Fix Fray hang when a thread exits but the monitor lock is acquired by… #151
Conversation
There was a problem hiding this comment.
PR Overview
This PR introduces a new test case and updates test scheduling to verify the fix for a hang that occurred when a thread exits while holding a monitor lock.
- Add a new test case (ThreadExitDeadlock.java) simulating a potential deadlock scenario.
- Update the test suite (FrayTestCase.java) to invoke the new test case and change the scheduler from RandomScheduler to POSScheduler.
Reviewed Changes
| File | Description |
|---|---|
| integration-test/src/main/java/org/pastalab/fray/test/fail/thread/ThreadExitDeadlock.java | Added a test case to simulate the thread exit deadlock scenario. |
| CHANGELOG.md | Updated changelog to include a fix for the monitor lock issue. |
| integration-test/src/test/java/org/pastalab/fray/test/FrayTestCase.java | Modified test to execute the new ThreadExitDeadlock test case and changed scheduler to POSScheduler. |
Copilot reviewed 5 out of 5 changed files in this pull request and generated 1 comment.
Comments suppressed due to low confidence (1)
integration-test/src/test/java/org/pastalab/fray/test/FrayTestCase.java:94
- [nitpick] The replacement of RandomScheduler with POSScheduler may affect the test behavior. Consider adding a brief comment or confirming that POSScheduler provides the intended scheduling for this test scenario.
new POSScheduler(),
| }); | ||
| synchronized (t) { | ||
| t.start(); | ||
| latch.await(); |
There was a problem hiding this comment.
The latch is never decremented in the thread's runnable, which causes latch.await() to block indefinitely. If this is intended to simulate a deadlock scenario, please add a comment clarifying the purpose; otherwise, consider calling latch.countDown() in the thread.
Copilot is powered by AI, so mistakes are possible. Review output carefully before use.
There was a problem hiding this comment.
Hey Copilot, good catch! But this is an intended behavior.
… another thread.