[Main2Main] Upgrade vllm commit to 0109 (#5752)

### What this PR does / why we need it?
Upgrade vllm commit to 0109 (bde38c11df0ea066a740efe9b77fff5418be45df)

1. remove `init_cached_hf_modules ` due to
https://github.com/vllm-project/vllm/pull/31786
2. fix spec_decode e2e test due to
https://github.com/vllm-project/vllm/pull/29821 break
3. fix `vllm.v1.attention.backends.utils` duo to
https://github.com/vllm-project/vllm/pull/31891
4. fix `self.seq_lens - query_lens` on same device due to
https://github.com/vllm-project/vllm/pull/31773
5. skip model_runner_v2 e2e test due to `'_OpNamespace' '_C' object has
no attribute 'get_cuda_view_from_cpu_tensor'`

- vLLM version: v0.13.0
- vLLM main:
2f4e6548ef

Signed-off-by: hfadzxy <starmoon_zhang@163.com>

tests/ut/ops/test_activation.py

@@ -13,10 +13,11 @@
 # This file is a part of the vllm-ascend project.
 #
 
-from unittest.mock import patch
+from unittest.mock import MagicMock, patch
 
 import pytest
 import torch
+from vllm.config import set_current_vllm_config
 from vllm.model_executor.layers.activation import QuickGELU, SiluAndMul
 
 from vllm_ascend.utils import AscendDeviceType
@@ -27,8 +28,20 @@ def dummy_tensor():
     return torch.randn(4, 8, dtype=torch.float16)
 
 
+@pytest.fixture
+def default_vllm_config():
+    mock_config = MagicMock()
+
+    mock_config.compilation_config.dispatch_forward_backend = "eager"
+
+    mock_config.compilation_config.custom_ops = ["all"]
+
+    with set_current_vllm_config(mock_config):
+        yield mock_config
+
+
 @patch("torch_npu.npu_fast_gelu", side_effect=lambda x: x + 1)
-def test_QuickGELU_forward(mock_gelu, dummy_tensor):
+def test_QuickGELU_forward(mock_gelu, dummy_tensor, default_vllm_config):
     layer = QuickGELU()
     out = layer.forward(dummy_tensor)
 
@@ -45,7 +58,7 @@ def test_QuickGELU_forward(mock_gelu, dummy_tensor):
        side_effect=lambda x: None)
 def test_SiluAndMul_forward(mock_maybe_prefetch_mlp_down_proj,
                             mock_maybe_wait_prefetch_done, mock_swiglu,
-                            is_310p, dummy_tensor):
+                            is_310p, dummy_tensor, default_vllm_config):
 
     with patch("vllm_ascend.utils.get_ascend_device_type",
                return_value=AscendDeviceType._310P

tests/ut/ops/test_layernorm.py

@@ -1,7 +1,8 @@
-from unittest.mock import patch
+from unittest.mock import MagicMock, patch
 
 import pytest
 import torch
+from vllm.config import set_current_vllm_config
 from vllm.model_executor.layers.layernorm import RMSNorm
 
 from vllm_ascend.utils import AscendDeviceType
@@ -20,13 +21,22 @@ def mock_add_rms_norm(x, residual, weight, eps):
     return 2 * x, None, 2 * residual
 
 
+@pytest.fixture(autouse=True)
+def default_vllm_config():
+    mock_config = MagicMock()
+    mock_config.compilation_config.custom_ops = ["all"]
+
+    with set_current_vllm_config(mock_config):
+        yield mock_config
+
+
 @pytest.mark.parametrize("is_310p", [True, False])
 @pytest.mark.parametrize("residual",
                          [None, torch.randn(4, 8, dtype=torch.float32)])
 @patch("torch_npu.npu_rms_norm", side_effect=mock_rms_norm)
 @patch("torch_npu.npu_add_rms_norm", side_effect=mock_add_rms_norm)
 def test_RMSNorm_forward(mock_add_rmsnorm, mock_rmsnorm, is_310p, residual,
-                         dummy_tensor):
+                         dummy_tensor, default_vllm_config):
 
