[Refactor]7/N Extract common code to common_cp (#5490)

RFC: https://github.com/vllm-project/vllm-ascend/issues/4629
Reason：
Eliminate duplicate code for two file(mla_cp.py attention_cp.py) to
common_cp.py.

vLLM version: 0.13.0rc3
vLLM main:
ad32e3e19c

vLLM version: release/v0.13.0
vLLM main:
5fbfa8d9ef

- vLLM version: v0.13.0
- vLLM main:
5326c89803

---------

Signed-off-by: wujinyuan1 <wjy9595@qq.com>
Signed-off-by: wujinyuan1 <wujinyuan1@huawei.com>
Co-authored-by: wujinyuan1 <wjy9595@qq.com>

vllm_ascend/attention/utils.py

@@ -35,52 +35,6 @@ def enable_cp():
                 or prefill_config.decode_context_parallel_size > 1
 
 
-@dataclass
-class AscendMetadataForPrefill:
-
-    @dataclass
-    class AscendPCPMetadata:
-        q_head_idx: torch.Tensor = None
-        q_tail_idx: torch.Tensor = None
-        kv_with_q_head_nomask_idx: torch.Tensor = None
-        kv_with_q_head_mask_idx: torch.Tensor = None
-        kv_with_q_tail_nomask_idx: torch.Tensor = None
-        kv_with_q_tail_mask_idx: torch.Tensor = None
-        attn_mask_seqlens: torch.Tensor = None
-        head_attn_nomask_seqlens: torch.Tensor = None
-        tail_attn_nomask_seqlens: torch.Tensor = None
-        q_full_idx: torch.Tensor = None
-        pcp_prefill_mask: torch.Tensor = None
-
-    @dataclass
-    class ChunkedContextMetadata:
-        actual_chunk_seq_lengths: torch.Tensor
-        actual_seq_lengths_kv: torch.Tensor
-        starts: torch.Tensor
-        chunk_seq_mask_filtered_indices: torch.Tensor
-        chunked_req_mask: Optional[list[bool]] = None
-        local_context_lens_allranks: Optional[list[list[int]]] = None
-        cp_kv_recover_idx_for_chunk: Optional[list[int]] = None
-        kv_inverse_idx_for_chunk: Optional[list[int]] = None
-        batch_chunk_seq_mask: Optional[list[bool]] = None
-        local_total_toks: Optional[int] = None
-
-    """ Prefill Specific Metadata for Ascend"""
-    pcp_metadata: Optional[AscendPCPMetadata] = None
-    pcp_allgather_restore_idx: Optional[List[int]] = None
-    chunked_context: Optional[ChunkedContextMetadata] = None
-    block_tables: torch.Tensor = None
-    actual_seq_lengths_q: torch.Tensor = None
-
-
-@dataclass
-class AscendMetadataForDecode:
-    """ Decode Specific Metadata for Ascend"""
-    num_computed_tokens_of_pcp_dcp: Optional[list[list[list[int]]]] = None
-    batch_seq_mask: torch.Tensor = None
-    block_tables: torch.Tensor = None
-
-
 @dataclass
 # class AscendCommonLongSequenceMetadata:
 class AscendPrefillContextParallelMetadata: