* try to fix the infer_shape of reshape

* Update reshape.c

* fix the bugs of unsqueeze for Mr.Da

* try to add coverage

* try to fix workflow bug

* try to fix the bugs of CI on arm32/64

* try to fix the bugs of CI on arm32/64, 1

* fix the bugs of x86 int8, enable mulit-thread of it

* fix error of compile on arm int8

* update implement of x86 int8 convdw

* fix the bugs of x86 int8, enable mulit-thread of it

* fix error of compile on arm int8

* update CMakeLists to install tengine cpp api header

* update actions scripts
Co-authored-by: Ndongdong <ddzhao@openailab.com>

* fix the error of compile

* Update common.h

add back
Co-authored-by: Nkalcohol <314377460@qq.com>

* prepare to release v1.2

* Add, relu ref op with omp

* support compile with vs2019

Co-authored-by: Nxlchen <xlchen@openailab.com>

* Fix, disable pooling x86 sse implement to avoid bugs

* update yolov4 tiny example

Co-authored-by: Nqwang <qwang02@openailab.com>

add sgemm_i8 avx2 support

./build/install/bin/tm_yolov3_tiny -m models/yolov3_tiny.tmfile -i images/ssd_dog.jpg -r 1 -t 1

* update doc

* Fix, error info in interp ref op

* 修复aarch64编译错误.
Co-authored-by: Nshitouren1994 <shihebing@bigo.sg>

Co-authored-by: Nqwang <qwang02@openailab.com>

* 1.修复neon实现的round bug 2.add email
Co-authored-by: Nshitouren1994 <shihebing@bigo.sg>

* add AutoKernel in README

Co-authored-by: Nwangquan <qwang02@openailab.com>

* 1.fix x86 benchmark bug 2.add int8 im2col+gemm
Co-authored-by: Nshitouren1994 <shihebing@bigo.sg>

Co-authored-by: Nxlchen <xlchen@openailab.com>

fix

* 添加int8convdw:3x3s1和3x3s2
Co-authored-by: Nshitouren1994 <shihebing@bigo.sg>

* pooling pack sse

* 添加int8全连接实现
Co-authored-by: Nshitouren1994 <shihebing@bigo.sg>

Co-authored-by: Nxlchen <xlchen@openailab.com>

* initial int8 perchannel forward mode

* Fix, deconv bias impl, when no bias tensor.