarch64-asm.o -lm

util.o: util.c util.h
${CC} -O2 ${CFLAGS} -c util.c
Expand Down
221 changes: 194 additions & 27 deletions main.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,6 +21,7 @@
* DEALINGS IN THE SOFTWARE.
*/

#define _GNU_SOURCE
#include <string.h>
#include <stdint.h>
#include <stdio.h>
Expand All @@ -29,8 +30,11 @@
#include <sys/time.h>

#ifdef __linux__
#include <pthread.h>
#include <sched.h>
#include <unistd.h>
#include <fcntl.h>
#include <semaphore.h>
#include <linux/fb.h>
#include <sys/mman.h>
#include <sys/ioctl.h>
Expand Down Expand Up @@ -74,9 +78,99 @@ static void *mmap_framebuffer(size_t *fbsize)
*fbsize = finfo.smem_len;
return p;
}

struct f_data
{
void (*func)(int64_t *, int64_t *, int);
int64_t *arg1;
int64_t *arg2;
int arg3;
};
pthread_cond_t p_ready;
pthread_cond_t p_start;
pthread_mutex_t p_lock;
pthread_t *p_worker = NULL;
struct f_data *worker_data = NULL;
int p_worker_not_ready;
int p_workers_ready;

void *thread_func(void *data)
{
struct f_data *data_ptr = data;

pthread_mutex_lock(&p_lock);
--p_worker_not_ready;

if (p_worker_not_ready == 0)
{
pthread_cond_signal(&p_ready);
}
while (p_workers_ready != 1)
{
pthread_cond_wait(&p_start, &p_lock);
}
pthread_mutex_unlock(&p_lock);

(data_ptr->func)(data_ptr->arg1, data_ptr->arg2, data_ptr->arg3);

pthread_exit(NULL);
}

static void parallel_run(void)
{
pthread_mutex_lock(&p_lock);
p_workers_ready = 1;
pthread_mutex_unlock(&p_lock);
pthread_cond_broadcast(&p_start);
}

static void parallel_init(int threads)
{
int i;
pthread_attr_t attr;
cpu_set_t cpus;

pthread_cond_init(&p_ready, NULL);
pthread_cond_init(&p_start, NULL);
pthread_mutex_init(&p_lock, NULL);
p_worker_not_ready = threads;
p_workers_ready = 0;
pthread_attr_init(&attr);

if (!p_worker || !worker_data)
{
p_worker = (pthread_t *)malloc(threads * sizeof(pthread_t));
worker_data = (struct f_data *)malloc(threads * sizeof(struct f_data));
}
if (!p_worker || !worker_data)
{
printf("malloc failed\n");
return;
}

for (i = 0; i < threads; ++i)
{
#if 1
CPU_ZERO(&cpus);
CPU_SET(i, &cpus);
pthread_attr_setaffinity_np(&attr, sizeof(cpu_set_t), &cpus);
pthread_create(p_worker + i, &attr, thread_func, worker_data + i);
#else
pthread_create(p_worker + i, NULL, thread_func, worker_data + i);
#endif
}

pthread_mutex_lock(&p_lock);
while (p_worker_not_ready != 0)
{
pthread_cond_wait(&p_ready, &p_lock);
}
pthread_mutex_unlock(&p_lock);
}
#endif

static double bandwidth_bench_helper(int64_t *dstbuf, int64_t *srcbuf,
static double bandwidth_bench_helper(int threads,
int64_t *dstbuf, int64_t *srcbuf,
int64_t *tmpbuf,
int size, int blocksize,
const char *indent_prefix,
Expand All @@ -85,53 +179,90 @@ static double bandwidth_bench_helper(int64_t *dstbuf, int64_t *srcbuf,
const char *description)
{
int i, j, loopcount, innerloopcount, n;
double t1, t2;
double t, t1, t2;
double speed, maxspeed;
double s, s0, s1, s2;
int pt;

