From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by gabe.freedesktop.org (Postfix) with ESMTPS id C31BE10E2FD for ; Tue, 6 Dec 2022 07:47:28 +0000 (UTC) From: =?UTF-8?q?Zbigniew=20Kempczy=C5=84ski?= To: igt-dev@lists.freedesktop.org Date: Tue, 6 Dec 2022 08:47:02 +0100 Message-Id: <20221206074708.30690-6-zbigniew.kempczynski@intel.com> In-Reply-To: <20221206074708.30690-1-zbigniew.kempczynski@intel.com> References: <20221206074708.30690-1-zbigniew.kempczynski@intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Subject: [igt-dev] [PATCH i-g-t v4 05/11] benchmarks/intel_upload_blit_*: Remove libdrm in upload blits List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: igt-dev-bounces@lists.freedesktop.org Sender: "igt-dev" List-ID: Few benchmarks still used libdrm so let's rewrite them to be libdrm free. I tried to mimic the libdrm behavior as much as possible but according how libdrm caches handles/mappings there may be some differences in performance execution. v2: Move benchmarks to scope without libdrm in meson.build Signed-off-by: Zbigniew Kempczyński Cc: Petri Latvala --- benchmarks/intel_upload_blit_large.c | 131 ++++++++++++--------- benchmarks/intel_upload_blit_large_gtt.c | 130 ++++++++++++--------- benchmarks/intel_upload_blit_large_map.c | 132 ++++++++++++--------- benchmarks/intel_upload_blit_small.c | 139 ++++++++++++++--------- benchmarks/meson.build | 13 +-- 5 files changed, 315 insertions(+), 230 deletions(-) diff --git a/benchmarks/intel_upload_blit_large.c b/benchmarks/intel_upload_blit_large.c index 12bbae3d06..af52d7a4e5 100644 --- a/benchmarks/intel_upload_blit_large.c +++ b/benchmarks/intel_upload_blit_large.c @@ -45,6 +45,7 @@ */ #include "igt.h" +#include "i915/gem_create.h" #include #include #include @@ -72,74 +73,99 @@ get_time_in_secs(void) } static void -do_render(drm_intel_bufmgr *bufmgr, struct intel_batchbuffer *batch, - drm_intel_bo *dst_bo, int width, int height) +do_render(int i915, uint32_t dst_handle) { - uint32_t data[width * height]; - drm_intel_bo *src_bo; - int i; + struct drm_i915_gem_execbuffer2 exec = {}; + struct drm_i915_gem_exec_object2 obj[3] = {}; + struct drm_i915_gem_relocation_entry reloc[2]; static uint32_t seed = 1; + uint32_t data[OBJECT_WIDTH * OBJECT_HEIGHT]; + uint64_t size = OBJECT_WIDTH * OBJECT_HEIGHT * 4, bb_size = 4096; + uint32_t src_handle, bb_handle, *bb; + uint32_t gen = intel_gen(intel_get_drm_devid(i915)); + const bool has_64b_reloc = gen >= 8; + int i; + + bb_handle = gem_create_from_pool(i915, &bb_size, REGION_SMEM); + src_handle = gem_create_from_pool(i915, &size, REGION_SMEM); - /* Generate some junk. Real workloads would be doing a lot more - * work to generate the junk. - */ - for (i = 0; i < width * height; i++) { + for (i = 0; i < OBJECT_WIDTH * OBJECT_HEIGHT; i++) data[i] = seed++; - } - /* Upload the junk. */ - src_bo = drm_intel_bo_alloc(bufmgr, "src", sizeof(data), 