Skip to content

[XPU] Fix XPU CI breaks from #41105 (KDA chunk_offsets) and #42588 (int4 test tp_group) #43334

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
mingfeima merged 3 commits into from
Oct 10, 2026
+12 −11
Merged

[XPU] Fix XPU CI breaks from #41105 (KDA chunk_offsets) and #42588 (int4 test tp_group) #43334

Changes from 2 commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
52a762c
[XPU] Accept chunk_offsets in the XPU chunk_gated_delta_rule_fwd_h
arathi-hlab Oct 9, 2026
4ba4d49
Merge branch 'main' into fix/xpu-kda-chunk-offsets
arathi-hlab Oct 9, 2026
c6962d8
[XPU] Fix int4 dense test to set TP size on the layer's tp_group
arathi-hlab Oct 9, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 6 additions & 5 deletions python/sglang/srt/hardware_backend/xpu/kernels/fla/chunk_delta_h.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -271,6 +271,7 @@ def chunk_gated_delta_rule_fwd_h(
inplace_update: bool = True,
track_state: Optional[torch.Tensor] = None,
track_chunk_idx: Optional[torch.Tensor] = None,
chunk_offsets: Optional[torch.Tensor] = None,
) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
if not inplace_update:
raise NotImplementedError(
Expand Down Expand Up @@ -298,11 +299,11 @@ def chunk_gated_delta_rule_fwd_h(
if cu_seqlens is None:
N, NT, chunk_offsets = B, triton.cdiv(T, BT), None
else:
N, NT, chunk_offsets = (
len(cu_seqlens) - 1,
len(chunk_indices),
prepare_chunk_offsets(cu_seqlens, BT),
)
# A caller-built chunk_offsets (CUDA-graph capture) must not go through
# prepare_chunk_offsets: its tensor_cache would replay capture-time values.
N, NT = len(cu_seqlens) - 1, len(chunk_indices)
if chunk_offsets is None:
chunk_offsets = prepare_chunk_offsets(cu_seqlens, BT)
assert K <= 256, "current kernel does not support head dimension larger than 256."

h = k.new_empty(B, NT, H, V, K)
Expand Down
Loading