Thanks, no big problem.  The thing is that the parts I wrote
(Modules/_decimal/*, Objects/memoryobject.c, Modules/_testbuffer.c)
have been audited rather heavily and have 100% code coverage
with a privately maintained patch that inserts allocation
failures.

There are even tests in test_memoryview() / test_buffer() for
breaking up reference cycles.

So I'd really prefer not to have things labeled as potential issues
when there are none. That also applies to the PyMem_NEW() "potential
overflow" fix in _testbuffer.c where overflow was impossible.

Thus, I'll probably revert that at some point, not out of
hostility, but just because it's not my approach to software
development (in the _testbuffer case, I view the 'len' parameter
as a constrained type in the Ada sense where 0 <= len <= ndim = 64).