4096); - drm_intel_bo_subdata(src_bo, 0, sizeof(data), data); + gem_write(i915, src_handle, 0, data, sizeof(data)); /* Render the junk to the dst. */ - BLIT_COPY_BATCH_START(0); - OUT_BATCH((3 << 24) | /* 32 bits */ + bb = gem_mmap__device_coherent(i915, bb_handle, 0, bb_size, PROT_WRITE); + i = 0; + bb[i++] = XY_SRC_COPY_BLT_CMD | + XY_SRC_COPY_BLT_WRITE_ALPHA | + XY_SRC_COPY_BLT_WRITE_RGB | + (6 + 2*(gen >= 8)); + bb[i++] = (3 << 24) | /* 32 bits */ (0xcc << 16) | /* copy ROP */ - (width * 4) /* dst pitch */); - OUT_BATCH(0); /* dst x1,y1 */ - OUT_BATCH((height << 16) | width); /* dst x2,y2 */ - OUT_RELOC(dst_bo, I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER, 0); - OUT_BATCH(0); /* src x1,y1 */ - OUT_BATCH(width * 4); /* src pitch */ - OUT_RELOC(src_bo, I915_GEM_DOMAIN_RENDER, 0, 0); - ADVANCE_BATCH(); - - intel_batchbuffer_flush(batch); - - drm_intel_bo_unreference(src_bo); + (OBJECT_WIDTH * 4) /* dst pitch */; + bb[i++] = 0; /* dst x1,y1 */ + bb[i++] = (OBJECT_HEIGHT << 16) | OBJECT_WIDTH; /* dst x2,y2 */ + + obj[0].handle = dst_handle; + obj[0].offset = dst_handle * size; + reloc[0].target_handle = dst_handle; + reloc[0].presumed_offset = obj[0].offset; + reloc[0].offset = sizeof(uint32_t) * i; + reloc[0].read_domains = I915_GEM_DOMAIN_RENDER; + reloc[0].write_domain = I915_GEM_DOMAIN_RENDER; + bb[i++] = obj[0].offset; + if (has_64b_reloc) + bb[i++] = obj[0].offset >> 32; + + bb[i++] = 0; /* src x1,y1 */ + bb[i++] = OBJECT_WIDTH * 4; /* src pitch */ + + obj[1].handle = src_handle; + obj[1].offset = src_handle * size; + reloc[1].target_handle = src_handle; + reloc[1].presumed_offset = obj[1].offset; + reloc[1].offset = sizeof(uint32_t) * i; + reloc[1].read_domains = I915_GEM_DOMAIN_RENDER; + reloc[1].write_domain = 0; + bb[i++] = obj[1].offset; + if (has_64b_reloc) + bb[i++] = obj[1].offset >> 32; + + obj[2].handle = bb_handle; + obj[2].relocs_ptr = to_user_pointer(reloc); + obj[2].relocation_count = 2; + + bb[i++] = MI_BATCH_BUFFER_END; + gem_munmap(bb, bb_size); + + exec.buffers_ptr = to_user_pointer(obj); + exec.buffer_count = 3; + exec.flags = gen >= 6 ? I915_EXEC_BLT : 0 | I915_EXEC_NO_RELOC; + + gem_execbuf(i915, &exec); } int main(int argc, char **argv) { - int fd; - int object_size = OBJECT_WIDTH * OBJECT_HEIGHT * 4; double start_time, end_time; - drm_intel_bo *dst_bo; - drm_intel_bufmgr *bufmgr; - struct intel_batchbuffer *batch; - int i; - - fd = drm_open_driver(DRIVER_INTEL); + uint32_t dst_handle; + int i915, i; - bufmgr = drm_intel_bufmgr_gem_init(fd, 4096); - drm_intel_bufmgr_gem_enable_reuse(bufmgr); - - batch = intel_batchbuffer_alloc(bufmgr, intel_get_drm_devid(fd)); - - dst_bo = drm_intel_bo_alloc(bufmgr, "dst", object_size, 4096); + i915 = drm_open_driver(DRIVER_INTEL); + dst_handle = gem_create(i915, OBJECT_WIDTH * OBJECT_HEIGHT * 4); /* Prep loop to get us warmed up. */ - for (i = 0; i < 