[Feature] support aclgraph for model runner v2 (#7110)

### What this PR does / why we need it?
This PR aims to support aclgraph for model runner v2, please see RFC
#5208. The PR contains these modifications:
- adapt to newest commit of vllm main branch.
- supply a unified interface of extra forward context for both model
runner v1 and model runner v2.
- implement graph mode for main model. 

### Does this PR introduce _any_ user-facing change?
no

### How was this patch tested?

- vLLM version: v0.16.0
- vLLM main:
4034c3d32e

---------

Signed-off-by: Ronald1995 <ronaldautomobile@163.com>

vllm_ascend/attention/context_parallel/mla_cp.py

@@ -10,7 +10,6 @@ from vllm.distributed import (
     get_decode_context_model_parallel_world_size,
     get_pcp_group,
 )
-from vllm.forward_context import ForwardContext, get_forward_context
 from vllm.utils.math_utils import cdiv
 from vllm.v1.attention.backend import AttentionCGSupport
 from vllm.v1.kv_cache_interface import AttentionSpec, MLAAttentionSpec
@@ -30,6 +29,7 @@ from vllm_ascend.attention.mla_v1 import (
 )
 # isort: on
 
+from vllm_ascend.ascend_forward_context import _EXTRA_CTX
 from vllm_ascend.attention.context_parallel.common_cp import (
     AscendPCPMetadata,
     CPChunkedContextMetadata,
@@ -294,7 +294,7 @@ class AscendMlaCPImpl(AscendMLAImpl):
         num_dcp_pcp_tokens=None,
         draft_attn_metadatas=None,
     ):
-        if forward_context.is_draft_model:
+        if _EXTRA_CTX.is_draft_model:
             graph_params = get_draft_graph_params()
         else:
             graph_params = get_graph_params()
@@ -659,12 +659,11 @@ class AscendMlaCPImpl(AscendMLAImpl):
             "softmax_lse_flag": True,
         }
 
-        forward_context: ForwardContext = get_forward_context()
-        if forward_context.is_draft_model:
+        if _EXTRA_CTX.is_draft_model:
             graph_params = get_draft_graph_params()
         else:
             graph_params = get_graph_params()
-        if forward_context.capturing:
+        if _EXTRA_CTX.capturing:
             stream = torch_npu.npu.current_stream()
             event = torch.npu.ExternalEvent()
             event.wait(stream)