[digikam] [Bug 326959] New: Digikam crash while scanning pictures for thumnails maintenance

[eldk]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #19 from [hidden email] ---
Created attachment 83267
  --> https://bugs.kde.org/attachment.cgi?id=83267&action=edit
starting script for launching digikam in local dev folder with GDB for testing,
debugging ..

Should be use in debug C++ config (debug properties panel) for launching
digikam with gdb in eclipse or standalone.
Source : http://www.baekhoej.com/node/195

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Gilles Caulier-4]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #20 from Gilles Caulier <[hidden email]> ---
e.longuemare,

I found something interesting. I see in your last backtrace some line about
color management.
Here, it's the same.

If i disable CM, crash is not reproducible. Can you confirm ?

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Gilles Caulier-4]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #21 from Gilles Caulier <[hidden email]> ---
Git commit ee22c2cc043b604698f5637d12437ecd909108a9 by Gilles Caulier.
Committed on 02/11/2013 at 08:36.
Pushed by cgilles into branch 'master'.

move event loop after thread host thread reassignement.

M  +2    -2    libs/progressmanager/progressmanager.cpp

http://commits.kde.org/digikam/ee22c2cc043b604698f5637d12437ecd909108a9

diff --git a/libs/progressmanager/progressmanager.cpp
b/libs/progressmanager/progressmanager.cpp
index 7f763b9..ac6a0f7 100644
--- a/libs/progressmanager/progressmanager.cpp
+++ b/libs/progressmanager/progressmanager.cpp
@@ -381,13 +381,13 @@ void ProgressManager::Private::removeItem(ProgressItem*
const t)
 ProgressManager::ProgressManager()
     : d(new Private)
 {
-    d->waitingLoop = new QEventLoop(this);
-
     if (thread() != QApplication::instance()->thread())
     {
         kWarning() << "Attention: ProgressManager was created from a thread.
Create it in the main thread!";
         moveToThread(QApplication::instance()->thread());
     }
+
+    d->waitingLoop = new QEventLoop();
 }

 ProgressManager::~ProgressManager()

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Gilles Caulier-4]

https://bugs.kde.org/show_bug.cgi?id=326959

Gilles Caulier <[hidden email]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|general                     |Maintenance

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[eldk]

https://bugs.kde.org/show_bug.cgi?id=326959

[hidden email] changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|Maintenance                 |general

--- Comment #22 from [hidden email] ---
Hello,

I launch the test now and see if it's the same :
1 - disable color management in digikam conf
2 - then launch a task (thumbnails maintenance)
3 - and reduce and restore windows

Thank you.

Eric

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Gilles Caulier-4]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #23 from Gilles Caulier <[hidden email]> ---
To disable CM, just go to digiKam CM setup panel and turn off lead CM option.
No need to touch digikamrc file. digiKam setup will do it for you...

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Gilles Caulier-4]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #24 from Gilles Caulier <[hidden email]> ---
Marcel,

My first investigations about QEventLoop has been a wrong way.

It sound like problem is about CM enabled on computer. This setting is used in
ThumbnailLoadThread. As instance is created in a separated thread and not main
application thread, through a ThreadWeaver::Job, it try to access on X11 API
which crash due to non re-entrance...

I see in ThumbnailLoadThread::Private::createLoadingDescription() that so
bracket play with CM. I forced displayingWidget to null in ThumbnailLoadThread
properties, but i think it's not enough to disable CM in ThumbnailLoadThread.

Perhaps it's the wrong way, and i must to move ThumbnailLoadThread instance
outside ThreadWeaver::Job ?

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Gilles Caulier-4]

https://bugs.kde.org/show_bug.cgi?id=326959

Gilles Caulier <[hidden email]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|general                     |Maintenance

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[eldk]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #25 from [hidden email] ---
Gilles,

For the moment I try but can't reproduce with cm enable or disable.

What have been changed since yesterday :

I have removed my second display monitor (it will be back in a few hour).
I look at digikamrc and the file was corrupted (but can't reproduce even with
this corrupted RC file).

