Skip to content

Latest commit

 

History

History
237 lines (156 loc) · 6.72 KB

manual-2.md

File metadata and controls

237 lines (156 loc) · 6.72 KB
Raw

Instructions for the usage

Start the Container

Pull the image

After the docker installation is complete, restart your computer.

Open a terminal and enter the following command to pull the course image from dockerhub.

docker pull zhangshen777/ics-ubuntu

In ubuntu, if you have got the problem "permission denied", you can try to pull it with administrator privileges.

sudo docker pull zhangshen777/ics-ubuntu

Basic Usage

In the terminal, enter the following command to start a container with the course image.

docker run -it -p 5001:22 zhangshen777/ics-ubuntu /bin/bash

If you fail to start the container, check whether you have pulled the image successfully using this command.

docker images -a

The image repository should include "zhangshen777/ics-ubuntu", like the following picture.

If the container is started successfully, it will enter the container's bash smoothly, as shown in the following image.

By enter exit, you can exit the bash and meanwhile the container will be shut down by the docker.

If you want to start the container again, you should check the container's ID firstly. The numbers in the red square is just your container's ID.

docker ps -a

Then, you can start your container with this command.

Note!!! The ID in the command is different for everyone, and you must replace the ID with your own ID. 
docker satrt a85e338feb2c

After starting the container in this way, it will not automatically go to the terminal of ubuntu. If needed, we need to start the terminal manually.

docker exec -it a85e338feb2c /bin/bash

If you want to close the container , you can use the following command to close it.

docker stop a85e338feb2c

Connect to the Container

Start the container.

docker satrt a85e338feb2c

Connect to the linux server with VS-Code via SSH.

Input the ssh address in the pop-up window. The "123456" in the command is the password for user "docker".

ssh://docker:123456@localhost:5001

Once the configuration is complete, click on the location in the image below to connect the container.

When you click on it, a new window will usually pop up and you will need to enter your password. Enter password "123456" to get authorization. After that, click on "Open Folder" and select the path, "/home/docker/workspace".

Necessary Setup

Config Git

After the connection to the container, open the terminal of the container in VS-Code.

In the terminal, config the git like the following command.

# config the user name with your real name
git config --global user.name "2021xxxxx-ZhangSan"
# config the user email with your real email
git config --global user.email "[email protected]"

Fill your information

After the connection to the container, you can view the files in the container.

First of all, you should fill your information in /home/docker/workspace/ics/myinfo.txt

Note!!! If you did not fill the information, you could not be able to use the environment.

Verilator Demo

Open a new terminal and input the following commands. Note: When executing the second command, the terminal will ask you to input the test cycles and you can randomly input a number.

cd ics
./build.sh -e demo -b -s

After that, there will be a new folder, build_test in the dir /home/docker/workspace/ics/projects/demo. There are some intermidate simulation files and the final vcd waveform file. With the help of wavetrace, we can directly view the vcd waveform in VS-Code.

You can also download the vcd waveform file and view the waveform in you local desktop using gtkwave.

Pytorch Test

The miniconda3 has been installed in the iamge and the pytorch has been set up. By executing the following command in the terminal, you can activate the specific python environment.

conda activate torch

Usually we develop the python code in jupyter. The jupyter has also been set up in the container. And there are two methods that you can use to start your work in jupyter.

Method 1

With the jupyter extension for VS-Code which has been installed, you can directly open the ipynb files in VS-Code and edit them. When you open one ipynb file, you can sellect the python version in the top right corner. Then after selecting the torch, the VS-Code will start the jupyter server thread.

Method 2

Another methond is to use the website browser. This method requires the help of the extension for VS-Code, ssh-extension. With the forwarded ports via ssh, we can use local desktop to view and edit the website page in the contianer.

Open a new terminal of the container, start the jupyter lab or jupyter notebook.

# activate torch
conda activate torch

# enter the directory
cd /home/docker/workspace/ics/projects/torch

# start the jupyter-lab server
jupyter lab

# If you prefer to use notebook, start the jupyter-notebook server
# jupyter notebook

Then in your local destop, open the website browser and visit the URL, http://127.0.0.1:8888/lab.

Then, you can see this page of jupyter-lab.

NOTE!!! If the password authentication screen pops up when you visit the website, you can enter the password 123456 for authentication.

After the usage of jupyter-lab or jupyter-notebook, you can enter ctrl+c to end the jupyter server.