Skip to content

HIP: disable sync warp shuffel operators from clr amd_warp_sync_functions.h #15273

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 1 commit into from
Aug 12, 2025

Conversation

IMbackK
Copy link
Collaborator

@IMbackK IMbackK commented Aug 12, 2025

We dont use these operators at all on hip and they are causing trouble with rocwmma's headers, so lets just disable them.

@github-actions github-actions bot added Nvidia GPU Issues specific to Nvidia GPUs ggml changes relating to the ggml tensor library for machine learning labels Aug 12, 2025
@slojosic-amd
Copy link
Contributor

slojosic-amd commented Aug 12, 2025

@IMbackK could you please add this change to your PR?
hipBLAS changes have been introduced with ROCm 6.5 (same as HIP_ENABLE_WARP_SYNC_BUILTINS change to HIP_DISABLE_WARP_SYNC_BUILTINS) which is 6.4 with 7.0 preview and some of TheRock rocm-sdk builds used by community have HIP_VERSION 605xxxxx. For example: #14634 (comment)
image

@IMbackK
Copy link
Collaborator Author

IMbackK commented Aug 12, 2025

sure but this will be a follow up pr

@IMbackK IMbackK merged commit b049315 into ggml-org:master Aug 12, 2025
47 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
ggml changes relating to the ggml tensor library for machine learning Nvidia GPU Issues specific to Nvidia GPUs
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants