[Patch] format patch module to make it more clear (#601)

Format patch module to make it more clear. 
Add the patch doc description, the new patch must follow this guide.

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>

vllm_ascend/patch/platform/patch_0_8_4/__init__.py

@@ -14,17 +14,5 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 #
-# What's Patched and how it works:
-# ** File: platform/patch_0_8_4/patch_config.py**
-#   1. `vllm.config.ModelConfig.__init__()`
-#    Why:
-#       It is hard coded for sleep mode to support cuda platform only
-#    How：
-#       Using a new method to check if sleep mode is available
-#    Related PR (if no, explain why): 1. refused by vllm. 2. vllm doesn't support 3. prepare to submit....
-#       https://github.com/vllm-project/vllm/pull/16562
-#    Future Plan:
-#       This patch is only used for 084 and can't be revert. just keep as it is.
 
 import vllm_ascend.patch.platform.patch_0_8_4.patch_config  # noqa
-import vllm_ascend.patch.platform.patch_0_8_4.patch_distributed  # noqa

vllm_ascend/patch/platform/patch_common/__init__.py

@@ -13,4 +13,6 @@
 # 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.
-#
+#
+
+import vllm_ascend.patch.platform.patch_common.patch_distributed  # noqa

vllm_ascend/patch/platform/patch_common/patch_distributed.py

@@ -27,40 +27,6 @@ from torch.distributed.distributed_c10d import (Backend, PrefixStore,
 from torch.distributed.rendezvous import rendezvous
 from vllm.config import ParallelConfig
 
-# What's Patched and how it works:
-# ** File: platform/patch_0_8_4/patch_distributed.py**
-#   1. `vllm.distributed.parallel_state.destroy_model_parallel()`
-#    Why:
-#       vllm dose not support outside platform maintain its own `CoordinatorGroup`, vllm-ascend maintain EP and ETP
-#       inside of the repo, and needs a common interface to destroy them, this patch add the interface of destroy
-#       platform owned `CoordinatorGroup` to make sure all the CoordinateGroup can be properly destroyed
-#    How：
-#       Call platform method `destroy_platform_model_parallel` to destroy all the `CoordinateGroup`
-#    Related PR (if no, explain why): no related PR, we want add this ability into vllm
-#    Future Plan:
-#       Remove those patch when vllm merged them
-#   2. `vllm.distributed.stateless_init_torch_distributed_process_group()`
-#    Why:
-#       The stateless process group can not be initialized except from gloo and nccl backend, vllm-ascend
-#       needs to initialize its own stateless process group for communication, so we add the platform related
-#       call to the `stateless_init_torch_distributed_process_group`, to enable other platform which may support
-#       stateless process group initialize method
-#    How：
-#       Call platform method `platform_has_backend_register` to judge if there is a stateless process group initialize
-#       method and call platform method `platform_register_backend` to initialize them
-#    Related PR (if no, explain why): no related PR, we want add this ability into vllm
-#    Future Plan:
-#       Remove those patch when vllm merged them
-#   3. `ParallelConfig.get_next_dp_init_port`
-#    Why:
-#       We want to get dp port from env variable, so the multi-node inference can be properly initialized and run.
-#    How：
-#       Get the dp port from env variable enable multi-mode dp inference
-#    Related PR (if no, explain why): no related PR, we want add this ability into vllm
-#    Future Plan:
-#       Its a workaround in vllm-ascend to enable multi-node dp inference, maybe removed if vllm have better plan
-#       on multi-node dp inference implementation
-
 
 def ascend_destroy_model_parallel():
     """Set the groups to none and destroy them."""

vllm_ascend/patch/platform/patch_main/__init__.py

@@ -14,4 +14,3 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 #
-import vllm_ascend.patch.platform.patch_main.patch_distributed  # noqa F401

vllm_ascend/patch/platform/patch_main/patch_distributed.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.
-# Adapted from vllm/model_executor/models/qwen2_vl.py
-# This file is a part of the vllm-ascend project.
-
-import vllm
-import vllm.distributed
-from vllm.config import ParallelConfig
-
-from vllm_ascend.patch.platform.patch_0_8_4.patch_distributed import (
-    ascend_destroy_model_parallel,
-    ascend_stateless_init_torch_distributed_process_group,
-    parallel_config_get_dp_port)
-
-# All details of those patch please refer to vllm_ascend/patch/platform/patch_0_8_4/patch_distributed.py
-vllm.distributed.parallel_state.destroy_model_parallel = ascend_destroy_model_parallel
-vllm.distributed.stateless_init_torch_distributed_process_group = ascend_stateless_init_torch_distributed_process_group
-ParallelConfig.get_next_dp_init_port = parallel_config_get_dp_port