     with patch("vllm_ascend.utils.get_ascend_device_type",
                return_value=AscendDeviceType._310P

tests/ut/ops/test_rotary_embedding.py

@@ -78,6 +78,12 @@ class TestAscendRotaryEmbedding(unittest.TestCase):
 
     def setUp(self):
         # Common setup for tests
+        self.config_patcher = patch('vllm.config.vllm.get_current_vllm_config')
+        self.mock_get_config = self.config_patcher.start()
+        mock_config = MagicMock()
+        mock_config.compilation_config.custom_ops = ["all"]
+
+        self.mock_get_config.return_value = mock_config
         self.positions = torch.tensor([1, 2, 3])
         self.query = torch.randn(3, 1, 32, dtype=torch.float16)
         self.key = torch.randn(3, 1, 32, dtype=torch.float16)
@@ -242,6 +248,12 @@ class TestAscendDeepseekScalingRotaryEmbedding(TestBase):
 
     def setUp(self):
         # Common setup for tests
+        self.config_patcher = patch('vllm.config.vllm.get_current_vllm_config')
+        self.mock_get_config = self.config_patcher.start()
+        mock_config = MagicMock()
+        mock_config.compilation_config.custom_ops = ["all"]
+
+        self.mock_get_config.return_value = mock_config
         self.positions = torch.tensor([1, 2, 3])
         self.query = torch.randn(3, 1, 32, dtype=torch.float16)
         self.key = torch.randn(3, 1, 32, dtype=torch.float16)
@@ -368,7 +380,11 @@ class TestAscendDeepseekScalingRotaryEmbedding(TestBase):
 class TestAscendMRotaryEmbedding(unittest.TestCase):
 
     def setUp(self):
-        # Common setup for tests
+        self.config_patcher = patch('vllm.config.vllm.get_current_vllm_config')
+        self.mock_get_config = self.config_patcher.start()
+        mock_config = MagicMock()
+        mock_config.compilation_config.custom_ops = ["all"]
+        self.mock_get_config.return_value = mock_config
         self.number_tokens = 3
         self.num_head = 8
         self.num_kvhead = 8

tests/ut/ops/test_token_dispatcher.py

@@ -29,6 +29,23 @@ from vllm_ascend.ops.fused_moe.token_dispatcher import (  # isort: skip
 class TestTokenDispatcherWithMC2(TestBase):
 
     def setUp(self):
+        self.config_patcher = patch(
+            'vllm_ascend.ops.fused_moe.token_dispatcher.get_current_vllm_config'
+        )
+        self.mock_get_config = self.config_patcher.start()
+
+        mock_config = MagicMock()
+
+        mock_config.scheduler_config.max_num_seqs = 256
+        mock_config.scheduler_config.decode_max_num_seqs = 256
+
+        mock_config.compilation_config.custom_ops = ["all"]
+
+        mock_config.speculative_config = None
+
+        mock_config.parallel_config.tensor_parallel_size = 1
+
+        self.mock_get_config.return_value = mock_config
         self.mc2_group = MagicMock()
         self.mc2_group.device_group.return_value._get_backend.return_value.get_hccl_comm_name.return_value = "hccl_123"
         self.mc2_group.rank_in_group = 0

tests/ut/ops/test_vocab_parallel_embedding.py

@@ -208,6 +208,15 @@ class TestCustomVocabParallelEmbedding(unittest.TestCase):
 class TestAscendLogitsProcessor(unittest.TestCase):
 
     def setUp(self):
+        self.mock_vllm_config = MagicMock()
+        self.mock_vllm_config.compilation_config.custom_ops = ["all"]
+
+        from vllm.config.vllm import set_current_vllm_config
+        set_current_vllm_config(self.mock_vllm_config)
+
+        self.config_patch = patch("vllm.config.vllm.get_current_vllm_config",
+                                  return_value=self.mock_vllm_config)
+        self.config_patch.start()
         self.vocab_size = 50
         self.num_embeddings = 50
         self.embedding_dim = 10