[digikam] [Bug 381877] New: Digikam start allocating all memory when scan a new collection.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

[hidden email] changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[hidden email]

--- Comment #19 from [hidden email] ---
The situation is quite similar with mine, described on DK user mailing list
("memory leak : valgrind backtrace", from Paulux).
If it's related, gdb should show nothing.

On my system, downgrading qtAV to version 1.11 fixed the problem (workaround).

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #20 from [hidden email] ---
QtAV is to handle video files. Do you have any video in your collection while
scanning which can explain why downgrading QtAv can fix the problem ?

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #21 from [hidden email] ---
I've installed digikam on a different PC (with ArchLinux) and it works.
The problems seems related to my installation.

So I must rebuild digikam in debug mode, and I'll do the same job with
> gdb digikam

?

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #22 from [hidden email] ---
(In reply to caulier.gilles from comment #20)
> QtAV is to handle video files. Do you have any video in your collection
> while scanning which can explain why downgrading QtAv can fix the problem ?

Yes I've a pair of MOV videos.

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #23 from [hidden email] ---
And if you scan a collection without video files, the problem is not
reproducible ?

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #24 from [hidden email] ---
(In reply to nvlplx from comment #19)
> The situation is quite similar with mine, described on DK user mailing list
> ("memory leak : valgrind backtrace", from Paulux).
> If it's related, gdb should show nothing.
>
> On my system, downgrading qtAV to version 1.11 fixed the problem
> (workaround).

I confirm!
I've downgraded to qtAV 1.11 (from 1.12) and digikam is still working.

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #25 from [hidden email] ---
When you downgrade qtav, ffmpeg packages are also downgraded or it still the
same for both QtAV versions ?

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #26 from [hidden email] ---
(In reply to caulier.gilles from comment #25)
> When you downgrade qtav, ffmpeg packages are also downgraded or it still the
> same for both QtAV versions ?

ffmpeg is the same (ffmpeg 1:3.3.2-1).

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #27 from [hidden email] ---
Ok, now we will investiguate the memory leak with valgrind. Start digiKam from
a console like this :

valgrind --tool=memcheck --leak-check=full --error-limit=no digikam

This will run DK in Valgrind which will catch the memory corruptions. This slow
down digiKam.

Just report here all the trace from the console until the end of digiKam
session.

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #28 from [hidden email] ---
Here are some extracts of valgrind outputs.
I don't have more, because I've not reinstalled qtAV 1.12, sorry for that.

[...]
==1397== 576 bytes in 1 blocks are possibly lost in loss record 46,312 of
49,097
==1397==    at 0x4C2DF05: calloc (in
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==1397==    by 0x40119F1: allocate_dtv (in /usr/lib/ld-2.25.so)
==1397==    by 0x401236D: _dl_allocate_tls (in /usr/lib/ld-2.25.so)
==1397==    by 0xE333DC9: pthread_create@@GLIBC_2.2.5 (in
/usr/lib/libpthread-2.25.so)
==1397==    by 0x879B875: QThread::start(QThread::Priority) (in
/usr/lib/libQt5Core.so.5.9.1)
==1397==    by 0x1232DC22:
QtAV::VideoFrameExtractor::VideoFrameExtractor(QObject*) (in
/usr/lib/libQtAV.so.1.12.0)
==1397==    by 0x5EF8824: Digikam::VideoThumbnailer::VideoThumbnailer(QObject*)
(in /usr/lib/libdigikamcore.so.5.6.0)
==1397==    by 0x5EF9DE4:
Digikam::VideoThumbnailerJob::VideoThumbnailerJob(QObject*) (in
/usr/lib/libdigikamcore.so.5.6.0)
==1397==    by 0x5EEAA06:
Digikam::ThumbnailLoadThread::ThumbnailLoadThread(QObject*) (in
/usr/lib/libdigikamcore.so.5.6.0)
==1397==    by 0x55A2A3D: ??? (in /usr/lib/libdigikamgui.so.5.6.0)
==1397==    by 0x559C53D: ??? (in /usr/lib/libdigikamgui.so.5.6.0)
==1397==    by 0x5106204: ??? (in /usr/lib/libdigikamgui.so.5.6.0)
==1397==  
==1397== 576 bytes in 1 blocks are possibly lost in loss record 46,313 of
49,097
==1397==    at 0x4C2DF05: calloc (in
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==1397==    by 0x40119F1: allocate_dtv (in /usr/lib/ld-2.25.so)
==1397==    by 0x401236D: _dl_allocate_tls (in /usr/lib/ld-2.25.so)
==1397==    by 0xE333DC9: pthread_create@@GLIBC_2.2.5 (in
/usr/lib/libpthread-2.25.so)
==1397==    by 0x879B875: QThread::start(QThread::Priority) (in
/usr/lib/libQt5Core.so.5.9.1)
==1397==    by 0x13B696B5: QNetworkConfigurationManagerPrivate::initialize()
(in /usr/lib/libQt5Network.so.5.9.1)
==1397==    by 0x13B638EC: qNetworkConfigurationManagerPrivate() (in
/usr/lib/libQt5Network.so.5.9.1)
==1397==    by 0x13B63986:
QNetworkConfigurationManager::QNetworkConfigurationManager(QObject*) (in
/usr/lib/libQt5Network.so.5.9.1)
==1397==    by 0x174C1FA4: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x89A9241: QObject::event(QEvent*) (in
/usr/lib/libQt5Core.so.5.9.1)
==1397==    by 0x788746B: QApplicationPrivate::notify_helper(QObject*, QEvent*)
(in /usr/lib/libQt5Widgets.so.5.9.1)
==1397==    by 0x788ECF3: QApplication::notify(QObject*, QEvent*) (in
/usr/lib/libQt5Widgets.so.5.9.1)

