Skip to contentSkip to conversation

cpufreq: intel_pstate: Address some HWP-related oddities

PATCH Superseded Rafael J. Wysocki 2 participants 9 messages Aug 20, 2020 – Aug 24, 2020

Subsystems (2)

v1 of 4 Superseded Latest is v4 →

lore
Share & tools
lore
#0 lore raw

cover cpufreq: intel_pstate: Address some HWP-related oddities

Hi All,

The purpose of this series is to address some peculiarities related to taking CPUs offline/online and switching between different operation modes with HWP enabled that have become visible after allowing the driver to work in the passive mode with HWP enabled in 5.9-rc1 (and one that was there earlier, but can be addressed easily after the changes madein 5.9-rc1).

Please refer to the patch changelogs for details.

For easier testing/review, the series is available from the git branch at:

git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git \
intel_pstate-testing

Thanks,
Rafael

#1 lore raw

PATCH 1/4 cpufreq: intel_pstate: Refuse to turn off with HWP enabled

From: "Rafael J. Wysocki" <[email protected]>

After commit f6ebbcf08f37 ("cpufreq: intel_pstate: Implement passive mode with HWP enabled") it is possible to change the driver status to "off" via sysfs with HWP enabled, which effectively causes the driver to unregister itself, but HWP remains active and it forces the minimum performance, so even if another cpufreq driver is loaded, it will not be able to control the CPU frequency.

For this reason, make the driver refuse to change the status to "off" with HWP enabled.

---
 drivers/cpufreq/intel_pstate.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)
drivers/cpufreq/intel_pstate.c +9 -3
index e0220a6fbc69..bcda1e700a73 100644
@@ -2716,9 +2716,15 @@ static int intel_pstate_update_status(const char *buf, size_t size)
 {
 	int ret;
 
-	if (size == 3 && !strncmp(buf, "off", size))
-		return intel_pstate_driver ?
-			intel_pstate_unregister_driver() : -EINVAL;
+	if (size == 3 && !strncmp(buf, "off", size)) {
+		if (!intel_pstate_driver)
+			return -EINVAL;
+
+		if (hwp_active)
+			return -EBUSY;
+
+		return intel_pstate_unregister_driver();
+	}
 
 	if (size == 6 && !strncmp(buf, "active", size)) {
 		if (intel_pstate_driver) {
Signature · 1 line
2.26.2
#2 lore raw

PATCH 2/4 cpufreq: intel_pstate: Always return last EPP value from sysfs

In reply to Rafael J. Wysocki

From: "Rafael J. Wysocki" <[email protected]>

Make the energy_performance_preference policy attribute in sysfs always return the last EPP value written to it instead of the one currently in the HWP Request MSR to avoid possible confusion when the performance scaling algorithm is used in the active mode with HWP enabled (in which case the EPP is forced to 0 regardless of what value it has been set to via sysfs).

---
 drivers/cpufreq/intel_pstate.c | 26 +++++++++++++++++---------
 1 file changed, 17 insertions(+), 9 deletions(-)
drivers/cpufreq/intel_pstate.c +17 -9
index bcda1e700a73..3d18934fa975 100644
@@ -606,13 +606,10 @@ static const unsigned int epp_values[] = {
 
 static int intel_pstate_get_energy_pref_index(struct cpudata *cpu_data, int *raw_epp)
 {
-	s16 epp;
+	s16 epp = cpu_data->epp_cached;
 	int index = -EINVAL;
 
 	*raw_epp = 0;
-	epp = intel_pstate_get_epp(cpu_data, 0);
-	if (epp < 0)
-		return epp;
 
 	if (boot_cpu_has(X86_FEATURE_HWP_EPP)) {
 		if (epp == HWP_EPP_PERFORMANCE)
@@ -644,6 +641,8 @@ static int intel_pstate_get_energy_pref_index(struct cpudata *cpu_data, int *raw
 
 static int intel_pstate_set_epp(struct cpudata *cpu, u32 epp)
 {
+	int ret;
+
 	/*
 	 * Use the cached HWP Request MSR value, because in the active mode the
 	 * register itself may be updated by intel_pstate_hwp_boost_up() or
@@ -659,7 +658,11 @@ static int intel_pstate_set_epp(struct cpudata *cpu, u32 epp)
 	 * function, so it cannot run in parallel with the update below.
 	 */
 	WRITE_ONCE(cpu->hwp_req_cached, value);
-	return wrmsrl_on_cpu(cpu->cpu, MSR_HWP_REQUEST, value);
+	ret = wrmsrl_on_cpu(cpu->cpu, MSR_HWP_REQUEST, value);
+	if (!ret)
+		cpu->epp_cached = epp;
+
+	return ret;
 }
 
