https://bugs.winehq.org/show_bug.cgi?id=37773
--- Comment #4 from felix.huber@schyf.de --- It seems like some recent changes to heap.c boosted the performance to about 90% of the original timing. That's good enough for me. This report can be closed.