Skip to content

test(xpu): skip cases blocked by missing paged decode kernel #15406

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
rmccorm4 merged 5 commits into from
Oct 14, 2026
Merged

test(xpu): skip cases blocked by missing paged decode kernel #15406

Show file tree
Hide file tree
Changes from 3 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
f932e78
test(xpu): skip cases affected by missing paged decode kernel
VincyZhang Sep 30, 2026
c5c8fba
test(xpu): skip Qwen3-VL router with missing paged decode kernel
VincyZhang Sep 30, 2026
ede3f6f
test(xpu): track skipped multimodal coverage in #15417
VincyZhang Sep 30, 2026
595484c
Merge remote-tracking branch 'upstream/main' into skip_paged_decode_k…
VincyZhang Oct 13, 2026
c66a26d
test(xpu): restore vLLM Omni audio coverage
VincyZhang Oct 13, 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
2 changes: 1 addition & 1 deletion tests/mm_router/test_vllm_mm_router_e2e.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -45,8 +45,8 @@
pytest.mark.e2e,
pytest.mark.vllm,
pytest.mark.multimodal,
# Restore xpu_1 after the paged decode kernel is available; see #15417.
pytest.mark.gpu_1,
pytest.mark.xpu_1,
pytest.mark.model(VLLM_MM_MODEL),
]

Expand Down
46 changes: 40 additions & 6 deletions tests/serve/multimodal_profiles/vllm_xpu.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -61,7 +61,13 @@
gpu_marker="xpu_1",
topologies={
"agg": TopologyConfig(
marks=[pytest.mark.post_merge, pytest.mark.xpu_1],
marks=[
Comment thread
VincyZhang marked this conversation as resolved.
Resolved
pytest.mark.skip(
Comment thread
VincyZhang marked this conversation as resolved.
Resolved
Comment thread
VincyZhang marked this conversation as resolved.
Resolved
reason="vLLM XPU v0.29.0 lacks paged decode kernel 8,128,16,false,false,false; see #15417"
),
pytest.mark.post_merge,
pytest.mark.xpu_1,
],
# TODO: re-enable XPU-parallel scheduling with
# profiled_vram_gib=9.6 once this has a bounded --kv-bytes profile.
timeout_s=300,
Expand Down Expand Up @@ -116,7 +122,13 @@
# (post_merge).
"agg_router": TopologyConfig(
health_check_workers=True,
marks=[pytest.mark.pre_merge, pytest.mark.xpu_2],
marks=[
pytest.mark.skip(
reason="vLLM XPU v0.29.0 lacks paged decode kernel 8,128,16,false,false,false; see #15417"
),
pytest.mark.pre_merge,
pytest.mark.xpu_2,
],
gpu_marker="xpu_2",
timeout_s=400,
profiled_vram_gib=18.7,
Expand All @@ -135,7 +147,13 @@
# single-GPU CI environment.
"agg_router_chat_processor": TopologyConfig(
health_check_workers=True,
marks=[pytest.mark.post_merge, pytest.mark.xpu_2],
marks=[
pytest.mark.skip(
Comment thread
VincyZhang marked this conversation as resolved.
Resolved
reason="vLLM XPU v0.29.0 lacks paged decode kernel 8,128,16,false,false,false; see #15417"
),
pytest.mark.post_merge,
pytest.mark.xpu_2,
],
gpu_marker="xpu_2",
timeout_s=400,
profiled_vram_gib=18.7,
Expand All @@ -153,7 +171,13 @@
# CI; the content-hash correctness assertion lives in
# tests/mm_router/test_router_rust_mm_frontend_decode_e2e.py.
"agg_router_frontend_decode": TopologyConfig(
marks=[pytest.mark.post_merge, pytest.mark.xpu_2],
marks=[
pytest.mark.skip(
Comment thread
VincyZhang marked this conversation as resolved.
Resolved
reason="vLLM XPU v0.29.0 lacks paged decode kernel 8,128,16,false,false,false; see #15417"
),
pytest.mark.post_merge,
pytest.mark.xpu_2,
],
gpu_marker="xpu_2",
timeout_s=400,
profiled_vram_gib=18.7,
Expand All @@ -180,7 +204,12 @@
topologies={
"agg_router": TopologyConfig(
health_check_workers=True,
marks=[pytest.mark.post_merge],
marks=[
pytest.mark.skip(
Comment thread
VincyZhang marked this conversation as resolved.
Resolved
reason="vLLM XPU v0.29.0 lacks paged decode kernel 8,128,16,false,false,false; see #15417"
),
pytest.mark.post_merge,
],
timeout_s=500,
gpu_marker="xpu_2",
profiled_vram_gib=19.0,
Expand All @@ -196,7 +225,12 @@
topologies={
"agg_router": TopologyConfig(
health_check_workers=True,
marks=[pytest.mark.post_merge],
marks=[
pytest.mark.skip(
Comment thread
VincyZhang marked this conversation as resolved.
Resolved
reason="vLLM XPU v0.29.0 lacks paged decode kernel 8,128,16,false,false,false; see #15417"
),
pytest.mark.post_merge,
],
timeout_s=500,
gpu_marker="xpu_2",
profiled_vram_gib=16.0,
Expand Down
2 changes: 1 addition & 1 deletion tests/serve/test_vllm_omni.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -176,8 +176,8 @@ class VLLMOmniConfig(EngineConfig):
directory=vllm_dir,
script_name="agg_omni_audio.sh",
marks=[
# Restore XPU with a two-card case after startup is fixed; see #15417.
pytest.mark.gpu_1,
pytest.mark.xpu_1,
Comment thread
VincyZhang marked this conversation as resolved.
Resolved
pytest.mark.pre_merge,
pytest.mark.timeout(1200),
],
Expand Down
Loading