Ronald G Minnich <[EMAIL PROTECTED]> writes: > Eric, for your PCI scheme, do you end up with an entry for every single > PCI device, or for classes of devices? I'm wondering how big linuxbios > will get to be if we have the whole PCI database in there.
At this point what I am looking at is not hard coding all of the pci devices into the LinuxBIOS image (though I may need some of that) but instead just adding some additional methods to what we have now. And adding code for those devices that need special setup. Most of the devices will probably suffice with generic setup code. All that really doesn't work at the moment is automatic pci bus speed detection. But if I can come up with some hooks roughly like the superio code but a little more general for the pci case, this is a really good time to do it. In addition I want to make an additional pass in the resource allocation so I can do optimal packing of the pci memory space. When everything is restricted to being power of 2 aligned and power of 2 sized the optimal algorithm is just allocating the smallest areas right above 4G, and then working our way up to 3G or however much pci memory space we need. > can linuxbios make everything slow and let linux handle it? Nope. Because Linux works on the strategy that it will do what it knows, but if it doesn't know it won't do anything. And there is a lot of stuff that really makes sense to do at the BIOS level because there isn't a generic layer to do it. > Will linux hot plug affect anything? It can help but I really don't want to count on it. If my code falls flat I will do something else but I'm pretty certain I can do this as a small evolutionary step without a noticeable impact on existing ports. Eric
