mirror of https://github.com/zeldaret/tp.git
59 lines
4.1 KiB
Markdown
59 lines
4.1 KiB
Markdown
Contributing
|
|
=================
|
|
Ghidra Setup
|
|
-----
|
|
1. Install [Java 11 64-bit Runtime and Development Kit](https://www.oracle.com/java/technologies/javase/jdk11-archive-downloads.html).
|
|
2. Download Ghidra from [here](https://github.com/NationalSecurityAgency/ghidra/releases).
|
|
3. Open a pull request against [this repo](https://github.com/zsrtp/tp-ghidra-server), adding your desired username and permissions level into users.yaml
|
|
4. Login using the following server info:
|
|
|
|
```
|
|
Server: ghidra.tpgz.io
|
|
Port: 13100
|
|
Username: <your_username>
|
|
Password: changeme
|
|
```
|
|
|
|
You will be prompted to change your password at login. After logging in, you will see a list of all the Twilight Princess DOLs for each version.
|
|
|
|
|
|
|
|
Getting Started
|
|
-----
|
|
To begin, select a file to begin work on. You can check what files are available on [Trello](https://trello.com/b/Y04SAxbw/twilight-princess-decompilation).
|
|
|
|
This project has been automatically generated using `dol2asm`. This process generates blank headers for every source file, default includes, necessary types to be used within the file, forward references, and external references of data / functions used within the file's functions. Additionally, all data and functions have been split and generated with default types.
|
|
|
|
When decompiling a file, you will need to manually fix structs/classes, function return types, data types, and data definitions. Parameters are already setup for any function with a mangled name. Some functions / data may need to be changed to be auto-generated by the compiler, such as by inlining functions or making data in-function static.
|
|
|
|
The TP Nvidia Shield Debug Rom contains extra useful info, as it was compiled with less aggressive function inlining, and also contains debug tools / debug strings. While the primary focus currently is on the NTSC-U GameCube version, it is highly recommended to reference the Debug rom for help / documentation.
|
|
|
|
### diff.py
|
|
`diff.py` is the tool we use to fix matching errors. You can invoke `diff.py` from command line like:
|
|
`python diff.py -mwo mangled_symbol_name`
|
|
- `-m` will run make before opening diff.
|
|
- `-w` will automatically update the diff when you save your changes.
|
|
- `-o` will diff only the necessary `.o` file, which is recommended for this project.
|
|
- `-f` can be used to specify a specific object file to use when diffing with `-o` (Used to diff rels)
|
|
- You can find the mangled symbol name in the function info comment. (Ex. `onEventBit__11dSv_event_cFUs`)
|
|
|
|
NOTE: You must run `./tp expected` with an OK build before diffing to have a correct binary to compare to.
|
|
|
|
You can install the requirements for `diff.py` by running `pip install -r tools/requirements.txt`.
|
|
|
|
### tp.py
|
|
`tp.py` is a multi-purpose tool that can be used for things such as checking SHA1 checksums, current decompilation progress, and removing unused assembly files.
|
|
You can invoke `tp.py` from command line either directly in the `tools/` folder, or by running the `tp` bash script.
|
|
Useful commands:
|
|
- `check` - Compares the SHA1 checksums of your current build to the ones of the expected build. Add `--rels` to check rel checksums.
|
|
- `expected` - Copies your current build folder to the expected build folder.
|
|
- `progress` - Calculates and displays the decompilation progress of the main dol. Add `--rels` to calculate rel progress.
|
|
- `remove-unused-asm` - Finds and removes all assembly files that are no longer being used.
|
|
- `pull-request` - Verifies that everything is OK before you set up a pull request.
|
|
|
|
### decomp.me
|
|
[decomp.me](https://decomp.me/) is a very useful website for sharing individual functions with other contributers.
|
|
To use decomp.me with TP, select the `GameCube / Wii` platform and `Twilight Princess` compiler preset. Enter your function's mangled name under `Function name`, and copy / paste the function assembly from the project into the `Target assembly` box. Provide any necessary context within the provided box, then create your scratch.
|
|
|
|
From here you can use the built-in editor to write your function code. Make sure to save the scratch when you are done, then you can copy and share the URL to receive help matching the function.
|