[XPU][UT]Enable ut qk_norm_rope_fusion (#42486)

Signed-off-by: Lai, Yejing <yejing.lai@intel.com>
Co-authored-by: Kunshang Ji <kunshang.ji@intel.com>
This commit is contained in:
Yejing Lai
2026-07-01 07:38:03 +00:00
committed by GitHub
co-authored by Kunshang Ji
parent 8f82be5705
commit f651a8a9a4
2 changed files with 10 additions and 7 deletions
@@ -18,12 +18,15 @@ from torch._inductor.pattern_matcher import PatternMatcherPass, PatternPrettyPri
from vllm.config import VllmConfig
from vllm.logger import init_logger
from vllm.platforms import current_platform
from .fx_utils import is_func
from .inductor_pass import InductorPass, enable_fake_mode
logger = init_logger(__name__)
DEVICE_TYPE = current_platform.device_type
@dataclass
class InductorCompilationConfig:
@@ -227,23 +230,23 @@ class VllmPatternReplacement(ABC, Generic[P, R]):
# Helpers for get_inputs: uninitialized tensors of common dtypes.
@staticmethod
def empty(*args, **kwargs) -> torch.Tensor:
return torch.empty(*args, device="cuda", **kwargs)
return torch.empty(*args, device=DEVICE_TYPE, **kwargs)
@staticmethod
def empty_bf16(*args, **kwargs) -> torch.Tensor:
return torch.empty(*args, dtype=torch.bfloat16, device="cuda", **kwargs)
return torch.empty(*args, dtype=torch.bfloat16, device=DEVICE_TYPE, **kwargs)
@staticmethod
def empty_fp16(*args, **kwargs) -> torch.Tensor:
return torch.empty(*args, dtype=torch.float16, device="cuda", **kwargs)
return torch.empty(*args, dtype=torch.float16, device=DEVICE_TYPE, **kwargs)
@staticmethod
def empty_fp32(*args, **kwargs) -> torch.Tensor:
return torch.empty(*args, dtype=torch.float32, device="cuda", **kwargs)
return torch.empty(*args, dtype=torch.float32, device=DEVICE_TYPE, **kwargs)
@staticmethod
def empty_i32(*args, **kwargs) -> torch.Tensor:
return torch.empty(*args, dtype=torch.int32, device="cuda", **kwargs)
return torch.empty(*args, dtype=torch.int32, device=DEVICE_TYPE, **kwargs)
def _fx_view_to_reshape(gm: fx.GraphModule) -> None: