Skip to content

Revert "Add flashinfer rmsnorm + quant fusion support SM90, SM100, SM120" - #33455

Merged
Fridge003 merged 1 commit into
mainfrom
revert-32994-dev/dlal/norm-quant-fusion
Aug 4, 2026
Merged

Fridge003 merged 1 commit into
mainfrom
revert-32994-dev/dlal/norm-quant-fusion

Conversation

@Fridge003

@Fridge003 Fridge003 commented Aug 4, 2026

Copy link
Copy Markdown
Collaborator

Reverts #32994

This commit breaks CI: https://github.com/sgl-project/sglang/actions/runs/30868463119/job/91865475888


CI States

Latest PR Test (Base): ⏳ Run #30870646431
Latest PR Test (Extra): ⏳ Run #30870646307

@gemini-code-assist

Copy link
Copy Markdown
Contributor

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.

@github-actions github-actions Bot added quant LLM Quantization sgl-kernel labels Aug 4, 2026
@Fridge003
Fridge003 merged commit eb31a53 into main Aug 4, 2026
56 of 85 checks passed
@DevashishLal-CB

Copy link
Copy Markdown
Contributor

@hnyls2002 this CI run https://github.com/sgl-project/sglang/actions/runs/30868463119/job/91865475888 didn't build the latest sgl-kernel, the assert that was hit should not even trigger on the latest kernel as I had modified it https://github.com/sgl-project/sglang/pull/32994/changes#diff-05fba125f0fda254854fe7de03d2e1b534b8948809fe696a7a90e84945f33317L1118

Test passed locally on my branch

➜ pytest -v ./test/registered/quant/test_modelopt_fp8.py
=============================================================================== test session starts ===============================================================================
platform linux -- Python 3.12.13, pytest-9.1.1, pluggy-1.6.0 -- /home/devcode/uv_env/sglang/bin/python3
cachedir: .pytest_cache
rootdir: /data/users/devcode/gitrepos/sglang/test
configfile: pytest.ini
plugins: anyio-4.14.2, cov-7.1.0
collected 2 items

test/registered/quant/test_modelopt_fp8.py::TestModeloptFP8::test_gsm8k PASSED                                                                                              [ 50%]
test/registered/quant/test_modelopt_fp8.py::TestModeloptFP8WithRustServer::test_gsm8k SKIPPED (embedded rust server extension not built)                                    [100%]

================================================================================ warnings summary =================================================================================
../../../../../home/devcode/uv_env/sglang/lib64/python3.12/site-packages/_pytest/config/__init__.py:1464
  /home/devcode/uv_env/sglang/lib64/python3.12/site-packages/_pytest/config/__init__.py:1464: PytestConfigWarning: Unknown config option: asyncio_mode

    self._warn_or_fail_if_strict(f"Unknown config option: {key}\n")

../../../../../home/devcode/uv_env/sglang/lib64/python3.12/site-packages/torch/jit/_script.py:365: 14 warnings
  /home/devcode/uv_env/sglang/lib64/python3.12/site-packages/torch/jit/_script.py:365: DeprecationWarning: `torch.jit.script_method` is deprecated. Please switch to `torch.compile` or `torch.export`.
    warnings.warn(

-- Docs: https://docs.pytest.org/en/stable/how-to/capture-warnings.html
============================================================== 1 passed, 1 skipped, 15 warnings in 152.75s (0:02:32) =============================================================

kfhfar pushed a commit to kfhfar/sglang that referenced this pull request Aug 7, 2026
@Fridge003
Fridge003 deleted the revert-32994-dev/dlal/norm-quant-fusion branch August 14, 2026 02:04
saturn-acc pushed a commit to saturn-acc/sglang that referenced this pull request Aug 16, 2026
jakki-amd pushed a commit to jakki-amd/sglang that referenced this pull request Sep 9, 2026
Atituiset pushed a commit to Atituiset/sglang that referenced this pull request Sep 10, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

quant LLM Quantization sgl-kernel

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants