[GSoC 2019] Weeks #1, #2 and #3 progress reports
Hello everybody! It's been a while since my previous post.
This post is a short summary of all the work done upto this point. For a detailed report of every day, visit this blog I set up using Blogger.
As mentioned in my proposal, the target for week #1 was document the current btrfs implementation. As part of this, I documented the following classes
Week #1 went smoothly without any problems.
Weeks 2 and 3
The target for weeks 2 and 3 was to implement write support for files that occupy less than one block. However, it turns out that before implementing this, creating files must be implemented. Therefore after a discussion with my mentor (Hy Che) over IRC, I decided to work on implementing file creation first.
Also during this week I learnt various advanced git concepts such as stashing, refspecs, rewriting history, filter branch etc. Before Google Summer of Code I just knew the basics like commit, clone and pull. For this used the textbook Pro Git. I would like to thank Haiku for this wonderful opportunity to improve my git skills.
I also went through the Haiku API documentation (a.k.a Haiku Book) during these two weeks. In particular, I thoroughly studied the documentation under Storage Kit and Device Drivers.
However, not all things went smoothly. In order to learn how to implement file creation, my mentor (Hy Che) suggested that I go through Linux's btrfs implementation. I spent most of week #3 going through Linux's btrfs code for creating, modifying, writing and linking files (see Blogger blog for detailed reports, including pseudocode I was able to deduce based on the C code). Even after looking into Linux's btrfs implementation, I wasn't confident enough to code up my own implementation. As a result, I might have to do things differently from what I mentioned in my proposal.
That brings me to what I'm currently doing - reading up the papers published by Ohad Rodeh on btrfs. So far I've read the following papers
- B-trees, Shadowing, and Range-operations
- B-trees, Shadowing, and Clones
- Deferred Reference Counters for Copy-On-Write B-trees
I plan on finishing reading BTRFS: The Linux B-tree Filesystem by tomorrow.
After reading these papers, I definitely feel more confident about my understanding of btrfs. I will try to finish implementing creating new files by the end of this week.
So far the learning curve has been quite steep, if I say so myself. I find GSoC to be both challenging and fun. Once again, I'd like to take a moment to thank the Haiku community for giving me this amazing opportunity.
- No, I'm not Haiku's lead developer
- Haiku monthly activity report - 12/2019
- Haiku almost-monthly activity report - October and November 2019
- Haiku monthly activity report - September 2019
- Node.js now available in Haiku
- Haiku monthly activity report - 08/2019
- GSOC 2019 Final Report
- Haiku Activity Report: Performance Edition
- new PVS studio scan
- Coding week 4,5,6