[Perf][V1] Fully overlap model execution (#2783)

This PR is based on top of
[#23569](https://github.com/vllm-project/vllm/pull/23569) and
[#24219](https://github.com/vllm-project/vllm/pull/24219).

### What this PR does / why we need it?
This PR allows the model runner to function asynchronously when using
async scheduling. This allows full overlap of the cpu operations
(including prepare_inputs) and the model forward pass. This diff is
functional and does not support speculative decoding, PP, or guided
decoding.

Expected speedup is 5-10% over the current async scheduling.

### Does this PR introduce _any_ user-facing change?

### How was this patch tested?
server
```
python -m vllm.entrypoints.openai.api_server --model=Qwen3-32B\
	--trust-remote-code --enforce-eager \
	--distributed-executor-backend=mp \
	-tp=4 \
	--port 8006 \
	--max-model-len 32000 \
	--block-size 128 \
	--gpu-memory-utilization 0.99
```
client
```
python $TEST_PY --backend vllm --trust-remote-code --model Qwen3-32B \
  --dataset-name random --random-input-len 2048 --random-output-len 2048 \
  --ignore-eos\
  --num-prompts 48 --max-concurrency 48  --request-rate inf --temperature 0 \
  --metric-percentiles 90  --base-url http://localhost:8006 --save-result \
  --result-dir $PROFILER_DIR
```

benchmark test based on Qwen3-32B TPOT result:
||forward async| scheduler async |sync|
|-|-|-|-|
|avg|41.73|41.86|44.20|
|improve0|0.3%|0|0|
|improve1|5.58%|0|0|

benchmark test based on Qwen2___5-VL-7B-Instruct TPOT result:
||forward async|sync|
|-|-|-|
|avg|23.22|29.16|
|improve|20.3%|0|


- vLLM version: main
- vLLM main:
e93f4cc9e3

Signed-off-by: jiangpeng36 <jiangpeng36@huawei.com>
Signed-off-by: Ronald1995 <ronaldautomobile@163.com>
Co-authored-by: jiangpeng36 <jiangpeng36@huawei.com>
Co-authored-by: Ronald1995 <ronaldautomobile@163.com>

tests/e2e/singlecard/test_ascend_scheduler.py

@@ -1,6 +1,7 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 import pytest
+from vllm import SamplingParams
 
 from tests.e2e.conftest import VllmRunner
 from tests.e2e.model_utils import check_outputs_equal
@@ -86,3 +87,25 @@ def test_chunked_prefill_with_ascend_scheduler(
         name_0="vllm_output",
         name_1="chunked_prefill_output",
     )
+
+
+def test_async_scheduling() -> None:
+    prompts = [
+        "Hello, my name is",
+        "The president of the United States is",
+        "The capital of France is",
+        "The future of AI is",
+    ] * 10
+    sampling_params = SamplingParams(temperature=0.2,
+                                     max_tokens=10,
+                                     stop_token_ids=None)
+
+    with VllmRunner(
+            "Qwen/Qwen2.5-0.5B-Instruct",
+            max_model_len=4096,
+            max_num_seqs=50,
+            dtype="bfloat16",
+            gpu_memory_utilization=0.9,
+            async_scheduling=True,
+    ) as vllm_model:
+        vllm_model.generate(prompts, sampling_params=sampling_params)