Re: [PATCH v7 004/120] x86/cpuid: Introduce <asm/cpuid/leaf_types.h>

[Borislav Petkov]

On Mon, Jun 08, 2026 at 08:39:32AM +0200, Ahmed S. Darwish wrote:
> I guess that's everyone's goal :)

Glad we're on the same page. :)
 
> You mentioned that several times.

I do sound like a broken record sometimes, sure. Better that than dealing with
nasty luserspace dependence.

> How can userspace depend on an internal x86 kernel header?

Well, the recent futex snafu teaches us again that luserspace can depend on
anything in the kernel.

What you're doing is a database of *all* CPUID leafs. How long until someone
copies or exports the header to something? Or starts generating the header
from your XML and using it in her/his project?

And what happens if someone has already started using the headers, you fix
those spelling mistakes, release a new revision and someone updates the
generated headers locally and code starts failing to build all of a sudden?

What do you do then?

I think with issues like that which can cause nasty workarounds, we cannot be
urgent enough in fixing them.

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette