On Tue, Mar 19, 2019 at 06:42:35PM +0100, Christophe Leroy wrote: > Well, that's what I started with, but at the end my main worry has been that > you bring a non exciting set of complicated macros and code to replace > simple code, and you break something out of generic OF code to a new brand > new generic one, instead of updating the existing generic OF code. Even if we update the generic OF code it only changes the powerpc changes slightly. Because in arch/powerpc/kernel/prom_init.c there is a second version of the same thing, which doesn't use OF.
We're not replacing simple macro's in powerpc with in-kind replacements, we're adding a feature which we want. So yes our macros are more complicated, but in the grand scheme of things they are very simple macros. If you think my stuff is complicated, you haven't seen complicated. I didn't see anyplace in your comments when you found code which would cause a problem ? Did you find breakage which I missed? > I like the idea behind your series very much, but I don't like too much the > way it is proposed to be implemented. If you give me one week or two, I will > come with a lighter proposal that should achieve the same goal. It's fine with us, we just want the feature set. We'll continue with our version tho, unless you decide to submit something. I will incorporate your comments now, but immediately prior to a pull request I couldn't add them. Daniel