本PR修复了以下两个tensor切片索引bug：

1.修复Tensor索引赋值调用set_value op出现的显存泄露问题，该问题主要是由Inplace策略的使用不当导致，本PR中已完成修复。

2.修复使用多个None类型索引时结果维度异常的问题

* support data_format='NHWC' for prelu channel mode (#37019)

* support data_format='NHWC' for prelu channel mode

* fix prelu weight shape for NHWC of static mode (#38310)

[CustomOp]Add RWLock to protect loading module under multi-thread and multi-process (#38128) (#38271)

[Cherry-pick][CustomOp]Add RWLock to protect loading module under multi-thread and multi-process (#38128)

* fix clip_extra when QAT export model

* clear up unittes modify

Add python interface of subgraph: 1. all_sub_graphs() 2. get_sub_graph(idx)

Fix svd shape error.

修复tensor_shape_transformer中的错误。
之前在类似if len(paddle.shape(x)[0]) > 0中，paddle会被当做一个变量被传入convert_var_shape函数中

Issue37932 反馈 layer.to 不支持paddle.dtype类型的问题，新增了对这类型的支持。详情见：#38018
同时，也一并cherry-pick了遗漏的 PR 36779 的commit。

修改内容：
Cherry-pick #36779
cherrry-pick #38018

动转静时，将函数中的注释行进行删除。
有函数体外的注释行，使用gast库进行func2ast转换时会导致出错，本PR之后将注释行（#开头的行）进行了删除

优化layer.to实现逻辑，相关pr：
Remove additional warnning in layer.to ( #36700)
Refine param conversion logic in layer.to ( #36862)
Fix Layer.to() of device bug ( #37156)

* fix_ele_convert: IElementWiseLayer can broadcast

* fix_ele_convert

fix: when ceil_model==true && Padding_algo!=SAME, (x-size)/stride != int, this convert is wrong (#37929) (#38033)
Co-authored-by: Nfeng_shuai <fengshuai03@baidu.com>

Polish for zip in dy2stat

Fix default behavior if block=None in static mode (#37827)

Fix cflags D_GLIBCXX_USE_CXX11_ABI takes no effect problem in customized op

* py2 to py3 bug and iface fix for pslib (#36102)

* avoid setting logging.basicConfig (#37031)

* fix dropout static when axis != None

* update dropout test

* add dropout test

* fix test

* Update test_dropout_op.py

* Update test_dropout_op.py

* fix testcase

* fix testcase

* Update test_dropout_op.py

* fix testcase

* fix testcase

* optimize perf

* add new test

* fix testcase

cherry-pick #37536

修复pass_desc.proto在编译时产生依赖问题。

cherry-pick of PR #37566:

Based on #37411, this PR:

    Continue to fix the bugs when bias add is none in static graph for fused_attention op.
    Polish and improve the unittests in test_fused_attention_op_api.py.

slice_grad op在选择kernel过程中出现错误，问题原因是在获取use_mkldnn属性时，map中未找到该键值，所以抛出out_of_range异常
本PR在map获取use_mkldnn属性数据前增加了是否存在该键值的判断逻辑，从而避免出现上述异常

cherry-pick #36714

* fix. test=develop

* fix. test=develop

* fix. test=develop

* fix. test=develop

* fix. test=develop

* [heterps]bug fix for _run_from_dataset

* fix heter_server.cc

* fix launch_utils.py

* fix heter_section_worker.cc

* fix. test=develop

* fix. test=develop

* fix data parallel when VOCAB var in program

* fix ci coverage

目前的fused_attention_op不支持attn_mask=None的输入，本PR对此进行了补充，并补充了相应的单测逻辑。