[Bugfix][DP] Add with_prefill_across_dp to AscendMetadata to fix dp (#1094)

### What this PR does / why we need it?
Add `with_prefill_across_dp` to AscendMetadata to fix dp

This pr fixes the bug introduced by #1012, which add an arg
`with_prefill_across_dp` when dp_size > 1.

Signed-off-by: MengqingCao <cmq0113@163.com>

Author: MengqingCao

vllm_ascend/attention/attention_v1.py

@@ -132,6 +132,8 @@ class AscendMetadata:
     # For logging.
     num_input_tokens: int = 0  # Number of tokens including padding.
 
+    with_prefill_across_dp: bool = False
+
 
 class AscendAttentionMetadataBuilder:
 
@@ -142,8 +144,12 @@ def reorder_batch(self, input_batch: "InputBatch",
                       scheduler_output: "SchedulerOutput") -> bool:
         return False
 
-    def build(self, num_reqs, num_actual_tokens, max_query_len,
-              common_prefix_len):
+    def build(self,
+              num_reqs,
+              num_actual_tokens,
+              max_query_len,
+              common_prefix_len,
+              with_prefill_across_dp: bool = False):
 
         block_table = self.runner.input_batch.block_table[0].get_device_tensor(
         )
@@ -160,15 +166,17 @@ def build(self, num_reqs, num_actual_tokens, max_query_len,
         query_start_loc = query_start_loc_cpu.to(self.runner.device,
                                                  non_blocking=True)
 
-        attn_metadata = AscendMetadata(num_actual_tokens=num_actual_tokens,
-                                       block_tables=block_table,
-                                       query_start_loc=query_start_loc,
-                                       query_lens=query_lens,
-                                       seq_lens=seq_lens,
-                                       max_query_len=max_query_len,
-                                       slot_mapping=slot_mapping,
-                                       attn_mask=attn_mask,
-                                       attn_state=attn_state)
+        attn_metadata = AscendMetadata(
+            num_actual_tokens=num_actual_tokens,
+            block_tables=block_table,
+            query_start_loc=query_start_loc,
+            query_lens=query_lens,
+            seq_lens=seq_lens,
+            max_query_len=max_query_len,
+            slot_mapping=slot_mapping,
+            attn_mask=attn_mask,
+            attn_state=attn_state,
+            with_prefill_across_dp=with_prefill_across_dp)
         return attn_metadata