Bug#599768: toshiba_acpi: backlight broken after resume.
 

Hi,

this was first sent to debian as bug 599768, but Ben Hutchings
recommended sending it to you guys too.

anyway, the drill: backlight on my R800 laptop works fine until I
suspend it, then, /sys/class/backlight/* doesn't change anything
anymore.

This patch is taken from the old toshset tool (that. btw. seems to
document a few HCI commands we're not using in the toshiba_acpi code,
maybe interesting to have a look), and the important part is the code
that sets HCI_LCD_BRIGHTNESS in set_lcd.

after this patch, backlight works through
/sys/class/backlight/toshiba/brightness but for some reason the keys
(that are still detected) don't trigger any backlight change (or call of
set_lcd), ... and I really don't know how to debug that. (they still get
to me in the .notify callback, but I fail to find documentation on the
link between that and get_lcd...

(do I need to call backlight_device_register() or something along the
lines of the sparse_keymap code on resume ?)

anyway, I understand this patch is nowhere near ready for inclusion, but
I send it in order to see if it gives someone a clue to fix this
longstanding issue with toshiba backlights.

thanks,


--
Niv Sardi