As i'm currently waiting for my host to supply the monster cloud instance i've ran a few tests to make sure things will go smoothly later on and initially everything looked perfectly fine. I've started a basic 1 core instance, installed my prepared NetBSD image and it happily booted. Now i've upped the ante a bit though and tried the same with a 16 core VM (likely basically the same Qemu configuration just with more cores/resources?) and bam... kernel panic during boot (as far as i can tell by looking at the VNC console it's happening in cdev_open/cnopen).
I'm a bit confused here as both VMs should really be more or less the same thing (besides cores/RAM/disk size) and sadly i don't have the slightest idea how to proceed. I figure this is likely a bug since 16 cores really aren't anything extreme these days but i thought i'd ask if this might be some kind of known problem.
I as far as i know cnopen is called to allocate the console device. Not that i'd have any idea why this would fail on a mundane Qemu VM but i guess switching to serial (which i have no access to but i don't really need the console anyways as my image is fully configured by default) would be at least something i could try to work around the panic but beyond that i'm pretty clueless.
Any ideas/pointers maybe?
Edit: Switching the console device to com0 fixed the crash. Pretty strange but oh well, at least it boots now.
