[Intel-gfx] BACKLIGHT property for KMS case

[Matthew Garrett]

On Tue, Aug 18, 2009 at 01:52:03PM +0200, Matthias Hopf wrote:

> I tried to use that on an hp netbook here, but somehow this had no
> effect.

It'll depend what interface is in use - for instance, ACPI typically 
won't let you disable the backlight completely.

> Anyway, I'd like to use the "0 is off, 1 is minimum" semantics in the
> RandR properties - we'd probably have to do some magic in the driver.

In that case you'll have to add 1 to the values when dealing with most 
/sys/class/backlight implementations.

> power is only allowed to have 0 and 1 as values?

According to the header files, 0 is on, 1-3 are DPMS states, 4 is fully 
off. See include/linux/fb.h - it's the FB_BLANK_* defines.

-- 
Matthew Garrett | mjg59 at srcf.ucam.org