Re: [PATCH 2/2] devicetable-offsets: Get struct cpu_feature from a device-id header
From: Nicolas Schier
Date: Thu Nov 19 2026 - 10:25:05 EST
Hi,
I have a question regarding best-practise:
On Mon, Nov 09, 2026 at 05:07:05PM +0100, Nicolas Schier wrote:
> On Mon, 09 Nov 2026 09:14:41 -0500, Sasha Levin wrote:
> > Since commit cdfdbc11552b ("Move struct cpu_feature to
> > <linux/cpufeature.h>"), devicetable-offsets.c includes
> > <linux/cpufeature.h>, which includes <asm/cpufeature.h> on architectures
> > that select GENERIC_CPU_AUTOPROBE. On loongarch and mips that header
> > pulls in <asm/elf.h>, and from there <linux/fs.h> and <linux/mmzone.h>,
> > which need <generated/bounds.h>. devicetable-offsets.c is built before
> > that header is generated, so even defconfig fails on both:
> >
> > [...]
>
> Applied to kbuild/linux.git (kbuild-next-unstable), thanks!
>
> [2/2] devicetable-offsets: Get struct cpu_feature from a device-id header
> https://git.kernel.org/kbuild/c/50ccc340
I applied this fix on kbuild-next-unstable (and rebased today [1]), two
commits after "Move struct cpu_feature to <linux/cpufeature.h>" [2],
which is being "fixed" by this patch.
If we keep it this way, a bisect may hit the issue; folding does not
look correct to me as that would mix-up related but different things.
So my questions:
* Shall we keep it the way it is to keep the original patch order
(bisect might fail on the two commits before this patch)?
* Or shall the patch be moved before Uwe's (removing the 'Fixes' tag,
and rewriting the commit msg)?
I am in favor of the latter. Are there any objections?
[1]: https://git.kernel.org/kbuild/c/f1dc66c38b7eb4b2fb946a2fd6929250ecd2d85a
[2]: https://patch.msgid.link/bc9fbf5784f6788e10c749a05d56c42bdb68f3d9.1792144769.git.u.kleine-koenig@xxxxxxxxxxxx
--
Nicolas