Re: [RFC PATCH v2] x86/ACPI: optimise C3 entry on AMD CPUs

[Olivier Lambert <olivier.lambert@xxxxxxxxxx>]

On 01.10.2026 09:14, Jan Beulich wrote:
> Why the RFC, btw?

It's my first Xen patch, and on staging it was only build-tested: I
booted and measured it on the 4.17 backport only. I should have dropped
the tag for v2.

> Linux also deals with modern Centaur (VIA) and Zhaoxin (Shanghai) CPUs.
> We will want to mirror that too, provided we want to continue this
> mirroring. Alternatively we could have Linux'es xen-acpi-processor
> driver invoke acpi_processor_power_init_bm_check() before passing up
> C-state info. That would imo be better overall, with one downside: Non-
> Linux Dom0-s would then need to follow suit.

Noted. I'll leave the choice between the two to you and the other x86
maintainers.

Thanks for committing it, and for fixing the attribution.

Olivier


--
Olivier Lambert | Vates CEO

XCP-ng & Xen Orchestra - Vates solutions
Book a meeting with me: https://calrs.vates.tech/u/olivier-lambert
web: https://vates.tech