Merge tag 'pm+acpi-3.18-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm

Pull ACPI and power management fixes from Rafael Wysocki:
 "These are fixes received after my previous pull request plus one that
  has been in the works for quite a while, but its previous version
  caused problems to happen, so it's been deferred till now.

  Fixed are two recent regressions (MFD enumeration and cpufreq-dt),
  ACPI EC regression introduced in 3.17, system suspend error code path
  regression introduced in 3.15, an older bug related to recovery from
  failing resume from hibernation and a cpufreq-dt driver issue related
  to operation performance points.

  Specifics:

   - Fix a crash on r8a7791/koelsch during resume from system suspend
     caused by a recent cpufreq-dt commit (Geert Uytterhoeven).

   - Fix an MFD enumeration problem introduced by a recent commit adding
     ACPI support to the MFD subsystem that exposed a weakness in the
     ACPI core causing ACPI enumeration to be applied to all devices
     associated with one ACPI companion object, although it should be
     used for one of them only (Mika Westerberg).

   - Fix an ACPI EC regression introduced during the 3.17 cycle causing
     some Samsung laptops to misbehave as a result of a workaround
     targeted at some Acer machines.  That includes a revert of a commit
     that went too far and a quirk for the Acer machines in question.
     From Lv Zheng.

   - Fix a regression in the system suspend error code path introduced
     during the 3.15 cycle that causes it to fail to take errors from
     asychronous execution of "late" suspend callbacks into account
     (Imre Deak).

   - Fix a long-standing bug in the hibernation resume error code path
     that fails to roll back everything correcty on "freeze" callback
     errors and leaves some devices in a "suspended" state causing more
     breakage to happen subsequently (Imre Deak).

   - Make the cpufreq-dt driver disable operation performance points
     that are not supported by the VR connected to the CPU voltage plane
     with acceptable tolerance instead of constantly failing voltage
     scaling later on (Lucas Stach)"

* tag 'pm+acpi-3.18-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  ACPI / EC: Fix regression due to conflicting firmware behavior between Samsung and Acer.
  Revert "ACPI / EC: Add support to disallow QR_EC to be issued before completing previous QR_EC"
  cpufreq: cpufreq-dt: Restore default cpumask_setall(policy->cpus)
  PM / Sleep: fix recovery during resuming from hibernation
  PM / Sleep: fix async suspend_late/freeze_late error handling
  ACPI: Use ACPI companion to match only the first physical device
  cpufreq: cpufreq-dt: disable unsupported OPPs

5 files changed, 126 insertions, 53 deletions

diff --git a/drivers/acpi/ec.c b/drivers/acpi/ec.c
index 3d304ff7f09..5f9b74b9b71 100644
--- a/drivers/acpi/ec.c
+++ b/drivers/acpi/ec.c
@@ -126,6 +126,7 @@ static int EC_FLAGS_MSI; /* Out-of-spec MSI controller */
 static int EC_FLAGS_VALIDATE_ECDT; /* ASUStec ECDTs need to be validated */
 static int EC_FLAGS_SKIP_DSDT_SCAN; /* Not all BIOS survive early DSDT scan */
 static int EC_FLAGS_CLEAR_ON_RESUME; /* Needs acpi_ec_clear() on boot/resume */
+static int EC_FLAGS_QUERY_HANDSHAKE; /* Needs QR_EC issued when SCI_EVT set */
 
 /* --------------------------------------------------------------------------
  *                           Transaction Management
@@ -236,13 +237,8 @@ static bool advance_transaction(struct acpi_ec *ec)
 		}
 		return wakeup;
 	} else {
-		/*
-		 * There is firmware refusing to respond QR_EC when SCI_EVT
-		 * is not set, for which case, we complete the QR_EC
-		 * without issuing it to the firmware.
-		 * https://bugzilla.kernel.org/show_bug.cgi?id=86211
-		 */
-		if (!(status & ACPI_EC_FLAG_SCI) &&
+		if (EC_FLAGS_QUERY_HANDSHAKE &&
+		    !(status & ACPI_EC_FLAG_SCI) &&
 		    (t->command == ACPI_EC_COMMAND_QUERY)) {
 			t->flags |= ACPI_EC_COMMAND_POLL;
 			t->rdata[t->ri++] = 0x00;
@@ -334,13 +330,13 @@ static int acpi_ec_transaction_unlocked(struct acpi_ec *ec,
 	pr_debug("***** Command(%s) started *****\n",
 		 acpi_ec_cmd_string(t->command));
 	start_transaction(ec);
-	spin_unlock_irqrestore(&ec->lock, tmp);
-	ret = ec_poll(ec);
-	spin_lock_irqsave(&ec->lock, tmp);
 	if (ec->curr->command == ACPI_EC_COMMAND_QUERY) {
 		clear_bit(EC_FLAGS_QUERY_PENDING, &ec->flags);
 		pr_debug("***** Event stopped *****\n");
 	}
+	spin_unlock_irqrestore(&ec->lock, tmp);
+	ret = ec_poll(ec);
+	spin_lock_irqsave(&ec->lock, tmp);
 	pr_debug("***** Command(%s) stopped *****\n",
 		 acpi_ec_cmd_string(t->command));
 	ec->curr = NULL;
@@ -1012,6 +1008,18 @@ static int ec_enlarge_storm_threshold(const struct dmi_system_id *id)
 }
 
 /*
+ * Acer EC firmware refuses to respond QR_EC when SCI_EVT is not set, for
+ * which case, we complete the QR_EC without issuing it to the firmware.
+ * https://bugzilla.kernel.org/show_bug.cgi?id=86211
+ */
+static int ec_flag_query_handshake(const struct dmi_system_id *id)
+{
+	pr_debug("Detected the EC firmware requiring QR_EC issued when SCI_EVT set\n");
+	EC_FLAGS_QUERY_HANDSHAKE = 1;
+	return 0;
+}
+
+/*
  * On some hardware it is necessary to clear events accumulated by the EC during
  * sleep. These ECs stop reporting GPEs until they are manually polled, if too
  * many events are accumulated. (e.g. Samsung Series 5/9 notebooks)
@@ -1085,6 +1093,9 @@ static struct dmi_system_id ec_dmi_table[] __initdata = {
 	{
 	ec_clear_on_resume, "Samsung hardware", {
 	DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD.")}, NULL},
+	{
+	ec_flag_query_handshake, "Acer hardware", {
+	DMI_MATCH(DMI_SYS_VENDOR, "Acer"), }, NULL},
 	{},
 };
 
diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
index d670158a26c..0476e90b209 100644
--- a/drivers/acpi/scan.c
+++ b/drivers/acpi/scan.c
@@ -142,6 +142,53 @@ static int create_modalias(struct acpi_device *acpi_dev, char *modalias,
 }
 
 /*
+ * acpi_companion_match() - Can we match via ACPI companion device
+ * @dev: Device in question
+ *
+ * Check if the given device has an ACPI companion and if that companion has
+ * a valid list of PNP IDs, and if the device is the first (primary) physical
+ * device associated with it.
+ *
+ * If multiple physical devices are attached to a single ACPI companion, we need
+ * to be careful.  The usage scenario for this kind of relationship is that all
+ * of the physical devices in question use resources provided by the ACPI
+ * companion.  A typical case is an MFD device where all the sub-devices share
+ * the parent's ACPI companion.  In such cases we can only allow the primary
+ * (first) physical device to be matched with the help of the companion's PNP
+ * IDs.
+ *
+ * Additional physical devices sharing the ACPI companion can still use
+ * resources available from it but they will be matched normally using functions
+ * provided by their bus types (and analogously for their modalias).
+ */
+static bool acpi_companion_match(const struct device *dev)
+{
+	struct acpi_device *adev;
+	bool ret;
+
+	adev = ACPI_COMPANION(dev);
+	if (!adev)
+		return false;
+
+	if (list_empty(&adev->pnp.ids))
+		return false;
+
+	mutex_lock(&adev->physical_node_lock);
+	if (list_empty(&adev->physical_node_list)) {
+		ret = false;
+	} else {
+		const struct acpi_device_physical_node *node;
+
+		node = list_first_entry(&adev->physical_node_list,
+					struct acpi_device_physical_node, node);
+		ret = node->dev == dev;
+	}
+	mutex_unlock(&adev->physical_node_lock);
+
+	return ret;
+}
+
+/*
  * Creates uevent modalias field for ACPI enumerated devices.
  * Because the other buses does not support ACPI HIDs & CIDs.
  * e.g. for a device with hid:IBM0001 and cid:ACPI0001 you get:
@@ -149,20 +196,14 @@ static int create_modalias(struct acpi_device *acpi_dev, char *modalias,
  */
 int acpi_device_uevent_modalias(struct device *dev, struct kobj_uevent_env *env)
 {
-	struct acpi_device *acpi_dev;
 	int len;
 
-	acpi_dev = ACPI_COMPANION(dev);
-	if (!acpi_dev)
-		return -ENODEV;
-
-	/* Fall back to bus specific way of modalias exporting */
-	if (list_empty(&acpi_dev->pnp.ids))
+	if (!acpi_companion_match(dev))
 		return -ENODEV;
 
 	if (add_uevent_var(env, "MODALIAS="))
 		return -ENOMEM;
-	len = create_modalias(acpi_dev, &env->buf[env->buflen - 1],
+	len = create_modalias(ACPI_COMPANION(dev), &env->buf[env->buflen - 1],
 				sizeof(env->buf) - env->buflen);
 	if (len <= 0)
 		return len;
@@ -179,18 +220,12 @@ EXPORT_SYMBOL_GPL(acpi_device_uevent_modalias);
  */
 int acpi_device_modalias(struct device *dev, char *buf, int size)
 {
-	struct acpi_device *acpi_dev;
 	int len;
 
-	acpi_dev = ACPI_COMPANION(dev);
-	if (!acpi_dev)
+	if (!acpi_companion_match(dev))
 		return -ENODEV;
 
-	/* Fall back to bus specific way of modalias exporting */
-	if (list_empty(&acpi_dev->pnp.ids))
-		return -ENODEV;
-
-	len = create_modalias(acpi_dev, buf, size -1);
+	len = create_modalias(ACPI_COMPANION(dev), buf, size -1);
 	if (len <= 0)
 		return len;
 	buf[len++] = '\n';
@@ -853,6 +888,9 @@ const struct acpi_device_id *acpi_match_device(const struct acpi_device_id *ids,
 	if (!ids || !handle || acpi_bus_get_device(handle, &adev))
 		return NULL;
 
+	if (!acpi_companion_match(dev))
+		return NULL;
+
 	return __acpi_match_device(adev, ids);
 }
 EXPORT_SYMBOL_GPL(acpi_match_device);
diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c
index 44973196d3f..9717d5f2013 100644
--- a/drivers/base/power/main.c
+++ b/drivers/base/power/main.c
@@ -1266,6 +1266,8 @@ int dpm_suspend_late(pm_message_t state)
 	}
 	mutex_unlock(&dpm_list_mtx);
 	async_synchronize_full();
+	if (!error)
+		error = async_error;
 	if (error) {
 		suspend_stats.failed_suspend_late++;
 		dpm_save_failed_step(SUSPEND_SUSPEND_LATE);
diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c
index 92c162af504..23aaf40cf37 100644
--- a/drivers/cpufreq/cpufreq-dt.c
+++ b/drivers/cpufreq/cpufreq-dt.c
@@ -187,6 +187,7 @@ static int cpufreq_init(struct cpufreq_policy *policy)
 	struct device *cpu_dev;
 	struct regulator *cpu_reg;
 	struct clk *cpu_clk;
+	unsigned long min_uV = ~0, max_uV = 0;
 	unsigned int transition_latency;
 	int ret;
 
@@ -206,16 +207,10 @@ static int cpufreq_init(struct cpufreq_policy *policy)
 	/* OPPs might be populated at runtime, don't check for error here */
 	of_init_opp_table(cpu_dev);
 
-	ret = dev_pm_opp_init_cpufreq_table(cpu_dev, &freq_table);
-	if (ret) {
-		dev_err(cpu_dev, "failed to init cpufreq table: %d\n", ret);
-		goto out_put_node;
-	}
-
 	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
 	if (!priv) {
 		ret = -ENOMEM;
-		goto out_free_table;
+		goto out_put_node;
 	}
 
 	of_property_read_u32(np, "voltage-tolerance", &priv->voltage_tolerance);
@@ -224,30 +219,51 @@ static int cpufreq_init(struct cpufreq_policy *policy)
 		transition_latency = CPUFREQ_ETERNAL;
 
 	if (!IS_ERR(cpu_reg)) {
-		struct dev_pm_opp *opp;
-		unsigned long min_uV, max_uV;
-		int i;
+		unsigned long opp_freq = 0;
 
 		/*
-		 * OPP is maintained in order of increasing frequency, and
-		 * freq_table initialised from OPP is therefore sorted in the
-		 * same order.
+		 * Disable any OPPs where the connected regulator isn't able to
+		 * provide the specified voltage and record minimum and maximum
+		 * voltage levels.
 		 */
-		for (i = 0; freq_table[i].frequency != CPUFREQ_TABLE_END; i++)
-			;
-		rcu_read_lock();
-		opp = dev_pm_opp_find_freq_exact(cpu_dev,
-				freq_table[0].frequency * 1000, true);
-		min_uV = dev_pm_opp_get_voltage(opp);
-		opp = dev_pm_opp_find_freq_exact(cpu_dev,
-				freq_table[i-1].frequency * 1000, true);
-		max_uV = dev_pm_opp_get_voltage(opp);
-		rcu_read_unlock();
+		while (1) {
+			struct dev_pm_opp *opp;
+			unsigned long opp_uV, tol_uV;
+
+			rcu_read_lock();
+			opp = dev_pm_opp_find_freq_ceil(cpu_dev, &opp_freq);
+			if (IS_ERR(opp)) {
+				rcu_read_unlock();
+				break;
+			}
+			opp_uV = dev_pm_opp_get_voltage(opp);
+			rcu_read_unlock();
+
+			tol_uV = opp_uV * priv->voltage_tolerance / 100;
+			if (regulator_is_supported_voltage(cpu_reg, opp_uV,
+							   opp_uV + tol_uV)) {
+				if (opp_uV < min_uV)
+					min_uV = opp_uV;
+				if (opp_uV > max_uV)
+					max_uV = opp_uV;
+			} else {
+				dev_pm_opp_disable(cpu_dev, opp_freq);
+			}
+
+			opp_freq++;
+		}
+
 		ret = regulator_set_voltage_time(cpu_reg, min_uV, max_uV);
 		if (ret > 0)
 			transition_latency += ret * 1000;
 	}
 
+	ret = dev_pm_opp_init_cpufreq_table(cpu_dev, &freq_table);
+	if (ret) {
+		pr_err("failed to init cpufreq table: %d\n", ret);
+		goto out_free_priv;
+	}
+
 	/*
 	 * For now, just loading the cooling device;
 	 * thermal DT code takes care of matching them.
@@ -277,7 +293,7 @@ static int cpufreq_init(struct cpufreq_policy *policy)
 	policy->cpuinfo.transition_latency = transition_latency;
 
 	pd = cpufreq_get_driver_data();
-	if (pd && !pd->independent_clocks)
+	if (!pd || !pd->independent_clocks)
 		cpumask_setall(policy->cpus);
 
 	of_node_put(np);
@@ -286,9 +302,9 @@ static int cpufreq_init(struct cpufreq_policy *policy)
 
 out_cooling_unregister:
 	cpufreq_cooling_unregister(priv->cdev);
-	kfree(priv);
-out_free_table:
 	dev_pm_opp_free_cpufreq_table(cpu_dev, &freq_table);
+out_free_priv:
+	kfree(priv);
 out_put_node:
 	of_node_put(np);
 out_put_reg_clk:
diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c
index a9dfa79b6ba..1f35a3478f3 100644
--- a/kernel/power/hibernate.c
+++ b/kernel/power/hibernate.c
@@ -502,8 +502,14 @@ int hibernation_restore(int platform_mode)
 	error = dpm_suspend_start(PMSG_QUIESCE);
 	if (!error) {
 		error = resume_target_kernel(platform_mode);
-		dpm_resume_end(PMSG_RECOVER);
+		/*
+		 * The above should either succeed and jump to the new kernel,
+		 * or return with an error. Otherwise things are just
+		 * undefined, so let's be paranoid.
+		 */
+		BUG_ON(!error);
 	}
+	dpm_resume_end(PMSG_RECOVER);
 	pm_restore_gfp_mask();
 	resume_console();
 	pm_restore_console();