>> Otherwise, over to you, DrumGizmo... ? Or what else can we do?
>
> Carla also had a problem with Drumgizmo, for a while it couldn't open
> the native GUI. Reported this, falkTX investigated and found that the
> LV2 manifest was not correct. So it seems sometimes a bit fishy.
>
> Also, there is a commit on the master branch in Drumgizmo which is
> named: "Alternative approach to semaphore timed wait." but as far as I
> can see this just uses clock_gettime instead of gettimeofday for the
> time to wait.
>
> Also the terminal reports a interrupted system call. In the code they
> don't check for EINTR and then just abort. Normal procedure would be to
> check for EINTR, recalculate the times and retry, right?
>
> Maybe I will ask them.

Ok, had a chat with them. In the Drumgizmo code in question I just added 
a check of EINTR and treated it like ETIMEDOUT, this makes the crash go 
away. So now they are aware of the problem and will provide a fix for it 
(hopefully soon, they are preparing 0.9.12)


But this brings me back to the other issue. So now I could create the 
video showing the behaviour with the drumlists (with the fixed Drumgizmo):

https://youtu.be/TocOwi2fxjw


I have also some new valgrind reports (formatting may be screwed up a 
bit in the mail):

==19777== Conditional jump or move depends on uninitialised value(s)
==19777==    at 0x54D0D58: MusECore::JackAudioDevice::stopTransport() 
(jack.cpp:2109)
==19777==    by 0x500BD74: MusECore::Audio::start() (audio.cpp:224)
==19777==    by 0x4FD92B7: MusEGui::MusE::seqStart() (app.cpp:158)
==19777==    by 0x40B36D: main (main.cpp:882)


==19777== Invalid read of size 8
==19777==    at 0xECE1E42: ??? (in 
/usr/lib/x86_64-linux-gnu/libgtk-x11-2.0.so.0.2400.23)
==19777==    by 0xECE95DD: ??? (in 
/usr/lib/x86_64-linux-gnu/libgtk-x11-2.0.so.0.2400.23)
==19777==    by 0x147598DA: ??? (in 
/usr/lib/x86_64-linux-gnu/libgio-2.0.so.0.4002.0)
==19777==    by 0x147598F8: ??? (in 
/usr/lib/x86_64-linux-gnu/libgio-2.0.so.0.4002.0)
==19777==    by 0xBFBBCE4: g_main_context_dispatch (in 
/lib/x86_64-linux-gnu/libglib-2.0.so.0.4002.0)
==19777==    by 0xBFBC047: ??? (in 
/lib/x86_64-linux-gnu/libglib-2.0.so.0.4002.0)
==19777==    by 0xBFBC0EB: g_main_context_iteration (in 
/lib/x86_64-linux-gnu/libglib-2.0.so.0.4002.0)
==19777==    by 0x727A98B: 
QEventDispatcherGlib::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) 
(in /usr/lib/x86_64-linux-gnu/libQt5Core.so.5.2.1)
==19777==    by 0x722C96A: 
QEventLoop::exec(QFlags<QEventLoop::ProcessEventsFlag>) (in 
/usr/lib/x86_64-linux-gnu/libQt5Core.so.5.2.1)
==19777==    by 0x6471E0C: QDialog::exec() (in 
/usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5.2.1)
==19777==    by 0x9C3C241: MusEGui::getOpenFileName(QString const&, char 
const**, QWidget*, QString const&, bool*, 
MusEGui::MFileDialog::ViewType) (in 
/opt/muse3/lib/muse-3.0/modules/libmuse_widgets.so)
==19777==    by 0x7FDA06C: 
MusEGui::TList::loadTrackDrummap(MusECore::MidiTrack*, char const*) 
(tlist.cpp:2321)
==19777==  Address 0x31e60cd0 is 608 bytes inside a block of size 864 free'd
==19777==    at 0x4C2BDEC: free (in 
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==19777==    by 0xBD53142: g_type_free_instance (in 
/usr/lib/x86_64-linux-gnu/libgobject-2.0.so.0.4002.0)
==19777==    by 0xECDB929: ??? (in 
/usr/lib/x86_64-linux-gnu/libgtk-x11-2.0.so.0.2400.23)
==19777==    by 0xECE95DD: ??? (in 
/usr/lib/x86_64-linux-gnu/libgtk-x11-2.0.so.0.2400.23)
==19777==    by 0x147598DA: ??? (in 
/usr/lib/x86_64-linux-gnu/libgio-2.0.so.0.4002.0)
==19777==    by 0x147598F8: ??? (in 
/usr/lib/x86_64-linux-gnu/libgio-2.0.so.0.4002.0)
==19777==    by 0xBFBBCE4: g_main_context_dispatch (in 
/lib/x86_64-linux-gnu/libglib-2.0.so.0.4002.0)
==19777==    by 0xBFBC047: ??? (in 
/lib/x86_64-linux-gnu/libglib-2.0.so.0.4002.0)
==19777==    by 0xBFBC0EB: g_main_context_iteration (in 
/lib/x86_64-linux-gnu/libglib-2.0.so.0.4002.0)
==19777==    by 0x727A98B: 
QEventDispatcherGlib::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) 
(in /usr/lib/x86_64-linux-gnu/libQt5Core.so.5.2.1)
==19777==    by 0x722C96A: 
QEventLoop::exec(QFlags<QEventLoop::ProcessEventsFlag>) (in 
/usr/lib/x86_64-linux-gnu/libQt5Core.so.5.2.1)
==19777==    by 0x6471E0C: QDialog::exec() (in 
/usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5.2.1)



