#1070755 akregator: Akregator hangs and crashed in Bookworm KDE on start (arm64 - VmWare Fusion on mac m2)

Package:
akregator
Source:
akregator
Description:
RSS/Atom feed aggregator
Submitter:
Vladimir Zivadinovic
Date:
2026-05-25 10:31:02 UTC
Severity:
normal
Tags:
#1070755#5
Date:
2024-05-08 14:02:31 UTC
From:
To:
Dear Maintainer,

*** Reporter, please consider answering these questions, where appropriate ***

   * What led up to the situation?

I have freshly installed the arm64 in VmWare Fusion, KDE desktop. Other apps
seem to work normally. I tried to launch Akregator, and it hand and crashed.

The window launches, looks normal, but hangs when I try clicking inside.The
title changes to Akregator (Not Responding).

After relaunch, I get a message "Akregator did not close correctly. Would you
like to restore the previous session?" with 3 buttons bellow. However, it hangs
again when I click inside the window, and the title changes again.

   * What exactly did you do (or not do) that was effective (or
     ineffective)?
I tried removing configuration and cache files.
   * What was the outcome of this action?

Same issue. Nothing changed.
   * What outcome did you expect instead?

To work normally.

*** End of the template - remove these template lines ***

#1070755#10
Date:
2024-05-28 11:20:35 UTC
From:
To:
Hi,

It seems we're also getting the reported behaviour when testing under openqa.

Sadly one doesn't get to see anything when it happens:

https://openqa.debian.net/tests/265985#step/akregator/7

but having jumped into one of those tests, and tried it again, I then got the
same "did not close" warning, as reported.

We are running this in the test as:

  QT_LOGGING_RULES="*=true" akregator

which will hopefully get you some logs to look at. I believe the resulting
output ends up here:

https://openqa.debian.net/tests/265985/logfile?filename=akregator-journal.txt

HTH

Cheers, Phil.

#1070755#15
Date:
2024-08-10 20:28:30 UTC
From:
To:
Can replicate on an M1 Pro MacBook Pro, running Debian Trixie through
Asahi. Appears to be hung up in totalCountFor, the less significant
frames of the trace change with every break but it's always in
totalCountFor.

Backtrace from gdb:

(gdb) bt
#0  0x0000fffff7ee78f8 in c4_ColOfInts::Get_8i (this=0xaaaaab524d28, index_=4) at ./src/storage/metakit/src/column.cpp:924
#1  0x0000fffff7ee8fa4 [PAC] in c4_ColOfInts::Get (this=0xaaaaab524d28, index_=<optimized out>, length_=@0xffffffffda84: 8) at ./src/storage/metakit/src/column.cpp:1353
#2  0x0000fffff7ef4f10 [PAC] in c4_Handler::GetBytes (this=<optimized out>, index_=index_@entry=4, buf_=..., copySmall_=copySmall_@entry=false)
    at ./src/storage/metakit/src/handler.cpp:56
#3  0x0000fffff7f076f0 [PAC] in c4_Sequence::Get (this=0xaaaaab728d40, index_=4, propId_=<optimized out>, buf_=...) at ./src/storage/metakit/src/viewx.cpp:363
#4  0x0000fffff7f08738 [PAC] in c4_Reference::GetData (buf_=..., this=<optimized out>) at ./src/storage/metakit/include/mk4.inl:420
#5  c4_IntRef::operator long (this=this@entry=0xffffffffdb88) at ./src/storage/metakit/src/viewx.cpp:461
#6  0x0000fffff7efcd3c [PAC] in c4_HashViewer::IsDummy (row_=4, this=0xaaaaab728e90) at ./src/storage/metakit/src/remap.cpp:160
#7  c4_HashViewer::LookDict (this=this@entry=0xaaaaab728e90, hash_=-7788384493399555046, cursor_=...) at ./src/storage/metakit/src/remap.cpp:303
#8  0x0000fffff7efcf18 [PAC] in c4_HashViewer::Lookup (this=0xaaaaab728e90, key_=..., count_=@0xffffffffdc94: 65535) at ./src/storage/metakit/src/remap.cpp:401
#9  0x0000fffff7eeb59c [PAC] in c4_CustomSeq::RestrictSearch (this=<optimized out>, cursor_=..., pos_=@0xffffffffdccc: 0, count_=@0xffffffffdcd4: 5)
    at ./src/storage/metakit/src/custom.cpp:126
#10 0x0000fffff7f05784 [PAC] in c4_View::Find (this=0xaaaaab5cc190, crit_=..., start_=<optimized out>, start_@entry=0) at ./src/storage/metakit/src/view.cpp:904
#11 0x0000fffff7f10420 [PAC] in Akregator::Backend::Storage::totalCountFor (this=0xaaaaab5c65e0, url=...) at /usr/include/c++/13/bits/unique_ptr.h:199
#12 0x0000fffff7f0aad4 [PAC] in Akregator::Backend::FeedStorage::totalCount (this=<optimized out>) at ./src/storage/feedstorage.cpp:160
#13 0x0000fffff7ecde78 in Akregator::Feed::fromOPML (e=..., storage=0xaaaaab5c65e0) at ./src/feed/feed.cpp:191
#14 0x0000fffff7ed71a8 [PAC] in Akregator::FeedList::parseChildNodes (this=0xaaaaabc35340, node=..., parent=0xaaaaac22ddb0) at /usr/include/c++/13/bits/unique_ptr.h:199
#15 0x0000fffff7ed7378 [PAC] in Akregator::FeedList::parseChildNodes (this=0xaaaaabc35340, node=..., parent=<optimized out>) at ./src/feed/feedlist.cpp:238
#16 0x0000fffff7ed7378 [PAC] in Akregator::FeedList::parseChildNodes (this=this@entry=0xaaaaabc35340, node=..., parent=<optimized out>) at ./src/feed/feedlist.cpp:238
#17 0x0000fffff7ed7ae0 [PAC] in Akregator::FeedList::readFromOpml (this=0xaaaaabc35340, doc=...) at ./src/feed/feedlist.cpp:275
#18 0x0000ffffd9c60de4 [PAC] in Akregator::LoadFeedListCommandPrivate::handleDocument (this=this@entry=0xaaaaab3a7510, doc=...)
    at /usr/include/aarch64-linux-gnu/qt5/QtCore/qsharedpointer_impl.h:307
#19 0x0000ffffd9c6198c [PAC] in Akregator::LoadFeedListCommandPrivate::doLoad (this=0xaaaaab3a7510) at ./src/command/loadfeedlistcommand.cpp:134
#20 0x0000fffff6586c38 [PAC] in QtPrivate::QSlotObjectBase::call (this=<optimized out>, r=<optimized out>, a=0xffffffffe340)
    at ../../include/QtCore/../../src/corelib/kernel/qobjectdefs_impl.h:398
#21 QSingleShotTimer::timerEvent (this=0xaaaaabce3a10) at kernel/qtimer.cpp:322
#22 0x0000fffff6573e80 [PAC] in QObject::event (this=0xaaaaabce3a10, e=0xffffffffe530) at kernel/qobject.cpp:1369
#23 0x0000fffff708d1a0 [PAC] in QApplicationPrivate::notify_helper (this=<optimized out>, receiver=0xaaaaabce3a10, e=0xffffffffe530) at kernel/qapplication.cpp:3640
#24 0x0000fffff653d78c [PAC] in QCoreApplication::notifyInternal2 (receiver=0xaaaaabce3a10, event=0xffffffffe530) at kernel/qcoreapplication.cpp:1064
#25 0x0000fffff653da18 [PAC] in QCoreApplication::sendEvent (receiver=<optimized out>, event=event@entry=0xffffffffe530) at kernel/qcoreapplication.cpp:1462
#26 0x0000fffff65a9374 in QTimerInfoList::activateTimers (this=0xaaaaaab17b80) at kernel/qtimerinfo_unix.cpp:643
#27 0x0000fffff65a9ec4 [PAC] in timerSourceDispatch (source=<optimized out>) at kernel/qeventdispatcher_glib.cpp:183
#28 idleTimerSourceDispatch (source=<optimized out>) at kernel/qeventdispatcher_glib.cpp:230
#29 0x0000fffff3d2d520 [PAC] in g_main_dispatch (context=context@entry=0xaaaaaab16da0) at ../../../glib/gmain.c:3344
#30 0x0000fffff3d2f8dc [PAC] in g_main_context_dispatch_unlocked (context=0xaaaaaab16da0) at ../../../glib/gmain.c:4152
#31 g_main_context_iterate_unlocked (context=context@entry=0xaaaaaab16da0, block=block@entry=1, dispatch=dispatch@entry=1, self=<optimized out>) at ../../../glib/gmain.c:4217
#32 0x0000fffff3d30088 [PAC] in g_main_context_iteration (context=0xaaaaaab16da0, may_block=1) at ../../../glib/gmain.c:4282
#33 0x0000fffff65aa298 [PAC] in QEventDispatcherGlib::processEvents (this=0xaaaaaab17fa0, flags=...) at kernel/qeventdispatcher_glib.cpp:423
#34 0x0000fffff653ba84 [PAC] in QEventLoop::exec (this=this@entry=0xffffffffe7f0, flags=..., flags@entry=...) at ../../include/QtCore/../../src/corelib/global/qflags.h:69
#35 0x0000fffff65458e0 [PAC] in QCoreApplication::exec () at ../../include/QtCore/../../src/corelib/global/qflags.h:121
#36 0x0000fffff69519bc [PAC] in QGuiApplication::exec () at kernel/qguiapplication.cpp:1863
#37 0x0000fffff708d148 in QApplication::exec () at kernel/qapplication.cpp:2832
#38 0x0000aaaaaaaab090 in main (argc=<optimized out>, argv=<optimized out>) at ./src/main.cpp:115

#1070755#20
Date:
2026-05-25 10:28:16 UTC
From:
To:
The upstream fix for this bug has landed in KDE, on both master and the
release/26.04 stable branch:

  master:        https://invent.kde.org/pim/akregator/-/commit/98fa7ab52f02e67700803b47ab85b7fb2064618a
  release/26.04: merged via MR !75

It's a two-line change in the bundled metakit (src/storage/metakit/include/mk4.h)
adding __aarch64__ / __arm64__ to the q4_LONG64 guard. Without it, t4_i32
resolves to an 8-byte long instead of a 4-byte int on 64-bit ARM, corrupting
the hash lookup in Storage::totalCountFor() and hanging akregator on launch.

The patch applies cleanly to the 24.12.3 source currently in Trixie.
Attached as a DEP-3 quilt patch ready for debian/patches/.

Please consider to include this in the next Trixie point release. The
bug has been open across KDE, Debian, and Ubuntu for several years and
affects every arm64 user.


Thanks,
Vladimir Zivadinovic