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

[David Arcari <darcari@redhat.com>]

Hi Artem,

On 2/12/25 6:32 AM, Artem Bityutskiy wrote:
> On Wed, 2025-02-12 at 18:09 +0800, kernel test robot wrote:
>>     drivers/idle/intel_idle.c: In function 'intel_idle_init':
>>>> drivers/idle/intel_idle.c:2289:27: error: 'no_acpi' undeclared (first use
>>>> in this function); did you mean 'no_action'?
>>      2289 |         if (no_native && !no_acpi) {
>>           |                           ^~~~~~~
>>           |                           no_action
>>     drivers/idle/intel_idle.c:2289:27: note: each undeclared identifier is
>> reported only once for each function it appears in
> 
> David, this must be the !CONFIG_ACPI_PROCESSOR_CSTATE case.
> 
> Thanks!

Oh - I see the problem.

After a quick look I see two options:

- #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?

Thanks,
-DA