refactor: template AttentionParams, rename .cuh to .h

- Convert AttentionParams to a template struct supporting arbitrary types
- Rename attn_common.cuh -> attn_common.h (no CUDA-specific code remains)
- Include standard headers explicitly in each .cuh instead of via attn_common.cuh
- Allow .h files in csrc/ via .gitignore

csrc/kernels/attn_entry_utils.cuh

@@ -1,7 +1,8 @@
 #pragma once
 #include <torch/extension.h>
-#include "attn_common.cuh"
+#include "attn_common.h"
 
+template<typename T>
 inline void attn_pack_params(
     torch::Tensor q,
     torch::Tensor k,
@@ -10,7 +11,7 @@ inline void attn_pack_params(
     bool is_causal,
     int64_t causal_offset,
     c10::optional<double> scale,
-    AttentionParams& p
+    AttentionParams<T>& p
 ) {
     TORCH_CHECK(q.is_cuda() && k.is_cuda() && v.is_cuda());
     TORCH_CHECK(q.dtype() == torch::kBFloat16);
@@ -27,9 +28,9 @@ inline void attn_pack_params(
     p.is_causal = is_causal ? 1 : 0;
     p.causal_offset = (int)causal_offset;
     p.scale = scale.has_value() ? (float)scale.value() : 1.0f / sqrtf((float)p.head_dim);
-    p.q = (const bf16*)q.data_ptr();
-    p.k = (const bf16*)k.data_ptr();
-    p.v = (const bf16*)v.data_ptr();
+    p.q = (const T*)q.data_ptr();
+    p.k = (const T*)k.data_ptr();
+    p.v = (const T*)v.data_ptr();
     if (p.use_mask) {
         TORCH_CHECK(mask.value().dtype() == torch::kBool);
         TORCH_CHECK(mask.value().dim() == 2);