[...]
==1397== 2,304 bytes in 4 blocks are possibly lost in loss record 48,185 of
49,097
==1397==    at 0x4C2DF05: calloc (in
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==1397==    by 0x40119F1: allocate_dtv (in /usr/lib/ld-2.25.so)
==1397==    by 0x401236D: _dl_allocate_tls (in /usr/lib/ld-2.25.so)
==1397==    by 0xE333DC9: pthread_create@@GLIBC_2.2.5 (in
/usr/lib/libpthread-2.25.so)
==1397==    by 0x450FFF4B: ??? (in /usr/lib/xorg/modules/dri/radeonsi_dri.so)
==1397==    by 0x451003BC: ??? (in /usr/lib/xorg/modules/dri/radeonsi_dri.so)
==1397==    by 0x454719F3: ??? (in /usr/lib/xorg/modules/dri/radeonsi_dri.so)
==1397==    by 0x4552B57E: amdgpu_winsys_create (in
/usr/lib/xorg/modules/dri/radeonsi_dri.so)
==1397==    by 0x44E50EDE: ??? (in /usr/lib/xorg/modules/dri/radeonsi_dri.so)
==1397==    by 0x4517FEE9: ??? (in /usr/lib/xorg/modules/dri/radeonsi_dri.so)
==1397==    by 0x4517BC1F: ??? (in /usr/lib/xorg/modules/dri/radeonsi_dri.so)
==1397==    by 0x43B77FA4: ??? (in /usr/lib/libGLX_mesa.so.0.0.0)

[...]
==1397== 4,360 bytes in 1 blocks are possibly lost in loss record 48,482 of
49,097
==1397==    at 0x4C2BEEF: malloc (in
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==1397==    by 0x27BA3F90: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BA3502: sqlite3Malloc (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BB1A53: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BB22C5: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BAC9FF: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27B41A26: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27B46A49: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27B48410: sqlite3BtreeMovetoUnpacked (in
/usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BD604D: sqlite3VdbeExec (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BD7F9E: sqlite3_step (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x5019375C: ??? (in
/usr/lib/qt/plugins/sqldrivers/libqsqlite.so)
==1397==  
==1397== 4,360 bytes in 1 blocks are possibly lost in loss record 48,483 of
49,097
==1397==    at 0x4C2BEEF: malloc (in
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==1397==    by 0x27BA3F90: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BA3502: sqlite3Malloc (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BB1A53: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BB22C5: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BAC9FF: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27B41A26: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27B48BB5: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27B486D6: sqlite3BtreeMovetoUnpacked (in
/usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BD92FD: ??? (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BD24AE: sqlite3VdbeExec (in /usr/lib/libsqlite3.so.0.8.6)
==1397==    by 0x27BD7F9E: sqlite3_step (in /usr/lib/libsqlite3.so.0.8.6)


