Ubuntu Dev

Install Development Environment

Install Nvidia driver (Ubuntu 20.04)

1. install graphic driver (2080Ti driver is already installed on Ubuntu 20.04)
2. install cuda toolkit (https://developer.nvidia.com/cuda-toolkit-archive)

   $ wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/cuda-ubuntu2004.pin
   $ sudo mv cuda-ubuntu2004.pin /etc/apt/preferences.d/cuda-repository-pin-600
   $ wget https://developer.download.nvidia.com/compute/cuda/11.3.1/local_installers/cuda-repo-ubuntu2004-11-3-local_11.3.1-465.19.01-1_amd64.deb
   $ sudo dpkg -i cuda-repo-ubuntu2004-11-3-local_11.3.1-465.19.01-1_amd64.deb
   $ sudo apt-key add /var/cuda-repo-ubuntu2004-11-3-local/7fa2af80.pub
   $ sudo apt-get update
   $ sudo apt-get -y install cuda
   

3. install cudnn (https://developer.nvidia.com/cudnn) (don’t use)

   $ tar xvzf cudnn-11.2-linux-x64-v8.1.0.77.tgz
   $ sudo cp cuda/include/cudnn* /usr/local/cuda/include
   $ sudo cp cuda/lib64/libcudnn* /usr/local/cuda/lib64
   $ sudo chmod a+r /usr/local/cuda/include/cudnn.h /usr/local/cuda/lib64/libcudnn*
   

4. Setup environment .profile ``` CUDA_HOME=/usr/local/cuda-11.3 export PATH=$CUDA_HOME/bin:$PATH

export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH

7. install cudnn toolkit (don't install nvidia-cuda-toolkit of apt)

$ # sudo apt install nvidia-cuda-toolkit (don’t install)

8. solve link warning

$ sudo ln -sf /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_adv_train.so.8.2.0 /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_adv_train.so.8 $ sudo ln -sf /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_ops_infer.so.8.2.0 /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_ops_infer.so.8 $ sudo ln -sf /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8.2.0 /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8 $ sudo ln -sf /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8.2.0 /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8 $ sudo ln -sf /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8.2.0 /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8 $ sudo ln -sf /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8.2.0 /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8 $ sudo ln -sf /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8.2.0 /usr/local/cuda/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8

## Build Essential

$ sudo apt install build-essential $ sudo apt install ubuntu-restricted-extras #(to intall extra media codecs)

## Install Java
### Install OpenJdk

$ sudo apt install openjdk-8-jdk $ sudo apt install openjdk-11-jdk

### change java version

$ sudo update-alternatives –config java

## Build Essential

$ sudo apt install build-essential $ sudo apt install ubuntu-restricted-extras #(to intall extra media codecs)

$ sudo apt install curl $ sudo apt install libcurl4-openssl-dev zlib1g-dev $ sudo apt install libssl-dev $ sudo apt install openssl

## Install Cmake (to install higher version)
0. install curl-devel zlib-devel (the cmake with Https support is needed for installing opencv) 
https://github.com/opencv/opencv_contrib/issues/1131 (check skvark's comment and ./bootstrap --system-curl. from the version of Cmake-3.12, --system-curls might be not needed for Https support)
    ```
    $ sudo yum -y install curl-devel zlib-devel
    ```
    For Ubuntu,
    ```
    $ sudo apt install libcurl4-openssl-dev zlib1g-dev
    ```
1. goto https://cmake.org/download/, and download the latest stable install file
2. tar -zxvf cmake-xxxversionxxx.tar.gz
3. cd cmake-xxxversionxxx
4. install
    ```
    $ ./bootstrap --system-curl
    $ make
    $ sudo make install
    ```
    
## Install Miscellaneous

$ sudo apt install maven $ sudo apt install ant $ sudo apt install libhiredis-dev $ sudo apt install -y libthrift-dev $ sudo apt install -y libprotobuf-dev libprotobuf-c-dev

