[Bugfix] Fix broken CI (#1848)

### What this PR does / why we need it?
- Fix broken commit by
[#20927](https://github.com/vllm-project/vllm/pull/20927)
- Fix broken commit by
[#20466](https://github.com/vllm-project/vllm/pull/20466)
- TODO: more fully adapt to the upstream reconstruction, let's first
make CI happy

- vLLM version: v0.9.2
- vLLM main:
11dfdf21bf

---------

Signed-off-by: wangli <wangli858794774@gmail.com>

vllm_ascend/pool/__init__.py

@@ -1,16 +0,0 @@
-#
-# Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved.
-#
-# 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.
-#

vllm_ascend/pool/metadata.py

@@ -1,32 +0,0 @@
-#
-# Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved.
-# Copyright 2023 The vLLM team.
-#
-# 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.
-# Adapted from vllm-project/vllm/vllm/v1/pool/metadata.py
-#
-from dataclasses import dataclass
-from typing import Optional
-
-import torch
-from vllm.pooling_params import PoolingParams
-
-
-@dataclass
-class PoolingMetadata:
-    """Tensors for pooling."""
-
-    prompt_lens: torch.Tensor
-    prompt_token_ids: Optional[torch.Tensor]
-    pooling_params: list[PoolingParams]

vllm_ascend/worker/model_runner_v1.py

@@ -59,6 +59,7 @@ from vllm.v1.kv_cache_interface import (FullAttentionSpec, KVCacheConfig,
                                         KVCacheSpec)
 from vllm.v1.outputs import (EMPTY_MODEL_RUNNER_OUTPUT, LogprobsTensors,
                              ModelRunnerOutput)
+from vllm.v1.pool.metadata import PoolingMetadata
 from vllm.v1.sample.metadata import SamplingMetadata
 from vllm.v1.sample.sampler import Sampler
 from vllm.v1.spec_decode.metadata import SpecDecodeMetadata
@@ -76,7 +77,6 @@ from vllm_ascend.attention.attention_v1_torchair import AscendTorchairMetadata
 from vllm_ascend.attention.mla_v1 import (AscendMLAMetadata,
                                           CommonAttentionMetadata)
 from vllm_ascend.platform import NPUPlatform
-from vllm_ascend.pool.metadata import PoolingMetadata
 from vllm_ascend.sample.rejection_sampler import AscendRejectionSampler
 from vllm_ascend.utils import (ACL_FORMAT_FRACTAL_ND, ACL_FORMAT_FRACTAL_NZ,
                                ProfileExecuteDuration,
@@ -571,7 +571,10 @@ class NPUModelRunner(LoRAModelRunnerMixin):
 
         # OPTIMIZATION: Start copying the block table first.
         # This way, we can overlap the copy with the following CPU operations.
-        self.input_batch.block_table.commit(num_reqs)
+        if vllm_version_is("0.9.2"):
+            self.input_batch.block_table.commit(num_reqs)
+        else:
+            self.input_batch.block_table.commit_block_table(num_reqs)
 
         # Get the number of scheduled tokens for each request.
         req_ids = self.input_batch.req_ids
@@ -902,7 +905,10 @@ class NPUModelRunner(LoRAModelRunnerMixin):
 
         # OPTIMIZATION: Start copying the block table first.
         # This way, we can overlap the copy with the following CPU operations.
-        self.input_batch.block_table.commit(num_reqs)
+        if vllm_version_is("0.9.2"):
+            self.input_batch.block_table.commit(num_reqs)
+        else:
+            self.input_batch.block_table.commit_block_table(num_reqs)
 
         # Get the number of scheduled tokens for each request.
         # TODO: The Python loop can be slow. Optimize.

vllm_ascend/worker/npu_input_batch.py

@@ -28,14 +28,13 @@ from vllm.pooling_params import PoolingParams
 from vllm.sampling_params import SamplingParams, SamplingType
 from vllm.utils import swap_dict_values
 from vllm.v1.outputs import LogprobsTensors
+from vllm.v1.pool.metadata import PoolingMetadata
 from vllm.v1.sample.logits_processor import init_builtin_logitsprocs
 from vllm.v1.sample.metadata import SamplingMetadata
 from vllm.v1.spec_decode.utils import is_spec_decode_unsupported
 from vllm.v1.utils import copy_slice
 from vllm.v1.worker.block_table import MultiGroupBlockTable
 
-from vllm_ascend.pool.metadata import PoolingMetadata
-
 _SAMPLING_EPS = 1e-5