[...]
==1397== Use of uninitialised value of size 8
==1397==    at 0x16399805: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x16399B9A: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1612DBCA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1610F5AC: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161291DA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161295E3: JSC::Heap::collect(JSC::HeapOperation) (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639C080: JSC::GCActivityCallback::doWork() (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639D050: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x89A91FA: QObject::event(QEvent*) (in
/usr/lib/libQt5Core.so.5.9.1)
==1397==    by 0x788746B: QApplicationPrivate::notify_helper(QObject*, QEvent*)
(in /usr/lib/libQt5Widgets.so.5.9.1)
==1397==    by 0x788ECF3: QApplication::notify(QObject*, QEvent*) (in
/usr/lib/libQt5Widgets.so.5.9.1)
==1397==    by 0x8979A27: QCoreApplication::notifyInternal2(QObject*, QEvent*)
(in /usr/lib/libQt5Core.so.5.9.1)
==1397==  
==1397== Conditional jump or move depends on uninitialised value(s)
==1397==    at 0x1639980C: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x16399B9A: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1612DBCA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1610F5AC: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161291DA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161295E3: JSC::Heap::collect(JSC::HeapOperation) (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639C080: JSC::GCActivityCallback::doWork() (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639D050: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x89A91FA: QObject::event(QEvent*) (in
/usr/lib/libQt5Core.so.5.9.1)
==1397==    by 0x788746B: QApplicationPrivate::notify_helper(QObject*, QEvent*)
(in /usr/lib/libQt5Widgets.so.5.9.1)
==1397==    by 0x788ECF3: QApplication::notify(QObject*, QEvent*) (in
/usr/lib/libQt5Widgets.so.5.9.1)
==1397==    by 0x8979A27: QCoreApplication::notifyInternal2(QObject*, QEvent*)
(in /usr/lib/libQt5Core.so.5.9.1)
==1397==  
==1397== Conditional jump or move depends on uninitialised value(s)
==1397==    at 0x1639989B: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x16399B9A: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1612DBCA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1610F5AC: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161291DA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161295E3: JSC::Heap::collect(JSC::HeapOperation) (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639C080: JSC::GCActivityCallback::doWork() (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639D050: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x89A91FA: QObject::event(QEvent*) (in
/usr/lib/libQt5Core.so.5.9.1)
==1397==    by 0x788746B: QApplicationPrivate::notify_helper(QObject*, QEvent*)
(in /usr/lib/libQt5Widgets.so.5.9.1)
==1397==    by 0x788ECF3: QApplication::notify(QObject*, QEvent*) (in
/usr/lib/libQt5Widgets.so.5.9.1)
==1397==    by 0x8979A27: QCoreApplication::notifyInternal2(QObject*, QEvent*)
(in /usr/lib/libQt5Core.so.5.9.1)
==1397==  
==1397== Use of uninitialised value of size 8
==1397==    at 0x1612C383: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x16399290: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x16399B9A: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1612DBCA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1610F5AC: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161291DA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161295E3: JSC::Heap::collect(JSC::HeapOperation) (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639C080: JSC::GCActivityCallback::doWork() (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639D050: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x89A91FA: QObject::event(QEvent*) (in
/usr/lib/libQt5Core.so.5.9.1)
==1397==    by 0x788746B: QApplicationPrivate::notify_helper(QObject*, QEvent*)
(in /usr/lib/libQt5Widgets.so.5.9.1)
==1397==    by 0x788ECF3: QApplication::notify(QObject*, QEvent*) (in
/usr/lib/libQt5Widgets.so.5.9.1)
==1397==  
==1397== Conditional jump or move depends on uninitialised value(s)
==1397==    at 0x1612C389: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x16399290: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x16399B9A: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1612DBCA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1610F5AC: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161291DA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161295E3: JSC::Heap::collect(JSC::HeapOperation) (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639C080: JSC::GCActivityCallback::doWork() (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639D050: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x89A91FA: QObject::event(QEvent*) (in
/usr/lib/libQt5Core.so.5.9.1)
==1397==    by 0x788746B: QApplicationPrivate::notify_helper(QObject*, QEvent*)
(in /usr/lib/libQt5Widgets.so.5.9.1)
==1397==    by 0x788ECF3: QApplication::notify(QObject*, QEvent*) (in
/usr/lib/libQt5Widgets.so.5.9.1)
==1397==  
==1397== Use of uninitialised value of size 8
==1397==    at 0x1612C3D8: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x16399290: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x16399B9A: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1612DBCA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1610F5AC: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161291DA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161295E3: JSC::Heap::collect(JSC::HeapOperation) (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639C080: JSC::GCActivityCallback::doWork() (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639D050: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x89A91FA: QObject::event(QEvent*) (in
/usr/lib/libQt5Core.so.5.9.1)
==1397==    by 0x788746B: QApplicationPrivate::notify_helper(QObject*, QEvent*)
(in /usr/lib/libQt5Widgets.so.5.9.1)
==1397==    by 0x788ECF3: QApplication::notify(QObject*, QEvent*) (in
/usr/lib/libQt5Widgets.so.5.9.1)
==1397==  
==1397== Conditional jump or move depends on uninitialised value(s)
==1397==    at 0x1612C3DE: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x16399290: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x16399B9A: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1612DBCA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1610F5AC: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161291DA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x161295E3: JSC::Heap::collect(JSC::HeapOperation) (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639C080: JSC::GCActivityCallback::doWork() (in
/usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x1639D050: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==1397==    by 0x89A91FA: QObject::event(QEvent*) (in
/usr/lib/libQt5Core.so.5.9.1)
==1397==    by 0x788746B: QApplicationPrivate::notify_helper(QObject*, QEvent*)
(in /usr/lib/libQt5Widgets.so.5.9.1)
==1397==    by 0x788ECF3: QApplication::notify(QObject*, QEvent*) (in
/usr/lib/libQt5Widgets.so.5.9.1)
==1397==