To install NATS,

$ sudo apt install snapd $ sudo snap install nats

To install GTest,

$ sudo apt install googletest $ sudo apt install googletest-tools

## Install Boost

$ sudo apt install libboost-all-dev

## Install OpenCV
### Preparation

$ sudo apt install -y libssl-dev $ sudo apt install -y git pkg-config $ sudo apt install -y libcurl4-gnutls-dev $ sudo apt install -y libcurl4-openssl-dev $ sudo apt install -y libgtk-3-dev libgtk2.0-dev $ sudo apt install -y libpng-dev libjpeg-dev libjpeg-turbo8-dev libtiff-dev libwebp-dev $ sudo apt install -y libdc1394-22-dev libv4l-dev v4l-utils libxine2-dev $ sudo apt install -y libavcodec-dev libavformat-dev $ sudo apt install -y libxvidcore-dev libx264-dev $ sudo apt install -y gstreamer1.0-opencv libgstreamer1.0-dev libgstreamer-plugins-base1.0-dev $ sudo apt install -y mesa-utils libgl1-mesa-dri $ sudo apt install -y libatlas-base-dev gfortran $ sudo apt install -y libtbb-dev libeigen3-dev $ sudo apt install -y ffmpeg libswscale-dev $ sudo apt install -y libavcodec-dev libavformat-dev libavutil-dev libavresample-dev

### Install OpenCV

$ sudo apt install -y libopencv-contrib4.2 (is already installed on Ubuntu 20.04? check!!) $ sudo apt install -y libopencv-dev $ sudo apt install -y libopencv-contrib-dev # check please

## Install darknet
1. download source

$ git clone https://github.com/AlexeyAB/darknet.git

3. modify MakeFile

$ cd darknet $ vi Makefile

Makefile

GPU=1 CUDNN=1 CUDNN_HALF=0 OPENCV=1 AVX=0 OPENMP=0 LIBSO=0 ZED_CAMERA=0 ZED_CAMERA_v2_8=0

4. build

$ make

5. use darknet folder and files

## Install Python3
python3.8 is already installed on Ubutu 20.04

$ sudo apt install python3-pip # python3-dev will be installed with python3-pip $ sudo pip3 install opencv-python

## Modify .zshrc

$ cd $ vi .zshrc … $ source .zshrc

.zshrc

JAVA_HOME=/usr/lib/jvm/java-1.8.0-openjdk-amd64

$ JAVA_HOME=/usr/lib/jvm/java-1.11.0-openjdk

export JAVA_HOME

export CPLUS_INCLUDE_PATH=”/usr/local/include” export LIBRARY_PATH=”/usr/local/lib”

CUDA_PATH=/usr/local/cuda-11.3 export CUDAT_PATH

export LD_LIBRARY_PATH=$CUDA_PATH/lib64:$LD_LIBRARY_PATH export LD_LIBRARY_PATH=”$LD_LIBRARY_PATH:/usr/lib/x86_64-linux-gnu/”

export PATH=$JAVA_HOME/bin:$PATH export PATH=$CUDA_PATH/bin:$PATH

## Install Dev. IDE
https://linuxconfig.org/how-to-install-intellij-idea-on-ubuntu-20-04-linux-desktop

## Opencascade
 - download .tar file and unzip
 - install 3rd party first
### Install 3rd party

$ sudo apt install ffmpeg $ sudo apt install libfreeimage-dev $ sudo apt install libfreetype-dev $ sudo apt install libvtk6-dev libvtk6-qt-dev $ sudo apt install libocct-visualization-7.3 $ sudo apt install libvtk7-dev libvtk7-qt-dev $ sudo apt install tcl-vtk7 install qt creator $ sudo apt install libtbb-dev $ sudo apt install tcl $ sudo apt install libtclcl1-dev $ sudo apt install libxmu-dev libxmu-headers libxi-dev $ sudo apt install liblxi-dev lxi-tools # ??

https://wiki.qt.io/Install_Qt_5_on_Ubuntu

### Build Opencascade
 - cmake
 - make
 - make install

$ mkdir build $ cd build $ cmake .. $ make $ sudo make install ```