[PATCH 0/2] ACPI / PM: Fix suspend regression related to button GPEs

[PATCH 0/2] ACPI / PM: Fix suspend regression related to button GPEs

[Rafael J. Wysocki]

Hi,

The first of the following patches fixes the suspend regression tracked in
References: https://bugzilla.kernel.org/show_bug.cgi?id=27372
and the second one corrects a small ugliness in the button driver.

Please apply!

Thanks,
Rafael

[PATCH 1/2] ACPI / Wakeup: Enable button GPEs unconditionally during initialization

[Rafael J. Wysocki]

From: Rafael J. Wysocki <rjw@sisk.pl>

Commit 9630bdd (ACPI: Use GPE reference counting to support shared
GPEs) introduced a suspend regression where boxes resume immediately
after being suspended due to the lid or sleep button wakeup status
not being cleared properly.  This happens if the GPEs corresponding
to those devices are not enabled all the time, which apparently is
expected by some BIOSes.

To fix this problem, enable button and lid GPEs unconditionally
during initialization and keep them enabled all the time, regardless
of whether or not the ACPI button driver is used.

References: https://bugzilla.kernel.org/show_bug.cgi?id=27372
Reported-and-tested-by: Ferenc Wágner <wferi@niif.hu>
Cc: stable@kernel.org
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
---
 drivers/acpi/button.c |   13 -------------
 drivers/acpi/wakeup.c |    6 +++++-
 2 files changed, 5 insertions(+), 14 deletions(-)

Index: linux-2.6/drivers/acpi/wakeup.c
===================================================================
--- linux-2.6.orig/drivers/acpi/wakeup.c
+++ linux-2.6/drivers/acpi/wakeup.c
@@ -86,8 +86,12 @@ int __init acpi_wakeup_device_init(void)
 		struct acpi_device *dev = container_of(node,
 						       struct acpi_device,
 						       wakeup_list);
-		if (device_can_wakeup(&dev->dev))
+		if (device_can_wakeup(&dev->dev)) {
+			/* Button GPEs are supposed to be always enabled. */
+			acpi_enable_gpe(dev->wakeup.gpe_device,
+					dev->wakeup.gpe_number);
 			device_set_wakeup_enable(&dev->dev, true);
+		}
 	}
 	mutex_unlock(&acpi_device_lock);
 	return 0;

[PATCH 2/2] ACPI / Button: Avoid disabling wakeup unnecessarily on remove

[Rafael J. Wysocki]

From: Rafael J. Wysocki <rjw@sisk.pl>

If a button device had already been enabled to wake up the system
from sleep states before the button driver saw it, the driver
shouldn't disable the device's wakeup capability when being detached
from the device.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
---
 drivers/acpi/button.c |    9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

Index: linux-2.6/drivers/acpi/button.c
===================================================================
--- linux-2.6.orig/drivers/acpi/button.c
+++ linux-2.6/drivers/acpi/button.c
@@ -98,6 +98,7 @@ struct acpi_button {
 	struct input_dev *input;
 	char phys[32];			/* for input device */
 	unsigned long pushed;
+	bool wakeup_enabled;
 };
 
 static const struct file_operations acpi_button_info_fops = {
@@ -430,7 +431,10 @@ static int acpi_button_add(struct acpi_d
 		/* Button's GPE is run-wake GPE */
 		acpi_enable_gpe(device->wakeup.gpe_device,
 				device->wakeup.gpe_number);
-		device_set_wakeup_enable(&device->dev, true);
+		if (!device_may_wakeup(&device->dev)) {
+			device_set_wakeup_enable(&device->dev, true);
+			button->wakeup_enabled = true;
+		}
 	}
 
 	printk(KERN_INFO PREFIX "%s [%s]\n", name, acpi_device_bid(device));
@@ -452,7 +456,8 @@ static int acpi_button_remove(struct acp
 	if (device->wakeup.flags.valid) {
 		acpi_disable_gpe(device->wakeup.gpe_device,
 				device->wakeup.gpe_number);
-		device_set_wakeup_enable(&device->dev, false);
+		if (button->wakeup_enabled)
+			device_set_wakeup_enable(&device->dev, false);
 	}
 
 	acpi_button_remove_fs(device);

[Updated][PATCH 1/2] ACPI / Wakeup: Enable button GPEs unconditionally during initialization

[Rafael J. Wysocki]

From: Rafael J. Wysocki <rjw@sisk.pl>
Subject: ACPI / Wakeup: Enable button GPEs unconditionally during initialization

Commit 9630bdd (ACPI: Use GPE reference counting to support shared
GPEs) introduced a suspend regression where boxes resume immediately
after being suspended due to the lid or sleep button wakeup status
not being cleared properly.  This happens if the GPEs corresponding
to those devices are not enabled all the time, which apparently is
expected by some BIOSes.

To fix this problem, enable button and lid GPEs unconditionally
during initialization and keep them enabled all the time, regardless
of whether or not the ACPI button driver is used.

References: https://bugzilla.kernel.org/show_bug.cgi?id=27372
Reported-and-tested-by: Ferenc Wágner <wferi@niif.hu>
Cc: stable@kernel.org
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
---

Hi,

This time with the correct diffstat.

Thanks,
Rafael

---
 drivers/acpi/wakeup.c |    6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

Index: linux-2.6/drivers/acpi/wakeup.c
===================================================================
--- linux-2.6.orig/drivers/acpi/wakeup.c
+++ linux-2.6/drivers/acpi/wakeup.c
@@ -86,8 +86,12 @@ int __init acpi_wakeup_device_init(void)
 		struct acpi_device *dev = container_of(node,
 						       struct acpi_device,
 						       wakeup_list);
-		if (device_can_wakeup(&dev->dev))
+		if (device_can_wakeup(&dev->dev)) {
+			/* Button GPEs are supposed to be always enabled. */
+			acpi_enable_gpe(dev->wakeup.gpe_device,
+					dev->wakeup.gpe_number);
 			device_set_wakeup_enable(&dev->dev, true);
+		}
 	}
 	mutex_unlock(&acpi_device_lock);
 	return 0;