October 2010 - cgma-dev - mailman3.cels.anl.gov

r4239 - cgm/trunk/geom/OCC
by jiangtao_ma＠yahoo.com 26 Oct '10

26 Oct '10

Author: janehu Date: 2010-10-26 09:27:13 -0500 (Tue, 26 Oct 2010) New Revision: 4239 Modified: cgm/trunk/geom/OCC/OCCModifyEngine.cpp Log: Fixed nightly run failure. Modified: cgm/trunk/geom/OCC/OCCModifyEngine.cpp =================================================================== --- cgm/trunk/geom/OCC/OCCModifyEngine.cpp 2010-10-25 14:21:09 UTC (rev 4238) +++ cgm/trunk/geom/OCC/OCCModifyEngine.cpp 2010-10-26 14:27:13 UTC (rev 4239) @@ -2451,7 +2451,7 @@ DLIList<CubitBox*> bs; edge_lists.reset(); CubitBox box; - for(int i = 0; i < size, size > 1; i++) + for(int i = 0; i < size && size > 1; i++) { edge_list = edge_lists.get_and_step(); for(int j = 0; j < edge_list->size(); j++) @@ -2493,7 +2493,7 @@ } } } - for(int i = 0; i < size, size > 1; i++) + for(int i = 0; i < size && size > 1; i++) delete bs.pop(); edge_lists.remove_all_with_value((DLIList<TopoDS_Edge*>*)NULL);

1 0

buildbot failure in Fathom on cgm-occ
by cgma-dev＠lists.mcs.anl.gov 26 Oct '10

26 Oct '10

The Fathomatic has detected a failed build of cgm-occ on Fathom. Full details are available at: http://gnep.mcs.anl.gov:8010/builders/cgm-occ/builds/280 Buildbot URL: http://gnep.mcs.anl.gov:8010/ Buildslave for this Build: gnep Build Reason: The Nightly scheduler named 'nightly' triggered this build Build Source Stamp: None Blamelist: BUILD FAILED: failed test LOG EXCERPT: ======================================= Constructed 2 Volumes: 30, 31 ======================================= Updated volume(s): 30 Destroyed volume(s): 31 Exported: 1 OCC Body to unite5.occ ======================================= Constructed 2 Volumes: 32, 33 ======================================= Updated volume(s): 33 Destroyed volume(s): 32 Exported: 1 OCC Body to unite6.occ PASS: r_w Geometry engine set to: OpenCascade 6.3 Constructed 1 Volume: 1 done... PASS: test_occ Geometry engine set to: OpenCascade 6.3 ======================================= Constructed 2 Volumes: 1, 2 ======================================= Total 0 cuts performed, with from_shape having 8 faces. Total 0 cuts performed, with from_shape having 8 faces. Created volume(s): 5 Updated volume(s): 1 PASS: operation Geometry engine set to: OpenCascade 6.3 PASS: brick_occ Geometry engine set to: OpenCascade 6.3 Surface 2 and 7 consolidated Consolidated 1 pair of surfaces Consolidated 0 pair of curves Consolidated 0 pairs of vertices Exported: 2 OCC Bodies to merge_test.OCC Constructed 2 Volumes: 3, 4 PASS: merge_occ ==================== 1 of 12 tests failed ==================== sincerely, -The Fathomatic