What have not changed :
Digikam can't be quit without ctrl+ c
I have not updated and rebuild digikam (same version which produced the crash
yesterday)

So, is it something with Color Management and dual screen ?
Or is it something with some files that aren't process at the moment ?

So I relaunch a thumbnail scan (all collection, all tag) and wait, then try to
reproduce.

component (no change) :
digiKam version 4.0.0-beta1
Codec PGF parallélisé: Non
Composant graphique Marble: 0.16.2 (stable version)
Demosaic GPL2 pack support: Inconnu
Demosaic GPL3 pack support: Inconnu
Dématriçage parallélisé: Inconnu
Exiv2 peut écrire dans un fichier JP2: Oui
Exiv2 peut écrire dans un fichier JPEG: Oui
Exiv2 peut écrire dans un fichier PGF: Oui
Exiv2 peut écrire dans un fichier PNG: Oui
Exiv2 peut écrire dans un fichier TIFF: Oui
Exiv2 prend en charge les métadonnées XMP: Oui
LibCImg: 130
LibEigen: 3.0.5
LibExiv2: 0.22
LibJPEG: 80
LibJasper: 1.900.1
LibKDE: 4.11.2
LibKExiv2: 2.3.1
LibKGeoMap: 2.0.0
LibKdcraw: 2.4.0
LibLCMS: 2040
LibLensFun: 0.2.7-0
LibPGF: 6.12.27 - bibliothèque interne
LibPNG: 1.2.46
LibQt: 4.8.2
LibRaw: 0.15.4
LibTIFF: LIBTIFF, Version 3.9.5 Copyright (c) 1988-1996 Sam Leffler Copyright
(c) 1991-1996 Silicon Graphics, Inc.
Prise en charge du codec RawSpeed: Inconnu
LibGphoto2: 2.4.14
LibKface: 3.0.0
LibKipi: 2.1.0
LibOpenCV: 2.4.5
Modules externes KIPI: 3.5.0
Moteur de base de données: QMYSQL
Serveur interne de base de données: Non

Digikam (no change):
Date de compilation : Oct 31 2013 (cible : debugfull)
Rév. : ea2847ec161df82f8d64e70afdbc5d12feb7708e

Database stats :
digiKam version 4.0.0-beta1
Images:
BMP: 5
GIF: 56
JP2: 16
JPG: 22735
PNG: 130
RAW-CR2: 3464
XCF: 209
total: 26615
:
:
Vidéos:
AVI: 29
MOV: 186
MP4: 43
MPEG: 2
WMV: 1
total: 261
:
:
Total des éléments: 26876
Albums: 386
Étiquettes: 495
Moteur de base de données: QMYSQL
Serveur interne de base de données: Non

Eric.

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[eldk]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #26 from [hidden email] ---
Created attachment 83282
  --> https://bugs.kde.org/attachment.cgi?id=83282&action=edit
diff between old digikamrc and new one

I made a mistake in thumbnail database name (digikam main database was the good
one). Use the production thumbnail database for the test this morning :
digikam_thumbnail (test one is digikam_thumbnail_clone). There are over
differences.

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[eldk]

https://bugs.kde.org/show_bug.cgi?id=326959

[hidden email] changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #83282|0                           |1
        is obsolete|                            |

--- Comment #27 from [hidden email] ---
Created attachment 83285
  --> https://bugs.kde.org/attachment.cgi?id=83285&action=edit
diff between old digikamrc and new one

diff between old digikamrc and new one I made a mistake in thumbnail database
name (digikam main database was the good one). Use the production thumbnail
database for the test this morning : digikam_thumbnail (test one is
digikam_thumbnail_clone). There are over differences.

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Gilles Caulier-4]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #28 from Gilles Caulier <[hidden email]> ---
Do you have a dual screen. For me, it's my case.

