Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/aaronc81/archer
Machine instruction search engine
https://github.com/aaronc81/archer
architecture assembly ruby
Last synced: about 2 months ago
JSON representation
Machine instruction search engine
- Host: GitHub
- URL: https://github.com/aaronc81/archer
- Owner: AaronC81
- License: mit
- Created: 2024-11-03T18:03:57.000Z (3 months ago)
- Default Branch: main
- Last Pushed: 2024-12-22T22:36:37.000Z (about 2 months ago)
- Last Synced: 2024-12-22T23:26:31.636Z (about 2 months ago)
- Topics: architecture, assembly, ruby
- Language: Ruby
- Homepage: https://archer.computer
- Size: 4.4 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Archer
> **Hosted at [archer.computer](https://archer.computer)**
## What is this?
Archer is a web app which provides a searchable, filterable directory of machine instructions for a
variety of computer architectures.
I grew frustrated at a lack of standardised, consistent documentation for people writing Assembly.
You're largely stuck with clunky vendor websites or ancient PDFs.
Archer lets you answer questions like...
- Which instructions can I use to write x87 floating point data to memory?
- Select only "x87 FPU" capability, filter to "Memory: Store"
- What flag-pushing instructions does x86 have?
- Filter to "Operands: None" for input and output, and "Memory: Store"
- How do I transfer a value from an Address Register to a Data Register on TriCore?
- Filter to "Operands: Input: Address register" and "Operands: Output: Data register"
## Data sources
Archer's data is a combination of:
* Machine information dumped from **LLVM and some forks of it**
* Latest copy saved in `llvm/dump`
* Hand-written YAML "supplementary" files, to imbue more human-friendly meaning to the LLVM info
* In `supp`
The latest list of forks used can always be found on the homepage of
[archer.computer](https://archer.computer).
## Building
Archer is a static site once built, and can be used offline if you like.
You will need a reasonably up-to-date Ruby to build the site (tested on 3.2).
```
bundle install
bundle exec build_site.rb
```
## Refreshing LLVM data
To fetch or refetch the LLVM data dumps, install Docker then run:
```
cd llvm
./generate_tablegen_dumps.sh
```
This will take a fair bit of time, disk space, and RAM, as it creates multiple checkouts of LLVM and
partially builds one of them.
## Things to do
- **Ideas**
- [ ] _(Easy)_ "Clear filters" button
- [x] _(Medium)_ Copy link to specific instructions
- [ ] _(Medium)_ Copy link to search filter sets
- [ ] _(Medium)_ Dark mode
- [ ] _(Hard)_ Group instructions with the same mnemonic, rather than displaying separately
- [ ] _(Hard)_ Properly collect and show register class members
- **Issues**
- [ ] Mnemonic search includes entire text, not just the mnemonic
- [ ] Instruction parsing seems broken - lots of errors for x86
- **Architecture Support**
- [ ] ARM
- [ ] Rest of the x86 capabilities
- [ ] Rest of the RISC-V capabilities
- [ ] More PowerPC
## License
All Archer source code and supplementary data is licensed under the [MIT License](./LICENSE).
LLVM dumped data in `llvm/dump` originates from the LLVM project and forks of it, therefore being
licensed under the [Apache License v2.0 with LLVM Exceptions](https://github.com/llvm/llvm-project/blob/main/LICENSE.TXT)
or the [University of Illinois/NCSA Open Source License](https://github.com/TriDis/llvm-tricore/blob/tricore/LICENSE.TXT)
depending on the age of the fork. I believe both of these licenses are compatible with MIT.