Re: [PATCH 1/3] x86/cpufreq: Clean up powernow registration

[Jan Beulich <jbeulich@xxxxxxxx>]

On 12.11.2021 19:28, Andrew Cooper wrote:
> --- a/xen/arch/x86/acpi/cpufreq/cpufreq.c
> +++ b/xen/arch/x86/acpi/cpufreq/cpufreq.c
> @@ -640,13 +640,19 @@ static int __init cpufreq_driver_init(void)
>  {
>      int ret = 0;
>  
> -    if ((cpufreq_controller == FREQCTL_xen) &&
> -        (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL))
> -        ret = cpufreq_register_driver(&acpi_cpufreq_driver);
> -    else if ((cpufreq_controller == FREQCTL_xen) &&
> -        (boot_cpu_data.x86_vendor &
> -         (X86_VENDOR_AMD | X86_VENDOR_HYGON)))
> -        ret = powernow_register_driver();
> +    if ( cpufreq_controller == FREQCTL_xen )
> +    {
> +        switch ( boot_cpu_data.x86_vendor )
> +        {
> +        case X86_VENDOR_INTEL:
> +            ret = cpufreq_register_driver(&acpi_cpufreq_driver);
> +            break;

I've been wondering why we continue to tie this to Intel. I don't
think there's much Intel specific in the ACPI driver, so I wonder
whether this shouldn't use "default:" instead. But I can agree
that's likely better to be done in a separate change.

> @@ -353,25 +349,13 @@ static const struct cpufreq_driver __initconstrel 
> powernow_cpufreq_driver = {
>      .update = powernow_cpufreq_update
>  };
>  
> -unsigned int __init powernow_register_driver()
> +unsigned int __init powernow_register_driver(void)
>  {
> -    unsigned int i, ret = 0;
> +    if ( !(boot_cpu_data.x86_vendor & (X86_VENDOR_AMD | X86_VENDOR_HYGON)) )
> +        return -ENODEV;

Ideally with this dropped (and of course with the issue pointed
out by Roger taken care of)

Acked-by: Jan Beulich <jbeulich@xxxxxxxx>

Jan