Bug#942219: systemd-backlight at backlight:dell_backlight.service fails at boot for dell d600, d620, d820

[Michael Biebl]

Hi Ben,

I'm bringing you into the loop here. Maybe you can help as kernel
maintainer (or know someone we could ask).

On Thu, 17 Oct 2019 11:26:05 +0200 Rado S <d-ebian at q3co.de> wrote:
> =- Michael Biebl wrote on Wed 16.Oct'19 at 13:06:29 +0200 -=
> 
> > >>> # cat /sys/class/backlight/intel_backlight/brightness
> > >>> 1500
> > > 
> > > I get 79560.
> > > 
> > >> and change it via
> > >>> # echo 2000 > /sys/class/backlight/intel_backlight/brightness
> > > 
> > > Setting it via intel to 66666 makes it darker.
> > > 
> > >> What's the output you get for your dell_backlight?
> > > 
> > >>From dell_backlight I get:
> > > # cat dell_backlight/brightness
> > > 6
> > > 
> > > But when setting it, it says:
> > > 
> > > # echo 5 > dell_backlight/brightness
> > > -ksh: echo: write to 1 failed [Input/output error]
> > > {...}
> > > What next?
> > 
> > Dunno. You'll probably need to ask someone who is familiar with
> > Dell hardware. I have no idea why these Dell laptops provide a
> > dell_backlight device and what this device supposed to control.
> > Apparently it can't be used to change the brightness despite it
> > exposing the brightness attribute.
> 
> For the record:
> the dell value for brightness reflects the scale given in the BIOS
> menu: (0-7)
> 
> The dell vostro works OK:
> despite having an Intel GM45 it has no intel_backlight device at
> the same place as the others.
> Instead it has acpi_video0 which scales like the dell_backlight
> (values 0-7) rather than the intel_backlight (*1000).
> At Installation time xorg also decided not to use a gpu specific
> driver (xorg-video-intel) with the vostro for the acpi_video0, which
> in return made X operate slow compared to the older dells, which
> still do use the xorg-video-intel (or radeon).
> 
> Which components of the installer decide whether to use xorg-drivers
> and which devices to create for backlight handling?
> 
> Is this still systemd or again a kernel issue?

Tbh, I don't know.
Ben, maybe you can help here:
If a backlight device exposes a brightness attribute, is it safe to
assume that this property is writable?
If not, is there a way to determine whether it is writable?

Regards,
Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <http://alioth-lists.debian.net/pipermail/pkg-systemd-maintainers/attachments/20191105/8c0e980c/attachment.sig>