Skip to content

[None][chore] promote flashinfer-prims-ts source from #20224 - #20870

Merged
yuxianq merged 1 commit into
NVIDIA:mainfrom
yuxianq:chore/flashinfer-prims-ts-promote-20224-0e0761ef5ef1
Oct 23, 2026
Merged

yuxianq merged 1 commit into
NVIDIA:mainfrom
yuxianq:chore/flashinfer-prims-ts-promote-20224-0e0761ef5ef1

Conversation

@yuxianq

@yuxianq yuxianq commented Oct 23, 2026 •

Copy link
Copy Markdown
Collaborator

Dev Engineer Review

The vendor lock selects https://github.com/yuxianq/flashinfer.git and branch trtllm-prims-ts-dev-20261012 for flashinfer-prims-ts. The commit, source and destination paths, patch, and digests remain unchanged. This config change does not alter runtime behavior or public APIs. Review findings were not supplied.

QA Engineer Review

No test changes.

Per-File QA Perspective

  • 3rdparty/vendor_sources.lock.yaml: The source URL and branch selection changed. The pinned commit and vendor content digests remain unchanged, so no runtime behavior change is indicated.

Description

Promote the exact source revision reviewed in #20224.
Only the selected vendor lock URL/branch changes. The source SHA, selected vendor files, compatibility patch, and digests are unchanged.

Test Coverage

Source materialization and lock-only diff verified by the promotion tool. No new kernel behavior. The tool posts /bot skip --comment "skip CI since no code change" to request the bot's no-code-change CI path. Automatic GitHub checks and required reviews still apply.

PR Checklist

  • No runtime/API/dependency changes.
  • Immutable source pin and vendor content preserved.
  • Please check this after reviewing the above items as appropriate for this PR.

Required squash commit message

Preserve the following message, including the provenance block and DCO sign-off.

[None][chore] promote flashinfer-prims-ts source from #20224

----- BEGIN VENDOR PROMOTION V1 -----
{
  "attribution": {
    "0e0761ef5ef10cb0762cc1b890a5c1a329bb49bf": {
      "method": "upstream-aggregate",
      "upstream_base": "ae312f7cc1723b06fda5168105ade0f328ee47d8",
      "upstream_head": "1bec389b38a1534b7da21c135dde001a6abfd123"
    },
    "6b00497398f491136be91fb0e60ac3f314da53ba": {
      "method": "same-edits",
      "upstream_base": "2895577720741ddfedff118202c89584996b8875",
      "upstream_head": "14770b419ec3261c2afc55e401051c92bb7f3fd9"
    },
    "ae3851ab5f958ba22f154059df5b38321f6369d2": {
      "comparison": {
        "source_base": "6b00497398f491136be91fb0e60ac3f314da53ba",
