Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Move LowerGpuRt into LGC #2852

Merged
merged 2 commits into from
Nov 30, 2023
Merged

Conversation

LLJJDD
Copy link
Contributor

@LLJJDD LLJJDD commented Nov 29, 2023

With some tweak to adapt the change.

With some tweak to adapt the change.
@LLJJDD LLJJDD requested a review from a team as a code owner November 29, 2023 01:39
@amdvlk-admin
Copy link

Test summary for commit d15e7f2

CTS tests (Failed: 1645/138443)
  • Built with version 1.3.5.2
  • Ubuntu navi3x, Srdcvk
    • Passed: 33639/69228 (48.6%)
    • Failed: 1587/69228 (2.3%)

      Failures:

      FAILURE: dEQP-VK.api.buffer.suballocation.transfer_src.transfer_dst.uniform_texel.storage_texel.uniform.storage.create.binding_residency
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.buffer.suballocation.uniform_texel.uniform.index.create.binding
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.buffer.suballocation.uniform_texel.uniform.index.indirect.create.binding_aliased
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.copy_and_blit.copy_commands2.blit_image.all_formats.generate_mipmaps.from_previous_level.layercount_6.r8g8_snorm.general_optimal_nearest
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.copy_and_blit.copy_commands2.buffer_to_depthstencil.buffer_offset_d16_unorm_s8_uint_D_S
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.copy_and_blit.core.blit_image.simple_tests.scaling_whole2.b8g8r8a8_unorm_linear
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.copy_and_blit.core.blit_image.simple_tests.without_scaling_partial_3d.b8g8r8a8_unorm_linear
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.copy_and_blit.core.buffer_to_buffer_with_offset.6_1
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.copy_and_blit.core.depth_stencil_msaa_copy.whole.d16_unorm_s8_uint_general_general_D_8_bit_bind_offset
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.copy_and_blit.dedicated_allocation.blit_image.simple_tests.without_scaling_partial_3d.linear
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.copy_and_blit.dedicated_allocation.image_to_image.dimensions.src4x8192_dst4x8192.r4g4_unorm_pack8.r8_srgb.optimal_optimal
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.copy_and_blit.dedicated_allocation.image_to_image.dimensions.src6x16384_dst6x16384.r32_sfloat.r32_sfloat.optimal_optimal
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.copy_and_blit.dedicated_allocation.resolve_image.whole_array_image_one_region.4_bit
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.external.fence.sync_fd.export_multiple_times_temporary
      Stack trace: Fail
      
      FAILURE: dEQP-VK.api.image_clearing.core.clear_color_attachment.single_layer.r8_srgb_64x11
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.image_clearing.dedicated_allocation.clear_color_image.1d.linear.remaining_array_layers_twostep.r16g16b16a16_sfloat_200x1
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.image_clearing.dedicated_allocation.clear_color_image.1d.optimal.remaining_array_layers_twostep.r8g8b8a8_sint_200x1
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.image_clearing.dedicated_allocation.clear_color_image.2d.optimal.remaining_array_layers.r8g8_sint_64x11
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.image_clearing.dedicated_allocation.partial_clear_color_attachment.cube_layers.r16g16_sint
      Stack trace: Flake
      
      FAILURE: dEQP-VK.api.image_clearing.dedicated_allocation.partial_clear_color_attachment.single_layer.a4b4g4r4_unorm_pack16_64x11_clamp_input
      Stack trace: Flake
      ...
      

    • Not Supported: 34002/69228 (49.1%)
    • Warnings: 0/69228 (0.0%)
    Ubuntu navi2x, Srdcvk
    • Passed: 35184/69215 (50.8%)
    • Failed: 58/69215 (0.1%)

      Failures:

