From: Andy D. <dou...@la...> - 2002-08-29 21:06:29
|
On Wed, 28 Aug 2002, rod holland wrote: > I am revisiting my compile problems. I gave up for a month or so but I > would really like to get VPython running on Mandrake 8.2. I have attached > the compilation output. In xgl.h; if I use /gtk/... then it makes it to > "convex" before segfaulting. If I use gtk-1.2/gtk/... as suggested on the > web site then it segfaults in the "box" module. I have included a full > output for a "convex" case and the last few lines of three successive runs. > The case where it fails in "box" is similar. > > Are my libraries incompatible - how can I test and fix? I really do not > know what to do from this point? > g++ -I. -I./CXX/Include -I/usr/local/include/python2.2 > -I/usr/local/include/python2.2/Numeric -I/usr/include/gtk-1.2 > -I/usr/include/glib-1.2 -I/usr/lib/glib/include -D_REENTRANT > -I/usr/X11R6/include -w -c -o convex.o convex.cpp > In file included from pvector.h:6, > from cvisual.h:6, > from display.h:5, > from arrprim.h:5, > from convex.cpp:2: > CXX/Include/CXX_Objects.h:1039: Internal error: Segmentation fault. > Please submit a full bug report. > See <URL:https://qa.mandrakesoft.com/> for instructions. My suspicion today is the same as it was the last time you reported this. It should be impossible for Vpython to cause g++ to generate an "Internal error" and cause a Segmentation fault. Did you follow the suggestion above and "submit a full bug report"? One very real possiblity is a compiler bug of some sort. Also, the last time you reported this, you noted that running the same command several times in succession gave different segmentation faults. That's often a sign of either a hardware problem or a disk file system corruption problem. Did you follow up on my suggestions last month to look into those issues? -- Andy Dougherty dou...@la... Dept. of Physics Lafayette College, Easton PA 18042 |