Hans de Graaff wrote:
> This is the problem, G_FREE_NULL should not have been used there to
> free the chunks. A fix is already in CVS for that, but I'm not sure if
> anonymous CVS has already synced up.

I think we could easily tag memory chunks so that it's obvious whether
they have been g_malloc'ed, walloc'ed or zalloc'ed and verify this in
the free-function (or even have only one and let the free-function
decide.

-- 
Christian

Attachment: pgp00000.pgp
Description: PGP signature

Reply via email to