@@ -127,22 +127,19 @@ void igt_panfrost_bo_mmap(int fd, struct panfrost_bo *bo)
igt_assert(bo->map);
}
-/* TODO: Make this more generic so it works on GPUs other than T760 */
struct panfrost_submit *igt_panfrost_trivial_job(int fd, bool do_crash, int width, int height, uint32_t color)
{
struct panfrost_submit *submit;
struct mali_job_descriptor_header header = {
.job_type = JOB_TYPE_FRAGMENT,
.job_index = 1,
-#ifdef __LP64__
.job_descriptor_size = 1,
-#endif
};
struct mali_payload_fragment payload = {
.min_tile_coord = MALI_COORDINATE_TO_TILE_MIN(0, 0),
.max_tile_coord = MALI_COORDINATE_TO_TILE_MAX(ALIGN(width, 16), height),
};
- struct bifrost_framebuffer framebuffer = {
+ struct bifrost_framebuffer mfbd_framebuffer = {
.unk0 = 0x0,
.unknown1 = 0x0,
.tiler_meta = 0xff00000000,
@@ -159,6 +156,20 @@ struct panfrost_submit *igt_panfrost_trivial_job(int fd, bool do_crash, int widt
.clear_depth = 0.000000,
.unknown2 = 0x1f,
};
+ struct mali_single_framebuffer sfbd_framebuffer = {
+ .unknown2 = 0x1f,
+ .width = MALI_POSITIVE(width),
+ .height = MALI_POSITIVE(height),
+ .stride = width * 4,
+ .resolution_check = ((width + height) / 3) << 4,
+ .tiler_flags = 0xfff,
+ .clear_color_1 = color,
+ .clear_color_2 = color,
+ .clear_color_3 = color,
+ .clear_color_4 = color,
+ .clear_flags = 0x101100 | MALI_CLEAR_SLOW,
+ .format = 0xb84e0281,
+ };
struct mali_rt_format fmt = {
.unk1 = 0x4000000,
.unk2 = 0x1,
@@ -179,31 +190,53 @@ struct panfrost_submit *igt_panfrost_trivial_job(int fd, bool do_crash, int widt
.clear_color_3 = color,
.clear_color_4 = color,
};
+ int gpu_prod_id = igt_panfrost_get_param(fd, DRM_PANFROST_PARAM_GPU_PROD_ID);
+ uint32_t *known_unknown;
uint32_t *bos;
submit = malloc(sizeof(*submit));
submit->fbo = igt_panfrost_gem_new(fd, ALIGN(width, 16) * height * 4);
rts.framebuffer = submit->fbo->offset;
+ sfbd_framebuffer.framebuffer = submit->fbo->offset;
- submit->tiler_heap_bo = igt_panfrost_gem_new(fd, 32768 * 4096);
- framebuffer.tiler_heap_start = submit->tiler_heap_bo->offset;
- framebuffer.tiler_heap_end = submit->tiler_heap_bo->offset + 32768 * 4096;
+ submit->tiler_heap_bo = igt_panfrost_gem_new(fd, 32768 * 128);
+ mfbd_framebuffer.tiler_heap_start = submit->tiler_heap_bo->offset;
+ mfbd_framebuffer.tiler_heap_end = submit->tiler_heap_bo->offset + 32768 * 128;
+ sfbd_framebuffer.tiler_heap_free = mfbd_framebuffer.tiler_heap_start;
+ sfbd_framebuffer.tiler_heap_end = mfbd_framebuffer.tiler_heap_end;
- submit->tiler_scratch_bo = igt_panfrost_gem_new(fd, 128 * 128 * 4096);
- framebuffer.tiler_scratch_start = submit->tiler_scratch_bo->offset;
- framebuffer.tiler_scratch_middle = submit->tiler_scratch_bo->offset + 0xf0000;
+ submit->tiler_scratch_bo = igt_panfrost_gem_new(fd, 128 * 128 * 128);
+ mfbd_framebuffer.tiler_scratch_start = submit->tiler_scratch_bo->offset;
+ mfbd_framebuffer.tiler_scratch_middle = submit->tiler_scratch_bo->offset + 0xf0000;
+ sfbd_framebuffer.unknown_address_0 = mfbd_framebuffer.tiler_scratch_start;
submit->scratchpad_bo = igt_panfrost_gem_new(fd, 64 * 4096);
- framebuffer.scratchpad = submit->scratchpad_bo->offset;
-
- submit->fb_bo = igt_panfrost_gem_new(fd, sizeof(framebuffer) + sizeof(struct bifrost_render_target));
- igt_panfrost_bo_mmap(fd, submit->fb_bo);
- memcpy(submit->fb_bo->map, &framebuffer, sizeof(framebuffer));
- memcpy(submit->fb_bo->map + sizeof(framebuffer), &rts, sizeof(struct bifrost_render_target));
- payload.framebuffer = submit->fb_bo->offset | MALI_MFBD;
-
- submit->submit_bo = igt_panfrost_gem_new(fd, sizeof(header) + sizeof(payload));
+ igt_panfrost_bo_mmap(fd, submit->scratchpad_bo);
+ mfbd_framebuffer.scratchpad = submit->scratchpad_bo->offset;
+ sfbd_framebuffer.unknown_address_1 = submit->scratchpad_bo->offset;
+ sfbd_framebuffer.unknown_address_2 = submit->scratchpad_bo->offset + 512;
+
+ known_unknown = ((void*)submit->scratchpad_bo->map) + 512;
+ *known_unknown = 0xa0000000;
+
+ if (gpu_prod_id >= 0x0750) {
+ submit->fb_bo = igt_panfrost_gem_new(fd, sizeof(mfbd_framebuffer) + sizeof(struct bifrost_render_target));
+ igt_panfrost_bo_mmap(fd, submit->fb_bo);
+ memcpy(submit->fb_bo->map, &mfbd_framebuffer, sizeof(mfbd_framebuffer));
+ memcpy(submit->fb_bo->map + sizeof(mfbd_framebuffer), &rts, sizeof(struct bifrost_render_target));
+ payload.framebuffer = submit->fb_bo->offset | MALI_MFBD;
+ } else {
+ // We don't know yet how to cause a hang on <=T720
+ // Should probably use an infinite loop to hang the GPU
+ igt_require(!do_crash);
+ submit->fb_bo = igt_panfrost_gem_new(fd, sizeof(sfbd_framebuffer));
+ igt_panfrost_bo_mmap(fd, submit->fb_bo);
+ memcpy(submit->fb_bo->map, &sfbd_framebuffer, sizeof(sfbd_framebuffer));
+ payload.framebuffer = submit->fb_bo->offset | MALI_SFBD;
+ }
+
+ submit->submit_bo = igt_panfrost_gem_new(fd, sizeof(header) + sizeof(payload) + 1024000);
igt_panfrost_bo_mmap(fd, submit->submit_bo);
memcpy(submit->submit_bo->map, &header, sizeof(header));
@@ -29,7 +29,7 @@
struct panfrost_bo {
int handle;
- uint32_t offset;
+ uint64_t offset;
uint32_t size;
void *map;
};
T720 GPUs and older don't have support for MFBD, so use the SFBD structs instead. We don't know yet how to hang the GPU with SFBD descriptors, so for now skip that test. Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com> --- lib/igt_panfrost.c | 71 +++++++++++++++++++++++++++++++++------------- lib/igt_panfrost.h | 2 +- 2 files changed, 53 insertions(+), 20 deletions(-)