[PATCH 0/2] cpupower: Allow control of boost feature on non-x86 based systems with boost support.

[PATCH 0/2] cpupower: Allow control of boost feature on non-x86 based systems with boost support.

[Shinji Nomoto]

The cpufreq subsystem has a generic sysfs interface for controlling boost
(/sys/devices/system/cpu/cpufreq/boost).
The sysfs interface can be used to enable boost control from the cpupower
command on non-x86 platforms as well. So, allow boost controlling on
non-x86 system if boost sysfs file exists.

The set subcommand enables/disables the boost feature using the following
syntax:
  cpupower set --boost 1
  cpupower set --boost 0

The --boost option is an alias for --turbo-boost. We provided the neutral
option name because the name "turbo boost" is specific to Intel technology.

The frequency-info subcommand displays the enabled/disabled state of
the boost feature as follows:
  boost state support:
    Active: yes (or no)

Additionally, there was an existing bug in the option handling,
which I fixed in the first commit.  The fix is independent of the added
functionality.

Shinji Nomoto (2):
  cpupower: Fix a bug where the -t option of the set subcommand was not
    working.
  cpupower: Allow control of boost feature on non-x86 based systems with
    boost support.

 tools/power/cpupower/man/cpupower-set.1      |  7 +-
 tools/power/cpupower/utils/cpufreq-info.c    | 16 ++++-
 tools/power/cpupower/utils/cpupower-set.c    |  5 +-
 tools/power/cpupower/utils/helpers/helpers.h | 14 ++--
 tools/power/cpupower/utils/helpers/misc.c    | 76 ++++++++++++++------
 5 files changed, 83 insertions(+), 35 deletions(-)

-- 
2.43.0

[PATCH 1/2] cpupower: Fix a bug where the -t option of the set subcommand was not working.

[Shinji Nomoto]

The set subcommand's -t option is documented as being available for boost
configuration, but it was not actually functioning due to a bug
in the option handling.

Signed-off-by: Shinji Nomoto <fj5851bi@fujitsu.com>
---
 tools/power/cpupower/utils/cpupower-set.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/power/cpupower/utils/cpupower-set.c b/tools/power/cpupower/utils/cpupower-set.c
index 0677b58374ab..59ace394cf3e 100644
--- a/tools/power/cpupower/utils/cpupower-set.c
+++ b/tools/power/cpupower/utils/cpupower-set.c
@@ -62,8 +62,8 @@ int cmd_set(int argc, char **argv)
 
 	params.params = 0;
 	/* parameter parsing */
-	while ((ret = getopt_long(argc, argv, "b:e:m:",
-						set_opts, NULL)) != -1) {
+	while ((ret = getopt_long(argc, argv, "b:e:m:t:",
+				  set_opts, NULL)) != -1) {
 		switch (ret) {
 		case 'b':
 			if (params.perf_bias)
-- 
2.43.0

[PATCH 2/2] cpupower: Allow control of boost feature on non-x86 based systems with boost support.

[Shinji Nomoto]

The cpufreq subsystem has a generic sysfs interface for controlling boost
(/sys/devices/system/cpu/cpufreq/boost).
The sysfs interface can be used to enable boost control from the cpupower
command on non-x86 platforms as well. So, allow boost controlling
on non-x86 system if boost sysfs file exists.

The set subcommand enables/disables the boost feature using the following
syntax:
    cpupower set --boost 1
    cpupower set --boost 0

The --boost option is an alias for --turbo-boost. We provided the neutral
option name because the name "turbo boost" is specific to Intel technology.

The frequency-info subcommand displays the enabled/disabled state of
the boost feature as follows:
    boost state support:
        Active: yes (or no)

Signed-off-by: Shinji Nomoto <fj5851bi@fujitsu.com>
---
 tools/power/cpupower/man/cpupower-set.1      |  7 +-
 tools/power/cpupower/utils/cpufreq-info.c    | 16 ++++-
 tools/power/cpupower/utils/cpupower-set.c    |  1 +
 tools/power/cpupower/utils/helpers/helpers.h | 14 ++--
 tools/power/cpupower/utils/helpers/misc.c    | 76 ++++++++++++++------
 5 files changed, 81 insertions(+), 33 deletions(-)

diff --git a/tools/power/cpupower/man/cpupower-set.1 b/tools/power/cpupower/man/cpupower-set.1
index 500653ef98c7..8ac82b6f9189 100644
--- a/tools/power/cpupower/man/cpupower-set.1
+++ b/tools/power/cpupower/man/cpupower-set.1
@@ -81,10 +81,11 @@ Refer to the AMD P-State kernel documentation for further information.
 .RE
 
 .PP
-\-\-turbo\-boost, \-t
+\-\-turbo\-boost, \-\-boost, \-t
 .RS 4
-This option is used to enable or disable the turbo boost feature on
-supported Intel and AMD processors.
+This option is used to enable or disable the boost feature on
+supported Intel and AMD processors, and other boost supported systems.
+(The --boost option is an alias for the --turbo-boost option)
 
 This option takes as parameter either \fB1\fP to enable, or \fB0\fP to disable the feature.
 
diff --git a/tools/power/cpupower/utils/cpufreq-info.c b/tools/power/cpupower/utils/cpufreq-info.c
index fc750e127404..7d3732f5f2f6 100644
--- a/tools/power/cpupower/utils/cpufreq-info.c
+++ b/tools/power/cpupower/utils/cpufreq-info.c
@@ -128,7 +128,7 @@ static int get_boost_mode_x86(unsigned int cpu)
 	/* ToDo: Make this more global */
 	unsigned long pstates[MAX_HW_PSTATES] = {0,};
 
-	ret = cpufreq_has_boost_support(cpu, &support, &active, &b_states);
+	ret = cpufreq_has_x86_boost_support(cpu, &support, &active, &b_states);
 	if (ret) {
 		printf(_("Error while evaluating Boost Capabilities"
 				" on CPU %d -- are you root?\n"), cpu);
@@ -204,6 +204,18 @@ static int get_boost_mode_x86(unsigned int cpu)
 	return 0;
 }
 
+static int get_boost_mode_generic(unsigned int cpu)
+{
+	bool active;
+
+	if (!cpufreq_has_generic_boost_support(&active)) {
+		printf(_("  boost state support:\n"));
+		printf(_("    Active: %s\n"), active ? _("yes") : _("no"));
+	}
+
+	return 0;
+}
+
 /* --boost / -b */
 
 static int get_boost_mode(unsigned int cpu)
@@ -214,6 +226,8 @@ static int get_boost_mode(unsigned int cpu)
 	    cpupower_cpu_info.vendor == X86_VENDOR_HYGON ||
 	    cpupower_cpu_info.vendor == X86_VENDOR_INTEL)
 		return get_boost_mode_x86(cpu);
+	else
+		get_boost_mode_generic(cpu);
 
 	freqs = cpufreq_get_boost_frequencies(cpu);
 	if (freqs) {
diff --git a/tools/power/cpupower/utils/cpupower-set.c b/tools/power/cpupower/utils/cpupower-set.c
index 59ace394cf3e..c2117e5650dd 100644
--- a/tools/power/cpupower/utils/cpupower-set.c
+++ b/tools/power/cpupower/utils/cpupower-set.c
@@ -21,6 +21,7 @@ static struct option set_opts[] = {
 	{"epp", required_argument, NULL, 'e'},
 	{"amd-pstate-mode", required_argument, NULL, 'm'},
 	{"turbo-boost", required_argument, NULL, 't'},
+	{"boost", required_argument, NULL, 't'},
 	{ },
 };
 
diff --git a/tools/power/cpupower/utils/helpers/helpers.h b/tools/power/cpupower/utils/helpers/helpers.h
index 95749b8ee475..82ea62bdf5a2 100644
--- a/tools/power/cpupower/utils/helpers/helpers.h
+++ b/tools/power/cpupower/utils/helpers/helpers.h
@@ -103,6 +103,9 @@ extern struct cpupower_cpu_info cpupower_cpu_info;
 
 /* cpuid and cpuinfo helpers  **************************/
 
+int cpufreq_has_generic_boost_support(bool *active);
+int cpupower_set_turbo_boost(int turbo_boost);
+
 /* X86 ONLY ****************************************/
 #if defined(__i386__) || defined(__x86_64__)
 
@@ -118,7 +121,6 @@ extern unsigned long long msr_intel_get_turbo_ratio(unsigned int cpu);
 
 extern int cpupower_set_epp(unsigned int cpu, char *epp);
 extern int cpupower_set_amd_pstate_mode(char *mode);
-extern int cpupower_set_turbo_boost(int turbo_boost);
 
 /* Read/Write msr ****************************/
 
@@ -139,8 +141,8 @@ extern int decode_pstates(unsigned int cpu, int boost_states,
 
 /* AMD HW pstate decoding **************************/
 
-extern int cpufreq_has_boost_support(unsigned int cpu, int *support,
-				     int *active, int * states);
+int cpufreq_has_x86_boost_support(unsigned int cpu, int *support,
+				  int *active, int *states);
 
 /* AMD P-State stuff **************************/
 bool cpupower_amd_pstate_enabled(void);
@@ -181,13 +183,11 @@ static inline int cpupower_set_epp(unsigned int cpu, char *epp)
 { return -1; };
 static inline int cpupower_set_amd_pstate_mode(char *mode)
 { return -1; };
-static inline int cpupower_set_turbo_boost(int turbo_boost)
-{ return -1; };
 
 /* Read/Write msr ****************************/
 
-static inline int cpufreq_has_boost_support(unsigned int cpu, int *support,
-					    int *active, int * states)
+static inline int cpufreq_has_x86_boost_support(unsigned int cpu, int *support,
+						int *active, int *states)
 { return -1; }
 
 static inline bool cpupower_amd_pstate_enabled(void)
diff --git a/tools/power/cpupower/utils/helpers/misc.c b/tools/power/cpupower/utils/helpers/misc.c
index 76e461ff4f74..166dc1e470ea 100644
--- a/tools/power/cpupower/utils/helpers/misc.c
+++ b/tools/power/cpupower/utils/helpers/misc.c
@@ -8,15 +8,14 @@
 #include "helpers/helpers.h"
 #include "helpers/sysfs.h"
 #include "cpufreq.h"
+#include "cpupower_intern.h"
 
 #if defined(__i386__) || defined(__x86_64__)
 
-#include "cpupower_intern.h"
-
 #define MSR_AMD_HWCR	0xc0010015
 
-int cpufreq_has_boost_support(unsigned int cpu, int *support, int *active,
-			int *states)
+int cpufreq_has_x86_boost_support(unsigned int cpu, int *support, int *active,
+				  int *states)
 {
 	int ret;
 	unsigned long long val;
@@ -124,24 +123,6 @@ int cpupower_set_amd_pstate_mode(char *mode)
 	return 0;
 }
 
-int cpupower_set_turbo_boost(int turbo_boost)
-{
-	char path[SYSFS_PATH_MAX];
-	char linebuf[2] = {};
-
-	snprintf(path, sizeof(path), PATH_TO_CPU "cpufreq/boost");
-
-	if (!is_valid_path(path))
-		return -1;
-
-	snprintf(linebuf, sizeof(linebuf), "%d", turbo_boost);
-
-	if (cpupower_write_sysfs(path, linebuf, 2) <= 0)
-		return -1;
-
-	return 0;
-}
-
 bool cpupower_amd_pstate_enabled(void)
 {
 	char *driver = cpufreq_get_driver(0);
@@ -160,6 +141,39 @@ bool cpupower_amd_pstate_enabled(void)
 
 #endif /* #if defined(__i386__) || defined(__x86_64__) */
 
+int cpufreq_has_generic_boost_support(bool *active)
+{
+	char path[SYSFS_PATH_MAX];
+	char linebuf[2] = {};
+	unsigned long val;
+	char *endp;
+
+	snprintf(path, sizeof(path), PATH_TO_CPU "cpufreq/boost");
+
+	if (!is_valid_path(path))
+		return -EACCES;
+
+	if (cpupower_read_sysfs(path, linebuf, 2) <= 0)
+		return -EINVAL;
+
+	val = strtoul(linebuf, &endp, 0);
+	if (endp == linebuf || errno == ERANGE)
+		return -EINVAL;
+
+	switch (val) {
+	case 0:
+		*active = false;
+		break;
+	case 1:
+		*active = true;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 /* get_cpustate
  *
  * Gather the information of all online CPUs into bitmask struct
@@ -259,3 +273,21 @@ void print_speed(unsigned long speed, int no_rounding)
 		}
 	}
 }
+
+int cpupower_set_turbo_boost(int turbo_boost)
+{
+	char path[SYSFS_PATH_MAX];
+	char linebuf[2] = {};
+
+	snprintf(path, sizeof(path), PATH_TO_CPU "cpufreq/boost");
+
+	if (!is_valid_path(path))
+		return -1;
+
+	snprintf(linebuf, sizeof(linebuf), "%d", turbo_boost);
+
+	if (cpupower_write_sysfs(path, linebuf, 2) <= 0)
+		return -1;
+
+	return 0;
+}
-- 
2.43.0

Re: [PATCH 0/2] cpupower: Allow control of boost feature on non-x86 based systems with boost support.

[John B. Wyatt IV]

Hello Shinji, ty for your patch.

On Thu, May 22, 2025 at 03:10:57PM +0900, Shinji Nomoto wrote:
> The cpufreq subsystem has a generic sysfs interface for controlling boost
> (/sys/devices/system/cpu/cpufreq/boost).
> The sysfs interface can be used to enable boost control from the cpupower
> command on non-x86 platforms as well. So, allow boost controlling on
> non-x86 system if boost sysfs file exists.

Did you test this on non-x86 systems? If so, would you please provide
details on those architectures and systems?

-- 
Sincerely,
John Wyatt
Software Engineer, Core Kernel
Red Hat

Re: [PATCH 0/2] cpupower: Allow control of boost feature on non-x86 based systems with boost support.

[Shuah Khan]

On 5/29/25 05:48, John B. Wyatt IV wrote:
> Hello Shinji, ty for your patch.
> 
> On Thu, May 22, 2025 at 03:10:57PM +0900, Shinji Nomoto wrote:
>> The cpufreq subsystem has a generic sysfs interface for controlling boost
>> (/sys/devices/system/cpu/cpufreq/boost).
>> The sysfs interface can be used to enable boost control from the cpupower
>> command on non-x86 platforms as well. So, allow boost controlling on
>> non-x86 system if boost sysfs file exists.
> 
> Did you test this on non-x86 systems? If so, would you please provide
> details on those architectures and systems?
> 

I applied these to coupower - I will go drop them now.

thanks,
-- Shuah

RE: [PATCH 0/2] cpupower: Allow control of boost feature on non-x86 based systems with boost support.

[Shinji Nomoto (Fujitsu)]

John Wyatt wrote:
> Hello Shinji, ty for your patch.
> 
> On Thu, May 22, 2025 at 03:10:57PM +0900, Shinji Nomoto wrote:
> > The cpufreq subsystem has a generic sysfs interface for controlling 
> > boost (/sys/devices/system/cpu/cpufreq/boost).
> > The sysfs interface can be used to enable boost control from the 
> > cpupower command on non-x86 platforms as well. So, allow boost 
> > controlling on
> > non-x86 system if boost sysfs file exists.
>
> Did you test this on non-x86 systems? If so, would you please provide details on those architectures and systems?

Hello,

This patch has been tested on Arm-based systems.

* It was tested on our internal simulator based on QEMU which supports boost.
* It was tested on the Nvidia grace system (which does not support boost).
* The cppc_cpufreq driver is working on both of the above systems.

We have also confirmed that it continues to work as expected on AMD systems.

Regards.
----
Shinji

Re: [PATCH 0/2] cpupower: Allow control of boost feature on non-x86 based systems with boost support.

[Shuah Khan]

On 6/3/25 01:16, Shinji Nomoto (Fujitsu) wrote:
> John Wyatt wrote:
>> Hello Shinji, ty for your patch.
>>
>> On Thu, May 22, 2025 at 03:10:57PM +0900, Shinji Nomoto wrote:
>>> The cpufreq subsystem has a generic sysfs interface for controlling
>>> boost (/sys/devices/system/cpu/cpufreq/boost).
>>> The sysfs interface can be used to enable boost control from the
>>> cpupower command on non-x86 platforms as well. So, allow boost
>>> controlling on
>>> non-x86 system if boost sysfs file exists.
>>
>> Did you test this on non-x86 systems? If so, would you please provide details on those architectures and systems?
> 
> Hello,
> 
> This patch has been tested on Arm-based systems.
> 
> * It was tested on our internal simulator based on QEMU which supports boost.
> * It was tested on the Nvidia grace system (which does not support boost).
> * The cppc_cpufreq driver is working on both of the above systems.
> 
> We have also confirmed that it continues to work as expected on AMD systems.
> 

John,

Let me know if you are good with this testing details.

thanks,
-- Shuah

Re: [PATCH 0/2] cpupower: Allow control of boost feature on non-x86 based systems with boost support.

