https://bugs.kde.org/show_bug.cgi?id=470026

Adam Fontenot <adam.m.fontenot+...@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|WAITINGFORINFO              |---
             Status|NEEDSINFO                   |REPORTED

--- Comment #2 from Adam Fontenot <adam.m.fontenot+...@gmail.com> ---
(In reply to Nate Graham from comment #1)
> Hi Adam! Sorry we were not able to get to this yet. Can you still reproduce
> the issue in Plasma 6.3.4 or later? Thanks a lot!

I think this needs a little more testing to confirm some specifics of the issue
(I don't remember how / why I downloaded so many files), but I was able to
reproduce it in principle - even saving a single file causes kactivitymanagerd
to noticeably spike.

However, I think this issue might be caused by the file picker. When I click
save, it's appears to be the *portal* that is updating
~/.local/share/recently-used.xbel, and kactivitymanagerd then reads / parses
this file, and that's the CPU activity I see. (Both the default Firefox file
picker and the KDE xdg-desktop-portal one seem to trigger this.) The issue
doesn't seem to be the result of the download itself - I see the CPU activity
even if the file fails to download.

I tried the most obvious thing I could think of, which was to disable the
"recent files / remember opened documents" option in system settings. However,
I ran into an issue that was rather disturbing - neither disabling the option
nor *explicitly* clearing the history ("forget everything") deleted the
contents of the file recently-used.xbel nor did they affect the behavior of
this issue. Am I misunderstanding what these settings are supposed to do, or is
this a bug that I need to report separately?

Manually deleting the recently-used.xbel file knocked the CPU time used by
kactivitymanagerd down to almost nothing, indicating that parsing this file
(which could get very large if history pruning is not enabled or is broken) is
probably responsible for the issue.

-- 
You are receiving this mail because:
You are watching all bug changes.

Reply via email to