[Bug 20920] New: Regression - read buffer overflow in StorageImpl_Construct in many msi tests

http://bugs.winehq.org/show_bug.cgi?id=20920

Summary: Regression - read buffer overflow in
                    StorageImpl_Construct in many msi tests
           Product: Wine
           Version: 1.1.34
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: ole32
        AssignedTo: [e-mail]
        ReportedBy: [e-mail]

http://kegel.com/wine/valgrind/logs/2009-12-04-18.45/diff-msi_msi.txt
http://kegel.com/wine/valgrind/logs/2009-12-04-18.45/vg-msi_msi.txt
and the rest of the msi logs for today show a regression like this:
 Invalid read of size 1
    at  memcpy (mc_replace_strmem.c:482)
    by  StorageImpl_Construct (storage32.c:2280)
    by  StgCreateDocfile (storage32.c:5696)
    by  MSI_OpenDatabaseW (database.c:121)
    by  MsiOpenDatabaseW (database.c:228)
    by  MsiOpenDatabaseA (database.c:263)
    by  create_package_db (msi.c:150)
    by  test_MsiOpenProduct (msi.c:7194)
  Address 0x7f0459a2 is 0 bytes after a block of size 26 alloc'd
    at  notify_alloc (heap.c:247)
    by  RtlAllocateHeap (heap.c:1697)
    by  msi_alloc (msipriv.h:1078)
    by  strdupAtoW (msipriv.h:1124)
    by  MsiOpenDatabaseA (database.c:249)
    by  create_package_db (msi.c:150)
    by  test_MsiOpenProduct (msi.c:7194)

Did today's ole32 storage changes introduce it?  This change looks like
it's related:

http://www.winehq.org/pipermail/wine-cvs/2009-December/062265.html

@@ -2295,6 +2271,9 @@ static HRESULT StorageImpl_Construct(
          goto end;
       }
       strcpyW(This->pwcsName, pwcsName);
+
+      memcpy(This->base.filename, pwcsName, DIRENTRY_NAME_BUFFER_LEN-1);
+      This->base.filename[DIRENTRY_NAME_BUFFER_LEN-1] = 0;