Add a dedicated page for Google Summer of Code 2026 by hsiangkao · Pull Request #19 · erofs/docs

hsiangkao · 2026-03-10T08:56:31Z

Adds a new page describing Google Summer of Code (GSoC) participation and listing initial project ideas for the 2026 program.

Copilot

Pull request overview

Adds a new documentation page describing Google Summer of Code (GSoC) participation and listing initial project ideas for the 2026 program.

Changes:

Introduces src/gsoc.md with an overview of GSoC and proposal guidelines.
Adds an AI-usage disclosure requirement for proposals.
Starts a “Project Ideas for Google Summer of Code 2026” section with an initial idea entry.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

src/gsoc.md

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 6 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

src/gsoc.md

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 3 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

src/gsoc.md

Copilot · 2026-03-10T14:48:45Z

src/gsoc.md

+
+## Project Proposal Guidelines
+
+Students are responsible for writing a proposal and submitting it to Google


This uses “Students” here, but the page otherwise refers to “contributors”. Please use consistent terminology throughout (e.g., “contributors/applicants”) to match the rest of the page and avoid mixed wording.

Suggested change

Students are responsible for writing a proposal and submitting it to Google

Contributors/applicants are responsible for writing a proposal and submitting it to Google

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 6 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-10T15:10:04Z

src/gsoc.md

+1. (erofs-utils) Multi-threaded Decompression Support in fsck.erofs
+
+    Proposed mentors: Yifan Zhao, Gao Xiang


Tool names like fsck.erofs are written as plain text here, but elsewhere in the docs command names are formatted as code (backticks). Please format fsck.erofs consistently throughout this section to improve readability.

Copilot · 2026-03-10T15:10:04Z

src/gsoc.md

+    Its userspace tool, fsck.erofs, is critical for: filesystem integrity
+    checking, image unpacking and regression testing.


This sentence uses an odd colon after “critical for”, which reads like a list label but continues mid-sentence. Consider rewriting as a normal sentence/list without the colon for clarity.

Suggested change

Its userspace tool, fsck.erofs, is critical for: filesystem integrity

checking, image unpacking and regression testing.

Its userspace tool, fsck.erofs, is critical for filesystem integrity

checking, image unpacking, and regression testing.

Copilot · 2026-03-10T15:10:05Z

src/gsoc.md

+
+    **Description**
+
+    *TODO*


This project idea is currently a placeholder (“TODO”). If this page is intended to be published as a complete GSoC 2026 landing page, please either fill in the description or explicitly mark it as “TBD” with a timeline/where applicants should look for updates.

Suggested change

*TODO*

TBD – The detailed scope for extending the `erofs-rs` functionality is still being finalized for GSoC 2026. We expect to publish a full project description closer to the application period. In the meantime, interested applicants should monitor the project repository and participate in the development mailing list or Matrix room (see “Project Proposal Guidelines” above) for updates and to discuss potential directions.

Copilot · 2026-03-10T15:10:05Z

src/gsoc.md

+
+    **Description**
+
+    For local builds, currently mkfs.erofs only supports generating from source


mkfs.erofs should be formatted as inline code (backticks) to match the style used elsewhere in the docs (e.g., mkfs.erofs in src/mkfs.md and src/faq.md).

Suggested change

For local builds, currently mkfs.erofs only supports generating from source

For local builds, currently `mkfs.erofs` only supports generating from source

src/gsoc.md

Copilot

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 5 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

src/gsoc.md

Copilot · 2026-03-10T15:24:44Z

src/gsoc.md

+*TODO*
+


This project idea section is left as *TODO*, which makes the GSoC ideas page incomplete. Please either flesh out the description (scope, goals, prerequisites) or remove the idea until it’s ready.

Suggested change

*TODO*

`erofs-rs` is a Rust implementation of the EROFS (Enhanced Read-Only File System)

on-disk format. The goal of this project is to significantly extend its basic

functionality so that it can be reliably used for reading and inspecting EROFS

images from userspace tools and automated pipelines.

The project will focus on improving core read-only functionality rather than

advanced or experimental features. Depending on your interests and prior

experience, potential areas of work include:

- Implementing support for additional on-disk features (e.g. compression

variants, inline data, extended attributes) that are currently missing

or only partially supported in `erofs-rs`;

- Improving directory and inode handling, including edge cases and error

reporting, so that a full EROFS tree can be traversed reliably;

- Enhancing metadata validation and consistency checking to make `erofs-rs`

robust when dealing with malformed or adversarial images;

- Adding or extending a small CLI tool (or library API) for inspecting and

dumping EROFS images using `erofs-rs`;

- Writing comprehensive unit tests and integration tests against a corpus of

real-world EROFS images (e.g. Android system images, container images)

and cross-validating behavior against the reference Linux implementation.

