[ACPI/CPUFREQ] Introduce bios_limit per cpu cpufreq sysfs interface

This interface is mainly intended (and implemented) for ACPI _PPC BIOS frequency limitations, but other cpufreq drivers can also use it for similar use-cases. Why is this needed: Currently it's not obvious why cpufreq got limited. People see cpufreq/scaling_max_freq reduced, but this could have happened by: - any userspace prog writing to scaling_max_freq - thermal limitations - hardware (_PPC in ACPI case) limitiations Therefore export bios_limit (in kHz) to: - Point the user that it's the BIOS (broken or intended) which limits frequency - Export it as a sysfs interface for userspace progs. While this was a rarely used feature on laptops, there will appear more and more server implemenations providing "Green IT" features like allowing the service processor to limit the frequency. People want to know about HW/BIOS frequency limitations. All ACPI P-state driven cpufreq drivers are covered with this patch: - powernow-k8 - powernow-k7 - acpi-cpufreq Tested with a patched DSDT which limits the first two cores (_PPC returns 1) via _PPC, exposed by bios_limit: # echo 2200000 >cpu2/cpufreq/scaling_max_freq # cat cpu*/cpufreq/scaling_max_freq 2600000 2600000 2200000 2200000 # #scaling_max_freq shows general user/thermal/BIOS limitations # cat cpu*/cpufreq/bios_limit 2600000 2600000 2800000 2800000 # #bios_limit only shows the HW/BIOS limitation CC: Pallipadi Venkatesh <venkatesh.pallipadi@intel.com> CC: Len Brown <lenb@kernel.org> CC: davej@codemonkey.org.uk CC: linux@dominikbrodowski.net Signed-off-by: Thomas Renninger <trenn@suse.de> Signed-off-by: Dave Jones <davej@redhat.com>

[ACPI/CPUFREQ] Introduce bios_limit per cpu cpufreq sysfs interface
This interface is mainly intended (and implemented) for ACPI _PPC BIOS frequency limitations, but other cpufreq drivers can also use it for similar use-cases. Why is this needed: Currently it's not obvious why cpufreq got limited. People see cpufreq/scaling_max_freq reduced, but this could have happened by: - any userspace prog writing to scaling_max_freq - thermal limitations - hardware (_PPC in ACPI case) limitiations Therefore export bios_limit (in kHz) to: - Point the user that it's the BIOS (broken or intended) which limits frequency - Export it as a sysfs interface for userspace progs. While this was a rarely used feature on laptops, there will appear more and more server implemenations providing "Green IT" features like allowing the service processor to limit the frequency. People want to know about HW/BIOS frequency limitations. All ACPI P-state driven cpufreq drivers are covered with this patch: - powernow-k8 - powernow-k7 - acpi-cpufreq Tested with a patched DSDT which limits the first two cores (_PPC returns 1) via _PPC, exposed by bios_limit: # echo 2200000 >cpu2/cpufreq/scaling_max_freq # cat cpu*/cpufreq/scaling_max_freq 2600000 2600000 2200000 2200000 # #scaling_max_freq shows general user/thermal/BIOS limitations # cat cpu*/cpufreq/bios_limit 2600000 2600000 2800000 2800000 # #bios_limit only shows the HW/BIOS limitation CC: Pallipadi Venkatesh <venkatesh.pallipadi@intel.com> CC: Len Brown <lenb@kernel.org> CC: davej@codemonkey.org.uk CC: linux@dominikbrodowski.net Signed-off-by: Thomas Renninger <trenn@suse.de> Signed-off-by: Dave Jones <davej@redhat.com>
Thomas Renninger · Dave Jones
1 parent cf3289d0e7
Showing 8 changed files with 81 additions and 24 deletions Side-by-side Diff
Documentation/cpu-freq/user-guide.txt
arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
arch/x86/kernel/cpu/cpufreq/powernow-k7.c
arch/x86/kernel/cpu/cpufreq/powernow-k8.c
drivers/acpi/processor_perflib.c
drivers/cpufreq/cpufreq.c
include/acpi/processor.h
include/linux/cpufreq.h
@@ -203,6 +203,17 @@
 				the frequency the kernel thinks the CPU runs
 				at.
  
+bios_limit :			If the BIOS tells the OS to limit a CPU to
+				lower frequencies, the user can read out the
+				maximum available frequency from this file.
+				This typically can happen through (often not
+				intended) BIOS settings, restrictions
+				triggered through a service processor or other
+				BIOS/HW based implementations.
+				This does not cover thermal ACPI limitations
+				which can be detected through the generic
+				thermal driver.
+
 If you have selected the "userspace" governor which allows you to
 set the CPU operating frequency to a specific value, you can read out
 the current frequency in
@@ -764,14 +764,15 @@
 };
  
 static struct cpufreq_driver acpi_cpufreq_driver = {
-	.verify = acpi_cpufreq_verify,
-	.target = acpi_cpufreq_target,
-	.init = acpi_cpufreq_cpu_init,
-	.exit = acpi_cpufreq_cpu_exit,
-	.resume = acpi_cpufreq_resume,
-	.name = "acpi-cpufreq",
-	.owner = THIS_MODULE,
-	.attr = acpi_cpufreq_attr,
+	.verify		= acpi_cpufreq_verify,
+	.target		= acpi_cpufreq_target,
+	.bios_limit	= acpi_processor_get_bios_limit,
+	.init		= acpi_cpufreq_cpu_init,
+	.exit		= acpi_cpufreq_cpu_exit,
+	.resume		= acpi_cpufreq_resume,
+	.name		= "acpi-cpufreq",
+	.owner		= THIS_MODULE,
+	.attr		= acpi_cpufreq_attr,
 };
  
 static int __init acpi_cpufreq_init(void)
@@ -714,14 +714,17 @@
 };
  
 static struct cpufreq_driver powernow_driver = {
-	.verify	= powernow_verify,
-	.target	= powernow_target,
-	.get	= powernow_get,
-	.init	= powernow_cpu_init,
-	.exit	= powernow_cpu_exit,
-	.name	= "powernow-k7",
-	.owner	= THIS_MODULE,
-	.attr	= powernow_table_attr,
+	.verify		= powernow_verify,
+	.target		= powernow_target,
+	.get		= powernow_get,
+#ifdef CONFIG_X86_POWERNOW_K7_ACPI
+	.bios_limit	= acpi_processor_get_bios_limit,
+#endif
+	.init		= powernow_cpu_init,
+	.exit		= powernow_cpu_exit,
+	.name		= "powernow-k7",
+	.owner		= THIS_MODULE,
+	.attr		= powernow_table_attr,
 };
  
 static int __init powernow_init(void)
@@ -1398,14 +1398,15 @@
 };
  
 static struct cpufreq_driver cpufreq_amd64_driver = {
-	.verify = powernowk8_verify,
-	.target = powernowk8_target,
-	.init = powernowk8_cpu_init,
-	.exit = __devexit_p(powernowk8_cpu_exit),
-	.get = powernowk8_get,
-	.name = "powernow-k8",
-	.owner = THIS_MODULE,
-	.attr = powernow_k8_attr,
+	.verify		= powernowk8_verify,
+	.target		= powernowk8_target,
+	.bios_limit	= acpi_processor_get_bios_limit,
+	.init		= powernowk8_cpu_init,
+	.exit		= __devexit_p(powernowk8_cpu_exit),
+	.get		= powernowk8_get,
+	.name		= "powernow-k8",
+	.owner		= THIS_MODULE,
+	.attr		= powernow_k8_attr,
 };
  
 /* driver entry point for init */
@@ -167,6 +167,19 @@
 		return cpufreq_update_policy(pr->id);
 }
  
+int acpi_processor_get_bios_limit(int cpu, unsigned int *limit)
+{
+	struct acpi_processor *pr;
+
+	pr = per_cpu(processors, cpu);
+	if (!pr || !pr->performance || !pr->performance->state_count)
+		return -ENODEV;
+	*limit = pr->performance->states[pr->performance_platform_limit].
+		core_frequency * 1000;
+	return 0;
+}
+EXPORT_SYMBOL(acpi_processor_get_bios_limit);
+
 void acpi_processor_ppc_init(void)
 {
 	if (!cpufreq_register_notifier
@@ -647,6 +647,21 @@
 	return policy->governor->show_setspeed(policy, buf);
 }
  
+/**
+ * show_scaling_driver - show the current cpufreq HW/BIOS limitation
+ */
+static ssize_t show_bios_limit(struct cpufreq_policy *policy, char *buf)
+{
+	unsigned int limit;
+	int ret;
+	if (cpufreq_driver->bios_limit) {
+		ret = cpufreq_driver->bios_limit(policy->cpu, &limit);
+		if (!ret)
+			return sprintf(buf, "%u\n", limit);
+	}
+	return sprintf(buf, "%u\n", policy->cpuinfo.max_freq);
+}
+
 #define define_one_ro(_name) \
 static struct freq_attr _name = \
 __ATTR(_name, 0444, show_##_name, NULL)
@@ -666,6 +681,7 @@
 define_one_ro(scaling_available_governors);
 define_one_ro(scaling_driver);
 define_one_ro(scaling_cur_freq);
+define_one_ro(bios_limit);
 define_one_ro(related_cpus);
 define_one_ro(affected_cpus);
 define_one_rw(scaling_min_freq);
@@ -902,6 +918,11 @@
 	}
 	if (cpufreq_driver->target) {
 		ret = sysfs_create_file(&policy->kobj, &scaling_cur_freq.attr);
+		if (ret)
+			goto err_out_kobj_put;
+	}
+	if (cpufreq_driver->bios_limit) {
+		ret = sysfs_create_file(&policy->kobj, &bios_limit.attr);
 		if (ret)
 			goto err_out_kobj_put;
 	}
@@ -295,6 +295,7 @@
 void acpi_processor_ppc_init(void);
 void acpi_processor_ppc_exit(void);
 int acpi_processor_ppc_has_changed(struct acpi_processor *pr);
+extern int acpi_processor_get_bios_limit(int cpu, unsigned int *limit);
 #else
 static inline void acpi_processor_ppc_init(void)
 {
@@ -316,6 +317,11 @@
 	}
 	return 0;
 }
+static inline int acpi_processor_get_bios_limit(int cpu, unsigned int *limit)
+{
+	return -ENODEV;
+}
+
 #endif				/* CONFIG_CPU_FREQ */
  
 /* in processor_throttling.c */
@@ -232,6 +232,7 @@
 	/* optional */
 	unsigned int (*getavg)	(struct cpufreq_policy *policy,
 				 unsigned int cpu);
+	int	(*bios_limit)	(int cpu, unsigned int *limit);
  
 	int	(*exit)		(struct cpufreq_policy *policy);
 	int	(*suspend)	(struct cpufreq_policy *policy, pm_message_t pmsg);
...	...	@@ -203,6 +203,17 @@
203	203	the frequency the kernel thinks the CPU runs
204	204	at.
205	205
	206	+bios_limit : If the BIOS tells the OS to limit a CPU to
	207	+ lower frequencies, the user can read out the
	208	+ maximum available frequency from this file.
	209	+ This typically can happen through (often not
	210	+ intended) BIOS settings, restrictions
	211	+ triggered through a service processor or other
	212	+ BIOS/HW based implementations.
	213	+ This does not cover thermal ACPI limitations
	214	+ which can be detected through the generic
	215	+ thermal driver.
	216	+
206	217	If you have selected the "userspace" governor which allows you to
207	218	set the CPU operating frequency to a specific value, you can read out
208	219	the current frequency in
...	...	@@ -764,14 +764,15 @@
764	764	};
765	765
766	766	static struct cpufreq_driver acpi_cpufreq_driver = {
767		- .verify = acpi_cpufreq_verify,
768		- .target = acpi_cpufreq_target,
769		- .init = acpi_cpufreq_cpu_init,
770		- .exit = acpi_cpufreq_cpu_exit,
771		- .resume = acpi_cpufreq_resume,
772		- .name = "acpi-cpufreq",
773		- .owner = THIS_MODULE,
774		- .attr = acpi_cpufreq_attr,
	767	+ .verify = acpi_cpufreq_verify,
	768	+ .target = acpi_cpufreq_target,
	769	+ .bios_limit = acpi_processor_get_bios_limit,
	770	+ .init = acpi_cpufreq_cpu_init,
	771	+ .exit = acpi_cpufreq_cpu_exit,
	772	+ .resume = acpi_cpufreq_resume,
	773	+ .name = "acpi-cpufreq",
	774	+ .owner = THIS_MODULE,
	775	+ .attr = acpi_cpufreq_attr,
775	776	};
776	777
777	778	static int __init acpi_cpufreq_init(void)
...	...	@@ -714,14 +714,17 @@
714	714	};
715	715
716	716	static struct cpufreq_driver powernow_driver = {
717		- .verify = powernow_verify,
718		- .target = powernow_target,
719		- .get = powernow_get,
720		- .init = powernow_cpu_init,
721		- .exit = powernow_cpu_exit,
722		- .name = "powernow-k7",
723		- .owner = THIS_MODULE,
724		- .attr = powernow_table_attr,
	717	+ .verify = powernow_verify,
	718	+ .target = powernow_target,
	719	+ .get = powernow_get,
	720	+#ifdef CONFIG_X86_POWERNOW_K7_ACPI
	721	+ .bios_limit = acpi_processor_get_bios_limit,
	722	+#endif
	723	+ .init = powernow_cpu_init,
	724	+ .exit = powernow_cpu_exit,
	725	+ .name = "powernow-k7",
	726	+ .owner = THIS_MODULE,
	727	+ .attr = powernow_table_attr,
725	728	};
726	729
727	730	static int __init powernow_init(void)
...	...	@@ -1398,14 +1398,15 @@
1398	1398	};
1399	1399
1400	1400	static struct cpufreq_driver cpufreq_amd64_driver = {
1401		- .verify = powernowk8_verify,
1402		- .target = powernowk8_target,
1403		- .init = powernowk8_cpu_init,
1404		- .exit = __devexit_p(powernowk8_cpu_exit),
1405		- .get = powernowk8_get,
1406		- .name = "powernow-k8",
1407		- .owner = THIS_MODULE,
1408		- .attr = powernow_k8_attr,
	1401	+ .verify = powernowk8_verify,
	1402	+ .target = powernowk8_target,
	1403	+ .bios_limit = acpi_processor_get_bios_limit,
