Thanks
Back to top

I installed the 3.2 kernel from the future repository. When I boot the resolution changes to what appears to be the correct resolution. This is new and initially gave me hope. But about halfway through the boot sequence, the screen goes blank. At that point syslog and kern.log stop recording. No error is reported.

The funny thing is that I can ssh into the system from an external computer, login and do anything I need to do from a terminal.
Back to top

I just noticed the following error in my kern.log


I suspect now that this is the problem.
Back to top

Some good news. I have achieved a perfect screen setup. But the screen still goes blank on boot. If I let it sit for several minutes it comes on. What I have observed is that the screen is not consistently blank. It appears to blink ever so faintly. And it does eventually turn on. It is as though some process is looping through various settings until it finally gets it right. I have looked in syslog and dmesg.log but haven't been able to spot what's going on yet. I feel that I'm getting close to a permanent solution.
Back to top

Thanks to Daniel Vetter & Co. over at intel-linux-graphics. They finally fixed my Sandy Bridge backlight problem or perhaps I should say problems. You can find the discussion and progress at bug 48435. The fix is in the 3.5-rc2+ kernel that can be found here at the bottom of the page:


Back to top

The patch listed in the bug report should be available on 3.4.0-3.dmz.2 and newer.
Back to top

Thanks, damentz. I'll try it out.
Back to top

I just installed the 3.4.0-3.dmz.2 kernel. Back light is still off on my machine. But it works on the 3.5-rc2+ kernel I compiled from the source given.
Back to top

Ok, well either that's the wrong bug, the patch has more dependencies, or the patch is simply incorrect.

I won't be backporting the drm for i915. We've done that in the past with zen-kernel and it always resulted in unpredictable behavior and broken graphics.

EDIT: On the bug report, they're still testing their stable backport to v3.2. I'll be watching for any successful patches. So far they're not having any good luck.
Back to top

Jonathan Nieder is a Debian kernel developer. It was he who first introduced me to Daniel Vetter and the intel linux graphics group. Both he and Daniel assisted me in weeding through the tasks of building and testing kernel patches. Daniel did the patch development. When he finally got a combination that proved a success on my sandy bridge machine, Jonathan picked up the torch, so to speak, hoping to bring the fix to the stable debian kernel. Since I had the machine that was broke and then fixed, Jonathan asked me to test a patch he had gotten from Daniel. Unfortunately it did not work. I suspect that the fix is more complicated than one patch and is probably due to many patches applied over the past several months while they worked on the bug.

In case you haven't guessed, I'm Joel the author of the bug report.
Back to top