Skip to content

[SYCL] Remove UNSUPPORTED in copy_subregion_2D.cpp for Windows MTL - #22969

Merged
sarnex merged 5 commits into
intel:syclfrom
lbushi25:enable_copy_subregion_test
Aug 19, 2026
Merged

[SYCL] Remove UNSUPPORTED in copy_subregion_2D.cpp for Windows MTL#22969
sarnex merged 5 commits into
intel:syclfrom
lbushi25:enable_copy_subregion_test

Conversation

@lbushi25

@lbushi25 lbushi25 commented Aug 18, 2026

Copy link
Copy Markdown
Contributor

The test copy_subregion_2D.cpp passes on Level Zero with the latest drivers so remove the UNSUPPORTED for Windows + MTL configuration.

Closes: #21380

@lbushi25
lbushi25 marked this pull request as ready for review August 18, 2026 19:37
@lbushi25
lbushi25 requested a review from a team as a code owner August 18, 2026 19:37
@lbushi25
lbushi25 requested a review from dm-vodopyanov August 18, 2026 19:37
@lbushi25 lbushi25 changed the title [SYCL] Re-enable copy_subregion_2D.cpp for Windows MTL [SYCL] Remove UNSUPPORTED in copy_subregion_2D.cpp for Windows MTL Aug 18, 2026
@dyniols

dyniols commented Aug 19, 2026

Copy link
Copy Markdown
Contributor

@lbushi25 I should have describe the failure better, my bad. I meant L0 because you're right that we don't support OpenCL for bindless backend. However the idea was to provide compute runtime version on which it started to fail.

Since test passes with new driver we can re-enable it.

@sys-ce-bb

Copy link
Copy Markdown
Contributor

@intel/llvm-gatekeepers please consider merging

@KornevNikita

Copy link
Copy Markdown
Contributor

@lbushi25 @dyniols do we have Windows+MTL runner here?

@lbushi25

Copy link
Copy Markdown
Contributor Author

@lbushi25 @dyniols do we have Windows+MTL runner here?

As far as i can tell, no.

@lbushi25

Copy link
Copy Markdown
Contributor Author

@lbushi25 I should have describe the failure better, my bad. I meant L0 because you're right that we don't support OpenCL for bindless backend. However the idea was to provide compute runtime version on which it started to fail.

Since test passes with new driver we can re-enable it.

I see. Yeah I tested Level Zero as well just for good measure and it reported no failures.

@dyniols

dyniols commented Aug 19, 2026

Copy link
Copy Markdown
Contributor

@lbushi25 @dyniols do we have Windows+MTL runner here?

I checked CI and couldn't find an MTL runner. @sarnex Do you know if we have MTL runner so we could modify pre-commit on windows just in this PR to test if re-enabled test doesn't fail on arch-intel_gpu_mtl_h?

@sarnex

sarnex commented Aug 19, 2026

Copy link
Copy Markdown
Contributor

The ARL runner shows up in sycl-ls as intel_gpu_mtl_u, but I don't know which is true. So we definitely don't have a intel_gpu_mtl_h runner so IMO we can just merge it

@sarnex
sarnex merged commit e6d007a into intel:sycl Aug 19, 2026
36 of 37 checks passed
@dyniols

dyniols commented Aug 20, 2026

Copy link
Copy Markdown
Contributor

Thanks

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.

Test bindless_images/copies/copy_subregion_2D.cpp fails for Meteor Lake on Windows

5 participants