Skip to content

[SYCL][CUDA] Support GroupBroadcast with 32-bit id #2759

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 2 commits into from
Nov 11, 2020

Conversation

Pennycook
Copy link
Contributor

Use of the broadcast algorithm with the sub_group class clamps the
sub-group local id into a uint32_t. libspirv was missing an entry
point for this case.

Signed-off-by: John Pennycook [email protected]

Use of the broadcast algorithm with the sub_group class clamps the
sub-group local id into a uint32_t.  libspirv was missing an entry
point for this case.

Signed-off-by: John Pennycook <[email protected]>
@Pennycook Pennycook added bug Something isn't working spec extension All issues/PRs related to extensions specifications cuda CUDA back-end libclc libclc project related issues labels Nov 10, 2020
@Pennycook Pennycook requested review from bader and a team as code owners November 10, 2020 23:01
@Pennycook Pennycook linked an issue Nov 10, 2020 that may be closed by this pull request
@Pennycook
Copy link
Contributor Author

Sorry for the link-unlink. I wanted to reference #2754 because this is a partial fix. I'll be addressing the load_store test with a separate PR.

@bader bader merged commit 6e3f244 into intel:sycl Nov 11, 2020
@Pennycook Pennycook deleted the cuda-sub-group-broadcast branch January 28, 2021 18:23
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working cuda CUDA back-end libclc libclc project related issues spec extension All issues/PRs related to extensions specifications
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants