Bug 382288

Summary: libQt5Core5 5.9.1 introduces memory leak and plasmashell crash
Product: [Plasma] plasmashell Reporter: Mircea Kitsune <sonichedgehog_hyperblast00>
Component: generalAssignee: David Edmundson <kde>
Status: RESOLVED DUPLICATE    
Severity: normal CC: plasma-bugs, sonichedgehog_hyperblast00
Priority: NOR    
Version: 5.10.3   
Target Milestone: 1.0   
Platform: Other   
OS: Linux   
Latest Commit: Version Fixed In:
Sentry Crash Report:
Attachments: Screenshot

Description Mircea Kitsune 2017-07-12 23:03:16 UTC 
Since upgrading from libQt5Core5 5.9.0 to 5.9.1, I experience two issues with plasmashell, which seem to be caused by the same core problem.

The first is that the desktop causes massive memory usage. In over an hour, my system goes from using 2GB of memory (after booting and opening my applications) to approximately 8GB! However this memory is not attributed to any one process; It seems to be disk cache memory, which is further indicated by the fact that issuing the command "echo 3 > /proc/sys/vm/drop_caches" causes some of it to be freed up (not all).

The second issue is that every few hours, plasmashell will crash and restart. I use an image slideshow wallpaper, and before the crash it always becomes black instead of an image... plasmashell then dies immediately after I attempt to click anything on the desktop.
Comment 1 Mircea Kitsune 2017-07-13 01:16:41 UTC 
Created attachment 106586 [details]
Screenshot

Attaching a screenshot of the desktop as it becomes broken. It shows both the Media Frame widget and slideshow wallpaper being replaced by a black color, whereas normally they should contain images... it's at this point that clicking on things causes Plasma to crash and restart. The only edits include me blurring the contents of a plasmoid, and adding the red text and arrows to better indicate what happens.
Comment 2 David Edmundson 2017-07-13 18:58:49 UTC 
Can you confirm that setting it to not a single image instead of a slideshow means this doesn't happen?

If so, it's the duplicate report, if not can you reopen this.

*** This bug has been marked as a duplicate of bug 381000 ***