Created attachment 307263 [details] build.log painting/qpaintengineex.cpp:776:5: error: narrowing conversion of ‘(((double)x2) - (((double)xRadius) * 4.47715250200000025593283226044150069355964660645e-1))’ from ‘double’ to ‘qreal {aka float}’ inside { } [-fpermissive]
Created attachment 307267 [details] environment
Created attachment 307269 [details] info
Created attachment 307271 [details] pqv
qreal is a typedef for float on arm platforms. I think this is an upstream bug, could you report it there please?
It should build with USE=-c++0x anyway...
Does 4.8.2 still fail?
With masked c++0x & gcc-4.6.3 - success.
(In reply to comment #7) > With masked c++0x & gcc-4.6.3 - success. Do you mean with USE=c++0x enabled?
=x11-libs/qt-gui-4.8.2 USE="accessibility dbus egl exceptions gif glib gtkstyle mng tiff xinerama xv (-aqua) (-c++0x) -cups -debug -nas -nis -pch (-qpa) -qt3support -trace" built with gcc-4.6.3
You haven't enabled the c++0x use flag. Your comment is irrelevant to this bug
(In reply to comment #9) > =x11-libs/qt-gui-4.8.2 USE="accessibility dbus egl exceptions gif glib > gtkstyle mng tiff xinerama xv (-aqua) (-c++0x) -cups -debug -nas -nis -pch > (-qpa) -qt3support -trace" > built with gcc-4.6.3 (-c++0x) means that you have c++0x USE flag disabled and masked, so this is totally irrelevant for this bug report, which is about building qt-gui with c++0x *enabled*. You have to unmask the flag locally and rebuild qt-gui.
Can you try with qt 4.8.4 (which has USE=c++0x unmasked) please?
(In reply to comment #12) > Can you try with qt 4.8.4 (which has USE=c++0x unmasked) please? ping?
Not reproduceable with recent in-tree Qt and GCC versions. Feel free to reopen if issue still happens.