Re: [pcp] [PATCH] Fix memory corruption in python support

[Nathan Scott <nathans@xxxxxxxxxx>]

Hi David,

----- Original Message -----
> [...]
> Unfortunately, pcp's python support was decrementing reference counts in
> some cases. I went through and audited the calls to PyArg_ParseTuple(),
> PyArg_ParseTupleAndKeywords(), and PyArg_Parse() and believe I've fixed
> the reference count issues.

Nice work!

> This patch fixed the memory corruption I was seeing.

Can we distil that memory corruption into a reproducible test case?  We
have several tests exercising various aspects of the PMDA wrapper but they
are not triggering this problem - it'd be good to come up with something
(e.g. minimal python script?) that demonstrates the problem and which we
can use to verify the fix.

thanks!

--
Nathan