Module: Mesa Branch: main Commit: f1e45f4bbdd8218af8aa8b25d71e51630ad23f19 URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=f1e45f4bbdd8218af8aa8b25d71e51630ad23f19
Author: Karmjit Mahil <[email protected]> Date: Mon Jun 26 13:09:53 2023 +0100 pvr: Use the SPM EOT on barrier stores Signed-off-by: Karmjit Mahil <[email protected]> Reviewed-by: Frank Binns <[email protected]> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24138> --- src/imagination/vulkan/pvr_cmd_buffer.c | 50 ++++++++++++++++++++++----------- src/imagination/vulkan/pvr_private.h | 4 +++ 2 files changed, 37 insertions(+), 17 deletions(-) diff --git a/src/imagination/vulkan/pvr_cmd_buffer.c b/src/imagination/vulkan/pvr_cmd_buffer.c index 71b7fa7c5b1..1cfe83ec463 100644 --- a/src/imagination/vulkan/pvr_cmd_buffer.c +++ b/src/imagination/vulkan/pvr_cmd_buffer.c @@ -1387,30 +1387,50 @@ static VkResult pvr_sub_cmd_gfx_job_init(const struct pvr_device_info *dev_info, struct pvr_framebuffer *framebuffer = render_pass_info->framebuffer; struct pvr_spm_bgobj_state *spm_bgobj_state = &framebuffer->spm_bgobj_state_per_render[sub_cmd->hw_render_idx]; - struct pvr_emit_state emit_state = { 0 }; struct pvr_render_target *render_target; VkResult result; - pvr_setup_emit_state(dev_info, hw_render, render_pass_info, &emit_state); + if (sub_cmd->barrier_store) { + /* There can only ever be one frag job running on the hardware at any one + * time, and a context switch is not allowed mid-tile, so instead of + * allocating a new scratch buffer we can reuse the SPM scratch buffer to + * perform the store. + * So use the SPM EOT program with the SPM PBE reg words in order to store + * the render to the SPM scratch buffer. + */ - memcpy(job->pbe_reg_words, - emit_state.pbe_reg_words, - sizeof(emit_state.pbe_reg_words)); + memcpy(job->pbe_reg_words, + &framebuffer->spm_eot_state_per_render[0].pbe_reg_words, + sizeof(job->pbe_reg_words)); + job->pds_pixel_event_data_offset = + framebuffer->spm_eot_state_per_render[0] + .pixel_event_program_data_offset; + } else { + struct pvr_emit_state emit_state = { 0 }; - result = pvr_sub_cmd_gfx_per_job_fragment_programs_create_and_upload( - cmd_buffer, - emit_state.emit_count, - emit_state.pbe_cs_words[0], - &pds_pixel_event_program); - if (result != VK_SUCCESS) - return result; + pvr_setup_emit_state(dev_info, hw_render, render_pass_info, &emit_state); - job->pds_pixel_event_data_offset = pds_pixel_event_program.data_offset; + memcpy(job->pbe_reg_words, + emit_state.pbe_reg_words, + sizeof(job->pbe_reg_words)); + + result = pvr_sub_cmd_gfx_per_job_fragment_programs_create_and_upload( + cmd_buffer, + emit_state.emit_count, + emit_state.pbe_cs_words[0], + &pds_pixel_event_program); + if (result != VK_SUCCESS) + return result; + + job->pds_pixel_event_data_offset = pds_pixel_event_program.data_offset; + } if (sub_cmd->barrier_load) { job->enable_bg_tag = true; job->process_empty_tiles = true; + /* Load the previously stored render from the SPM scratch buffer. */ + STATIC_ASSERT(ARRAY_SIZE(job->pds_bgnd_reg_values) == ARRAY_SIZE(spm_bgobj_state->pds_reg_values)); typed_memcpy(job->pds_bgnd_reg_values, @@ -1683,10 +1703,6 @@ static VkResult pvr_sub_cmd_gfx_job_init(const struct pvr_device_info *dev_info, * case SPM is hit so set the flag unconditionally. */ job->requires_spm_scratch_buffer = true; - /* FIXME: We should be using the SPM PBE reg words and the SPM EOT PDS data - * section instead of the regular, but for now we don't have an actual - * SPM EOT USC shader so that would cause problems. - */ return VK_SUCCESS; } diff --git a/src/imagination/vulkan/pvr_private.h b/src/imagination/vulkan/pvr_private.h index 37a84c897ff..00be6f8438d 100644 --- a/src/imagination/vulkan/pvr_private.h +++ b/src/imagination/vulkan/pvr_private.h @@ -459,7 +459,11 @@ struct pvr_sub_cmd_gfx { bool modifies_depth; bool modifies_stencil; + /* Store the render to a scratch buffer. */ bool barrier_store; + /* Load the render (stored with a `barrier_store`) as a background to the + * current render. + */ bool barrier_load; const struct pvr_query_pool *query_pool;