 static int intel_pstate_set_energy_pref_index(struct cpudata *cpu_data,
@@ -762,10 +765,8 @@ static ssize_t store_energy_performance_preference(
 			cpufreq_stop_governor(policy);
 			ret = intel_pstate_set_epp(cpu, epp);
 			err = cpufreq_start_governor(policy);
-			if (!ret) {
-				cpu->epp_cached = epp;
+			if (!ret)
 				ret = err;
-			}
 		}
 	}
 
@@ -2378,6 +2379,13 @@ static int intel_pstate_cpu_init(struct cpufreq_policy *policy)
 	 */
 	policy->policy = CPUFREQ_POLICY_POWERSAVE;
 
+
+	if (hwp_active) {
+		struct cpudata *cpu = all_cpu_data[policy->cpu];
+
+		cpu->epp_cached = intel_pstate_get_epp(cpu, 0);
+	}
+
 	return 0;
 }
 
@@ -2585,7 +2593,7 @@ static int intel_cpufreq_cpu_init(struct cpufreq_policy *policy)
 		policy->transition_delay_us = INTEL_CPUFREQ_TRANSITION_DELAY_HWP;
 		rdmsrl_on_cpu(cpu->cpu, MSR_HWP_REQUEST, &value);
 		WRITE_ONCE(cpu->hwp_req_cached, value);
-		cpu->epp_cached = (value & GENMASK_ULL(31, 24)) >> 24;
+		cpu->epp_cached = intel_pstate_get_epp(cpu, value);
 	} else {
 		turbo_max = cpu->pstate.turbo_pstate;
 		policy->transition_delay_us = INTEL_CPUFREQ_TRANSITION_DELAY;
Signature · 1 line
2.26.2
#3 lore raw

PATCH 3/4 cpufreq: intel_pstate: Add ->offline and ->online callbacks

In reply to Rafael J. Wysocki

From: "Rafael J. Wysocki" <[email protected]>

Add ->offline and ->online driver callbacksto to do the cleanup before taking a CPU offline and to restore its working configuration when it goes back online, respectively, to avoid invoking the ->init callback on every CPU online which is quite a bit of unnecessary overhead.

Define ->offline and ->online so that they can be used in the passive as well as in the active mode and because ->offline will do the majority of ->stop_cpu work, the passive mode does not need that callback any more, so drop it.

---
 drivers/cpufreq/intel_pstate.c | 38 ++++++++++++++++++++++++++++------
 1 file changed, 32 insertions(+), 6 deletions(-)
drivers/cpufreq/intel_pstate.c +32 -6
index 3d18934fa975..aca0587b176f 100644
@@ -2297,28 +2297,51 @@ static int intel_pstate_verify_policy(struct cpufreq_policy_data *policy)
 	return 0;
 }
 
-static void intel_cpufreq_stop_cpu(struct cpufreq_policy *policy)
+static int intel_pstate_cpu_offline(struct cpufreq_policy *policy)
 {
+	pr_debug("CPU %d going offline\n", policy->cpu);
+
+	intel_pstate_exit_perf_limits(policy);
+
+	/*
+	 * If the CPU is an SMT thread and it goes offline with the performance
+	 * settings different from the minimum, it will prevent its sibling
+	 * from getting to lower performance levels, so force the minimum
+	 * performance on CPU offline to prevent that form happening.
+	 */
 	if (hwp_active)
 		intel_pstate_hwp_force_min_perf(policy->cpu);
 	else
 		intel_pstate_set_min_pstate(all_cpu_data[policy->cpu]);
+
+	return 0;
+}
+
+static int intel_pstate_cpu_online(struct cpufreq_policy *policy)
+{
+	pr_debug("CPU %d going online\n", policy->cpu);
+
+	intel_pstate_init_acpi_perf_limits(policy);
+
+	if (hwp_active)
+		wrmsrl_on_cpu(policy->cpu, MSR_HWP_REQUEST,
+			      all_cpu_data[policy->cpu]->hwp_req_cached);
+
+	return 0;
 }
 
