Re: linux-next: manual merge of the kbuild-current tree with the origin tree
From: Magnus Lindholm
Date: Wed Oct 21 2026 - 16:28:24 EST
Hi Nathan,
On Wed, Oct 21, 2026 at 8:59 PM Nathan Chancellor <nathan@xxxxxxxxxx> wrote:
>
> On Wed, Oct 21, 2026 at 04:40:22PM +0200, Magnus Lindholm wrote:
> > On Wed, Oct 21, 2026 at 1:01 PM Mark Brown <broonie@xxxxxxxxxx> wrote:
> > >
> > > Hi all,
> > >
> > > Today's linux-next merge of the kbuild-current tree got a conflict in:
> > >
> > > arch/alpha/boot/Makefile
> > >
> > > between commit:
> > >
> > > c437950dceb88 ("alpha: add self-extracting compressed kernel boot stub")
> > >
> > > from the origin tree and commit:
> > >
> > > 0d4d70da348b0 ("kbuild: Mark the gzip recipes as recursive")
> > >
> > > from the kbuild-current tree.
>
> Thank you for the heads up, I have let Linus know about this one since
> it is a little tricky.
>
> https://lore.kernel.org/20261021184723.GA2608581@ax162/
>
> > Thanks for resolving this.
> >
> > This recipe now only copies the completed image, so I think it
> > should stay without '+'. Matt's series moved compression into
> > arch/alpha/boot/compressed/Makefile, where the change should be:
> >
> > $(obj)/vmlinux.bin.gz: $(obj)/vmlinux.bin FORCE
> > - $(call if_changed,gzip)
> > + +$(call if_changed,gzip)
> >
> > The existing vmlinux.nh.gz gzip recipe in arch/alpha/boot/Makefile
> > still needs the '+' from the kbuild patch.
>
> Correct, thank you for catching this.
>
> > Maybe adjust the resolution accordingly?
>
> Since the conflict came from Linus's tree, I have resolved this in
> kbuild-fixes-for-next so that Mark does not have to worry about this:
>
> https://git.kernel.org/kbuild/c/83deedcf85af511dc120b3cd8ee597d211cdf2fb
>
Thanks so much for handling this.
> On a related note, is your tree flowing into -next? The recent
> pull requests for alpha have come from your lindholm/alpha tree but
> Next/Trees shows Matt's mattst88/alpha. Should that be changed? It would
> not have saved much here since I send my pull request to Linus before
> c437950dceb88 was applied in your tree as far as I can tell but it is
> always good to catch these things early.
>
You are absolutely right, I'll wire up my lindholm/alpha tree into -next before
the next merge window.
Thanks again,
Magnus