Merge tag 'pm-5.2-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm

Pull more power management updates from Rafael Wysocki:
 "These fix a recent regression causing kernels built with CONFIG_PM
  unset to crash on systems that support the Performance and Energy Bias
  Hint (EPB), clean up the cpufreq core and some users of transition
  notifiers and introduce a new power domain flag into the generic power
  domains framework (genpd).

  Specifics:

   - Fix recent regression causing kernels built with CONFIG_PM unset to
     crash on systems that support the Performance and Energy Bias Hint
     (EPB) by avoiding to compile the EPB-related code depending on
     CONFIG_PM when it is unset (Rafael Wysocki).

   - Clean up the transition notifier invocation code in the cpufreq
     core and change some users of cpufreq transition notifiers
     accordingly (Viresh Kumar).

   - Change MAINTAINERS to cover the schedutil governor as part of
     cpufreq (Viresh Kumar).

   - Simplify cpufreq_init_policy() to avoid redundant computations (Yue
     Hu).

   - Add explanatory comment to the cpufreq core (Rafael Wysocki).

   - Introduce a new flag, GENPD_FLAG_RPM_ALWAYS_ON, to the generic
     power domains (genpd) framework along with the first user of it
     (Leonard Crestez)"

* tag 'pm-5.2-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  soc: imx: gpc: Use GENPD_FLAG_RPM_ALWAYS_ON for ERR009619
  PM / Domains: Add GENPD_FLAG_RPM_ALWAYS_ON flag
  cpufreq: Update MAINTAINERS to include schedutil governor
  cpufreq: Don't find governor for setpolicy drivers in cpufreq_init_policy()
  cpufreq: Explain the kobject_put() in cpufreq_policy_alloc()
  cpufreq: Call transition notifier only once for each policy
  x86: intel_epb: Take CONFIG_PM into account

3 files changed, 97 insertions, 64 deletions

diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c
index 7a6aa2318915..33c30c1e6a30 100644
--- a/drivers/base/power/domain.c
+++ b/drivers/base/power/domain.c
@@ -128,6 +128,7 @@ static const struct genpd_lock_ops genpd_spin_ops = {
 #define genpd_is_always_on(genpd)	(genpd->flags & GENPD_FLAG_ALWAYS_ON)
 #define genpd_is_active_wakeup(genpd)	(genpd->flags & GENPD_FLAG_ACTIVE_WAKEUP)
 #define genpd_is_cpu_domain(genpd)	(genpd->flags & GENPD_FLAG_CPU_DOMAIN)
+#define genpd_is_rpm_always_on(genpd)	(genpd->flags & GENPD_FLAG_RPM_ALWAYS_ON)
 
 static inline bool irq_safe_dev_in_no_sleep_domain(struct device *dev,
 		const struct generic_pm_domain *genpd)
@@ -515,7 +516,9 @@ static int genpd_power_off(struct generic_pm_domain *genpd, bool one_dev_on,
 	 * (1) The domain is configured as always on.
 	 * (2) When the domain has a subdomain being powered on.
 	 */
-	if (genpd_is_always_on(genpd) || atomic_read(&genpd->sd_count) > 0)
+	if (genpd_is_always_on(genpd) ||
+			genpd_is_rpm_always_on(genpd) ||
+			atomic_read(&genpd->sd_count) > 0)
 		return -EBUSY;
 
 	list_for_each_entry(pdd, &genpd->dev_list, list_node) {
@@ -1812,7 +1815,8 @@ int pm_genpd_init(struct generic_pm_domain *genpd,
 	}
 
 	/* Always-on domains must be powered on at initialization. */
-	if (genpd_is_always_on(genpd) && !genpd_status_on(genpd))
+	if ((genpd_is_always_on(genpd) || genpd_is_rpm_always_on(genpd)) &&
+			!genpd_status_on(genpd))
 		return -EINVAL;
 
 	if (genpd_is_cpu_domain(genpd) &&
diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
index db779b650fce..85ff958e01f1 100644
--- a/drivers/cpufreq/cpufreq.c
+++ b/drivers/cpufreq/cpufreq.c
@@ -340,11 +340,14 @@ static void cpufreq_notify_transition(struct cpufreq_policy *policy,
 				      struct cpufreq_freqs *freqs,
 				      unsigned int state)
 {
+	int cpu;
+
 	BUG_ON(irqs_disabled());
 
 	if (cpufreq_disabled())
 		return;
 
+	freqs->policy = policy;
 	freqs->flags = cpufreq_driver->flags;
 	pr_debug("notification %u of frequency transition to %u kHz\n",
 		 state, freqs->new);
@@ -364,10 +367,8 @@ static void cpufreq_notify_transition(struct cpufreq_policy *policy,
 			}
 		}
 
-		for_each_cpu(freqs->cpu, policy->cpus) {
-			srcu_notifier_call_chain(&cpufreq_transition_notifier_list,
-						 CPUFREQ_PRECHANGE, freqs);
-		}
+		srcu_notifier_call_chain(&cpufreq_transition_notifier_list,
+					 CPUFREQ_PRECHANGE, freqs);
 
 		adjust_jiffies(CPUFREQ_PRECHANGE, freqs);
 		break;
@@ -377,11 +378,11 @@ static void cpufreq_notify_transition(struct cpufreq_policy *policy,
 		pr_debug("FREQ: %u - CPUs: %*pbl\n", freqs->new,
 			 cpumask_pr_args(policy->cpus));
 
-		for_each_cpu(freqs->cpu, policy->cpus) {
-			trace_cpu_frequency(freqs->new, freqs->cpu);
-			srcu_notifier_call_chain(&cpufreq_transition_notifier_list,
-						 CPUFREQ_POSTCHANGE, freqs);
-		}
+		for_each_cpu(cpu, policy->cpus)
+			trace_cpu_frequency(freqs->new, cpu);
+
+		srcu_notifier_call_chain(&cpufreq_transition_notifier_list,
+					 CPUFREQ_POSTCHANGE, freqs);
 
 		cpufreq_stats_record_transition(policy, freqs->new);
 		policy->cur = freqs->new;
@@ -618,50 +619,52 @@ static struct cpufreq_governor *find_governor(const char *str_governor)
 	return NULL;
 }
 
+static int cpufreq_parse_policy(char *str_governor,
+				struct cpufreq_policy *policy)
+{
+	if (!strncasecmp(str_governor, "performance", CPUFREQ_NAME_LEN)) {
+		policy->policy = CPUFREQ_POLICY_PERFORMANCE;
+		return 0;
+	}
+	if (!strncasecmp(str_governor, "powersave", CPUFREQ_NAME_LEN)) {
+		policy->policy = CPUFREQ_POLICY_POWERSAVE;
+		return 0;
+	}
+	return -EINVAL;
+}
+
 /**
- * cpufreq_parse_governor - parse a governor string
+ * cpufreq_parse_governor - parse a governor string only for !setpolicy
  */
 static int cpufreq_parse_governor(char *str_governor,
 				  struct cpufreq_policy *policy)
 {
-	if (cpufreq_driver->setpolicy) {
-		if (!strncasecmp(str_governor, "performance", CPUFREQ_NAME_LEN)) {
-			policy->policy = CPUFREQ_POLICY_PERFORMANCE;
-			return 0;
-		}
+	struct cpufreq_governor *t;
 
-		if (!strncasecmp(str_governor, "powersave", CPUFREQ_NAME_LEN)) {
-			policy->policy = CPUFREQ_POLICY_POWERSAVE;
-			return 0;
-		}
-	} else {
-		struct cpufreq_governor *t;
+	mutex_lock(&cpufreq_governor_mutex);
 
-		mutex_lock(&cpufreq_governor_mutex);
+	t = find_governor(str_governor);
+	if (!t) {
+		int ret;
 
-		t = find_governor(str_governor);
-		if (!t) {
-			int ret;
-
-			mutex_unlock(&cpufreq_governor_mutex);
+		mutex_unlock(&cpufreq_governor_mutex);
 
-			ret = request_module("cpufreq_%s", str_governor);
-			if (ret)
-				return -EINVAL;
+		ret = request_module("cpufreq_%s", str_governor);
+		if (ret)
+			return -EINVAL;
 
-			mutex_lock(&cpufreq_governor_mutex);
+		mutex_lock(&cpufreq_governor_mutex);
 
-			t = find_governor(str_governor);
-		}
-		if (t && !try_module_get(t->owner))
-			t = NULL;
+		t = find_governor(str_governor);
+	}
+	if (t && !try_module_get(t->owner))
+		t = NULL;
 
-		mutex_unlock(&cpufreq_governor_mutex);
+	mutex_unlock(&cpufreq_governor_mutex);
 
-		if (t) {
-			policy->governor = t;
-			return 0;
-		}
+	if (t) {
+		policy->governor = t;
+		return 0;
 	}
 
 	return -EINVAL;
@@ -783,8 +786,13 @@ static ssize_t store_scaling_governor(struct cpufreq_policy *policy,
 	if (ret != 1)
 		return -EINVAL;
 
-	if (cpufreq_parse_governor(str_governor, &new_policy))
-		return -EINVAL;
+	if (cpufreq_driver->setpolicy) {
+		if (cpufreq_parse_policy(str_governor, &new_policy))
+			return -EINVAL;
+	} else {
+		if (cpufreq_parse_governor(str_governor, &new_policy))
+			return -EINVAL;
+	}
 
 	ret = cpufreq_set_policy(policy, &new_policy);
 
@@ -1050,32 +1058,39 @@ __weak struct cpufreq_governor *cpufreq_default_governor(void)
 
 static int cpufreq_init_policy(struct cpufreq_policy *policy)
 {
-	struct cpufreq_governor *gov = NULL;
+	struct cpufreq_governor *gov = NULL, *def_gov = NULL;
 	struct cpufreq_policy new_policy;
 
 	memcpy(&new_policy, policy, sizeof(*policy));
 
-	/* Update governor of new_policy to the governor used before hotplug */
-	gov = find_governor(policy->last_governor);
-	if (gov) {
-		pr_debug("Restoring governor %s for cpu %d\n",
+	def_gov = cpufreq_default_governor();
+
+	if (has_target()) {
+		/*
+		 * Update governor of new_policy to the governor used before
+		 * hotplug
+		 */
+		gov = find_governor(policy->last_governor);
+		if (gov) {
+			pr_debug("Restoring governor %s for cpu %d\n",
 				policy->governor->name, policy->cpu);
+		} else {
+			if (!def_gov)
+				return -ENODATA;
+			gov = def_gov;
+		}
+		new_policy.governor = gov;
 	} else {
-		gov = cpufreq_default_governor();
-		if (!gov)
-			return -ENODATA;
-	}
-
-	new_policy.governor = gov;
-
-	/* Use the default policy if there is no last_policy. */
-	if (cpufreq_driver->setpolicy) {
-		if (policy->last_policy)
+		/* Use the default policy if there is no last_policy. */
+		if (policy->last_policy) {
 			new_policy.policy = policy->last_policy;
-		else
-			cpufreq_parse_governor(gov->name, &new_policy);
+		} else {
+			if (!def_gov)
+				return -ENODATA;
+			cpufreq_parse_policy(def_gov->name, &new_policy);
+		}
 	}
-	/* set default policy */
+
 	return cpufreq_set_policy(policy, &new_policy);
 }
 
@@ -1133,6 +1148,11 @@ static struct cpufreq_policy *cpufreq_policy_alloc(unsigned int cpu)
 				   cpufreq_global_kobject, "policy%u", cpu);
 	if (ret) {
 		pr_err("%s: failed to init policy->kobj: %d\n", __func__, ret);
+		/*
+		 * The entire policy object will be freed below, but the extra
+		 * memory allocated for the kobject name needs to be freed by
+		 * releasing the kobject.
+		 */
 		kobject_put(&policy->kobj);
 		goto err_free_real_cpus;
 	}
diff --git a/drivers/soc/imx/gpc.c b/drivers/soc/imx/gpc.c
index 7d14a4b4e82a..29b43651c261 100644
--- a/drivers/soc/imx/gpc.c
+++ b/drivers/soc/imx/gpc.c
@@ -431,10 +431,19 @@ static int imx_gpc_probe(struct platform_device *pdev)
 		return ret;
 	}
 
-	/* Disable PU power down in normal operation if ERR009619 is present */
+	/*
+	 * Disable PU power down by runtime PM if ERR009619 is present.
+	 *
+	 * The PRE clock will be paused for several cycles when turning on the
+	 * PU domain LDO from power down state. If PRE is in use at that time,
+	 * the IPU/PRG cannot get the correct display data from the PRE.
+	 *
+	 * This is not a concern when the whole system enters suspend state, so
+	 * it's safe to power down PU in this case.
+	 */
 	if (of_id_data->err009619_present)
 		imx_gpc_domains[GPC_PGC_DOMAIN_PU].base.flags |=
-				GENPD_FLAG_ALWAYS_ON;
+				GENPD_FLAG_RPM_ALWAYS_ON;
 
 	/* Keep DISP always on if ERR006287 is present */
 	if (of_id_data->err006287_present)