Skip to content

[GDN] remove XPU path of causal_conv1d_fn and causal_conv1d_update - #35775

Merged
mingfeima merged 1 commit into
sgl-project:mainfrom
jiayisunx:jiayi/fix_bug
Aug 24, 2026
Merged

mingfeima merged 1 commit into
sgl-project:mainfrom
jiayisunx:jiayi/fix_bug

Conversation

@jiayisunx

@jiayisunx jiayisunx commented Aug 21, 2026

Copy link
Copy Markdown
Contributor

Motivation

Fix #34720.
There will be two backend for linear attention on XPU: triton and intel_xpu. Our plan:
For the triton backend, we always use triton kernels. That means we need to remove the standalong SYCL conv1d kernel from SGLang.
For the intel_xpu backend, we will use the fused SYCL GDN kernel (#33354)

Modifications

Remove the SYCL conv1d kernels and fallback to triton kernels.

Accuracy Tests

Speed Tests and Profiling

Checklist

Review and Merge Process

  1. Ping Merge Oncalls to start the process. See the PR Merge Process.
  2. Get approvals from CODEOWNERS and other reviewers.
  3. Trigger CI tests with comments or contact authorized users to do so.
    • Common commands include /tag-and-rerun-ci, /tag-run-ci-label, /rerun-failed-ci
  4. After green CI and required approvals, ask Merge Oncalls or people with Write permission to merge the PR.

CI States

Latest PR Test (Base): ✅ Run #32438029441
Latest PR Test (Extra): ✅ Run #32703340685
Latest PR Test (AMD ROCm 7.2): ❌ Run #32438029391

@jiayisunx

Copy link
Copy Markdown
Contributor Author

/tag-and-rerun-ci

@mingfeima
mingfeima merged commit 5c46223 into sgl-project:main Aug 24, 2026
274 of 317 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

intel run-ci xpu intel gpu with device `torch.xpu`

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[Bug] [XPU] Qwen3.5 GDN + speculative decode: causal_conv1d_update_xpu() got an unexpected keyword argument 'intermediate_conv_window'

2 participants