Skip to contentSkip to conversation

ACPI: CPPC: add sysfs entry for guaranteed performance

PATCH In review Petr Tesařík 3 participants 4 messages Jun 6, 2024 – Jun 21, 2024

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 ACPI: CPPC: add sysfs entry for guaranteed performance

Expose the CPPC guaranteed performance as reported by the platform through GuaranteedPerformanceRegister.

The current value is already read in cppc_get_perf_caps() and stored in struct cppc_perf_caps (to be used by the intel_pstate driver), so only the attribute itself needs to be defined.

---
 drivers/acpi/cppc_acpi.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/acpi/cppc_acpi.c +2 -0
index 1d857978f5f4..9976bb57356e 100644
@@ -160,6 +160,7 @@ show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, highest_perf);
 show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, lowest_perf);
 show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, nominal_perf);
 show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, lowest_nonlinear_perf);
+show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, guaranteed_perf);
 show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, lowest_freq);
 show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, nominal_freq);
 
@@ -196,6 +197,7 @@ static struct attribute *cppc_attrs[] = {
 	&highest_perf.attr,
 	&lowest_perf.attr,
 	&lowest_nonlinear_perf.attr,
+	&guaranteed_perf.attr,
 	&nominal_perf.attr,
 	&nominal_freq.attr,
 	&lowest_freq.attr,
Signature · 1 line
2.45.1
#1 lore raw
Quoted 6 lines · On Thu, 6 Jun 2024 13:55:41 +0200 Petr Tesarik <[email protected]> wrote:

Expose the CPPC guaranteed performance as reported by the platform through GuaranteedPerformanceRegister.

The current value is already read in cppc_get_perf_caps() and stored in struct cppc_perf_caps (to be used by the intel_pstate driver), so only the attribute itself needs to be defined.

Are there any objections to exposing this CPPC register through sysfs? I mean, if everybody is OK with it, the patch could be acked and queued for 6.11, right?

Petr T

Quoted 25 lines
---
 drivers/acpi/cppc_acpi.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/acpi/cppc_acpi.c +2 -0
index 1d857978f5f4..9976bb57356e 100644
@@ -160,6 +160,7 @@ show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, highest_perf);
 show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, lowest_perf);
 show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, nominal_perf);
 show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, lowest_nonlinear_perf);
+show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, guaranteed_perf);
 show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, lowest_freq);
 show_cppc_data(cppc_get_perf_caps, cppc_perf_caps, nominal_freq);
 
@@ -196,6 +197,7 @@ static struct attribute *cppc_attrs[] = {
 	&highest_perf.attr,
 	&lowest_perf.attr,
 	&lowest_nonlinear_perf.attr,
+	&guaranteed_perf.attr,
 	&nominal_perf.attr,
 	&nominal_freq.attr,
 	&lowest_freq.attr,
#2 lore raw
Quoted 7 lines · Petr Tesarik
Quoted 3 lines · On Thu, 6 Jun 2024 13:55:41 +0200 Petr Tesarik <[email protected]> wrote:

Expose the CPPC guaranteed performance as reported by the platform through GuaranteedPerformanceRegister.

Quoted 7 lines
Quoted 3 lines

The current value is already read in cppc_get_perf_caps() and stored in struct cppc_perf_caps (to be used by the intel_pstate driver), so only the attribute itself needs to be defined.

Are there any objections to exposing this CPPC register through sysfs? I mean, if everybody is OK with it, the patch could be acked and queued for 6.11, right?

It actually has been queued already, sorry for the missing notice.

It's been in linux-next for some time even.

#3 lore raw
Quoted 12 lines · On Fri, 21 Jun 2024 12:16:06 +0200 "Rafael J. Wysocki" <[email protected]>…
Quoted 11 lines · Petr Tesarik
Quoted 6 lines · On Thu, 6 Jun 2024 13:55:41 +0200 Petr Tesarik <[email protected]> wrote:

Expose the CPPC guaranteed performance as reported by the platform through GuaranteedPerformanceRegister.

The current value is already read in cppc_get_perf_caps() and stored in struct cppc_perf_caps (to be used by the intel_pstate driver), so only the attribute itself needs to be defined.

Quoted 7 lines
Quoted 3 lines

Are there any objections to exposing this CPPC register through sysfs? I mean, if everybody is OK with it, the patch could be acked and queued for 6.11, right?

It actually has been queued already, sorry for the missing notice.

It's been in linux-next for some time even.

Oh, OK, I could have checked myself. ;-)

Thanks!
Petr T