Plasmashell memory leak

[pemartins]

Mine is bigger than yours!

I've got a memory leak in plasmashell, it has at this moment more than 1 GB:

pemartins@pemartins-X55U: ~ $ ps aux | grep plasmashell
pemarti+ 18186  3.7 33.0 6041604 1195304 ? Sl   05:22   9:54 /usr/bin/plasmashell
pemarti+ 31834  0.0  0.0  14432   964 pts/1    S+   09:46   0:00 grep --color=auto plasmashell
pemartins@pemartins-X55U: ~ $ apt policy plasma-desktop
plasma-desktop:
Installed: 4:5.12.5-0ubuntu0.1
Candidate: 4:5.12.5-0ubuntu0.1
Version table:
*** 4:5.12.5-0ubuntu0.1 500
    500 http://archive.ubuntu.com/ubuntu bionic-updates/universe amd64 Packages
    100 /var/lib/dpkg/status
 4:5.12.4-0ubuntu1 500
    500 http://archive.ubuntu.com/ubuntu bionic/universe amd64 Packages
pemartins@pemartins-X55U: ~ $ uptime
09:46:46 up 12:07,  3 users,  load average: 1,52, 1,97, 1,99
pemartins@pemartins-X55U: ~ $ uname -a
Linux pemartins-X55U 4.15.0-23-generic #25-Ubuntu SMP Wed May 23 18:02:16 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux
pemartins@pemartins-X55U: ~ $ free
          total        used        free      shared  buff/cache   available
Mem:        3613648     2358588      391360       23656      863700     1170584
Swap:       3761148       26124     3735024

How can I identify the root of the leak?

[chimak111]

Can you disable redshift-gtk and see if that helps?

You could also look at what's causing your CPU usage to be 57%.

Including the output of

inxi -Fxz

may be informative as well. If you don't have inxi installed, get it with sudo apt install inxi.

See https://www.kubuntuforums.net/showth...ll-Memory-leak for widgets that could be misbehaving.

[SpecialEd]

Are you using "Slideshow" as your screensaver?

[pemartins]

You could also look at what's causing your CPU usage to be 57%.

Oh that was the infamous update-apt-xapi.

$ inxi -Fxz
System:    Host: pemartins-X55U Kernel: 4.15.0-23-generic x86_64 bits: 64 gcc: 7.3.0
          Desktop: KDE Plasma 5.12.5 (Qt 5.9.5) Distro: Custom Live CD
Machine:   Device: laptop System: ASUSTeK product: X55U v: 1.0 serial: N/A
          Mobo: ASUSTeK model: X55U v: 1.0 serial: N/A
          UEFI: American Megatrends v: X55U.423 date: 08/06/2013
CPU:       Dual core AMD E2-1800 APU with Radeon HD Graphics (-MCP-)
          arch: Bobcat rev.0 cache: 512 KB
          flags: (lm nx sse sse2 sse3 sse4a ssse3 svm) bmips: 6787
          clock speeds: max: 1700 MHz 1: 1002 MHz 2: 851 MHz
Graphics:  Card: Advanced Micro Devices [AMD/ATI] Wrestler [Radeon HD 7340]
          bus-ID: 00:01.0
          Display Server: x11 (X.Org 1.19.6 ) driver: radeon
          Resolution: 1366x768@60.01hz
          OpenGL: renderer: AMD PALM (DRM 2.50.0 / 4.15.0-23-generic, LLVM 6.0.0)
          version: 3.3 Mesa 18.0.4 Direct Render: Yes
Audio:     Card-1 Advanced Micro Devices [AMD] FCH Azalia Controller
          driver: snd_hda_intel bus-ID: 00:14.2
          Card-2 Advanced Micro Devices [AMD/ATI] Wrestler HDMI Audio
          driver: snd_hda_intel bus-ID: 00:01.1
          Sound: Advanced Linux Sound Architecture v: k4.15.0-23-generic
Network:   Card-1: Ralink RT5390 Wireless 802.11n 1T/1R PCIe
          driver: rt2800pci v: 2.3.0 bus-ID: 01:00.0
          IF: wlp1s0 state: up mac: <filter>
          Card-2: Qualcomm Atheros AR8161 Gigabit Ethernet
          driver: alx port: e000 bus-ID: 02:00.0
          IF: enp2s0 state: down mac: <filter>
Drives:    HDD Total Size: 500.1GB (6.5% used)
          ID-1: /dev/sda model: ST500LM012_HN size: 500.1GB
Partition: ID-1: / size: 56G used: 27G (52%) fs: ext4 dev: /dev/sda10
          ID-2: swap-1 size: 3.85GB used: 0.00GB (0%) fs: swap dev: /dev/sda6
RAID:      No RAID devices: /proc/mdstat, md_mod kernel module present
Sensors:   System Temperatures: cpu: 59.0C mobo: N/A gpu: 59.0
          Fan Speeds (in rpm): cpu: 2800
Info:      Processes: 160 Uptime: 12 min Memory: 1154.7/3529.0MB
          Init: systemd runlevel: 5 Gcc sys: 7.3.0
          Client: Shell (bash 4.4.191) inxi: 2.3.56

I exited Redshift.

---

Are you using "Slideshow" as your screensaver?

Yes I am. When I browsed the issue I read that indeed that could be the cause of the memory leak but I do not think that it is in my particular situation. I have slideshow as screensaver set for about a week now and I never had any significant memory leak issues.

It's usual that plasmashell keeps increasing size during a session but nothing significant, I do not recall seeing it above 300 MB (usually my system starts with it around 180 MB).
Yesterday I remember doing stuff like adding themes and widgets to Plasma, changing themes (but I kept no new widgets besides the ones I already was using), testing different panel layouts, using Wine programs to look for duplicate files, I remember also using Firefox for normal browsing and watching a video stream, also used Kodi for a bit. Maybe I did some more stuff but this is what I can remember.

Recalling my actions and the abnormal plasmashell size I'm tempted to jump to the conclusion that the cause could be the customization I was fooling around with, all the other actions I have done before without similar plasmashell behave.

[SpecialEd]

<<<<<<

Yes I am. When I browsed the issue I read that indeed that could be the cause of the memory leak but I do not think that it is in my particular situation. I have slideshow as screensaver set for about a week now and I never had any significant memory leak issues.<<<<<

I'm not saying Slideshow is your problem but seemingly unrelated updates can affect the way Slideshow and programs like Discover, etc., behave. Ask me how I know.

[pemartins]

I just remembered that yesterday I also did a lot of moving and deleting files on Dolphin. Usually when I do that and when a lot of files (and big amount of data) are moved or deleted, I notice that the panel hangs for a bit (noticed once that the clock was delayed a couple of minutes) after the action is done; then I get the activity notification of files being moved, but actually the job is already done and nothing is being moved and I have to close (press stop on the icon in the notifications) the notification because it only shows a transfer but no file names are displayed.

Today I've been using the pc normally and plasmashell remains as usual without signs of any memory leak. So I'll mark this one as solved since the abnormal activity seems to be what caused the memory leak and it's hard to reproduce the steps in order to try to replicate it.

[GreyGeek]

Was it "abnormal" activity, or normal indexing?

[pemartins]

I really can't tell. I have file search disabled, also unchecked search folder updater from autostart, both are off since I installed Kubuntu. But I do not know if there are other indexing services running like updatedb or something similar, nor do I know if they are stored at plasmashell.

[GreyGeek]

You can use ksysguard (K System Monitor) and sort by %CPU usage or total memory, or run top in a Konsole, to see which process is using the most memory or CPU. If you set ksysguard to the tree display mode then you can follow a process down the tree to locate the guilty party (if it is in a treed process).

[pemartins]

The math is not right in my ram usage, if I do a total on all processes shown it should perhaps return a sum of about 600 MB or so; but as it shows in the printscreen, how can it be taking 1.6 GB of ram? What is using the extra 1 GB of ram not shown anywhere?



It just does not add up...

edit:

$ ps -u $USER -wo rss=,comm= --sort -rss | while read -r rss comm ; do echo $((rss/1024))"MB -" $comm; done && free -m
265MB - plasmashell
80MB - kded5
73MB - konsole
56MB - kwin_x11
27MB - krunner
27MB - ksmserver
24MB - http.so
24MB - http.so
22MB - firetools
15MB - kactivitymanage
15MB - klauncher
11MB - kdeconnectd
10MB - org_kde_powerde
10MB - python
10MB - polkit-kde-auth
10MB - kglobalaccel5
9MB - kwalletd5
9MB - kaccess
9MB - kuiserver5
9MB - pulseaudio
6MB - kdeinit5
5MB - xembedsniproxy
5MB - kscreen_backend
4MB - bash
3MB - xbindkeys
3MB - kio_http_cache_
3MB - ps
2MB - dbus-daemon
2MB - at-spi2-registr
2MB - gvfsd
2MB - redshift
2MB - systemd
2MB - at-spi-bus-laun
2MB - (sd-pam)
2MB - gconfd-2
1MB - dconf-service
1MB - bash
1MB - dbus-daemon
1MB - ksysguardd
0MB - gconf-helper
0MB - startkde
0MB - kwrapper5
0MB - ssh-agent
0MB - start_kdeinit
             total        used        free      shared  buff/cache   available
Mem:           3528        1649        1296           3         583        1654
Swap:          3672           2        3670

[chimak111]

$ ps -u $USER -wo rss=,comm= --sort -rss | while read -r rss comm ; do echo $((rss/1024))"MB -" $comm; done && free -m

Could you please explain ^^^

I've usually just looked at free -m. Understanding RAM usage in Linux can get pretty difficult for us non-technical users: https://www.linuxatemyram.com/

[pemartins]

@chimak111 thank you so much for the info, I really hope that link about the disk cache on ram is the explanation. Stuff like a 1 GB virus running on the background crossed my mind...

I'd love to be able to explain that command but unfortunately is most probable that we see Trump and Hillary dancing salsa together before I can understand what's on it It's a copy and paste from somewhere made some time ago, all I recall is that the command intents to display the ram use per application in the user session.

[GreyGeek]

$ ps -u $USER -wo rss=,comm= --sort -rss | while read -r rss comm ; do echo $((rss/1024))"MB -" $comm; done && free -m

Could you please explain ^^^

I've usually just looked at free -m. Understanding RAM usage in Linux can get pretty difficult for us non-technical users: https://www.linuxatemyram.com/

It's output is a listing of the total RAM used for each running process, and totals for each of the parts RAM is divided into. For a complete listing of possible command sequences see "man ps". It is a lot easier to get totals this way rather than using a calculator on ksysguard's output and trying to total them manually.

:~$ ps -u $USER -wo rss=,comm= --sort -rss | while read -r rss comm ; do echo $((rss/1024))"MB -" $comm; done && free -m
319MB - plasmashell
301MB - firefox
199MB - Web Content
103MB - kwin_x11
84MB - Web Content
78MB - krunner
63MB - kgpg
62MB - konsole
61MB - hp-systray
60MB - blueman-applet
58MB - evolution-calen
55MB - kded5
52MB - evolution-alarm
52MB - klauncher
51MB - ksmserver
50MB - kdeconnectd
48MB - evolution-calen
48MB - evolution-calen
46MB - polkit-kde-auth
45MB - kwalletd5
44MB - kaccess
43MB - org_kde_powerde
42MB - kglobalaccel5
32MB - kactivitymanage
27MB - kuiserver
27MB - gmenudbusmenupr
27MB - xembedsniproxy
25MB - distro-release-
24MB - kscreen_backend
24MB - hp-systray
24MB - kio_http_cache_
23MB - evolution-sourc
22MB - kdeinit5
18MB - evolution-addre
18MB - evolution-addre
16MB - hp-systray
11MB - pulseaudio
9MB - obexd
7MB - gnome-keyring-d
6MB - kwrapper5
6MB - gnome-keyring-d
6MB - at-spi2-registr
6MB - gvfsd
5MB - gconfd-2
5MB - bash
5MB - at-spi-bus-laun
5MB - bash
4MB - systemd
4MB - dconf-service
4MB - dbus-daemon
4MB - gconf-helper
3MB - dbus-daemon
3MB - ps
2MB - dbus-launch
2MB - (sd-pam)
1MB - ksysguardd
1MB - bash
1MB - startkde
0MB - gpg-agent
0MB - ssh-agent
0MB - start_kdeinit
              total        used        free      shared  buff/cache   available
Mem:           5801        1094        2440          56        2266        4290
Swap:             0           0           0

[bschelst]

mine is even bigger...

ptime;ps -u $USER -wo rss=,comm= --sort -rss | while read -r rss comm ; do echo $((rss/1024))"MB -" $comm; done && free -m
20:06:35 up 7:30, 3 users, load average: 1,02, 0,91, 0,90
3112MB - plasmashell
713MB - akonadi_ews_res
629MB - chrome
532MB - TeamViewer
408MB - chrome
402MB - mysqld
383MB - kmail
368MB - chrome
364MB - chrome

[GreyGeek]

I have only 6GB of RAM. Do you have more?