Since some days I have the problem, that "plasmashell" from plasma5-workspace-5.8.4-1.1 (latest Tumbleweed updates) freezes often (around every 30 minutes).
"top" shows then:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 4340 myuser 20 0 8881100 4.116g 158576 R 100.0 26.37 18:57.14 plasmashell
Killing plasmashell und restarting it with "plasmashell --shut-up" resolves the problem until next time.
Anyone else with this problem?
Greetings, Björn
On 01/06/2017 11:00 AM, Bjoern Voigt wrote:
Since some days I have the problem, that "plasmashell" from plasma5-workspace-5.8.4-1.1 (latest Tumbleweed updates) freezes often (around every 30 minutes).
"top" shows then:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 4340 myuser 20 0 8881100 4.116g 158576 R 100.0 26.37 18:57.14 plasmashell
Killing plasmashell und restarting it with "plasmashell --shut-up" resolves the problem until next time.
Anyone else with this problem?
I do. I noticed that it normally manifests once the system resumes from hibernate but it can just be a coincidence. But I am having this problem for a while so it may not be the related to yours.
I do not see this issue here.
2017-01-06 12:04 GMT+01:00 Markos Chandras mchandras@suse.de:
On 01/06/2017 11:00 AM, Bjoern Voigt wrote:
Since some days I have the problem, that "plasmashell" from plasma5-workspace-5.8.4-1.1 (latest Tumbleweed updates) freezes often (around every 30 minutes).
"top" shows then:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 4340 myuser 20 0 8881100 4.116g 158576 R 100.0 26.37 18:57.14 plasmashell
Killing plasmashell und restarting it with "plasmashell --shut-up" resolves the problem until next time.
Anyone else with this problem?
I do. I noticed that it normally manifests once the system resumes from hibernate but it can just be a coincidence. But I am having this problem for a while so it may not be the related to yours.
-- markos
I did notice this also several times after resuming from sleeping, but not exclusively. I have a hardware with NVidia graphics (Optimus) so I played around with the drivers and Mesa. The problem happened also when using the native nvidia driver, apparently there is no nouveau-related cause.
René
René Krell wrote:
I did notice this also several times after resuming from sleeping, but not exclusively. I have a hardware with NVidia graphics (Optimus) so I played around with the drivers and Mesa. The problem happened also when using the native nvidia driver, apparently there is no nouveau-related cause.
I am using the latest stable NVidia driver 375.26 with a standalone graphics card. I have to observe, if the issue also comes after fresh boot or only after suspend/resume.
Greetings, Björn
2017-01-06 12:00 GMT+01:00 Bjoern Voigt bjoernv@arcor.de:
Since some days I have the problem, that "plasmashell" from plasma5-workspace-5.8.4-1.1 (latest Tumbleweed updates) freezes often (around every 30 minutes).
"top" shows then:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 4340 myuser 20 0 8881100 4.116g 158576 R 100.0 26.37 18:57.14 plasmashell
Killing plasmashell und restarting it with "plasmashell --shut-up" resolves the problem until next time.
Anyone else with this problem?
Greetings, Björn
I can confirm this issue, I prepared already a short script to restart plasmashell due to this problem. Some time ago, I created an issue here: https://bugs.kde.org/show_bug.cgi?id=373368
René
René Krell wrote:
I can confirm this issue, I prepared already a short script to restart plasmashell due to this problem. Some time ago, I created an issue here: https://bugs.kde.org/show_bug.cgi?id=373368
Thanks. I subscribed this bug too. Personally I do not use Kmail, but Akonadi for calendar functions in Korganizer. The problem may be unrelated to KMail.
Greetings, Björn
Hi Bjoern,
Since some days I have the problem, that "plasmashell" from plasma5-workspace-5.8.4-1.1 (latest Tumbleweed updates) freezes often (around every 30 minutes).
"top" shows then:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 4340 myuser 20 0 8881100 4.116g 158576 R 100.0 26.37 18:57.14 plasmashell
Killing plasmashell und restarting it with "plasmashell --shut-up" resolves the problem until next time.
Anyone else with this problem?
