* fix pooling3x3s2 max. test=develop

* fix format. test=devleop

* fix format. test=develop

* add device info for 865. test=develop

* [LITE][BM] support multiclass_nms2 and fix some issues, test=develop

* create

* [LITE][BM] fix input shape order changed issue,test=develop

* add reduce op on arm. test=develop

* fix format. test=develop

* fix acccording to comments. test=develop

* fix bilinear_resize result not equal with fluid. test=develop

* fix cv build error. test=develop


* fix format. test=develop

* fix conv_3x3s1_dw v7-compute nan. test=develop

* fix compute. tets=develop

* set sgemm basic_test is false. test=develop

* [X86] add box_coder kernel and fix compile on MAC, test=develop

* [DOC] remove debug info, test=develop

* [X86] add new op of density_prior_box, test=develop

* add scatter op on arm. test=develop

* fix cmake error. test=develop

* test=develop

[LITE][XPU] 1. Add sequence_unpad kernel for XPU; 2. Bugfix in sequence_unpad kernel for x86, as InferShapeImpl() is now empty in lite/operators/sequence_unpad_op.cc; 3. Refine TargetWrapperXPU; (#4237)

* [Bugfix][OpenCL] fix depthwise_conv2d_3x3 with dilation > 1. test=develop

* platform portability of tls, test=develop

* update build_ios.sh, test=develop

* add static keyword for tls, test=develop

* rename the alias of tls, test=develop

* [OpenCL] Delete redundant calling: InitializeDevice. test=develop

* fix conv3x3 group. test=develop
* remove useless. test=develop
* optimize 3x3 group

* add v8 4x8 implment. test=devvelop

* fix run error

* change sgemm compute. test=develop

* fix format. test=develop

* [KERNEL][OPENCL] support opencl concat 3inputs, 4inputs of cases; enable conv3x3 with group; fix buffer opencl concat; clean and clear cmake of opencl kernels. test=develop

* fix deconv+act compute error. test=develop

* fix format test=develop

* fix pooling bug and speed

* fix build error

* delete VLOGin pool, test=develop

* add openmp, test=develop

* fix lite/kernels/arm/pool_compute_test basic_pooling compute error bug, test=develop

* update pooling 2-pad to 4-pad, test=develop

* fix 2-pad to 4-pad in operators/pool_op.h, AttachKernel will set param, so 2-pad to 4-pad funcs should put in AttachKernel. test=ddevellop

* put 2-pad to 4-pad in AttachImpl, test=develop

* according to reviews, fix some format error. test=develop

* fix format errorr, add (). test=develop

* change paddings type to support dynamically modify, test=develop

* update padding type int other devices, test=develop

* fix x8d build error on shared_ptr, test=ddevelop

* fix formmat in operators pool_op.cc, test=develop

* fix conflict

* fix conv_dw

* add relu

* fix build

* fix build

* fix compute

* fix compute error

* fix conv3x3 compute error

* fix conv3x3s2

* fix conv3x3s2  kl

* fix format, test=develop

* add some op infershape implement, test=develop

* add reshape infershape, test=develop

* fix format， test=develop

* fix format, test=develop

* fix space format. test=develop

* add conv_transpose+bn fusion. test=develop

* delete note, test=develop

* fix format, test=develop

* fix format space, test=develop

* fix opt run error, test=develop

* add boxcoder opencl kernel, test=develop

* fix format, test=develop

* add cmake, test=develop

* fix format. test=develop

* fix format. test=develop

* fix format aa. test=develop

* fix , test=develop

* update profile info(add new element), test=develop

* fix clang ut build error

* add gemm+relu6

* fix build error

* fix .h

* fix gemm_s8

* fix ut conv+leakyRelu

* improve 3x3s1 direct profile

* fix format, test=develop

* add gemv+relu6/lleakyRelu

* fix v7 build bug

* fix relu6 bug

* fix gemm ut bug

* fix ut

* fix ut

* fi format. test=develop

* fix format. test=develop

* fic format. test=develop

* ff. test=develop

* fix v7 clang build error, test=develop

* fix v7 build register error, test=develop

* fix format.  test=develop

* add 2x2s2p1 pooling. test=develop

* fix conflict, test=develop

* fix test=develop

* fix conflict, test=develop

* rm other info, test=develop

* rm other info, test=develop

* fix build register error, ttest=develop

* fix format, test=develop

* ff format,test=develop

* fix relu6 problem, test=develop

* fix form, test=develop

* fix format, test=develop

* ff, test=develop

* add deformable conv op

* add six / scale , test=develop

* fix 1x1 deformable conv

* add convparam to deformabelconv param.

* fix other conv kernel size

* delete printf info

* fix format, test=develop

* fix formatt. test=develop

* delete exttra info, test=develop

* test=develop

* ff, test=develop

* fix ut error. test=develop

* fix ut, test=develop

* fix format. test=develop

* test=develop

* fix pooling overflow, test=develop

* fix conflict test=develop

* delete unuseful message. test=develop

* add grouup_norm

* fix format. test=develop

* fix foormat, test=develop

* fix format. test=develop

* fix ff.test=develoop

* fix xiaodu crash. test=develop

* format. test=develop

* fix concatt axis < 0 errorr,ttest=develop

* fix format. test=develop

* fix conv int8 kernel choose and sooftmax compute bug

* change axis_size = 4 kernel choose, test=develop

* fix format. test=develop

* uupdate sequence_pool and sequence_conv profiler, test=develop

* uupdate sequence_pool and sequence_conv profiler, test=develop

* fix format, testt=develop

* fix format, test=develop

* fix format test=develop

* fix compute error. test=develop

* dd

* fix compute error

* fix compute error, test=develop

* delete warning and extra info, test=develop

* update sequence_conv profile

* delete extra file, test=develop

* delete extra file test=develop

* fix format test=develop

* d

* delete test=develop

* test=develop

* add elu act

* update build

* fix elu act not find error, test=develop

* fix format.test=develop

* test=develop

* add sequence_pool_grad op on arm

* test=develop

* fix build error

* fix int8 model opt error iin conv+conv fusion, test=develop

* fix format. test=develop

* fiix build error, test=develop

* fix format, test=develop

* fix build,test=develop

* test=develop

* fix arm winograd compute segment. test=develop

* fix ttfnet bug. test=develop

* fix format. test=develop

* fix format. test=develop

* fix compute error

* fix format, test=develop

* fix 3x3s2p0 max compute error when input_padding is not equal. test=develop

* fix format test=develop

* fix format test=develop

* fix compute error. test=develop

* test=develop

* test=develop

* fix compute. test=develop

* add coonv+conv fusion requirment. test=develop

* test=develop

* fix format. test=develop

* refresh coonv+conv. test=develop

* [OPENCL] Add hard sigmoid opencl kernel. test=develop

* optimize hard sigmoid. test=develop

* [OPENCL] if specify addr , do not need to modify. avoid mem copy

* test=develop

* [LITE][BM] Fixed dependency path issues for compiling third-party libraries. test=develop

* Update CMakeLists.txt

* optimize conv_dw profiler

* fix build conv_dw_3x3s1 bug

* update conv_dw_3x3s2

* fxi foormat test=develop

* fix compute error

* fix format, test=develop

* fix sequence pool grad compute error, test=develop

* fix sequence pool grad run, test=develop

* fix format test=develop

* add support for three-dimentional input in reduce max test=develop

* [ascend] add pool and elementwise op, test=develop

* [ascend] add batch_norm op and simpify tensor update, test=develop

* change pooling3x3s2max padding right to intrisic. test=develop