Add unit test local cpu guide and enable base testcase (#1566)

### What this PR does / why we need it?
Use Base test and cleanup all manaul patch code
- Cleanup EPLB config to avoid tmp test file
- Use BaseTest with global cache
- Add license
- Add a doc to setup unit test in local env 

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

### How was this patch tested?
CI passed

Signed-off-by: Yikun Jiang <yikunkero@gmail.com>

Author: Yikun

docs/source/developer_guide/contribution/testing.md

@@ -9,8 +9,52 @@ The fastest way to setup test environment is to use the main branch container im
 :::::{tab-set}
 :sync-group: e2e
 
-::::{tab-item} Single card
+::::{tab-item} Local (CPU)
 :selected:
+:sync: cpu
+
+You can run the unit tests on CPU with the following steps:
+
+```{code-block} bash
+   :substitutions:
+
+cd ~/vllm-project/
+# ls
+# vllm  vllm-ascend
+
+# Use mirror to speedup download
+# docker pull quay.nju.edu.cn/ascend/cann:|cann_image_tag|
+export IMAGE=quay.io/ascend/cann:|cann_image_tag|
+docker run --rm --name vllm-ascend-ut \
+    -v $(pwd):/vllm-project \
+    -v ~/.cache:/root/.cache \
+    -ti $IMAGE bash
+
+# (Optional) Configure mirror to speedup download
+sed -i 's|ports.ubuntu.com|mirrors.huaweicloud.com|g' /etc/apt/sources.list
+pip config set global.index-url https://mirrors.huaweicloud.com/repository/pypi/simple/
+
+# For torch-npu dev version or x86 machine
+export PIP_EXTRA_INDEX_URL="https://download.pytorch.org/whl/cpu/ https://mirrors.huaweicloud.com/ascend/repos/pypi"
+
+apt-get update -y
+apt-get install -y python3-pip git vim wget net-tools gcc g++ cmake libnuma-dev curl gnupg2
+
+# Install vllm
+cd /vllm-project/vllm
+VLLM_TARGET_DEVICE=empty python3 -m pip -v install .
+
+# Install vllm-ascend
+cd /vllm-project/vllm-ascend
+# [IMPORTANT] Import LD_LIBRARY_PATH to enumerate the CANN environment under CPU
+export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/Ascend/ascend-toolkit/latest/$(uname -m)-linux/devlib
+python3 -m pip install -r requirements-dev.txt
+python3 -m pip install -v .
+```
+
+::::
+
+::::{tab-item} Single card
 :sync: single
 
 ```{code-block} bash
@@ -36,6 +80,16 @@ docker run --rm \
     -it $IMAGE bash
 ```
 
+After starting the container, you should install the required packages:
+
+```bash
+# Prepare
+pip config set global.index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple
+
+# Install required packages
+pip install -r requirements-dev.txt
+```
+
 ::::
 
 ::::{tab-item} Multi cards
@@ -63,20 +117,23 @@ docker run --rm \
     -p 8000:8000 \
     -it $IMAGE bash
 ```
-::::
-
-:::::
 
 After starting the container, you should install the required packages:
 
 ```bash
+cd /vllm-workspace/vllm-ascend/
+
 # Prepare
 pip config set global.index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple
 
 # Install required packages
 pip install -r requirements-dev.txt
 ```
 
+::::
+
+:::::
+
 ## Running tests
 
 ### Unit test
@@ -89,14 +146,48 @@ There are several principles to follow when writing unit tests:
 - Example: [tests/ut/test_ascend_config.py](https://github.com/vllm-project/vllm-ascend/blob/main/tests/ut/test_ascend_config.py).
 - You can run the unit tests using `pytest`:
 
-    ```bash
-    cd /vllm-workspace/vllm-ascend/
-    # Run all single card the tests
-    pytest -sv tests/ut
+:::::{tab-set}
+:sync-group: e2e
 
