feat: paged decode attention with split-KV (scalar + MMA)

- PagedAttentionParams merged into attn_common.h
- Scalar variant: warp-per-query-head split-KV, resolves page table per-position for the K/V shared-memory tile load
- MMA variant (sm_80+): tensor-core head-packing with cp.async, single page-table lookup per tile (BC=32 fits within page_size>=32)
- Standalone test: 14 cases across head_dim 32/64/128/256, GQA, multi-batch, both paths verified against CPU reference

csrc/kernels/attn_common.h

@@ -14,11 +14,39 @@ struct AttentionParams {
     int causal_offset;
     int num_splits;
     float scale;
-    
+
     const T* __restrict__ q;
     const T* __restrict__ k;
     const T* __restrict__ v;
     const bool* __restrict__ mask;
+
+    T* __restrict__ o;
+    AT* __restrict__ o_part;
+    AT* __restrict__ ml_part;
+};
+
+template<typename T, typename AT = float>
+struct PagedAttentionParams {
+    int batch;
+    int q_head;
+    int kv_head;
+    int q_len;
+    int kv_len;
+    int head_dim;
+    int use_mask;
+    int is_causal;
+    int causal_offset;
+    float scale;
+
+    int num_splits;
+    int page_size;
+    int max_pages;
+
+    const T* __restrict__ q;
+    const T* __restrict__ k_cache;
+    const T* __restrict__ v_cache;
+    const bool* __restrict__ mask;
+    const int64_t* __restrict__ page_table;
     
     T* __restrict__ o;
     AT* __restrict__ o_part;