Re: [XEN PATCH v2 1/5] x86/Kconfig: introduce CENTAUR, HYGON & SHANGHAI config options

[Jan Beulich <jbeulich@xxxxxxxx>]

On 16.08.2024 13:10, Sergiy Kibrik wrote:
> --- a/xen/arch/x86/Kconfig.cpu
> +++ b/xen/arch/x86/Kconfig.cpu
> @@ -10,6 +10,25 @@ config AMD
>         May be turned off in builds targetting other vendors.  Otherwise,
>         must be enabled for Xen to work suitably on AMD platforms.
>  
> +config CENTAUR
> +     bool "Support Centaur CPUs"
> +     default y
> +     help
> +       Detection, tunings and quirks for VIA platforms.
> +
> +       May be turned off in builds targeting other vendors. Otherwise, must
> +          be enabled for Xen to work suitably on VIA platforms.
> +
> +config HYGON
> +     bool "Support Hygon CPUs"
> +     depends on AMD
> +     default y
> +     help
> +       Detection, tunings and quirks for Hygon platforms.
> +
> +       May be turned off in builds targeting other vendors. Otherwise, must
> +          be enabled for Xen to work suitably on Hygon platforms.
> +
>  config INTEL
>       bool "Support Intel CPUs"
>       default y
> @@ -19,4 +38,14 @@ config INTEL
>         May be turned off in builds targetting other vendors.  Otherwise,
>         must be enabled for Xen to work suitably on Intel platforms.
>  
> +config SHANGHAI
> +     bool "Support Shanghai CPUs"
> +     depends on INTEL
> +     default y
> +     help
> +       Detection, tunings and quirks for Zhaoxin platforms.
> +
> +       May be turned off in builds targeting other vendors. Otherwise, must
> +          be enabled for Xen to work suitably on Zhaoxin platforms.
> +
>  endmenu

Imo this re-raises the question of whether it is a good idea to leave out
"CPU" from the names: The more names there are, the more likely it'll become
that going forward we'll run into a naming collision. Andrew, iirc you
were the main proponent for omitting "CPU" - may I ask that you re-consider?

Furthermore I wonder whether "depends on" is appropriate here. This way one
won't be offered e.g. SHANGHAI if earlier on one de-selected INTEL.

Plus it is mere luck that the alphabetic ordering ends up with the dependents
after their dependencies (things would be somewhat odd the other way around).

Finally please check indentation of help text - there is one inconsistency
repeated for all three entries being added.

Jan