A successful project should deliver:

- A set of concrete new features in `erofs-rs` that are documented and

covered by tests;

- Demonstrated compatibility with standard EROFS images produced by existing

tooling;

- Improved documentation explaining how to use `erofs-rs` in other Rust

projects or from the command line.

While prior experience with EROFS is not required, you should:

- Be comfortable writing idiomatic Rust and working with low-level data

structures and binary formats;

- Have a basic understanding of filesystem concepts (inodes, directories,

block layout, metadata vs. data blocks);

- Be willing to read and understand existing code and upstream EROFS

documentation and specifications.

The exact scope will be refined with your mentors based on your experience and

the available time, with an emphasis on shipping well-tested, maintainable

functionality rather than a large but fragile feature set.

src/gsoc.md

Copilot

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

src/gsoc.md

Copilot

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 6 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

src/roadmap.md

src/gsoc.md

Copilot · 2026-03-11T02:04:29Z

src/gsoc.md

+Proposed mentors: Chunhai Guo, Yifan Zhao, Gao Xiang  
+Languages: C  
+Estimated project length: 350 hours  
+Difficulty: hard  


The roadmap previously linked this idea to a tracking issue. Since roadmap now points to gsoc.md, consider adding a “Tracking issue:” link here (e.g., to erofs-utils issue #33) so readers can find background discussion and current status.

Suggested change

Difficulty: hard

Difficulty: hard

Tracking issue: [erofs-utils#33](https://github.com/erofs/erofs-utils/issues/33)

Copilot · 2026-03-11T02:04:29Z

src/gsoc.md

+Languages: C  
+Estimated project length: 350 hours  
+Difficulty: medium  
+Skills:  


The roadmap previously linked this idea to a FreeBSD-src tracking issue. Since roadmap now points to gsoc.md, consider adding a “Tracking issue:” link here to preserve that context and make it easy to follow ongoing work.

Suggested change

Languages: C

Estimated project length: 350 hours

Difficulty: medium

Skills:

Tracking issue: _TBD_

Languages: C

Estimated project length: 350 hours

Difficulty: medium

Copilot

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

src/gsoc.md

speedan1

Looks good to me.
Thanks,

Copilot

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 3 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

src/roadmap.md

src/gsoc.md

src/roadmap.md

Copilot

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 3 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-11T02:44:20Z

src/gsoc.md

+`This proposal was written with the assistance of [ChatGPT/Gemini/etc] to [check spelling / check accuracy / format text].`
+
+or:
+
+`This proposal was written without the use of any AI tools.`


The AI-disclosure statements are currently formatted as long inline code spans, which tend to render poorly (no wrapping / horizontal scrolling) and are harder to copy as a block. Consider converting these to a fenced code block (or a {code-block}) with two separate lines for the alternatives.

src/gsoc.md

Copilot · 2026-03-11T02:44:20Z

src/gsoc.md

+*TBD*
+


This project idea currently has no details beyond "TBD". If it’s intended to be listed publicly as an initial GSoC idea, consider adding at least a short problem statement and a couple of concrete deliverables (or remove it until it’s fleshed out) so applicants can scope proposals.

Suggested change

*TBD*

erofs-rs is a Rust implementation of the EROFS read-only filesystem, intended

to provide a safe and modern codebase that can be integrated into container

runtimes, user space tooling, and other systems that benefit from EROFS’s

space-efficiency and performance characteristics.

Today, erofs-rs lacks several features that are present in the upstream

Linux kernel EROFS implementation, as well as the surrounding tooling. This

limits its applicability for production use and makes it harder to adopt as a

drop-in alternative in ecosystems that expect “full” EROFS support.

The primary goal of this project is to extend erofs-rs so that it supports a

significantly more complete set of EROFS on-disk features, along with robust

testing and documentation, bringing it closer to production readiness.

Key objectives:

- Identify and implement a set of missing core EROFS features in erofs-rs

(for example: additional compression modes, inline/xattrs handling,

extended metadata features, or layout optimizations), in coordination with

mentors to define a realistic scope;

- Add or extend automated tests (unit tests and integration tests) that

validate compatibility of erofs-rs with images produced and consumed by

existing EROFS tooling;

- Introduce basic performance benchmarks and profiling to compare erofs-rs

with reference implementations on representative workloads, and optimize

identified hot paths where feasible;

- Improve developer and user documentation for erofs-rs, including build and

usage instructions, feature matrix, and guidance on integration into

higher-level systems (e.g., container runtimes or image builders).

Signed-off-by: Gao Xiang <xiang@kernel.org>

Copilot AI review requested due to automatic review settings March 10, 2026 08:56

Copilot started reviewing on behalf of hsiangkao March 10, 2026 08:56 View session