Hello,
Starting a couple of days ago, pamac-manager is crashing with a segfault.
I can , however, start it from command line with a parameter ( pamac-manager --updates ) and it opens on the updates page, where I can perform the refresh and updates normally.
Navigating on the Installed page works as expected as well.
Navigating on the Browse page, however, results in a segfault . I do not get any meaningful output when running with strace.
I looked at the installed packages versions and there seems to be a discrepancy (not sure if it is normal):
- pamac from galaxy is at version 11.7.1-1
- libpamac from galaxy is at version 11.6.4-4
Is there a conflict introduced by the latest updates of GTK libraries ?
It works here.
But I only use pamac to notify me of updates.
The rest, I do mostly in pacman or yay.
Hi. I pushed new version for pamac (11.7.3-1) and libpamac (11.7.3-1) to galaxy-gremlins repo, can you test it please?
Hi
@marzosh , thank you for the update.
I installed the new version from galaxy-gremlins and I get exactly the same behavior :
- direct launch of
pamac-manager results in a segfault
- launch with
pamac-manager --updates opens the window with the
Updates tab active; I can search for updates, refresh updates and visit the
Installed tab where I can use
Search ; as soon as I click on
Browse tab, pamac-manager stops with a segfault.
Is there a way to enable meaningful traces to catch where the problem lies ?
Couldn't reproduce your problem on a clean setup here – Artix repos (system, world, galaxy) worked fine. Try this: back up your config (if you have), wipe pamac and libpamac, then reinstall from scratch and update.
@marzosh Your reply gave me an idea - I manually edited the config file (~/.config/pamac/config) and set the
software_mode flag to
true.
After that,
pamac-manager started without segfault and I could change all tabs.
I disabled the software mode from the GUI, closed and started again
pamac-manager and this time it started normally, without any segfault.
I have inspected the config file before modification and it seems identical to the current one.
Weird - but hey, it works!
Thank you for the time spent so far !
Good idea, glad it worked! I’ll go ahead and mark this as resolved then.