Skip to content

[UR][L0] Refactor IL code handling allowing future extension #18441

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

Merged
merged 3 commits into from
May 20, 2025

Conversation

Maetveis
Copy link
Contributor

Refactor the code to make it easier to add support for different IL formats (besides SPIR-V) in the future.
The only functional change is that SPIR-V binaries with invalid magic number are now rejected by returning UR_RESULT_ERROR_INVALID_BINARY from urProgramCreateWithIL.

Refactor the code to make it easier to add support for different
IL formats (besides SPIR-V) in the future.
The only functional change is that SPIR-V binaries with invalid magic
number are now rejected by returning UR_RESULT_ERROR_INVALID_BINARY
from urProgramCreateWithIL.
@Maetveis Maetveis requested a review from a team as a code owner May 13, 2025 12:28
@Maetveis Maetveis temporarily deployed to WindowsCILock May 13, 2025 12:28 — with GitHub Actions Inactive
@Maetveis Maetveis temporarily deployed to WindowsCILock May 13, 2025 12:50 — with GitHub Actions Inactive
@Maetveis Maetveis temporarily deployed to WindowsCILock May 13, 2025 12:50 — with GitHub Actions Inactive
@Maetveis Maetveis temporarily deployed to WindowsCILock May 14, 2025 07:20 — with GitHub Actions Inactive
@Maetveis
Copy link
Contributor Author

I believe the CI failure (https://github.com/intel/llvm/actions/runs/14996606351/job/42133884358 ) is unrelated: #18416

@Maetveis Maetveis requested a review from kbenzie May 14, 2025 07:27
@Maetveis Maetveis temporarily deployed to WindowsCILock May 14, 2025 07:40 — with GitHub Actions Inactive
@Maetveis Maetveis temporarily deployed to WindowsCILock May 14, 2025 07:40 — with GitHub Actions Inactive
@Maetveis
Copy link
Contributor Author

I don't believe the CI failure is related:

[ RUN      ] urCommandBufferAppendMemBufferWriteRectTestWithParam.Success/Intel_R__oneAPI_Unified_Runtime_over_Level_Zero_V2__Intel_R__Data_Center_GPU_Max_1100_ID0ID______________________write_3D_with_offsets
Segmentation fault from GPU at 0x8057ffc01000, ctx_id: 1 (CCS) type: 0 (NotPresent), level: 0 (PTE), access: 0 (Read), banned: 1, aborting.
Segmentation fault from GPU at 0x8057ffc01000, ctx_id: 1 (CCS) type: 0 (NotPresent), level: 0 (PTE), access: 0 (Read), banned: 1, aborting.
Abort was called at 276 line in file:
../../neo/shared/source/os_interface/linux/drm_neo.cpp

@Maetveis
Copy link
Contributor Author

ping @intel/unified-runtime-reviewers-level-zero

Copy link
Contributor

@pbalcer pbalcer left a comment

Choose a reason for hiding this comment

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

mostly lgtm, just a small nit.

@Maetveis Maetveis requested a review from pbalcer May 19, 2025 13:57
@Maetveis Maetveis temporarily deployed to WindowsCILock May 19, 2025 13:57 — with GitHub Actions Inactive
@pbalcer
Copy link
Contributor

pbalcer commented May 19, 2025

I don't believe the CI failure is related:

[ RUN      ] urCommandBufferAppendMemBufferWriteRectTestWithParam.Success/Intel_R__oneAPI_Unified_Runtime_over_Level_Zero_V2__Intel_R__Data_Center_GPU_Max_1100_ID0ID______________________write_3D_with_offsets
Segmentation fault from GPU at 0x8057ffc01000, ctx_id: 1 (CCS) type: 0 (NotPresent), level: 0 (PTE), access: 0 (Read), banned: 1, aborting.
Segmentation fault from GPU at 0x8057ffc01000, ctx_id: 1 (CCS) type: 0 (NotPresent), level: 0 (PTE), access: 0 (Read), banned: 1, aborting.
Abort was called at 276 line in file:
../../neo/shared/source/os_interface/linux/drm_neo.cpp

This is a known driver bug.

@Maetveis
Copy link
Contributor Author

Maetveis commented May 20, 2025

@intel/llvm-gatekeepers please merge :)

CI failure is unrelated: #18544
I believe pre-commit failure is also unrelated, looks to be an infrastructure issue:

lib/fallback-imf-bf16-host.o /workspace/SYCL_CI/intel/Build_PR_RHEL/llvm.obj/lib/fallback-imf-bf16-host.o 

[2025-05-19T14:12:17.306Z] cd /workspace/SYCL_CI/intel/Build_PR_RHEL/llvm.obj/tools/libdevice && /workspace/SYCL_CI/intel/Build_PR_RHEL/llvm.obj/bin/clang-21 -Wno-sycl-strict -Wno-undefined-internal -sycl-std=2020 --target=x86_64-unknown-linux-gnu --gcc-install-dir=<redacted> -c --target=x86_64-unknown-linux-gnu -D__LIBDEVICE_HOST_IMPL__ --gcc-install-dir=<redacted> -fPIC -I /workspace/SYCL_CI/intel/Build_PR_RHEL/llvm.src/libdevice/imf /workspace/SYCL_CI/intel/Build_PR_RHEL/llvm.obj/./lib/libdevice/imf_bf16_fallback.cpp -o /workspace/SYCL_CI/intel/Build_PR_RHEL/llvm.obj/./lib/fallback-imf-bf16-host.o

[2025-05-19T14:12:17.306Z] clang-21: error: '<redacted>' does not contain a GCC installation

@steffenlarsen steffenlarsen merged commit 4e535c8 into intel:sycl May 20, 2025
26 of 33 checks passed
@Maetveis Maetveis deleted the l0-refactor-ir-handling branch May 20, 2025 06:09
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.

3 participants