[Platform][Kernel] platform-specific kernel loading (#25823)

Signed-off-by: Hank <[email protected]>

Author: ILikeIneine

vllm/_custom_ops.py

@@ -1,7 +1,6 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-import contextlib
 from typing import TYPE_CHECKING, Optional, Union
 
 import torch
@@ -13,16 +12,8 @@
 
 logger = init_logger(__name__)
 
-if not current_platform.is_tpu() and not current_platform.is_xpu():
-    try:
-        import vllm._C
-    except ImportError as e:
-        logger.warning("Failed to import from vllm._C with %r", e)
-
-supports_moe_ops = False
-with contextlib.suppress(ImportError):
-    import vllm._moe_C  # noqa: F401
-    supports_moe_ops = True
+current_platform.import_core_kernels()
+supports_moe_ops = current_platform.try_import_moe_kernels()
 
 if TYPE_CHECKING:
 

vllm/platforms/interface.py

@@ -1,5 +1,6 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
+import contextlib
 import enum
 import os
 import platform
@@ -163,6 +164,22 @@ def device_id_to_physical_device_id(cls, device_id: int):
         else:
             return device_id
 
+    @classmethod
+    def import_core_kernels(cls) -> None:
+        """ Import any platform-specific C kernels. """
+        try:
+            import vllm._C  # noqa: F401
+        except ImportError as e:
+            logger.warning("Failed to import from vllm._C: %r", e)
+
+    @classmethod
+    def try_import_moe_kernels(cls) -> bool:
+        """ Import any platform-specific MoE kernels. """
+        with contextlib.suppress(ImportError):
+            import vllm._moe_C  # noqa: F401
+            return True
+        return False
+
     @classmethod
     def get_vit_attn_backend(cls, head_size: int,
                              dtype: torch.dtype) -> "_Backend":

vllm/platforms/tpu.py

@@ -47,6 +47,10 @@ class TpuPlatform(Platform):
         "TPU_CHIPS_PER_HOST_BOUNDS", "TPU_HOST_BOUNDS"
     ]
 
+    @classmethod
+    def import_core_kernels(cls) -> None:
+        pass
+
     @classmethod
     def get_attn_backend_cls(cls, selected_backend: "_Backend", head_size: int,
                              dtype: torch.dtype, kv_cache_dtype: Optional[str],

vllm/platforms/xpu.py

@@ -34,6 +34,10 @@ class XPUPlatform(Platform):
     dist_backend: str = "ccl"  # ccl | xccl
     device_control_env_var: str = "ZE_AFFINITY_MASK"
 
+    @classmethod
+    def import_core_kernels(cls) -> None:
+        pass
+
     @classmethod
     def get_attn_backend_cls(cls, selected_backend: "_Backend", head_size: int,
                              dtype: torch.dtype, kv_cache_dtype: Optional[str],