Re: [PATCH v3] intel_idle: introduce 'no_native' module parameter

[David Arcari <darcari@redhat.com>]

On 2/12/25 7:46 AM, Artem Bityutskiy wrote:
> On Wed, 2025-02-12 at 07:41 -0500, David Arcari wrote:
>> - #ifdef the code that doesn't compile
>> - default no_acpi=true in the !CONFIG_ACPI_PROCESSOR_CSTATE case
>>
>> I sort of like the second option better, but I worry about the
>> documentation.  Specifically:
>>
>> "In the case that ACPI is not configured these flags have no impact
>> +on functionality."
>>
>> I guess that is still true.
>>
>> Perhaps there is a better option.  What do you think?
> 
> I've not been involved into kernel that much for long time. In old days
> sprinkling #ifdefs around was an anti-pattern. Most probably nowadays too. So
> the second option sounds better to me.

Another option would be to change the offending code to a function call:

if (ignore_native()) {

And have ignore_native() always return false when ACPI is not configured.

And yes I should have built and tested the kernel with ACPI disabled. 
My apologies.

I will do that for v4.

-DA

> 
> Artem.
>