1 0

r4238 - cgm/trunk/geom/OCC
by jiangtao_ma＠yahoo.com 25 Oct '10

25 Oct '10

Author: janehu Date: 2010-10-25 09:21:09 -0500 (Mon, 25 Oct 2010) New Revision: 4238 Modified: cgm/trunk/geom/OCC/OCCModifyEngine.cpp Log: Fixed for fill2 test in mcnp2cad/tests/INP-fill2. Added a check to find arc-curve overlap and made sure if imprinting with a face which has inner loops, first use outer loop to split. Modified: cgm/trunk/geom/OCC/OCCModifyEngine.cpp =================================================================== --- cgm/trunk/geom/OCC/OCCModifyEngine.cpp 2010-10-22 20:28:40 UTC (rev 4237) +++ cgm/trunk/geom/OCC/OCCModifyEngine.cpp 2010-10-25 14:21:09 UTC (rev 4238) @@ -2307,48 +2307,9 @@ sort_curves(common_curves, temp_edge_lists); DLIList<TopoDS_Edge*>* edge_list; int size = temp_edge_lists.size(); - //if size > 1 , check to make sure they are not outer wire - // and inner wires. - DLIList<CubitBox*> bs; - temp_edge_lists.reset(); - CubitBox box, box1; - for(int i = 0; i < size; i++) - { - edge_list = temp_edge_lists.get_and_step(); - box.reset(box1); - for(int j = 0; j < edge_list->size(); j++) - { - TopoDS_Edge e = *(edge_list->get_and_step()); - int k = OCCQueryEngine::instance()->OCCMap->Find(e); - Curve* curve = NULL; - curve = (Curve*)(OCCQueryEngine::instance()->OccToCGM->find(k))->second; - box |= curve->bounding_box(); - } - bs.append(&box); - } - for(int i = 0; i < bs.size()-1; i++) - { - for(int j = i + 1; j < bs.size(); j ++) - { - if(bs[j] < bs[i]) - { - bs[size] = bs[i]; - bs[i] = bs[j]; - bs[j] = bs[size]; - } - if(bs[i] < bs[j]) - { - temp_edge_lists.step(i); - temp_edge_lists.change_to((DLIList<TopoDS_Edge*>*)NULL); - continue; - } - } - } for(int i = 0; i < size; i++) { - edge_list = temp_edge_lists.pop(); - if(edge_list == NULL) - continue; + edge_list = temp_edge_lists.get_and_step();

1 0

Re: [cgma-dev] Seg. fault with gcc 4.4.5 and Ubuntu 10.10
by Jason Kraftcheck 22 Oct '10

22 Oct '10

On 10/22/2010 09:34 AM, Rajeev Jain wrote: > I'm getting a segmentation fault with CGM built with Cubit10.2 on Ubuntu > 10.10 using gcc version 4.4.5 There appears to be some heap corruption happening that causes a free() call to abort(). But that could be happening anywhere. For example, the ancient ACIS libraries used by Cubit 10.x have many mis-matched new[] and delete (no brackets) calls that weren't a problem in the past but may be for a newer version of libstdc++. - jason

1 0

r4235 - cgm/trunk/itaps
by kraftche＠cae.wisc.edu 22 Oct '10

22 Oct '10

Author: kraftche Date: 2010-10-22 11:45:02 -0500 (Fri, 22 Oct 2010) New Revision: 4235 Modified: cgm/trunk/itaps/iGeom_CGMA.cc Log: replace a little bit if C string handling with a nicer C++ version Modified: cgm/trunk/itaps/iGeom_CGMA.cc =================================================================== --- cgm/trunk/itaps/iGeom_CGMA.cc 2010-10-22 16:14:33 UTC (rev 4234) +++ cgm/trunk/itaps/iGeom_CGMA.cc 2010-10-22 16:45:02 UTC (rev 4235) @@ -313,22 +313,17 @@ // scan options for default engine option std::string engine; if (options && options_size) { - char* opt = (char*)malloc( options_size + 1 ); - memcpy(opt, options, options_size); - opt[options_size] = '\0'; - const char delim[] = { opt[0], '\0' }; - for (char* s = strtok(opt+1, delim); s; s = strtok(0,delim)) { - char* v = strchr( s, '=' ); - if (v) { - *v = '\0'; - ++v; - } - if (!strcmp(s,"engine")) { - engine = std::string(v); - break; - } + std::string tmp(options, options_size); + char f[] = ";engine="; f[0]=tmp[0]; // correct delimiter + size_t p = tmp.find( f ); + if (p != std::string::npos) { // if we found engine option + p += strlen(f); // advance to value (past '=') + size_t e = tmp.find( tmp[0], p ); // find end delimiter + if (e == std::string::npos) // if no end delim, then must be last option + engine = tmp.substr( p, std::string::npos ); + else + engine = tmp.substr( p, e-p ); } - free(opt); } // initialize static var with result so that call happens only once

