So this turns out to be a bit of a problem... If i do: make modules_install make install It generates a 900MB+ initrd that doesn't fit - and without it i can't get it to boot properly.. I don't know what ubuntu does with mkinitrdfs but it seems quite excessive. Also, the kernel still seems to crash (doesn't respond to capslock) and it doesn't seem to like dyndbg="file power.c p+" - it said something about p+ being wrong (Company laptop, lvm, crypted fs and all that fun stuff... :/) On Thu, Nov 18, 2021 at 2:52 PM Rafael J. Wysocki <rafael@xxxxxxxxxx> wrote: > > On Thu, Nov 18, 2021 at 2:41 PM Ian Kumlien <ian.kumlien@xxxxxxxxx> wrote: > > > > So, hard to capture, crashes faster with minimal output: > > https://photos.app.goo.gl/9p56395EdhNBuikQA > > > > First time i didn't even see the oops at all... > > The dynamic debug output is still not present. > > Please apply the attached patch (you may need to adjust it for 5.15 - > the idea is to comment out the execution of _ON and _OFF) and boot > with the dyndbg in power.c enabled. > > It should not crash this time, so you should be able to collect full > dmesg and send it (as an attachment, please).
