On Friday, September 9, 2016 5:55:30 AM CEST you wrote: > On Friday, September 9, 2016 3:38:44 AM CEST you wrote: > > On Thursday, September 8, 2016 9:55:40 PM CEST you wrote: > > > On Thursday, September 8, 2016 7:00:42 PM CEST you wrote: > > > > I just did a "Aptitude GUI" testing upgrade (aptitude -R -t stretch); > > > > Then I used command line: > > > > aptitude search -F"%p; %v; %V; %d" -t sid > > > > '?installed(?maintainer(debian-qt- kde))'|egrep 5.7|egrep > > > > '5.6|5.7.0'|less -i (goal here is empty output) to see what packages I > > > > should also upgrade from sid > > > > (Which I did after recommendations I've seen on this list); > > > > which I've upgraded using the same GUI way (aptitude -R -t sid). > > > > I had one glitch with "qml-module-qtmultimedia (5.6.1-2 and others)". > > > > > > > > After reboot I can't see anything unusual. > > > > > > Then After a While the screen got locked: it didn't seem unusual at the > > > time even though the duration after which it occurred was unusually > > > short... Then when I came back to my computer it was in a suspend to ram > > > condition. The computer is a closed lid laptop with external monitor and > > > of course keyboard... > > > It is always on AC power. > > > I join a screenshot of the relevant settings screen that shows, I > > > believe, > > > that everything is as it should be. And that therefore the computer > > > shouldn't have behaved as it did. > > > Maybe it won't happen again. > > > > It did happen again. So what I did is tick then untick the "on AC power" > > suspend session thing in settings/power management before clicking to > > apply. > > > > I'll see what comes from it. > > > > (I've got lock screen automatically after 60 min set in Desktop Behavior - > > Screen locking timeout... > > There is some redundancy in settings there. > > Also I wonder if the right thing to do wouldn't be to tick the suspend > > session thing in power management and specifying that the thing to do > > would > > then be to just lock the screen; maybe if it's just not ticked there is a > > default behavior consisting in suspend to ram... > > All that pure speculations I'm afraid.) > > I've found more precisely what causes the suspend to ram thing: it is when I > use the power button of the external monitor to turn it off. > However my settings say: "button event handling > when laptop lid closed > > do nothing", and "even when an external monitor is connected", that last > one is unticked. > > I tried to tick the "even when external monitor is connected" thing, so that > the semantic would be: do nothing, even in that specific case; but the > outcome was just the same. > > So finally I've got so far not solution at all but allowing the laptop to > suspend to ram, and reach it to open the lid temporarily, when I want to use > it again. Which is definitely not a solution. > > It might be important in order to reproduce the bug, the external monitor is > connected to the laptop (Lenovo x230) through "mini displayport". > > There are 2 upstream bugs: > https://bugs.kde.org/show_bug.cgi?id=361022 > https://bugs.kde.org/show_bug.cgi?id=359142 > But none seem directly related, because mine is specifically "When I power > off the external monitor". > > It seems very very bad.
Here is the workaround: Get rid of Powerdevil altogether. Get rid of acpid just in case. Get rid of anything else if it's relevant. No need to remove Upower or acpi. Then I got help from #debian-next: Edit /etc/systemd/logind.conf and make HandelLidSwitch=ignore (remove the comment) Then 'systemctl restart systemd-logind' Ta-da! I rebooted to see if it was resilient... >From this point it might be that it'd still work if reinstalling Powerdevil... But for the present it fits my needs. > > > > > Chris
