7 files changed, 162 insertions, 123 deletions

diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c
index e82d0976a5d0..2bfd53cbfe80 100644
--- a/drivers/acpi/apei/ghes.c
+++ b/drivers/acpi/apei/ghes.c
@@ -729,10 +729,10 @@ static struct llist_head ghes_estatus_llist;
 static struct irq_work ghes_proc_irq_work;
 
 /*
- * NMI may be triggered on any CPU, so ghes_nmi_lock is used for
- * mutual exclusion.
+ * NMI may be triggered on any CPU, so ghes_in_nmi is used for
+ * having only one concurrent reader.
  */
-static DEFINE_RAW_SPINLOCK(ghes_nmi_lock);
+static atomic_t ghes_in_nmi = ATOMIC_INIT(0);
 
 static LIST_HEAD(ghes_nmi);
 
@@ -797,73 +797,75 @@ static void ghes_print_queued_estatus(void)
 	}
 }
 
+/* Save estatus for further processing in IRQ context */
+static void __process_error(struct ghes *ghes)
+{
+#ifdef CONFIG_ARCH_HAVE_NMI_SAFE_CMPXCHG
+	u32 len, node_len;
+	struct ghes_estatus_node *estatus_node;
+	struct acpi_hest_generic_status *estatus;
+
+	if (ghes_estatus_cached(ghes->estatus))
+		return;
+
+	len = cper_estatus_len(ghes->estatus);
+	node_len = GHES_ESTATUS_NODE_LEN(len);
+
+	estatus_node = (void *)gen_pool_alloc(ghes_estatus_pool, node_len);
+	if (!estatus_node)
+		return;
+
+	estatus_node->ghes = ghes;
+	estatus_node->generic = ghes->generic;
+	estatus = GHES_ESTATUS_FROM_NODE(estatus_node);
+	memcpy(estatus, ghes->estatus, len);
+	llist_add(&estatus_node->llnode, &ghes_estatus_llist);
+#endif
+}
+
+static void __ghes_panic(struct ghes *ghes)
+{
+	oops_begin();
+	ghes_print_queued_estatus();
+	__ghes_print_estatus(KERN_EMERG, ghes->generic, ghes->estatus);
+
+	/* reboot to log the error! */
+	if (panic_timeout == 0)
+		panic_timeout = ghes_panic_timeout;
+	panic("Fatal hardware error!");
+}
+
 static int ghes_notify_nmi(unsigned int cmd, struct pt_regs *regs)
 {
-	struct ghes *ghes, *ghes_global = NULL;
-	int sev, sev_global = -1;
-	int ret = NMI_DONE;
+	struct ghes *ghes;
+	int sev, ret = NMI_DONE;
+
+	if (!atomic_add_unless(&ghes_in_nmi, 1, 1))
+		return ret;
 
-	raw_spin_lock(&ghes_nmi_lock);
 	list_for_each_entry_rcu(ghes, &ghes_nmi, list) {
 		if (ghes_read_estatus(ghes, 1)) {
 			ghes_clear_estatus(ghes);
 			continue;
 		}
-		sev = ghes_severity(ghes->estatus->error_severity);
-		if (sev > sev_global) {
-			sev_global = sev;
-			ghes_global = ghes;
-		}
-		ret = NMI_HANDLED;
-	}
-
-	if (ret == NMI_DONE)
-		goto out;
 
-	if (sev_global >= GHES_SEV_PANIC) {
-		oops_begin();
-		ghes_print_queued_estatus();
-		__ghes_print_estatus(KERN_EMERG, ghes_global->generic,
-				     ghes_global->estatus);
-		/* reboot to log the error! */
-		if (panic_timeout == 0)
-			panic_timeout = ghes_panic_timeout;
-		panic("Fatal hardware error!");
-	}
+		sev = ghes_severity(ghes->estatus->error_severity);
+		if (sev >= GHES_SEV_PANIC)
+			__ghes_panic(ghes);
 
-	list_for_each_entry_rcu(ghes, &ghes_nmi, list) {
-#ifdef CONFIG_ARCH_HAVE_NMI_SAFE_CMPXCHG
-		u32 len, node_len;
-		struct ghes_estatus_node *estatus_node;
-		struct acpi_hest_generic_status *estatus;
-#endif
 		if (!(ghes->flags & GHES_TO_CLEAR))
 			continue;
-#ifdef CONFIG_ARCH_HAVE_NMI_SAFE_CMPXCHG
-		if (ghes_estatus_cached(ghes->estatus))
-			goto next;
-		/* Save estatus for further processing in IRQ context */
-		len = cper_estatus_len(ghes->estatus);
-		node_len = GHES_ESTATUS_NODE_LEN(len);
-		estatus_node = (void *)gen_pool_alloc(ghes_estatus_pool,
-						      node_len);
-		if (estatus_node) {
-			estatus_node->ghes = ghes;
-			estatus_node->generic = ghes->generic;
-			estatus = GHES_ESTATUS_FROM_NODE(estatus_node);
-			memcpy(estatus, ghes->estatus, len);
-			llist_add(&estatus_node->llnode, &ghes_estatus_llist);
-		}
-next:
-#endif
+
+		__process_error(ghes);
 		ghes_clear_estatus(ghes);
+
+		ret = NMI_HANDLED;
 	}
+
 #ifdef CONFIG_ARCH_HAVE_NMI_SAFE_CMPXCHG
 	irq_work_queue(&ghes_proc_irq_work);
 #endif
-
-out:
-	raw_spin_unlock(&ghes_nmi_lock);
+	atomic_dec(&ghes_in_nmi);
 	return ret;
 }
 
diff --git a/drivers/acpi/device_pm.c b/drivers/acpi/device_pm.c
index 735db11a9b00..717afcdb5f4a 100644
--- a/drivers/acpi/device_pm.c
+++ b/drivers/acpi/device_pm.c
@@ -98,17 +98,16 @@ int acpi_device_get_power(struct acpi_device *device, int *state)
 
 		/*
 		 * The power resources settings may indicate a power state
-		 * shallower than the actual power state of the device.
+		 * shallower than the actual power state of the device, because
+		 * the same power resources may be referenced by other devices.
 		 *
-		 * Moreover, on systems predating ACPI 4.0, if the device
-		 * doesn't depend on any power resources and _PSC returns 3,
-		 * that means "power off".  We need to maintain compatibility
-		 * with those systems.
+		 * For systems predating ACPI 4.0 we assume that D3hot is the
+		 * deepest state that can be supported.
 		 */
 		if (psc > result && psc < ACPI_STATE_D3_COLD)
 			result = psc;
 		else if (result == ACPI_STATE_UNKNOWN)
-			result = psc > ACPI_STATE_D2 ? ACPI_STATE_D3_COLD : psc;
+			result = psc > ACPI_STATE_D2 ? ACPI_STATE_D3_HOT : psc;
 	}
 
 	/*
@@ -153,8 +152,8 @@ static int acpi_dev_pm_explicit_set(struct acpi_device *adev, int state)
  */
 int acpi_device_set_power(struct acpi_device *device, int state)
 {
+	int target_state = state;
 	int result = 0;
-	bool cut_power = false;
 
 	if (!device || !device->flags.power_manageable
 	    || (state < ACPI_STATE_D0) || (state > ACPI_STATE_D3_COLD))
@@ -169,11 +168,21 @@ int acpi_device_set_power(struct acpi_device *device, int state)
 		return 0;
 	}
 
-	if (!device->power.states[state].flags.valid) {
+	if (state == ACPI_STATE_D3_COLD) {
+		/*
+		 * For transitions to D3cold we need to execute _PS3 and then
+		 * possibly drop references to the power resources in use.
+		 */
+		state = ACPI_STATE_D3_HOT;
+		/* If _PR3 is not available, use D3hot as the target state. */
+		if (!device->power.states[ACPI_STATE_D3_COLD].flags.valid)
+			target_state = state;
+	} else if (!device->power.states[state].flags.valid) {
 		dev_warn(&device->dev, "Power state %s not supported\n",
 			 acpi_power_state_string(state));
 		return -ENODEV;
 	}
+
 	if (!device->power.flags.ignore_parent &&
 	    device->parent && (state < device->parent->power.state)) {
 		dev_warn(&device->dev,
@@ -183,39 +192,38 @@ int acpi_device_set_power(struct acpi_device *device, int state)
 		return -ENODEV;
 	}
 
-	/* For D3cold we should first transition into D3hot. */
-	if (state == ACPI_STATE_D3_COLD
-	    && device->power.states[ACPI_STATE_D3_COLD].flags.os_accessible) {
-		state = ACPI_STATE_D3_HOT;
-		cut_power = true;
-	}
-
-	if (state < device->power.state && state != ACPI_STATE_D0
-	    && device->power.state >= ACPI_STATE_D3_HOT) {
-		dev_warn(&device->dev,
-			 "Cannot transition to non-D0 state from D3\n");
-		return -ENODEV;
-	}
-
 	/*
 	 * Transition Power
 	 * ----------------
-	 * In accordance with the ACPI specification first apply power (via
-	 * power resources) and then evaluate _PSx.
+	 * In accordance with ACPI 6, _PSx is executed before manipulating power
+	 * resources, unless the target state is D0, in which case _PS0 is
+	 * supposed to be executed after turning the power resources on.
 	 */
-	if (device->power.flags.power_resources) {
-		result = acpi_power_transition(device, state);
+	if (state > ACPI_STATE_D0) {
+		/*
+		 * According to ACPI 6, devices cannot go from lower-power
+		 * (deeper) states to higher-power (shallower) states.
+		 */
+		if (state < device->power.state) {
+			dev_warn(&device->dev, "Cannot transition from %s to %s\n",
+				 acpi_power_state_string(device->power.state),
+				 acpi_power_state_string(state));
+			return -ENODEV;
+		}
+
+		result = acpi_dev_pm_explicit_set(device, state);
 		if (result)
 			goto end;
-	}
-	result = acpi_dev_pm_explicit_set(device, state);
-	if (result)
-		goto end;
 
-	if (cut_power) {
-		device->power.state = state;
-		state = ACPI_STATE_D3_COLD;
-		result = acpi_power_transition(device, state);
+		if (device->power.flags.power_resources)
+			result = acpi_power_transition(device, target_state);
+	} else {
+		if (device->power.flags.power_resources) {
+			result = acpi_power_transition(device, ACPI_STATE_D0);
+			if (result)
+				goto end;
+		}
+		result = acpi_dev_pm_explicit_set(device, ACPI_STATE_D0);
 	}
 
  end:
@@ -264,13 +272,24 @@ int acpi_bus_init_power(struct acpi_device *device)
 		return result;
 
 	if (state < ACPI_STATE_D3_COLD && device->power.flags.power_resources) {
+		/* Reference count the power resources. */
 		result = acpi_power_on_resources(device, state);
 		if (result)
 			return result;
 
-		result = acpi_dev_pm_explicit_set(device, state);
-		if (result)
-			return result;
+		if (state == ACPI_STATE_D0) {
+			/*
+			 * If _PSC is not present and the state inferred from
+			 * power resources appears to be D0, it still may be
+			 * necessary to execute _PS0 at this point, because
+			 * another device using the same power resources may
+			 * have been put into D0 previously and that's why we
+			 * see D0 here.
+			 */
+			result = acpi_dev_pm_explicit_set(device, state);
+			if (result)
+				return result;
+		}
 	} else if (state == ACPI_STATE_UNKNOWN) {
 		/*
 		 * No power resources and missing _PSC?  Cross fingers and make
@@ -603,12 +622,12 @@ int acpi_pm_device_sleep_state(struct device *dev, int *d_min_p, int d_max_in)
 	if (d_max_in < ACPI_STATE_D0 || d_max_in > ACPI_STATE_D3_COLD)
 		return -EINVAL;
 
-	if (d_max_in > ACPI_STATE_D3_HOT) {
+	if (d_max_in > ACPI_STATE_D2) {
 		enum pm_qos_flags_status stat;
 
 		stat = dev_pm_qos_flags(dev, PM_QOS_FLAG_NO_POWER_OFF);
 		if (stat == PM_QOS_FLAGS_ALL)
-			d_max_in = ACPI_STATE_D3_HOT;
+			d_max_in = ACPI_STATE_D2;
 	}
 
 	adev = ACPI_COMPANION(dev);
@@ -953,6 +972,7 @@ EXPORT_SYMBOL_GPL(acpi_subsys_prepare);
  */
 void acpi_subsys_complete(struct device *dev)
 {
+	pm_generic_complete(dev);
 	/*
 	 * If the device had been runtime-suspended before the system went into
 	 * the sleep state it is going out of and it has never been resumed till
diff --git a/drivers/acpi/fan.c b/drivers/acpi/fan.c
index 7a36f02598a6..bea0bbaafa97 100644
--- a/drivers/acpi/fan.c
+++ b/drivers/acpi/fan.c
@@ -158,8 +158,9 @@ static int fan_get_state(struct acpi_device *device, unsigned long *state)
 	if (result)
 		return result;
 
-	*state = (acpi_state == ACPI_STATE_D3_COLD ? 0 :
-		 (acpi_state == ACPI_STATE_D0 ? 1 : -1));
+	*state = acpi_state == ACPI_STATE_D3_COLD
+			|| acpi_state == ACPI_STATE_D3_HOT ?
+		0 : (acpi_state == ACPI_STATE_D0 ? 1 : -1);
 	return 0;
 }
 
diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c
index 5226a8b921ae..a3acd84f2533 100644
--- a/drivers/acpi/osl.c
+++ b/drivers/acpi/osl.c
@@ -536,7 +536,7 @@ static char acpi_os_name[ACPI_MAX_OVERRIDE_LEN];
 
 acpi_status
 acpi_os_predefined_override(const struct acpi_predefined_names *init_val,
-			    acpi_string * new_val)
+			    char **new_val)
 {
 	if (!init_val || !new_val)
 		return AE_BAD_PARAMETER;
diff --git a/drivers/acpi/pci_irq.c b/drivers/acpi/pci_irq.c
index b1def411c0b8..03e4b6c49c6d 100644
--- a/drivers/acpi/pci_irq.c
+++ b/drivers/acpi/pci_irq.c
@@ -44,7 +44,6 @@
 ACPI_MODULE_NAME("pci_irq");
 
 struct acpi_prt_entry {
-	struct list_head	list;
 	struct acpi_pci_id	id;
 	u8			pin;
 	acpi_handle		link;
diff --git a/drivers/acpi/power.c b/drivers/acpi/power.c
index e0bcfb642b52..93eac53b5110 100644
--- a/drivers/acpi/power.c
+++ b/drivers/acpi/power.c
@@ -684,7 +684,8 @@ int acpi_power_get_inferred_state(struct acpi_device *device, int *state)
 		}
 	}
 
-	*state = ACPI_STATE_D3_COLD;
+	*state = device->power.states[ACPI_STATE_D3_COLD].flags.valid ?
+		ACPI_STATE_D3_COLD : ACPI_STATE_D3_HOT;
 	return 0;
 }
 
@@ -710,8 +711,6 @@ int acpi_power_transition(struct acpi_device *device, int state)
 	    || (device->power.state > ACPI_STATE_D3_COLD))
 		return -ENODEV;
 
-	/* TBD: Resources must be ordered. */
-
 	/*
 	 * First we reference all power resources required in the target list
 	 * (e.g. so the device doesn't lose power while transitioning).  Then,
@@ -761,6 +760,25 @@ static void acpi_power_sysfs_remove(struct acpi_device *device)
 	device_remove_file(&device->dev, &dev_attr_resource_in_use);
 }
 
+static void acpi_power_add_resource_to_list(struct acpi_power_resource *resource)
+{
+	mutex_lock(&power_resource_list_lock);
+
+	if (!list_empty(&acpi_power_resource_list)) {
+		struct acpi_power_resource *r;
+
+		list_for_each_entry(r, &acpi_power_resource_list, list_node)
+			if (r->order > resource->order) {
+				list_add_tail(&resource->list_node, &r->list_node);
+				goto out;
+			}
+	}
+	list_add_tail(&resource->list_node, &acpi_power_resource_list);
+
+ out:
+	mutex_unlock(&power_resource_list_lock);
+}
+
 int acpi_add_power_resource(acpi_handle handle)
 {
 	struct acpi_power_resource *resource;
@@ -811,9 +829,7 @@ int acpi_add_power_resource(acpi_handle handle)
 	if (!device_create_file(&device->dev, &dev_attr_resource_in_use))
 		device->remove = acpi_power_sysfs_remove;
 
-	mutex_lock(&power_resource_list_lock);
-	list_add(&resource->list_node, &acpi_power_resource_list);
-	mutex_unlock(&power_resource_list_lock);
+	acpi_power_add_resource_to_list(resource);
 	acpi_device_add_finalize(device);
 	return 0;
 
@@ -844,7 +860,22 @@ void acpi_resume_power_resources(void)
 		    && resource->ref_count) {
 			dev_info(&resource->device.dev, "Turning ON\n");
 			__acpi_power_on(resource);
-		} else if (state == ACPI_POWER_RESOURCE_STATE_ON
+		}
+
+		mutex_unlock(&resource->resource_lock);
+	}
+	list_for_each_entry_reverse(resource, &acpi_power_resource_list, list_node) {
+		int result, state;
+
+		mutex_lock(&resource->resource_lock);
+
+		result = acpi_power_get_state(resource->device.handle, &state);
+		if (result) {
+			mutex_unlock(&resource->resource_lock);
+			continue;
+		}
+
+		if (state == ACPI_POWER_RESOURCE_STATE_ON
 		    && !resource->ref_count) {
 			dev_info(&resource->device.dev, "Turning OFF\n");
 			__acpi_power_off(resource);
diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
index 0a099917a006..0ca4574e89ea 100644
--- a/drivers/acpi/scan.c
+++ b/drivers/acpi/scan.c
@@ -1768,15 +1768,9 @@ static void acpi_bus_init_power_state(struct acpi_device *device, int state)
 	if (acpi_has_method(device->handle, pathname))
 		ps->flags.explicit_set = 1;
 
-	/*
-	 * State is valid if there are means to put the device into it.
-	 * D3hot is only valid if _PR3 present.
-	 */
-	if (!list_empty(&ps->resources)
-	    || (ps->flags.explicit_set && state < ACPI_STATE_D3_HOT)) {
+	/* State is valid if there are means to put the device into it. */
+	if (!list_empty(&ps->resources) || ps->flags.explicit_set)
 		ps->flags.valid = 1;
-		ps->flags.os_accessible = 1;
-	}
 
 	ps->power = -1;		/* Unknown - driver assigned */
 	ps->latency = -1;	/* Unknown - driver assigned */
@@ -1812,21 +1806,13 @@ static void acpi_bus_get_power_flags(struct acpi_device *device)
 		acpi_bus_init_power_state(device, i);
 
 	INIT_LIST_HEAD(&device->power.states[ACPI_STATE_D3_COLD].resources);
+	if (!list_empty(&device->power.states[ACPI_STATE_D3_HOT].resources))
+		device->power.states[ACPI_STATE_D3_COLD].flags.valid = 1;
 
-	/* Set defaults for D0 and D3 states (always valid) */
+	/* Set defaults for D0 and D3hot states (always valid) */
 	device->power.states[ACPI_STATE_D0].flags.valid = 1;
 	device->power.states[ACPI_STATE_D0].power = 100;
-	device->power.states[ACPI_STATE_D3_COLD].flags.valid = 1;
-	device->power.states[ACPI_STATE_D3_COLD].power = 0;
-
-	/* Set D3cold's explicit_set flag if _PS3 exists. */
-	if (device->power.states[ACPI_STATE_D3_HOT].flags.explicit_set)
-		device->power.states[ACPI_STATE_D3_COLD].flags.explicit_set = 1;
-
-	/* Presence of _PS3 or _PRx means we can put the device into D3 cold */
-	if (device->power.states[ACPI_STATE_D3_HOT].flags.explicit_set ||
-			device->power.flags.power_resources)
-		device->power.states[ACPI_STATE_D3_COLD].flags.os_accessible = 1;
+	device->power.states[ACPI_STATE_D3_HOT].flags.valid = 1;
 
 	if (acpi_bus_init_power(device))
 		device->flags.power_manageable = 0;