Port libtorchaudio GPU extensions to build under ROCm/HIP#16
Closed
ethanwee1 wants to merge 2 commits into
Closed
Conversation
The torch CUDAExtension HIPIFY path only rewrites source files, not headers, so CUDA symbols used in headers survived un-hipified and broke the ROCm build (e.g. "cuda_runtime_api.h file not found", and the rnnt/forced_align GPU kernels were entirely gated behind #ifdef USE_CUDA). - Add cuda_compat.h mapping the CUDA runtime names used in headers (cudaStream_t, cudaError_t, cudaSuccess, cudaGetErrorString, cudaGetLastError) to their HIP equivalents under USE_ROCM. HIP natively supports the <<<>>> launch syntax and warp shuffle intrinsics already guarded in the kernels. - Extend the #ifdef USE_CUDA guards in the rnnt GPU headers, macros.h, options.h, and cuctc headers to also cover USE_ROCM, including HIP fp16/runtime headers.
ethanwee1
added a commit
to ROCm/pytorch
that referenced
this pull request
Jul 1, 2026
Temporarily points torchaudio at the ROCm/audio#16 HIP-port commit (ethanwee1/audio fork) so TheRock can build/validate it. Repoint to the merged ROCm/audio release/2.11.0.1 commit once ROCm/audio#16 lands.
Author
|
TheRock test build (Windows gfx110X-all) exercising this HIP port via ROCm/pytorch#3382: |
- cuda_compat.h: alias the CUDA runtime functions/enums used by the sources (cudaSetDevice, cudaMemcpy*, cudaMemset*, cudaStreamSynchronize, cudaGetErrorName, cudaMemcpy kinds) to HIP under USE_ROCM. - Route cub through hipcub (namespace cub = hipcub) under USE_ROCM in forced_align and cuctc kernels. - Enable Options::stream_ for USE_ROCM. - ctc_prefix_decoder.cpp includes cuda_compat.h instead of cuda_runtime.h.
Author
|
Build with hipcub + runtime shim + Options::stream_ (via ROCm/pytorch#3382): Fixes this round: cuda_compat.h now aliases the CUDA runtime funcs used by the sources (cudaSetDevice/cudaMemcpy*/cudaMemset*/cudaStreamSynchronize/cudaGetErrorName), cub routed through hipcub, and Options::stream_ enabled for USE_ROCM. |
jithunnair-amd
pushed a commit
to ROCm/pytorch
that referenced
this pull request
Jul 1, 2026
https://amd-hub.atlassian.net/browse/ROCM-24803?focusedCommentId=500051 Interim unblock: pins torchaudio to the stable release/2.11 (2.11.0), which already builds and publishes on Windows, replacing the floating pytorch/audio main (2.11.0a0) pin. The torchaudio 2.11.0.1 ROCm GPU port is tracked separately in #3382 and ROCm/audio#16.
Author
|
Superseded by the HIPIFY-based approach (mirrors torchvision's setup.py) instead of a hand-written CUDA->HIP compat shim. |
ethanwee1
added a commit
to ROCm/pytorch
that referenced
this pull request
Jul 6, 2026
…io#16) Revert the temporary HIPIFY-based pin (ROCm/audio#17) back to the compat-shim port commit that passed TheRock Windows validation (28529085509). The HIPIFY setup.py approach still leaves CUDA headers unconverted; the cuda_compat.h port in ROCm/audio#16 is the validated path. The Windows cross-drive HIPIFY fix remains in torch/utils/cpp_extension.py (this PR); it is unrelated to the audio-side port strategy.
Author
|
Reopened — this is the validated port path. ROCm/pytorch#3382 pin repointed back here ( Validated TheRock Windows build: https://github.com/ROCm/TheRock/actions/runs/28529085509 |
Author
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
https://amd-hub.atlassian.net/browse/ROCM-24803?focusedCommentId=500051