 static void intel_pstate_stop_cpu(struct cpufreq_policy *policy)
 {
-	pr_debug("CPU %d exiting\n", policy->cpu);
+	pr_debug("CPU %d stopping\n", policy->cpu);
 
 	intel_pstate_clear_update_util_hook(policy->cpu);
 	if (hwp_active)
 		intel_pstate_hwp_save_state(policy);
-
-	intel_cpufreq_stop_cpu(policy);
 }
 
 static int intel_pstate_cpu_exit(struct cpufreq_policy *policy)
 {
-	intel_pstate_exit_perf_limits(policy);
+	pr_debug("CPU %d exiting\n", policy->cpu);
 
 	policy->fast_switch_possible = false;
 
@@ -2398,6 +2421,8 @@ static struct cpufreq_driver intel_pstate = {
 	.init		= intel_pstate_cpu_init,
 	.exit		= intel_pstate_cpu_exit,
 	.stop_cpu	= intel_pstate_stop_cpu,
+	.offline	= intel_pstate_cpu_offline,
+	.online		= intel_pstate_cpu_online,
 	.update_limits	= intel_pstate_update_limits,
 	.name		= "intel_pstate",
 };
@@ -2652,7 +2677,8 @@ static struct cpufreq_driver intel_cpufreq = {
 	.fast_switch	= intel_cpufreq_fast_switch,
 	.init		= intel_cpufreq_cpu_init,
 	.exit		= intel_cpufreq_cpu_exit,
-	.stop_cpu	= intel_cpufreq_stop_cpu,
+	.offline	= intel_pstate_cpu_offline,
+	.online		= intel_pstate_cpu_online,
 	.update_limits	= intel_pstate_update_limits,
 	.name		= "intel_cpufreq",
 };
Signature · 1 line
2.26.2
#4 lore raw

RE: [PATCH 3/4] cpufreq: intel_pstate: Add ->offline and ->online callbacks

Hi Rafael,

Just annoying typo type feedback.

Quoted 4 lines · On 2020.08.20 09:38 Rafael J. Wysocki wrote:

From: "Rafael J. Wysocki" <[email protected]>

Add ->offline and ->online driver callbacksto to do the cleanup

"to to" and suggest this:

Add ->offline and ->online driver callbacks to cleanup

Quoted 7 lines

before taking a CPU offline and to restore its working configuration when it goes back online, respectively, to avoid invoking the ->init callback on every CPU online which is quite a bit of unnecessary overhead.

Define ->offline and ->online so that they can be used in the passive as well as in the active mode and because ->offline will

passive mode

Quoted 18 lines

do the majority of ->stop_cpu work, the passive mode does not need that callback any more, so drop it.

---
 drivers/cpufreq/intel_pstate.c | 38 ++++++++++++++++++++++++++++------
 1 file changed, 32 insertions(+), 6 deletions(-)
drivers/cpufreq/intel_pstate.c +1 -1
index 3d18934fa975..aca0587b176f 100644
@@ -2297,28 +2297,51 @@ static int intel_pstate_verify_policy(struct cpufreq_policy_data *policy)
 	return 0;
 }
 
-static void intel_cpufreq_stop_cpu(struct cpufreq_policy *policy)
+static int intel_pstate_cpu_offline(struct cpufreq_policy *policy)

drivers/cpufreq/intel_pstate.c static int intel_pstate_verify_policy(struct cpufreq_policy_data *policy) · line 2310

Quoted 10 lines
drivers/cpufreq/intel_pstate.c +9 -0
 {
+	pr_debug("CPU %d going offline\n", policy->cpu);
+
+	intel_pstate_exit_perf_limits(policy);
+
+	/*
+	 * If the CPU is an SMT thread and it goes offline with the performance
+	 * settings different from the minimum, it will prevent its sibling
+	 * from getting to lower performance levels, so force the minimum
+	 * performance on CPU offline to prevent that form happening.

