add mc2 mask

Signed-off-by: weiguihua2 <weiguihua2@huawei.com>

Author: weiguihua2

vllm_ascend/attention/mla_v1.py

@@ -208,7 +208,7 @@ def __init__(self,
         ascend_config = get_ascend_config()
         self.torchair_graph_enabled = ascend_config.torchair_graph_config.enabled
 
-    def generate_active_mask(self, actual_seqs_num, batch_size):
+    def generate_activate_mask(self, actual_seqs_num, batch_size):
         mc2_mask = torch.zeros(batch_size,
                                dtype=torch.bool,
                                device=current_platform.device_type)
@@ -345,7 +345,7 @@ def build_torchair_graph_dummy(
         else:
             attn_state = AscendAttentionState.DecodeOnly
             num_decode_tokens = 1
-        mc2_mask = self.generate_active_mask(num_actual_tokens, num_reqs)
+        mc2_mask = self.generate_activate_mask(num_actual_tokens, num_reqs)
         decode_metadata = AscendMLADecodeMetadata(
             input_positions=input_positions,
             block_table=block_table,
@@ -511,7 +511,7 @@ def build(
                                                   num_reqs_pad_size]
             else:
                 seq_lens_list = seq_lens.tolist()
-            mc2_mask = self.generate_active_mask(num_actual_tokens,
+            mc2_mask = self.generate_activate_mask(num_actual_tokens,
                                                  num_reqs + num_reqs_pad_size)
 
             decode_metadata = AscendMLADecodeMetadata(