[Insight-developers] Remaning Valgrind error in TimeVaryingBSpline...

[Bradley Lowekamp]

Hello,

I have managed to whittle down the valgrind errors to just one test on the new valgrind build. This has been just customizing a suppression file up to this point with mostly NrddIO, and a couple dlopen errors, which I have presumed are false positives.

What is now remaining is reported int the itkTimeVaryingBSplineVelocityFieldImageRegistrationTest:

http://open.cdash.org/viewDynamicAnalysisFile.php?id=2891716

The majority of these memory error are coming from ITK code, and a few Nifti zlib io. My intuition is telling me that these are not false positive and need some in depth investigation.

Brad