[Bug 36135] New: valgrind shows some errors in ole32/tests/ole2.c

https://bugs.winehq.org/show_bug.cgi?id=36135

Bug ID: 36135
           Summary: valgrind shows some errors in ole32/tests/ole2.c
           Product: Wine
           Version: 1.7.17
          Hardware: x86
                OS: Linux
            Status: NEW
          Keywords: download, source, testcase
          Severity: normal
          Priority: P2
         Component: ole32
          Assignee: [e-mail]
          Reporter: [e-mail]

==29824== Conditional jump or move depends on uninitialised value(s)
==29824==    at 0x5A3F653: order_rect (gdi_private.h:443)
==29824==    by 0x5A4539D: ExtTextOutW (font.c:2258)
==29824==    by 0x5A44DBB: ExtTextOutA (font.c:2113)
==29824==    by 0x5A6BABF: MF_Play_MetaExtTextOut (metafile.c:1437)
==29824==    by 0x5A69A00: PlayMetaFileRecord (metafile.c:853)
==29824==    by 0x5A68002: PlayMetaFile (metafile.c:429)
==29824==    by 0x52D107C: DataCache_Draw (datacache.c:1594)
==29824==    by 0x4995F58: test_data_cache (oleidl.h:3818)
==29824==    by 0x499BF61: func_ole2 (ole2.c:2387)
==29824==    by 0x49BE544: run_test (test.h:584)
==29824==    by 0x49BE933: main (test.h:654)
==29824==  Uninitialised value was created by a stack allocation
==29824==    at 0x5A6B84B: MF_Play_MetaExtTextOut (metafile.c:1397)
==29824==

==29824== Conditional jump or move depends on uninitialised value(s)
==29824==    at 0x5A3F67F: order_rect (gdi_private.h:449)
==29824==    by 0x5A4539D: ExtTextOutW (font.c:2258)
==29824==    by 0x5A44DBB: ExtTextOutA (font.c:2113)
==29824==    by 0x5A6BABF: MF_Play_MetaExtTextOut (metafile.c:1437)
==29824==    by 0x5A69A00: PlayMetaFileRecord (metafile.c:853)
==29824==    by 0x5A68002: PlayMetaFile (metafile.c:429)
==29824==    by 0x52D107C: DataCache_Draw (datacache.c:1594)
==29824==    by 0x4995F58: test_data_cache (oleidl.h:3818)
==29824==    by 0x499BF61: func_ole2 (ole2.c:2387)
==29824==    by 0x49BE544: run_test (test.h:584)
==29824==    by 0x49BE933: main (test.h:654)
==29824==  Uninitialised value was created by a stack allocation
==29824==    at 0x5A6B84B: MF_Play_MetaExtTextOut (metafile.c:1397)
==29824==

Also a possible leak:
==29824== 8 bytes in 1 blocks are possibly lost in loss record 47 of 708
==29824==    at 0x7BC4C735: notify_alloc (heap.c:255)
==29824==    by 0x7BC50F79: RtlAllocateHeap (heap.c:1716)
==29824==    by 0x7B84586B: HeapAlloc (heap.c:271)
==29824==    by 0x7B845B40: GlobalAlloc (heap.c:376)
==29824==    by 0x4994A77: DataObject_DAdvise (ole2.c:1281)
==29824==    by 0x52D1933: setup_sink (objidl.h:7541)
==29824==    by 0x52D21E9: DataCache_OnRun (datacache.c:2085)
==29824==    by 0x49978AA: test_data_cache (oleidl.h:2656)
==29824==    by 0x499BF61: func_ole2 (ole2.c:2387)
==29824==    by 0x49BE544: run_test (test.h:584)
==29824==    by 0x49BE933: main (test.h:654)
==29824==

[Bug 36135] valgrind shows some errors in ole32/tests/ole2.c

https://bugs.winehq.org/show_bug.cgi?id=36135

Austin English  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |valgrind

[Bug 36135] valgrind shows some errors in ole32/tests/ole2.c

https://bugs.winehq.org/show_bug.cgi?id=36135

Thomas Faller  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

--- Comment #1 from Thomas Faller  ---
I've submitted a patch for the two jump issues:

http://source.winehq.org/patches/data/111712

And it seems to me that the leak was fixed earlier, can you confirm this?

[Bug 36135] valgrind shows some errors in ole32/tests/ole2.c

https://bugs.winehq.org/show_bug.cgi?id=36135

Austin English  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
      Fixed by SHA1|                            |9244cac3afe8723375d2e9e5f0d
                   |                            |3ffaf30b751dc
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED

--- Comment #2 from Austin English  ---
(In reply to Thomas Faller from comment #1)
> I've submitted a patch for the two jump issues:
>
> http://source.winehq.org/patches/data/111712

Works here as well, thanks.
> And it seems to me that the leak was fixed earlier, can you confirm this?

No, I still see the possible leak:
==11516== 112 bytes in 1 blocks are possibly lost in loss record 470 of 745
==11516==    at 0x7BC4A9F1: notify_alloc (heap.c:254)
==11516==    by 0x7BC4EC94: RtlAllocateHeap (heap.c:1715)
==11516==    by 0x5021EEC: DefaultHandler_Construct (defaulthandler.c:1890)
==11516==    by 0x50225E0: OleCreateEmbeddingHelper (defaulthandler.c:2085)
==11516==    by 0x50226D8: OleCreateDefaultHandler (defaulthandler.c:2112)
==11516==    by 0x4B3155B: test_default_handler (ole2.c:1834)
==11516==    by 0x4B357BC: func_ole2 (ole2.c:2578)
==11516==    by 0x4B57E78: run_test (test.h:584)
==11516==    by 0x4B582C0: main (test.h:666)
==11516==

I filed bug 38716 for that.

[Bug 36135] valgrind shows some errors in ole32/tests/ole2.c

https://bugs.winehq.org/show_bug.cgi?id=36135

Alexandre Julliard  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |CLOSED

--- Comment #3 from Alexandre Julliard  ---
Closing bugs fixed in 1.7.45.