[PATCH i-g-t 2/4] benchmark: Measure allocation time for objects
Pravalika Gurram
pravalika.gurram at intel.com
Wed Apr 2 06:37:21 UTC 2025
A basic measurement, how fast can we create and populate an object with
backing storage
Signed-off-by: Pravalika Gurram <pravalika.gurram at intel.com>
---
benchmarks/meson.build | 1 +
benchmarks/xe_create.c | 233 +++++++++++++++++++++++++++++++++++++++++
2 files changed, 234 insertions(+)
create mode 100644 benchmarks/xe_create.c
diff --git a/benchmarks/meson.build b/benchmarks/meson.build
index 4421ede86..00203c62e 100644
--- a/benchmarks/meson.build
+++ b/benchmarks/meson.build
@@ -22,6 +22,7 @@ benchmark_progs = [
'prime_lookup',
'vgem_mmap',
'xe_blt',
+ 'xe_create',
]
benchmarksdir = join_paths(libexecdir, 'benchmarks')
diff --git a/benchmarks/xe_create.c b/benchmarks/xe_create.c
new file mode 100644
index 000000000..cd97d6920
--- /dev/null
+++ b/benchmarks/xe_create.c
@@ -0,0 +1,233 @@
+/*
+ * Copyright © 2025 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ * Authors:
+ * Pravalika Gurram <pravalika.gurram at intel.com>
+ *
+ */
+
+#include <unistd.h>
+#include <stdlib.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <string.h>
+#include <fcntl.h>
+#include <inttypes.h>
+#include <errno.h>
+#include <sys/stat.h>
+#include <sys/ioctl.h>
+#include <sys/time.h>
+#include <time.h>
+
+#include "drm.h"
+#include "drmtest.h"
+#include "i915/gem_create.h"
+#include "igt_aux.h"
+#include "igt_stats.h"
+#include "intel_reg.h"
+#include "ioctl_wrappers.h"
+
+#include "igt.h"
+#include "igt_core.h"
+#include "igt_syncobj.h"
+#include "intel_reg.h"
+#include "xe/xe_ioctl.h"
+#include "xe/xe_query.h"
+
+
+#define OBJECT_SIZE (1<<23)
+
+struct data {
+ uint32_t batch[16];
+ uint64_t pad;
+ uint32_t data;
+ uint64_t addr;
+};
+
+static double elapsed(const struct timespec *start,
+ const struct timespec *end)
+{
+ return (end->tv_sec - start->tv_sec) + 1e-9*(end->tv_nsec - start->tv_nsec);
+}
+
+static void store_dword_batch(struct data *data, uint64_t addr, int value)
+{
+ int b;
+ uint64_t batch_offset = (char *)&(data->batch) - (char *)data;
+ uint64_t batch_addr = addr + batch_offset;
+ uint64_t sdi_offset = (char *)&(data->data) - (char *)data;
+ uint64_t sdi_addr = addr + sdi_offset;
+
+ b = 0;
+
+ data->batch[b++] = MI_BATCH_BUFFER_END;
+ igt_assert(b <= ARRAY_SIZE(data->batch));
+
+ data->addr = batch_addr;
+}
+static void test_exec(int fd, int busy)
+{
+ uint32_t vm;
+ size_t bo_size;
+ uint32_t bo = 0;
+ struct data *data;
+ uint32_t exec_queue;
+ uint32_t syncobj;
+ int value = 0x123456;
+ uint64_t addr = 0x100000;
+
+ struct drm_xe_sync sync = {
+ .flags = DRM_XE_SYNC_TYPE_SYNCOBJ | DRM_XE_SYNC_FLAG_SIGNAL
+ };
+ struct drm_xe_exec exec = {
+ .num_batch_buffer = 1,
+ .num_syncs = 1,
+ .syncs = to_user_pointer(&sync),
