[Tegra] 如何安裝 CUDA , cuDNN 和 OpenCV－熊熊的部落格

如何安裝 CUDA , cuDNN 和 OpenCV

1.下載 CUDA

版本 8.0

wget "http://developer.download.nvidia.com/devzone/devcenter/mobile/jetpack_l4t/013/linux-x64/cuda-repo-l4t-8-0-local_8.0.84-1_arm64.deb"

版本 9.0

wget "https://developer.download.nvidia.com/devzone/devcenter/mobile/jetpack_l4t/3.2.1/m8u2ki/JetPackL4T_321_b23/cuda-repo-l4t-9-0-local_9.0.252-1_arm64.deb"

版本 10.0

wget "https://developer.nvidia.com/assets/embedded/secure/tools/files/jetpack-sdks/jetpack-4.2/JETPACK_42_b158/cuda-repo-l4t-10-0-local-10.0.166_1.0-1_arm64.deb"

2. 安裝 CUDA

執行下面指令後再重新開機:

# Version 8.0

./cuda-l4t.sh cuda-repo-l4t-8-0-local_8.0.84-1_arm64.deb 8.0 8-0

# Version 9.0

./cuda-l4t.sh cuda-repo-l4t-9-0-local_9.0.252-1_arm64.deb 9.0 9-0

檔案 cuda-l4t.sh

#!/bin/bash
# $1: cuda debian file name 
# $2: cuda_ver (example 6.0, 6.5)
# $3: cuda_dash_ver (6-0, 6-5)

if [ $# != 3 ] ; then
    echo "Incorrect arguments, use following command to install cuda"
    echo "$0 cuda_deb_file_name 6.5 6-5"
    exit 1
fi

while sudo fuser /var/lib/dpkg/lock > /dev/null 2>&1; do
    echo "Waiting for other apt-get command to finish"
    sleep 3
done

# 原始檔案
#sudo dpkg --force-all -i ~/cuda-l4t/$1
# 修改過
sudo dpkg --force-all -i $1
sleep 5
sudo apt-key add /var/cuda-repo-$3-local/*.pub
sleep 2
sudo sed -i.bak 's/\(^deb.*main restricted\)\s*$/\1 universe multiverse/g' /etc/apt/sources.list
sudo sed -i.bak 's/\(^deb.*main restricted universe\)\s*$/\1 multiverse/g' /etc/apt/sources.list

sudo apt-get -y update
sudo apt-get -y --allow-downgrades install cuda-toolkit-$3 libgomp1 libfreeimage-dev libopenmpi-dev openmpi-bin

grep -q "export PATH=.*/usr/local/cuda-$2/bin" ~/.bashrc || echo "export PATH=/usr/local/cuda-"$2"/bin:$PATH">>~/.bashrc

if dpkg --print-architecture | grep -q arm64; then 
    lib_dir=lib64
else
    lib_dir=lib
fi
grep -q "export LD_LIBRARY_PATH=/usr/local/cuda-$2/$lib_dir" ~/.bashrc || echo "export LD_LIBRARY_PATH=/usr/local/cuda-"$2"/"$lib_dir":$LD_LIBRARY_PATH" >> ~/.bashrc
export LD_LIBRARY_PATH=/usr/local/cuda-$2/$lib_dir:$LD_LIBRARY_PATH

3. 驗證 CUDA 是否安裝成功

nvcc --version

4. 下載 cuDNN

# Version 6.0

wget "http://developer.download.nvidia.com/devzone/devcenter/mobile/jetpack_l4t/013/linux-x64/libcudnn6_6.0.21-1+cuda8.0_arm64.deb"
wget "http://developer.download.nvidia.com/devzone/devcenter/mobile/jetpack_l4t/013/linux-x64/libcudnn6-dev_6.0.21-1+cuda8.0_arm64.deb"
wget "http://developer.download.nvidia.com/devzone/devcenter/mobile/jetpack_l4t/013/linux-x64/libcudnn6-doc_6.0.21-1+cuda8.0_arm64.deb"

# Version 7.1

wget "https://developer.download.nvidia.com/devzone/devcenter/mobile/jetpack_l4t/3.2.1/m8u2ki/JetPackL4T_321_b23/libcudnn7_7.1.5.14-1+cuda9.0_arm64.deb"
wget "https://developer.download.nvidia.com/devzone/devcenter/mobile/jetpack_l4t/3.2.1/m8u2ki/JetPackL4T_321_b23/libcudnn7-dev_7.1.5.14-1+cuda9.0_arm64.deb"
wget "https://developer.download.nvidia.com/devzone/devcenter/mobile/jetpack_l4t/3.2.1/m8u2ki/JetPackL4T_321_b23/libcudnn7-doc_7.1.5.14-1+cuda9.0_arm64.deb"

5. 安裝 cuDNN

# Version 6.0

sudo dpkg -i ./libcudnn6_6.0.21-1+cuda8.0_arm64.deb
sudo dpkg -i ./libcudnn6-dev_6.0.21-1+cuda8.0_arm64.deb
sudo dpkg -i ./libcudnn6-doc_6.0.21-1+cuda8.0_arm64.deb

# Version 7.1

sudo dpkg -i ./libcudnn7_7.1.5.14-1+cuda9.0_arm64.deb
sudo dpkg -i ./libcudnn7-dev_7.1.5.14-1+cuda9.0_arm64.deb
sudo dpkg -i ./libcudnn7-doc_7.1.5.14-1+cuda9.0_arm64.deb

6. 確認是否有安裝

確認這個檔案 /usr/include/cudnn.h 是否存在