1 0

r4234 - cgm/trunk
by kraftche＠cae.wisc.edu 22 Oct '10

22 Oct '10

Author: kraftche Date: 2010-10-22 11:14:33 -0500 (Fri, 22 Oct 2010) New Revision: 4234 Modified: cgm/trunk/configure.ac Log: fix configure bug where we need libtool during configure but it isn't created until end of script for libtool 2.x.x Modified: cgm/trunk/configure.ac =================================================================== --- cgm/trunk/configure.ac 2010-10-22 16:13:20 UTC (rev 4233) +++ cgm/trunk/configure.ac 2010-10-22 16:14:33 UTC (rev 4234) @@ -25,6 +25,14 @@ AC_FC_WRAPPERS fi +################################################################################ +# Create libtool script +################################################################################ +# We need the libtool script both for the calls to the ITAPS_LIBTOOL_VAR +# macro and for us in testing for ACIS libraries. If we're using a newer +# version of libtool, the script normally isn't generated until AC_OUTPUT +# at the end of the configure script. Ask that it be created now instead. +m4_ifdef([LT_OUTPUT],[LT_OUTPUT]) ################################################################################

1 0

Seg. fault with gcc 4.4.5 and Ubuntu 10.10
by Rajeev Jain 22 Oct '10

22 Oct '10

I'm getting a segmentation fault with CGM built with Cubit10.2 on Ubuntu 10.10 using gcc version 4.4.5 Error message and gdb debug message is :- rajeev@nix:~/FATHOM/source/CGM/itaps$ gdb testgeom GNU gdb (GDB) 7.2-ubuntu Copyright (C) 2010 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "i686-linux-gnu". For bug reporting instructions, please see: <http://www.gnu.org/software/gdb/bugs/>... Reading symbols from /home/rajeev/FATHOM/source/CGM/itaps/testgeom...done. (gdb) r Starting program: /home/rajeev/FATHOM/source/CGM/itaps/testgeom [Thread debugging using libthread_db enabled] Using default input file: ./testgeom.sat ITAPS GEOMETRY INTERFACE TEST PROGRAM: gLoad: Read 1 ACIS Entities from the input file Constructed 1 Volume: 1 Model contents: vertices: 36 edges: 60 faces: 33 regions: 1 Success tags: Tags defined on model: GLOBAL_ID, NAME Success Program received signal SIGSEGV, Segmentation fault. 0x00d2de39 in __libc_free (mem=0x80f4e20) at malloc.c:3724 3724 malloc.c: No such file or directory. in malloc.c (gdb) bt #0 0x00d2de39 in __libc_free (mem=0x80f4e20) at malloc.c:3724 #1 0x001ba441 in operator delete(void*) () from /usr/lib/libstdc++.so.6 #2 0x0019cacd in std::basic_string<char, std::char_traits<char>, std::allocator<char> >::_Rep::_M_destroy(std::allocator<char> const&) () from /usr/lib/libstdc++.so.6 #3 0x0019cd61 in std::basic_string<char, std::char_traits<char>, std::allocator<char> >::_M_mutate(unsigned int, unsigned int, unsigned int) () from /usr/lib/libstdc++.so.6 #4 0x0019cdba in std::basic_string<char, std::char_traits<char>, std::allocator<char> >::_M_replace_safe(unsigned int, unsigned int, char const*, unsigned int) () from /usr/lib/libstdc++.so.6 #5 0x0019ce7e in std::basic_string<char, std::char_traits<char>, std::allocator<char> >::assign(char const*, unsigned int) () from /usr/lib/libstdc++.so.6 #6 0x0019d085 in std::basic_string<char, std::char_traits<char>, std::allocator<char> >::operator=(char const*) () from /usr/lib/libstdc++.so.6 #7 0x08072065 in iGeom_setLastError (error_type=15, description=0x0) at iGeomError.cc:54 #8 0x0807314d in CGMTagManager::getTagHandle (this=0x808c860, tag_name=0x807d204 "__cgm_parent_child_tag") at CATag.cpp:337 #9 0x080727f7 in CGMTagManager::pc_tag (this=0x808c860, create_if_missing=true) at CATag.cpp:163 #10 0x0807441a in CGMTagManager::pc_list (this=0x808c860, gentity=0x81bbc98, list_no=0, create_if_missing=true) at CATag.cpp:699 #11 0x0805b24f in igeom_addprntchld_ (instance=0x808c860, parent_entity_set=0x81b9218, child_entity_set=0x81bbc98, err=0xbfffeea0) at iGeom_CGMA.cc:827 #12 0x08051014 in gentityset_test (geom=0x808c860) at testgeom.cc:676 #13 0x0804e96b in main (argc=1, argv=0xbfffefd4) at testgeom.cc:194 (gdb) My config line is: ./configure --prefix=/home/rajeev/FATHOM/lib/CGM --with-cubit=/home/rajeev/cubit-32 --enable-debug LDFLAGS=-L/home/rajeev/cubit-32/bin LIBS=-lcubiti19 --disable-optimize Rajeev

