Results 1 to 10 of 10
I recently upgraded to Debian unstable and Linux kernel 22.214.171.124 in order to get my new Wacom tablet working. (Needed a new version of X.org in order to build the ...
Enjoy an ad free experience by logging in. Not a member yet? Register.
- 01-03-2011 #1
EEE901 power management problem after upgrade
If I suspend the machine by closing the lid or running pm-suspend, the screen goes black but it doesn't appear to go into suspend mode (i.e. no blinky lights) - and I can't wake it up again, either.
Something similar happens when I'm running on battery and plug in the AC power - laptop suddenly goes black.
I know a lot about a lot, but not much about laptop power management... So I'm kind of stuck here. Any help would be greatly appreciated.
- 01-03-2011 #2
- 01-03-2011 #3
All the power management stuff worked before the upgrade - but so much changed in the upgrade (kernel and lots of software) that it's hard to guess at exactly what's gone wrong.
I guess there's a fair chance that something went wrong with my kernel build, maybe - as far as I can tell the power management scripts just call an ioctl on a file in /sys in order to initiate suspend - if that's true then it seems a very strong indication that the kernel is to blame somehow. But I'm a bit averse to rebuilding the kernel as a way to try and fix things (unless I'm sure it's the problem) because it takes a long time to build the kernel on this machine... (I would build it on my desktop machine, but it's running a 64-bit system and last time I tried a cross-compile it didn't work right...)
But then, plugging in AC is causing problems, too... That shouldn't trigger much other than a notification on dbus or something... I don't really get what's going on and I don't really know how to figure it out - given that the system apparently crashes when I do anything PM-related...
- 01-03-2011 #4All the power management stuff worked before the upgrade - but so much changed in the upgrade (kernel and lots of software) that it's hard to guess at exactly what's gone wrong. I guess there's a fair chance that something went wrong with my kernel build
I used a 2.6.36 Liqourix kernel instead to fix it (uninstalled aptosid 2.6.36). I use smxi script for debian upgrading and kernel upgrades. That kernel works just fine on my EEE 701SD as far as power management and hotkeys where the aptosid kernel would not.
smxi Manual :: Configuration and Operation Information
- 01-05-2011 #5
Still don't know what went wrong... Best guess is something went wrong with my build or installation of the kernel... But as luck would have it, kernel version 2.6.37 has the version of the Wacom driver that I needed, so probably I'll grab a kernel package from the Debian experimental repo and see if that solves my problem.
- 01-06-2011 #6
My kernel on my 701SD which has been running for hours.
Kernel 2.6.36-2.dmz.5-liquorix-686 i686 Sensors: System Temperatures: cpu: 52.0C Fan Speeds (in rpm): cpu: 1390
- 01-06-2011 #7
Something else has gone wrong - now any time I try to boot into my build of 2.6.36 or the Debian build of 2.6.32, the machine crashes with a black-screen sometime during the boot procedure... I don't understand what happened. I was able to boot the thing just fine on Monday... <sigh> I still have 2.6.26 installed so I can boot into that, but that's not providing any insight into why the newer kernel versions aren't booting properly... Maybe there's a piece of software on the system that's crashing the machine when run under 2.6.3x and refusing to run or something when run on 2.6.26...
I think I installed some new packages or something on Monday, but I can't think of what I did... I think I spent some time trying to solve my power management problems, but I don't remember what I tried specifically... <sigh>
- 01-06-2011 #8Debian build of 2.6.32, the machine crashes with a black-screen sometime during the boot procedure... I don't understand what happened. I was able to boot the thing just fine on Monday... <sigh>
I still have 2.6.26 installed so I can boot into that,
- 01-06-2011 #9
I guess I could mention. If I was in Your shoes. Instead of reinstall. I would use Synaptic or apt-get to delete all kernels (headers and image) but the one that can boot. Then hook up land line and run smxi (which I run) to get a list of available kernels. Pick and choose till I find one that works. Delete the ones that don't work. But that is just me.
- 01-06-2011 #10
Next up, I guess, is to move away the pm-utils stuff, on the assumption that something, somewhere is trying to call that crap during startup...
(EDIT): I did dpkg --purge acpi-support eeepc-acpi-scripts pm-utils, and I was able to boot in to 2.6.36 (and use wireless) again. Then I reinstalled those packages via apt and that seems to have cleared things up... So I guess it was some kind of problem with the software upgrade, maybe I kept a config file or two that I should have overwritten...
shutdown still doesn't work right, but suspending works again...
(EDIT): ...and, it failed again. There must be something wrong with the ACPI stuff in Debian Unstable at this point. I tried changing the EEE hotkey config and reloading it, and the system blew up... So I removed/reinstalled the ACPI stuff again and this time around, I could boot the system but it blew up as soon as GNOME started (because of the gnome power manager, no doubt...)
I'll have to see if I can get working ACPI stuff for my kernel version from a slightly more stable version of Debian... Or maybe they'll fix the problem in the unstable repo soon. Or maybe there's still something mysteriously wrong with some other piece of software on my system... I don't know. <sigh>
(EDIT): OK, it was eeepc-acpi-scripts. I removed it and things are working great now. I don't know why specifically it was causing the machine to crash, but the package is apparently entirely unnecessary if you've got a current version of acpid and acpi-support installed. All the hotkeys (well, with the exception of USR3 and USR4 anyway) work as expected.
And damned if I know why... The rules in /etc/acpi/events apparently don't handle the EEEPC hotkeys or the generic functions they represent... I guess, maybe, it's a combination of key remapping done by the eeepc_laptop kernel module along with default rules present in acpid itself? Or maybe I'm just not understanding the acpi-support rule set as well as I think...
Last edited by tetsujin; 01-07-2011 at 05:30 AM.