-    # Run 
-    pytest -sv tests/ut/test_ascend_config.py
-    ```
+::::{tab-item} Local (CPU)
+:selected:
+:sync: cpu
+
+```bash
+# Run unit tests
+export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/Ascend/ascend-toolkit/latest/$(uname -m)-linux/devlib
+VLLM_USE_V1=1 TORCH_DEVICE_BACKEND_AUTOLOAD=0 pytest -sv tests/ut
+```
+
+::::
+
+::::{tab-item} Single card
+:sync: single
+
+```bash
+cd /vllm-workspace/vllm-ascend/
+# Run all single card the tests
+pytest -sv tests/ut
+
+# Run single test
+pytest -sv tests/ut/test_ascend_config.py
+```
+::::
+
+::::{tab-item} Multi cards test
+:sync: multi
+
+```bash
+cd /vllm-workspace/vllm-ascend/
+# Run all single card the tests
+pytest -sv tests/ut
+
+# Run single test
+pytest -sv tests/ut/test_ascend_config.py
+```
+::::
+
+:::::
 
 ### E2E test
 
@@ -106,6 +197,12 @@ locally.
 :::::{tab-set}
 :sync-group: e2e
 
+::::{tab-item} Local (CPU)
+:sync: cpu
+
+You can't run e2e test on CPU.
+::::
+
 ::::{tab-item} Single card
 :selected:
 :sync: single

tests/ut/attention/test_attention_v1.py

@@ -1,5 +1,3 @@
-import vllm_ascend.patch.worker.patch_common.patch_utils  # type: ignore[import]  # isort: skip  # noqa
-
 from unittest.mock import MagicMock, patch
 
 import torch

tests/ut/base.py

@@ -1,7 +1,26 @@
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# This file is a part of the vllm-ascend project.
+#
+
 import unittest
 
 from vllm_ascend.utils import adapt_patch
 
+# fused moe ops test will hit the infer_schema error, we need add the patch
+# here to make the test pass.
+import vllm_ascend.patch.worker.patch_common.patch_utils  # type: ignore[import]  # isort: skip  # noqa
+
 
 class TestBase(unittest.TestCase):
 

tests/ut/distributed/test_parallel_state.py

@@ -1,16 +1,31 @@
-import unittest
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# This file is a part of the vllm-ascend project.
+#
+
 from unittest.mock import MagicMock, patch
 
 import pytest
 from vllm.distributed.parallel_state import GroupCoordinator
 
 import vllm_ascend
+from tests.ut.base import TestBase
 from vllm_ascend.distributed.parallel_state import (
     destory_ascend_model_parallel, get_ep_group, get_etp_group,
     init_ascend_model_parallel, model_parallel_initialized)
 
 
-class TestParallelState(unittest.TestCase):
+class TestParallelState(TestBase):
 
     @patch('vllm_ascend.distributed.parallel_state._EP',
            new_callable=lambda: MagicMock(spec=GroupCoordinator))

tests/ut/ops/expert_map.json

@@ -0,0 +1,17 @@
+{
+    "moe_layer_count":
+    1,
+    "layer_list": [{
+        "layer_id":
+        0,
+        "device_count":
+        2,
+        "device_list": [{
+            "device_id": 0,
+            "device_expert": [7, 2, 0, 3, 5]
+        }, {
+            "device_id": 1,
+            "device_expert": [6, 1, 4, 7, 2]
+        }]
+    }]
+}

tests/ut/ops/test_expert_load_balancer.py

@@ -1,14 +1,26 @@
-# fused moe ops test will hit the infer_schema error, we need add the patch
-# here to make the test pass.
-import vllm_ascend.patch.worker.patch_common.patch_utils  # type: ignore[import]  # isort: skip  # noqa
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# This file is a part of the vllm-ascend project.
+#
 
 import json
-import unittest
+import os
 from typing import List, TypedDict
 from unittest import mock
 
 import torch
 
+from tests.ut.base import TestBase
 from vllm_ascend.ops.expert_load_balancer import ExpertLoadBalancer
 
 
@@ -28,31 +40,13 @@ class MockData(TypedDict):
     layer_list: List[Layer]
 
 
-MOCK_DATA: MockData = {
-    "moe_layer_count":
-    1,
-    "layer_list": [{
-        "layer_id":
-        0,
-        "device_count":
-        2,
-        "device_list": [{
-            "device_id": 0,
-            "device_expert": [7, 2, 0, 3, 5]
-        }, {
-            "device_id": 1,
-            "device_expert": [6, 1, 4, 7, 2]
-        }]
-    }]
-}
-
-
-class TestExpertLoadBalancer(unittest.TestCase):
+class TestExpertLoadBalancer(TestBase):
 
     def setUp(self):
-        json_file = "expert_map.json"
-        with open(json_file, 'w') as f:
-            json.dump(MOCK_DATA, f)
+        _TEST_DIR = os.path.dirname(__file__)
+        json_file = _TEST_DIR + "/expert_map.json"
+        with open(json_file, 'r') as f:
+            self.expert_map: MockData = json.load(f)
 
         self.expert_load_balancer = ExpertLoadBalancer(json_file,
                                                        global_expert_num=8)
@@ -62,9 +56,9 @@ def test_init(self):
         self.assertIsInstance(self.expert_load_balancer.expert_map_tensor,
                               torch.Tensor)
         self.assertEqual(self.expert_load_balancer.layers_num,
-                         MOCK_DATA["moe_layer_count"])
+                         self.expert_map["moe_layer_count"])
         self.assertEqual(self.expert_load_balancer.ranks_num,
-                         MOCK_DATA["layer_list"][0]["device_count"])
+                         self.expert_map["layer_list"][0]["device_count"])
 
     def test_generate_index_dicts(self):
         tensor_2d = torch.tensor([[7, 2, 0, 3, 5], [6, 1, 4, 7, 2]])
@@ -142,6 +136,6 @@ def test_get_rank_log2phy_map(self):
     def test_get_global_redundant_expert_num(self):
         redundant_expert_num = self.expert_load_balancer.get_global_redundant_expert_num(
         )
-        expected_redundant_expert_num = len(MOCK_DATA["layer_list"][0]["device_list"][0]["device_expert"]) * \
-                                        MOCK_DATA["layer_list"][0]["device_count"] - 8
+        expected_redundant_expert_num = len(self.expert_map["layer_list"][0]["device_list"][0]["device_expert"]) * \
+                                        self.expert_map["layer_list"][0]["device_count"] - 8
         self.assertEqual(redundant_expert_num, expected_redundant_expert_num)

tests/ut/patch/worker/patch_common/test_patch_distributed.py

@@ -1,3 +1,18 @@
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# This file is a part of the vllm-ascend project.
+#
+
 from tests.ut.base import TestBase
 
 

tests/ut/patch/worker/patch_common/test_patch_sampler.py

@@ -1,17 +1,20 @@
 import importlib
 import os
-import unittest
 from unittest import mock
 
 import torch
 from vllm.v1.sample.ops import topk_topp_sampler
 
+from tests.ut.base import TestBase
 
-class TestTopKTopPSamplerOptimize(unittest.TestCase):
+
+class TestTopKTopPSamplerOptimize(TestBase):
 
     @mock.patch.dict(os.environ, {"VLLM_ASCEND_ENABLE_TOPK_OPTIMIZE": "1"})
     @mock.patch("torch_npu.npu_top_k_top_p")
     def test_npu_topk_topp_called_when_optimized(self, mock_npu_op):
+        # We have to patch and reload because the patch will take effect
+        # only after VLLM_ASCEND_ENABLE_TOPK_OPTIMIZE is set.
         import vllm_ascend.patch.worker.patch_0_9_1.patch_sampler
         importlib.reload(vllm_ascend.patch.worker.patch_0_9_1.patch_sampler)
 

tests/ut/quantization/test_quant_config.py

@@ -1,5 +1,3 @@
-import vllm_ascend.patch.worker.patch_common.patch_utils  # type: ignore[import]  # isort: skip  # noqa
-
 from unittest.mock import MagicMock, patch
 
 import torch

tests/ut/quantization/test_quantizer.py

@@ -1,5 +1,3 @@
-import vllm_ascend.patch.worker.patch_common.patch_utils  # type: ignore[import]  # isort: skip  # noqa
-
 from unittest.mock import MagicMock, patch
 
 from tests.ut.base import TestBase