Registered Member
|
I do not use plasma-nm. However, I still have plasmashell using 15% of CPU just by enabling the CPU monitor and the memory monitor, which refreshes (only) about once per second. Is this an universal problem of animations in plasma 5 now?
|
Registered Member
|
This is a known, general, problem: https://bugs.kde.org/show_bug.cgi?id=345696 https://bugs.kde.org/show_bug.cgi?id=346134 But it's not universal for all animations AIUI... |
Registered Member
|
In Opensuse Thumbleweed i solve it , i just remove yahoo weather widget.
Hope to help.. |
Registered Member
|
aha,only 25% my plasmashell use 46% cpu and opening 2 or more progress with very hig cpu load .I like kde very much ,I just do not like this plasmashell anymore.
|
Registered Member
|
in fact ,if you open any window in plasma,cpu load increase to a high level, if use some widget high cpu load will sustain until close you widget.
|
Registered Member
|
I'm also experiencing plasmashell 5.4.0 stealing lots of CPU.
The problem: Right now I'm writing with plasmashell using 25%CPU, and kactivitymanagerd another 25%. I have a dual core, non-hyperthreaded CPU, so Plasma is using a full core. Plasma is completely unresponsive (fortunately KWin works fine). Sometimes a popup window manages to appear several minutes after I clicked it, but it will not render its contents. The setup: just a Haenau desktop with a minimal bar with only a system tray. I use standard widgets (nm, klipper, battery etc.). The only non-standard is the redshift widget. I'm suspecting the culprit to be the Haenau desktop. Question: As Plasma is blocked I cannot try to remove the current widgets / panel, and I don't want to remove the whole .kde directory and lose my settings. Anyone knows how to reset plasma only, or which config file can I edit manually? Terminal output: This is what plasmashell reports when started from a terminal:
Any idea? |
Registered Member
|
Hi OdinEidolon,
I'm having the exact same problem as you, and I was able to remove some plasmoids thanks to @wolfi323's help. I removed the plasmoid which I suspected to be cause for this trouble, since I installed them today and never had such problems before, but my plasmashell is still consuming a full CPU core and is totally unreponsive. The plasmoids configuration is saved here, quoting wolfi323:
Hope this helps and let me know if you still have high CPU load, because I do and it's annoying as hell. My plasma is not usable at all at the moment |
Registered Member
|
Thanks. For me it was the application dashboard plasmoid for some reason. Will file a bug. |
Registered Member
|
You are right, for me it was also the new app dashboard! Will you post the link once you filed this bug please? |
Registered Member
|
There it is: https://bugs.kde.org/show_bug.cgi?id=352649 Please fell free to edit and add information. |
Registered Member
|
Late to the party but yes I just upgraded to KDE5 and Sabayon. I see 20-25% CPU load as well until I disable my Analog Clock plasmoid. This was never an issue in KDE4. Switching back until the kinks get worked out of 5.
|
Registered Member
|
Hi!
I observe the same problem; plasmashell process having 25% CPU usage shown in systemmonitor. I'm on Fedora 22 with `plasmashell --version` reporting version 5.4.1 Trying to attach to plasmashell in gdb resulted in the following error:
Any hints on how to find out what's going on? Regards, Piotr Dobrogost |
|
you ran an update between starting plasmashell 6080 and trying to gdb attach to it.
|
Registered Member
|
Thanks. After restarting plasmashell and running gdb again I got error about missing symbols. After I installed missing symbols I'm now getting this after `tb` in gdb:
Still no traceback. I'm wondering if KDE has some page with description of all steps necessary to be able to provide needed information like traceback and others? |
Registered Member
|
These are backtraces of plasmashell when it had 20%-30% of CPU utilization according to systemmonitor. Please not that it's visible that after issuing `attach` command in gdb utilization of CPU by plasmashell dropx to a few % before it changes to "unavailable". Can this mean that backtraces do not reflect the state of plasmashell when there as high CPU utilization?
|
Registered users: bancha, Bing [Bot], Google [Bot], Sogou [Bot]