I calibrated one screen and as both are the same, i set color profile file in
digiKam CM settings (there is no way that i know under linux to assign a
profile for each monitor)

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[eldk]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #29 from [hidden email] ---
Yes Gilles.
I have lend it to a friend. It will be back this afternoon.
I will try to reproduce.

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[eldk]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #30 from [hidden email] ---
Yes I use dual screen. But not this morning.

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Gilles Caulier-4]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #31 from Gilles Caulier <[hidden email]> ---
And you want mean that using only one screen and to enable CM, stressing
digiKam to reduce/expand main window during thumbnails task do not crash it ?

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[eldk]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #32 from [hidden email] ---
Yes

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[eldk]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #33 from [hidden email] ---
Gilles,

I reconnect my second display monitor.
Launch digikam.
Launch thumbnail scan
#####
then clicking on reduce and restore
################
digikam(4595)/digikam (core) Digikam::MaintenanceMngr::stage1: stage1
digikam(4595)/digikam (core) Digikam::MaintenanceMngr::stage2: stage2
digikam(4595)/KDCRAW KDcrawIface::RActionThreadBase::RActionThreadBase:
Starting Main Thread
[Thread 0x95594b40 (LWP 4637) exited]
[Thread 0x94592b40 (LWP 4639) exited]
[Thread 0x94d93b40 (LWP 4638) exited]
[New Thread 0x94d93b40 (LWP 4657)]
digikam(4595)/KDCRAW KDcrawIface::RActionThreadBase::run: In action thread Run
[New Thread 0x94592b40 (LWP 4658)]
[New Thread 0x95594b40 (LWP 4659)]
##########
clic and clic waiting scan is really launch - no crash
###################
scan is really launch - then crash
################
digikam(4595)/digikam (core) Digikam::DImg::load:
"/media/DiskphotosTests/age_taille_mortalite_maturite_cabillaud.jpg"  : JPEG
file identified
[New Thread 0x93d91b40 (LWP 4660)]
digikam(4595)/digikam (core) Digikam::DImg::load:
"/media/DiskphotosTests/age_taille_mortalite_maturite_cabillaud.jpg"  : JPEG
file identified
[xcb] Unknown request in queue while dequeuing
[xcb] Most likely this is a multi-threaded client and XInitThreads has not been
called
[xcb] Aborting, sorry about that.
digikam: ../../src/xcb_io.c :179 : dequeue_pending_request:  l'assertion «
!xcb_xlib_unknown_req_in_deq » a échoué.

Program received signal SIGABRT, Aborted.
[Switching to Thread 0x94592b40 (LWP 4658)]
0xb7fdd424 in __kernel_vsyscall ()

Try now with only one display monitor.

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[eldk]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #34 from [hidden email] ---
###########
digikam(4595)/digikam (core) Digikam::MaintenanceMngr::stage1: stage1
digikam(4595)/digikam (core) Digikam::MaintenanceMngr::stage2: stage2
digikam(4595)/KDCRAW KDcrawIface::RActionThreadBase::RActionThreadBase:
Starting Main Thread
[Thread 0x95594b40 (LWP 4637) exited]
[Thread 0x94592b40 (LWP 4639) exited]
[Thread 0x94d93b40 (LWP 4638) exited]
[New Thread 0x94d93b40 (LWP 4657)]
digikam(4595)/KDCRAW KDcrawIface::RActionThreadBase::run: In action thread Run
[New Thread 0x94592b40 (LWP 4658)]
[New Thread 0x95594b40 (LWP 4659)]
digikam(4595)/digikam (core) Digikam::DImg::load:
"/media/DiskphotosTests/age_taille_mortalite_maturite_cabillaud.jpg"  : JPEG
file identified
[New Thread 0x93d91b40 (LWP 4660)]
digikam(4595)/digikam (core) Digikam::DImg::load:
"/media/DiskphotosTests/age_taille_mortalite_maturite_cabillaud.jpg"  : JPEG
file identified
[xcb] Unknown request in queue while dequeuing
[xcb] Most likely this is a multi-threaded client and XInitThreads has not been
called
[xcb] Aborting, sorry about that.
digikam: ../../src/xcb_io.c :179 : dequeue_pending_request:  l'assertion «
!xcb_xlib_unknown_req_in_deq » a échoué.