        "source_head": "ae3851ab5f958ba22f154059df5b38321f6369d2",
        "upstream_base": "873d21fd565adfa98748c53f90fd4f8aa4d5ed60",
        "upstream_head": "a9883a5e8ada0d25754ecb31ae7eb83109c403cc"
      },
      "method": "similar-edits",
      "refresh_policy": "retain-until-verified",
      "similarity": 0.9859154929577465,
      "similarity_metric": "sequence-matcher-lines-v1",
      "similarity_threshold": 0.9,
      "upstream_base": "2895577720741ddfedff118202c89584996b8875",
      "upstream_head": "14770b419ec3261c2afc55e401051c92bb7f3fd9",
      "upstream_pr": 6482
    }
  },
  "attribution_scope": {
    "include": [
      "**/*.py"
    ],
    "source": "flashinfer/attention/prims_ts"
  },
  "canonical_branch": "trtllm-prims-ts-dev-20261012",
  "canonical_url": "https://github.com/yuxianq/flashinfer.git",
  "changes": [
    {
      "commits": [
        "0e0761ef5ef10cb0762cc1b890a5c1a329bb49bf"
      ],
      "upstream_base": "ae312f7cc1723b06fda5168105ade0f328ee47d8",
      "upstream_head": "1bec389b38a1534b7da21c135dde001a6abfd123",
      "upstream_merge": "388e7a311e9f79941134dcc298cbc942b5c607a1",
      "upstream_pr": "https://github.com/flashinfer-ai/flashinfer/pull/6014"
    },
    {
      "commits": [
        "6b00497398f491136be91fb0e60ac3f314da53ba",
        "ae3851ab5f958ba22f154059df5b38321f6369d2"
      ],
      "upstream_base": "2895577720741ddfedff118202c89584996b8875",
      "upstream_head": "14770b419ec3261c2afc55e401051c92bb7f3fd9",
      "upstream_merge": "f57c58b898cc0f96ddea03066a3fd66ad201337a",
      "upstream_pr": "https://github.com/flashinfer-ai/flashinfer/pull/6482"
    }
  ],
  "consumer_base": "main",
  "consumer_merge": "791a402fc8d514ba73ece0fe37af06036b08a18e",
  "consumer_pr": "https://github.com/NVIDIA/TensorRT-LLM/pull/20224",
  "contributor_metadata": {
    "metadata_digest": "ca3752e66016a893957f16ade23d2cf69bf0b75d087817f241402f29d47a5d6c",
    "pr_author": "heyuhhh",
    "source_head": "08d8cfa731414f27a20c31bbe9967f254a711cd3"
  },
  "kind": "promotion",
  "previous": "ae312f7cc1723b06fda5168105ade0f328ee47d8",
  "promoted": "0e0761ef5ef10cb0762cc1b890a5c1a329bb49bf",
  "source_branch": "trtllm-prims-ts-dev",
  "source_url": "https://github.com/heyuhhh/flashinfer.git",
  "upstream_base": "ae312f7cc1723b06fda5168105ade0f328ee47d8",
  "upstream_branch": "main",
  "upstream_observed_main": "13d7fcf4041635f5b7f573bb0dbab23570ae1034",
  "upstream_repository": "flashinfer-ai/flashinfer",
  "vendor": "flashinfer-prims-ts",
  "version": 1
}
----- END VENDOR PROMOTION V1 -----

