[Bug 10186] New: Scary heap warnings in ole32 tests

http://bugs.winehq.org/show_bug.cgi?id=10186

Summary: Scary heap warnings in ole32 tests
           Product: Wine
           Version: CVS/GIT
          Platform: Other
        OS/Version: other
            Status: NEW
          Severity: normal
          Priority: P2
         Component: wine-ole
        AssignedTo: [e-mail]
        ReportedBy: [e-mail]

Doing 
  cd dlls/ole32/tests
  rm *.ok
  export WINEDEBUG=warn+heap
  make -k test
yields a couple scary heap warnings.  In particular,

../../../tools/runtest -P wine -M ole32.dll -T ../../.. -p ole32_test.exe.so
hglobalstream.c && touch hglobalstream.ok
warn:heap:GlobalLock invalid handle 0x125852 (Magic: 0x0098)

../../../tools/runtest -P wine -M ole32.dll -T ../../.. -p ole32_test.exe.so
moniker.c
warn:heap:HEAP_ValidateInUseArena Heap 0x110000: unaligned arena pointer
0x12ba7c
warn:heap:HEAP_ValidateInUseArena Heap 0x110000: unaligned arena pointer
0x12bafc

In tests/hglobalstream.c near line 286, there is the code
    hr = CreateStreamOnHGlobal(hglobal, FALSE, &pStream); ...
    GlobalFree(hglobal); ...
    hr = IStream_Read(pStream, buffer, sizeof(buffer), &read);
and the IStream_Read is what triggers the invalid handle warning.
The test code seems quite intentional, and the test passes on wine;
is the warning spurious?

In tests/moniker.c, there's a double release of a pointer at line 937, and
removing one of them makes one warning go away; is this a typo, or intended?
  IEnumMoniker_Release(spEM1); 
  IEnumMoniker_Release(spEM1);