https://bugs.winehq.org/show_bug.cgi?id=47380
--- Comment #10 from Henri Verbeet hverbeet@gmail.com --- (In reply to Tim Schumacher from comment #9)
It seems like the finish method is constantly called from the wrong thread (thead 9 seems to be the main thread), causing the thread to constantly stall instead of calling the singlethread finish function.
That's pretty much expected when csmt is enabled. The more interesting information in that log is "queue_id 1" (WINED3D_CS_QUEUE_MAP), which probably means these are caused by wined3d_cs_map() and wined3d_cs_unmap(). (wined3d_cs_emit_update_sub_resource() is possible too, but less likely.)
This is an area that's known to have room for improvement. In some cases it may be possible to avoid going through the CS for the map, in others it may be possible to avoid mapping the resource at all.