/* do up to MAXREPEATS measurements */
s = s0 = s1 = s2 = 0;
maxspeed = 0;
s = s0 = s1 = s2 = 0.0;
maxspeed = 0.0;
for (n = 0; n < MAXREPEATS; n++)
{
#if 1
parallel_init(threads);
for (pt = 0; pt < threads; ++pt)
{
(worker_data + pt)->func = f;
(worker_data + pt)->arg1 = dstbuf + size * pt / sizeof(int64_t);
(worker_data + pt)->arg2 = srcbuf + size * pt / sizeof(int64_t);
(worker_data + pt)->arg3 = size;
}
parallel_run();
for (pt = 0; pt < threads; ++pt)
pthread_join(p_worker[pt], NULL);
#else
f(dstbuf, srcbuf, size);
#endif
loopcount = 0;
innerloopcount = 1;
t1 = gettime();
t = 0.0;
do
{
loopcount += innerloopcount;
if (use_tmpbuf)
{
for (i = 0; i < innerloopcount; i++)
{
t1 = gettime();
for (j = 0; j < size; j += blocksize)
{
{
f(tmpbuf, srcbuf + j / sizeof(int64_t), blocksize);
f(dstbuf + j / sizeof(int64_t), tmpbuf, blocksize);
}
t2 = gettime();
t += t2 - t1;
}
}
else
{
for (i = 0; i < innerloopcount; i++)
{
#if 1
parallel_init(threads);
for (pt = 0; pt < threads; ++pt)
{
(worker_data + pt)->func = f;
(worker_data + pt)->arg1 = dstbuf + size * pt / sizeof(int64_t);
(worker_data + pt)->arg2 = srcbuf + size * pt / sizeof(int64_t);
(worker_data + pt)->arg3 = size;
}

t1 = gettime();
parallel_run();
for (pt = 0; pt < threads; ++pt)
pthread_join(p_worker[pt], NULL);
t2 = gettime();
#else
t1 = gettime();
f(dstbuf, srcbuf, size);
t2 = gettime();
#endif
t += t2 - t1;
}
}
innerloopcount *= 2;
t2 = gettime();
} while (t2 - t1 < 0.5);
speed = (double)size * loopcount / (t2 - t1) / 1000000.;
} while (t < 0.5);
speed = (double)size * (use_tmpbuf ? 1 : threads) * loopcount / t / 1000000.;

s0 += 1;
s0 += 1.;
s1 += speed;
s2 += speed * speed;

if (speed > maxspeed)
maxspeed = speed;

if (s0 > 2)
if (s0 > 2.)
{
s = sqrt((s0 * s2 - s1 * s1) / (s0 * (s0 - 1)));
if (s < maxspeed / 1000.)
Expand All @@ -141,13 +272,18 @@ static double bandwidth_bench_helper(int64_t *dstbuf, int64_t *srcbuf,

if (maxspeed > 0 && s / maxspeed * 100. >= 0.1)
{
printf("%s%-52s : %8.1f MB/s (%.1f%%)\n", indent_prefix, description,
maxspeed, s / maxspeed * 100.);
printf("%s%-52s : %8.1f MB/s (%.1f%%)",
indent_prefix, description, maxspeed, s / maxspeed * 100.);
}
else
{
printf("%s%-52s : %8.1f MB/s\n", indent_prefix, description, maxspeed);
printf("%s%-52s : %8.1f MB/s ", indent_prefix, description, maxspeed);
}
if (use_tmpbuf || threads == 1)
printf("\n");
else
printf(" @ %d thread%c\n", (use_tmpbuf ? 1 : threads), (use_tmpbuf || threads == 1 ? ' ' : 's'));

return maxspeed;
}

Expand All @@ -172,27 +308,40 @@ static bench_info c_benchmarks[] =
{ "C 2-pass copy", 1, aligned_block_copy },
{ "C 2-pass copy prefetched (32 bytes step)", 1, aligned_block_copy_pf32 },
{ "C 2-pass copy prefetched (64 bytes step)", 1, aligned_block_copy_pf64 },
{ "C fetch", 0, aligned_block_fetch },
{ "C fetch unaligned", 0, unaligned_block_fetch },
{ "C fill", 0, aligned_block_fill },
{ "C fill unaligned", 0, unaligned_block_fill },
{ "C fill (shuffle within 16 byte blocks)", 0, aligned_block_fill_shuffle16 },
{ "C fill (shuffle within 32 byte blocks)", 0, aligned_block_fill_shuffle32 },
{ "C fill (shuffle within 64 byte blocks)", 0, aligned_block_fill_shuffle64 },
{ NULL, 0, NULL }
};

static bench_info c_unaligned_benchmarks[] =
{
{ "C fetch unaligned", 0, unaligned_block_fetch },
{ "C fill unaligned", 0, unaligned_block_fill },
{ NULL, 0, NULL }
};

static bench_info libc_benchmarks[] =
{
{ "standard memcpy", 0, memcpy_wrapper },
{ "standard memset", 0, memset_wrapper },
{ NULL, 0, NULL }
};

