[opensuse-kernel] Samsung NP300V5A -- no backlight brightness change
Hello all! Fn + F2(down) and Fn + F3(up) don't change brightness level of backlight on my Samsung NP300V5A laptop. 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) With no special setup, I can't lower brightness (screen is flickering), and have 3 entries under /sys/class/backlight: Code: acpi0_backlight acpi1_backlight intel_backlight Haven't tried to manipulate them through console. With acpi_backlight=vendor, I can't up brightness, and have 2 entries: Code: intel_backlight samsung Changing brightness with intel_backlight through console seems OK, but changing it with samsung don't work, I can set only 0 or 1 brightness level(same with Fn-##) Code: dmesg | grep samsung [ 16.423716] samsung_laptop: enabled workaround for brightness stepping quirk [ 16.424060] samsung_laptop: detected SABI interface: SwSmi@ I have solved problem for me, by deleting strange workaround from samsung-laptop.c and recompiling module. I think that it is a bug. I can see that we have some notebooks models hardcoded in module for quirk workaround (there is no my model). So I think that actual bug is inside code that determines whether workaround is needed or not (it is check_for_stepping_quirk(..) function)). Also I don't clearly understand what #ifdef CONFIG_ACPI do. But I have no samsung-laptop-acpi related messages in logs. Diff for samsung-laptop.c: http://zalil.ru/34466753 Some related links: https://bugs.launchpad.net/ubuntu/+s...ux/+bug/810093 https://lists.ubuntu.com/archives/ke...er/018086.html -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
Kernel has been updated recently to 3.7.10-1.4, no changes in samsung-laptop. And patch still works. 2013/4/26 Денис Купляков <dener.kup@gmail.com>:
Hello all!
Fn + F2(down) and Fn + F3(up) don't change brightness level of backlight on my Samsung NP300V5A laptop. 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)
With no special setup, I can't lower brightness (screen is flickering), and have 3 entries under /sys/class/backlight: Code:
acpi0_backlight acpi1_backlight intel_backlight
Haven't tried to manipulate them through console.
With acpi_backlight=vendor, I can't up brightness, and have 2 entries: Code:
intel_backlight samsung
Changing brightness with intel_backlight through console seems OK, but changing it with samsung don't work, I can set only 0 or 1 brightness level(same with Fn-##) Code:
dmesg | grep samsung [ 16.423716] samsung_laptop: enabled workaround for brightness stepping quirk [ 16.424060] samsung_laptop: detected SABI interface: SwSmi@
I have solved problem for me, by deleting strange workaround from samsung-laptop.c and recompiling module.
I think that it is a bug. I can see that we have some notebooks models hardcoded in module for quirk workaround (there is no my model). So I think that actual bug is inside code that determines whether workaround is needed or not (it is check_for_stepping_quirk(..) function)).
Also I don't clearly understand what #ifdef CONFIG_ACPI do. But I have no samsung-laptop-acpi related messages in logs.
Diff for samsung-laptop.c: http://zalil.ru/34466753
Some related links: https://bugs.launchpad.net/ubuntu/+s...ux/+bug/810093 https://lists.ubuntu.com/archives/ke...er/018086.html -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
On Tue, May 14, 2013 at 10:46:05PM +0400, Денис Купляков wrote:
Kernel has been updated recently to 3.7.10-1.4, no changes in samsung-laptop. And patch still works.
2013/4/26 Денис Купляков <dener.kup@gmail.com>:
Hello all!
Fn + F2(down) and Fn + F3(up) don't change brightness level of backlight on my Samsung NP300V5A laptop. 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)
With no special setup, I can't lower brightness (screen is flickering), and have 3 entries under /sys/class/backlight: Code:
acpi0_backlight acpi1_backlight intel_backlight
Haven't tried to manipulate them through console.
With acpi_backlight=vendor, I can't up brightness, and have 2 entries: Code:
intel_backlight samsung
Changing brightness with intel_backlight through console seems OK, but changing it with samsung don't work, I can set only 0 or 1 brightness level(same with Fn-##) Code:
dmesg | grep samsung [ 16.423716] samsung_laptop: enabled workaround for brightness stepping quirk [ 16.424060] samsung_laptop: detected SABI interface: SwSmi@
I have solved problem for me, by deleting strange workaround from samsung-laptop.c and recompiling module.
I think that it is a bug. I can see that we have some notebooks models hardcoded in module for quirk workaround (there is no my model). So I think that actual bug is inside code that determines whether workaround is needed or not (it is check_for_stepping_quirk(..) function)).
Also I don't clearly understand what #ifdef CONFIG_ACPI do. But I have no samsung-laptop-acpi related messages in logs.
Diff for samsung-laptop.c: http://zalil.ru/34466753
Some related links: https://bugs.launchpad.net/ubuntu/+s...ux/+bug/810093 https://lists.ubuntu.com/archives/ke...er/018086.html
Any reason you haven't pushed this kernel patch to the upstream developers? That's the best way to get this resolved. thanks, greg k-h -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
Can you assist me, I don't know where should I post it. 2013/5/14 Greg KH <greg@kroah.com>:
On Tue, May 14, 2013 at 10:46:05PM +0400, Денис Купляков wrote:
Kernel has been updated recently to 3.7.10-1.4, no changes in samsung-laptop. And patch still works.
2013/4/26 Денис Купляков <dener.kup@gmail.com>:
Hello all!
Fn + F2(down) and Fn + F3(up) don't change brightness level of backlight on my Samsung NP300V5A laptop. 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)
With no special setup, I can't lower brightness (screen is flickering), and have 3 entries under /sys/class/backlight: Code:
acpi0_backlight acpi1_backlight intel_backlight
Haven't tried to manipulate them through console.
With acpi_backlight=vendor, I can't up brightness, and have 2 entries: Code:
intel_backlight samsung
Changing brightness with intel_backlight through console seems OK, but changing it with samsung don't work, I can set only 0 or 1 brightness level(same with Fn-##) Code:
dmesg | grep samsung [ 16.423716] samsung_laptop: enabled workaround for brightness stepping quirk [ 16.424060] samsung_laptop: detected SABI interface: SwSmi@
I have solved problem for me, by deleting strange workaround from samsung-laptop.c and recompiling module.
I think that it is a bug. I can see that we have some notebooks models hardcoded in module for quirk workaround (there is no my model). So I think that actual bug is inside code that determines whether workaround is needed or not (it is check_for_stepping_quirk(..) function)).
Also I don't clearly understand what #ifdef CONFIG_ACPI do. But I have no samsung-laptop-acpi related messages in logs.
Diff for samsung-laptop.c: http://zalil.ru/34466753
Some related links: https://bugs.launchpad.net/ubuntu/+s...ux/+bug/810093 https://lists.ubuntu.com/archives/ke...er/018086.html
Any reason you haven't pushed this kernel patch to the upstream developers? That's the best way to get this resolved.
thanks,
greg k-h -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
On Tue, May 14, 2013 at 11:19:50PM +0400, Денис Купляков wrote:
Can you assist me, I don't know where should I post it.
First off, where is the patch itself? I can't seem to get any of the links you provided below to work. Can you just send it to this mailing list? thanks, greg k-h -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
I can't say it is a patch, it is fix for my notebook model. It simply comments workaround in code, so it will break functionality on some notebooks. PS: link provided at first post still works for me. 2013/5/14 Greg KH <gregkh@linux.com>:
On Tue, May 14, 2013 at 11:19:50PM +0400, Денис Купляков wrote:
Can you assist me, I don't know where should I post it.
First off, where is the patch itself? I can't seem to get any of the links you provided below to work. Can you just send it to this mailing list?
thanks,
greg k-h
A: No. Q: Should I include quotations after my reply? http://daringfireball.net/2007/07/on_top On Wed, May 15, 2013 at 12:04:26AM +0400, Денис Купляков wrote:
I can't say it is a patch, it is fix for my notebook model. It simply comments workaround in code, so it will break functionality on some notebooks.
Then a patch like that could never be accepted. You can't break someone else's working machines, right?
PS: link provided at first post still works for me.
What exact link would that be? thanks, greg k-h -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
2013/5/15 Greg KH <gregkh@linux.com>:
A: No. Q: Should I include quotations after my reply?
OK, will use it in future :)
Then a patch like that could never be accepted. You can't break someone else's working machines, right?
I don't pretend to post a patch now, I simply trying to make it. Can you tell me what is difference between brightness and backlight in terms of module? Or is there any documentation for samsung-laptop.c code and conception?
PS: link provided at first post still works for me.
What exact link would that be?
It is http://zalil.ru/34466753, I know that posting attachments allowed so it doesn't make sense now.
thanks,
greg k-h -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
I have done more investigation using debugfs commands, it seems for me that 0x11 command (set_brightness) works very unusual. There is a comment in code about quirk workaround for some notebook models. It says that if you will set brightness to/from 0 all work like real "set" function. But it is false for me, in spite of that my notebook have stepping mode. I'm sending you attachment with console output and some comments from me. I propose to rewrite set_brightness to something like this: while (don't have need brightness && number_of_iterations <= MAX_BRIGHTNESS) { change brightness } if (don't have need brightness) { some error found } It will work good for all notebook models (step-mode with 0-set, step-mode(like me) and set-mode) and it will be possible to remove stepping quirk workaround at all. Do you know other behaviors for set command? Best regards, Denis Kuplyakov.
On Wed, May 15, 2013 at 05:58:58PM +0400, Денис Купляков wrote:
I have done more investigation using debugfs commands, it seems for me that 0x11 command (set_brightness) works very unusual. There is a comment in code about quirk workaround for some notebook models. It says that if you will set brightness to/from 0 all work like real "set" function. But it is false for me, in spite of that my notebook have stepping mode. I'm sending you attachment with console output and some comments from me.
I propose to rewrite set_brightness to something like this:
while (don't have need brightness && number_of_iterations <= MAX_BRIGHTNESS) { change brightness }
if (don't have need brightness) { some error found }
It will work good for all notebook models (step-mode with 0-set, step-mode(like me) and set-mode) and it will be possible to remove stepping quirk workaround at all. Do you know other behaviors for set command?
Best regards, Denis Kuplyakov.
I don't know the behaviors at all, sorry. I suggest contacting the upstream maintainer of the driver about this, as they are the one that can fix this the best for you. The name of the maintainer and email address is in the MAINTAINERS file in the kernel source tree. Hope this helps, greg k-h -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
participants (3)
-
Greg KH
-
Greg KH
-
Денис Купляков