[Insight-developers] Memory leak question.

[Mark Foskey]

I'm getting the following memory leak message from Purify off the 
dashboard.  I'm including relevant lines of code so that you can see 
what's going on:

----------------------------------

[W] MLK: Memory leak of 20 bytes from 1 block allocated in 
itk::CellBoundary::itk,VectorContainer::itk>::itk,set::std,allocator::std>::std>::itk>::itk>::itk>::AddUsingCell(D
         Distribution of leaked blocks
                 20 bytes from 1 block of 20 bytes (0x05aa2c60)
         Allocation location
 
std::_Tree::std,allocator::std>::std,less::std,allocator::std>::_Insert(_Node::_Tree::std,allocator::std>::std,less::std,allocator:: 
[xtree:462]
 
itk::CellBoundary::itk,VectorContainer::itk>::itk,set::std,allocator::std>::std>::itk>::itk>::itk>::AddUsingCell(D 
[itkCellBoundary.txx:44]

void
CellBoundary<TCell>
::AddUsingCell(CellIdentifier cellId)
{
   m_UsingCells.insert(cellId);  // cellID is an unsigned long
}
 
itk::AutomaticTopologyMeshSource::itk>::itk>::AddHexahedron(Array::itk 
const&) [itkAutomaticTopologyMeshSource.txx:672]

:      boundaryFeature->AddUsingCell( cellID );
 
itk::AutomaticTopologyMeshSource::itk>::itk>::AddHexahedron(DWORD,DWORD,DWORD,DWORD,DWORD,DWORD,DWORD,DWORD) 
[itkAutomaticTopologyMeshSource.txx:777]

:  return AddHexahedron( pointIDs );
:  // pointIDs is an itk::Array of unsigned longs


             itkAutomaticTopologyMeshSourceTest(int,char * * const) 
[itkAutomaticTopologyMeshSourceTest.cxx:165]
             main           [itkTestMain.h:127]
             mainCRTStartup [chkstk.obj]

-------------------------------------------

 From the purify report, it looks like the allocation is happening 
inside the xtree standard library code, which is part of the 
implementation of std::set.  I believe there have been problems with 
Purify detecting spurious memory leaks with at least some versions of 
the STL, and it certainly looks to me like that's what's happening 
here.  Is there something obvious that I'm missing, or, if not, is 
there a standard way to keep these Purify errors from appearing?

-- 
Mark Foskey    (919) 843-5436  Computer-Aided Diagnosis and Display Lab
mark_foskey at unc . edu            Department of Radiology, CB 7515, UNC
http://www . cs . unc . edu/~foskey  Chapel Hill, NC  27599-7515