fxlibc/README.md
2022-08-21 17:11:28 +02:00

126 lines
4.5 KiB
Markdown
Raw Permalink Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

# fxlibc: The FX C Library
This directory contains the sources of the FxLibc Library. See `CMakeLists.txt`
to see what release version you have.
The FxLibc is the standard system C library implementation for all Casio
fx calculators, and is an important part of what makes up programs on these
devices. It provides the system API for all programs written in C and
C-compatible languages such as C++ and Objective-C; the runtime facilities of
other programming languages use the C library to access the underlying operating
system.
---
## Dependencies
FxLibc requires a GCC compiler toolchain the PATH to build for any calculator.
You cannot build with your system compiler! The tutorial on Planète Casio
builds an `sh-elf` toolchain that supports all models using multilib. See also
[Lephenixnoir/sh-elf-gcc](/Lephenixnoir/sh-elf-gcc).
For Vhex targets, the headers of the kernel are also required (but not for gint; the fxlibc is installed before gint).
---
## Building and installing FxLibc
FxLibc supports several targets:
* Vhex on SH targets (`vhex-sh`)
* CASIOWIN for fx-9860G-like calculators (`casiowin-fx`)
* CASIOWIN for fx-CG-series calculators (`casiowin-cg`)
* gint for all calculators (`gint`)
Each target supports different features depending on what the kernel/OS
provides.
For automated gint/fxSDK setups using [GiteaPC](/Lephenixnoir/GiteaPC) is recommended. The instructions below are for manual installs.
#### Configuration
Configure with CMake; specify the target with `-DFXLIBC_TARGET`. For SH
platforms, set the toolchain to `cmake/toolchain-sh.cmake`.
You can either install FxLibc in the compiler's `include` folder (for Vhex), or another folder of your choice (eg. the fxSDK sysroot). If you choose non-standard folders you might need `-I` and `-L` options to use the library.
```bash
# Install in the compiler's include folder
% PREFIX="$(sh-elf-gcc -print-file-name=.)"
# Install in the fxSDK sysroot
% PREFIX="$(fxsdk path sysroot)"
# Custom target
% PREFIX="$HOME/.sh-prefix/"
# For gint, do not specify anything, the fxSDK will be used dynamically
```
Example for a static Vhex build:
```bash
% cmake -B build-vhex-sh -DFXLIBC_TARGET=vhex-sh -DCMAKE_TOOLCHAIN_FILE=cmake/toolchain-sh.cmake -DCMAKE_INSTALL_PREFIX="$PREFIX"
```
Or for a traditional gint/fxSDK build:
```bash
% cmake -B build-gint -DFXLIBC_TARGET=gint -DCMAKE_TOOLCHAIN_FILE=cmake/toolchain-sh.cmake
```
#### Build and install
Build in the directory specified in `cmake -B`.
```bash
% make -C build-X
% make -C build-X install
```
---
### Contributing
Bug reports, feature suggestions and especially code contributions are most
welcome.
If you are interested in doing a port, or want to contribute to this project,
please, try to respect these constraints:
* Document your code.
* One function per file (named like the function).
* Use the same formatting conventions as the rest of the code.
* Only use hardware-related code (DMA, SPU, etc) in target-specified files
when the target explicitly supports it.
---
### Using FxLibc
Include headers normally (`#include <stdio.h>`); on SH platforms where
`sh-elf-gcc` is used, link with `-lc` (by default the `-nostdlib` flag is
used for a number of reasons).
If you're installing in a custom folder, you also need `-I "$PREFIX/include"`
and `-L "$PREFIX/lib"`. If you're installing in the GCC install folder, you
don't need `-I` but you still need the `-L` as the default location for
libraries is at the root instead of in a `lib` subfolder.
---
### Licences
This work is licensed under a CC0 1.0 Universal License. To view a copy of this
license, visit: https://creativecommons.org/publicdomain/zero/1.0/legalcode.txt
Or see the LICENSE file.
FxLibc also includes third-party code that is distributed under its own
license. Currently, this includes:
* A stripped-down version of the [TinyMT random number generator](http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/TINYMT/index.html)
([GitHub repository](https://github.com/MersenneTwister-Lab/TinyMT)) by
Mutsuo Saito and Makoto Matsumoto. See `3rdparty/tinymt32/LICENSE.txt`.
* A stripped-down version of the [Grisu2b floating-point representation
algorithm](https://www.cs.tufts.edu/~nr/cs257/archive/florian-loitsch/printf.pdf)
with α=-59 and γ=-56, by Florian Loitsch. See `3rdparty/grisu2b_59_56/README`
for details, and [the original code here](https://drive.google.com/open?id=0BwvYOx00EwKmejFIMjRORTFLcTA).
---
### Special thanks to
* Lephenixnoir - For all <3
* Kristaba - For the idea with the shared libraries workaround !