support cp&dcp (#3260)

### What this PR does / why we need it?
This PR adds the Prefill Context Parallelism (PCP) feature, which
corresponds to DCP. For specific implementation details, please refer to
the RFC https://github.com/vllm-project/vllm/issues/25749.
TL;DR: PCP enhances long-sequence inference capabilities by partitioning
the sequence dimension during the prefill stage.
### Does this PR introduce _any_ user-facing change?
The current implementation primarily includes the following changes:

Modified ModelRunner.py for CP partitioning logic for tokens;
Modified attention_v1.py and mla_v1.py to adapt the GQA/MLA backend to
PCP.
Modified block_tables.py to extend the KV cache storage based on
DCP&PCP;
Added necessary command-line arguments to control parallelism for PCP;
### How was this patch tested?


- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0

---------

Signed-off-by: LookAround <lixushi@huawei.com>
Signed-off-by: chenjie <chenjie137@huawei.com>
Signed-off-by: Delphine-Nic <tanwenqin@huawei.com>
Signed-off-by: zhangsicheng5 <zhangsicheng5@huawei.com>
Signed-off-by: Feng Liu <liufeng248@huawei.com>
Signed-off-by: gaojc <1055866782@qq.com>
Signed-off-by: weiguihua2 <weiguihua2@huawei.com>
Signed-off-by: z50049692 <zhangmingwei11@huawei.com>
Co-authored-by: chenjie <chenjie137@huawei.com>
Co-authored-by: Delphine-Nic <tanwenqin@huawei.com>
Co-authored-by: zhangsicheng5 <zhangsicheng5@huawei.com>
Co-authored-by: Feng Liu <liufeng248@huawei.com>
Co-authored-by: gaojc <1055866782@qq.com>
Co-authored-by: weiguihua2 <weiguihua2@huawei.com>
Co-authored-by: z50049692 <zhangmingwei11@huawei.com>
Co-authored-by: w00896881 <wangzixuan40@huawei.com>

vllm_ascend/attention/utils.py

@@ -1,5 +1,5 @@
 from dataclasses import dataclass
-from typing import Any, List
+from typing import Any, List, Optional
 
 import torch
 import torch.nn.functional as F
@@ -9,6 +9,39 @@ from vllm.distributed.kv_transfer import (get_kv_transfer_group,
 from vllm.forward_context import ForwardContext, get_forward_context
 
 
+@dataclass
+# class AscendCommonLongSequenceMetadata:
+class AscendPrefillContextParallelMetadata:
+    pcp_allgather_restore_idx: torch.Tensor = None
+
+    num_actual_tokens_pcp_padded: Optional[int] = None
+
+    num_computed_tokens_of_pcp_dcp: Optional[list[Optional[list[Optional[
+        list[int]]]]]] = None
+
+    q_head_idx_tensor: torch.Tensor = None
+
+    q_tail_idx_tensor: torch.Tensor = None
+
+    kv_with_q_head_nomask_idx_tensor: torch.Tensor = None
+
+    kv_with_q_head_mask_idx_tensor: torch.Tensor = None
+
+    kv_with_q_tail_nomask_idx_tensor: torch.Tensor = None
+
+    kv_with_q_tail_mask_idx_tensor: 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 AscendCommonAttentionMetadata:
     """
@@ -72,6 +105,9 @@ class AscendCommonAttentionMetadata:
     cos: torch.Tensor = None
     sin: torch.Tensor = None
 
+    prefill_context_parallel_metadata: Optional[
+        AscendPrefillContextParallelMetadata] = None
+
 
 def split_decodes_and_prefills(
     common_attn_metadata: AscendCommonAttentionMetadata,