adopt rope in vllm-ascend (#530)

### What this PR does / why we need it?
Adopt custom kernel rotary embedding in actual model inference,
customized rotary_embedding will generate contiguous query and key in
the cpp side to reduce the overhead of two contiguous and index_select
compared with rotary_embedding in torch_npu. For now, rotary_embedding
can only support the scenario of `is_neox = true`, non-neox version rope
will be updated soon in the future.
---------

Signed-off-by: ganyi <pleaplusone.gy@gmail.com>

vllm_ascend/platform.py

@@ -23,7 +23,9 @@ import torch
 import torch_npu  # noqa: F401
 import vllm.envs as envs
 from vllm.logger import logger
+from vllm.platforms import Platform, PlatformEnum
 
+CUSTOM_OP_ENABLED = False
 try:
     # register custom ops into torch_library here
     import vllm_ascend.vllm_ascend_C  # type: ignore  # noqa: F401
@@ -35,8 +37,8 @@ except ImportError as e:
         logging.warning(
             "Warning: Failed to register custom ops, all custom ops will be disabled"
         )
-
-from vllm.platforms import Platform, PlatformEnum
+    else:
+        CUSTOM_OP_ENABLED = True
 
 if TYPE_CHECKING:
     from vllm.config import ModelConfig, VllmConfig