http://bugs.winehq.org/show_bug.cgi?id=16934
--- Comment #5 from Susan Cragin susancragin@earthlink.net 2009-01-15 08:06:16 --- I will try that later today.
I did have trouble running the regression test. A couple of times the program crashed, but not exactly as I expected it to. So perhaps there are two regressions involved rather than one.
I reported git bad to all crashes, and your patch is what came up.