Arch Linux

Please read this before reporting a bug:

Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.

REPEAT: Do NOT report bugs for outdated packages!

FS#36698 - [digikam] segfaults on start

Attached to Project: Arch Linux
Opened by phanisvara das (phani00) - Tuesday, 27 August 2013, 22:36 GMT
Last edited by Ronald van Haren (pressh) - Monday, 28 October 2013, 13:38 GMT
Task Type Bug Report
Category Packages: Extra
Status Closed
Assigned To Ronald van Haren (pressh)
Architecture x86_64
Severity Low
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 1
Private No


i haven't used digikam in a while, but trying to start it now results in a seg. fault. i'm not sure if i have used digikam since KDE 4.11 made it into extra, probably not. previously i used digikam-dev from AUR, but that also didn't work today, so i installed the version from [extra]. output from the command line:

[phani@laptop ~]$ digikam
Fontconfig warning: "/etc/fonts/conf.d/50-user.conf", line 14: reading configurations from ~/.fonts.conf is deprecated.
QSqlDatabasePrivate::removeDatabase: connection 'ConnectionTest' is still in use, all queries will cease to work.
"/org/freedesktop/UDisks2/drives/ST9750420AS_5WS34Y4W" : property "DeviceNumber" does not exist
"/org/freedesktop/UDisks2/drives/ST9750420AS_5WS34Y4W" : property "Device" does not exist
KCrash: Application 'digikam' crashing...
KCrash: Attempting to start /usr/lib/kde4/libexec/drkonqi from kdeinit

[1]+ Stopped digikam
[phani@laptop ~]$ QSocketNotifier: Invalid socket 18 and type 'Read', disabling...

[1]+ Exit 253 digikam
[phani@laptop ~]$
This task depends upon

Closed by  Ronald van Haren (pressh)
Monday, 28 October 2013, 13:38 GMT
Reason for closing:  Works for me
Additional comments about closing:  user is happy with current git version installed on his machine and doesn't want to follow up on this bug at this moment.
Comment by phanisvara das (phani00) - Wednesday, 28 August 2013, 18:50 GMT
here is what drkonqi from KDE has to say, in case that helps. seems to make more sense than the CLI output above:

Application: digiKam (digikam), signal: Segmentation fault
Using host libthread_db library "/usr/lib/".
[Current thread is 1 (Thread 0x7fd5927ff980 (LWP 3980))]

Thread 3 (Thread 0x7fd56884e700 (LWP 3984)):
#0 0x00007fd589e4a03f in pthread_cond_wait@@GLIBC_2.3.2 () from /usr/lib/
#1 0x00007fd58cee7266 in QWaitCondition::wait(QMutex*, unsigned long) () from /usr/lib/
#2 0x00000000005d35ae in ?? ()
#3 0x00007fd58cee6d8f in ?? () from /usr/lib/
#4 0x00007fd589e460a2 in start_thread () from /usr/lib/
#5 0x00007fd58c10ea2d in clone () from /usr/lib/

Thread 2 (Thread 0x7fd563fff700 (LWP 3985)):
#0 0x00007fd589e49500 in __pthread_mutex_unlock_usercnt () from /usr/lib/
#1 0x00007fd5849a6151 in g_mutex_unlock () from /usr/lib/
#2 0x00007fd5849667b8 in g_main_context_prepare () from /usr/lib/
#3 0x00007fd584966fa3 in ?? () from /usr/lib/
#4 0x00007fd58496718c in g_main_context_iteration () from /usr/lib/
#5 0x00007fd58d013b46 in QEventDispatcherGlib::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) () from /usr/lib/
#6 0x00007fd58cfe5b1f in QEventLoop::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) () from /usr/lib/
#7 0x00007fd58cfe5e15 in QEventLoop::exec(QFlags<QEventLoop::ProcessEventsFlag>) () from /usr/lib/
#8 0x00007fd58cee46ef in QThread::exec() () from /usr/lib/
#9 0x00007fd58cfc72c3 in ?? () from /usr/lib/
#10 0x00007fd58cee6d8f in ?? () from /usr/lib/
#11 0x00007fd589e460a2 in start_thread () from /usr/lib/
#12 0x00007fd58c10ea2d in clone () from /usr/lib/

Thread 1 (Thread 0x7fd5927ff980 (LWP 3980)):
[KCrash Handler]
#5 0x000000000139f290 in ?? ()
#6 0x00007fd586f1676a in cmsReadTag () from /usr/lib/
#7 0x00007fd58fe3587a in ?? () from /usr/lib/
#8 0x00007fd58fea0e78 in Digikam::IccProfile::description() () from /usr/lib/
#9 0x00007fd58feb788a in Digikam::IccSettings::loadAllProfilesProperties() () from /usr/lib/
#10 0x0000000000596044 in ?? ()
#11 0x0000000000491a2a in ?? ()
#12 0x00007fd58c04bbc5 in __libc_start_main () from /usr/lib/
#13 0x0000000000494351 in _start ()
Comment by phanisvara das (phani00) - Sunday, 27 October 2013, 21:08 GMT
  • Field changed: Percent Complete (100% → 0%)
digikam 3.4 still segfaults with the same or similar message :(
Comment by phanisvara das (phani00) - Sunday, 27 October 2013, 21:28 GMT
um, i've seen that the latest digikam version seems to work for others; there have been complaints on ML or forum that this or that doesn't work. for me it doesn't even start up, with all DBs and config files removed, so i guess it's something screwed up on my machine. i don't have any other problems, but quite a few AUR packages, and when i looked at digikam's pactree, i sort of lost interest. as long as it's only me, why bother? i built digikam from their GIT repo just now, and everything seems to work fine, that's good enough for me.

so unless somebody else has similar problems (and it doesn't look that way), i propose to close this again...
Comment by Ronald van Haren (pressh) - Monday, 28 October 2013, 13:25 GMT
Not sure if the stable and git version use the same config directory, but do you have any color profiles set for the stable (non-working) digikam?

If I'm reading the backtrace correctly digikam is crashing on lcms2, which is used for managing color profiles.
Comment by phanisvara das (phani00) - Monday, 28 October 2013, 13:31 GMT
i don't have the arch version of digikam installed anymore; installed the GIT version system-wide. [edit: yes, they use the same config under ~/.kde4]

i don't remember setting any color profiles; but anyway, after deleting all config files, DB, etc., shouldn't those color profiles be gone, too?

as i said, i seem to be the only one with this problem, and right now i'm not too eager to follow this up. unless there's some other reason than my installation, i'd rather wait for the next digikam release and try again then.
Comment by Ronald van Haren (pressh) - Monday, 28 October 2013, 13:37 GMT
yes, suppose that after deleting the config files any color profile preference should be gone.

I'll close this for now if you're happy with the git version you have installed now. Feel free to re-open this again in the future, although it indeed seems to be something specific to your machine/installation.