form/from

Quoted 64 lines
drivers/cpufreq/intel_pstate.c +22 -6
+	 */
 	if (hwp_active)
 		intel_pstate_hwp_force_min_perf(policy->cpu);
 	else
 		intel_pstate_set_min_pstate(all_cpu_data[policy->cpu]);
+
+	return 0;
+}
+
+static int intel_pstate_cpu_online(struct cpufreq_policy *policy)
+{
+	pr_debug("CPU %d going online\n", policy->cpu);
+
+	intel_pstate_init_acpi_perf_limits(policy);
+
+	if (hwp_active)
+		wrmsrl_on_cpu(policy->cpu, MSR_HWP_REQUEST,
+			      all_cpu_data[policy->cpu]->hwp_req_cached);
+
+	return 0;
 }
 
 static void intel_pstate_stop_cpu(struct cpufreq_policy *policy)
 {
-	pr_debug("CPU %d exiting\n", policy->cpu);
+	pr_debug("CPU %d stopping\n", policy->cpu);
 
 	intel_pstate_clear_update_util_hook(policy->cpu);
 	if (hwp_active)
 		intel_pstate_hwp_save_state(policy);
-
-	intel_cpufreq_stop_cpu(policy);
 }
 
 static int intel_pstate_cpu_exit(struct cpufreq_policy *policy)
 {
-	intel_pstate_exit_perf_limits(policy);
+	pr_debug("CPU %d exiting\n", policy->cpu);
 
 	policy->fast_switch_possible = false;
 
@@ -2398,6 +2421,8 @@ static struct cpufreq_driver intel_pstate = {
 	.init		= intel_pstate_cpu_init,
 	.exit		= intel_pstate_cpu_exit,
 	.stop_cpu	= intel_pstate_stop_cpu,
+	.offline	= intel_pstate_cpu_offline,
+	.online		= intel_pstate_cpu_online,
 	.update_limits	= intel_pstate_update_limits,
 	.name		= "intel_pstate",
 };
@@ -2652,7 +2677,8 @@ static struct cpufreq_driver intel_cpufreq = {
 	.fast_switch	= intel_cpufreq_fast_switch,
 	.init		= intel_cpufreq_cpu_init,
 	.exit		= intel_cpufreq_cpu_exit,
-	.stop_cpu	= intel_cpufreq_stop_cpu,
+	.offline	= intel_pstate_cpu_offline,
+	.online		= intel_pstate_cpu_online,
 	.update_limits	= intel_pstate_update_limits,
 	.name		= "intel_cpufreq",
 };
--

2.26.2

#5 lore raw

PATCH v2 3/4 cpufreq: intel_pstate: Add ->offline and ->online callbacks

Quoted 3 lines · CEST Doug Smythies

Hi Rafael,

Just annoying typo type feedback.

Thanks!

Please find a corrected patch below.

Cheers!

---
From: Rafael J. Wysocki <[email protected]>
Subject: [PATCH] cpufreq: intel_pstate: Add ->offline and ->online callbacks

Add ->offline and ->online driver callbacks to prepare for taking a CPU offline and to restore its working configuration when it goes back online, respectively, to avoid invoking the ->init callback on every CPU online which is quite a bit of unnecessary overhead.

Define ->offline and ->online so that they can be used in the passive mode as well as in the active mode and because ->offline will do the majority of ->stop_cpu work, the passive mode does not need that callback any more, so drop it.

---
 drivers/cpufreq/intel_pstate.c |   38 ++++++++++++++++++++++++++++++++------
 1 file changed, 32 insertions(+), 6 deletions(-)
drivers/cpufreq/intel_pstate.c +32 -6
Index: linux-pm/drivers/cpufreq/intel_pstate.c
===================================================================
@@ -2297,28 +2297,51 @@ static int intel_pstate_verify_policy(st
 	return 0;
 }
 
-static void intel_cpufreq_stop_cpu(struct cpufreq_policy *policy)
+static int intel_pstate_cpu_offline(struct cpufreq_policy *policy)
 {
+	pr_debug("CPU %d going offline\n", policy->cpu);
+
+	intel_pstate_exit_perf_limits(policy);
+
+	/*
+	 * If the CPU is an SMT thread and it goes offline with the performance
+	 * settings different from the minimum, it will prevent its sibling
+	 * from getting to lower performance levels, so force the minimum
+	 * performance on CPU offline to prevent that from happening.
+	 */
 	if (hwp_active)
 		intel_pstate_hwp_force_min_perf(policy->cpu);
 	else
 		intel_pstate_set_min_pstate(all_cpu_data[policy->cpu]);
+
+	return 0;
+}
+
+static int intel_pstate_cpu_online(struct cpufreq_policy *policy)
+{
+	pr_debug("CPU %d going online\n", policy->cpu);
+
+	intel_pstate_init_acpi_perf_limits(policy);
+
+	if (hwp_active)
+		wrmsrl_on_cpu(policy->cpu, MSR_HWP_REQUEST,
+			      all_cpu_data[policy->cpu]->hwp_req_cached);
+
+	return 0;
 }
 
 static void intel_pstate_stop_cpu(struct cpufreq_policy *policy)
 {
-	pr_debug("CPU %d exiting\n", policy->cpu);
+	pr_debug("CPU %d stopping\n", policy->cpu);
 
 	intel_pstate_clear_update_util_hook(policy->cpu);
 	if (hwp_active)
 		intel_pstate_hwp_save_state(policy);
-
-	intel_cpufreq_stop_cpu(policy);
 }
 
 static int intel_pstate_cpu_exit(struct cpufreq_policy *policy)
 {
-	intel_pstate_exit_perf_limits(policy);
+	pr_debug("CPU %d exiting\n", policy->cpu);
 
 	policy->fast_switch_possible = false;
 
@@ -2398,6 +2421,8 @@ static struct cpufreq_driver intel_pstat
 	.init		= intel_pstate_cpu_init,
 	.exit		= intel_pstate_cpu_exit,
 	.stop_cpu	= intel_pstate_stop_cpu,
+	.offline	= intel_pstate_cpu_offline,
+	.online		= intel_pstate_cpu_online,
 	.update_limits	= intel_pstate_update_limits,
 	.name		= "intel_pstate",
 };
@@ -2652,7 +2677,8 @@ static struct cpufreq_driver intel_cpufr
 	.fast_switch	= intel_cpufreq_fast_switch,
 	.init		= intel_cpufreq_cpu_init,
 	.exit		= intel_cpufreq_cpu_exit,
-	.stop_cpu	= intel_cpufreq_stop_cpu,
+	.offline	= intel_pstate_cpu_offline,
+	.online		= intel_pstate_cpu_online,
 	.update_limits	= intel_pstate_update_limits,
 	.name		= "intel_cpufreq",
 };
#6 lore raw

PATCH 4/4 cpufreq: intel_pstate: Free memory only when turning off

In reply to Rafael J. Wysocki

From: "Rafael J. Wysocki" <[email protected]>

When intel_pstate switches the operation mode from "active" to "passive" or the other way around, freeing its data structures representing CPUs and allocating them again from scratch is not necessary and wasteful. Moreover, if these data structures are preserved, the cached HWP Request MSR value from there may be written to the MSR to start with to reinitialize it and help to restore the EPP value set previously (it is set to 0xFF when CPUs go offline to allow their SMT siblings to use the full range of EPP values and that also happens when the driver gets unregistered).

Accordingly, modify the driver to only do a full cleanup on driver object registration errors and when its status is changed to "off" via sysfs and to write the cached HWP Request MSR value back to the MSR on CPU init if the data structure representing the given CPU is still there.

---
 drivers/cpufreq/intel_pstate.c | 51 +++++++++++++---------------------
 1 file changed, 19 insertions(+), 32 deletions(-)
drivers/cpufreq/intel_pstate.c +19 -32
index aca0587b176f..a7c6491f2b4a 100644
@@ -2091,30 +2091,32 @@ static int intel_pstate_init_cpu(unsigned int cpunum)
 
 	cpu = all_cpu_data[cpunum];
 
-	if (!cpu) {
+	if (cpu) {
+		if (hwp_active)
+			wrmsrl_on_cpu(cpunum, MSR_HWP_REQUEST,
+				      cpu->hwp_req_cached);
+	} else {
 		cpu = kzalloc(sizeof(*cpu), GFP_KERNEL);
 		if (!cpu)
 			return -ENOMEM;
 
 		all_cpu_data[cpunum] = cpu;
 
+		cpu->cpu = cpunum;
+
 		cpu->epp_default = -EINVAL;
 		cpu->epp_powersave = -EINVAL;
 		cpu->epp_saved = -EINVAL;
-	}
-
-	cpu = all_cpu_data[cpunum];
 
-	cpu->cpu = cpunum;
+		if (hwp_active) {
+			const struct x86_cpu_id *id;
 
-	if (hwp_active) {
-		const struct x86_cpu_id *id;
+			intel_pstate_hwp_enable(cpu);
 
