Skip to content

[https://nvbugs/6427411][test] Re-enable PP regression tests - #17117

Merged
mingyangHao merged 3 commits into
NVIDIA:mainfrom
mingyangHao:user/mingyangh/unwaive-nvbug-6427411
Aug 14, 2026
Merged

[https://nvbugs/6427411][test] Re-enable PP regression tests#17117
mingyangHao merged 3 commits into
NVIDIA:mainfrom
mingyangHao:user/mingyangh/unwaive-nvbug-6427411

Conversation

@mingyangHao

@mingyangHao mingyangHao commented Jul 31, 2026

Copy link
Copy Markdown
Collaborator

Remove the remaining PP regression waivers added after #15920. The host-side greedy stop change was reverted by #16163, so these tests should run in CI again.

Dev Engineer Review

  • tests/integration/test_lists/waives.txt removes obsolete waiver entries.
  • The change re-enables affected pipeline parallelism regression tests in CI.
  • No production-code, API, or configuration changes are included.
  • No test-list format, path, duplicate-entry, or bug-reference issues were identified.

QA Engineer Review

  • No test-db/ or qa/ files were modified.
  • CI waiver entries were removed from tests/integration/test_lists/waives.txt.
  • The affected pipeline parallelism regression tests should run in CI again.
  • CBTS coverage data is unavailable.

Verdict: needs follow-up

Description

Test Coverage

PR Checklist

Please review the following before submitting your PR:

  • PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.

  • PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.

  • Test cases are provided for new code paths (see test instructions)

  • If PR introduces API changes, an appropriate PR label is added - either api-compatible or api-breaking. For api-breaking, include BREAKING in the PR title.

  • Any new dependencies have been scanned for license and vulnerabilities

  • CODEOWNERS updated if ownership changes

  • Documentation updated as needed

  • Update tava architecture diagram if there is a significant design change in PR.

  • The reviewers assigned automatically/manually are appropriate for the PR.

  • Please check this after reviewing the above items as appropriate for this PR.

GitHub Bot Help

To see a list of available CI bot commands, please comment /bot help.

@mingyangHao

Copy link
Copy Markdown
Collaborator Author

/bot run

@coderabbitai

coderabbitai Bot commented Jul 31, 2026

Copy link
Copy Markdown
Contributor

Review Change Stack

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: b04cc397-22f8-40c1-8a84-5c40b680af65

📥 Commits

Reviewing files that changed from the base of the PR and between 84d2fb6 and c3d3b68.

📒 Files selected for processing (1)
  • tests/integration/test_lists/waives.txt

Walkthrough

Updated integration test waiver entries for disaggregated serving, model configurations, Ray, and LLM API cases. Added Llama3.3 FP4 skips and removed obsolete skips.

Changes

Integration waiver maintenance

Layer / File(s) Summary
Refresh integration test waivers
tests/integration/test_lists/waives.txt
Updated Llama, DeepSeekV3Lite, Nemotron, Ray, statistics, additional-model-output, and logits-processor entries. Added Llama3.3 FP4 skips and removed obsolete waivers.

Estimated code review effort: 1 (Trivial) | ~5 minutes

Merge Risk: ⚪ Minimal · up to c3d3b

This localized test-configuration change re-enables the intended PP regression tests without changing production behavior. No actionable merge-blocking risk remains beyond normal checks and review.

Suggested reviewers: jieli-matrix, schetlur-nv, yuanjingx87, zhanruisunch

🚥 Pre-merge checks | ✅ 5
✅ Passed checks (5 passed)
Check name Status Explanation
Title check ✅ Passed The title clearly states that PP regression tests are re-enabled and follows the repository's ticket and type format.
Description check ✅ Passed The description explains the reason for removing the waivers and identifies the related changes, although the formal test coverage section is empty.
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.
✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests

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

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #63003 [ run ] triggered by Bot. Commit: 01c40d9 Link to invocation

@mingyangHao
mingyangHao requested a review from shuyixiong July 31, 2026 09:08

@BowenFu BowenFu left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Verified the premise rather than taking it on trust: #16163 is an exact line-for-line inverse of #15920's commit 4f1c3cd (179 changed lines each, the revert diff is the precise complement), nothing on main has re-introduced the host-side greedy-stop path, and main has exactly 30 6427411 lines — this removes all 30 and adds no new waiver. NVBugs 6427411 is already Bug - Fixed / Verify-to-Close with FixInfo pointing at #16163.

One non-blocking coverage note: 28 of the 30 are scheduled pre-merge (individually in l0_dgx_h100.yml / l0_dgx_b200.yml / l0_gb200_multi_gpus.yml, or via the -m "gpu2"/"gpu4" aggregates), so this PR's own /bot run re-exercises them. The two exceptions are post-merge only — TestNemotronV3Super::test_nvfp4_parallelism[TP4_PP2] (l0_dgx_b200.yml:247) and TestNemotronV3Ultra::test_nvfp4_parallelism[ADP2_PP2] (l0_gb200_multi_gpus.yml:85). The bug's own checklist asks for --extra-stage/--stage-list in exactly this case, so consider re-running with those two stages before merge.

FYI three of the removed entries also have a full:GB300/ or full:B300/ line under a different bug (6388153, 6322073); those are post-merge platform-scoped, so the tests correctly stay skipped there while returning to pre-merge.

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #63003 [ run ] completed with state SUCCESS. Commit: 01c40d9
/LLM/main/L0_MergeRequest_PR pipeline #51111 completed with status: 'SUCCESS'

CI Report

Link to invocation

@mingyangHao
mingyangHao force-pushed the user/mingyangh/unwaive-nvbug-6427411 branch from 01c40d9 to a702ae9 Compare August 14, 2026 09:13
Remove the remaining PP regression waivers added after NVIDIA#15920. The host-side greedy stop change was reverted by NVIDIA#16163, so these tests should run in CI again.

Signed-off-by: Mingyang Hao <200044211+mingyangHao@users.noreply.github.com>
@mingyangHao mingyangHao reopened this Aug 14, 2026
@coderabbitai

coderabbitai Bot commented Aug 14, 2026

Copy link
Copy Markdown
Contributor

Note

GitHub couldn't provide a complete incremental comparison for this pull request, so CodeRabbit is performing a full review instead. This review may take a little longer.

@mingyangHao

Copy link
Copy Markdown
Collaborator Author

/bot skip --comment 'unwaive test'

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #66242 [ skip ] triggered by Bot. Commit: 165bbb4 Link to invocation

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #66242 [ skip ] completed with state SUCCESS. Commit: 165bbb4
Skipping testing for commit 165bbb4

Link to invocation

@mingyangHao

Copy link
Copy Markdown
Collaborator Author

/bot skip --comment 'unwaive test'

@mingyangHao
mingyangHao enabled auto-merge (squash) August 14, 2026 12:47
@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #66276 [ skip ] triggered by Bot. Commit: c3d3b68 Link to invocation

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #66276 [ skip ] completed with state SUCCESS. Commit: c3d3b68
Skipping testing for commit c3d3b68

Link to invocation

@mingyangHao
mingyangHao merged commit 87aa25c into NVIDIA:main Aug 14, 2026
11 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants