diff mbox series

ACPI: Only evaluate the Intel _OSC and _PDC on platforms with HWP

Message ID 20240614193241.7341-1-mario.limonciello@amd.com (mailing list archive)
State Rejected, archived
Headers show
Series ACPI: Only evaluate the Intel _OSC and _PDC on platforms with HWP | expand

Commit Message

Mario Limonciello June 14, 2024, 7:32 p.m. UTC
The UUID 4077A616-290C-47BE-9EBD-D87058713953 and _PDC methods are
only used on Intel platforms with HWP support.

Attempting to evaluate them and showing messages on hardware without
HWP is pointless needlessly noisy.

Gate the code on X86_FEATURE_HWP.

Cc: PradeepVineshReddy (Pradeep Vinesh Reddy) Kodamati <PradeepVineshReddy.Kodamati@amd.com>
Suggested-by: CL Lin <clin41@lenovo.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
---
 drivers/acpi/acpi_processor.c | 2 ++
 1 file changed, 2 insertions(+)

Comments

Rafael J. Wysocki June 15, 2024, 10:38 a.m. UTC | #1
On Fri, Jun 14, 2024 at 9:33 PM Mario Limonciello
<mario.limonciello@amd.com> wrote:
>
> The UUID 4077A616-290C-47BE-9EBD-D87058713953 and _PDC methods are
> only used on Intel platforms with HWP support.

I beg to differ.  See arch_acpi_set_proc_cap_bits() definition.

> Attempting to evaluate them and showing messages on hardware without
> HWP is pointless needlessly noisy.
>
> Gate the code on X86_FEATURE_HWP.

Not really.

> Cc: PradeepVineshReddy (Pradeep Vinesh Reddy) Kodamati <PradeepVineshReddy.Kodamati@amd.com>
> Suggested-by: CL Lin <clin41@lenovo.com>
> Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
> ---
>  drivers/acpi/acpi_processor.c | 2 ++
>  1 file changed, 2 insertions(+)
>
> diff --git a/drivers/acpi/acpi_processor.c b/drivers/acpi/acpi_processor.c
> index 7a0dd35d62c9..84848b5e65d6 100644
> --- a/drivers/acpi/acpi_processor.c
> +++ b/drivers/acpi/acpi_processor.c
> @@ -597,6 +597,8 @@ static bool __init acpi_early_processor_osc(void)
>
>  void __init acpi_early_processor_control_setup(void)
>  {
> +       if (!boot_cpu_has(X86_FEATURE_HWP))
> +               return;
>         if (acpi_early_processor_osc()) {
>                 pr_info("_OSC evaluated successfully for all CPUs\n");
>         } else {
> --
> 2.43.0
>
Mario Limonciello June 18, 2024, 7:33 p.m. UTC | #2
On 6/15/2024 05:38, Rafael J. Wysocki wrote:
> On Fri, Jun 14, 2024 at 9:33 PM Mario Limonciello
> <mario.limonciello@amd.com> wrote:
>>
>> The UUID 4077A616-290C-47BE-9EBD-D87058713953 and _PDC methods are
>> only used on Intel platforms with HWP support.
> 
> I beg to differ.  See arch_acpi_set_proc_cap_bits() definition.

Ah I see; thanks for sharing.  I agree with you.

> 
>> Attempting to evaluate them and showing messages on hardware without
>> HWP is pointless needlessly noisy.
>>
>> Gate the code on X86_FEATURE_HWP.
> 
> Not really.

I guess this really started getting noisy in 6.6 (9527264).

Would you be open to downgrading to debug instead?

> 
>> Cc: PradeepVineshReddy (Pradeep Vinesh Reddy) Kodamati <PradeepVineshReddy.Kodamati@amd.com>
>> Suggested-by: CL Lin <clin41@lenovo.com>
>> Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
>> ---
>>   drivers/acpi/acpi_processor.c | 2 ++
>>   1 file changed, 2 insertions(+)
>>
>> diff --git a/drivers/acpi/acpi_processor.c b/drivers/acpi/acpi_processor.c
>> index 7a0dd35d62c9..84848b5e65d6 100644
>> --- a/drivers/acpi/acpi_processor.c
>> +++ b/drivers/acpi/acpi_processor.c
>> @@ -597,6 +597,8 @@ static bool __init acpi_early_processor_osc(void)
>>
>>   void __init acpi_early_processor_control_setup(void)
>>   {
>> +       if (!boot_cpu_has(X86_FEATURE_HWP))
>> +               return;
>>          if (acpi_early_processor_osc()) {
>>                  pr_info("_OSC evaluated successfully for all CPUs\n");
>>          } else {
>> --
>> 2.43.0
>>
Rafael J. Wysocki June 18, 2024, 7:46 p.m. UTC | #3
On Tue, Jun 18, 2024 at 9:33 PM Mario Limonciello
<mario.limonciello@amd.com> wrote:
>
> On 6/15/2024 05:38, Rafael J. Wysocki wrote:
> > On Fri, Jun 14, 2024 at 9:33 PM Mario Limonciello
> > <mario.limonciello@amd.com> wrote:
> >>
> >> The UUID 4077A616-290C-47BE-9EBD-D87058713953 and _PDC methods are
> >> only used on Intel platforms with HWP support.
> >
> > I beg to differ.  See arch_acpi_set_proc_cap_bits() definition.
>
> Ah I see; thanks for sharing.  I agree with you.
>
> >
> >> Attempting to evaluate them and showing messages on hardware without
> >> HWP is pointless needlessly noisy.
> >>
> >> Gate the code on X86_FEATURE_HWP.
> >
> > Not really.
>
> I guess this really started getting noisy in 6.6 (9527264).
>
> Would you be open to downgrading to debug instead?

That would be fine, thanks!


> >
> >> Cc: PradeepVineshReddy (Pradeep Vinesh Reddy) Kodamati <PradeepVineshReddy.Kodamati@amd.com>
> >> Suggested-by: CL Lin <clin41@lenovo.com>
> >> Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
> >> ---
> >>   drivers/acpi/acpi_processor.c | 2 ++
> >>   1 file changed, 2 insertions(+)
> >>
> >> diff --git a/drivers/acpi/acpi_processor.c b/drivers/acpi/acpi_processor.c
> >> index 7a0dd35d62c9..84848b5e65d6 100644
> >> --- a/drivers/acpi/acpi_processor.c
> >> +++ b/drivers/acpi/acpi_processor.c
> >> @@ -597,6 +597,8 @@ static bool __init acpi_early_processor_osc(void)
> >>
> >>   void __init acpi_early_processor_control_setup(void)
> >>   {
> >> +       if (!boot_cpu_has(X86_FEATURE_HWP))
> >> +               return;
> >>          if (acpi_early_processor_osc()) {
> >>                  pr_info("_OSC evaluated successfully for all CPUs\n");
> >>          } else {
> >> --
> >> 2.43.0
> >>
>
diff mbox series

Patch

diff --git a/drivers/acpi/acpi_processor.c b/drivers/acpi/acpi_processor.c
index 7a0dd35d62c9..84848b5e65d6 100644
--- a/drivers/acpi/acpi_processor.c
+++ b/drivers/acpi/acpi_processor.c
@@ -597,6 +597,8 @@  static bool __init acpi_early_processor_osc(void)
 
 void __init acpi_early_processor_control_setup(void)
 {
+	if (!boot_cpu_has(X86_FEATURE_HWP))
+		return;
 	if (acpi_early_processor_osc()) {
 		pr_info("_OSC evaluated successfully for all CPUs\n");
 	} else {