Fwd: Samsung NP300V5A -- no backlight brightness change.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hello all!

Fn + F2(down) and Fn + F3(up) doen't change brightness level of
backlight on my Samsung NP300V5A laptop. Display is simply flickering.
I have OpenSUSE 12.3 with
bumblebee installed as described in Wiki ( Repo URL: Index of
/repositories/home:/Overman79:/Laptop/openSUSE_12.3 ). There we no
brightness change before installation of bumblebee (so I don't think
it has influence).

I have used attached patch (samsung-laptop-newline.patch), it only add
missing newline to make debug info readable. And I think it already
can be inserted to real code.

So there is different information about problem. I have used attached
script(samsung-generate-report.sh) to make reports with huge amount of
different information that could be very helpful. See it to know how
report is organized.

************first mode with NO acpi_backlight=vendor *******************
*****There are 3 backlight drivers:
acpi_video0
acpi_video1
intel_backlight

*****Fn behavior:
Brightness can be lowered. Trying to increase it causes set *real*
brightness to MAX value.

*****See attached samsung-brightness-report-acpivideo.txt
Additional comments: brightness has been changing *real* (when I say
real I mean brightness that I can see with my eyes) correctly with
setting brightness through acpi_video0, acpi_video1 and
intel_backlight. But sending commands through debugfs of
samsung-laptop making screen only flickering.

It seems that problem is in handling of Fn keys, maybe....

************second mode with acpi_backlight=vendor *******************
*****There are 2 backlight drivers:
intel_backlight
samsung

*****Fn behavior:
Brightness can only be lowered. Trying to increase it causes flickering.

*****See attached samsung-brightness-report-acpivendor.txt
Additional comments: brightness *really* correctly has been changing
only through intel_backlight driver. Behavior of samsung driver and
debugfs commands is very specific: if you have brightness for ex. 4
and want to make it 7, first such command will set it to 5, second to
6 and >=3rd will make it 7. This behavior has been mentioned in module
code as stepping_quirk, but comments says that "if you will place
brightness to 0 and then to needed all will be OK", but my system
seems to be an exception.

************second mode with acpi_backlight=vendor and whileloop patch *****
So knowing it, I have decided to make a patch
(samsung-laptop-whileloop.patch), that will call set_brightness until
it get needed result. And you can see appropriate
report(samsung-brightness-report-whileloop.patch) for it. All still is
good with intel_backlight and debugfs behavior is same, but changing
samsung driver brightness causes to set brightness randomly to MAX or
MIN, and in log you can see that brightness doesn't change after same
0x11 command and fallbacks with "Brightness can't be set" message. And
it is pretty strange because through debugfs all is fine. Fn causes to
decrease/increase brightness with very big steps. Maybe driver 0x11
commands are buffered and it makes such behavior??

So now questions I have:
1) Should system work properly with both acpi_backlight=vendor and without?
2) Where I can found code that handles Fn keys?
3) Is SABI commands are buffered?

I really want to get rid with this bug. Here you can see previous
thread at openSUSE mailing list, to know history of my investigations:
http://lists.opensuse.org/opensuse-kernel/2013-05/msg00063.html

Best regards,
Denis Kuplyakov

Attachment: samsung-files.tar.gz
Description: GNU Zip compressed data


[Index of Archives]     [Linux Kernel Development]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux