Arch Linux

gatecrasher

Member

Registered: 2025-01-28

Posts: 2

Sticking an M2 with arch into a new laptop?

Hello, considering that I'll have to get a laptop for uni soon, and I've had arch on my PC for over a year or so, would it be possible to take out the M2 out of my PC and put it into the laptop M2 slot, provided I get a laptop that has the same M2 module width?

I'm asking because a post that I've found already answering the question claimed it wouldn't work. Does anyone have experience with this? Will it boot perfectly or will I have to tinker with it?

V1del

Forum Moderator

Registered: 2012-10-16

Posts: 24,094

Re: Sticking an M2 with arch into a new laptop?

it's going to generally work, maybe not get to a GUI if you have a broken xorg.conf autogenerated from nvidia-xconfig which only exposes your PCs nvdia card but in that case you should get rid of that anyway.

You'll likely need to do the first boot off of the fallback initramfs image (and regenerate the initramfs after) depending on HW differences. And since you won't have NVRAM entries populate any ESP partitions fallpack path might come in handy if you don't want to accompany this with a live disk to fix the bootloader. Obviously, if you're using secureboot/LUKS keys on the TPM ths will significantly complicate the matter, but if not there's nothing that would inherently not make this possible.

Last edited by V1del (2025-01-28 12:03:38)

gatecrasher

Member

Registered: 2025-01-28

Posts: 2

Re: Sticking an M2 with arch into a new laptop?

Thank you a lot for the reply!

You mention both xorg and nvidia, though I'm on wayland and AMD (not that you could've known, maybe I should've clarified that immediately). Would sticking with AMD make the process easier? Will wayland spew out errors or will it be fine?

Thank you in advance!

V1del

Forum Moderator

Registered: 2012-10-16

Posts: 24,094

Re: Sticking an M2 with arch into a new laptop?

No, it's not an inherent problem on nvidia nor has anything to do with wayland and xorg per se, but a potential correlation. If someone generated a static xorg config (or written one themselves) that hard requires devices to be at a certain spot, then xorg will fail to start. That's something that's relatively extraordinary, since most usecases should be covered by the autodetection xorg does anyway.

Since wayland isn't a "thing", this highly depends on your compositor, but I'd also consider it unlikely they implement this in a hardcoded way. Potentially some graphics stutters initially for shaders to get rebuilt.