	1404	+ .init = powernowk8_cpu_init,
	1405	+ .exit = __devexit_p(powernowk8_cpu_exit),
	1406	+ .get = powernowk8_get,
	1407	+ .name = "powernow-k8",
	1408	+ .owner = THIS_MODULE,
	1409	+ .attr = powernow_k8_attr,
1409	1410	};
1410	1411
1411	1412	/* driver entry point for init */
...	...	@@ -167,6 +167,19 @@
167	167	return cpufreq_update_policy(pr->id);
168	168	}
169	169
	170	+int acpi_processor_get_bios_limit(int cpu, unsigned int *limit)
	171	+{
	172	+ struct acpi_processor *pr;
	173	+
	174	+ pr = per_cpu(processors, cpu);
	175	+ if (!pr \|\| !pr->performance \|\| !pr->performance->state_count)
	176	+ return -ENODEV;
	177	+ *limit = pr->performance->states[pr->performance_platform_limit].
	178	+ core_frequency * 1000;
	179	+ return 0;
	180	+}
	181	+EXPORT_SYMBOL(acpi_processor_get_bios_limit);
	182	+
170	183	void acpi_processor_ppc_init(void)
171	184	{
172	185	if (!cpufreq_register_notifier
...	...	@@ -647,6 +647,21 @@
647	647	return policy->governor->show_setspeed(policy, buf);
648	648	}
649	649
	650	+/**
	651	+ * show_scaling_driver - show the current cpufreq HW/BIOS limitation
	652	+ */
	653	+static ssize_t show_bios_limit(struct cpufreq_policy policy, char buf)
	654	+{
	655	+ unsigned int limit;
	656	+ int ret;
	657	+ if (cpufreq_driver->bios_limit) {
	658	+ ret = cpufreq_driver->bios_limit(policy->cpu, &limit);
	659	+ if (!ret)
	660	+ return sprintf(buf, "%u\n", limit);
	661	+ }
	662	+ return sprintf(buf, "%u\n", policy->cpuinfo.max_freq);
	663	+}
	664	+
650	665	#define define_one_ro(_name) \
651	666	static struct freq_attr _name = \
652	667	__ATTR(_name, 0444, show_##_name, NULL)
...	...	@@ -666,6 +681,7 @@
666	681	define_one_ro(scaling_available_governors);
667	682	define_one_ro(scaling_driver);
668	683	define_one_ro(scaling_cur_freq);
	684	+define_one_ro(bios_limit);
669	685	define_one_ro(related_cpus);
670	686	define_one_ro(affected_cpus);
671	687	define_one_rw(scaling_min_freq);
...	...	@@ -902,6 +918,11 @@
902	918	}
903	919	if (cpufreq_driver->target) {
904	920	ret = sysfs_create_file(&policy->kobj, &scaling_cur_freq.attr);
	921	+ if (ret)
	922	+ goto err_out_kobj_put;
	923	+ }
	924	+ if (cpufreq_driver->bios_limit) {
	925	+ ret = sysfs_create_file(&policy->kobj, &bios_limit.attr);
905	926	if (ret)
906	927	goto err_out_kobj_put;
907	928	}
...	...	@@ -295,6 +295,7 @@
295	295	void acpi_processor_ppc_init(void);
296	296	void acpi_processor_ppc_exit(void);
297	297	int acpi_processor_ppc_has_changed(struct acpi_processor *pr);
	298	+extern int acpi_processor_get_bios_limit(int cpu, unsigned int *limit);
298	299	#else
299	300	static inline void acpi_processor_ppc_init(void)
300	301	{
...	...	@@ -316,6 +317,11 @@
316	317	}
317	318	return 0;
318	319	}
	320	+static inline int acpi_processor_get_bios_limit(int cpu, unsigned int *limit)
	321	+{
	322	+ return -ENODEV;
	323	+}
	324	+
319	325	#endif /* CONFIG_CPU_FREQ */
320	326
321	327	/* in processor_throttling.c */
...	...	@@ -232,6 +232,7 @@
232	232	/* optional */
233	233	unsigned int (getavg) (struct cpufreq_policy policy,
234	234	unsigned int cpu);
	235	+ int (bios_limit) (int cpu, unsigned int limit);
235	236
236	237	int (exit) (struct cpufreq_policy policy);
237	238	int (suspend) (struct cpufreq_policy policy, pm_message_t pmsg);