[Insight-developers] bug in itkPeriodicBoundaryCondition

[Torsten Rohlfing]

Hi --

I think I ran into a bug in the 
itk::PeriodicBoundaryCondition::operator() implementations. The problem 
is that a local ConstNeighborhoodIterator is created, but the type of 
this iterator does not get the PeriodicBoundaryCondition type as its 
second template parameter. Thus, a dynamic cast from the function 
parameter to the local pointer gives NULL, and the code segfaults. This 
occurs twice, once in each of the two operator() functions.

Here's a diff against the latest CVS version:

Index: itkPeriodicBoundaryCondition.txx
===================================================================
RCS file: 
/cvsroot/Insight/Insight/Code/Common/itkPeriodicBoundaryCondition.txx,v
retrieving revision 1.6
diff -u -r1.6 itkPeriodicBoundaryCondition.txx
--- itkPeriodicBoundaryCondition.txx    20 Apr 2006 17:01:27 -0000    1.6
+++ itkPeriodicBoundaryCondition.txx    15 Jan 2007 23:35:58 -0000
@@ -27,8 +27,8 @@
              const NeighborhoodType *data) const
 {
   typedef typename OffsetType::OffsetValueType OffsetValueType;
-  const ConstNeighborhoodIterator<TImage> * iterator
-    = dynamic_cast<const ConstNeighborhoodIterator<TImage> *>(data);
+  const ConstNeighborhoodIterator<TImage,Self> * iterator
+    = dynamic_cast<const ConstNeighborhoodIterator<TImage,Self> *>(data);
   typename TImage::PixelType *ptr;
   int linear_index = 0;
   unsigned int i;
@@ -81,8 +81,8 @@
              const NeighborhoodAccessorFunctorType 
&neighborhoodAccessorFunctor) const
 {
   typedef typename OffsetType::OffsetValueType OffsetValueType;
-  const ConstNeighborhoodIterator<TImage> * iterator
-    = dynamic_cast<const ConstNeighborhoodIterator<TImage> *>(data);
+  const ConstNeighborhoodIterator<TImage,Self> * iterator
+    = dynamic_cast<const ConstNeighborhoodIterator<TImage,Self> *>(data);
   typename TImage::PixelType *ptr;
   int linear_index = 0;
   unsigned int i;

-- 
Torsten Rohlfing, PhD          SRI International, Neuroscience Program
 Research Scientist             333 Ravenswood Ave, Menlo Park, CA 94025
  Phone: ++1 (650) 859-3379      Fax: ++1 (650) 859-2743
   torsten at synapse.sri.com        http://www.stanford.edu/~rohlfing/

     "Though this be madness, yet there is a method in't"

-------------- next part --------------
A non-text attachment was scrubbed...
Name: torsten.vcf
Type: text/x-vcard
Size: 366 bytes
Desc: not available
Url : http://www.itk.org/mailman/private/insight-developers/attachments/20070115/99f80ab6/torsten.vcf