add opencl prelu kernel

Merge pull request !4261 from yangruoqi713/test_dw

Merge pull request !4289 from xiefangqi/fix_pad_validate

Merge pull request !4258 from leopz/fix_doc

Merge pull request !4285 from yanghaitao/yht_fix_doc

Merge pull request !4268 from liubuyu/master

Merge pull request !4270 from songhonglei413/roi

Merge pull request !4296 from chenjianping/lite_dev2

Merge pull request !4275 from yuchaojie/add_allreduce_group_for_resnet_gpu

Merge pull request !4263 from chenjianping/lite_dev

Merge pull request !4283 from wangqiuliang/add-insert-gradient-of-operator-support

Merge pull request !4233 from fuzhiye/tmp

Merge pull request !4241 from wangshaocong/lite_wsc

Merge pull request !4209 from lixian/master

Merge pull request !4282 from qianlong21st/fix_cache_compile_error

Merge pull request !3874 from qianlong21st/soft_dvpp

Merge pull request !4253 from lianliguang/master

Merge pull request !4255 from lianliguang/unify-primitive

Merge pull request !4097 from gengdongjie/master

Merge pull request !4272 from yoonlee666/bugfix

Merge pull request !4264 from liuxiao93/check-fp16-fp32

Merge pull request !4274 from zhaozhenlong/lite/op/int8/resize

2. enable int8 conv multi thread execute

Merge pull request !4234 from ling/conv1x1_post

Merge pull request !4207 from yeyunpeng2020/master_time_profile

Merge pull request !4189 from lyvette/tflite_parser

Merge pull request !4238 from guozhijian/update_run_c75b100

Merge pull request !4213 from fangzehua/resize_near

Merge pull request !4271 from hangq/master

adjust resize fp32 using NHWC

Merge pull request !4226 from ZPaC/master-add-gpu-div-kernel

Merge pull request !4266 from chenzhongming/lite

Merge pull request !4053 from panbingao/master

Merge pull request !4256 from hangq/master