Hi Michael, On Tue, Jan 20, 2015 at 8:05 AM, Michael Ellerman <mpe@xxxxxxxxxxxxxx> wrote: > It looks like it's ~4 years since we updated some of these, so do a bulk > update. > > Verified that the before and after generated configs are exactly the > same. Yep, that's guaranteed by the tool (modulo bugs). > Which begs the question why update them? The answer is that it can be > confusing when the stored defconfig drifts too far from the generated > result. And it becomes hard to update them, e.g. for enabling or disabling a single option. Now the big question is: do they still work? Plainly ignoring any bugs introduced in the code, there are probably several "select" statements that have been removed. 4 years ago, they automatically selected some critical options, while now, they're no longer selected, causing breakage. I'm afraid the only safe way is (for all 108 defconfig files): - start from the full .config file, as generated from the minimal defconfig file in a source tree from 4 years ago, - run "make oldconfig" (boring), - run "make savedefconfig", - update the minimal defconfig. The boring step can be made a little bit less boring by making a diff between the .config versions before and after running the first "make oldconfig", and applying that to the later .config files first, but it's still a tedious task, especially after 4 years. That's why I do this (for m68k) on every rc release. Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds -- To unsubscribe from this list: send the line "unsubscribe linux-kbuild" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html