diff --git a/vllm_ascend/attention/mla_v1.py b/vllm_ascend/attention/mla_v1.py
index 4044126..c0e175c 100644
--- a/vllm_ascend/attention/mla_v1.py
+++ b/vllm_ascend/attention/mla_v1.py
@@ -1166,6 +1166,8 @@ class AscendMLAImpl(MLAAttentionImpl):
                 dim=-1,
             )
             q_c = self.q_a_layernorm(q_c)
+            # allgather need contiguous data
+            kv_no_split = kv_no_split.contiguous()
         else:
             q_c = hidden_states
             kv_no_split = self.kv_a_proj_with_mqa(hidden_states)[0]