|
From: skunkwerk <sku...@gm...> - 2008-08-19 03:12:38
|
sorry Paolo, it turns out the problem was not with the allocator but my code... it just took a lot of trial and error to pinpoint it as the segfault wasn't giving me a line number to where the problem was. it's still segfaulting, however, when I try to access a vertex of a face outside the vector's range (no surprise there); I think this is because the new face's vertices are not properly initialized, as it ends up requesting a vertex index in the billions... I've tried two different methods of initializing vertices, but neither fixes the problem: vcg::tri::Allocator<CMeshO>::AddVertices(m,2);//2 new vertices, 1 old vertex the first way: m.face[m.face.size()-1].V(0) = &m.vert[m.vert.size()-2];//new vertex m.face[m.face.size()-1].V(1) = &m.vert[m.vert.size()-1];//new vertex m.face[m.face.size()-1].V(2) = &vertices[2];//old vertex: vertices[2] = m.vert[f.V(2)-&(*m.vert.begin())]; the second way: vi=m.vert.end();--vi;--vi;//vertices added to the back of the vector, decrement because end() is past the last element, and second time because added 2 vertices ivp[0]=&*vi;(*vi).P()=intersections[0];++vi;//new vertex, intersections[0] = Point3f(....) ivp[1]=&*vi;(*vi).P()=intersections[1];//new vertex, intersections[1] = Point3f(....) ivp[2]=&vertices[2];//old vertex: vertices[2] = m.vert[f.V(2)-&(*m.vert.begin())]; fi = m.face.end();--fi;//faces added to back of vector, decrement because end() is past the last element (*fi).V(0)=ivp[0]; (*fi).V(1)=ivp[1]; (*fi).V(2)=ivp[2]; when I tried using CoordType(0.0f,-1.0f,1.0f) as used in the documentation I got the error 'no matching function call to... (Point3<float>) (float, float, float) even though I've declared CoordType and included allocate.h as the cause of the segfault is known in this case, I'm not attaching code for people to replicate the issue - I'm just asking for the correct way of doing this... appreciate the help, imran On Mon, Aug 18, 2008 at 12:52 AM, Paolo Cignoni <pao...@is...>wrote: > Could you provide a 10~30 lines self contained command-line program that > is AS SHORT AS POSSIBLE and that tries to duplicate your issue? > Something in the style of the vcglib/apps/sample/trimesh_XXXX examples > (just a .cpp and a .pro). > > >From just the two lines that you are repeatedly posting is impossible to > understand what is your problem. > If you post just these two lines nobody can help you. The context is > necessary. You should know that... > > The program should be AS SHORT AS POSSIBLE that is should be impossible to > remove any statement (any statement) without breaking its ability of > duplicating the issue. > If someone post a long bug-duplication program that could be made shorter > anyone will think that the submitter is either lazy or that the submitter > considers his time more precious that the others one. > > cheers > p. > > > > > > skunkwerk wrote: > > i'm still struggling with how to add and delete faces... I no longer get > the previous assertion error, but am now getting a segfault whenever I do: > > vcg::tri::Allocator<CMeshO>::Addvertices(m,2); > vcg::tri::Allocator<CMeshO>::AddFaces(m,1); > > when I comment out those two lines, I don't get any crashes... > what am I doing wrong? > > thanks, > imran > > On Sat, Aug 16, 2008 at 11:07 AM, skunkwerk <sku...@gm...> wrote: > > i managed to track down the error once the compiler gave me the correct > line... > > now i'm getting this assertion failure though: > meshlab: ../../../../vcglib/vcg/complex/trimesh/allocate.h:513: static void > vcg::tri::Allocator<AllocateMeshType>::CompactFaceVector(AllocateMeshType&) > [with AllocateMeshType = CMeshO]: Assertion `(int)pos==m.fn' failed. > > this is how i'm adding/deleting faces: > vcg::tri::Allocator<CMeshO>::AddFaces(m,1); > vcg::tri::Allocator<CMeshO>::DeleteFace(m,fp);//delete old face, leave > vertices > vcg::tri::Allocator<CMeshO>::CompactFaceVector(m);//compact faces to remove > those marked as deleted > > am i doing something wrong? > > thanks, > imran > > On Thu, Aug 14, 2008 at 4:58 PM, skunkwerk <sku...