From 3a8708f60a8f0456818b786adf68dce4b1e5da38 Mon Sep 17 00:00:00 2001
From: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Date: Tue, 19 Aug 2025 18:17:08 -0400
Subject: [PATCH] [BugFix] fix CUTLASS MLA full cudagraph  (#23200)

Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: simon-mo <simon.mo@hey.com>
---
 vllm/v1/attention/backends/mla/cutlass_mla.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/vllm/v1/attention/backends/mla/cutlass_mla.py b/vllm/v1/attention/backends/mla/cutlass_mla.py
index 6e1e5d6533dab..024f38201970b 100644
--- a/vllm/v1/attention/backends/mla/cutlass_mla.py
+++ b/vllm/v1/attention/backends/mla/cutlass_mla.py
@@ -21,7 +21,7 @@ logger = init_logger(__name__)
 
 class CutlassMLAMetadataBuilder(MLACommonMetadataBuilder[MLACommonMetadata]):
     # enable full CUDA Graph support for decode-only capture
-    attn_cudagraph_support: ClassVar[
+    cudagraph_support: ClassVar[
         AttentionCGSupport] = AttentionCGSupport.UNIFORM_SINGLE_TOKEN_DECODE