Merge pull request !27980 from lilei/insert_virtualdataset_for_master

2) changing CumSum from composition to a single Operator;
3) add InferMirrorOps for CumSumInfo.

Merge pull request !28073 from huanghui/fix-warning

Merge pull request !27862 from Xiaoda/119-adapting-sharding-propagation

Merge pull request !27856 from yangzhenzhang/use-neighborexchangev2-for-conv2d

Merge pull request !27525 from yangzhenzhang/support-opt-parallel-for-adafactor

Merge pull request !27439 from Xiaoda/117-add-cumsum-op

Merge pull request !27428 from lilei/modify_virtualdataset_for_master

Merge pull request !27401 from jiahongQian/master

Merge pull request !27237 from lilei/modify_virtualdataset_for_master

Merge pull request !27251 from lilei/add_parallel_ut

Merge pull request !27024 from jiahongQian/master

However, to be forward compatitable, it is suitable to keep 'auto_parallel_search_mode' available. This PR recovers the 'auto_parallel_search_mode' interface and adds a warning when using this old interface.

This PR also deals with other codestyle things.

Merge pull request !26807 from Xiaoda/113-auto-parallel-search-mode-changes-to-search-mode

Merge pull request !26804 from yao_yf/virtual_dataset_avoid_auto_parallel

Merge pull request !26619 from 王志伟/parallel_ut_refactoring1

2) 'sharding_propagation' moves to 'search_mode';

Merge pull request !26006 from yao_yf/add_transformer_slice_activation_config

Merge pull request !26414 from yao_yf/fault_recover_by_mirror_group_fix_opt_shard

slice recompute activation

1) when propagate sharding strategy from one op to another, try to find the strategy with zero communication cost;
2) if there is no such strategy, find the strategy with minimum communication cost, and raise a warning;

Merge pull request !26350 from yangzhenzhang/add-check-for-resize-op

Merge pull request !26036 from yangzhenzhang/add-output-strategy-for-op-init

1) using 'swc index of strategy_cost_' as reshape's selected strategy;
2) when encountering reshape in BFS, select the 'swc index' with zero communication cost;
3) when encountering a reshape that is already visited, check whether there exists communication between reshape and current operator. It is OK if communication happens between two configured operators;
4) currently, two consecutive reshapes are not supported;
5) adjusting BFS structure in graph_costmodel.cc;
6) adjusting some code in step_auto_parallel.cc to avoid cyclomatic complexity.