I observe two incarnations of this problem on my machines. On my notebook plasmashell freezes whenever a notification pops up. I see an empty frame and then everything freezes. This can be worked around by disabling notifications at all. On this machine an Intel graphics adapater is used. The problem happens with the Intel driver and also with the modesetting driver. On my desktop plasmashell also freezes randomly, but I couldn't identify a single reason. This machine has an Nvidia graphics adapter and the problem occurs with the proprietary Nvidia driver and also with the modesetting driver. I supect the problems stemming from Mesa, but I have no proof of this.
Greetings, Björn
Bye. Michael.
Am 07 Jan 2017 11:24:00 +0100 schrieb mh@mike.franken.de (Michael Hirmke):
Hi Bjoern,
Since some days I have the problem, that "plasmashell" from plasma5-workspace-5.8.4-1.1 (latest Tumbleweed updates) freezes often (around every 30 minutes).
"top" shows then:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 4340 myuser 20 0 8881100 4.116g 158576 R 100.0 26.37 18:57.14 plasmashell
Killing plasmashell und restarting it with "plasmashell --shut-up" resolves the problem until next time.
Anyone else with this problem?
Yes, here. Been happening every once in a while since about summer, I guess. Restart helped. I could not really detect the source, but i have several suspicions: a) It seemed to happen only on systems that had self-updating widgets (cpu load, network load, ...) b) removing plasma config files (incl. the widget stuff and definitions) helped... on some systems. c) Plasmashell didn't freeze, it just used one CPU core fully. Never two or more, always only one - but that also doesn't say anything helpful. Multicore systems remain usable, but slow down.
It's especially annoying on laptops, since it drains battery very fast, and that is sad, because KDE is doing very very very well with battery consumption (even with akonadi running I often manage less than 4 Watts on two different laptops - reported by powertop/system and confirmed by runtime)
I observe two incarnations of this problem on my machines. On my notebook plasmashell freezes whenever a notification pops up. I see an empty frame and then everything freezes. This can be worked around by disabling notifications at all. On this machine an Intel graphics adapater is used. The problem happens with the Intel driver and also with the modesetting driver. On my desktop plasmashell also freezes randomly, but I couldn't identify a single reason. This machine has an Nvidia graphics adapter and the problem occurs with the proprietary Nvidia driver and also with the modesetting driver. I supect the problems stemming from Mesa, but I have no proof of this.
Greetings, Björn
Bye. Michael.
I think this is the well known and annoying bug in plasma https://bugs.kde.org/show_bug.cgi?id=356479 (+many other duplicates) Someone (Lindsay Roberts) found at least part of the solution, will be in plasma 5.9 and should be backported to 5.8 (rant - it is hard to believe how many years this could stay on kde buglist when it is so prevalent and easy to determine from profiling, it was a sitting duck and they did nothing)
On Sunday, 8 January 2017 20:29:01 GMT nicholas wrote:
I think this is the well known and annoying bug in plasma https://bugs.kde.org/show_bug.cgi?id=356479 (+many other duplicates) Someone (Lindsay Roberts) found at least part of the solution, will be in plasma 5.9 and should be backported to 5.8 (rant - it is hard to believe how many years this could stay on kde buglist when it is so prevalent and easy to determine from profiling, it was a sitting duck and they did nothing)
As its an open source world, software development/maintenance relies on skilled volunteers so i wouldn't rant too much as this work is done by people giving up their free time for no reward. I know its frustrating but thats the reality of it
On Monday, 9 January 2017 09:50:36 CET ianseeks wrote:
On Sunday, 8 January 2017 20:29:01 GMT nicholas wrote:
I think this is the well known and annoying bug in plasma https://bugs.kde.org/show_bug.cgi?id=356479 (+many other duplicates) Someone (Lindsay Roberts) found at least part of the solution, will be in plasma 5.9 and should be backported to 5.8 (rant - it is hard to believe how many years this could stay on kde buglist when it is so prevalent and easy to determine from profiling, it was a sitting duck and they did nothing)
As its an open source world, software development/maintenance relies on skilled volunteers so i wouldn't rant too much as this work is done by people giving up their free time for no reward. I know its frustrating but thats the reality of it
[my main point was highlight of the bug - the rant was an added bonus] point taken, but does not dismiss the issue. There are several paid members of the plasma team. More importantly the way plasma is structured, no introspection, the near complete lack of documentation mean that casual contributions are very difficult, the system is close to being a black-box.
-
Bjoern Voigt -
ianseeks -
Markos Chandras -
Markus Feilner -
mh＠mike.franken.de -
nicholas -
René Krell -
Robby Engelmann