SUMMARY *** When i select 82 images and try Image Quality sort, digikam crashes *** STEPS TO REPRODUCE 1. Select folder from left menu 2. Select all photos 3. Right click ImageQualitySorter 4. Setup with deep learning 5. Wait some time - CRASH OBSERVED RESULT Im digikam console Qt Concurrent has caught an exception thrown from a worker thread. This is not supported, exceptions thrown in worker threads must be caught before control returns to Qt Concurrent. terminate called after throwing an instance of 'cv::Exception' what(): OpenCV(4.5.4) ./modules/core/src/matrix.cpp:768: error: (-215:Assertion failed) 0 <= _rowRange.start && _rowRange.start <= _rowRange.end && _rowRange.end <= m.rows in function 'Mat' in GDB #0 0x00007ffff2f2edb0 in cv::Mat::Mat(cv::Mat const&) () at /lib/x86_64-linux-gnu/libopencv_core.so.4.5d #1 0x00007ffff2f2f713 in cv::Mat::reshape(int, int) const () at /lib/x86_64-linux-gnu/libopencv_core.so.4.5d #2 0x00007ffff3976903 in () at /lib/x86_64-linux-gnu/libopencv_dnn.so.4.5d #3 0x00007ffff3977384 in () at /lib/x86_64-linux-gnu/libopencv_dnn.so.4.5d #4 0x00007ffff393883a in () at /lib/x86_64-linux-gnu/libopencv_dnn.so.4.5d #5 0x00007ffff3933742 in () at /lib/x86_64-linux-gnu/libopencv_dnn.so.4.5d #6 0x00007ffff3936fa3 in cv::dnn::dnn4_v20211004::Net::forward(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) () at /lib/x86_64-linux-gnu/libopencv_dnn.so.4.5d #7 0x00007ffff6110c1c in Digikam::AestheticDetector::detect(cv::Mat const&) const (this=0x7ffe76ed8080, image=...) at /home/pschonmann/temp/digikam/core/libs/imgqsort/detectors/aesthetic_detector.cpp:48 #8 0x00007ffff610c514 in Digikam::ImageQualityParser::startAnalyse() (this=0x7ffe7400cb40) at /home/pschonmann/temp/digikam/core/libs/imgqsort/imagequalityparser.cpp:92 #9 0x00007ffff78c9d15 in Digikam::ImageQualityTask::run() (this=0x555561ba5ae0) at /home/pschonmann/temp/digikam/core/utilities/maintenance/imagequalitytask.cpp:118 #10 0x00007ffff46cff92 in () at /lib/x86_64-linux-gnu/libQt5Core.so.5 #11 0x00007ffff46ccca1 in () at /lib/x86_64-linux-gnu/libQt5Core.so.5 #12 0x00007ffff3e94b43 in start_thread (arg=<optimized out>) at ./nptl/pthread_create.c:442 #13 0x00007ffff3f26a00 in clone3 () at ../sysdeps/unix/sysv/linux/x86_64/clone3.S:81 EXPECTED RESULT SOFTWARE/OS VERSIONS Windows: macOS: Linux/KDE Plasma: (available in About System) KDE Plasma Version: KDE Frameworks Version: Qt Version: ADDITIONAL INFORMATION
It crashes in OpenCV, your OpenCV-4.5.4 version is probably too old. Maik
Git commit 65d25dacab3c0f6379b2c8fe23bbca656c98568a by Gilles Caulier. Committed on 02/04/2023 at 04:56. Pushed by cgilles into branch 'master'. add missing C++ exception with opencv code in DNN part of aesthetic detector M +1 -6 core/libs/imgqsort/detectors/abstract_detector.cpp M +1 -1 core/libs/imgqsort/detectors/abstract_detector.h M +23 -8 core/libs/imgqsort/detectors/aesthetic_detector.cpp M +1 -1 core/libs/imgqsort/detectors/aesthetic_detector.h https://invent.kde.org/graphics/digikam/commit/65d25dacab3c0f6379b2c8fe23bbca656c98568a
Petr, AppImage 8.0.0 pre-release bundle is updated with last changes from Maik: https://files.kde.org/digikam/ Please test and give us a feedback. Best regards Gilles Caulier
Thanks for new 8.0.0 builds. But DK is still crashing. --- Thread 99 "Thread (pooled)" received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x7ffe567fc640 (LWP 264083)] 0x00007ffff1ea547e in __GI___libc_free (mem=0x7ffe00000004) at ./malloc/malloc.c:3368 3368 ./malloc/malloc.c: Adresář nebo soubor neexistuje. (gdb) bt #0 0x00007ffff1ea547e in __GI___libc_free (mem=0x7ffe00000004) at ./malloc/malloc.c:3368 #1 0x00007fffedb50499 in cv::dnn::dnn4_v20220524::Layer::~Layer() () from /tmp/.mount_digiKapnXfc9/usr/lib/libopencv_dnn.so.406 #2 0x00007fffedb50519 in cv::dnn::dnn4_v20220524::Layer::~Layer() () from /tmp/.mount_digiKapnXfc9/usr/lib/libopencv_dnn.so.406 #3 0x00007ffff72b2fba in