CUDA: General GEMV fusion (ggml-org#16715)

Author: am17an

ggml/src/ggml-cuda/common.cuh

@@ -1005,3 +1005,16 @@ struct ggml_backend_cuda_context {
         return pool(device);
     }
 };
+
+struct ggml_cuda_mm_fusion_args_host {
+    const ggml_tensor * x_bias = nullptr;
+    const ggml_tensor * gate = nullptr;
+    const ggml_tensor * gate_bias = nullptr;
+    ggml_glu_op glu_op;
+};
+struct ggml_cuda_mm_fusion_args_device {
+    const void * x_bias = nullptr;
+    const void * gate = nullptr;
+    const void * gate_bias = nullptr;
+    ggml_glu_op glu_op;
+};

ggml/src/ggml-cuda/convert.cuh

@@ -1,3 +1,4 @@
+#pragma once
 #include "common.cuh"
 
 #define CUDA_DEQUANTIZE_BLOCK_SIZE 256