Re: RFC: Backlight Class sysfs attribute behaviour

[Richard Purdie <rpurdie@rpsys.net>]

On Tue, 2006-03-07 at 00:07 +0300, Andrew Zabolotny wrote: 
> On Mon, 06 Mar 2006 08:45:28 +0000
> Richard Purdie <rpurdie@rpsys.net> wrote:
> 
> > * the user supplied power sysfs attribute
> > * the user supplied brightness sysfs attribute
> > * the current FB blanking state
> > * any other driver specific factors
> As far I see the only real concern here is the console blanking. So why
> not make it just another device state flag, which doesn't influence the
> 'power' attribute and which isn't visible from user space (what for?).
> And the 'real' power state will be computed as "blank && power"
> attributes. The entire logic could be hidden in backlight.c so that no
> driver will have to be modified for this. Maybe a 'hw_power' or such
> would be needed to see the 'real' hardware state (and possibly
> modify, if you really really want to).
> 
> Is there any need for a broader-range solution here?

You've ignored the "driver specific factors" and one of the drivers
already has such a thing (borgi_bl's low battery backlight limiting).

My driver_brightness is really a more generic version of your "hw_power"
which allows other factors to be taken into consideration as well. We
may as well have the broader-range solution as it costs us nothing.

(I don't expect each factor to be visible to userspace, just the end
result).

Richard