[ROCm][Perf] Enable `sparse_mla`'s cudagraph on ROCm platform (#35719)

Signed-off-by: ganyi <ygan@amd.com>

vllm/v1/attention/backends/mla/rocm_aiter_mla_sparse.py

@@ -151,7 +151,9 @@ class ROCMAiterMLASparseMetadata(AttentionMetadata):
 class ROCMAiterMLASparseMetadataBuilder(
     AttentionMetadataBuilder[ROCMAiterMLASparseMetadata]
 ):
-    _cudagraph_support: ClassVar[AttentionCGSupport] = AttentionCGSupport.NEVER
+    _cudagraph_support: ClassVar[AttentionCGSupport] = (
+        AttentionCGSupport.UNIFORM_SINGLE_TOKEN_DECODE
+    )
 
     def __init__(
         self,

vllm/v1/attention/ops/rocm_aiter_mla_sparse.py

@@ -327,9 +327,6 @@ def rocm_fp8_paged_mqa_logits(
     aiter_paged_mqa_logits_module = None
     if rocm_aiter_ops.is_enabled():
         aiter_paged_mqa_logits_module = paged_mqa_logits_module()
-    # FIXME(ganyi): Temporarily disable the aiter path until nightly docker
-    # update aiter to the fix PR.
-    aiter_paged_mqa_logits_module = None
 
     if aiter_paged_mqa_logits_module is not None:
         deepgemm_fp8_paged_mqa_logits_stage1 = (