add join for ref

adjust env eliminate to eliminate all env ops

add partial app cache

resolve while endless

fix env eliminate

support for "for while" cases

fix join shape error

Merge pull request !4460 from zongha/master

!4441 [MS][LITE] optimize arm cpu fp32 op: add assembly function for conv depthwise border calculating
Merge pull request !4441 from yangruoqi713/lite

Merge pull request !4382 from zhangzheng/master

Merge pull request !4449 from fuzhiye/tmp

2.fix conv 3x3 bug

Merge pull request !4461 from chenjianping/lite_dev

Merge pull request !4284 from yao_yf/remove_to_full_tensor_and_load_inputs_in_exexute_stage

Merge pull request !4443 from vlne-v1/I1RDCY-slice_shape_update_initializer

Merge pull request !4451 from yankai10/merge

Merge pull request !4402 from luoyang/c-api-pyfunc

2nd for update thor bert

Merge pull request !4444 from JoyLvliang/enlarge-time-threshold-of-resnet50-st-in-pynative

Merge pull request !4389 from yanghaitao/yht_fix_doc2

Merge pull request !4429 from ZhangQinghua/master1

Merge pull request !4450 from ghzl/adjust-test

* add support to update by scalar* fix update error for by Initializer when slice_shape is* add ut for compute and update

Merge pull request !4387 from 徐安越/master

Merge pull request !4453 from yangruoqi713/dw_int8

Merge pull request !4437 from 张学同/to_merge

Merge pull request !4421 from ms_yan/api_comment

Merge pull request !4236 from baihuawei/heter

Merge pull request !4391 from chenfei_mindspore/master

Merge pull request !4440 from jianghui/master

Merge pull request !4438 from mengchunyang/master

Merge pull request !4403 from byweng/master

Merge pull request !4403 from byweng/master

Merge pull request !4419 from zhaodezan/master

Merge pull request !4432 from fuzhiye/tmp