Conversation
…alls JIT load_inline against TheRock wheels was resolving ROCM_HOME to runtime-only _rocm_sdk_core, so hipBLAS headers in _rocm_sdk_devel were never on the include path. Prefer devel via find_spec.
… core Trunk ROCm CI exports ROCM_HOME, so TestMemPool never hits Guess #2. Mock find_spec with both packages and the env unset so CI covers the TheRock JIT include path.
Author
|
Superseded by #3635 (the functional fix, already merged). Tests-only follow-up for trunk is pytorch#197449. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
QA 2.13 Docker images fail four
TestMemPooltests becauseload_inlinebuilds
dummy_allocatorwith-isystem …/_rocm_sdk_core/include. Devel isalready installed; torch never puts it on the include path. Tracked in
ROCM-30466.
PyTorch ROCm CI exports
ROCM_HOMEfromrocm-sdk path --root, so it neverhits Guess #2.
Changes
Backport of pytorch#195726:
_find_rocm_home()prefers_rocm_sdk_develover_rocm_sdk_coreviafind_spec. AddsTestFindRocmHomeso CI covers Guess #2 with the env unset.
Validation
Equivalent 2.14 QA image (container
rocm30466), env unset: fourTestMemPooltests FAILED (errors=4) unpatched, OK patched. Same
_find_rocm_homehunk.TestFindRocmHomein that container, 3/3 passed in 4.667s (devel-over-core,core fallback, env wins).
Upstream: pytorch#195726