KDE 5 power management doesn't seem to work at all

English Applications
1

No matter what I do in Power Management module in KDE settings to change “dim screen”, “screen energy saving” etc nothing seems to be applied to the system at all.

There’s an old, I presume, ~/.kde4/share/config/powermanagementprofilesrc file, it gets ignored and tweaking power management module doesn’t write to it anymore and I don’t know where else to look.

I’ve tried manually playing with “xset dpms 180 300 600” and various possibilities there, it sort of works but after a while the screen comes back from dim/sleep and turns back on.

What’s worse is that something spams the system with some sort of keyboard input on wake up and even when the screen is on and I’m working. It doesn’t put any actual letters out but I’ve noticed that if I have some process running in konsole it would appear as if somebody was hitting the keyboard repeteadly and konsole, not waiting for actual input, shows a weird combination of */ or some such. Like if zypper is refreshing repos and waiting for repomod file to download the spinner gets interrupted and screen gets filled with this repeated garbage. In other programs it might appear as if a button gets hit over and over again, in browser address field might get focused and start blinking like crazy, and does not respond to any clicks.

What stops this is hitting Esc on keyboard, or even simply turning keyboard on and off. I thought I might have some sticky key but no, can’t find anything unusual there.

In /var/log/messages I constantly get things like:

2016-02-24T00:27:15.360372+07:00 linux-pwfe org.kde.KScreen[1158]: message repeated 199 times:  ks
creen: Primary output changed from KScreen::Output(Id: 69 , Name: "VGA1" ) ( "VGA1" ) to KScreen::O
utput(Id: 69 , Name: "VGA1" ) ( "VGA1" )]

These messages seem to correspond with my mini freezes and are usually followed by reports of CPU overheating. Here’s a longer chunk from messages:

2016-02-24T00:27:15.360372+07:00 linux-pwfe org.kde.KScreen[1158]: message repeated 199 times:  ks
creen: Primary output changed from KScreen::Output(Id: 69 , Name: "VGA1" ) ( "VGA1" ) to KScreen::O
utput(Id: 69 , Name: "VGA1" ) ( "VGA1" )]
2016-02-24T00:30:01.601292+07:00 linux-pwfe cron[18291]: pam_unix(crond:session): session opened fo
r user root by (uid=0)
2016-02-24T00:30:01.611777+07:00 linux-pwfe systemd: pam_unix(systemd-user:session): session opened
 for user root by (uid=0)
2016-02-24T00:30:01.641745+07:00 linux-pwfe CRON[18291]: pam_unix(crond:session): session closed fo
r user root
2016-02-24T00:30:01.650163+07:00 linux-pwfe systemd: pam_unix(systemd-user:session): session closed
 for user root
2016-02-24T00:31:22.626348+07:00 linux-pwfe org.kde.KScreen[1158]: kscreen: Primary output changed 
from KScreen::Output(Id: 69 , Name: "VGA1" ) ( "VGA1" ) to KScreen::Output(Id: 69 , Name: "VGA1" ) 
( "VGA1" )
2016-02-24T00:33:33.412981+07:00 linux-pwfe kernel: [53593.514161] CPU0: Core temperature above thr
eshold, cpu clock throttled (total events = 31)
2016-02-24T00:33:33.413011+07:00 linux-pwfe kernel: [53593.514172] CPU4: Core temperature above thr
eshold, cpu clock throttled (total events = 31)
2016-02-24T00:33:33.413012+07:00 linux-pwfe kernel: [53593.514173] CPU5: Package temperature above 
threshold, cpu clock throttled (total events = 40)
2016-02-24T00:33:33.413012+07:00 linux-pwfe kernel: [53593.514174] CPU1: Package temperature above 
threshold, cpu clock throttled (total events = 40)
2016-02-24T00:33:33.413013+07:00 linux-pwfe kernel: [53593.514175] CPU3: Package temperature above 
threshold, cpu clock throttled (total events = 40)
2016-02-24T00:33:33.413014+07:00 linux-pwfe kernel: [53593.514176] CPU7: Package temperature above 
threshold, cpu clock throttled (total events = 40)
2016-02-24T00:33:33.413024+07:00 linux-pwfe kernel: [53593.514178] CPU4: Package temperature above 
threshold, cpu clock throttled (total events = 40)
2016-02-24T00:33:33.413026+07:00 linux-pwfe kernel: [53593.514179] CPU2: Package temperature above 
threshold, cpu clock throttled (total events = 40)
2016-02-24T00:33:33.413027+07:00 linux-pwfe kernel: [53593.514180] CPU6: Package temperature above 
threshold, cpu clock throttled (total events = 40)
2016-02-24T00:33:33.413028+07:00 linux-pwfe kernel: [53593.514185] CPU0: Package temperature above 
threshold, cpu clock throttled (total events = 40)
2016-02-24T00:33:33.414880+07:00 linux-pwfe kernel: [53593.519190] CPU0: Core temperature/speed nor
mal
2016-02-24T00:33:33.414887+07:00 linux-pwfe kernel: [53593.519191] CPU4: Core temperature/speed nor
mal
2016-02-24T00:33:33.414888+07:00 linux-pwfe kernel: [53593.519192] CPU7: Package temperature/speed 
normal
2016-02-24T00:33:33.414888+07:00 linux-pwfe kernel: [53593.519203] CPU1: Package temperature/speed 
normal
2016-02-24T00:33:33.414889+07:00 linux-pwfe kernel: [53593.519204] CPU3: Package temperature/speed 
normal
2016-02-24T00:33:33.414889+07:00 linux-pwfe kernel: [53593.519205] CPU5: Package temperature/speed 
normal
2016-02-24T00:33:33.414891+07:00 linux-pwfe kernel: [53593.519206] CPU4: Package temperature/speed 
normal
2016-02-24T00:33:33.414891+07:00 linux-pwfe kernel: [53593.519208] CPU2: Package temperature/speed 
normal
2016-02-24T00:33:33.414892+07:00 linux-pwfe kernel: [53593.519209] CPU6: Package temperature/speed 
normal
2016-02-24T00:33:33.414893+07:00 linux-pwfe kernel: [53593.519209] CPU0: Package temperature/speed 
normal
2016-02-24T00:33:54.474401+07:00 linux-pwfe org.kde.KScreen[1158]: message repeated 127 times:  ks
creen: Primary output changed from KScreen::Output(Id: 69 , Name: "VGA1" ) ( "VGA1" ) to KScreen::O
utput(Id: 69 , Name: "VGA1" ) ( "VGA1" )]

This leads me to think that it’s KDE’s power management that is the underlying culprit behind it all and it’s getting serious. It looks as if it’s kernel or system itself that corrects runaway KScreen spam several times an hour.

Any ideas? Disable power management from startup services altogether? But I still need the monitor to dim and turn off.

It’s a desktop, btw, the video card is integrated Intel and the monitor is an old LCD with VGA connection, no multiple monitors. After boot dpms settings don’t take effect unless I run them explicitly even if quering them shows they are always on. I don’t know how xset and dpms relate to KDE, using dpms to force monitor to sleep is just something I saw on the internet.

2

Reading it all, I suspect a hardware issue, but have no idea what. A (partly) broken VGA cable or connector, motherboard dying, keyboard internals broken.

I don’t suspect the KDE powermanagement module, but the Plasma5 config files live in ~/.config

3

I turned off Power Management in Background Services in Startup/Shutdown and minifreeze issue went away. I still get “KScreen sent 32 messages about changing monitor” log entries, though, but without any visual impairment to the machine. They do not lead to overheating and they appear rather rarely. Is there a better log to monitor this kind of events than /var/log/messages?

Without power management module on, the monitor goes to sleep and turns off according to dpms settings in “xset q”. Somehow all three values (standby, suspend, off) are set at 600 seconds and I need to issue my desired values manually. On reboot they get overwritten again and I don’t know where xset reads this settings from. I don’t have ~/.xinitrc or similar files in my home directory.

Initially I suspected a dying battery in keyboard but now it’s clearly not an issue. VGA cable connectors are fastened with screws and if the cable was faulty itself I’d immediately see deterioration in picture quality (because it’s analog, not digital). I’ve had faulty cables before, this one looks perfect. The machine itself is less than two years old so motherboard should not be dying. I still use VGA because I just hasn’t gotten around to buying a new monitor for it.

Settings for power management are indeed in ~/.config/powermanagementprofilesrc, thanks, that’s where all Plasma5 files are! I recognize my changes there, but, as I said in the OP, they don’t have any affect on the machine (when the module was on). No matter what I set there nothing was recognized and the machine even developed these freeze issue.

I’d rather have Plasma module working as intended, messing with dpms is just a workaround.