[John B. Wyatt IV]

On Wed, Jun 18, 2025 at 03:54:56PM -0600, Shuah Khan wrote:
> On 6/3/25 01:16, Shinji Nomoto (Fujitsu) wrote:
> > John Wyatt wrote:
> > > Did you test this on non-x86 systems? If so, would you please provide details on those architectures and systems?
> > 
> > Hello,
> > 
> > This patch has been tested on Arm-based systems.
> > 
> > * It was tested on our internal simulator based on QEMU which supports boost.
> > * It was tested on the Nvidia grace system (which does not support boost).
> > * The cppc_cpufreq driver is working on both of the above systems.
> > 
> > We have also confirmed that it continues to work as expected on AMD systems.
> > 
> 
> John,
> 
> Let me know if you are good with this testing details.

I have not tested it, but I the testing details are fine.

-- 
Sincerely,
John Wyatt
Software Engineer, Core Kernel
Red Hat

RE: [PATCH 0/2] cpupower: Allow control of boost feature on non-x86 based systems with boost support.

[Shinji Nomoto (Fujitsu)]

Hi, John and Shuah

Regarding the subject, I submitted patches to cpupower, but so far I haven't received confirmation that the content has been approved, nor does it appear to have been merged into the shuah/linux.git master branch.
Could you please inform me about the current status of the patches?

Thank you for your time and consideration.

----
Shinji Nomoto

-----Original Message-----
From: John B. Wyatt IV <jwyatt@redhat.com> 
Sent: Saturday, June 21, 2025 2:27 AM
To: Shuah Khan <skhan@linuxfoundation.org>
Cc: Nomoto, Shinji/野本 真二 <fj5851bi@fujitsu.com>; Thomas Renninger <trenn@suse.com>; Shuah Khan <shuah@kernel.org>; John Kacur <jkacur@redhat.com>; linux-pm@vger.kernel.org; linux-kernel@vger.kernel.org
Subject: Re: [PATCH 0/2] cpupower: Allow control of boost feature on non-x86 based systems with boost support.

On Wed, Jun 18, 2025 at 03:54:56PM -0600, Shuah Khan wrote:
> On 6/3/25 01:16, Shinji Nomoto (Fujitsu) wrote:
> > John Wyatt wrote:
> > > Did you test this on non-x86 systems? If so, would you please provide details on those architectures and systems?
> > 
> > Hello,
> > 
> > This patch has been tested on Arm-based systems.
> > 
> > * It was tested on our internal simulator based on QEMU which supports boost.
> > * It was tested on the Nvidia grace system (which does not support boost).
> > * The cppc_cpufreq driver is working on both of the above systems.
> > 
> > We have also confirmed that it continues to work as expected on AMD systems.
> > 
> 
> John,
> 
> Let me know if you are good with this testing details.

I have not tested it, but I the testing details are fine.

-- 
Sincerely,
John Wyatt
Software Engineer, Core Kernel
Red Hat

Re: [PATCH 0/2] cpupower: Allow control of boost feature on non-x86 based systems with boost support.

[John B. Wyatt IV]

On Tue, Jul 22, 2025 at 07:30:27AM +0000, Shinji Nomoto (Fujitsu) wrote:
> Hi, John and Shuah
> 
> Regarding the subject, I submitted patches to cpupower, but so far I haven't received confirmation that the content has been approved, nor does it appear to have been merged into the shuah/linux.git master branch.
> Could you please inform me about the current status of the patches?
> 
> Thank you for your time and consideration.

I do not see any issues with the answer Shinji gave on testing.

I am fine with it being merged.

-- 
Sincerely,
John Wyatt
Software Engineer, Core Kernel
Red Hat

Re: [PATCH 0/2] cpupower: Allow control of boost feature on non-x86 based systems with boost support.

[Shuah Khan]

On 7/22/25 11:59, John B. Wyatt IV wrote:
> On Tue, Jul 22, 2025 at 07:30:27AM +0000, Shinji Nomoto (Fujitsu) wrote:
>> Hi, John and Shuah
>>
>> Regarding the subject, I submitted patches to cpupower, but so far I haven't received confirmation that the content has been approved, nor does it appear to have been merged into the shuah/linux.git master branch.
>> Could you please inform me about the current status of the patches?
>>
>> Thank you for your time and consideration.
> 
> I do not see any issues with the answer Shinji gave on testing.
> 
> I am fine with it being merged.
> 

Thank you. I thought I pulled these in for Linux 6.17-rc1.
It appears I didn't. I applied these now to my cpupower
branch.

These will be in my next pr to Rafael which will be after
rc1 comes out.

thanks,
-- Shuah