Skip to content

[Triton/Gluon] : Remove bad mla_decode.py op - #6633

Merged
azaidy merged 1 commit into
mainfrom
triton_rm_bad_mla_decode_op
Oct 19, 2026
Merged

azaidy merged 1 commit into
mainfrom
triton_rm_bad_mla_decode_op

Conversation

@rahulbatra85

@rahulbatra85 rahulbatra85 commented Oct 16, 2026 •

Copy link
Copy Markdown
Contributor

Motivation

  1. mla_decode.py op is not MLA specific. More of generic paged attention decode
  2. mla_decode_rope.py has the MLA decode kernel with support for rope as well
  3. Missing unit tests, benchmarks, configs and also doesn't break the op and kernels into separate files

Technical Details

Remove duplicate kernel

Test Plan

Test Result

Submission Checklist

@rahulbatra85
rahulbatra85 requested a review from a team October 16, 2026 16:38
@github-actions github-actions Bot changed the title [Triton/Gluon]: Remove bad mla_decode.py op [Triton/Gluon] : Remove bad mla_decode.py op Oct 16, 2026
@github-actions

Copy link
Copy Markdown
Contributor

🏷️ CI Guide

Runs automatically on every PR:

  • ✅ Pre-checks (submodule verification, code formatting)
  • ✅ Aiter op tests (gfx942 + gfx950)
  • ✅ Triton tests on MI35X (only when aiter/ops/triton/** or related paths are changed)

Extended tests (opt-in via labels):

Label Tests
ci:gfx1250-ffm-triton Run the five-shard gfx1250 FFM Triton test suite
ci:triton-300x Run an additional Triton test job on MI300X in PRs (added automatically when gfx942 configs change); main branch always runs both MI35X and MI300X
ci:triton-355 Run the full Triton test suite on MI35X, not only the tests the change affects
ci:windows Windows JIT op smoke test on the gfx1201 ROCm runner
ci:multi-gpu Aiter multi-GPU tests on the 8-GPU runner
ci:sglang SGLang integration tests: DeepSeek-R1-MXFP4 accuracy, Qwen 3.5 accuracy
ci:atom ATOM benchmark: DeepSeek-R1-0528, GPT-OSS-120B
ci:atom_full ATOM accuracy suite for PR and main models from ATOM models_accuracy.json
ci:vllm vLLM benchmark: DeepSeek-V4-Pro, MiniMax-M3-MXFP4
ci:all All standard extended tests (excludes ci:atom_full)

Only add ci:atom_full for FlyDSL or Triton upgrades.
Add labels via the sidebar or gh pr edit 6633 --add-label <label>

One backend per PR:
A PR changes one kernel backend: [Triton/Gluon] (Triton and Gluon count as one), [HIP], [ASM], [CK], [OPUS] or [FlyDSL]. If the title ends up with two backend tags, split the PR -- as stacked pull requests when one part cannot merge without the other.

PR title tags & labels:
Component tags ([Triton/Gluon], [HIP], [CK], [ASM], ...) are added to the PR title and as PR labels automatically from the changed files and re-synced on every push — change-type tags like [fix]/[Perf], op tags like [MLA], and human labels (ci:*) are left untouched. Add the no-auto-title label to stop the title rewrites; labels stay in sync either way.

Copilot AI 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.

🔵 Needs a closer look

A stale compatibility entry targets the deleted module and causes attribute checks to raise ModuleNotFoundError.

0 open findings

What changed in this PR

Removes the generic paged-attention decode implementation from the Triton attention package because it is not MLA-specific.

Changes:

  • Deletes the decode wrapper, split-K kernels, and CSR-to-dense helper.
  • Leaves the existing mla_decode_rope.py implementation unchanged.
File Description
aiter/​ops/​triton/​attention/​mla_decode.py Deletes the decode implementation and supporting helpers.

🧠 Review effort: Balanced


💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

@zufayu
zufayu requested review from a team and vgokhale October 18, 2026 23:31
@rahulbatra85
rahulbatra85 force-pushed the triton_rm_bad_mla_decode_op branch from 6f1817b to 85bb8b2 Compare October 19, 2026 16:46

@azaidy azaidy 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.

LGTM!

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