@gm...> wrote: > > thanks Paolo, > i had to write 200 lines of code to get the face splitting working - > and it almost compiles, except for these 3 errors: > > colladaio.cpp:264: instantiated from here > /usr/include/c++/4.3/bits/vector.tcc:293: error: array must be initialized > with a brace-enclosed initializer > /usr/include/c++/4.3/bits/vector.tcc:299: error: invalid array assignment > /usr/include/c++/4.3/bits/stl_construct.h:88: error: request for member > '~float [2]' in '* pointer', which is of non-class type 'float [2]' > > but i have no idea where in the code these errors are coming from, as those > are all stl things... any suggestions on how to trace this? i'm using gcc > > OR if someone could tell me how to use the existing code in meshlab, that'd > be great too > > imran > > On Thu, Aug 14, 2008 at 1:04 AM, Paolo Cignoni <pao...@is...>wrote: > > no, at the end. > read this page (and referrred subpages too) > > http://vcg.sourceforge.net/index.php/Tutorial > > p, > > skunkwerk wrote: > > quick question: i was reading 'creating elements' on this wiki page: > http://meshlab.sourceforge.net/wiki/index.php/Accessing_the_mesh > > when you call Allocator<CMeshO>::AddVertices or AddFaces, are > vertices/faces added to the front of the vector? > because the next line that describes specifying the vertex coordinates, > etc. uses a pointer to the front of the vector > > thanks, > imran > > > ------------------------------------------------------------------------- > This SF.Net email is sponsored by the Moblin Your Move Developer's challenge > Build the coolest Linux based applications with Moblin SDK & win great prizes > Grand prize is a trip for two to an Open Source event anywhere in the worldhttp://moblin-contest.org/redirect.php?banner_id=100&url=/ > > > _______________________________________________ > Meshlab-devel mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/meshlab-devel > > > > -- > Paolo Cignoni -- Senior Researcher > Visual Computing Laboratory - ISTI - CNR http://vcg.isti.cnr.it/~cignoni <http://vcg.isti.cnr.it/%7Ecignoni> > > ISTI - CNR > Via Moruzzi 1, > 56124 Pisa > ITALY > > > ------------------------------------------------------------------------- > This SF.Net email is sponsored by the Moblin Your Move Developer's > challenge > Build the coolest Linux based applications with Moblin SDK & win great > prizes > Grand prize is a trip for two to an Open Source event anywhere in the world > http://moblin-contest.org/redirect.php?banner_id=100&url=/ > _______________________________________________ > Meshlab-devel mailing list > Mes...@li... > https://lists.sourceforge.net/lists/listinfo/meshlab-devel > > > > > ------------------------------ > > ------------------------------------------------------------------------- > This SF.Net email is sponsored by the Moblin Your Move Developer's challenge > Build the coolest Linux based applications with Moblin SDK & win great prizes > Grand prize is a trip for two to an Open Source event anywhere in the worldhttp://moblin-contest.org/redirect.php?banner_id=100&url=/ > > ------------------------------ > > _______________________________________________ > Meshlab-devel mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/meshlab-devel > > > > -- > Paolo Cignoni -- Senior Researcher > Visual Computing Laboratory - ISTI - CNR http://vcg.isti.cnr.it/~cignoni <http://vcg.isti.cnr.it/%7Ecignoni> > > ISTI - CNR > Via Moruzzi 1, > 56124 Pisa > ITALY > > > ------------------------------------------------------------------------- > This SF.Net email is sponsored by the Moblin Your Move Developer's > challenge > Build the coolest Linux based applications with Moblin SDK & win great > prizes > Grand prize is a trip for two to an Open Source event anywhere in the world > http://moblin-contest.org/redirect.php?banner_id=100&url=/ > _______________________________________________ > Meshlab-devel mailing list > Mes...@li... > https://lists.sourceforge.net/lists/listinfo/meshlab-devel > > |