Signed-off-by: Yuxian Qiu <142763828+yuxianq@users.noreply.github.com>

----- BEGIN VENDOR PROMOTION V1 -----
{
  "attribution": {
    "0e0761ef5ef10cb0762cc1b890a5c1a329bb49bf": {
      "method": "upstream-aggregate",
      "upstream_base": "ae312f7cc1723b06fda5168105ade0f328ee47d8",
      "upstream_head": "1bec389b38a1534b7da21c135dde001a6abfd123"
    },
    "6b00497398f491136be91fb0e60ac3f314da53ba": {
      "method": "same-edits",
      "upstream_base": "2895577720741ddfedff118202c89584996b8875",
      "upstream_head": "14770b419ec3261c2afc55e401051c92bb7f3fd9"
    },
    "ae3851ab5f958ba22f154059df5b38321f6369d2": {
      "comparison": {
        "source_base": "6b00497398f491136be91fb0e60ac3f314da53ba",
        "source_head": "ae3851ab5f958ba22f154059df5b38321f6369d2",
        "upstream_base": "873d21fd565adfa98748c53f90fd4f8aa4d5ed60",
        "upstream_head": "a9883a5e8ada0d25754ecb31ae7eb83109c403cc"
      },
      "method": "similar-edits",
      "refresh_policy": "retain-until-verified",
      "similarity": 0.9859154929577465,
      "similarity_metric": "sequence-matcher-lines-v1",
      "similarity_threshold": 0.9,
      "upstream_base": "2895577720741ddfedff118202c89584996b8875",
      "upstream_head": "14770b419ec3261c2afc55e401051c92bb7f3fd9",
      "upstream_pr": 6482
    }
  },
  "attribution_scope": {
    "include": [
      "**/*.py"
    ],
    "source": "flashinfer/attention/prims_ts"
  },
  "canonical_branch": "trtllm-prims-ts-dev-20261012",
  "canonical_url": "https://github.com/yuxianq/flashinfer.git",
  "changes": [
    {
      "commits": [
        "0e0761ef5ef10cb0762cc1b890a5c1a329bb49bf"
      ],
      "upstream_base": "ae312f7cc1723b06fda5168105ade0f328ee47d8",
      "upstream_head": "1bec389b38a1534b7da21c135dde001a6abfd123",
      "upstream_merge": "388e7a311e9f79941134dcc298cbc942b5c607a1",
      "upstream_pr": "flashinfer-ai/flashinfer#6014"
    },
    {
      "commits": [
        "6b00497398f491136be91fb0e60ac3f314da53ba",
        "ae3851ab5f958ba22f154059df5b38321f6369d2"
      ],
      "upstream_base": "2895577720741ddfedff118202c89584996b8875",
      "upstream_head": "14770b419ec3261c2afc55e401051c92bb7f3fd9",
      "upstream_merge": "f57c58b898cc0f96ddea03066a3fd66ad201337a",
      "upstream_pr": "flashinfer-ai/flashinfer#6482"
    }
  ],
  "consumer_base": "main",
  "consumer_merge": "791a402fc8d514ba73ece0fe37af06036b08a18e",
  "consumer_pr": "NVIDIA#20224",
  "contributor_metadata": {
    "metadata_digest": "ca3752e66016a893957f16ade23d2cf69bf0b75d087817f241402f29d47a5d6c",
    "pr_author": "heyuhhh",
    "source_head": "08d8cfa731414f27a20c31bbe9967f254a711cd3"
  },
  "kind": "promotion",
  "previous": "ae312f7cc1723b06fda5168105ade0f328ee47d8",
  "promoted": "0e0761ef5ef10cb0762cc1b890a5c1a329bb49bf",
  "source_branch": "trtllm-prims-ts-dev",
  "source_url": "https://github.com/heyuhhh/flashinfer.git",
  "upstream_base": "ae312f7cc1723b06fda5168105ade0f328ee47d8",
  "upstream_branch": "main",
  "upstream_observed_main": "13d7fcf4041635f5b7f573bb0dbab23570ae1034",
  "upstream_repository": "flashinfer-ai/flashinfer",
  "vendor": "flashinfer-prims-ts",
  "version": 1
}
----- END VENDOR PROMOTION V1 -----

Signed-off-by: Yuxian Qiu <142763828+yuxianq@users.noreply.github.com>
@yuxianq
yuxianq requested review from a team as code owners October 23, 2026 05:34
@yuxianq

yuxianq commented Oct 23, 2026

Copy link
Copy Markdown
Collaborator Author

/bot skip --comment "skip CI since no code change"

@yuxianq
yuxianq enabled auto-merge (squash) October 23, 2026 05:34
@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #80069 [ skip ] triggered by Bot. Commit: 72491ba Link to invocation

@coderabbitai

coderabbitai Bot commented Oct 23, 2026

Copy link
Copy Markdown
Contributor

Review in Change Stack →

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info
⚙️ Run configuration
📥 Commits

Reviewing files that changed from the base of the PR and between 791a402 and 72491ba.

📒 Files selected for processing: 1
  • 3rdparty/vendor_sources.lock.yaml

Included review availability: This review used your included allowance. Your plan provides up to 12 included reviews per hour; 11 remain after this review.

Walkthrough

The flashinfer-prims-ts vendor source now points to a different repository URL and branch. Its pinned commit and remaining vendor configuration are unchanged.

Changes

FlashInfer source configuration

Layer / File(s) Summary
Update vendor source reference
3rdparty/vendor_sources.lock.yaml
The source URL changed to yuxianq/flashinfer, and the branch changed to trtllm-prims-ts-dev-20261012. The pinned commit is unchanged.

Priority: ⬇️ Low

Estimated code review effort: 1 (Trivial) | ~3 minutes

Change: Other

Possibly related PRs

Suggested reviewers: heyuhhh

Merge Risk: ⚪ Minimal · up to 72491

The vendor reference points to the retained source revision, with no identified issue blocking merge after normal checks.

