Bug 2270892 - [abrt] krfb: std::__glibcxx_assert_fail(): krfb killed by SIGABRT
Summary: [abrt] krfb: std::__glibcxx_assert_fail(): krfb killed by SIGABRT
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: Fedora
Classification: Fedora
Component: krfb
Version: 40
Hardware: x86_64
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Than Ngo
QA Contact: Fedora Extras Quality Assurance
URL: https://retrace.fedoraproject.org/faf...
Whiteboard: abrt_hash:d157aa06920b8136259acd81a9a...
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2024-03-22 00:51 UTC by 80myol6v2
Modified: 2024-05-08 03:31 UTC (History)
6 users (show)

Fixed In Version: krfb-24.02.2-3.fc40
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2024-05-08 03:31:29 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
File: proc_pid_status (1.49 KB, text/plain)
2024-03-22 00:51 UTC, 80myol6v2
no flags Details
File: maps (3.98 KB, text/plain)
2024-03-22 00:51 UTC, 80myol6v2
no flags Details
File: limits (1.29 KB, text/plain)
2024-03-22 00:51 UTC, 80myol6v2
no flags Details
File: open_fds (11.80 KB, text/plain)
2024-03-22 00:51 UTC, 80myol6v2
no flags Details
File: mountinfo (2.88 KB, text/plain)
2024-03-22 00:51 UTC, 80myol6v2
no flags Details
File: os_info (721 bytes, text/plain)
2024-03-22 00:51 UTC, 80myol6v2
no flags Details
File: core_backtrace (34.03 KB, text/plain)
2024-03-22 00:51 UTC, 80myol6v2
no flags Details
File: dso_list (165 bytes, text/plain)
2024-03-22 00:51 UTC, 80myol6v2
no flags Details
File: var_log_messages (719 bytes, text/plain)
2024-03-22 00:51 UTC, 80myol6v2
no flags Details
File: cpuinfo (3.05 KB, text/plain)
2024-03-22 00:51 UTC, 80myol6v2
no flags Details
File: backtrace (136.68 KB, text/plain)
2024-03-22 00:51 UTC, 80myol6v2
no flags Details
File: environ (2.21 KB, text/plain)
2024-03-22 00:51 UTC, 80myol6v2
no flags Details

Description 80myol6v2 2024-03-22 00:51:25 UTC
Description of problem:
Connected with Remmina on Vannila OS. Worked for a while then Krfb crashed. Remmina reportedly working fine

Version-Release number of selected component:
krfb-24.02.0-1.fc40

Additional info:
reporter:       libreport-2.17.15
type:           CCpp
reason:         krfb killed by SIGABRT
journald_cursor: s=395d69511ecb477cbde3b3fa91c0d1f8;i=cab86;b=cdd5e29a01fc4363b45b0ac49e069710;m=1799f68e04;t=613f5ecf55cd0;x=75f47b0781cd071d
executable:     /usr/bin/krfb
cmdline:        /usr/bin/krfb -qwindowtitle Krfb
cgroup:         0::/user.slice/user-1000.slice/user/app.slice/app-org.kde.krfb-fc6ad8dc9a264daeb286d82f41afd9d5.scope
rootdir:        /
uid:            1000
kernel:         6.8.0-0.rc6.49.fc40.x86_64
package:        krfb-24.02.0-1.fc40
runlevel:       N 5
backtrace_rating: 4
crash_function: std::__glibcxx_assert_fail
comment:        Connected with Remmina on Vannila OS. Worked for a while then Krfb crashed. Remmina reportedly working fine

