fix: Ensure that Python stub generation works against libnvidia-ml stubs #6188

MartinMarciniszyn · 2025-07-18T16:04:26Z

Summary by CodeRabbit

Chores
- Improved Docker image build efficiency and reliability by consolidating commands and ensuring correct library linking for GPU support.
- Simplified build scripts by removing redundant logic related to library file checks and symbolic linking.
- Updated Docker image tags for LLM-related components to reflect the latest builds.

coderabbitai · 2025-07-18T16:04:33Z

Walkthrough

The Docker multi-stage build process was updated to handle the creation of a symbolic link for libnvidia-ml.so directly within the Dockerfile. Corresponding logic for this symlink creation was removed from the Python build script, centralizing the handling of this library in the container build process. Additionally, LLM-related Docker image tags were updated to newer versions.

Changes

File(s)	Change Summary
docker/Dockerfile.multi	Consolidated OpenCV uninstall/reinstall commands; added RUN command to symlink `libnvidia-ml.so`.
scripts/build_wheel.py	Removed logic for searching and symlinking `libnvidia-ml.so`; environment variable logic remains.
jenkins/current_image_tags.properties	Updated LLM Docker image tags to newer timestamped versions.

Suggested reviewers

DomBrown

Poem

In Docker’s warren, links are made anew,
While Python scripts shed tasks they once knew.
No more searching for libraries in fright—
The Dockerfile now makes symlinks right.
A tidy warren, a build more bright!
🐇✨

📜 Recent review details

Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between a6f21ae and 44c545c.

📒 Files selected for processing (1)

jenkins/current_image_tags.properties (1 hunks)

✅ Files skipped from review due to trivial changes (1)

jenkins/current_image_tags.properties

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)

GitHub Check: Pre-commit Check

🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

Review comments: Directly reply to a review comment made by CodeRabbit. Example:
- I pushed a fix in commit <commit_id>, please review it.
- Explain this complex logic.
- Open a follow-up GitHub issue for this discussion.
Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
- @coderabbitai explain this code block.
- @coderabbitai modularize this function.
PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
- @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
- @coderabbitai read src/utils.ts and explain its main purpose.
- @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
- @coderabbitai help me debug CodeRabbit configuration file.

Support

Need help? Create a ticket on our support page for assistance with any issues or questions.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (Invoked using PR comments)

@coderabbitai pause to pause the reviews on a PR.
@coderabbitai resume to resume the paused reviews.
@coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
@coderabbitai full review to do a full review from scratch and review all the files again.
@coderabbitai summary to regenerate the summary of the PR.
@coderabbitai generate docstrings to generate docstrings for this PR.
@coderabbitai generate sequence diagram to generate a sequence diagram of the changes in this PR.
@coderabbitai resolve resolve all the CodeRabbit review comments.
@coderabbitai configuration to show the current CodeRabbit configuration for the repository.
@coderabbitai help to get help.

Other keywords and placeholders

Add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.
Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
Add @coderabbitai anywhere in the PR title to generate the title automatically.

Documentation and Community

Visit our Documentation for detailed information on how to use CodeRabbit.
Join our Discord Community to get help, request features, and share feedback.
Follow us on X/Twitter for updates and announcements.

coderabbitai

Actionable comments posted: 1

📜 Review details

Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 07e8813 and a6f21ae.

📒 Files selected for processing (2)

docker/Dockerfile.multi (2 hunks)
scripts/build_wheel.py (0 hunks)

💤 Files with no reviewable changes (1)

scripts/build_wheel.py

🔇 Additional comments (1)

docker/Dockerfile.multi (1)

73-77: Uninstall command may miss the actual package name

opencv is not the canonical PyPI package; the wheel in NVIDIA images is usually opencv-python / opencv-python-headless.
If the base layer was installed with opencv-python, the pip uninstall -y opencv line becomes a no-op and leaves the old wheel on disk, defeating the purpose of the clean re-install.

Consider uninstalling all known OpenCV wheels in one shot:
-RUN pip3 uninstall -y opencv &&  \
+RUN pip3 uninstall -y opencv opencv-python opencv-python-headless opencv-contrib-python || true && \
This will work regardless of which flavour was pre-installed.

docker/Dockerfile.multi

MartinMarciniszyn · 2025-07-18T17:19:19Z

/bot run --skip-test

tensorrt-cicd · 2025-07-18T17:25:35Z

PR_Github #12339 [ run ] triggered by Bot

MartinMarciniszyn · 2025-07-18T17:28:35Z

Supersedes #5940 .

xavier-nvidia · 2025-07-18T17:39:19Z

These changes look good to me.

tensorrt-cicd · 2025-07-18T18:05:43Z

PR_Github #12339 [ run ] completed with state FAILURE
/LLM/main/L0_MergeRequest_PR pipeline #9166 (Partly Tested) completed with status: 'FAILURE'

scripts/build_wheel.py

Signed-off-by: Martin Marciniszyn Mehringer <11665257+MartinMarciniszyn@users.noreply.github.com>

MartinMarciniszyn · 2025-08-08T13:03:55Z

/bot run --skip-tests

tensorrt-cicd · 2025-08-08T13:09:40Z

PR_Github #14608 Bot args parsing error: usage: /bot [-h]
{run,kill,skip,submit,reviewers,reuse-pipeline,reuse-review} ...
/bot: error: unrecognized arguments: --skip-tests

MartinMarciniszyn · 2025-08-08T13:21:17Z

/bot run --skip-test

tensorrt-cicd · 2025-08-08T13:26:35Z

PR_Github #14611 [ run ] triggered by Bot

tensorrt-cicd · 2025-08-08T15:35:07Z

PR_Github #14611 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #11037 (Partly Tested) completed with status: 'SUCCESS'
Pipeline passed with automatic retried tests. Check the rerun report for details.

juney-nvidia · 2025-08-12T06:37:25Z

@MartinMarciniszyn Hi Martin, I noticed that you use "run --skip-test" for this PR, and this PR actually increase the post-merge failure ratios.

June

…ia-ml stubs (NVIDIA#6188)" This reverts commit 9a8195e.

MartinMarciniszyn · 2025-08-12T07:23:22Z

@juney-nvidia, yes, I did not expect that the Python stubs would have any impact on the tests. I'll follow up and check what happened.

juney-nvidia · 2025-08-12T07:52:26Z

@juney-nvidia, yes, I did not expect that the Python stubs would have any impact on the tests. I'll follow up and check what happened.

It may even affect the pre-merge. Now there is already a revert PR from Freddy to try to fix this:

#6814

Thanks
June

scripts/build_wheel.py

MartinMarciniszyn · 2025-08-12T08:45:23Z

/but run

MartinMarciniszyn self-assigned this Jul 18, 2025

coderabbitai bot reviewed Jul 18, 2025

View reviewed changes

docker/Dockerfile.multi Outdated Show resolved Hide resolved

MartinMarciniszyn requested a review from jiaganc July 18, 2025 17:24

jiaganc approved these changes Jul 21, 2025

View reviewed changes

jiaganc mentioned this pull request Jul 21, 2025

fix: Fix Build Error in Local User Container #5940

Closed

jiaganc self-requested a review July 21, 2025 07:57

jiaganc requested changes Jul 21, 2025

View reviewed changes

scripts/build_wheel.py Show resolved Hide resolved

fix: Ensure that Python stub generation works against libnvidia-ml stubs

be36d7a

Signed-off-by: Martin Marciniszyn Mehringer <11665257+MartinMarciniszyn@users.noreply.github.com>

MartinMarciniszyn force-pushed the user/martinma/fix_libnvidia-ml branch from 44c545c to be36d7a Compare August 7, 2025 15:10

MartinMarciniszyn requested review from a team as code owners August 7, 2025 15:10

MartinMarciniszyn requested review from Shixiaowei02 and yuanjingx87 August 7, 2025 15:10

Refactor stub generation in build_wheel.py

49c5345

Signed-off-by: Martin Marciniszyn Mehringer <11665257+MartinMarciniszyn@users.noreply.github.com>

jiaganc self-requested a review August 9, 2025 09:31

jiaganc approved these changes Aug 9, 2025

View reviewed changes

MartinMarciniszyn merged commit 9a8195e into NVIDIA:main Aug 11, 2025
2 of 4 checks passed

QiJune added a commit to QiJune/TensorRT-LLM that referenced this pull request Aug 12, 2025

Revert "fix: Ensure that Python stub generation works against libnvid…

49beea1

…ia-ml stubs (NVIDIA#6188)" This reverts commit 9a8195e.

coderabbitai bot mentioned this pull request Aug 12, 2025

[None][fix] fix ci #6814

Merged

Linda-Stadter reviewed Aug 12, 2025

View reviewed changes

scripts/build_wheel.py Show resolved Hide resolved

coderabbitai bot mentioned this pull request Aug 12, 2025

[None][fix] Clean up linking to CUDA stub libraries in build_wheel.py #6823

Merged

fix: Ensure that Python stub generation works against libnvidia-ml stubs #6188

fix: Ensure that Python stub generation works against libnvidia-ml stubs #6188

Uh oh!

Conversation

MartinMarciniszyn commented Jul 18, 2025 • edited by coderabbitai bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary by CodeRabbit

Uh oh!

coderabbitai bot commented Jul 18, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Suggested reviewers

Poem

Chat

Support

CodeRabbit Commands (Invoked using PR comments)

Other keywords and placeholders

Documentation and Community

Uh oh!

coderabbitai bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

MartinMarciniszyn commented Jul 18, 2025

Uh oh!

tensorrt-cicd commented Jul 18, 2025

Uh oh!

MartinMarciniszyn commented Jul 18, 2025

Uh oh!

xavier-nvidia commented Jul 18, 2025

Uh oh!

tensorrt-cicd commented Jul 18, 2025

Uh oh!

Uh oh!

MartinMarciniszyn commented Aug 8, 2025

Uh oh!

tensorrt-cicd commented Aug 8, 2025

Uh oh!

MartinMarciniszyn commented Aug 8, 2025

Uh oh!

tensorrt-cicd commented Aug 8, 2025

Uh oh!

tensorrt-cicd commented Aug 8, 2025

Uh oh!

Uh oh!

juney-nvidia commented Aug 12, 2025

Uh oh!

MartinMarciniszyn commented Aug 12, 2025

Uh oh!

juney-nvidia commented Aug 12, 2025

Uh oh!

Uh oh!

MartinMarciniszyn commented Aug 12, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

MartinMarciniszyn commented Jul 18, 2025 •

edited by coderabbitai bot

Loading

coderabbitai bot commented Jul 18, 2025 •

edited

Loading