Skip to content
This repository was archived by the owner on Mar 28, 2023. It is now read-only.

[SYCL][CUDA] Enables sub_group_sycl2020 for CUDA #316

Merged

Conversation

steffenlarsen
Copy link

sub_group_sycl2020.cpp test is currently marked as unsupported for CUDA due to missing support for kernel-bundles. This patch adds a special CUDA path for avoiding the use of kernel-bundles, using the known size of sub-groups instead of querying for it.

This is a temporary work-around and can be reverted once kernel-bundle support is added in the PI CUDA backend.

sub_group_sycl2020.cpp test is currently marked as unsupported for CUDA
due to missing support for kernel-bundles. This patch adds a special
CUDA path for avoiding the use of kernel-bundles, using the known size
of sub-groups instead of querying for it.

Signed-off-by: Steffen Larsen <[email protected]>
@romanovvlad romanovvlad merged commit c540803 into intel:intel Jun 9, 2021
smaslov-intel pushed a commit to smaslov-intel/llvm-test-suite that referenced this pull request Aug 12, 2021
sub_group_sycl2020.cpp test is currently marked as unsupported for CUDA
due to missing support for kernel-bundles. This patch adds a special
CUDA path for avoiding the use of kernel-bundles, using the known size
of sub-groups instead of querying for it.

Signed-off-by: Steffen Larsen <[email protected]>
myler pushed a commit to myler/llvm-test-suite that referenced this pull request Apr 12, 2022
CMPLRLLVM-29952: add bin-llvm to PATH for SYCL testing
aelovikov-intel pushed a commit to aelovikov-intel/llvm that referenced this pull request Mar 27, 2023
…te#316)

sub_group_sycl2020.cpp test is currently marked as unsupported for CUDA
due to missing support for kernel-bundles. This patch adds a special
CUDA path for avoiding the use of kernel-bundles, using the known size
of sub-groups instead of querying for it.

Signed-off-by: Steffen Larsen <[email protected]>
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants