Re: [patch 2.6.25-rc6 5/7] ACPI sets up device.power.can_wakeup flags

[Zhao Yakui <yakui.zhao@intel.com>]

On Thu, 2008-03-20 at 14:20 -0700, David Brownell wrote:
> This exports the "acpi_device.wakeup.flags.valid" flag to the driver
> model as "device.power.can_wakeup".  That is, each "ACPI device" in
> the /proc/acpi/wakeup file that corresponds to a "real" device will
> initialize the can_wakeup flag to true.
> 
For the button device it is good to set the wakeup flag. 
Good. When the acpi device supports the _PRW object, the wakeup flag bit
is set.( Can_wakeup; should_wakeup).
> A separate patch is needed to make ACPI pay attention to the userspace
> input:  devices that can_wakeup have a should_wakeup flag, initially
> true, which is settable via sysfs and tested using device_may_wakeup().
> That's intended to be the primary policy input to the system, not the
> ACPI-specific /proc/acpi/wakeup file.


> Note also that the ACPI tables list only motherboard devices, but many
> systems include more peripherals than those.  While the USB framework
> handles USB peripherals, nothing in Linux currently sets those flags
> for wakeup-capable devices on other expansion busses (like PCI).
> 
> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
> ---
>  drivers/acpi/button.c |    3 +++
>  drivers/acpi/glue.c   |    1 +
>  drivers/acpi/power.c  |    4 ++++
>  3 files changed, 8 insertions(+)
> 
> --- g26.orig/drivers/acpi/button.c	2008-03-20 03:02:45.000000000 -0700
> +++ g26/drivers/acpi/button.c	2008-03-20 03:27:24.000000000 -0700
> @@ -476,6 +476,9 @@ static int acpi_button_add(struct acpi_d
>  	if (button->type == ACPI_BUTTON_TYPE_LID)
>  		acpi_lid_send_state(button);
>  
> +	/* for these devices the ACPI node *IS* the "physical" device */
> +	device_init_wakeup(&device->dev, device->wakeup.flags.valid);
> +

>  	if (device->wakeup.flags.valid) {
>  		/* Button's GPE is run-wake GPE */
>  		acpi_set_gpe_type(device->wakeup.gpe_device,
> --- g26.orig/drivers/acpi/glue.c	2008-03-20 03:27:11.000000000 -0700
> +++ g26/drivers/acpi/glue.c	2008-03-20 03:27:25.000000000 -0700
> @@ -162,6 +162,7 @@ static int acpi_bind_one(struct device *
>  	if (!ACPI_FAILURE(status)) {
>  		int ret;
>  
> +		device_init_wakeup(dev, acpi_dev->wakeup.flags.valid);
>  		ret = sysfs_create_link(&dev->kobj, &acpi_dev->dev.kobj,
>  				"firmware_node");
>  		if (ret != 0)
> --- g26.orig/drivers/acpi/power.c	2008-03-20 03:02:45.000000000 -0700
> +++ g26/drivers/acpi/power.c	2008-03-20 03:27:25.000000000 -0700
> @@ -313,6 +313,7 @@ int acpi_enable_wakeup_device_power(stru
>  		ret = acpi_power_on(dev->wakeup.resources.handles[i], dev);
>  		if (ret) {
>  			printk(KERN_ERR PREFIX "Transition power state\n");
> +			device_init_wakeup(&dev->dev, 0);
>  			dev->wakeup.flags.valid = 0;
>  			return -1;
>  		}
> @@ -322,6 +323,7 @@ int acpi_enable_wakeup_device_power(stru
>  	status = acpi_evaluate_object(dev->handle, "_PSW", &arg_list, NULL);
>  	if (ACPI_FAILURE(status) && (status != AE_NOT_FOUND)) {
>  		printk(KERN_ERR PREFIX "Evaluate _PSW\n");
> +		device_init_wakeup(&dev->dev, 0);
>  		dev->wakeup.flags.valid = 0;
>  		ret = -1;
>  	}
> @@ -351,6 +353,7 @@ int acpi_disable_wakeup_device_power(str
>  	status = acpi_evaluate_object(dev->handle, "_PSW", &arg_list, NULL);
>  	if (ACPI_FAILURE(status) && (status != AE_NOT_FOUND)) {
>  		printk(KERN_ERR PREFIX "Evaluate _PSW\n");
> +		device_init_wakeup(&dev->dev, 0);
>  		dev->wakeup.flags.valid = 0;
>  		return -1;
>  	}
> @@ -360,6 +363,7 @@ int acpi_disable_wakeup_device_power(str
>  		ret = acpi_power_off_device(dev->wakeup.resources.handles[i], dev);
>  		if (ret) {
>  			printk(KERN_ERR PREFIX "Transition power state\n");
> +			device_init_wakeup(&dev->dev, 0);
>  			dev->wakeup.flags.valid = 0;
>  			return -1;
>  		}
> --
> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html