https://bugs.kde.org/show_bug.cgi?id=374568
--- Comment #13 from RJVB <rjvber...@gmail.com> --- I've been running qtcurve_utils.cpp through a few analysers. Cppcheck doesn't complain about anything, nor does the clang 3.9 static and coverage analyser (but I never used the latter before so may have misused it). Clazy however finds this, with level3 (all relevant and related messages): qtcq5-lnx-work/QtCurve-1.8.18/qt5/style/qtcurve.h:62:1: warning: QtCurve::Style should take QObject parent argument in CTOR [-Wclazy-ctor-missing-parent-argument] class Style: public QCommonStyle { ^ >>> should it? <<< qtcq5-lnx-work/QtCurve-1.8.18/qt5/style/qtcurve_utils.cpp:226:5: warning: Polymorphic class is copyable. Potential slicing. [-Wclazy-copyable-polymorphic] struct HackEvent: public QMouseEvent { ^ qtcq5-lnx-work/QtCurve-1.8.18/qt5/style/qtcurve_utils.cpp:241:5: warning: HackedMenu should take QWidget parent argument in CTOR [-Wclazy-ctor-missing-parent-argument] struct HackedMenu: public QMenuBar { ^ qtcq5-lnx-work/QtCurve-1.8.18/qt5/style/qtcurve_utils.cpp:241:5: warning: HackedMenu is missing a Q_OBJECT macro [-Wclazy-missing-qobject-macro] The missing Q_OBJECT macro might explain why casts to HackedMenu can go subtly wrong. -- You are receiving this mail because: You are watching all bug changes.