fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-05-08 22:08:26 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

v3dv/pipeline: don't clone the nir shader at pipeline_state_create_binning

Browse source 
At that point we didn't call all the v3dv lowerings. So the reference
nir shader used to call the v3d compiler could be different.

Note that at that point the nir shader is only available for internal
shaders (like gs multiview).

This specifically affected multiview tests that wrote gl_PointSize, as
the nir shader for the geometry shader were wrongly exposing
per_vertex_point_size as false, as we were basing our check on the
nir_shader_info, and that was gathered calling nir_shader_gather_info
at pipeline_lower_nir.

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13325>
(cherry picked from commit 19894bec1f)
This commit is contained in:
Alejandro Piñeiro 2021-10-12 23:33:43 +02:00 • committed by Eric Engestrom
parent 3fc4ccde2d
commit 4286b1ebed
3 changed files with 6 additions and 29 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
.pick_status.json
View file

@ -7555,7 +7555,7 @@
"description": "v3dv/pipeline: don't clone the nir shader at pipeline_state_create_binning",
"nominated": false,
"nomination_type": null,
"resolution": 4,
"resolution": 1,
"main_sha": null,
"because_sha": null
},

27
src/broadcom/ci/deqp-v3dv-rpi4-fails.txt
View file

@ -3,30 +3,3 @@ dEQP-VK.geometry.input.basic_primitive.line_strip_adjacency,Fail
# CTS bug; fix submitted
dEQP-VK.spirv_assembly.instruction.graphics.variable_pointers.graphics.writes_single_buffer_geom,Fail
# Multiview doesn't work with points
dEQP-VK.multiview.point_size.15,Fail
dEQP-VK.multiview.point_size.8,Fail
dEQP-VK.multiview.point_size.1_2_4_8,Fail
dEQP-VK.multiview.point_size.15_15_15_15,Fail
dEQP-VK.multiview.point_size.8_1_1_8,Fail
dEQP-VK.multiview.point_size.5_10_5_10,Fail
dEQP-VK.multiview.point_size.1_2_4_8_16_32,Fail
dEQP-VK.multiview.point_size.max_multi_view_view_count,Fail
dEQP-VK.draw.instanced.draw_vk_primitive_topology_point_list_attrib_divisor_1_multiview,Fail
dEQP-VK.draw.instanced.draw_vk_primitive_topology_point_list_attrib_divisor_2_multiview,Fail
dEQP-VK.draw.instanced.draw_vk_primitive_topology_point_list_attrib_divisor_4_multiview,Fail
dEQP-VK.draw.instanced.draw_vk_primitive_topology_point_list_attrib_divisor_20_multiview,Fail
dEQP-VK.draw.instanced.draw_indexed_vk_primitive_topology_point_list_attrib_divisor_1_multiview,Fail
dEQP-VK.draw.instanced.draw_indexed_vk_primitive_topology_point_list_attrib_divisor_2_multiview,Fail
dEQP-VK.draw.instanced.draw_indexed_vk_primitive_topology_point_list_attrib_divisor_4_multiview,Fail
dEQP-VK.draw.instanced.draw_indexed_vk_primitive_topology_point_list_attrib_divisor_20_multiview,Fail
dEQP-VK.draw.instanced.draw_indirect_vk_primitive_topology_point_list_attrib_divisor_1_multiview,Fail
dEQP-VK.draw.instanced.draw_indirect_vk_primitive_topology_point_list_attrib_divisor_2_multiview,Fail
dEQP-VK.draw.instanced.draw_indirect_vk_primitive_topology_point_list_attrib_divisor_4_multiview,Fail
dEQP-VK.draw.instanced.draw_indirect_vk_primitive_topology_point_list_attrib_divisor_20_multiview,Fail
dEQP-VK.draw.instanced.draw_indexed_indirect_vk_primitive_topology_point_list_attrib_divisor_1_multiview,Fail
dEQP-VK.draw.instanced.draw_indexed_indirect_vk_primitive_topology_point_list_attrib_divisor_2_multiview,Fail
dEQP-VK.draw.instanced.draw_indexed_indirect_vk_primitive_topology_point_list_attrib_divisor_4_multiview,Fail
dEQP-VK.draw.instanced.draw_indexed_indirect_vk_primitive_topology_point_list_attrib_divisor_20_multiview,Fail

6
src/broadcom/vulkan/v3dv_pipeline.c
View file

@ -1412,7 +1412,11 @@ pipeline_stage_create_binning(const struct v3dv_pipeline_stage *src,
p_stage->stage = bin_stage;
p_stage->entrypoint = src->entrypoint;
p_stage->module = src->module;
p_stage->nir = src->nir ? nir_shader_clone(NULL, src->nir) : NULL;
/* For binning shaders we will clone the NIR code from the corresponding
* render shader later, when we call pipeline_compile_xxx_shader. This way
* we only have to run the relevant NIR lowerings once for render shaders
*/
p_stage->nir = NULL;
p_stage->spec_info = src->spec_info;
p_stage->feedback = (VkPipelineCreationFeedbackEXT) { 0 };
memcpy(p_stage->shader_sha1, src->shader_sha1, 20);