Re: [coreboot] [PATCH] drivers vs. obj

[Stefan Reinauer]

On 2/23/10 5:29 PM, Patrick Georgi wrote:
> Am 23.02.2010 17:28, schrieb Myles Watson:
> > -driver-y += mainboard.o
> > -driver-y += m3885.o
> > -driver-y += ec.o
> > -driver-y += rtl8168.o
> > +obj-y += mainboard.o
> > +obj-y += m3885.o
> > +obj-y += ec.o
> > +obj-y += rtl8168.o
> >
> > It looks like rtl8168 should stay a driver, since it's a PCI device. I
> > don't know about m3885.o and ec.o. I'm also not sure about the debug
> > devices.  Are they considered PCI?  It's been a long time since I
> > thought about this.
>
> As I understand it, the main criterion was if they're using the linker
> magic. I couldn't find any of that in there, but maybe that's just too
> hidden for my (short, I admit) glance?

No, that's right.. the EC code is only called from mainboard.c and the
SMM handler