earlephilhower / esp-quick-toolchain

GCC toolchain for esp8266/arduino on MacOS, Linux, ARM64, Raspberry Pi, and Windows

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

esp-quick-toolchain - Build ESP8266 toolchain for multiple architectures

Allows building Win32, Win64, OSX, Linux x86_64, ARM64 (aarch64) and Raspberry Pi ESP8266 toolchains in a Docker container.

Builds work for GCC 4.8, 7.2, 9.3, 10.1, and 10.2.

Downloading GCC/etc. sources

Run

make download

to clone the GCC and libs needed to the repo/ directory. This takes a while, so be patient, but it only is done once (and allows us to switch between GCC versions without redownloading anything in the future).

Building only native mode binaries

If you're only compiling natively, you can just clone this repo and run

make GCC={4.8|4.9|5.2|7.2|9.3|10.1|10.2} REL=x.x.x SUBREL=x -jx  # I like -j32 on a 16-core server, adjust according to your CPU power

Note that to build a non-linux toolchain, you first need to build a linux chain in the directory. This is because the cross compiler requires a local host executable gcc for the target architecture to build properly.

Building full suite architectures

To build all architectures use the commands

git clone https://github.com/earlephilhower/esp-quick-toolchain
cd esp-quick-toolchain
docker run --user $(id -u):$(id -g) --rm -v $(pwd):/workdir earlephilhower/gcc-cross bash -c "cd /workdir; make -j32 GCC={4.8|4.9|5.2|7.2|9.3|10.1|10.2} REL=2.5.0 SUBREL=3 all"

To make a draft release of the binaries:

make GCC={4.8|4.9|5.2|7.2|9.3|10.1|10.2} REL=2.5.0 SUBREL=3 upload

You then promote the draft to a pre-release so it becomes visible and can then make a PR against the Arduino core to merge it.

Then to install the libraries and headers into the Arduino core (not including the toolchain exes) just

make GCC={4.8|4.9|5.2|7.2|9.3|10.1|10.2} REL=2.5.0 SUBREL=3 install  (INSTALLBRANCH=xxx may be added to apply against a predefined branch other than master)
<in Arduino dir>
git commit -a

About

GCC toolchain for esp8266/arduino on MacOS, Linux, ARM64, Raspberry Pi, and Windows


Languages

Language:Makefile 80.9%Language:Shell 11.4%Language:Python 7.7%