Fix bug in test executor where manual aborts were not letting the test clean up properly. #1267
+21
−5
Conversation
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
copybara-service
bot
force-pushed
the
test_846000293
branch
from
370df34 to
bb71e9f
Compare
…t clean up properly. To understand this bug, note that: 1. OpenHTF registers a signal handler for SIGINT 2. This handler re-raises KeyboardInterrupt 3. The `execute` function in TestDescriptor expects this, and effectively waits on the test runner thread to join - twice in the case of a manual abort. This would have worked until [Python was "fixed"](python/cpython@a22be49#diff-b250d3d0e0d909784a67083e9069ce507513a67e8514a011e2f364027e25e0b0). (Equivalent code doesn't exist in Python 3.14, but at least today, in Python 3.12, this bites us). After this, the second wait() call would return immediately while the thread was still running. is_alive() also indicates the thread is dead when in fact it's still alive. It's debatable whether the "fix" above is the right thing, but we can avoid this hairy issue by not using `Thread.join()` and instead use `Event.wait()` which we can explicitly set at the point we want the thread to be considered "done". Without this fix, the test executor thread is still running teardowns while the output callbacks have started processing in the main thread, causing output callbacks to receive a test descriptor that hasn't been finalized - among other issues. PiperOrigin-RevId: 846387423
copybara-service
bot
force-pushed
the
test_846000293
branch
from
bb71e9f to
7a363fb
Compare
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Outdated suggestions cannot be applied.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.
Fix bug in test executor where manual aborts were not letting the test clean up properly.
To understand this bug, note that:
executefunction in TestDescriptor expects this, and effectively waits on the test runner thread to join - twice in the case of a manual abort.This would have worked until Python was "fixed". (Equivalent code doesn't exist in Python 3.14, but at least today, in Python 3.12, this bites us). After this, the second wait() call would return immediately while the thread was still running. is_alive() also indicates the thread is dead when in fact it's still alive.
It's debatable whether the "fix" above is the right thing, but we can avoid this hairy issue by not using
Thread.join()and instead useEvent.wait()which we can explicitly set at the point we want the thread to be considered "done".Without this fix, the test executor thread is still running teardowns while the output callbacks have started processing in the main thread, causing output callbacks to receive a test descriptor that hasn't been finalized - among other issues.