On Mon, Jun 21, 2010 at 12:08 PM, Alberto Milone alberto.milone@canonical.com wrote:

On 21 June 2010 16:39, Alex Deucher alexdeucher@gmail.com wrote:

...

On Mon, Jun 21, 2010 at 10:21 AM, Alberto Milone alberto.milone@canonical.com wrote:

...

Hi all,

Thanks to the help of Alex Deucher and Matthew Garrett, I've added support for calling the ATIF ACPI method to the radeon driver. This makes the video switch hotkey work properly, as we get an ACPI event when the key is pressed.

Note: I guess it depends on the bios I'm working with, but I also need to pass acpi_osi="!Windows 2009" on boot in order to get the ACPI event.

This is due to windows 7 wanting win-p for display switch events: http://mjg59.livejournal.com/121851.html

Couple of minor nits:

+/* radeon_acpi.c */ +#if defined(CONFIG_ACPI) +extern int radeon_acpi_init(struct radeon_device *rdev); +#else +static inline int radeon_acpi_init(struct radeon_device *rdev) { return 1; } +#endif

We probably want to return 0 in the non CONFIG_ACPI case?  On systems like ppc without acpi, there's no need for the acpi stuff so the warning would be needless.  On systems with acpi that are compiled without it, well, they will probably have issues anyway...

• /* No need to proceed if we're sure that ATIF is not supported */
• if (!ASIC_IS_AVIVO(rdev) || !rdev->bios)
• return 0;

Move this check to radeon_acpi_init() for now and there's no need to pass rdev to radeon_atif_call() anymore.

With those changes:

Reviewed-by: Alex Deucher alexdeucher@gmail.com

Alex

You made some really good points.

Thanks for reviewing the patch. Here's the attached patch with the changes that you requested.

Looks good. Thanks!

Reviewed-by: Alex Deucher alexdeucher@gmail.com

Alex