60; i++) { - do_render(bufmgr, batch, dst_bo, OBJECT_WIDTH, OBJECT_HEIGHT); - } - drm_intel_bo_wait_rendering(dst_bo); + for (i = 0; i < 60; i++) + do_render(i915, dst_handle); + gem_sync(i915, dst_handle); /* Do the actual timing. */ start_time = get_time_in_secs(); - for (i = 0; i < 200; i++) { - do_render(bufmgr, batch, dst_bo, OBJECT_WIDTH, OBJECT_HEIGHT); - } - drm_intel_bo_wait_rendering(dst_bo); + for (i = 0; i < 200; i++) + do_render(i915, dst_handle); + gem_sync(i915, dst_handle); + end_time = get_time_in_secs(); printf("%d iterations in %.03f secs: %.01f MB/sec\n", i, @@ -147,10 +173,5 @@ int main(int argc, char **argv) (double)i * OBJECT_WIDTH * OBJECT_HEIGHT * 4 / 1024.0 / 1024.0 / (end_time - start_time)); - intel_batchbuffer_free(batch); - drm_intel_bufmgr_destroy(bufmgr); - - close(fd); - - return 0; + close(i915); } diff --git a/benchmarks/intel_upload_blit_large_gtt.c b/benchmarks/intel_upload_blit_large_gtt.c index 0b704b57d4..1e991a6b24 100644 --- a/benchmarks/intel_upload_blit_large_gtt.c +++ b/benchmarks/intel_upload_blit_large_gtt.c @@ -45,6 +45,7 @@ */ #include "igt.h" +#include "i915/gem_create.h" #include #include #include @@ -69,74 +70,98 @@ get_time_in_secs(void) } static void -do_render(drm_intel_bufmgr *bufmgr, struct intel_batchbuffer *batch, - drm_intel_bo *dst_bo, int width, int height) +do_render(int i915, uint32_t dst_handle) { - uint32_t *data; - drm_intel_bo *src_bo; - int i; + struct drm_i915_gem_execbuffer2 exec = {}; + struct drm_i915_gem_exec_object2 obj[3] = {}; + struct drm_i915_gem_relocation_entry reloc[2]; static uint32_t seed = 1; + uint64_t size = OBJECT_WIDTH * OBJECT_HEIGHT * 4, bb_size = 4096; + uint32_t *data, src_handle, bb_handle, *bb; + uint32_t gen = intel_gen(intel_get_drm_devid(i915)); + const bool has_64b_reloc = gen >= 8; + int i; - src_bo = drm_intel_bo_alloc(bufmgr, "src", width * height * 4, 4096); - - drm_intel_gem_bo_map_gtt(src_bo); + bb_handle = gem_create_from_pool(i915, &bb_size, REGION_SMEM); + src_handle = gem_create_from_pool(i915, &size, REGION_SMEM); - data = src_bo->virtual; - for (i = 0; i < width * height; i++) { + data = gem_mmap__gtt(i915, src_handle, size, PROT_WRITE); + for (i = 0; i < OBJECT_WIDTH * OBJECT_HEIGHT; i++) data[i] = seed++; - } - - drm_intel_gem_bo_unmap_gtt(src_bo); + gem_munmap(data, size); /* Render the junk to the dst. */ - BLIT_COPY_BATCH_START(0); - OUT_BATCH((3 << 24) | /* 32 bits */ + bb = gem_mmap__device_coherent(i915, bb_handle, 0, bb_size, PROT_WRITE); + i = 0; + bb[i++] = XY_SRC_COPY_BLT_CMD | + XY_SRC_COPY_BLT_WRITE_ALPHA | + XY_SRC_COPY_BLT_WRITE_RGB | + (6 + 2*(gen >= 8)); + bb[i++] = (3 << 24) | /* 32 bits */ (0xcc << 16) | /* copy ROP */ - (width * 4) /* dst pitch */); - OUT_BATCH(0); /* dst x1,y1 */ - OUT_BATCH((height << 16) | width); /* dst x2,y2 */ - OUT_RELOC(dst_bo, I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER, 0); - OUT_BATCH(0); /* src x1,y1 */ - OUT_BATCH(width * 4); /* src pitch */ - OUT_RELOC(src_bo, I915_GEM_DOMAIN_RENDER, 0, 0); - ADVANCE_BATCH(); - - intel_batchbuffer_flush(batch); - - drm_intel_bo_unreference(src_bo); + (OBJECT_WIDTH * 4) /* dst pitch */; + bb[i++] = 0; /* dst x1,y1 */ + bb[i++] = (OBJECT_HEIGHT << 16) | OBJECT_WIDTH; /* dst x2,y2 */ + + obj[0].handle = dst_handle; + obj[0].offset = dst_handle * size; + reloc[0].target_handle = dst_handle; + reloc[0].presumed_offset = obj[0].offset; + reloc[0].offset = sizeof(uint32_t) * i; + reloc[0].read_domains = I915_GEM_DOMAIN_RENDER; + reloc[0].write_domain = I915_GEM_DOMAIN_RENDER; + bb[i++] = obj[0].offset; + if (has_64b_reloc) + bb[i++] = obj[0].offset >> 32; + + bb[i++] = 0; /* src x1,y1 */ + bb[i++] = OBJECT_WIDTH * 4; /* src pitch */ + + obj[1].handle = src_handle; + obj[1].offset = src_handle * size; + reloc[1].target_handle = src_handle; + reloc[1].presumed_offset = obj[1].offset; + reloc[1].offset = sizeof(uint32_t) * i; + reloc[1].read_domains = I915_GEM_DOMAIN_RENDER; + reloc[1].write_domain = 0; + bb[i++] = obj[1].offset; + if (has_64b_reloc) + bb[i++] = obj[1].offset >> 32; + + obj[2].handle = bb_handle; + obj[2].relocs_ptr = to_user_pointer(reloc); + obj[2].relocation_count = 2; + + bb[i++] = MI_BATCH_BUFFER_END; + gem_munmap(bb, bb_size); + + exec.buffers_ptr = to_user_pointer(obj); + exec.buffer_count = 3; + exec.flags = gen >= 6 ? I915_EXEC_BLT : 0 | I915_EXEC_NO_RELOC; + + gem_execbuf(i915, &exec); } int main(int argc, char **argv) { - int fd; - int object_size = OBJECT_WIDTH * OBJECT_HEIGHT * 4; double start_time, end_time; - drm_intel_bo *dst_bo; - drm_intel_bufmgr *bufmgr; - struct intel_batchbuffer *batch; - int i; - - fd = drm_open_driver(DRIVER_INTEL); + uint32_t dst_handle; + int i915, i; - bufmgr = drm_intel_bufmgr_gem_init(fd, 4096); - drm_intel_bufmgr_gem_enable_reuse(bufmgr); - - batch = intel_batchbuffer_alloc(bufmgr, intel_get_drm_devid(fd)); - - dst_bo = drm_intel_bo_alloc(bufmgr, "dst", object_size, 4096); + i915 = drm_open_driver(DRIVER_INTEL); + dst_handle = gem_create(i915, OBJECT_WIDTH * OBJECT_HEIGHT * 4); /* Prep loop to get us warmed up. */ - for (i = 0; i < 60; i++) { - do_render(bufmgr, batch, dst_bo, OBJECT_WIDTH, OBJECT_HEIGHT); - } - drm_intel_bo_wait_rendering(dst_bo); + for (i = 0; i < 60; i++) + do_render(i915, dst_handle); + gem_sync(i915, dst_handle); /* Do the actual timing. */ start_time = get_time_in_secs(); - for (i = 0; i < 200; i++) { - do_render(bufmgr, batch, dst_bo, OBJECT_WIDTH, OBJECT_HEIGHT); - } - drm_intel_bo_wait_rendering(dst_bo); + for (i = 0; i < 200; i++) + do_render(i915, dst_handle); + gem_sync(i915, dst_handle); + end_time = get_time_in_secs(); printf("%d iterations in %.03f secs: %.01f MB/sec\n", i, @@ -144,10 +169,5 @@ int main(int argc, char **argv) (double)i * OBJECT_WIDTH * OBJECT_HEIGHT * 4 / 1024.0 / 1024.0 / (end_time - start_time)); - intel_batchbuffer_free(batch); - drm_intel_bufmgr_destroy(bufmgr); - - close(fd); - - return 0; + close(i915); } diff --git a/benchmarks/intel_upload_blit_large_map.c b/benchmarks/intel_upload_blit_large_map.c index ae05434f26..6d3cd748cf 100644 --- a/benchmarks/intel_upload_blit_large_map.c +++ b/benchmarks/intel_upload_blit_large_map.c @@ -48,6 +48,7 @@ */ #include "igt.h" +#include "i915/gem_create.h" #include #include #include @@ -72,74 +73,99 @@ get_time_in_secs(void) } static void -do_render(drm_intel_bufmgr *bufmgr, struct intel_batchbuffer *batch, - drm_intel_bo *dst_bo, int width, int height) +do_render(int i915, uint32_t dst_handle) { - uint32_t *data; - drm_intel_bo *src_bo; - int i; + struct drm_i915_gem_execbuffer2 exec = {}; + struct drm_i915_gem_exec_object2 obj[3] = {}; + struct drm_i915_gem_relocation_entry reloc[2]; static uint32_t seed = 1; + uint64_t size = OBJECT_WIDTH * OBJECT_HEIGHT * 4, bb_size = 4096; + uint32_t *data, src_handle, bb_handle, *bb; + uint32_t gen = intel_gen(intel_get_drm_devid(i915)); + const bool has_64b_reloc = gen >= 8; + int i; - src_bo = drm_intel_bo_alloc(bufmgr, "src", width * height * 4, 4096); - - drm_intel_bo_map(src_bo, 1); + bb_handle = gem_create_from_pool(i915, &bb_size, REGION_SMEM); + src_handle = gem_create_from_pool(i915, &size, REGION_SMEM); - data = src_bo->virtual; - for (i = 0; i < width * height; i++) { + data = gem_mmap__cpu(i915, src_handle, 0, size, PROT_WRITE); + for (i = 0; i < OBJECT_WIDTH * OBJECT_HEIGHT; i++) data[i] = seed++; - } - - drm_intel_bo_unmap(src_bo); + gem_set_domain(i915, src_handle, I915_GEM_DOMAIN_CPU, 0); + gem_munmap(data, size); /* Render the junk to the dst. */ - BLIT_COPY_BATCH_START(0); - OUT_BATCH((3 << 24) | /* 32 bits */ + bb = gem_mmap__device_coherent(i915, bb_handle, 0, bb_size, PROT_WRITE); + i = 0; + bb[i++] = XY_SRC_COPY_BLT_CMD | + XY_SRC_COPY_BLT_WRITE_ALPHA | + XY_SRC_COPY_BLT_WRITE_RGB | + (6 + 2*(gen >= 8)); + bb[i++] = (3 << 24) | /* 32 bits */ (0xcc << 16) | /* copy ROP */ - (width * 4) /* dst pitch */); - OUT_BATCH(0); /* dst x1,y1 */ - OUT_BATCH((height << 16) | width); /* dst x2,y2 */ - OUT_RELOC(dst_bo, I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER, 0); - OUT_BATCH(0); /* src x1,y1 */ - OUT_BATCH(width * 4); /* src pitch */ - OUT_RELOC(src_bo, I915_GEM_DOMAIN_RENDER, 0, 0); - ADVANCE_BATCH(); - - intel_batchbuffer_flush(batch); - - drm_intel_bo_unreference(src_bo); + (OBJECT_WIDTH * 4) /* dst pitch */; + bb[i++] = 0; /* dst x1,y1 */ + bb[i++] = (OBJECT_HEIGHT << 16) | OBJECT_WIDTH; /* dst x2,y2 */ + + obj[0].handle = dst_handle; + obj[0].offset = dst_handle * size; + reloc[0].target_handle = dst_handle; + reloc[0].presumed_offset = obj[0].offset; + reloc[0].offset = sizeof(uint32_t) * i; + reloc[0].read_domains = I915_GEM_DOMAIN_RENDER; + reloc[0].write_domain = I915_GEM_DOMAIN_RENDER; + bb[i++] = obj[0].offset; + if (has_64b_reloc) + bb[i++] = obj[0].offset >> 32; + + bb[i++] = 0; /* src x1,y1 */ + bb[i++] = OBJECT_WIDTH * 4; /* src pitch */ + + obj[1].handle = src_handle; + obj[1].offset = src_handle * size; + reloc[1].target_handle = src_handle; + reloc[1].presumed_offset = obj[1].offset; + reloc[1].offset = sizeof(uint32_t) * i; + reloc[1].read_domains = I915_GEM_DOMAIN_RENDER; + reloc[1].write_domain = 0; + bb[i++] = obj[1].offset; + if (has_64b_reloc) + bb[i++] = obj[1].offset >> 32; + + obj[2].handle = bb_handle; + obj[2].relocs_ptr = to_user_pointer(reloc); + obj[2].relocation_count = 2; + + bb[i++] = MI_BATCH_BUFFER_END; + gem_munmap(bb, bb_size); + + exec.buffers_ptr = to_user_pointer(obj); + exec.buffer_count = 3; + exec.flags = gen >= 6 ? I915_EXEC_BLT : 0 | I915_EXEC_NO_RELOC; + + gem_execbuf(i915, &exec); } int main(int argc, char **argv) { - int fd; - int object_size = OBJECT_WIDTH * OBJECT_HEIGHT * 4; double start_time, end_time; - drm_intel_bo *dst_bo; - drm_intel_bufmgr *bufmgr; - struct intel_batchbuffer *batch; - int i; - - fd = drm_open_driver(DRIVER_INTEL); + uint32_t dst_handle; + int i915, i; - bufmgr = drm_intel_bufmgr_gem_init(fd, 4096); - drm_intel_bufmgr_gem_enable_reuse(bufmgr); - - batch = intel_batchbuffer_alloc(bufmgr, intel_get_drm_devid(fd)); - - dst_bo = drm_intel_bo_alloc(bufmgr, "dst", object_size, 4096); + i915 = drm_open_driver(DRIVER_INTEL); + dst_handle = gem_create(i915, OBJECT_WIDTH * OBJECT_HEIGHT * 4); /* Prep loop to get us warmed up. */ - for (i = 0; i < 60; i++) { - do_render(bufmgr, batch, dst_bo, OBJECT_WIDTH, OBJECT_HEIGHT); - } - drm_intel_bo_wait_rendering(dst_bo); + for (i = 0; i < 60; i++) + do_render(i915, dst_handle); + gem_sync(i915, dst_handle); /* Do the actual timing. */ start_time = get_time_in_secs(); - for (i = 0; i < 200; i++) { - do_render(bufmgr, batch, dst_bo, OBJECT_WIDTH, OBJECT_HEIGHT); - } - drm_intel_bo_wait_rendering(dst_bo); + for (i = 0; i < 200; i++) + do_render(i915, dst_handle); + gem_sync(i915, dst_handle); + end_time = get_time_in_secs(); printf("%d iterations in %.03f secs: %.01f MB/sec\n", i, @@ -147,10 +173,6 @@ int main(int argc, char **argv) (double)i * OBJECT_WIDTH * OBJECT_HEIGHT * 4 / 1024.0 / 1024.0 / (end_time - start_time)); - intel_batchbuffer_free(batch); - drm_intel_bufmgr_destroy(bufmgr); - - close(fd); - - return 0; + close(i915); } + diff --git a/benchmarks/intel_upload_blit_small.c b/benchmarks/intel_upload_blit_small.c index 7e3346eb74..525d68e36a 100644 --- a/benchmarks/intel_upload_blit_small.c +++ b/benchmarks/intel_upload_blit_small.c @@ -41,6 +41,7 @@ */ #include "igt.h" +#include "i915/gem_create.h" #include #include #include @@ -66,21 +67,27 @@ get_time_in_secs(void) } static void -do_render(drm_intel_bufmgr *bufmgr, struct intel_batchbuffer *batch, - drm_intel_bo *dst_bo, int width, int height) +do_render(int i915, uint32_t dst_handle) { - uint32_t data[64]; - drm_intel_bo *src_bo; - int i; + struct drm_i915_gem_execbuffer2 exec = {}; + struct drm_i915_gem_exec_object2 obj[3] = {}; + struct