Program received signal SIGABRT, Aborted.
[Switching to Thread 0x94592b40 (LWP 4658)]
0xb7fdd424 in __kernel_vsyscall ()
#####################
between those lines nothing as been cut (complete log)

So as soon as scan start scanning pictures, digikam crash

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[eldk]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #35 from [hidden email] ---
Now with only one main monitor no crash.

It's look like digikam quit normally when you launch a process and quit :

digikam(4782)/digikam (core) Digikam::DImg::load: "/media/DiskphotosTests/14
juillet 2006 - Tardinghem/tardinghem_2006 022.jpg"  : JPEG file identified
digikam(4782)/KEXIV2 KExiv2Iface::KExiv2::getImageOrientation: Orientation =>
Exif.Image.Orientation =>  1
#######################
Clic on quit while scan process is in progress
###################################
digikam(4782)/KDCRAW KDcrawIface::RActionThreadBase::cancel: Cancel Main Thread
digikam(4782)/digikam (core) Digikam::MaintenanceMngr::stage3: stage3
digikam(4782)/digikam (core) Digikam::MaintenanceMngr::stage4: stage4
digikam(4782)/digikam (core) Digikam::MaintenanceMngr::stage5: stage5
digikam(4782)/digikam (core) Digikam::MaintenanceMngr::stage6: stage6
digikam(4782)/digikam (core) Digikam::MaintenanceMngr::stage7: stage7
digikam(4782)/KDCRAW KDcrawIface::RActionThreadBase::run: Exiting Action Thread
[Thread 0x94d11b40 (LWP 4844) exited]
digikam(4782)/KDCRAW KDcrawIface::RWeaverObserver::slotThreadSuspended:
RWeaverObserver: thread  1  suspended
digikam(4782)/digikam (core) Digikam::MaintenanceThread::slotThreadFinished:
List of Pending Jobs is empty
digikam(4782)/KGEOMAP KGeoMap::ItemMarkerTiler::slotSourceModelReset: ----
[Thread 0xab654b40 (LWP 4785) exited]
digikam(4782)/KDCRAW KDcrawIface::RActionThreadBase::cancel: Cancel Main Thread
Thread::requestAbort: not running.
digikam(4782)/KDCRAW KDcrawIface::RActionThreadBase::~RActionThreadBase:
calling action thread destructor
digikam(4782)/KDCRAW KDcrawIface::RActionThreadBase::cancel: Cancel Main Thread
Thread::requestAbort: not running.
[Thread 0x94510b40 (LWP 4845) exited]
digikam(4782)/KDCRAW KDcrawIface::RActionThreadBase::cancel: Cancel Main Thread
digikam(4782)/KDCRAW KDcrawIface::RActionThreadBase::~RActionThreadBase:
calling action thread destructor
digikam(4782)/KDCRAW KDcrawIface::RActionThreadBase::cancel: Cancel Main Thread
[Thread 0xaa2ffb40 (LWP 4790) exited]
[Thread 0x93d0fb40 (LWP 4846) exited]
[Thread 0x95512b40 (LWP 4847) exited]
[Thread 0xaacffb40 (LWP 4786) exited]
#########################
Digikam quit (close) normally, not hang indefinitely
####################################
[Inferior 1 (process 4782) exited normally]

I try now with dual monitors and no color management.

