D. Barbier
unread,May 26, 2013, 7:23:40 AM5/26/13Sign in to reply to author
Sign in to forward
You do not have permission to delete messages in this group
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to oce...@googlegroups.com, pumpki...@gmail.com
On 2013/5/26 Luke Parry wrote:
> Dear OCE Team,
>
> Hi, I am one of the developers from FreeCAD. Having recently updated from
> OCE master, after the 6.6 patch, I have come across some segfaults that are
> guaranteed every time. The segfaults appear to occur when the object
> destructor is called automatically at the end of the function scope.
>
> This occurs in our own source code at the end of our function scope:
>
> TopoDS_Face face = validateFace(mkFace.Face());
> BRepAdaptor_Surface adapt(face);
> IntTools_FClass2d class2d(face, prec);
> Handle_Geom_Surface surf = new Geom_Plane(adapt.Plane());
> Handle_ShapeAnalysis_Surface as = new ShapeAnalysis_Surface(surf);
>
> TopExp_Explorer xp(wire2,TopAbs_VERTEX);
> while (xp.More()) {
> TopoDS_Vertex v = TopoDS::Vertex(xp.Current());
> gp_Pnt p = BRep_Tool::Pnt(v);
> gp_Pnt2d uv = as->ValueOfUV(p, prec);
> if (class2d.Perform(uv) == TopAbs_IN)
> return true;
>
> with the following stack trace from GDB.
>
> #0 Handle_Standard_Transient::EndScope()() at /usr/lib/libTKernel.so.5
[...]
> I have spoke to another developer who has compiled from the standard OCC 6.6
> release and he does not experience any segfaults.
>
> What are your thoughts on this? Is there something obvious that we have
> missed out specific to OCE changes or are these regressions specific to OCE?
Hello Luke,
There is some library mismatch, current master builds libTKernel.so.7
and not libTKernel.so.5.
Denis