[main] mlp weight prefetch in Qwen Dense Models (#2816)

### What this PR does / why we need it?
This PR prefetchs the weight of mlp layers in Qwen Dense Models to
optimize the performance in Decode phase mainly.

### Does this PR introduce _any_ user-facing change?
 No.

### How was this patch tested?
CI passed with new added/existing test.

- vLLM version: main
- vLLM main:
a1213fae5f

Signed-off-by: rjg-lyh <1318825571@qq.com>
Co-authored-by: Shuming19 <313093131@qq.com>

tests/ut/torchair/models/test_torchair_deepseek_v2.py

@@ -275,7 +275,12 @@ def test_torchair_deepseek_v2_mla_attention(mock_rms_norm, mock_distributed,
 
 @patch("torch_npu.npu_add_rms_norm")
 @patch("torch_npu.npu_rms_norm")
-def test_torchair_deepseek_v2_decoder_layer(mock_rms_norm, mock_add_norm,
+@patch("torch.ops.vllm.maybe_wait_prefetch_done", side_effect=lambda x: None)
+@patch("torch.ops.vllm.maybe_chunk_residual",
+       side_effect=lambda x, residual: residual)
+def test_torchair_deepseek_v2_decoder_layer(mock_maybe_chunk_residual,
+                                            mock_maybe_wait_prefetch_done,
+                                            mock_rms_norm, mock_add_norm,
                                             mock_distributed, base_config,
                                             vllm_config):
     mock_rms_norm.return_value = (torch.randn(2, 128), torch.randn(2, 128))