[CI/Build][LoRA] Temporarily fix long context failure issue

[jeejeelee]

#7049 added the parameter disable_async_output_proc with a default value of False, which led to test_long_context raising an IndexError: list index out of range error, see: https://buildkite.com/vllm/ci-aws/builds/7645#01919198-199d-488c-9834-8c48ae675338/193-420. Since then, this test has consistently shown this error, which has hidden errors introduced by other PRs. For example, when setting disable_async_output_proc=True, inference works without errors but generates incorrect results due to the cudagraph issue from #8645 (which has been fixed by #9549).

This PR is a temporary solution to make this test pass correctly, so that issues like #8645 can be exposed earlier in CI.

ping @DarkLight1337 , also cc @megha95 @alexm-neuralmagic

[github-actions]

👋 Hi! Thank you for contributing to the vLLM project.
Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging.

To run CI, PR reviewers can do one of these:

• Add ready label to the PR
• Enable auto-merge.

🚀

[DarkLight1337]

Thanks for fixing!

[jeejeelee]

Mark: test_long_context.py test passed,see: https://buildkite.com/vllm/ci-aws/builds/10183#0192b31a-7610-4295-b9e8-8c18fb417560