https://bugs.winehq.org/show_bug.cgi?id=49256
--- Comment #2 from Alan A. grsfdhj@tiscali.it --- Good catch, I created a debug trace with +relay,+ntdll etc. but I got a 2GB+ file that compresses to a ~75MB xz file with maximum compression settings... I will try to shrink it down to the minimum and upload it ASAP.
From what I can grasp from your trace, we have a single thread (tid=0098) that
throws a number of exceptions (nested exceptions, I presume) -- all with code "c0000005", which should be the code for access violation. Before the first exception (traced back by searching for the "raise_exception code" string) there is a call to ntdll.NtUnmapViewOfSection which returns a NTSTATUS=0 value (which should mean a success status: https://docs.microsoft.com/en-us/openspecs/windows_protocols/ms-erref/596a10...):
140202.520:015f:0098:Call ntdll.NtUnmapViewOfSection(ffffffffffffffff,016d0000) ret=14097752a 140202.520:015f:0098:Ret ntdll.NtUnmapViewOfSection() retval=00000000 ret=14097752a 140202.520:015f:0098:trace:seh:raise_exception code=c0000005 flags=0 addr=0x1582f17da ip=1582f17da tid=0098
Judging from the "ret" and exception "ip" addresses, I guess there's some other code between returning from the ntdll.NtUnmapViewOfSection() function and raising the exception, however.