分享 TinyVision 板子SDK编译的docker文件

[iamseer]

最近折腾了两天终于搞定了SDK编译，而且成功生成了img也烧录启动了。分享兼备份一下docker文件。毕竟docker比vmware这类的方便迁移多了。

这里我跳过docker的介绍和安装过程。首先需要确认docker的内存要给足。我观察到编译到opencv的dnn模块时内存占用能上升到26G，如果内存不够，编译会失败。我使用的这台电脑有16G内存而docker默认是使用50%的内存以及25%的swap空间。这样一共12G是不够的。我设置内存分配12G和20G的swap空间，这样一共32G，可以编译。

另外说明一点，这台主机的所有源全部可以直连，如果你不能的话需要添加或者更改源。另外Dockerfile里面的用户名密码要改成你的。

如果你不用docker，也可以参考一下Dockerfile来手工安装。

Dockerfile

Compose.yaml

容器启动以后，首先可以执行top看看内存和swap够不够。然后依次执行：

# supress some git warning
git config --global --add safe.directory /tina-v853-open/brandy/brandy-2.0/u-boot-2018
source build/envsetup.sh
echo "Y" | lunch v851se-tinyvision-tina
# setup needed, can just save and exit
m menuconfig 
# for debug: make -j1 V=s
make -j4
pack
#then copy the compiled img to host and use Phoenixcard 

然后就可以把生成的img文件复制到主机上，然后用Phoenixcard烧写TF卡。如果你是用随附的yaml构建容器，那复制到/opt/script里就能出现在主机里了。

另外附上Dockerfile内容，谨防附件失效。

# Use the Ubuntu 20.04 minimal image as the base
FROM ubuntu:20.04

# Set environment variable to prevent tzdata asking for geographic area
ENV DEBIAN_FRONTEND=noninteractive

# Install any dependencies or packages required by your application
RUN apt-get -y update

# install dependencies
RUN apt-get install -y build-essential subversion git libncurses5-dev zlib1g-dev gawk flex bison quilt libssl-dev xsltproc libxml-parser-perl mercurial bzr ecj cvs unzip lsof python3 python2 python3-dev android-tools-mkbootimg python2 libpython3-dev

# add i386 architecture
RUN dpkg --add-architecture i386
RUN apt-get update
RUN apt-get install -y gcc-multilib 
RUN apt-get install -y libc6:i386 libstdc++6:i386 lib32z1

# Install Repo tool
RUN curl https://storage.googleapis.com/git-repo-downloads/repo > /usr/local/bin/repo
RUN chmod a+x /usr/local/bin/repo

# Start to set up the tina-v853 SDK
RUN mkdir /tina-v853-open
WORKDIR /tina-v853-open

# # set up git credential
RUN git config --global credential.helper store

RUN git config --global user.name "名字"
RUN git config --global user.email "邮箱"

# Add credentials to credential store
RUN echo "https://aw-ol的用户名:aw-ol的密码@sdk.aw-ol.com" > ~/.git-credentials

ENV REPO_COLOR=0

#initialize the repo
RUN repo init -u https://sdk.aw-ol.com/git_repo/V853Tina_Open/manifest.git -b master -m tina-v853-open.xml

#sync the repo, took about 30 minutes in midnight
RUN repo sync

#start SDK
RUN repo start devboard-v853-tina-for-awol --all

RUN curl -LO https://github.com/YuzukiHD/TinyVision/releases/download/tina.0.0.2/tina-bsp-tinyvision.tar.gz

RUN tar xvf tina-bsp-tinyvision.tar.gz

#init SDK

# Use bash to run commands
RUN apt-get install -y python wget rsync vim bc u-boot-tools cpio busybox

#RUN ls openwrt/package/allwinner/vision/libAWIspApi/machinfo/v851se/
#make sure it exists
RUN cat openwrt/package/allwinner/vision/libAWIspApi/machinfo/v851se/build.mk

#root can compile
ENV FORCE_UNSAFE_CONFIGURE=1