1 0

r4225 - cgm/trunk/geom/OCC
by jiangtao_ma＠yahoo.com 19 Oct '10

19 Oct '10

Author: janehu Date: 2010-10-19 12:20:14 -0500 (Tue, 19 Oct 2010) New Revision: 4225 Modified: cgm/trunk/geom/OCC/OCCQueryEngine.cpp cgm/trunk/geom/OCC/OCCQueryEngine.hpp Log: Add for buffer read in or wirte out for occ shapes. Modified: cgm/trunk/geom/OCC/OCCQueryEngine.cpp =================================================================== --- cgm/trunk/geom/OCC/OCCQueryEngine.cpp 2010-10-18 22:10:22 UTC (rev 4224) +++ cgm/trunk/geom/OCC/OCCQueryEngine.cpp 2010-10-19 17:20:14 UTC (rev 4225) @@ -1203,189 +1203,8 @@ return CUBIT_SUCCESS; } -/* -CubitStatus -OCCQueryEngine::write_topology( //const unsigned char* p_buffer, - //char* p_buffer, - ofstream& os, - DLIList<OCCBody*> &OCC_bodies, - DLIList<OCCSurface*> &OCC_surfaces, - DLIList<OCCCurve*> &OCC_curves, - DLIList<OCCPoint*> &OCC_points) -{ - int i; - //Create a compound shape to export - BRep_Builder B; - TopoDS_Compound Co; - B.MakeCompound(Co); - - //Add every shape to the compound - for (i = 0; i < OCC_bodies.size(); i++) - { - OCCBody* body = OCC_bodies.get_and_step(); - TopoDS_CompSolid *shape = body->get_TopoDS_Shape(); - - if (shape == NULL || shape->IsNull()) //singel lump or sheet or shell body - { - OCCSurface* surface = body->my_sheet_surface(); - OCCShell* shell = body->shell(); - DLIList<Lump*> lumps = occ_body->lumps(); - if(surface) - B.Add(Co,*(surface->get_TopoDS_Face())); - else if (shell) - B.Add(Co,*(shell->get_TopoDS_Shell())); - else - B.Add(Co,*(CAST_TO(lumps.get(),OCCLump)->get_TopoDS_Solid())); - continue; - } - - //check if this body is build backwards from lump. if so, - //the body and its CompSolid doesn't have bounded relationship - //established. In this case, each individual lump of the body - // will be exported as TopoDS_Solid without a CompSolid - if(OCCMap->IsBound(*shape)) - B.Add(Co, *shape);

