[1/N][Attention] Restructure attention: move files (#31916)

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

tests/kernels/attention/test_flashmla.py

@@ -7,12 +7,12 @@ import random
 import pytest
 import torch
 
-from vllm.attention.ops.flashmla import (
+from vllm.triton_utils import triton
+from vllm.v1.attention.ops.flashmla import (
     flash_mla_with_kvcache,
     get_mla_metadata,
     is_flashmla_dense_supported,
 )
-from vllm.triton_utils import triton
 
 
 def cal_diff(