I have this bug on Ubuntu Intrepid Ibex with intel driver 2.4.1, every time I open a movie with vlc or mplayer the screen darkens and brightness is set to 0. I have a GMAX3100 on a Hasee laptop, I had a similar problem with driver 2.2.2 on Hardy Heron, the difference was that brightness was reset to xbacklight value, and I was able to workaround this bug by setting xbacklight to 100 on startup. But on Intrepid Ibex this doesn't seem to work anymore! I guess this is a different bug. Any help would be much appreciated!
please provide necessary information as required on http://www.intellinuxgraphics.org/how_to_report_bug.html.
We are not a OS Distirbutor's bugzilla, so just tell a Ubuntu code name isn't enough.. thanks!
You're right, sorry these are the required infos:
-- chipset: i965
-- system architecture: i686
-- xf86-video-intel version: 2:2.4.1-1ubuntu6
-- drm version: 2.3.1-0build1
-- mesa version: 7.2
-- xserver version: 7.4
-- kernel version: 2.6.27-6-generic
-- Distribution: Ubuntu 8.10
-- Machine: Hasee W230R laptop
-- Display connector: VGA (?)
How do I obtain a Xorg.0.log file?
Created attachment 19524 [details]
Created attachment 19525 [details]
Is there any additional info needed? Is there anybody working on this?
Does this still occur with the git bits of the driver?
When I find a .deb of intel driver 2.5.0 I'll try it out, I've no idea on how I could compile from git..
I remember zhenyu fixed a very similar issue.. re-assign to him for comment.
Please try current 2.4.3 to see if it fixes your problem. Could you paste X log and xrandr --verbose output?
Michael, this looks like not the bug we met before, as that bug showed brightness was set to max value.
Hi, I'd like to try it out, but I don't know how. Wouldn't it be easier for everybody if you made a .deb package for every release of the driver? You would have ten times the people trying and testing each release. I'm sorry, I don't have any idea of how I can try it: can you address me to some site where there is a clear explanation of how I can compile it and try it? I'm no noob, not a geek either but I can do things like compiling ect.
Thanks a lot
Ok, I could not find a .deb package of 2.4.3, but I found and installed the latest 2.5.1, and I can confirm that the bug is still there. I'm not the only one with this bug have a look here:
Could you tell me the command line you use for vlc and mplayer? I can't produce this problem on T61 here.
In the link I posted, it seems that this bug occurs with just some (rev 03) integrated intel cards only. My card is:
lspci | grep VGA:
00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 03)
I don't use any particular command line with mplayer or vlc, just open any kind of movie file. I use xvideo as video output on both. Totem, which is based on gstreamer, is not affected by this bug. Also some OpenGL apps are affected, like hedgewars, openarena and some others. They reset brightness.
It's the exact same apps and players that were affected by the xbacklight bug, but now xbacklight -set 100 does not work as a workaround anymore.
could this be a player version? Alessio, what's your mplayer version? And, Zhenyu, are you using Ubuntu Intrepid Ibex? I have a liveCD..
(In reply to comment #14)
> could this be a player version? Alessio, what's your mplayer version? And,
> Zhenyu, are you using Ubuntu Intrepid Ibex? I have a liveCD..
Hi, I don't think it's a player version bug, beacause as I said also some opengl apps and vlc do the same, and they have nothing to do with mplayer. However, mplayer version is 1.0 rc2. This bug is present in Fedora Core 10 also.
I noticed a strange behaviour on system startup: I turn on the laptop, and brightness is max. Then Ubuntu (or Fedora) starts loading, and brightness goes down. After the login screen, while the desktop is loading, brightness goes max again. I'm dual-boot with win xp: when I boot win xp, brightness stays max all the time.
To report backlight related issue, you should provide X log with ModeDebug option on, and "xrandr --verbose" output, and dmesg log after you've seen wrong backlight behavior. This smells like a platform acpi video backlight control issue. Please provide the info, thanks!
It seems you are right, the dmesg output has these lines, right after playing a a movie with VLC:
[ 382.536060] ACPI: EC: missing confirmations, switch off interrupt mode.
[ 383.037100] ACPI Exception (evregion-0419): AE_TIME, Returned by Handler for [EmbeddedControl] 
[ 383.037128] ACPI Error (psparse-0530): Method parse/execution failed [\_SB_.PCI0.LPCB.EC0_.BAT0._BST] (Node f7418f00), AE_TIME
[ 383.037219] ACPI Exception (battery-0360): AE_TIME, Evaluating _BST 
[ 584.853181] ACPI Error (utobject-0508): Unsupported Reference opcode=88 in object f116e460 
I'm attaching the files you requested, thanks for your help!
(In reply to comment #16)
> To report backlight related issue, you should provide X log with ModeDebug
> option on, and "xrandr --verbose" output, and dmesg log after you've seen wrong
> backlight behavior. This smells like a platform acpi video backlight control
> issue. Please provide the info, thanks!
Created attachment 21202 [details]
xrandr verbose ouput
Created attachment 21203 [details]
dmesg output after VLC
Created attachment 21205 [details]
xorg0.log with modedebug on in xorg.conf
Disabling DPMS option for Monitor in xorg.conf works for me.
(In reply to comment #21)
> Disabling DPMS option for Monitor in xorg.conf works for me.
Great!! It works for me as well!! It would be nice to have this bug solved but this workaround is perfect! Thank you!!
(In reply to comment #23)
Ok, maybe compiling mplayer from source solves the issue for mplayer without the need of disabling DPMS, but there's too many apps with this issue: virtualbox, VLC, hedgewars, xine, openarena just to mention some. It think it should be solved in the graphic driver, if possible, or in the ACPI driver. Also, with DPMS disabled, I have screen brightness set to max after resuming from suspend (well, not every time, it's a bit random), so I will keep it disabled for now. Thanks again!
cc Yakui for ACPI EC issue.
And I think this can be work around by using another backlight control method like "combination", xrandr --output LVDS --set BACKLIGHT_CONTROL combination will do that.
The following warning message is related with EC.
>382.536060] ACPI: EC: missing confirmations, switch off interrupt mode.
>[ 383.037100] ACPI Exception (evregion-0419): AE_TIME, Returned by Handler for[EmbeddedControl] 
>[ 383.037128] ACPI Error (psparse-0530): Method parse/execution failed
[\_SB_.PCI0.LPCB.EC0_.BAT0._BST] (Node f7418f00), AE_TIME
>[ 383.037219] ACPI Exception (battery-0360): AE_TIME, Evaluating _BST
Will you please try the latest kernel(2.6.28-rc8) and see whether the above issue still exists?
Of course please attach the output of acpidump. The latest acpidump tool(pmtool-20071116) can be found in http://www.kernel.org/pub/linux/kernel/people/lenb/acpi/utils/
Alessio, any update on this?
Sorry I was just waiting for the next Ubuntu Alpha Release, but maybe this weekend I'll look for another distro with this kernel and update my results, sorry had no time these days because of my job, I'll be back on sunday! Thanks for your patience!
(In reply to comment #27)
> Alessio, any update on this?
(In reply to comment #28)
> Sorry I was just waiting for the next Ubuntu Alpha Release, but maybe this
> weekend I'll look for another distro with this kernel and update my results,
> sorry had no time these days because of my job, I'll be back on sunday! Thanks
> for your patience!
> (In reply to comment #27)
> > Alessio, any update on this?
ping again...especially for ykzhao's comment in comment# 26..
Created attachment 22495 [details]
dmesg output in kernel 2.6.28
Created attachment 22498 [details]
acpidump in kernel 2.6.28
Created attachment 22499 [details]
acpidump in kernel 2.6.28
Created attachment 22500 [details]
acpidump in kernel 2.6.28
Hi, I uploaded the dmesg and acpidump outputs in Ubuntu 9.04 Alpha 3 with Kernel 2.6.28. Has I said before, brightness control is not working anymore, it is always at the lowest level, no way to turn it up. Sorry for triple postings but freedesktop.org was returning me an error after each time. I you need anything else just let me know, and thanks for your efforts I really appreciate!
Yakui, any clue from acpidump?
From the acpidump it seems that there exists the an error in the ACPI _BQC object:
>Method (_BQC, 0, NotSerialized)
Return (Index (BTBL, \_SB.PCI0.LPCB.EC0.BKLG))
In theory the actual brightness level should be returned. But in this object the index is returned. It is incorrect.
I will attach the custom DSDT and let the bug reporter try it.
Created attachment 22557 [details] [review]
try the custom DSDT
Will you please try the custom DSDT and see whether the issue still exists?
In the custom DSDT the actual brightness level is returned by the _BQC object instead of index field.
How to use the custom DSDT can be found in:
I just tryed the latest 2.6.28 kernel update in the latest ubuntu 9.04 alpha release and now brightness control is back and working, but I couldn't install vlc or mplayer because of a strange ubuntu error, I will keep you up to date as soon as I can sort it out. Thanks!
Alessio, does mplayer work now?
Sorry, I can't test anything until the next ubuntu alpha live cd, because I use this computer for work and I can't install beta software on it, and in the latest live cd of ubuntu alpha release I couldn't install mplayer, so I will update everything on thursday or friday I think when ubuntu 9.04 alpha 5 is out.
ubuntu jaunty alpha 5 is out. The bug should be fixed in 2.6.28. Please reopen if this issue still exists.