[3/N][Attention] Move AttentionMetadata-related code from utils.py to backend.py (#32054)

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>

vllm/v1/attention/backends/gdn_attn.py

@@ -7,12 +7,14 @@ from dataclasses import dataclass
 import torch
 
 from vllm.config import VllmConfig
-from vllm.v1.attention.backend import AttentionBackend
-from vllm.v1.attention.backends.utils import (
-    PAD_SLOT_ID,
+from vllm.v1.attention.backend import (
+    AttentionBackend,
     AttentionCGSupport,
     AttentionMetadataBuilder,
     CommonAttentionMetadata,
+)
+from vllm.v1.attention.backends.utils import (
+    PAD_SLOT_ID,
     compute_causal_conv1d_metadata,
     split_decodes_and_prefills,
 )