-		intel_pstate_hwp_enable(cpu);
-
-		id = x86_match_cpu(intel_pstate_hwp_boost_ids);
-		if (id && intel_pstate_acpi_pm_profile_server())
-			hwp_boost = true;
+			id = x86_match_cpu(intel_pstate_hwp_boost_ids);
+			if (id && intel_pstate_acpi_pm_profile_server())
+				hwp_boost = true;
+		}
 	}
 
 	intel_pstate_get_cpu_pstates(cpu);
@@ -2701,9 +2703,6 @@ static void intel_pstate_driver_cleanup(void)
 	}
 	put_online_cpus();
 
-	if (intel_pstate_driver == &intel_pstate)
-		intel_pstate_sysfs_hide_hwp_dynamic_boost();
-
 	intel_pstate_driver = NULL;
 }
 
@@ -2729,14 +2728,6 @@ static int intel_pstate_register_driver(struct cpufreq_driver *driver)
 	return 0;
 }
 
-static int intel_pstate_unregister_driver(void)
-{
-	cpufreq_unregister_driver(intel_pstate_driver);
-	intel_pstate_driver_cleanup();
-
-	return 0;
-}
-
 static ssize_t intel_pstate_show_status(char *buf)
 {
 	if (!intel_pstate_driver)
@@ -2748,8 +2739,6 @@ static ssize_t intel_pstate_show_status(char *buf)
 
 static int intel_pstate_update_status(const char *buf, size_t size)
 {
-	int ret;
-
 	if (size == 3 && !strncmp(buf, "off", size)) {
 		if (!intel_pstate_driver)
 			return -EINVAL;
@@ -2757,7 +2746,8 @@ static int intel_pstate_update_status(const char *buf, size_t size)
 		if (hwp_active)
 			return -EBUSY;
 
-		return intel_pstate_unregister_driver();
+		cpufreq_unregister_driver(intel_pstate_driver);
+		intel_pstate_driver_cleanup();
 	}
 
 	if (size == 6 && !strncmp(buf, "active", size)) {
@@ -2765,9 +2755,7 @@ static int intel_pstate_update_status(const char *buf, size_t size)
 			if (intel_pstate_driver == &intel_pstate)
 				return 0;
 
-			ret = intel_pstate_unregister_driver();
-			if (ret)
-				return ret;
+			cpufreq_unregister_driver(intel_pstate_driver);
 		}
 
 		return intel_pstate_register_driver(&intel_pstate);
@@ -2778,9 +2766,8 @@ static int intel_pstate_update_status(const char *buf, size_t size)
 			if (intel_pstate_driver == &intel_cpufreq)
 				return 0;
 
-			ret = intel_pstate_unregister_driver();
-			if (ret)
-				return ret;
+			cpufreq_unregister_driver(intel_pstate_driver);
+			intel_pstate_sysfs_hide_hwp_dynamic_boost();
 		}
 
 		return intel_pstate_register_driver(&intel_cpufreq);
Signature · 1 line
2.26.2
#7 lore raw
In reply to Rafael J. Wysocki

Hi Rafael,

Quoted 8 lines · On 2020.08.20 09:35 Rafael J. Wysocki wrote:

The purpose of this series is to address some peculiarities related to taking CPUs offline/online and switching between different operation modes with HWP enabled that have become visible after allowing the driver to work in the passive mode with HWP enabled in 5.9-rc1 (and one that was there earlier, but can be addressed easily after the changes madein 5.9-rc1).

Quoted 6 lines

Please refer to the patch changelogs for details.

For easier testing/review, the series is available from the git branch at:

git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git \
intel_pstate-testing

Thanks for that.

There still seems to be a problem with EPP getting messed up. I have not been able to find the exact spot in the code.

One problem is that EPP can end up as 0, and thereafter stays at 0. In sysfs terms, it ends up as "performance" and thereafter stays as "performance". Meanwhile I never modified it, and it started as "balance_performance".

It happens when changing from active to passive if the governor is performance. If the governor is not performance things work as expected.

Another problem is that EPP will end up as 128 when changing from passive to active. This erroneous condition is cleared by changing the governor to powersave and back to performance. It also doesn't occur the first time after boot, when booting to intel_cpufreq/performance/HWP. (confused yet?) The sysfs value is O.K. during this.

