cuda : add FILL op support (#17851)

* cuda : add FILL op support

* cuda : add missing FILL op files

ggml/src/ggml-cuda/ggml-cuda.cu

@@ -56,6 +56,7 @@
 #include "ggml-cuda/solve_tri.cuh"
 #include "ggml-cuda/tri.cuh"
 #include "ggml-cuda/cumsum.cuh"
+#include "ggml-cuda/fill.cuh"
 #include "ggml.h"
 
 #include <algorithm>
@@ -2730,6 +2731,9 @@ static bool ggml_cuda_compute_forward(ggml_backend_cuda_context & ctx, struct gg
         case GGML_OP_SOLVE_TRI:
             ggml_cuda_op_solve_tri(ctx, dst);
             break;
+        case GGML_OP_FILL:
+            ggml_cuda_op_fill(ctx, dst);
+            break;
         default:
             return false;
     }
@@ -4617,6 +4621,7 @@ static bool ggml_backend_cuda_device_supports_op(ggml_backend_dev_t dev, const g
         case GGML_OP_CROSS_ENTROPY_LOSS_BACK:
         case GGML_OP_OPT_STEP_ADAMW:
         case GGML_OP_OPT_STEP_SGD:
+        case GGML_OP_FILL:
         case GGML_OP_CUMSUM:
         case GGML_OP_TRI:
             return true;