https://bugs.kde.org/show_bug.cgi?id=459047
ronaldaj <vanelbur...@hetnet.nl> changed:
What |Removed |Added
----------------------------------------------------------------------------
Resolution|NOT A BUG |---
Status|RESOLVED |REPORTED
--- Comment #4 from ronaldaj <vanelbur...@hetnet.nl> ---
The testarray is not of the unsigned long type. It is of the type double. If
the code I provided would complain about an operator new[](double) I would
agree that we would be talking about the correct detection of an error.
In the context of Swig it is under Python memory management and I get the same
error.
| You then free pTestarray, but that is an array of raw pointers so freeing it
has no effect on the pointers it contains
| and they remain allocated, so the fact that the first one points to testarray
does not mean testarray gets freed.
The memory the pointers in pTestArray point to are subsets of testarray and do
not represent separately allocated blocks.
delete [] testarray should suffice to free that memory, and it does. But than
the problem doesn't show. I can therefore make it go away in a cpp context but
not when the array comes in from Python.
--
You are receiving this mail because:
You are watching all bug changes.
