[SciPy-user] Memory leak in delaunay interpolator

[Pauli Virtanen]

Thu, 17 Jan 2008 10:02:02 +0100, Rob Hetland wrote:

> On Jan 17, 2008, at 7:10 AM, Robert Kern wrote:
> 
> 
>>> I am not sure if the patch from #382 has been applied to svn trunk, I
>>> am
>>> still finding my way around scipy.
>>
>> It got applied in the scikits tree, where it now lives.
> 
> 
> I just reinstalled the latest scikits.delaunay package, and I am getting
> a segfault right away, in the first interp statement.  With the older
> version from the scipy.sandbox, I was getting a slow memory leak that
> would segfault after many (thousands of) interpolator/ interp calls.
> 
> These code bases are different, primarily in the _delaunay.cpp file.
> However, my C is bad, and my C++ is worse, so I am useless beyond
> diffing the files...
> 
> I just wanted to point out that there are perhaps two separate issues
> here.

The current version [1] of 
scikits.delaunay:_delaunay.cpp:nn_interpolate_unstructured_method
appears to XDECREF the variable intz before returning it. Is this a 
reference count error?

[1] http://svn.scipy.org/svn/scikits/trunk/delaunay/scikits/delaunay/
_delaunay.cpp

-- 
Pauli Virtanen