[RFC] Drop Non-smp PowerPC Kernel Flavor
benh at kernel.crashing.org
Tue Mar 20 04:07:40 UTC 2012
On Tue, 2012-03-20 at 14:06 +1100, Benjamin Herrenschmidt wrote:
> On Mon, 2012-03-19 at 07:48 +0000, Adam Conrad wrote:
> > On Mon, Mar 19, 2012 at 04:59:44PM +1100, Benjamin Herrenschmidt wrote:
> > >
> > > Any more data about where it stops booting ? Can you try making sure you
> > > have udbg btext enabled in config and boot with "udb-immortal debug
> > > initcall_debug" on the command line ?
> > Unfortunately, it boots from BootX, because I could never make quik run
> > on my broken OF (granted, I never tried that hard, so I'm sure it can
> > be done, but if it ain't broke?). The problem with booting from BootX,
> > though, is that there's no chance for debug output, because I just get
> > a slightly corrupt MacOS desktop, and then the machine hangs. It's not
> > getting far enough to init a framebuffer or anything shiny.
> That's odd ... initializing a FB comes very very early ... oh well, the
> wallstreet is also a bootx machine, so I'll give that a go.
A quick test with my -next branch SMP (I haven't tried your kernel &
config yet) and it seems to work fine (well, it doesn't properly mount
root afterward but that's a different issue).
Have you tried the BootX app instead of extension ? That might make the
display work better ?
The kernel should be setting up a raw text display for early output very
very early on, before the main framebuffer driver shows up, I'm
surprised that doesn't even happen.
I'll try the ubuntu kernel & config next, tho I'm a bit unwell today so
I suspect I'll just go home & do that tomorrow.
More information about the ubuntu-devel