1 0

r4222 - cgm/trunk/geom/OCC
by jiangtao_ma＠yahoo.com 18 Oct '10

18 Oct '10

Author: janehu Date: 2010-10-18 11:20:49 -0500 (Mon, 18 Oct 2010) New Revision: 4222 Modified: cgm/trunk/geom/OCC/OCCBody.cpp cgm/trunk/geom/OCC/OCCModifyEngine.cpp Log: Changes to correspond with body holding multiple volumes or shells or surfaces when treating Attributes. Modified: cgm/trunk/geom/OCC/OCCBody.cpp =================================================================== --- cgm/trunk/geom/OCC/OCCBody.cpp 2010-10-16 01:35:04 UTC (rev 4221) +++ cgm/trunk/geom/OCC/OCCBody.cpp 2010-10-18 16:20:49 UTC (rev 4222) @@ -189,6 +189,11 @@ void OCCBody::append_simple_attribute_virt(CubitSimpleAttrib *csa) { + if (myTopoDSShape != NULL) + { + OCCAttribSet::append_attribute(csa, *myTopoDSShape); + return; + } int i = 0; for (i; i < mySheetSurfaces.size();i++) { @@ -206,12 +211,15 @@ TopoDS_Solid* solid = lump->get_TopoDS_Solid(); OCCAttribSet::append_attribute(csa, *solid); } - if (myTopoDSShape != NULL) - OCCAttribSet::append_attribute(csa, *myTopoDSShape); } void OCCBody::remove_simple_attribute_virt(CubitSimpleAttrib *csa) { + if (myTopoDSShape != NULL) + { + OCCAttribSet::remove_attribute(csa, *myTopoDSShape); + return; + } for(int i = 0; i < myShells.size(); i++) { TopoDS_Shell* shell = myShells.get_and_step()->get_TopoDS_Shell(); @@ -228,8 +236,6 @@ TopoDS_Solid* solid = lump->get_TopoDS_Solid(); OCCAttribSet::remove_attribute(csa, *solid); } - if (myTopoDSShape != NULL) - OCCAttribSet::remove_attribute(csa, *myTopoDSShape); } @@ -238,6 +244,9 @@ CubitStatus OCCBody::get_simple_attribute(DLIList<CubitSimpleAttrib*>& csa_list) { + if (myTopoDSShape != NULL) + return OCCAttribSet::get_attributes(*myTopoDSShape,csa_list); + for (int i=0; i < mySheetSurfaces.size();i++)

