Hello,

Got a perplexing problem. Seems when I resume from suspend, my LCD backlight doesn't come on. If I tilt my screen just right I can see what's there... but... this is a little silly.

I'm running Fedora Core 6 x86_64, with the latest stock kernel (2.6.20-1.2933.fc6). My laptop is a spankin' new Asus z35f with a Core 2 Duo. It's got an intel 945g video card. Basically this is the System76 Darter, but I got it from someone else who sold it cheaper (and sold me the gray version; can't do the white PC thing). I'm suspending to ram via the KDE logoff menu. I tried dowloading a special kernel for suspend2 from ATRpms, but it has the same problem.

Has anybody else fought this problem and won? I'd love to hear how... I'm still going at it, if it cracks I'll send some cheers this way.

What happens if you switch to TTY1 (ALT+CTRL+1) and back to TTY7 (ALT+CTRL+7)? Does it still do the same?

Sorry for the late reply... been busy weekending.

Switching to TTL1 and back to 7 has no effect whatsoever.

I'm starting to think this is some sort of vbetool bug? Not really sure. I'm looking at scripts in /etc/pm/, particularly functions-intel. There's a line "/usr/sbin/vbetool dpms post" which is supposedly simulating the VGA commands sent at boot; from the manpages it sounds like it has problems with laptops. Trying to play around with the different options and see if one works; no major progress to report yet, though.

Hi,

I would recommend not using software suspend if you can help it. It's not very reliable in many linux distributions yet. Hibernate is reported to be more reliable.

It may be that the software isn't initializing your backlight inverter correctly. When a laptop of mine experienced the same symptoms (except our symptoms were permanent), we traced the problem to a failed inverter chip.

They can be expensive to purchase, and I don't know if continuing to experiment with yours might make the problem permanent.

Just a word of caution. I hope that isn't the case, and that it's a simple software problem.

Sam

Thanks for the warning on the inverter.

You know, I feel like a bit of a fool; I was looking for the hibernate command to be "hibernate", turns out it's "pm-hibernate". I'd prefer hibernate anyway; now that it works (minus some wireless junk; nothing surprising) I'm happy to stick with it! Now to test my acpid event to hibernate when the lid's closed...