drm_i915_gem_relocation_entry reloc[2]; static uint32_t seed = 1; + uint32_t data[OBJECT_WIDTH * OBJECT_HEIGHT]; + uint64_t size = OBJECT_WIDTH * OBJECT_HEIGHT * 4, bb_size = 4096; + uint32_t src_handle, bb_handle, *bb; + uint32_t gen = intel_gen(intel_get_drm_devid(i915)); + const bool has_64b_reloc = gen >= 8; + int i; - src_bo = drm_intel_bo_alloc(bufmgr, "src", width * height * 4, 4096); + bb_handle = gem_create_from_pool(i915, &bb_size, REGION_SMEM); + src_handle = gem_create_from_pool(i915, &size, REGION_SMEM); /* Upload some junk. Real workloads would be doing a lot more * work to generate the junk. */ - for (i = 0; i < width * height;) { - int size, j; + for (i = 0; i < OBJECT_WIDTH * OBJECT_HEIGHT; i++) { + int subsize, j; /* Choose a size from 1 to 64 dwords to upload. * Normal workloads have a distribution of sizes with a @@ -88,68 +95,92 @@ do_render(drm_intel_bufmgr *bufmgr, struct intel_batchbuffer *batch, * pile of vertices, most likely), but I'm trying to get at * the cost of the small uploads here. */ - size = random() % 64 + 1; - if (i + size > width * height) - size = width * height - i; + subsize = random() % 64 + 1; + if (i + subsize > OBJECT_WIDTH * OBJECT_HEIGHT) + subsize = OBJECT_WIDTH * OBJECT_HEIGHT - i; - for (j = 0; j < size; j++) + for (j = 0; j < subsize; j++) data[j] = seed++; /* Upload the junk. */ - drm_intel_bo_subdata(src_bo, i * 4, size * 4, data); + //drm_intel_bo_subdata(src_bo, i * 4, size * 4, data); + gem_write(i915, src_handle, i * 4, data, subsize * 4); - i += size; + i += subsize; } /* Render the junk to the dst. */ - BLIT_COPY_BATCH_START(0); - OUT_BATCH((3 << 24) | /* 32 bits */ + bb = gem_mmap__device_coherent(i915, bb_handle, 0, bb_size, PROT_WRITE); + i = 0; + bb[i++] = XY_SRC_COPY_BLT_CMD | + XY_SRC_COPY_BLT_WRITE_ALPHA | + XY_SRC_COPY_BLT_WRITE_RGB | + (6 + 2*(gen >= 8)); + bb[i++] = (3 << 24) | /* 32 bits */ (0xcc << 16) | /* copy ROP */ - (width * 4) /* dst pitch */); - OUT_BATCH(0); /* dst x1,y1 */ - OUT_BATCH((height << 16) | width); /* dst x2,y2 */ - OUT_RELOC(dst_bo, I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER, 0); - OUT_BATCH(0); /* src x1,y1 */ - OUT_BATCH(width * 4); /* src pitch */ - OUT_RELOC(src_bo, I915_GEM_DOMAIN_RENDER, 0, 0); - ADVANCE_BATCH(); - - intel_batchbuffer_flush(batch); - - drm_intel_bo_unreference(src_bo); + (OBJECT_WIDTH * 4) /* dst pitch */; + bb[i++] = 0; /* dst x1,y1 */ + bb[i++] = (OBJECT_HEIGHT << 16) | OBJECT_WIDTH; /* dst x2,y2 */ + + obj[0].handle = dst_handle; + obj[0].offset = dst_handle * size; + reloc[0].target_handle = dst_handle; + reloc[0].presumed_offset = obj[0].offset; + reloc[0].offset = sizeof(uint32_t) * i; + reloc[0].read_domains = I915_GEM_DOMAIN_RENDER; + reloc[0].write_domain = I915_GEM_DOMAIN_RENDER; + bb[i++] = obj[0].offset; + if (has_64b_reloc) + bb[i++] = obj[0].offset >> 32; + + bb[i++] = 0; /* src x1,y1 */ + bb[i++] = OBJECT_WIDTH * 4; /* src pitch */ + + obj[1].handle = src_handle; + obj[1].offset = src_handle * size; + reloc[1].target_handle = src_handle; + reloc[1].presumed_offset = obj[1].offset; + reloc[1].offset = sizeof(uint32_t) * i; + reloc[1].read_domains = I915_GEM_DOMAIN_RENDER; + reloc[1].write_domain = 0; + bb[i++] = obj[1].offset; + if (has_64b_reloc) + bb[i++] = obj[1].offset >> 32; + + obj[2].handle = bb_handle; + obj[2].relocs_ptr = to_user_pointer(reloc); + obj[2].relocation_count = 2; + + bb[i++] = MI_BATCH_BUFFER_END; + gem_munmap(bb, bb_size); + + exec.buffers_ptr = to_user_pointer(obj); + exec.buffer_count = 3; + exec.flags = gen >= 6 ? I915_EXEC_BLT : 0 | I915_EXEC_NO_RELOC; + + gem_execbuf(i915, &exec); } int main(int argc, char **argv) { - int fd; - int object_size = OBJECT_WIDTH * OBJECT_HEIGHT * 4; double start_time, end_time; - drm_intel_bo *dst_bo; - drm_intel_bufmgr *bufmgr; - struct intel_batchbuffer *batch; - int i; - - fd = drm_open_driver(DRIVER_INTEL); - - bufmgr = drm_intel_bufmgr_gem_init(fd, 4096); - drm_intel_bufmgr_gem_enable_reuse(bufmgr); - - batch = intel_batchbuffer_alloc(bufmgr, intel_get_drm_devid(fd)); + uint32_t dst_handle; + int i915, i; - dst_bo = drm_intel_bo_alloc(bufmgr, "dst", object_size, 4096); + i915 = drm_open_driver(DRIVER_INTEL); + dst_handle = gem_create(i915, OBJECT_WIDTH * OBJECT_HEIGHT * 4); /* Prep loop to get us warmed up. */ - for (i = 0; i < 20; i++) { - do_render(bufmgr, batch, dst_bo, OBJECT_WIDTH, OBJECT_HEIGHT); - } - drm_intel_bo_wait_rendering(dst_bo); + for (i = 0; i < 60; i++) + do_render(i915, dst_handle); + gem_sync(i915, dst_handle); /* Do the actual timing. */ start_time = get_time_in_secs(); - for (i = 0; i < 1000; i++) { - do_render(bufmgr, batch, dst_bo, OBJECT_WIDTH, OBJECT_HEIGHT); - } - drm_intel_bo_wait_rendering(dst_bo); + for (i = 0; i < 1000; i++) + do_render(i915, dst_handle); + gem_sync(i915, dst_handle); + end_time = get_time_in_secs(); printf("%d iterations in %.03f secs: %.01f MB/sec\n", i, @@ -157,10 +188,6 @@ int main(int argc, char **argv) (double)i * OBJECT_WIDTH * OBJECT_HEIGHT * 4 / 1024.0 / 1024.0 / (end_time - start_time)); - intel_batchbuffer_free(batch); - drm_intel_bufmgr_destroy(bufmgr); - - close(fd); - - return 0; + close(i915); } + diff --git a/benchmarks/meson.build b/benchmarks/meson.build index 98a08e25cd..c451268bc4 100644 --- a/benchmarks/meson.build +++ b/benchmarks/meson.build @@ -13,20 +13,15 @@ benchmark_progs = [ 'gem_syslatency', 'gem_userptr_benchmark', 'gem_wsim', + 'intel_upload_blit_large', + 'intel_upload_blit_large_gtt', + 'intel_upload_blit_large_map', + 'intel_upload_blit_small', 'kms_vblank', 'prime_lookup', 'vgem_mmap', ] -if libdrm_intel.found() - benchmark_progs += [ - 'intel_upload_blit_large', - 'intel_upload_blit_large_gtt', - 'intel_upload_blit_large_map', - 'intel_upload_blit_small', - ] -endif - benchmarksdir = join_paths(libexecdir, 'benchmarks') foreach prog : benchmark_progs -- 2.34.1