std::_Sp_counted_base<(__gnu_cxx::_Lock_policy)2>::_M_release (this=0x7ffe60003290) at ././/include/c++/7/bits/shared_ptr_base.h:154 #4 0x00007fffedc3801f in cv::dnn::dnn4_v20220524::Net::Impl::getLayerShapesRecursively(int, std::map<int, cv::dnn::dnn4_v20220524::detail::LayerShapes, std::less<int>, std::allocator<std::pair<int const, cv::dnn::dnn4_v20220524::detail::LayerShapes> > >&) () from /tmp/.mount_digiKapnXfc9/usr/lib/libopencv_dnn.so.406 #5 0x00007fffedc3a076 in cv::dnn::dnn4_v20220524::Net::Impl::getLayersShapes(std::vector<std::vector<int, std::allocator<int> >, std::allocator<std::vector<int, std::allocator<int> > > > const&, std::map<int, cv::dnn::dnn4_v20220524::detail::LayerShapes, std::less<int>, std::allocator<std::pair<int const, cv::dnn::dnn4_v20220524::detail::LayerShapes> > >&) () from /tmp/.mount_digiKapnXfc9/usr/lib/libopencv_dnn.so.406 #6 0x00007fffedc3eb99 in cv::dnn::dnn4_v20220524::Net::Impl::allocateLayers(std::vector<cv::dnn::dnn4_v20220524::detail::LayerPin, std::allocator<cv::dnn::dnn4_v20220524::detail::LayerPin> > const&) () from /tmp/.mount_digiKapnXfc9/usr/lib/libopencv_dnn.so.406 #7 0x00007fffedc42ad8 in cv::dnn::dnn4_v20220524::Net::Impl::setUpNet(std::vector<cv::dnn::dnn4_v20220524::detail::LayerPin, std::allocator<cv::dnn::dnn4_v20220524::detail::LayerPin> > const&) () from /tmp/.mount_digiKapnXfc9/usr/lib/libopencv_dnn.so.406 #8 0x00007fffedc4316d in cv::dnn::dnn4_v20220524::Net::Impl::forward(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) () from /tmp/.mount_digiKapnXfc9/usr/lib/libopencv_dnn.so.406 #9 0x00007fffedc280f9 in cv::dnn::dnn4_v20220524::Net::forward(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) () from /tmp/.mount_digiKapnXfc9/usr/lib/libopencv_dnn.so.406 #10 0x00007ffff58742a5 in Digikam::AestheticDetector::detect (this=this@entry=0x7ffe5a7d16c0, image=...) at /b/master/dktemp/digikam-master/core/libs/imgqsort/detectors/aesthetic_detector.cpp:50 #11 0x00007ffff5871de1 in Digikam::ImageQualityParser::startAnalyse (this=this@entry=0x7ffe58001550) at /b/master/dktemp/digikam-master/core/libs/imgqsort/imagequalityparser.cpp:92 #12 0x00007ffff747310f in Digikam::ImageQualityTask::run (this=0x55555910ea30) at /b/master/dktemp/digikam-master/core/utilities/maintenance/imagequalitytask.cpp:118 #13 0x00007ffff26b9bca in ?? () from /tmp/.mount_digiKapnXfc9/usr/lib/libQt5Core.so.5 #14 0x00007ffff26b61a2 in ?? () from /tmp/.mount_digiKapnXfc9/usr/lib/libQt5Core.so.5 #15 0x00007ffff1e94b43 in start_thread (arg=<optimized out>) at ./nptl/pthread_create.c:442 #16 0x00007ffff1f26a00 in clone3 () at ../sysdeps/unix/sysv/linux/x86_64/clone3.S:81
Maik, The crash is reproducible by Petr with the AppImage. It's not an OpenCV version problem. The message "Qt Concurrent has caught an exception thrown from a worker thread. This is not supported, exceptions thrown in worker threads must be caught before control returns to Qt Concurrent." means that the C++ exception must be handle in the top level where Qt::Concurrent is created ? Gilles
Petr, Did you downloaded the DNN model file for the aesthetic detector? (weights_inceptionv3_299.pb). What's the contents from ~/.local/share/digikam/facesengine ? Can you copy and paste the debug statement for the console when digikam start (not the gdb trace) ? Did you use multi-core option from maintenance tool ? Gilles Caulier
I can re produce the crash: it's due to use multicore with static member to host dnn model. I will fix it Gilles
Git commit e016a2797e78ebfc6bfdc320435caa50af925127 by Gilles Caulier. Committed on 03/04/2023 at 07:35. Pushed by cgilles into branch 'master'. Use muxtex to protext model in aesthetic parser when multicore is used. FIXED-IN: 8.0.0 M +52 -23 core/libs/imgqsort/detectors/aesthetic_detector.cpp M +10 -4 core/libs/imgqsort/detectors/aesthetic_detector.h M +1 -1 core/libs/imgqsort/imagequalityparser.cpp https://invent.kde.org/graphics/digikam/commit/e016a2797e78ebfc6bfdc320435caa50af925127