[https://nvbugs/6120535][fix] Only call _verify_ctx_response and _get_gen_request when generation will actuall#13701
Open
tensorrt-cicd wants to merge 1 commit intoNVIDIA:mainfrom
Open
Conversation
…st completes When a context-only request finishes during the context phase (finish_reason='stop'), the C++ executor does not populate ctx_request_id in disaggregated_params because no KV cache transfer is needed. The _verify_ctx_response validation was called unconditionally, causing a ValueError for these completed requests. Only verify disaggregated params and prepare the generation request when generation is actually needed (finish_reason is 'length' or 'not_finished'), matching the _need_gen logic that already handles this case correctly downstream. Signed-off-by: tensorrt-cicd <90828364+tensorrt-cicd@users.noreply.github.com>
Contributor
📝 WalkthroughWalkthroughThe PR modifies context-response handling in ChangesContext Response Handling
Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 minutes 🚥 Pre-merge checks | ✅ 3 | ❌ 2❌ Failed checks (2 warnings)
✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Review rate limit: 9/10 reviews remaining, refill in 6 minutes. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@tensorrt_llm/serve/openai_disagg_service.py`:
- Around line 148-153: The code dereferences ctx_response.choices[0] before
validating structure, risking IndexError; change the logic to validate the
response structure first by calling self._verify_ctx_response(ctx_response) (or
explicitly guard that ctx_response and ctx_response.choices exist and are
non-empty) before accessing choices[0] or checking its finish_reason, then
proceed to call self._get_gen_request(request, ctx_response, disagg_request_id)
only after validation passes.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Enterprise
Run ID: 61e69492-507f-4b02-a4f2-079e7efbbec9
📒 Files selected for processing (1)
tensorrt_llm/serve/openai_disagg_service.py
Comment on lines
+148
to
+153
| if not ctx_response or ctx_response.choices[0].finish_reason in ( | ||
| "length", | ||
| "not_finished", | ||
| ): | ||
| await self._verify_ctx_response(ctx_response) | ||
| gen_req = self._get_gen_request(request, ctx_response, disagg_request_id) |
Contributor
There was a problem hiding this comment.
Guard choices[0] access before finish-reason check.
Line 148 dereferences ctx_response.choices[0] before structural validation. If the ctx payload is malformed (e.g., zero choices), this throws IndexError instead of the controlled ValueError from _verify_ctx_response.
Proposed fix
ctx_response = await self._ctx_client.send_request(
ctx_req, server=ctx_server, hooks=hooks
)
- if not ctx_response or ctx_response.choices[0].finish_reason in (
+ if ctx_response and len(ctx_response.choices) != 1:
+ await self._verify_ctx_response(ctx_response)
+
+ if not ctx_response or ctx_response.choices[0].finish_reason in (
"length",
"not_finished",
):
await self._verify_ctx_response(ctx_response)
gen_req = self._get_gen_request(request, ctx_response, disagg_request_id)🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@tensorrt_llm/serve/openai_disagg_service.py` around lines 148 - 153, The code
dereferences ctx_response.choices[0] before validating structure, risking
IndexError; change the logic to validate the response structure first by calling
self._verify_ctx_response(ctx_response) (or explicitly guard that ctx_response
and ctx_response.choices exist and are non-empty) before accessing choices[0] or
checking its finish_reason, then proceed to call self._get_gen_request(request,
ctx_response, disagg_request_id) only after validation passes.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Test plan
Links
Summary by CodeRabbit
Release Notes