Add Custom Kernels For LoRA Performance (#1884)

### What this PR does / why we need it?
Add two custom kernels(bgmv_shrink and bgmv expand) to solve the
performance of LoRA
### Does this PR introduce _any_ user-facing change?
no user-facing change
### How was this patch tested?
we add Unit Test file to test the custom ascendc kernel. See
vllm-ascend/tests/e2e/singlecard/ops/test_bgmv_expand.py and
vllm-ascend/tests/e2e/singlecard/ops/test_bgmv_expand.py
Based on the actual test of the QWen2.5 7B model using vllm-ascend
version v0.9.2.rc1, the TTFT, TPOT and throughput have increased by
about 70%.

- vLLM version: v0.9.2
- vLLM main:
40d86ee412

---------

Signed-off-by: taoxudonghaha <justsheldon@163.com>

csrc/ops.h

@@ -60,4 +60,32 @@ namespace vllm_ascend {
     auto new_tensor = at_npu::native::from_blob(data_ptr, sizes, strides, options);
     return new_tensor;
   }
+
+  extern void bgmv_shrink_impl(
+        AscendType type,
+        void *stream,
+        void *x,
+        void *weight,
+        void *indices,
+        void *y, 
+        uint32_t batch_size,
+        uint32_t num_tokens_per_core,
+        uint32_t input_hidden_dim,
+        uint32_t lora_rank,
+        float scale);
+
+    extern void bgmv_expand_impl(
+        AscendType type,
+        void *stream,
+        void *x,
+        void *weight,
+        void *indices,
+        void *y,
+        void *y_out,
+        uint32_t batch_size,
+        uint32_t num_tokens_per_core,
+        uint32_t lora_rank,
+        uint32_t output_hidden_dim,
+        uint32_t slice_offset,
+        uint32_t output_full_dim);
 }