Truncated backtrace:
Thread no. 1 (24 frames)
 #4 std::__glibcxx_assert_fail at ../../../../../libstdc++-v3/src/c++11/assert_fail.cc:41
 #5 std::_Optional_base_impl<PipeWireCursor, std::_Optional_base<PipeWireCursor, false, false> >::_M_get at /usr/include/c++/14/optional:475
 #7 std::optional<PipeWireCursor>::operator-> at /usr/include/c++/14/optional:964
 #8 PWFrameBuffer::cursorPosition at /usr/src/debug/krfb-24.02.0-1.fc40.x86_64/framebuffers/pipewire/pw_framebuffer.cpp:514
 #9 RfbServerManager::updateScreens at /usr/src/debug/krfb-24.02.0-1.fc40.x86_64/krfb/rfbservermanager.cpp:156
 #10 QtPrivate::QSlotObjectBase::call at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/corelib/kernel/qobjectdefs_impl.h:433
 #11 doActivate<false> at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/corelib/kernel/qobject.cpp:4039
 #12 QMetaObject::activate at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/corelib/kernel/qobject.cpp:4099
 #13 QTimer::timeout at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/redhat-linux-build/src/corelib/Core_autogen/include/moc_qtimer.cpp:272
 #14 QObject::event at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/corelib/kernel/qobject.cpp:1459
 #15 QApplicationPrivate::notify_helper at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/widgets/kernel/qapplication.cpp:3296
 #16 QCoreApplication::notifyInternal2 at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/corelib/kernel/qcoreapplication.cpp:1121
 #17 QCoreApplication::sendEvent at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/corelib/kernel/qcoreapplication.cpp:1539
 #18 QTimerInfoList::activateTimers at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/corelib/kernel/qtimerinfo_unix.cpp:507
 #19 timerSourceDispatch at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/corelib/kernel/qeventdispatcher_glib.cpp:149
 #20 idleTimerSourceDispatch at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/corelib/kernel/qeventdispatcher_glib.cpp:196
 #22 g_main_context_dispatch_unlocked at ../glib/gmain.c:4152
 #23 g_main_context_iterate_unlocked.isra.0 at ../glib/gmain.c:4217
 #24 g_main_context_iteration at ../glib/gmain.c:4282
 #25 QEventDispatcherGlib::processEvents at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/corelib/kernel/qeventdispatcher_glib.cpp:393
 #26 QEventLoop::exec at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/corelib/global/qflags.h:34
 #27 QCoreApplication::exec at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/corelib/global/qflags.h:74
 #28 QGuiApplication::exec at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/gui/kernel/qguiapplication.cpp:1925
 #29 QApplication::exec at /usr/src/debug/qt6-qtbase-6.6.2-6.fc40.x86_64/src/widgets/kernel/qapplication.cpp:2574

Comment 1 80myol6v2 2024-03-22 00:51:29 UTC
Created attachment 2022992 [details]
File: proc_pid_status

Comment 2 80myol6v2 2024-03-22 00:51:30 UTC
Created attachment 2022993 [details]
File: maps

Comment 3 80myol6v2 2024-03-22 00:51:32 UTC
Created attachment 2022994 [details]
File: limits

Comment 4 80myol6v2 2024-03-22 00:51:34 UTC
Created attachment 2022995 [details]
File: open_fds

Comment 5 80myol6v2 2024-03-22 00:51:36 UTC
Created attachment 2022996 [details]
File: mountinfo

Comment 6 80myol6v2 2024-03-22 00:51:38 UTC
Created attachment 2022997 [details]
File: os_info

Comment 7 80myol6v2 2024-03-22 00:51:39 UTC
Created attachment 2022998 [details]
File: core_backtrace

Comment 8 80myol6v2 2024-03-22 00:51:41 UTC
Created attachment 2022999 [details]
File: dso_list

Comment 9 80myol6v2 2024-03-22 00:51:43 UTC
Created attachment 2023000 [details]
File: var_log_messages

Comment 10 80myol6v2 2024-03-22 00:51:45 UTC
Created attachment 2023001 [details]
File: cpuinfo

Comment 11 80myol6v2 2024-03-22 00:51:47 UTC
Created attachment 2023002 [details]
File: backtrace

Comment 12 80myol6v2 2024-03-22 00:51:48 UTC
Created attachment 2023003 [details]
File: environ

Comment 13 Yaroslav Sidlovsky 2024-04-20 09:19:53 UTC
I've just submitted patch that can fix this bug.
See patch attached to KDE bug https://bugs.kde.org/show_bug.cgi?id=472453.

Comment 14 Konstantin Vlasov 2024-04-27 22:11:28 UTC
1. Booted Fedora.
2. Opened Konsole, started Krfb.
3. Connected to this VNC session from outside (reproduced using Krdc in CentOS Linux, and RealVNC in Windows).
4. Inside the VNC session window, moved the windows around a little bit; opened Kwrite.
5. As soon as I press any key on the keyboard, Krfb crashes, and VNC session terminates. The key remains "pressed" inside the target system. If this was an alphanumeric key, it begins to be infinitely typed in the active window, until I myself press the same key. If this was a modifier key, like Shift, it becomes "stuck" and affects all the actions as if the real modifier key was being held pressed constantly.


reporter:       libreport-2.17.15
type:           CCpp
reason:         krfb killed by SIGABRT
journald_cursor: s=7a5d0ef34ed049a18d0925cf4a25ee6d;i=937a;b=953cdee3fa0c49d5a5bef0bb411436ef;m=cd949939;t=616cec3ba9a73;x=fe6f0791da6c89fd
executable:     /usr/bin/krfb
cmdline:        /usr/bin/krfb -qwindowtitle Krfb
cgroup:         0::/user.slice/user-1000.slice/user/app.slice/app-org.kde.krfb-ef09579b85d6440db1f2bb52ee250c44.scope
rootdir:        /
uid:            1000
kernel:         6.8.7-300.fc40.x86_64
package:        krfb-24.02.1-1.fc40
runlevel:       N 5
backtrace_rating: 4
crash_function: std::__glibcxx_assert_fail

Comment 15 Konstantin Vlasov 2024-04-28 01:27:02 UTC
After sending the report, I noticed there was an update ready, which also included krfb. I updated it, so now it has version 24.02.2, and tried the new version. It still crashes on the first keypress, but now the pressed key does not remain "stuck", it gets properly released.

