[vtk-developers] 3DMorph

[Sean McBride]

On Fri, 1 Mar 2013 10:29:31 -0500, Bill Lorensen said:

>When you get a chance, can you run either the Tcl or Python version of
>3DMorph on RogueResearch11? They are both segfaulting.

I looked a little more closely.  launchd has nothing to do with it.  It segfaults at night, and not 'by hand', only because 'by hand' I was not setting all the same debug variables as my script does.  I can repro the crash easily as follows:

$ export MallocGuardEdges=1

$ ctest -R 3DMorph
ctest(53119) malloc: protecting edges
Test project /Users/builder/external/VTK-clang-dbg-x86_64
    Start 664: vtkFiltersHybridPython-3DMorph
1/2 Test #664: vtkFiltersHybridPython-3DMorph ...***Exception: SegFault  6.83 sec
    Start 698: vtkFiltersHybridTcl-3DMorph
2/2 Test #698: vtkFiltersHybridTcl-3DMorph ......***Exception: SegFault  6.03 sec

Still, I use MallocGuardEdges on all my debug builds, and it's only crashing on this one machine, and crashing in the OpenGL driver code, so it's probably Apple/ATI's fault.

Can someone check/confirm that the 3DMorph tests pass with valgrind on some other machine?  Then I'll suppress them on Rogue11.  BTW, how con one check that?  I don't see 'morph' on this page:

<http://open.cdash.org/viewDynamicAnalysis.php?buildid=2834731>

But that page only shows errors, right?  How do I know if it even ran there?

Cheers,

-- 
____________________________________________________________
Sean McBride, B. Eng                 sean at rogue-research.com
Rogue Research                        www.rogue-research.com 
Mac Software Developer              Montréal, Québec, Canada