[Bugfix] Add missing sink tensor into flash attn cascade attn implementation (#26325)

vllm/v1/attention/backends/flash_attn.py

@@ -607,6 +607,7 @@ class FlashAttentionImpl(AttentionImpl):
             q_descale=layer._q_scale,
             k_descale=layer._k_scale,
             v_descale=layer._v_scale,
+            s_aux=self.sinks,
         )
         return output
 
@@ -767,6 +768,7 @@ def cascade_attention(
     q_descale: Optional[torch.Tensor] = None,
     k_descale: Optional[torch.Tensor] = None,
     v_descale: Optional[torch.Tensor] = None,
+    s_aux: Optional[torch.Tensor] = None,
 ) -> torch.Tensor:
     assert alibi_slopes is None, "Cascade attention does not support ALiBi."
     # TODO: Support sliding window.
@@ -801,6 +803,9 @@ def cascade_attention(
         q_descale=q_descale.expand(descale_shape) if q_descale is not None else None,
         k_descale=k_descale.expand(descale_shape) if k_descale is not None else None,
         v_descale=v_descale.expand(descale_shape) if v_descale is not None else None,
+        # s_aux is incorporated into prefix_lse inside the GPU kernel,
+        # enabling its effect during the final attention merge.
+        s_aux=s_aux,
     )
 
     descale_shape = (cu_query_lens.shape[0] - 1, key_cache.shape[-2])