relu_kernel test

add #include<random> in util.h

Modified tests with kernel_test_common

define kernel test common

boxing op need midlle blob

data loader kernel

* initial draft for boxing kernel

* draft fix common boxing kernel design

* draft fix common boxing kernel design

* draft fix common boxing kernel design

* fix boxing_kernel.cu and some compiling errors

* fix some compiling errors, still have issues

* compile errors fixed, need refine && test

* fix issues of copy rules in clone box case

* Add simple tests, still debugging...

* fixed bug in InferCopyRules

* fixed issues with memcpy, passed manual test for concat-split boxes

* Add concat-split test case, issues fixed

* add test cases for concat-clone && add-clone boxes of cpu version

* fix typos

* fix types

* fix some issues identified by Will

* add function CopyDataFromRules

* fix add-split forward && concat-clone backward, add test case for add-split forward cpu version

* Modified boxing_kernel by KernelUtil, test cases not finished yet.

* Modified boxing_kernel by KernelUtil, test cases not finished yet.

* re-write by KernelUtil

* delete DeviceType::kALL

* delete DeviceType::kALL

* fix include files issue

* fixed issues found by Will

* fix issues

* fix floating_point_type to FloatingPointType

* fix issues, delete add-split box rules

* fix issues with comments

* change 'slice' to 'concat_dim'

* fix issues

* add NOTE for mutable vars

* fix test file issues

* fix formats

* add a test case with nullptr on concat-split box

* fix comments

* delete test cases

* change BnInOp2BlobPtr to BnInOp2Blob

* fix typos

* refine copy actor

* using unix path separator

* separate Init for CopyHdActor and CopyCommNetActor

* refine CopyHdActor with different stages

* refine CopyCommNetActor with different stages

* remove useless code

* remove useless code

* using cuda stream in thread context

* using OF_SET_MSG_HANDLE

* remove useless code

* remote regst warpper

* fix blob ptr in RemoteRegstWarpper

* fix memory leak

* add model diff accumulate actor

* implement model diff accumulate actor

* fix style

* alignment

* fix MdDiffAccCompTask

* implement model diff acc kernel

* test for model diff acc kernel

fix bug in relu kernel util

* clonekernel

* fix namespace error

* fix clonekernel some problem

* fix coding style and lambda capture

* stack space to heap space

* fix coding style problem in clonekernel

* add a line below namespace

* fix coding style problems

* fix coding style problems in clonekernel

* delete a line, change 1.0 => {1.0f}, change var out_blob => odbn_blob

* blob.h char* to void*

* fix spell mistake

* move template function ti blas dict

* clone_kernel_test

* test fix some problem of clonekernel cpu version

* clone kernel test passed

* fix some coding style problem and add a random number test

* refactoring clone_kernel_test.cpp

* fix some function in clone_kernel_test

* fix coding style clone_kernel_test

* clang-format clone_kernel_test

* add double test

* floating_point_type -> FloatingPointType

relu_kernel, circle_read_line resolve warning