On Fri, Jun 19, 2009 at 12:06:51PM +0100, Harrison, Jon (SELEX GALILEO, UK) wrote:
Received and understood, then the basic issue is that at this point my coreboot build is still broken and will not boot, so there are no tables. (I had not realised the tables were read from RAM rather than the ROM image)
The upshot is that the flashrom ids are required to allow development from the original bios.... for the time being.
Once the build is booting properly and flashrom can autodetect I will remove the ids. In the mean time I'm hoping that having this early flashrom support will help others to help in the effort for the CN400/epia-n(l)/Luke.
If this is still all to immature for general release then fair enough, it can wait until things are working better.
Thanks, Jon
No, this is reasonable, i will add an extra comment and commit.
Plus, i think that we need to track down some badly filled out boards, pciid wise, so that we can change behaviour around the ids a bit. Make it a bit tighter: * require 2 main ids, always. * don't allow -m when a valid subsystem id is there.
But that's a discussion for some other time, and anyway, it has a rather tough prerequisite.
Luc Verhaegen.