plasmashell --replace & sudo swapoff -a && sudo swapon -a
any news about the bug report?
]]>That's one of the primary reasons I absolutely loathe doing any kind of support on reddit, it's incredibly clunky to basically have to use external sites for anything that's bigger than a few lines.
I just like it since I feel it's more intuitive. /shrug.
Then I'm out of immediate ideas as well and you might have to check with valgrind
I tried looking into Valgrind, but I can't attach it at runtime and instead have to run it through Valgrind at boot, and Valgrind really slows it down.
I also looked at gdb, but I can't really figure that out. I'm way out of my depth here, since my normal debugging process is using the built-in debugger for IntelliJ.
FWIW you don't seem to be an isolated case, but it seems to be the resource monitors in the relevant current case as well. : https://bugs.kde.org/show_bug.cgi?id=426539
I'll look into making a bug report and see if they can help.
Also, I do have one remaining widget which I forgot to remove and hoped it wasn't causing an issue, since I use it semi-frequently to quicky check my ram/swap levels. I'll remove it and hopefully, that should fix the issue.
I thought it was like reddit (and markdown in general), where it'd be the height of the whole thing, but if it was too long then it'd add a horizontal scroll bar (but never a vertical one)
That's one of the primary reasons I absolutely loathe doing any kind of support on reddit, it's incredibly clunky to basically have to use external sites for anything that's bigger than a few lines.
Then I'm out of immediate ideas as well and you might have to check with valgrind, FWIW you don't seem to be an isolated case, but it seems to be the resource monitors in the relevant current case as well. : https://bugs.kde.org/show_bug.cgi?id=426539
]]>I'll update if this works or not.
I've rebooted since then and with an uptime of about 11 hours, my ram is at 1.0GiB. Doesn't seem to be fixed.
I'm going to revert my theme back to what it was before because I liked how it was configured and the current icon for my file manager confuses me.
Seth and V1del, you have the patience of Job.
Yes, I'm very thankful for their patience. I'd like to think I'm rather good with computers (in relation to the average user), but debugging mem leaks in kde isn't exactly my speciality.
]]>If you wrap outputs in [ code ] [ /code ] tags you don't have to worry about posting long outputs here as long as they aren't as long as to have too many characters to post
I thought it was like reddit (and markdown in general), where it'd be the height of the whole thing, but if it was too long then it'd add a horizontal scroll bar (but never a vertical one)
From what we see here you have multiple of the potentially problematic monitors I was talking about enabled somewhere.
I'll delete those right now since I don't use them frequently.
In addition to that, shader based wallpapers (... that could just be a left over config node I have a similar one for the worldmap but if they are actually enabled I wouldn't be surprised regarding memory usage).
I installed those after I noticed the mem leak, so it's probably not what caused it though I've also disabled that.
I'll update if this works or not.
]]>From what we see here you have multiple of the potentially problematic monitors I was talking about enabled somewhere. In addition to that, shader based wallpapers (... that could just be a left over config node I have a similar one for the worldmap but if they are actually enabled I wouldn't be surprised regarding memory usage).
What happens if you remove those.
]]>If it leaks, the memory will not be freed by changing the theme (you'd have to reboot to free the RAM)
I've rebooted since then and after half a day of it being up, I have ~1.3GiB used by the plasmashell process.
Maybe so we can have a general idea of what you actually have active throw in the output of
qdbus org.kde.plasmashell /PlasmaShell org.kde.PlasmaShell.dumpCurrentLayoutJS #Needs qt5-tools
here is a pastebin with the output of that commands. (It was long and I don't want to spam the thread.)
]]>qdbus org.kde.plasmashell /PlasmaShell org.kde.PlasmaShell.dumpCurrentLayoutJS #Needs qt5-tools
If we can't isolate it this way, you'll have to go for https://en.wikipedia.org/wiki/Valgrind (it's oc. in the repos)
]]>https://wiki.archlinux.org/index.php/NVIDIA_Optimus
This requires data transfer between the units which might or not go through the process RAM for GL data (and leak there)
I'm not using nvidia optimus since I only have a single gpu in my system. (desktop, not laptop)
Change We10XOS plasma style to (probably) breeze (whatever is the current default) to rule it out as cause.
I've changed the global theme to Breeze Dark to overwrite everything else, as I'm 99% sure breeze dark comes preinstalled with kde, so it should have no issues there.
(After changing it there was no immediate effect and after waiting a bit, the ram usage for plasmashell is still at 1.1/1.2GiB)
300MB here. which widgets/plasmoids are you using? Historically the network/cpu/io bandwith monitors were prone to leaks, they have been rewritten however and I haven't heard any reports in that direction yet. FWIW there might be a relation to the 300 chrome tabs if you have the plasma-integration extension installed.
Also which graphics chip/driver? Integrated graphics might lead to some more RAM being used, though I don't think anything here would directly show up as process memory
how did you get plasma down to 300MB ram usage?
in my plasma install, i get about 500 to 600mb ram in idle, with nothing running, kvantum theme and standard stock widgets running, and a few services like zram ananicy and firefox-sync.
im not too woried, its fast and snappy, and wont break even after updating just today, about 7 months without updating.
Change We10XOS plasma style to (probably) breeze (whatever is the current default) to rule it out as cause.
]]>