fix: Improve cancellation handling for gRPC non-decoupled inference

[pskiran1]

What does the PR do?

This PR is intended to fix the following issue:

• In gRPC inference for a non-decoupled model, where the response and final flag are sent separately, if the client cancels the request before the final flag is sent (while sending the response), the related response_release_payload will be deleted during the cancellation handling. This causes a problem because the next call to InferResponseComplete (sending the final flag) will attempt to use a state that no longer exists.

Made changes to the existing test cases to cover more test data. Also added new test cases, test_grpc_async_cancel and test_grpc_async, to test the code changes.

Checklist

• PR title reflects the change and is of format <commit_type>: <Title>
• Changes are described in the pull request.
• Related issues are referenced.
• Populated github labels field
• Added test plan and verified test passes.
• Verified that the PR passes existing CI.
• Verified copyright is correct on all changed files.
• Added succinct git squash message before merging ref.
• All template sections are filled out.
• Optional: Additional screenshots for behavior/output changes with before/after.

Commit Type:

Check the conventional commit type
box here and add the label to the github PR.

• build
• ci
• docs
• feat
• fix
• perf
• refactor
• revert
• style
• test

Related PRs:

Where should the reviewer start?

Test plan:

• CI Pipeline ID: 28983978

Caveats:

Background

Related Issues: (use one of the action keywords Closes / Fixes / Resolves / Relates to)

• closes GitHub issue: #xxx

[yinggeh]

Nice fix