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[eldk]

https://bugs.kde.org/show_bug.cgi?id=326959

--- Comment #36 from [hidden email] ---
So,

With only one monitor, CM check, no crash but if nothing is in progress digikam
doesn't quit.
With only one monitor, CM check and a task in progress digikam quit.
With two monitor, CM uncheck, no crash and quit normally while a task is in
progress.
With two monitor, CM check, digikam crash on reduce restore, and crash or don't
quit while a task is in progress :

####
quit while digikam scan pictures for thumbnails
####################
digikam(4951)/digikam (core) Digikam::DImg::load: "/media/DiskphotosTests/photo
8 288.jpg"  : JPEG file identified
digikam(4951)/KEXIV2 KExiv2Iface::KExiv2::getImageOrientation: Orientation =>
Exif.Image.Orientation =>  1
digikam(4951)/KDCRAW KDcrawIface::RWeaverObserver::slotThreadBusy:
RWeaverObserver: thread  1  busy
digikam(4951)/KEXIV2 KExiv2Iface::KExiv2::Private::printExiv2ExceptionError:
Cannot load metadata from file   (Error # 11 :
/media/DiskphotosTests/planche_poisson_IFREMER.xcf: The file contains data of
an unknown image type
digikam(4951)/KDCRAW KDcrawIface::KDcraw::loadEmbeddedPreview: Failed to load
embedded RAW preview
digikam(4951)/digikam (core) Digikam::DImg::load:
"/media/DiskphotosTests/planche_poisson_IFREMER.xcf"  : QIMAGE file identified
##################
QUIT
####################
digikam(4951)/KDCRAW KDcrawIface::RActionThreadBase::cancel: Cancel Main Thread
digikam(4951)/digikam (core) Digikam::MaintenanceMngr::stage3: stage3
digikam(4951)/digikam (core) Digikam::MaintenanceMngr::stage4: stage4
digikam(4951)/digikam (core) Digikam::MaintenanceMngr::stage5: stage5
digikam(4951)/digikam (core) Digikam::MaintenanceMngr::stage6: stage6
digikam(4951)/digikam (core) Digikam::MaintenanceMngr::stage7: stage7
digikam(4951)/KDCRAW KDcrawIface::RActionThreadBase::run: Exiting Action Thread
[Thread 0x93d91b40 (LWP 5013) exited]
digikam(4951)/KDCRAW KDcrawIface::RWeaverObserver::slotThreadSuspended:
RWeaverObserver: thread  1  suspended
digikam(4951)/digikam (core) Digikam::MaintenanceThread::slotThreadFinished:
List of Pending Jobs is empty
digikam(4951)/KGEOMAP KGeoMap::ItemMarkerTiler::slotSourceModelReset: ----
[Thread 0xab654b40 (LWP 4953) exited]
digikam(4951)/KDCRAW KDcrawIface::RActionThreadBase::cancel: Cancel Main Thread
Thread::requestAbort: not running.
digikam(4951)/KDCRAW KDcrawIface::RActionThreadBase::~RActionThreadBase:
calling action thread destructor
digikam(4951)/KDCRAW KDcrawIface::RActionThreadBase::cancel: Cancel Main Thread
Thread::requestAbort: not running.
[Thread 0x94592b40 (LWP 5014) exited]
digikam(4951)/KDCRAW KDcrawIface::RActionThreadBase::cancel: Cancel Main Thread
digikam(4951)/KDCRAW KDcrawIface::RActionThreadBase::~RActionThreadBase:
calling action thread destructor
digikam(4951)/KDCRAW KDcrawIface::RActionThreadBase::cancel: Cancel Main Thread
[Thread 0xaa2ffb40 (LWP 4958) exited]
digikam(4951)/digikam (core) Digikam::KMemoryInfo::update: Platform identified
:  "LINUX"
digikam(4951)/digikam (core) Digikam::KMemoryInfo::bytes: AvailableRam:
141615104  (cache:  640679936 )
digikam(4951)/digikam (core) Digikam::KMemoryInfo::bytes: AvailableSwap:
1934073856
Fatal Error: Accessed global static 'FileReadWriteLockStaticPrivate
*static_d()' after destruction. Defined at
/home/eric/digikamgit2/dkgitdigikamsoftwarecompilation/core/libs/threadimageio/filereadwritelock.cpp:385

Program received signal SIGABRT, Aborted.
[Switching to Thread 0x95594b40 (LWP 5016)]
0xb7fdd424 in __kernel_vsyscall ()

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel