HPCRunner Contribution Guide

The HPCRunner project welcomes your expertise and enthusiasm!

Small improvements or bug fixes are always appreciated. Updating documentation is a great place to start. If you are considering a major contribution, please open an issue or start a discussion at HPCRunner.

Writing code is not the only way to contribute to Jarvis. You can also:

Contribute installation scripts.
Help test new HPC applications.
Develop tutorials and demos.
Spread the word about Jarvis.
Help onboard new contributors.

Features

Supports Arm and x86 architectures. Uses industry-standard dependency directory structures to manage massive dependencies and automatically generate environment module files.
One-click compilation, execution, CPU/GPU performance profiling, and benchmarking based on HPC configurations.
Stores all settings in a single file. Deploying HPC applications to different machines only requires modifying this configuration file.
Adopts an integrated log management system to automatically record all information during the HPC application deployment process.
Requires no compilation and runs directly with a Python environment.
Supported for Quantum Espresso (QE). For details, refer to the container/ directory.
Roadmap (future features)
- Integration of common performance tuning methods and core algorithms in the HPC field.
- Cluster performance analysis tools.
- Intelligent auto-tuning.

Code Contribution Guide

Preparations

Configure the SSH key.
```
cd ~/.ssh
ssh-keygen
cat id_rsa.pub
```
Copy the output content and add it to your profile via SSH Key.

Configure user settings.

# Ignore file mode (permission) changes.
git config core.fileMode false
git config --global user.name "XXX"
git config --global user.email "XXX"

Fork the repository and clone your fork. The following uses iotwins as an example. Replace it with your actual username.
```
git clone git@gitee.com:iotwins/hpcrunner.git
```

Link your local repository to the upstream remote.

git remote add upstream git@gitee.com:openeuler/hpcrunner.git

Code Contribution Workflow

Switch to the master branch and pull the latest upstream changes.
```
git checkout master
git pull upstream master
```
Create and switch to a new feature branch. The branch name is custom.
```
git checkout -b new_branch
```
Make your code modifications.

Commit and push your changes

git add .
git commit --no-verif -m "Add XXX function"
git push origin new_branch

Go to your forked repository page (for example, https://gitee.com/iotwins/hpcrunner) and create a Pull Request (PR).

FAQ

How Do I Fix an Incorrect Commit Message?

Method 1: Using git stash

If you have uncommitted changes in your workspace or staging area, use git stash to temporarily shelf them. This leaves you with a clean working directory to handle fixes or syncs. Afterward, restore your work using git stash pop to continue development.
Method 2: Using git rebase

Run the interactive rebase command targeting the parent commit of the one you want to modify:
```
git rebase -i <commitID>  # Example: git rebase -i sd98dsf89sdf
```
In the interactive rebase editor that opens:
1. Locate the commit line you want to change.
2. Change the command from pick to edit at the beginning of the line (you can do this for one or multiple commits).
3. Save and close the editor.
Depending on your goal, use --amend to modify the commit:
- Modify the commit message only.
```
git commit --amend
```
- Modify the author and email only.
```
git commit --amend --author="zhangsan <hello@gmail.com>" --no-edit
```
- Modify the message, author, and email simultaneously.
```
git commit --amend --author="zhangsan <hello@gmail.com>"
```
Once modified, resume the rebase process:
```
git rebase --continue
```
Repeat until you see the following confirmation message, indicating the rebase is complete:
```
Successfully rebased and updated xxx
```
Method 3: Using git push --force

To apply the rewritten history to your remote repository, force push to your branch:
```
git push --force origin master 
```

Developer Notes

Using HPCKit Instead of Standalone Installations of Hyper MPI, BiSheng Compiler, and KML

Recommended approach for installing and enabling HPCKit:

./jarvis -install hpckit/${HPCKIT_VERSION} any
module use software/utils/hpckit/224.0.0/HPCKit/24.0.0/modulefiles
module purge
module load bisheng/compiler${BISHENG_VERSION}/bishengmodule bisheng/hmpi${HMPI_VERSION}/hmpi

Not recommended:

./jarvis -install hmpi/1.1.1 clang
module use software/moduledeps/bisheng2.1.0
module load hmpi/1.1.1

Avoiding Hardcoded Absolute Paths in Development Templates

Not recommended:

export JARVIS_ROOT=/hpcrunner

Technical Articles & Resources

Demystifying HPC Applications

A Date with Containers

Jarvis: The Ultimate HPC Application Steward

For more information on HPC deployment and performance optimization, please scan and join the openEuler HPC SIG WeChat Group.

WeChat Group