> This dangerously seems like an atempt to be able to create proprietary drivers and bypass the GPL.
Probably only me, probably because IANAL, but I fail to see how an UIO driver would not be a derivative work of the kernel.
That would obviously be different if this interface was implemented in *BSD, Solaris, Windows... but I do not see that being the intent.
And implementing a GPL layer to connect a non-GPL driver would not change that the non-GPL driver is a derivative work of a GPL driver, which is a derivative work of the kernel.
Now if the law were applied...