Skip to content

[https://nvbugs/6160629][fix] Add both test filenames to EXCLUDE_TEST_FILES in `examples/auto_deploy/llmc/cr - #13943

Closed
tensorrt-cicd wants to merge 3 commits into
NVIDIA:mainfrom
tensorrt-cicd:repair-bot-bug6160629
Closed

[https://nvbugs/6160629][fix] Add both test filenames to EXCLUDE_TEST_FILES in `examples/auto_deploy/llmc/cr#13943
tensorrt-cicd wants to merge 3 commits into
NVIDIA:mainfrom
tensorrt-cicd:repair-bot-bug6160629

Conversation

@tensorrt-cicd

@tensorrt-cicd tensorrt-cicd commented May 9, 2026 •
edited by coderabbitai Bot
Loading

Copy link
Copy Markdown
Collaborator

Summary

  • Root cause: Two TRT-LLM-only test files (test_quant_fusion.py, test_finegrained_fp8_swiglu.py) are shipped in the standalone package but can't run without the full TRT-LLM runtime — causing 1 collection error (ModuleNotFoundError: utils.util) and 2 failures (KeyError: fuse_fp8_linear because fuse_quant.py fails to import tensorrt_llm.quantization.utils.fp8_utils).
  • Fix: Add both test filenames to EXCLUDE_TEST_FILES in examples/auto_deploy/llmc/create_standalone_package.py, matching the pattern used for other TRT-LLM-kernel-dependent tests; verified with a full run of the outer standalone wrapper.
  • Automated fix generated by repair-bot

Test plan

  • Verify fix on the same GPU type as the original failure
  • Check for regressions in related tests

Links

Summary by CodeRabbit

  • Chores
    • Updated standalone package exclusion list to include two additional test files, preventing them from being included in the standalone distribution.

Review Change Stack

@tensorrt-cicd
[nvbugs/6160629][fix] Exclude TRT-LLM-only quant tests from standalon…
e5e613c 
…e package

test_quant_fusion.py uses the fuse_fp8_linear transform, which is registered in
fuse_quant.py. That module imports tensorrt_llm.quantization.utils.fp8_utils at
module level and is silently skipped in standalone mode, so the transform never
registers and test_fuse_quant_rewrites_fp8_linear hits KeyError: 'fuse_fp8_linear'.
Sibling tests in the same file have a trtllm_ops_available guard and are already
skipped correctly.

test_finegrained_fp8_swiglu.py imports utils.util.skip_pre_blackwell (not shipped
in the standalone _utils_test directory) and exercises fuse_finegrained_fp8_swiglu
which also requires TRT-LLM runtime, so its collection fails in standalone.

Add both files to EXCLUDE_TEST_FILES alongside the existing TRT-LLM-kernel
exclusions.

Signed-off-by: tensorrt-cicd <90828364+tensorrt-cicd@users.noreply.github.com>
@tensorrt-cicd
tensorrt-cicd requested a review from a team as a code owner May 9, 2026 13:08
@tensorrt-cicd
tensorrt-cicd requested a review from MrGeva May 9, 2026 13:08
@coderabbitai

coderabbitai Bot commented May 9, 2026

Copy link
Copy Markdown
Contributor

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info
⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 978f482c-3fc7-492b-b315-80e145f090d0

📥 Commits

Reviewing files that changed from the base of the PR and between c163c72 and e5e613c.

📒 Files selected for processing (1)
  • examples/auto_deploy/llmc/create_standalone_package.py
📝 Walkthrough

Walkthrough

The PR adds two test file exclusions to the standalone package creation script, preventing tests that depend on unavailable transforms or utilities from being included in the standalone build. Each exclusion is documented with an inline comment.

Changes

Standalone Package Test Exclusions

Layer / File(s) Summary
Exclusion List Update
examples/auto_deploy/llmc/create_standalone_package.py		 EXCLUDE_TEST_FILES is extended with test_quant_fusion.py and test_finegrained_fp8_swiglu.py, each with comments explaining dependencies on unavailable transforms or utilities in standalone mode.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~2 minutes

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 inconclusive)

Check name Status Explanation Resolution
Description check ❓ Inconclusive The description provides root causes, the fix applied, and test verification, but lacks explicit sections matching the template structure (Description, Test Coverage, PR Checklist). Restructure the description to follow the template with explicit sections: Description (explaining what and why), Test Coverage (listing test safeguards), and PR Checklist confirmation.
✅ Passed checks (4 passed)
Check name Status Explanation
Title check ✅ Passed The title clearly identifies the fix (adding test filenames to exclusion list) and references the specific file being modified, directly matching the changeset.
Docstring Coverage ✅ Passed No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check ✅ Passed Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check ✅ Passed Check skipped because no linked issues were found for this pull request.

✏️ Tip: You can configure your own custom pre-merge checks in the settings.

✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests

Comment @coderabbitai help to get the list of available commands and usage tips.

@suyoggupta
suyoggupta requested a review from bmarimuthu-nv May 10, 2026 01:16
@suyoggupta

suyoggupta commented May 10, 2026

Copy link
Copy Markdown
Collaborator

@bmarimuthu-nv : please review

@suyoggupta

suyoggupta commented May 10, 2026

Copy link
Copy Markdown
Collaborator

/bot run

@tensorrt-cicd

tensorrt-cicd commented May 10, 2026

Copy link
Copy Markdown
Collaborator Author

PR_Github #47553 [ run ] triggered by Bot. Commit: 02711eb Link to invocation

@tensorrt-cicd

tensorrt-cicd commented May 10, 2026

Copy link
Copy Markdown
Collaborator Author

PR_Github #47553 [ run ] completed with state FAILURE. Commit: 02711eb
/LLM/main/L0_MergeRequest_PR pipeline #37468 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

  • Please check the failed tests and fix your PR
  • If you cannot view the failures, ask the CI triggerer to share details
  • Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

@suyoggupta

suyoggupta commented May 11, 2026

Copy link
Copy Markdown
Collaborator

/bot run

@tensorrt-cicd

tensorrt-cicd commented May 11, 2026

Copy link
Copy Markdown
Collaborator Author

PR_Github #47770 [ run ] triggered by Bot. Commit: f6f5ab8 Link to invocation

@suyoggupta

suyoggupta commented May 11, 2026

Copy link
Copy Markdown
Collaborator

@bmarimuthu-nv : can you please push a commit that also unwaives the tests. Looks like all of single gpu transforms are waived at the moment! https://sourcegraph.com/r/github.com/NVIDIA/TensorRT-LLM/-/blob/tests/integration/test_lists/waives.txt?L403

@tensorrt-cicd

tensorrt-cicd commented May 11, 2026

Copy link
Copy Markdown
Collaborator Author

PR_Github #47770 [ run ] completed with state SUCCESS. Commit: f6f5ab8
/LLM/main/L0_MergeRequest_PR pipeline #37662 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

  • Please check the failed tests and fix your PR
  • If you cannot view the failures, ask the CI triggerer to share details
  • Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

@bmarimuthu-nv

bmarimuthu-nv commented May 11, 2026

Copy link
Copy Markdown
Collaborator

@suyoggupta PTAL fixed in #14011

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@suyoggupta suyoggupta suyoggupta approved these changes
@bmarimuthu-nv bmarimuthu-nv bmarimuthu-nv approved these changes
@MrGeva MrGeva Awaiting requested review from MrGeva MrGeva is a code owner automatically assigned from NVIDIA/trt-llm-torch-autodeploy-devs

Assignees

@tensorrt-cicd tensorrt-cicd

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@tensorrt-cicd @suyoggupta @bmarimuthu-nv