On Wed, Dec 7, 2011 at 3:10 AM, Peter Maydell <peter.maydell at linaro.org> wrote: > On 7 December 2011 05:48, Christoffer Dall <cdall at cs.columbia.edu> wrote: >> On Mon, Dec 5, 2011 at 12:47 PM, Peter Maydell <peter.maydell at linaro.org> wrote: >>> Minor tweaks to the $(KERNEL) target: >>> ?* use $(MAKE) so -n &c work better >>> ?* depend on the uImage which we use, not the zImage which we don't >> >> I don't appreciate this. Simple 'make' in the kernel dir produces a >> zImage, which is what I try to pick up - I think hack, make, hack, >> make, etc. is a pretty standard work flow and uImage needs zImage to >> be written as well, so 'make uImage' will also produce a zImage... > > But this boot-wrapper makefile doesn't (and didn't) use the zImage, > it uses the uImage, and that's what it asks the kernel makefile > to create. I don't understand why you'd want to depend on something > you're not using -- surely that can only lead to problems with > not always rebuilding when necessary? > I don't think it can lead to any problems. The dependency graph is this: zImage <- uImage <- boot-wrapper so if uImage has changed, so has zImage. Or am I missing something? Anyway, it's not a big issue for me, I am just trying to avoid someone building their kernel using plain 'make' and not understanding why the boot-wrapper doen't reflect those changes.