==19777== Conditional jump or move depends on uninitialised value(s)
==19777==    at 0x8D40D57: MusEGui::DrumCanvas::rebuildOurDrumMap() 
(dcanvas.cpp:1967)
==19777==    by 0x8D382DF: 
MusEGui::DrumCanvas::DrumCanvas(MusEGui::MidiEditor*, QWidget*, int, 
int, char const*) (dcanvas.cpp:144)
==19777==    by 0x8D583A6: 
MusEGui::DrumEdit::DrumEdit(MusECore::PartList*, QWidget*, char const*, 
unsigned int) (drumedit.cpp:530)
==19777==    by 0x4FE47EB: 
MusEGui::MusE::startDrumEditor(MusECore::PartList*, bool) (app.cpp:1995)
==19777==    by 0x4FE4792: MusEGui::MusE::startDrumEditor() (app.cpp:1990)
==19777==    by 0x4FE3F91: MusEGui::MusE::startEditor(MusECore::Track*) 
(app.cpp:1838)
==19777==    by 0x4FD315B: MusEGui::MusE::qt_static_metacall(QObject*, 
QMetaObject::Call, int, void**) (moc_app.cpp:593)
==19777==    by 0x72552A5: QMetaObject::activate(QObject*, int, int, 
void**) (in /usr/lib/x86_64-linux-gnu/libQt5Core.so.5.2.1)
==19777==    by 0x7FE4420: MusEGui::Arranger::editPart(MusECore::Track*) 
(moc_arranger.cpp:281)
==19777==    by 0x7FE3E30: 
MusEGui::Arranger::qt_static_metacall(QObject*, QMetaObject::Call, int, 
void**) (moc_arranger.cpp:170)
==19777==    by 0x72552A5: QMetaObject::activate(QObject*, int, int, 
void**) (in /usr/lib/x86_64-linux-gnu/libQt5Core.so.5.2.1)
==19777==    by 0x7FE5B3C: 
MusEGui::PartCanvas::dclickPart(MusECore::Track*) (moc_pcanvas.cpp:296)


There are also tons of other reports, but they concern the VSTs and QT 
itself.


lg,
Michael


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Lmuse-developer mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/lmuse-developer

Reply via email to