add test for longjump with 0 as return value

[realprocrastinator]

Summary

This is a re-submitted patch reverted @#2720 due to a dependency issue with PR apache/nuttx#14152.

In this PR, we add a testcase specifically checking the behaviour of setjmp() after passing 0 as the return value (aka the second argument) to the longjmp().

Impact

Results in the incorrect behaviour which doesn't meet the POSIX specification.

Testing

ostest

[nuttxpr]

[Experimental Bot, please feedback here]

Let's analyze this Pull Request (PR) against the NuttX requirements:

Strengths:

• Clear Summary: The PR explains the "why" (re-submission, dependency fix) and the "what" (testcase for setjmp/longjmp).
• Impact Acknowledged: The PR identifies an incorrect behavior violating POSIX, which is crucial information.
• Testing Mentioned: The use of "ostest" implies testing was done, though more details would strengthen this.

Areas for Improvement:

• Functional Change Details: The summary should be more specific. How does the testcase work? What exact code is modified?
• Impact Specificity:
  • User Impact: Will this change how applications using setjmp/longjmp behave? If so, how?
  • Other Impacts: Most are marked "NO" without explanation. While likely accurate, brief justifications (e.g., "No new hardware features used") enhance clarity.
• Testing Details:
  • Insufficient Logs: "ostest" alone isn't enough. Provide snippets showing the failing behavior before your change and the correct behavior after.
  • Environment Context: List the specific build host and target(s) used for verification. This helps reviewers reproduce results.

Verdict:

The PR is on the right track but needs more detail to meet NuttX standards. Address the "Areas for Improvement" to make it easier for reviewers to assess and merge.