      FAILURE: dEQP-VK.ray_query.builtin.getintersectionfrontfaceCandidate.geom.triangles
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.builtin.getintersectionfrontfaceCommitted.frag.triangles
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.builtin.getintersectiongeometryindexCommitted.frag.aabbs
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.builtin.getintersectiongeometryindexCommitted.geom.aabbs
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.builtin.getworldrayorigin.geom.triangles
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.builtin.getworldrayorigin.vert.triangles
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.builtin.rayqueryterminate.geom.aabbs
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.coarse.cubic_quadratic.32x64.64x32
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.coarse.cubic_quadratic.32x64.64x64
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.coarse.cubic_quadratic.64x64.64x32
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.coarse.cubic_quadratic.64x64.64x64
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.coarse.linear_cubic.32x64.64x32
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.coarse.linear_cubic.32x64.64x64
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.coarse.linear_cubic.64x64.64x32
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.coarse.linear_cubic.64x64.64x64
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.coarse.linear_quadratic.32x64.64x32
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.coarse.linear_quadratic.32x64.64x64
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.coarse.linear_quadratic.64x64.64x32
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.coarse.linear_quadratic.64x64.64x64
      Stack trace: Fail
      
      FAILURE: dEQP-VK.ray_query.helper_invocations.gpu.fine.cubic_quadratic.32x64.64x32
      Stack trace: Fail
      ...
      

    • Not Supported: 33973/69215 (49.1%)
    • Warnings: 0/69215 (0.0%)

@LLJJDD
Copy link
Contributor Author

LLJJDD commented Nov 29, 2023

retest this please

@tsymalla-AMD
Copy link
Contributor

Some comments, apart from that it looks good. Are the Jenkins / CTS failures related?

The resource mapping seems to be messed up when using CreateReadBuiltInInput
out of builder replayer.
@LLJJDD
Copy link
Contributor Author

LLJJDD commented Nov 29, 2023

Some comments, apart from that it looks good. Are the Jenkins / CTS failures related?

Yes, hopefully my latest commit fixes them.

@amdvlk-admin
Copy link

Test summary for commit b8614d2

CTS tests (Failed: 0/138378)
  • Built with version 1.3.5.2
  • Ubuntu navi3x, Srdcvk
    • Passed: 35162/69163 (50.8%)
    • Failed: 0/69163 (0.0%)
    • Not Supported: 34001/69163 (49.2%)
    • Warnings: 0/69163 (0.0%)
    Ubuntu navi2x, Srdcvk
    • Passed: 35242/69215 (50.9%)
    • Failed: 0/69215 (0.0%)
    • Not Supported: 33973/69215 (49.1%)
    • Warnings: 0/69215 (0.0%)

@@ -342,6 +342,10 @@ Options PipelineContext::computePipelineOptions() const {
options.enablePrimGeneratedQuery = getPipelineOptions()->enablePrimGeneratedQuery;
options.enableFragColor = getPipelineOptions()->enableFragColor;

options.rtBoxSortHeuristicMode = m_rtState.boxSortHeuristicMode;
options.rtStaticPipelineFlags = m_rtState.staticPipelineFlags;
options.rtTriCompressMode = m_rtState.triCompressMode;
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do you need to default-initialize useGpurt here as well?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The = {} initialization at the top should take care of that, right?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes. My question is based on the fact that we initialize other members with a constant value here as well, so it is merely a question about coding style.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think we don't need to because it is always false (unknown) at here.

Copy link
Member

@nhaehnle nhaehnle left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Following yesterday's discussion, luckily @mbrkusanin is going to work on moving the descriptor load lowering out of the builder replayer. Once that's done, it should be possible to remove this workaround and move LowerGpuRt to after the BuilderReplayer and in fact after LowerRaytracingPipeline, where it should be in the long run.

Maybe you can double-check Thomas' question, but apart from that this LGTM.

@@ -342,6 +342,10 @@ Options PipelineContext::computePipelineOptions() const {
options.enablePrimGeneratedQuery = getPipelineOptions()->enablePrimGeneratedQuery;
options.enableFragColor = getPipelineOptions()->enableFragColor;

options.rtBoxSortHeuristicMode = m_rtState.boxSortHeuristicMode;
options.rtStaticPipelineFlags = m_rtState.staticPipelineFlags;
options.rtTriCompressMode = m_rtState.triCompressMode;
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The = {} initialization at the top should take care of that, right?

@LLJJDD LLJJDD merged commit 7eb4d8a into GPUOpen-Drivers:dev Nov 30, 2023
9 checks passed
@LLJJDD LLJJDD deleted the move-lower-gpurt branch November 30, 2023 06:47
qiaojbao pushed a commit that referenced this pull request Jan 2, 2024
With some tweak to adapt the change.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants