Avoid multiple redefinition (#1817)

csrc/cache.h

+#pragma once
+
 #include <torch/extension.h>
 
 #include <map>

csrc/cuda_utils.h

+#pragma once
+
 #include <torch/extension.h>
 
 int get_device_attribute(

csrc/dispatch_utils.h

@@ -2,6 +2,8 @@
  * Adapted from
  * https://github.com/pytorch/pytorch/blob/v2.0.1/aten/src/ATen/Dispatch.h
  */
+#pragma once
+
 #include <torch/extension.h>
 
 #define VLLM_DISPATCH_CASE_FLOATING_TYPES(...)              \

csrc/ops.h

+#pragma once
+
 #include <torch/extension.h>
 
 void paged_attention_v1(