[PATCH 0/3] Use acpi_video_unregister_backlight instead of acpi_video_unregister in platfrom backlight drivers
aaron.lu at intel.com
Tue Jun 2 07:33:50 PDT 2015
On Tue, Jun 02, 2015 at 12:14:28PM +0200, Hans de Goede wrote:
> On 02-06-15 11:59, Jani Nikula wrote:
> >On Mon, 01 Jun 2015, Hans de Goede <hdegoede at redhat.com> wrote:
> >>Hi All,
> >>I'm working on cleaning up the currently somewhat convoluted logic to
> >>select which backlight interfaces to register on x86 systems, see:
> >>For a rought outline (details will change in the actual patch-set).
> >>These 3 patches are a preparation for that work, as the behavior of the
> >>current code is not always consistent (it changes depending on module
> >>loading order in some cases). These 3 patches remove this inconsistency
> >>which in some cases may result in a behavior change.
> >>This way the cleanup can have a consistent base to build upon, and I can
> >>ensure that the cleanup itself does not cause any functional changes.
> >Thanks for doing this.
> >Slightly unrelated, but you'll end up calling:
> > acpi_video_bus_unregister_backlight
> > acpi_video_dev_unregister_backlight
> > thermal_cooling_device_unregister
> >Uhm, err, what? What business does *that* have in the backlight
> >unregister call chain?!
> I think the idea is that having the backlight on can cause the main board
> to heat up as there is often a regulator / pwm for the backlight on the
> main board. If that is turned off it may help to cool down the main board,
> so the backlight is a cooling device in the sense that when it is turned
> off it stops generating heat.
I think so.
> If this makes sense (*) then the cooling device stuff should maybe moved
> from the acpi/video.c code to the backlight core code.
It's just a specific driver that provides thermal capability, no need to
move to the core code.
> Or does the acpi-video code use some info from the BIOS to determine
> whether or not to register a cooling device ?
As long as we create the ACPI video backlight interface, we will create
the thermal cooling device. If we somehow knows that the ACPI methods to
adjust the backlight level doesn't work(a good hint is that the
acpi_video_unregister_backlight is called), we should make that cooling
> *) This is tricky, e.g. going low brightness may cause the main board
> to heat up more then full brightness depending on how things are hooked
> Should that be untangled from the mess too?
> >>I hope to post a v1 of the actual cleanup patch-set in 1-2 weeks.
> >>dri-devel mailing list
> >>dri-devel at lists.freedesktop.org
More information about the dri-devel