Ühel kenal päeval, R, 2009-01-09 kell 07:22, kirjutas Myles Watson: > > > -----Original Message----- > > From: [email protected] [mailto:[email protected]] > > On Behalf Of Mart Raudsepp > > > Also, what if we had the possibility to run code for disabled devices - > > then the vpci hiding could cleanly happen when the device is disabled > > instead of checks from northbridge code. Like a .disable functor in > > device_ops? > > We do. If you run it in phase2 it runs independent of the enabled flag, and > could use that to do any disabling that was needed. It's a little confusing > to put the device in the tree then, though, because it will never show up in > lspci.
I guess I mean a way to run disabling code if a device doesn't appear in the mainboard dts at all. E.g, there is a northbridge/amd/geodelx/graphics.dts and if the mainboard dts doesn't source it or it's disabled, a disabling function would be ran. This is just something that optionally wired built-in PCI devices might be interested in, not the typical case of PCI slots. I can imagine it would get tricky to implement and quite likely it's just a bad idea. Mart Raudsepp -- coreboot mailing list: [email protected] http://www.coreboot.org/mailman/listinfo/coreboot