Pre-merge checks: Passed 5
Check name Status Explanation
Title check Passed The title follows the repository format and clearly describes promoting the flashinfer-prims-ts source from PR #20224.
Description check Passed The description includes the required Description, Test Coverage, and PR Checklist sections. It explains the promotion, verification, and lack of runtime or API changes, and includes the required squa…
Docstring Coverage Passed No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 0…
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.
Full details: Description check

Explanation

The description includes the required Description, Test Coverage, and PR Checklist sections. It explains the promotion, verification, and lack of runtime or API changes, and includes the required squash commit message and provenance.

Full details: Docstring Coverage

Explanation

No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 0 files. (1 skipped: 1 unsupported.)

✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create a new PR
  • Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

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

@yuxianq

yuxianq commented Oct 23, 2026 •

Copy link
Copy Markdown
Collaborator Author

@yihwang-nv, could you help to review this PR as trt-llm-torch-attention-devs? Thanks~

Assignment explanation
Componentsintegration/dependencies (100%)
TagsNone identified
PR effort1 points — A two-field lockfile update requires a small provenance check, without attention algorithm or kernel reasoning.
Affinity (0–1)0.525 component/tag match
Outstanding points at selection41 → 42 (+1)
Points already reserved on this PR0
Effective weight1
Completed review points (preceding 168 hours)18
Combined points after assignment60
Demand-derived target59.00 points (soft share)
Additional fair share before assignment0.00 points
Share explanationWhole PR / qualification or coverage constraint exceeds a soft share; assignment is allowed.
Selection rulebalance-first maintenance
Assignment scope (yihwang-nv)all

Affinity and workload are snapshots at selection, not live counters. Existing reservations on this PR are counted once.

@yuxianq
yuxianq requested review from yihwang-nv and removed request for heyuhhh and yunruis October 23, 2026 05:44
@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #80069 [ skip ] completed with state SUCCESS. Commit: 72491ba
Skipping testing for commit 72491ba

Link to invocation

@yuxianq
yuxianq merged commit 12af626 into NVIDIA:main Oct 23, 2026
15 checks passed
@github-actions

Copy link
Copy Markdown

Automatically added "ci: full pre-merge approved" because this PR has satisfied the required GitHub review approvals. Unresolved review conversations and other required checks remain independent merge requirements.

yunruis added a commit to yunruis/TensorRT-LLM that referenced this pull request Oct 23, 2026
…ted canonical tip

TensorRT-LLM main (NVIDIA#20224, promoted by NVIDIA#20870) moved the flashinfer-prims-ts
canonical pin from ae3851ab to 0e0761ef (the NVIDIA#6014 SM103 block-sparse/sage
tuning). The linear source branch is rebased onto that tip as
yunruis/flashinfer trtllm-prims-ts-dev-20261023 @ aa709061: the duplicate
NVIDIA#6014 cherry-pick drops out and the remaining five flashinfer main PrimTS
commits (NVIDIA#6089, NVIDIA#5351, NVIDIA#6482, NVIDIA#6529, NVIDIA#5434) replay without conflicts. The
vendored tree, patch and digests are byte-identical to the previous pin
(f924fd45); only the lock's branch and commit change.

AI-assisted.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Signed-off-by: yunruis <205571022+yunruis@users.noreply.github.com>
yunruis added a commit to yunruis/TensorRT-LLM that referenced this pull request Oct 24, 2026
…ted canonical tip

TensorRT-LLM main (NVIDIA#20224, promoted by NVIDIA#20870) moved the flashinfer-prims-ts
canonical pin from ae3851ab to 0e0761ef (the NVIDIA#6014 SM103 block-sparse/sage
tuning). The linear source branch is rebased onto that tip as
yunruis/flashinfer trtllm-prims-ts-dev-20261023 @ aa709061: the duplicate
NVIDIA#6014 cherry-pick drops out and the remaining five flashinfer main PrimTS
commits (NVIDIA#6089, NVIDIA#5351, NVIDIA#6482, NVIDIA#6529, NVIDIA#5434) replay without conflicts. The
vendored tree, patch and digests are byte-identical to the previous pin
(f924fd45); only the lock's branch and commit change.

AI-assisted.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Signed-off-by: yunruis <205571022+yunruis@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants