[release/1.12.0] Cherry-pick: Fix 'Memory access fault' for fused dense on MI350 (#325)#331
Merged
pragupta merged 1 commit intoMay 26, 2026
Conversation
Use getCurrentCUDABlasLtHandle() instead of getCurrentCUDABlasHandle() in gemm_lt() to get the correct hipBLASLt handle, fixing a write-to-read-only-page fault on gfx950. Cherry-pick of af25af4
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Cherry-pick of #325 to release/1.12.0
JIRA: ROCM-24646
Original PR: #325
Original commit: af25af4
Summary
Cherry-picks the fix for GPU memory access fault in
fused_dense_cudaon MI350 (gfx950).Root cause:
gemm_lt()was usingat::cuda::getCurrentCUDABlasHandle()(regular hipBLAS handle) instead ofat::cuda::getCurrentCUDABlasLtHandle()(hipBLASLt handle) for hipBLASLt operations, causing a write-to-read-only-page fault.Fix: Replace
getCurrentCUDABlasHandle()withgetCurrentCUDABlasLtHandle()ingemm_lt().