[Bugfix][NIXL] Fix Mamba prefill truncation ordering - #53523
Merged
ivanium merged 2 commits intoAug 25, 2026
Merged
Conversation
ZeldaHuang
requested review from
ApostaC,
NickLucche,
ivanium,
orozery and
xuechendi
as code owners
August 24, 2026 04:39
![claude[bot]](https://avatars.githubusercontent.com/in/1236702?s=60&v=4){kind=small}
Move producer-side N-1 truncation into on_new_request so it runs before the scheduler prefix-cache lookup. This keeps the request length stable while computing local cache hits and prevents zero-new-token assertions. Fixes vllm-project#53514 Co-authored-by: OpenAI Codex <noreply@openai.com> Signed-off-by: Ziming Huang <zelda.huanghuang@gmail.com>
ZeldaHuang
force-pushed
the
fix/nixl-mamba-prefill-truncation
branch
from
August 24, 2026 04:51
6357ee0 to
cce91dd
Compare
ywang96
approved these changes
Aug 24, 2026
Member
|
/ci run |
|
✅ Triggered Buildkite CI #85299 for commit |
ivanium
approved these changes
Aug 25, 2026
ivanium
left a comment
ivanium
left a comment
Collaborator
There was a problem hiding this comment.
LGTM. Also I realized that we need the same fix for MooncakeConnector (for PD, not MooncakeStoreConnector for kv offloading) too. We can merge this one first but could you also check and submit another follow-up one for MooncakeConnector 🙏
TomerBN-Nvidia
added a commit
to TomerBN-Nvidia/vllm
that referenced
this pull request
Aug 27, 2026
Move producer-side N-1 truncation into on_new_request so it runs before the scheduler prefix-cache lookup. This keeps the request length stable while computing local cache hits and prevents zero-new-token EngineCore assertions. Backport of vllm-project#53523 (d4c4ceb). Co-authored-by: Ziming Huang <zelda.huanghuang@gmail.com> Co-authored-by: Roger Wang <hey@rogerw.io> Co-authored-by: OpenAI Codex <codex@openai.com> Signed-off-by: tbarnatan <tbarnatan@nvidia.com>
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
3 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.
Move producer-side N-1 truncation into on_new_request so it runs before the scheduler prefix-cache lookup. This keeps the request length stable while computing local cache hits and prevents zero-new-token assertions.
Fixes #53514
Purpose
Test Plan
Test Result
Essential Elements of an Effective PR Description Checklist
supported_models.mdandexamplesfor a new model.