[PyTorch] [CI] Capture subprocess stderr in distributed tests for better CI error re…

[sudhakarsingh27]

…porting

Distributed tests launch subprocesses via torch.distributed.launch/torchrun. When these fail, pytest only captures the CalledProcessError from the parent process, not the actual worker traceback. This makes CI JUnit XML reports show "exit code 1" with no useful error detail.

Add run_distributed() utility to tests/pytorch/utils.py that captures stderr while letting stdout stream to the terminal. On failure, the worker's stderr (containing the actual Python traceback) is included in the AssertionError, which pytest writes into the JUnit XML report.

Behavior:

• Interactive use: stdout streams in real time (unchanged), stderr shown on failure
• CI/JUnit XML: failure reports now include the actual worker traceback

Description

Please include a brief summary of the changes, relevant motivation and context.

Fixes # (issue)

Type of change

• Documentation change (change only to the documentation, either a fix or a new content)
• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Infra/Build change
• Code refactoring

Changes

Please list the changes introduced in this PR:

• Change A
• Change B

Checklist:

• I have read and followed the contributing guidelines
• The functionality is complete
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes