[PATCH] radeon: Expose backlight class device for legacy LVDS encoder

[Alex Deucher]

On Wed, Sep 8, 2010 at 1:03 PM, Matthew Garrett <mjg59 at srcf.ucam.org> wrote:
> On Wed, Sep 08, 2010 at 12:58:32PM -0400, Alex Deucher wrote:
>
>> The only problem with this is that not all oems use the internal
>> backlight controller; systems that don't need to use the acpi methods.
>
> That's why we expose the backlight type. Userspace should use the acpi
> or platform mechanism when available, with this being a last-ditch
> fallback.

Ah, gotcha.

>
>>  On atombios systems there is a bit in the
>> ATOM_FIRMWARE_CAPABILITY_ACCESS struct in the FirmwareInfo data table
>> to determine whether the backlight is controlled by the GPU or some
>> external mechanism.  Combios may have something similar.  If the
>> backlight is controlled via the GPU, it can be adjusting using the
>> atom OutputControl and TransmitterControl control tables depending on
>> the GPU family.  However, if the driver chooses to control the
>> backlight itself, it needs to set the appropriate bit in the bios
>> scratch regs to tell the firmware not to attempt to change the
>> backlight itself.
>
> If there's support for probing this more reliably then I'm all for that,
> but I'm not keen on taking over control if the BIOS has previous
> asserted it.

Agreed.

Alex