[Feat] support basic pcp&dcp for qwen3next (#6091)

### What this PR does / why we need it?
This PR implements Context Parallelism (CP) support for the Qwen3-Next
model, including PCP (Parallel Context Parallelism) and DCP
(Dynamic/Data Context Parallelism).

- vLLM version: v0.15.0
- vLLM main:
f176443446

---------

Signed-off-by: SunnyLee219 <3294305115@qq.com>
Signed-off-by: Jingchun Gao <gaojingchun1@huawei.com>
Signed-off-by: 白永斌 <baiyongbin3@h-partners.com>
Signed-off-by: Bai Yongbin <845473182@qq.com>
Co-authored-by: SunnyLee219 <3294305115@qq.com>
Co-authored-by: Jingchun Gao <gaojingchun1@huawei.com>
Co-authored-by: 白永斌 <baiyongbin3@h-partners.com>
Co-authored-by: Mengqing Cao <cmq0113@163.com>

vllm_ascend/ops/triton/fla/utils.py

@@ -24,10 +24,19 @@ def prepare_chunk_indices(cu_seqlens: torch.LongTensor, chunk_size: int) -> torc
     return torch.stack([indices.eq(0).cumsum(0) - 1, indices], 1).to(cu_seqlens)
 
 
+def prepare_final_chunk_indices(cu_seqlens: torch.LongTensor, chunk_size: int) -> torch.LongTensor:
+    indices = triton.cdiv(prepare_lens(cu_seqlens), chunk_size) + 1
+    return torch.cumsum(indices, 0) - 1
+
+
 def prepare_chunk_offsets(cu_seqlens: torch.LongTensor, chunk_size: int) -> torch.LongTensor:
     return torch.cat([cu_seqlens.new_tensor([0]), triton.cdiv(prepare_lens(cu_seqlens), chunk_size)]).cumsum(-1)
 
 
+def prepare_update_chunk_offsets(cu_seqlens: torch.LongTensor, chunk_size: int) -> torch.LongTensor:
+    return torch.cat([cu_seqlens.new_tensor([0]), triton.cdiv(prepare_lens(cu_seqlens), chunk_size) + 1]).cumsum(-1)
+
+
 def input_guard(fn: Callable[..., torch.Tensor]) -> Callable[..., torch.Tensor]:
     """
     A decorator to make sure all input tensors are contiguous and set the device based on input tensors.