Supporting data:
Example 1:

Grub:
GRUB_CMDLINE_LINUX_DEFAULT="ipv6.disable=1 consoleblank=450 msr.allow_writes=on cpuidle.governor=teo"

So I boot to intel_pstate/performance/HWP:

# /home/doug/c/msr-decoder (always edited for only relevant parts)
6.) 0x774: IA32_HWP_REQUEST:    CPU 0-5 :
    raw: 00002E2E : 00002E2E : 00002E2E : 00002E2E : 00002E2E : 00002E2E :
    epp:        0 :        0 :        0 :        0 :        0 :        0 :
# cat /sys/devices/system/cpu/cpu3/cpufreq/scaling_driver
intel_pstate
# cat /sys/devices/system/cpu/cpu3/cpufreq/scaling_governor
performance
# cat  /sys/devices/system/cpu/cpufreq/policy3/energy_performance_preference
balance_performance
# echo passive > /sys/devices/system/cpu/intel_pstate/status

Note: the following results are incorrect:

# cat  /sys/devices/system/cpu/cpufreq/policy3/energy_performance_preference
performance
# echo "ondemand" | tee /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
ondemand
# cat  /sys/devices/system/cpu/cpufreq/policy3/energy_performance_preference
performance
# /home/doug/c/msr-decoder
6.) 0x774: IA32_HWP_REQUEST:    CPU 0-5 :
    raw: 00002E08 : 00002E08 : 00002E08 : 00002E0B : 00002E13 : 00002E08 :
    epp:        0 :        0 :        0 :        0 :        0 :        0 :
# echo active > /sys/devices/system/cpu/intel_pstate/status
# cat /sys/devices/system/cpu/cpu3/cpufreq/scaling_governor
performance
# cat  /sys/devices/system/cpu/cpufreq/policy3/energy_performance_preference
performance
# echo "powersave" | tee /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
powersave
# cat  /sys/devices/system/cpu/cpufreq/policy3/energy_performance_preference
performance
# cat /sys/devices/system/cpu/cpu3/cpufreq/scaling_governor
powersave
# cat /sys/devices/system/cpu/cpu3/cpufreq/scaling_driver
intel_pstate

Example 2:
Grub:
GRUB_CMDLINE_LINUX_DEFAULT="ipv6.disable=1 consoleblank=450 intel_pstate=passive msr.allow_writes=on cpuidle.governor=teo"

So I boot to intel_cpufreq/performance/HWP:

# /home/doug/c/msr-decoder
6.) 0x774: IA32_HWP_REQUEST:    CPU 0-5 :
    raw: 80002E2E : 80002E2E : 80002E2E : 80002E2E : 80002E2E : 80002E2E :
    epp:      128 :      128 :      128 :      128 :      128 :      128 :
# cat /sys/devices/system/cpu/cpu3/cpufreq/scaling_driver
intel_cpufreq
# cat /sys/devices/system/cpu/cpu3/cpufreq/scaling_governor
performance
# cat /sys/devices/system/cpu/cpufreq/policy3/energy_performance_preference
balance_performance
# echo active > /sys/devices/system/cpu/intel_pstate/status
# cat /sys/devices/system/cpu/cpufreq/policy3/energy_performance_preference
balance_performance
# cat /sys/devices/system/cpu/cpu3/cpufreq/scaling_governor
performance
# cat /sys/devices/system/cpu/cpu3/cpufreq/scaling_driver
intel_pstate
# /home/doug/c/msr-decoder
6.) 0x774: IA32_HWP_REQUEST:    CPU 0-5 :
    raw: 00002E2E : 00002E2E : 00002E2E : 00002E2E : 00002E2E : 00002E2E :
    epp:        0 :        0 :        0 :        0 :        0 :        0 :
# echo passive > /sys/devices/system/cpu/intel_pstate/status
# /home/doug/c/msr-decoder
6.) 0x774: IA32_HWP_REQUEST:    CPU 0-5 :
    raw: 80002E2E : 80002E2E : 80002E2E : 80002E2E : 80002E2E : 80002E2E :
    epp:      128 :      128 :      128 :      128 :      128 :      128 :
# echo active > /sys/devices/system/cpu/intel_pstate/status

Note: the following results are incorrect:

# /home/doug/c/msr-decoder
6.) 0x774: IA32_HWP_REQUEST:    CPU 0-5 :
    raw: 80002E2E : 80002E2E : 80002E2E : 80002E2E : 80002E2E : 80002E2E :
    epp:      128 :      128 :      128 :      128 :      128 :      128 :
# cat /sys/devices/system/cpu/cpu3/cpufreq/scaling_governor
performance
# cat /sys/devices/system/cpu/cpu3/cpufreq/scaling_driver
intel_pstate
# cat /sys/devices/system/cpu/cpufreq/policy3/energy_performance_preference
balance_performance

Note: But the problem can be cleared:

# echo "powersave" | tee /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
powersave
# echo "performance" | tee /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
performance
# /home/doug/c/msr-decoder
6.) 0x774: IA32_HWP_REQUEST:    CPU 0-5 :
    raw: 00002E2E : 00002E2E : 00002E2E : 00002E2E : 00002E2E : 00002E2E :
    epp:        0 :        0 :        0 :        0 :        0 :        0 :

... Doug

#8 lore raw
Quoted 23 lines · CEST Doug Smythies

Hi Rafael,

Quoted 8 lines · On 2020.08.20 09:35 Rafael J. Wysocki wrote:

The purpose of this series is to address some peculiarities related to taking CPUs offline/online and switching between different operation modes with HWP enabled that have become visible after allowing the driver to work in the passive mode with HWP enabled in 5.9-rc1 (and one that was there earlier, but can be addressed easily after the changes madein 5.9-rc1).

Quoted 6 lines

Please refer to the patch changelogs for details.

For easier testing/review, the series is available from the git branch at:

git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git \
intel_pstate-testing

Thanks for that.

There still seems to be a problem with EPP getting messed up. I have not been able to find the exact spot in the code.

Quoted 7 lines

One problem is that EPP can end up as 0, and thereafter stays at 0. In sysfs terms, it ends up as "performance" and thereafter stays as "performance". Meanwhile I never modified it, and it started as "balance_performance".

It happens when changing from active to passive if the governor is performance. If the governor is not performance things work as expected.

One change is missing from the patches in the $subject series and IMO it doesn't belong to any of them, so please find it appended below (on top of the $subject series).

With it applied, this particular issue should go away.

Quoted 2 lines

Another problem is that EPP will end up as 128 when changing from passive to active.

I don't seem to be able to reproduce this (at least not without involving system-wide suspend/resume).

Cheers!

---
From: Rafael J. Wysocki <[email protected]>
Subject: [PATCH] cpufreq: intel_pstate: Restore cached EPP value during offline

Because hwp_req_cached contains the effective EPP value (0) when the "performance" scaling algorithm is used in the active mode, replace it with the cached EPP value during CPU offline to prevent it from being used (unexpectedly) after switching over from the active mode to the passive mode.

Also rename intel_pstate_hwp_force_min_perf() because it will do more than just forcing the minimum performance now.

---
 drivers/cpufreq/intel_pstate.c |   19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)
drivers/cpufreq/intel_pstate.c +15 -4
Index: linux-pm/drivers/cpufreq/intel_pstate.c
===================================================================
@@ -904,12 +904,23 @@ skip_epp:
 	wrmsrl_on_cpu(cpu, MSR_HWP_REQUEST, value);
 }
 
-static void intel_pstate_hwp_force_min_perf(int cpu)
+static void intel_pstate_hwp_offline(int cpu)
 {
-	u64 value;
+	struct cpudata *cpudata = all_cpu_data[cpu];
+	u64 value = READ_ONCE(cpudata->hwp_req_cached);
 	int min_perf;
 
-	value = all_cpu_data[cpu]->hwp_req_cached;
+	if (boot_cpu_has(X86_FEATURE_HWP_EPP)) {
+		/*
+		 * In case the EPP has been set to "performance" by the
+		 * active mode "performance" scaling algorithm, replace that
+		 * temporary value with the cached EPP one.
+		 */
+		value &= ~GENMASK_ULL(31, 24);
+		value |= HWP_ENERGY_PERF_PREFERENCE(cpudata->epp_cached);
+		WRITE_ONCE(cpudata->hwp_req_cached, value);
+	}
+
 	value &= ~GENMASK_ULL(31, 0);
 	min_perf = HWP_LOWEST_PERF(all_cpu_data[cpu]->hwp_cap_cached);
 
@@ -2312,7 +2323,7 @@ static int intel_pstate_cpu_offline(stru
 	 * performance on CPU offline to prevent that from happening.
 	 */
 	if (hwp_active)
-		intel_pstate_hwp_force_min_perf(policy->cpu);
+		intel_pstate_hwp_offline(policy->cpu);
 	else
 		intel_pstate_set_min_pstate(all_cpu_data[policy->cpu]);