[...]

==13178== Use of uninitialised value of size 8
==13178==    at 0x1612C383: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x16399290: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x16399B9A: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x1612DBCA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x1610F5AC: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x161291DA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x161295E3: JSC::Heap::collect(JSC::HeapOperation) (in
/usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x1639C080: JSC::GCActivityCallback::doWork() (in
/usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x1639D050: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x89A91FA: QObject::event(QEvent*) (in
/usr/lib/libQt5Core.so.5.9.1)
==13178==    by 0x788746B: QApplicationPrivate::notify_helper(QObject*,
QEvent*) (in /usr/lib/libQt5Widgets.so.5.9.1)
==13178==    by 0x788ECF3: QApplication::notify(QObject*, QEvent*) (in
/usr/lib/libQt5Widgets.so.5.9.1)
==13178==  Uninitialised value was created by a stack allocation
==13178==    at 0x4051A2: ??? (in /usr/bin/digikam)
==13178==  
==13178== Conditional jump or move depends on uninitialised value(s)
==13178==    at 0x1612C389: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x16399290: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x16399B9A: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x1612DBCA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x1610F5AC: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x161291DA: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x161295E3: JSC::Heap::collect(JSC::HeapOperation) (in
/usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x1639C080: JSC::GCActivityCallback::doWork() (in
/usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x1639D050: ??? (in /usr/lib/libQt5WebKit.so.5.212.0)
==13178==    by 0x89A91FA: QObject::event(QEvent*) (in
/usr/lib/libQt5Core.so.5.9.1)
==13178==    by 0x788746B: QApplicationPrivate::notify_helper(QObject*,
QEvent*) (in /usr/lib/libQt5Widgets.so.5.9.1)
==13178==    by 0x788ECF3: QApplication::notify(QObject*, QEvent*) (in
/usr/lib/libQt5Widgets.so.5.9.1)
==13178==  Uninitialised value was created by a stack allocation
==13178==    at 0x4051A2: ??? (in /usr/bin/digikam)

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #29 from [hidden email] ---
A pair of logs from valgrind
The second one is incomplete.


> valgrind --tool=memcheck --leak-check=full --error-limit=no  digikam

https://pastebin.com/JwvwiNd3

-------------------------------------------------------------------
-------------------------------------------------------------------


> valgrind --tool=memcheck --leak-check=full --error-limit=no --show-leak-kinds=all digikam

https://pastebin.com/q88sw263

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #30 from [hidden email] ---
2,874,814 bytes in 59 blocks are still reachable in loss record 42,450 of
42,454
==14781==    by 0xE57BA04: cmsOpenIOhandlerFromMem (in
/usr/lib/liblcms2.so.2.0.8)
==14781==    by 0xE57D2C4: cmsOpenProfileFromMemTHR (in
/usr/lib/liblcms2.so.2.0.8)
==14781==    by 0x5D19CA8: Digikam::IccProfile::open() (in
/usr/lib/libdigikamcore.so.5.6.0)
==14781==    by 0x5D24328: Digikam::IccSettings::Private::scanDirectory(QString
const&, QStringList const&, QList<Digikam::IccProfile>*) (in
/usr/lib/libdigikamcore.so.5.6.0)
==14781==
==14781== 2,876,289 bytes in 59 blocks are still reachable in loss record
42,451 of 42,454
==14781==    by 0x87AF923: QArrayData::allocate(unsigned long, unsigned long,
unsigned long, QFlags<QArrayData::AllocationOption>) (in
/usr/lib/libQt5Core.so.5.9.1)
==14781==    by 0x87B23C3: QByteArray::resize(int) (in
/usr/lib/libQt5Core.so.5.9.1)
==14781==    by 0x88B410D: QIODevice::readAll() (in
/usr/lib/libQt5Core.so.5.9.1)
==14781==    by 0x5D1970F: Digikam::IccProfile::data() (in
/usr/lib/libdigikamcore.so.5.6.0)
==14781==    by 0x5D19BAF: Digikam::IccProfile::open() (in
/usr/lib/libdigikamcore.so.5.6.0)

This happen when digiKAm scan ICC color profiles from your system. Typically,
you have the color management turned on.

Turn off CM and try again. Perhaps you have corrupted ICC color profile on your
system.

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

[hidden email] changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|Searches                    |Database-Scan

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #31 from [hidden email] ---
How do you turn off the ICC color profile?

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #32 from [hidden email] ---
Go to DK setup/color-management and turn off the first option on top of the
page.

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #33 from [hidden email] ---
(In reply to caulier.gilles from comment #32)
> Go to DK setup/color-management and turn off the first option on top of the
> page.
>
> Gilles Caulier

That's the result.
https://pastebin.com/kH7hx05K

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #34 from [hidden email] ---
(In reply to caulier.gilles from comment #32)
> Go to DK setup/color-management and turn off the first option on top of the
> page.
>
> Gilles Caulier

The problem has not been solved (with qtav 1.12)

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #35 from [hidden email] ---
You have a huge memory link about your NVidia video driver :

282,927 bytes in 1 blocks are possibly lost in loss record 47,954 of 47,969
==24333==    at 0x4C2DF55: calloc (in
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==24333==    by 0x43F4DE92: ??? (in /usr/lib/libGLX_nvidia.so.381.22)
==24333==    by 0x454783A0: ??? (in /usr/lib/libnvidia-glcore.so.381.22)

It's about OpenGL layer used... everywhere, as QtAV, OpenCV, Qt...

This is why it's not only relevant of on part of digiKam. The leak is
everywhere.

My tips to confirm :

- Create a VM with virtual box, and install a secondary system inside.
- install digiKam in this VM.
- run and check memory (here is work perfectly)

The VM will emulate OpenGL with an additional layer which will not use the
Nvidia driver directly.

About the original problem, you nivia package from your Linux has a serious
problem. I use also at work exclusively the NVidia cards for the Linux box
(nothing about DK), and i already seen this kind of problem. Updating the
driver or fixing the OpenGL settings can be a solution. There is a huge Nvidia
control panel to tune all the configurations.

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #36 from [hidden email] ---
Any feedback here ?

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=381877

--- Comment #37 from [hidden email] ---
(In reply to caulier.gilles from comment #36)
> Any feedback here ?
>
> Gilles Caulier

Digikam works again, I've done nothing.
Probably it was a problem with an external lib.

--
You are receiving this mail because:
You are the assignee for the bug.