Bug 499100 - Plasma Discover built with GCC 15 crashes on start
Summary: Plasma Discover built with GCC 15 crashes on start
Status: RESOLVED UPSTREAM
Alias: None
Product: Discover
Classification: Applications
Component: discover (show other bugs)
Version: 6.2.91
Platform: Fedora RPMs Linux
: NOR crash
Target Milestone: ---
Assignee: Plasma Bugs List
URL:
Keywords: drkonqi
Depends on:
Blocks:
 
Reported: 2025-01-24 17:19 UTC by Steve Cossette
Modified: 2025-02-15 16:45 UTC (History)
7 users (show)

See Also:
Latest Commit:
Version Fixed In:
Sentry Crash Report: https://crash-reports.kde.org/organizations/kde/issues/123248/events/37a914a779174354a56b4ccce6476e3a/


Attachments
New crash information added by DrKonqi (180.63 KB, text/plain)
2025-01-24 17:19 UTC, Steve Cossette
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Steve Cossette 2025-01-24 17:19:54 UTC
Application: plasma-discover (6.2.91)

ApplicationNotResponding [ANR]: false
Qt Version: 6.8.1
Frameworks Version: 6.10.0
Operating System: Linux 6.13.0-0.rc7.20250114gitc45323b7560e.56.fc42.x86_64 x86_64
Windowing System: Wayland
Distribution: "Fedora Linux 42 (KDE Plasma Prerelease)"
DrKonqi: 6.2.91 [CoredumpBackend]

-- Information about the crash:
Discover simply crashes on start on 6.2.91. That's it. Launch it, it crashes.

Note that this is on a brand new virtual machine on virt-manager from a Fedora rawhide snapshot.

The crash can be reproduced every time.

-- Backtrace (Reduced):
#5  std::__atomic_base<int>::fetch_add (this=0x8548013140c63878, __i=1, __m=std::memory_order_acq_rel) at /usr/include/c++/15/bits/atomic_base.h:630
#6  QAtomicOps<int>::ref<int> (_q_value=<error reading variable: Cannot access memory at address 0x8548013140c63878>) at /usr/src/debug/qt6-qtbase-6.8.1-11.fc42.x86_64/src/corelib/thread/qatomic_cxx11.h:259
[...]
#9  0x00007fc1ed356b68 in QObjectPrivate::connectImpl (sender=sender@entry=0x5643b4827f10, signal_index=11, receiver=receiver@entry=0x5643b2bd4230, slot=slot@entry=0x0, slotObjRaw=slotObjRaw@entry=0x5643b2bd4230, type=2, types=0x0, senderMetaObject=0x7fc0cc230f60 <PackageKitBackend::staticMetaObject>) at /usr/src/debug/qt6-qtbase-6.8.1-11.fc42.x86_64/src/corelib/kernel/qobject.cpp:5300
#10 0x00007fc1ed357320 in QObject::connectImpl (sender=0x5643b4827f10, signal=<optimized out>, receiver=0x5643b2bd4230, slot=0x0, slotObjRaw=0x5643b2bd4230, type=Qt::QueuedConnection, types=0x0, senderMetaObject=<optimized out>) at /usr/src/debug/qt6-qtbase-6.8.1-11.fc42.x86_64/src/corelib/kernel/qobject.cpp:5230
#14 0x00007fc0cc1c8b4d in PackageKitBackend::deferredResultStream(QString const&, std::function<void (PKResultsStream*)>) () from /usr/lib64/qt6/plugins/discover/packagekit-backend.so


Reported using DrKonqi
Comment 1 Steve Cossette 2025-01-24 17:19:55 UTC
Created attachment 177649 [details]
New crash information added by DrKonqi

DrKonqi auto-attaching complete backtrace.
Comment 2 Alessandro Astone 2025-01-24 17:54:23 UTC
gcc 15 is most likely to blame here.
Comment 3 Harald Sitter 2025-01-25 11:15:33 UTC
Can't reproduce on KDE Linux with a master build nor with a 6.3 build. Not sure what to do. Maybe it is indeed a compiler bug.
Comment 4 Steve Cossette 2025-01-25 13:43:41 UTC
FYI, running the update on Fedora 41 results in a 100% functional discover. So GCC15 being a culprit is very likely.
Comment 5 Neal Gompa 2025-01-26 10:21:01 UTC
This can be verified with the COPR build for Fedora 41: https://copr.fedorainfracloud.org/coprs/farchord/kde-test/builds/
Comment 6 Nate Graham 2025-01-31 16:23:43 UTC
Apparently it's a GCC 15 bug: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=116506
Comment 7 Sam James 2025-02-15 16:45:28 UTC
Fixed in GCC now.