Hi Hans, On Fri, 09 Nov 2007 09:03:31 +0100, Hans de Goede wrote: > Jean Delvare wrote: > > While it is possible to force SMBus-based hardware monitoring chip > > drivers to drive a not officially supported device, we do not have this > > possibility for Super-I/O-based drivers. That's unfortunate because > > sometimes newer chips are fully compatible and just forcing the driver > > to load would work. Instead of that we have to tell the users to > > recompile the kernel driver, which isn't an easy task for everyone. > > > > So, I propose that we add a module parameter to all Super-I/O based > > hardware monitoring drivers, letting advanced users force the driver > > to load on their machine. The user has to provide the device ID of a > > supposedly compatible device. This requires looking at the source code or > > a datasheet, so I am confident that users can't randomly force a driver > > without knowing what they are doing. Thus this should be relatively safe. > > > > As you can see from the code, the implementation is pretty simple and > > unintrusive. > > > > Signed-off-by: Jean Delvare <khali at linux-fr.org> > > Sounds like a good plan to me and the patch looks good to me too, notice that > it touches lots of files, so despite a best effort I may have overlooked a typo > of some kind. Well most of it is cut'n'paste from one driver to the next, there's not that much room for typos. Only the vt1211 and w83627ehf drivers are a bit different because the original code was a bit different. Just one thing that I'd like to change now that I think of it: "overrule" is probably not the correct word, "override" would probably be better. I'll submit an updated patch with this wording. > Acked-by: Hans de Goede <j.w.r.degoede at hhs.nl> Thanks for the quick review :) -- Jean Delvare