1 0

Re: [cgma-dev] r4190 - in cgm/trunk: geom/OCC test
by Jiangtao Hu 15 Oct '10

15 Oct '10

Hi, Hongjun Now the code is changed for bodies to hold multiple volumes, it should be able to contain a body attribute. I haven't gone through this part of code yet. But I can work on it if you need. Thanks for bringing up this topic. Jane --- On Fri, 10/15/10, hongjun(a)mcs.anl.gov <hongjun(a)mcs.anl.gov> wrote: From: hongjun(a)mcs.anl.gov <hongjun(a)mcs.anl.gov> Subject: Re: [cgma-dev] r4190 - in cgm/trunk: geom/OCC test To: jiangtao_ma(a)yahoo.com Cc: "Developer information for cgma" <cgma-dev(a)lists.mcs.anl.gov> Date: Friday, October 15, 2010, 11:59 AM Hello, Jane I am developing parallel meshing and I need to distribute OCC geometry to multiple processors. Since I am using Body as a unit of geometry distribution, I want to store distribution information as attributes to Body entities. So, I made several import/export functions from/to memory buffer in "OCCQueryTool". However, they are changed for "export_solid_model" function not to export parent Body entity of single Volume lump geometry and I lose the attribute information assigned to Body entity. So, can you change for single volume geometry to be exported with its parent Body attribute information? Otherwise, do I have to use Volume and Surface as a distribution unit and assign the distribution information to Volume entity? ----------------------------- Hong-Jun Kim Post-doc researcher 9700 S. Cass Ave. B240/R2147 Argonne, IL 60439 630-252-4791 hongjun(a)mcs.anl.gov ----------------------------- ----- "jiangtao ma" <jiangtao_ma(a)yahoo.com> wrote: > Author: janehu > Date: 2010-10-04 09:47:41 -0500 (Mon, 04 Oct 2010) > New Revision: 4190 > > Modified: > cgm/trunk/geom/OCC/OCCBody.cpp > cgm/trunk/geom/OCC/OCCBody.hpp > cgm/trunk/geom/OCC/OCCCurve.cpp > cgm/trunk/geom/OCC/OCCCurve.hpp > cgm/trunk/geom/OCC/OCCDrawTool.cpp > cgm/trunk/geom/OCC/OCCLump.cpp > cgm/trunk/geom/OCC/OCCLump.hpp > cgm/trunk/geom/OCC/OCCModifyEngine.cpp > cgm/trunk/geom/OCC/OCCPoint.hpp > cgm/trunk/geom/OCC/OCCQueryEngine.cpp > cgm/trunk/geom/OCC/OCCQueryEngine.hpp > cgm/trunk/geom/OCC/OCCShell.cpp > cgm/trunk/geom/OCC/OCCShell.hpp > cgm/trunk/geom/OCC/OCCSurface.cpp > cgm/trunk/geom/OCC/OCCSurface.hpp > cgm/trunk/test/Makefile.am > cgm/trunk/test/makept.cpp > cgm/trunk/test/modify.cpp > cgm/trunk/test/r_w.cpp > Log: > Body now has compound structure: can hold multiple surfaces, shells > and lumps. Single surface or shell or lump body will not have > TopoDS_Coumpound pointer in Body, and boolean operation is working > only on these kinds of bodies. The result of the boolean can be > compound. > So far, it passes all 12 tests under CGM/test, and it passed two more > tests in mcnp2cad: cone and fill1. Other tests are under investigation > now. > > Modified: cgm/trunk/geom/OCC/OCCBody.cpp > =================================================================== > --- cgm/trunk/geom/OCC/OCCBody.cpp 2010-10-04 13:09:41 UTC (rev 4189) > +++ cgm/trunk/geom/OCC/OCCBody.cpp 2010-10-04 14:47:41 UTC (rev 4190) > @@ -43,6 +43,7 @@ > #include "TopTools_ListIteratorOfListOfShape.hxx" > #include "gp_Ax1.hxx" > #include "gp_Ax2.hxx" > +#include "gp_GTrsf.hxx" > #include "Bnd_Box.hxx" > #include "BRepBndLib.hxx" > #include "TopExp_Explorer.hxx" > @@ -52,75 +53,133 @@ > #include "BRepGProp.hxx" > #include "Standard_Boolean.hxx" > #include "LocOpe_SplitShape.hxx" > +#include "TopoDS_Compound.hxx" > > //------------------------------------------------------------------------- > // Purpose : A constructor with a list of lumps that are > attached. > // > // Special Notes : > // > > //------------------------------------------------------------------------- > -OCCBody::OCCBody(TopoDS_CompSolid *theShape, CubitBoolean > isSheetBody, > - OCCSurface* surface, OCCShell* shell) > +OCCBody::OCCBody(TopoDS_Compound *theShape, > + OCCSurface* surface, OCCShell* shell, Lump* lump) > { > myTopoDSShape = theShape; > - IsSheetBody = isSheetBody; > - mySheetSurface = surface; > - myShell = shell; > + if (surface != NULL) > + mySheetSurfaces.append(surface); > + if( shell != NULL) > + myShells.append(shell); > + if (lump != NULL) > + myLumps.append(lump); > update_bounding_box(); > + if (myTopoDSShape && !myTopoDSShape->IsNull()) > + assert(myTopoDSShape->ShapeType() == TopAbs_COMPOUND); > } > > -void OCCBody::set_sheet_surface(OCCSurface* surface) > +TopoDS_Compound* OCCBody::get_TopoDS_Shape() > { > - IsSheetBody = CUBIT_TRUE; > - mySheetSurface = surface; > - if(surface == NULL) > - IsSheetBody = CUBIT_FALSE; > + if (myTopoDSShape && !myTopoDSShape->IsNull())

1 0