#3525 closed defect (fixed)
OMEdit segfaults in libQtGui.so.4
Reported by: | perost | Owned by: | adeas31 |
---|---|---|---|
Priority: | high | Milestone: | 1.11.0 |
Component: | OMEdit | Version: | |
Keywords: | Cc: | sjoelund.se, adrpo |
Description
OMEdit segfaults immediately for me, with gdb saying this:
Program received signal SIGSEGV, Segmentation fault. 0x00007fffec791c34 in ?? () from /usr/lib/libQtGui.so.4 (gdb) bt #0 0x00007fffec791c34 in ?? () from /usr/lib/libQtGui.so.4 #1 0x00007ffff7dea27a in call_init.part () from /lib64/ld-linux-x86-64.so.2 #2 0x00007ffff7dea38b in _dl_init () from /lib64/ld-linux-x86-64.so.2 #3 0x00007ffff7ddbdba in _dl_start_user () from /lib64/ld-linux-x86-64.so.2 #4 0x0000000000000001 in ?? () #5 0x00007fffffffe0b3 in ?? () #6 0x0000000000000000 in ?? ()
This is on Arch Linux 64-bit, with both qt4 and qt5 installed. ldd OMEdit gives (with non-OM and non-Qt stuff removed):
libQt5WebKitWidgets.so.5 => /usr/lib/libQt5WebKitWidgets.so.5 (0x00007f5e607c9000) libQt5WebKit.so.5 => /usr/lib/libQt5WebKit.so.5 (0x00007f5e5e129000) libQt5Svg.so.5 => /usr/lib/libQt5Svg.so.5 (0x00007f5e5ded2000) libQt5PrintSupport.so.5 => /usr/lib/libQt5PrintSupport.so.5 (0x00007f5e5dc5f000) libQt5Widgets.so.5 => /usr/lib/libQt5Widgets.so.5 (0x00007f5e5d3ab000) libQt5Gui.so.5 => /usr/lib/libQt5Gui.so.5 (0x00007f5e5cc3f000) libQt5XmlPatterns.so.5 => /usr/lib/libQt5XmlPatterns.so.5 (0x00007f5e5c63c000) libQt5Network.so.5 => /usr/lib/libQt5Network.so.5 (0x00007f5e5c2e4000) libQt5Xml.so.5 => /usr/lib/libQt5Xml.so.5 (0x00007f5e5c0a7000) libQt5Core.so.5 => /usr/lib/libQt5Core.so.5 (0x00007f5e5b9ba000) libQtGui.so.4 => /usr/lib/libQtGui.so.4 (0x00007f5e583c8000) libQtCore.so.4 => /usr/lib/libQtCore.so.4 (0x00007f5e57ec0000) libQt5Sensors.so.5 => /usr/lib/libQt5Sensors.so.5 (0x00007f5e57c83000) libQt5OpenGL.so.5 => /usr/lib/libQt5OpenGL.so.5 (0x00007f5e57a28000) libQt5Positioning.so.5 => /usr/lib/libQt5Positioning.so.5 (0x00007f5e54182000) libQt5Quick.so.5 => /usr/lib/libQt5Quick.so.5 (0x00007f5e53ba2000) libQt5Qml.so.5 => /usr/lib/libQt5Qml.so.5 (0x00007f5e53565000) libQt5WebChannel.so.5 => /usr/lib/libQt5WebChannel.so.5 (0x00007f5e53347000) libQt5Sql.so.5 => /usr/lib/libQt5Sql.so.5 (0x00007f5e53104000)
So OMEdit is linked against both libQtGui.so.4 and libQt5Gui.so.5, i.e. a mix of Qt 4 and Qt 5 libraries. That seems bad, or is it supposed to be like that?
Change History (7)
comment:1 Changed 8 years ago by sjoelund.se
comment:2 follow-up: ↓ 3 Changed 8 years ago by hkiel
Is this problem solved?
The corresponding ticket should be #3504 and the fix in https://trac.openmodelica.org/OpenModelica/changeset/7fd0367c6223e12217495b28ba59f193b4f3763d/OMEdit
comment:3 in reply to: ↑ 2 Changed 8 years ago by perost
Replying to hkiel:
Is this problem solved?
The corresponding ticket should be #3504 and the fix in https://trac.openmodelica.org/OpenModelica/changeset/7fd0367c6223e12217495b28ba59f193b4f3763d/OMEdit
No, this has not been solved. OMEdit uses qmake instead of CMake, so the fix isn't immediately applicable. I tried some ways to force the QT version with qmake before reporting the issue, but couldn't come up with a solution that worked (don't remember what I tried though, too long ago).
comment:4 Changed 8 years ago by perost
Removing set(QT4_BUILD ON) from qjson/CMakeLists.txt fixes the issue for me, but Martin isn't happy with that solution (need to be able to select which Qt version to build with).
comment:5 follow-up: ↓ 6 Changed 7 years ago by adeas31
Isn't it fixed now?
comment:6 in reply to: ↑ 5 Changed 7 years ago by perost
- Resolution set to fixed
- Status changed from new to closed
comment:7 Changed 7 years ago by adeas31
- Milestone changed from Future to 1.11.0
The problem I suppose is that qmake is found, but cmake prefers qt4 (there was another ticket for this...)