Any updates on the support for Manjaro on M1 running either in a VM or native?
A few distros have support for the M1 hardware by now in VMWare Fusion such as Ubuntu, Fedora, Photon OS, Debian.
Can duck duck go for Fusion-for-Apple-Silicon-Tech/Fusion-Tech-Preview-Testing-Guide for details.
2 Likes
I think you are mistakenly assume that hypervisor support is bare-metal support.
There is no point in trying to get Manjaro running on Apple branded hardware - If you want to use Apple branded hardware - use macOS - it is a much better solution as they are built for each other and quite frankly - I don’t think Apple has no intention of making it possible to run other operating systems on Apple branded hardware.
1 Like
I was hoping to test out Manjaro in a VM powered by Fusion on M1 hardware the same way as I have been testing other distros as Manjaro seems to have many nice features.
If there is a way to get Manjaro running in a similar fashion as described in the testing guidelines from VMware that would be great.
1 Like
If you can run other Linux - it is likely possible to run Manjaro.
Can you please point me to the most suitable .iso file for Arm64 or is there another process to follow to get this working?
ARM images has no iso file - but comes as a xz compressed archive - just unpack it - it works the same way…
1 Like
qemu based, user friendly, you can run any os on M1
2 Likes
ARM images are per device. So now, we don’t support the M1 yet. This would also mean some of our developers have to buy M1s to prepare and test the images they build. There is Asahi Linux, which is trying to get Linux on the new Mac’s. They may eventually release a remix of Arch Linux ARM, packaged for installation by end-users, as a distribution of the same name. When that happens we can adopt the effort easily. You can browse thru their github repos, to see what they are up to.
7 Likes
Thanks for the status update on the M1, I will wait until official support is hopefully announced.
1 Like
I run Parallels Desktop on my M1 MacBook Pro. Today I was able to build a custom generic aarch64 UEFI image and get Manjaro running on my Mac by modifying a few lines of the manjaro-arm-installer script. The changes were to create a slightly larger boot partition and to flag it as an ESP and format it in FAT32, as well as reference a custom device profile I made. I then had to enter the UEFI Shell of the VM on the first boot in order to boot into Manjaro where I could then use efibootmgr to add the boot entry (used instructions from the EFISTUB page on the ArchWiki)
8 Likes
Thank you for the explanation.
I think that’s right but have you tried it natively without vm?
I haven’t tried running natively yet. I’ll give that a try on my M1 Mini using Corellium’s custom boot object method: How We Ported Linux to the M1 | Corellium & GitHub - corellium/preloader-m1: Preloader for Linux on M1
However, I have a feeling that the use of Plasma is going to require the GPU work of the Asahi project to progress a bit more.
Yeah, the graphics aren’t far enough along to run a desktop environment yet.
I actually boot macOS and Manjaro KDE on my MacBook Pro, granted it is an Intel model but I have had no issue dual booting the two operating systems.
This topic is about ARM cpu version macs.
ARM images are per device.
Why is that exactly? The Intel images work on any Intel machine. Is there something unique about the ARM architecture that makes such a generic approach impossible?
Yessir I get that but I was merely responding to your blanket statement were you do not differentiate what Apple Branded hardware you are talking about. You basically stated that running Linux on Apple Branded hardware is not possible, nor should it be tried.
“There is no point in trying to get Manjaro running on Apple branded hardware - If you want to use Apple branded hardware - use macOS - it is a much better solution as they are built for each other and quite frankly - I don’t think Apple has no intention of making it possible to run other operating systems on Apple branded hardware”.
x86_64 architecture depends on ACPI ( I think it is) to detect what hardware is present. ACPI has not been implemented for ARM (yet), so on ARM we use a device tree file to define what hardware is present, so Linux knows what to expect. But this also means that what device tree to use, needs to be defined in a boot script somewhere. These boot scripts, often named boot.scr or extlinux.conf, therefore needs to be made on a per device basis. This also trickles down to the board firmware (often u-boot), which makes sure the board is actually looking for a kernel to boot and initializes the hardware. On x86_64 this process is already done by the manufacturer of the motherboard, often referred to as the BIOS or UEFI firmware, but most vendors of ARM boards, do not do this. So we (the operating system) are required to provide this firmware, which is, again, board specific.
7 Likes
I’m still dreaming of a Boot Camp method to dual boot MacOS and Manjaro on a M1 based Mac.
Like on previous intel based Macs.
It’s great, Linux for home, MacOS for graphics design at school/work on a single device.
I was able to build a custom generic aarch64 UEFI image and get Manjaro running on my Mac
@GeekPower0 would you be willing to share your image (or a small write-up explaining the process?)
I too would like to run Manjaro on my M1 Mac in Parallels. I currently have Windows 11, Fedora 34, Ubuntu 20.04 and Ubuntu 22.04 running Aarch64 versions, which is nice, it allows me to fire up these Operating Systems in a Hypervisor for various Linux-only (or Windows-only) tasks when needed. I’m trying to troubleshoot an issue that someone is having with Manjaro and although I can fire it up on Intel without an issue, getting it running in a VM on the M1 would be very convenient. 