Re: [PATCH] x86/acpi: Remove indirect jump from wakeup_long64()

[Ard Biesheuvel]

On Sun, 18 Jan 2026 at 20:17, Brian Gerst <brgerst@xxxxxxxxx> wrote:
>
> On Sun, Jan 18, 2026 at 1:09 PM Ard Biesheuvel <ardb@xxxxxxxxxx> wrote:
> >
> > On Sun, 18 Jan 2026 at 18:05, Brian Gerst <brgerst@xxxxxxxxx> wrote:
> > >
> > > wakeup_long64() is called from common_startup_64() via inital_code, so
> >
> > initial_code
> >
> > > it is already running on the normal virtual mapping.  There is no need
> > > to use an indirect jump since it is not switching mappings.
> > >
> >
> > By the same reasoning (i.e., that everything executes in the kernel
> > virtual mapping) you might also convert the movq on line 55 into a
> > RIP-relative leaq, as it is the only remaining non-RIP relative access
> > in the file. That way, I can drop this file from my PIE series too.
>
> This is more of a general cleanup than specifically for PIE.  When
> resume was converted to using the SMP trampoline, a lot of the old
> cruft was left in place and I am working on additional cleanups.

Fair enough.