EMBA on ARM64 based MacOS

The first time someone from our community was interested in getting EMBA fully up and running on ARM dates back to the beginning of 2023 (see here). While it would be great to see EMBA running on ARM we never had the need to get it up and running on ARM. So, this issue was rated as wontfix and unsupported and no further effort was spent on it.

A few days ago our well know EMBA user, contributor and open-source maintainer @floyd-fuh brought this topic back to the table. The great thing is that he found a quite simple way to run EMBA on a Mac.

The following description is from the original post:

I'm using Ubuntu 24.04 ARM64 with Rosetta (default Parallels click and install). Rosetta is the translation layer of Apple to run amd64 on ARM64.

The EMBA installer only failed because docker pull by default tries to pull the host architecture and there is no ARM64 image for EMBA, but with docker pull embeddedanalyzer/emba:2.0.1a --platform linux/amd64 before running the installer it works. Please insert the latest version of the docker base image from here.

Anything outside the docker container that prevents usage of ARM64 for EMBA? So far (progress 57/74, 10 hours in) it looks like it runs just fine... So I think you can change the documentation:

ARM64 machines are only supported as long as the docker container can run as an amd64 container. This was tested on MacOS with Parallels and Parallel's current "Ubuntu 24.04 (with Rosetta)" one-click-install VM that uses Apple's "Rosetta dynamic binary translator" to run amd64 binaries and docker containers. Before running the EMBA installer, reconfigure the VM to use 24GB RAM, 10 cpu cores and 200GB of disc space (make sure you resize the disc partition in the VM to use the free space, e.g. with gparted). Also make sure you pull the amd64 docker container with docker pull embeddedanalyzer/emba:2.0.1a --platform linux/amd64, then run the installer in classic mode (with docker, sudo ./installer.sh -d) and ignore the warnings about unsupported architecture and missing SSSE3 support. Please remember the RAM and CPU requirements, they apply as usually with EMBA.