http://bugs.winehq.org/show_bug.cgi?id=20358
--- Comment #19 from Juan Lang juan_lang@yahoo.com 2009-11-03 13:42:00 --- I sent a patch that demonstrates the problem a little more clearly: http://www.winehq.org/pipermail/wine-patches/2009-November/080871.html
It's still a little hard to tell the exact problem, but it appears to be a bug in how integers with the high bit are exported. They should have a 0 byte prepended to them to prevent them from being interpreted as signed. I believe the bug is in mp_to_unsigned in rsaenh's mpi.c, but that's as far as I've gotten.