+ };
+
+ struct drm_xe_engine_class_instance inst = {
+ .engine_class = DRM_XE_ENGINE_CLASS_COPY,
+ };
+
+ vm = xe_vm_create(fd, 0, 0);
+ bo_size = sizeof(*data);
+ bo_size = xe_bb_size(fd, bo_size);
+
+ bo = xe_bo_create(fd, vm, bo_size,
+ vram_if_possible(fd, 0),
+ DRM_XE_GEM_CREATE_FLAG_NEEDS_VISIBLE_VRAM);
+ if (busy) {
+ syncobj = syncobj_create(fd, 0);
+ sync.handle = syncobj;
+
+ xe_vm_bind_async(fd, vm, 0, bo, 0, addr, bo_size, &sync, 1);
+ data = xe_bo_map(fd, bo, bo_size);
+ store_dword_batch(data, addr, value);
+
+ igt_assert(syncobj_wait(fd, &syncobj, 1, INT64_MAX, 0, NULL));
+ syncobj_reset(fd, &syncobj, 1);
+
+ exec_queue = xe_exec_queue_create(fd, vm, &inst, 0);
+ exec.exec_queue_id = exec_queue;
+ exec.address = data->addr;
+ sync.flags &= DRM_XE_SYNC_FLAG_SIGNAL;
+ xe_exec(fd, &exec);
+
+ igt_assert(syncobj_wait(fd, &syncobj, 1, INT64_MAX, 0, NULL));
+
+ syncobj_destroy(fd, syncobj);
+
+ xe_exec_queue_destroy(fd, exec_queue);
+ }
+ munmap(data, bo_size);
+ gem_close(fd, bo);
+
+ xe_vm_destroy(fd, vm);
+}
+int main(int argc, char **argv)
+{
+ int fd = drm_open_driver(DRIVER_XE);
+ int size = 0;
+ int busy = 0;
+ int reps = 13;
+ int ncpus = 1;
+ int c, n, s;
+
+ while ((c = getopt (argc, argv, "bs:r:f")) != -1) {
+ switch (c) {
+ case 's':
+ size = atoi(optarg);
+ break;
+
+ case 'r':
+ reps = atoi(optarg);
+ if (reps < 1)
+ reps = 1;
+ break;
+
+ case 'f':
+ ncpus = sysconf(_SC_NPROCESSORS_ONLN);
+ break;
+
+ case 'b':
+ busy = true;
+ break;
+
+ default:
+ break;
+ }
+ }
+
+ if (size == 0) {
+ for (s = 4096; s <= OBJECT_SIZE; s <<= 1) {
+ igt_stats_t stats;
+
+ igt_stats_init_with_size(&stats, reps);
+ for (n = 0; n < reps; n++) {
+ struct timespec start, end;
+ uint64_t count = 0;
+
+ clock_gettime(CLOCK_MONOTONIC, &start);
+ do {
+ for (c = 0; c < 1000; c++)
+ test_exec(fd, busy);
+ count += c;
+ clock_gettime(CLOCK_MONOTONIC, &end);
+ } while (end.tv_sec - start.tv_sec < 2);
+
+ igt_stats_push_float(&stats, count / elapsed(&start, &end));
+ }
+ printf("%f\n", igt_stats_get_trimean(&stats));
+ igt_stats_fini(&stats);
+ }
+ } else {
+ double *shared;
+
+ shared = mmap(0, 4096, PROT_WRITE, MAP_SHARED | MAP_ANON, -1, 0);
+ for (n = 0; n < reps; n++) {
+ memset(shared, 0, 4096);
+
+ igt_fork(child, ncpus) {
+ struct timespec start, end;
+ uint64_t count = 0;
+
+ clock_gettime(CLOCK_MONOTONIC, &start);
+ do {
+ for (c = 0; c < 1000; c++)
+ test_exec(fd, busy);
+ count += c;
+ clock_gettime(CLOCK_MONOTONIC, &end);
+ } while (end.tv_sec - start.tv_sec < 2);
+
+ shared[child] = count / elapsed(&start, &end);
+ }
+ igt_waitchildren();
+
+ for (int child = 0; child < ncpus; child++)
+ shared[ncpus] += shared[child];
+
+ printf("%7.3f\n", shared[ncpus]);
+ }
+ }
+
+ return 0;
+}
--
2.34.1
More information about the igt-dev
mailing list