Comment 16 Yaroslav Sidlovsky 2024-04-28 07:57:41 UTC
Patch is not included in this version. Ask package maintainer to include it or wait for new release.

Comment 17 Konstantin Vlasov 2024-04-30 13:03:46 UTC
Is there a place where I could get the ready patch? I'd like to try building the fixed version and checking it out, but I'm not sure what patch you mean. Rawhide has the same version, and I can't find the relevant commits in the krfb upstream.

Comment 19 Fedora Update System 2024-04-30 16:01:30 UTC
FEDORA-2024-4a61fcd745 (krfb-24.02.2-3.fc40) has been submitted as an update to Fedora 40.
https://bodhi.fedoraproject.org/updates/FEDORA-2024-4a61fcd745

Comment 20 Konstantin Vlasov 2024-04-30 21:48:15 UTC
Thanks!
I've installed that test version into Fedora 40, and while the crash indeed seems to be gone, unfortunately, it has other critical issues.

I'm not too much familiar with bug reporting policy here, sorry. So should I report them as separate bugs or continue here?

Comment 21 Fedora Update System 2024-04-30 22:09:51 UTC
FEDORA-2024-4a61fcd745 has been pushed to the Fedora 40 testing repository.
Soon you'll be able to install the update with the following command:
`sudo dnf upgrade --enablerepo=updates-testing --refresh --advisory=FEDORA-2024-4a61fcd745`
You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2024-4a61fcd745

See also https://fedoraproject.org/wiki/QA:Updates_Testing for more information on how to test updates.

Comment 22 Yaroslav Sidlovsky 2024-05-01 08:41:39 UTC
> I'm not too much familiar with bug reporting policy here, sorry. So should I report them as separate bugs or continue here?

Please report it as separate issue.
And search for existing bug reports: https://bugs.kde.org/buglist.cgi?component=general&list_id=2699446&product=krfb&resolution=---

Comment 23 serg 2024-05-04 09:51:34 UTC
1. Start krfb in wayland session.
2. Connect from another pc to krfb by krdc or remmina.
3. Try keyboard input in any text field.

Result: krfb crashes.
Expected: keyboard input works normally.


reporter:       libreport-2.17.15
type:           CCpp
reason:         krfb killed by SIGABRT
journald_cursor: s=ed77079b78f042abace4c37c8c4f13fb;i=2a33f6;b=e198d8a5ff0845ef81bc82638a7ac662;m=1b27fd080;t=6179d7acc2f21;x=592da076c9aee114
executable:     /usr/bin/krfb
cmdline:        /usr/bin/krfb -qwindowtitle Krfb
cgroup:         0::/user.slice/user-1000.slice/user/app.slice/app-org.kde.krfb-4c09a30040f44e898e119c3306004e88.scope
rootdir:        /
uid:            1000
kernel:         6.8.8-300.fc40.x86_64
package:        krfb-24.02.2-2.fc40
runlevel:       N 5
backtrace_rating: 4
crash_function: std::__glibcxx_assert_fail

Comment 24 Yaroslav Sidlovsky 2024-05-04 12:58:13 UTC
According to file /usr/src/debug/krfb-24.02.2-2.fc40.x86_64/framebuffers/pipewire/pw_framebuffer.cpp, line 515 my patch still is not applied.
I'll try to figure out why.

Comment 25 Yaroslav Sidlovsky 2024-05-04 12:58:55 UTC
I mean those lines is unchanged  but should (from krfb-24.02.2-2.fc40.x86_64):
```
QPoint PWFrameBuffer::cursorPosition()
{
    return d->cursor->position;
}
```

Comment 26 Yaroslav Sidlovsky 2024-05-04 13:01:52 UTC
Build 24.02.2-2.fc40 doesn't have this patch applied, while 24.02.2-3.fc40 got it.
So wait for version 24.02.2-3.fc40 in stable repo.

Comment 27 Fedora Update System 2024-05-08 03:31:29 UTC
FEDORA-2024-4a61fcd745 (krfb-24.02.2-3.fc40) has been pushed to the Fedora 40 stable repository.
If problem still persists, please make note of it in this bug report.


Note You need to log in before you can comment on or make changes to this bug.