void bandwidth_bench(int64_t *dstbuf, int64_t *srcbuf, int64_t *tmpbuf,
void bandwidth_bench(int threads,
int64_t *dstbuf, int64_t *srcbuf, int64_t *tmpbuf,
int size, int blocksize, const char *indent_prefix,
bench_info *bi)
{
while (bi->f)
{
bandwidth_bench_helper(dstbuf, srcbuf, tmpbuf, size, blocksize,
bandwidth_bench_helper(threads,
dstbuf, srcbuf, tmpbuf,
size, blocksize,
indent_prefix, bi->use_tmpbuf,
bi->f,
bi->description);
Expand Down Expand Up @@ -378,8 +527,8 @@ static uint32_t rand32()
int latency_bench(int size, int count, int use_hugepage)
{
double t, t2, t_before, t_after, t_noaccess, t_noaccess2;
double xs, xs0, xs1, xs2;
double ys, ys0, ys1, ys2;
double xs, xs1, xs2;
double ys, ys1, ys2;
double min_t, min_t2;
int nbits, n;
char *buffer, *buffer_alloc;
Expand Down Expand Up @@ -480,24 +629,38 @@ int latency_bench(int size, int count, int use_hugepage)
return 1;
}

int main(void)
int main(int argc, char *argv[])
{
int latbench_size = SIZE * 2, latbench_count = LATBENCH_COUNT;
int64_t *srcbuf, *dstbuf, *tmpbuf;
void *poolbuf;
size_t bufsize = SIZE;
int threads;
#ifdef __linux__
size_t fbsize = 0;
int64_t *fbbuf = mmap_framebuffer(&fbsize);
fbsize = (fbsize / BLOCKSIZE) * BLOCKSIZE;
#endif

printf("tinymembench v" VERSION " (simple benchmark for memory throughput and latency)\n");
printf("tinymembench-pthread v" VERSION " (simple benchmark for memory throughput and latency)\n");

if (argc == 1)
{
threads = 1;
printf("Single thread test\n");
}
else
{
int total_cpu;

total_cpu = sysconf(_SC_NPROCESSORS_ONLN);
threads = atoi(argv[1]);
printf("%d threads on %d CPU\n", threads, total_cpu);
}

poolbuf = alloc_four_nonaliased_buffers((void **)&srcbuf, bufsize,
(void **)&dstbuf, bufsize,
(void **)&tmpbuf, BLOCKSIZE,
poolbuf = alloc_four_nonaliased_buffers((void **)&srcbuf, bufsize * threads,
(void **)&dstbuf, bufsize * threads,
(void **)&tmpbuf, BLOCKSIZE * threads,
NULL, 0);
printf("\n");
printf("==========================================================================\n");
Expand All @@ -514,13 +677,17 @@ int main(void)
printf("== brackets ==\n");
printf("==========================================================================\n\n");

bandwidth_bench(dstbuf, srcbuf, tmpbuf, bufsize, BLOCKSIZE, " ", c_benchmarks);
#if 0
bandwidth_bench(threads, dstbuf, srcbuf, tmpbuf, bufsize, BLOCKSIZE, " ", c_benchmarks);
printf(" ---\n");
#endif
bandwidth_bench(threads, dstbuf, srcbuf, tmpbuf, bufsize, BLOCKSIZE, " ", c_unaligned_benchmarks);
printf(" ---\n");
bandwidth_bench(dstbuf, srcbuf, tmpbuf, bufsize, BLOCKSIZE, " ", libc_benchmarks);
bandwidth_bench(threads, dstbuf, srcbuf, tmpbuf, bufsize, BLOCKSIZE, " ", libc_benchmarks);
bench_info *bi = get_asm_benchmarks();
if (bi->f) {
printf(" ---\n");
bandwidth_bench(dstbuf, srcbuf, tmpbuf, bufsize, BLOCKSIZE, " ", bi);
bandwidth_bench(threads, dstbuf, srcbuf, tmpbuf, bufsize, BLOCKSIZE, " ", bi);
}

#ifdef __linux__
Expand Down Expand Up @@ -551,7 +718,7 @@ int main(void)
srcbuf = fbbuf;
if (bufsize > fbsize)
bufsize = fbsize;
bandwidth_bench(dstbuf, srcbuf, tmpbuf, bufsize, BLOCKSIZE, " ", bi);
bandwidth_bench(1, dstbuf, srcbuf, tmpbuf, bufsize, BLOCKSIZE, " ", bi);
}
#endif

Expand Down
Loading