[UnitTest][NVPTX] Avoid cascading failures from CUDA postproc #15136
Conversation
This file contains 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
|
Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews from Reviewers by @-ing them in a comment.
Generated by tvm-bot |
Prior to this commit, the tests in `test_tir_transform_inject_ptx_async_copy.py` registered the `"tvm_callback_cuda_postproc"` function during pytest collection, and used a global variable to disable its functionality outside of the tests in this file. This had two major issues. First, if any other test also installs a postproc function, these postproc function required by the NVPTX tests would be overwritten. Second, if one of the NTPTX tests fails, the global variable controlling the postproc function would not be reset, causing any subsequent CUDA-related tests to also fail. This commit updates these NVPTX tests to conditionally install the postproc function, to de-register it after the test instead of disabling its functionality, and to de-register it regardless of the test result. This issue was initially found when debugging apache#15103, when a failure in `test_tir_transform_inject_ptx_async_copy.py::test_cp_async_in_if_then_else` caused failures in 32 unrelated tests ([CI link](https://ci.tlcpack.ai/blue/organizations/jenkins/tvm-gpu/detail/PR-15103/7/tests)).
Lunderberg
force-pushed
the
pytest_avoid_cuda_postproc_cascading_failures
branch
from
4044f53 to
f302f3a
Compare
junrushao
pushed a commit
to junrushao/tvm
that referenced
this pull request
Jul 15, 2023
…#15136) Prior to this commit, the tests in `test_tir_transform_inject_ptx_async_copy.py` registered the `"tvm_callback_cuda_postproc"` function during pytest collection, and used a global variable to disable its functionality outside of the tests in this file. This had two major issues. First, if any other test also installs a postproc function, these postproc function required by the NVPTX tests would be overwritten. Second, if one of the NTPTX tests fails, the global variable controlling the postproc function would not be reset, causing any subsequent CUDA-related tests to also fail. This commit updates these NVPTX tests to conditionally install the postproc function, to de-register it after the test instead of disabling its functionality, and to de-register it regardless of the test result. This issue was initially found when debugging apache#15103, when a failure in `test_tir_transform_inject_ptx_async_copy.py::test_cp_async_in_if_then_else` caused failures in 32 unrelated tests ([CI link](https://ci.tlcpack.ai/blue/organizations/jenkins/tvm-gpu/detail/PR-15103/7/tests)).
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Prior to this commit, the tests in
test_tir_transform_inject_ptx_async_copy.pyregistered the"tvm_callback_cuda_postproc"function during pytest collection, and used a global variable to disable its functionality outside of the tests in this file. This had two major issues. First, if any other test also installs a postproc function, these postproc function required by the NVPTX tests would be overwritten. Second, if one of the NTPTX tests fails, the global variable controlling the postproc function would not be reset, causing any subsequent CUDA-related tests to also fail.This commit updates these NVPTX tests to conditionally install the postproc function, to de-register it after the test instead of disabling its functionality, and to de-register it regardless of the test result.
This issue was initially found when debugging #15103, when a failure in
test_tir_transform_inject_ptx_async_copy.py::test_cp_async_in_if_then_elsecaused failures in 32 unrelated tests (CI link).