http://bugs.winehq.org/show_bug.cgi?id=17996
--- Comment #49 from Juan Lang juan_lang@yahoo.com 2009-05-04 22:53:55 --- Well, your valgrind run doesn't show a memory leak either, so I was wrong about that part. I have a handful of patches in inetmib1 that'll make it respect errors from snmpapi. With any luck they'll at least create some test failures before the crash to help narrow it down a little. Posting them in a sec.