[Bug 36615] New: valgrind shows a definite leak in mshtml/tests/htmldoc.c

https://bugs.winehq.org/show_bug.cgi?id=36615

Bug ID: 36615
           Summary: valgrind shows a definite leak in
                    mshtml/tests/htmldoc.c
           Product: Wine
           Version: 1.7.19
          Hardware: x86
                OS: Linux
            Status: NEW
          Keywords: download, source, testcase
          Severity: normal
          Priority: P2
         Component: mshtml
          Assignee: [e-mail]
          Reporter: [e-mail]

==3134== 112 bytes in 1 blocks are definitely lost in loss record 8,688 of
12,948
==3134==    at 0x7BC4C6DF: notify_alloc (heap.c:255)
==3134==    by 0x7BC515C4: RtlReAllocateHeap (heap.c:1883)
==3134==    by 0x7B845879: HeapReAlloc (heap.c:281)
==3134==    by 0x7B8465DA: GlobalReAlloc (heap.c:679)
==3134==    by 0x4FC8BD2: HGLOBALStreamImpl_SetSize (hglobalstream.c:381)
==3134==    by 0x4FC886D: HGLOBALStreamImpl_Write (objidl.h:4391)
==3134==    by 0x6DD059D: PersistHistory_SaveHistory (objidl.h:4384)
==3134==    by 0x4CEC11B: test_save_history (perhist.h:141)
==3134==    by 0x4CEE5F8: OleCommandTarget_Exec (htmldoc.c:3097)
==3134==    by 0x6DCD298: notify_travellog_update (docobj.h:981)
==3134==    by 0x6DCD46F: set_current_mon (persist.c:134)
==3134==    by 0x6DA5B11: navigate_proc (navigate.c:2093)
==3134==    by 0x6DE4D8C: hidden_proc (task.c:316)
==3134==    by 0x55A14F5: ??? (winproc.c:173)
==3134==    by 0x55A166A: call_window_proc (winproc.c:244)
==3134==    by 0x55A2858: WINPROC_CallProcAtoW (winproc.c:603)
==3134==    by 0x55A38FD: WINPROC_call_window (winproc.c:911)
==3134==    by 0x5564859: DispatchMessageA (message.c:3948)
==3134==    by 0x4CF697E: test_download (htmldoc.c:5753)
==3134==    by 0x4CF8BB9: test_put_href (htmldoc.c:6068)
==3134==

[Bug 36615] valgrind shows a definite leak in mshtml/tests/htmldoc.c

https://bugs.winehq.org/show_bug.cgi?id=36615

--- Comment #1 from Austin English  ---
There's also:
==16363== 152 bytes in 1 blocks are definitely lost in loss record 5,149 of
6,194
==16363==    at 0x4008DDA: calloc (vg_replace_malloc.c:619)
==16363==    by 0x4E852BE3: _dl_allocate_tls (in /usr/lib/ld-2.18.so)
==16363==    by 0x4EA7C7B9: pthread_create@@GLIBC_2.1 (in
/usr/lib/libpthread-2.18.so)
==16363==    by 0x7BC8EA23: RtlCreateUserThread (thread.c:544)
==16363==    by 0x7B87CFF0: CreateRemoteThread (thread.c:88)
==16363==    by 0x7B87CF4D: CreateThread (thread.c:54)
==16363==    by 0x727E156: _beginthreadex (thread.c:129)
==16363==    by 0x80C7BBA2: ??? (in
/home/austin/.wine-valgrind/drive_c/windows/system32/gecko/2.24/wine_gecko/nspr4.dll)
==16363==    by 0x80C796AD: ??? (in
/home/austin/.wine-valgrind/drive_c/windows/system32/gecko/2.24/wine_gecko/nspr4.dll)
==16363==

but that's more likely gecko

[Bug 36615] valgrind shows a definite leak in mshtml/tests/htmldoc.c

https://bugs.winehq.org/show_bug.cgi?id=36615

Austin English  changed:

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

[Bug 36615] valgrind shows a definite leak in mshtml/tests/htmldoc.c

http://bugs.winehq.org/show_bug.cgi?id=36615

lizhenbo  changed:

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

--- Comment #2 from lizhenbo  ---
Created attachment 48893
  --> http://bugs.winehq.org/attachment.cgi?id=48893
patch.txt

Hi Austin,
Could you have a look with this patch?
I'm not familiar with valgrind.
Thank you.

[Bug 36615] valgrind shows a definite leak in mshtml/tests/htmldoc.c

https://bugs.winehq.org/show_bug.cgi?id=36615

--- Comment #3 from Austin English  ---
(In reply to lizhenbo from comment #2)
> Created attachment 48893 [details]
> patch.txt
>
> Hi Austin,
> Could you have a look with this patch?
> I'm not familiar with valgrind.
> Thank you.

I can't get the bug to reliably occur to test the patch :/. It only seems to
occur with a full make test, but not the past two I've run (and it takes
several hours to run).

[Bug 36615] valgrind shows a definite leak in mshtml/tests/htmldoc.c

https://bugs.winehq.org/show_bug.cgi?id=36615

Nikolay Sivov  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #48893|0                           |1
        is obsolete|                            |

[Bug 36615] valgrind shows a definite leak in mshtml/tests/htmldoc.c

[WineHQ Bugzilla]

https://bugs.winehq.org/show_bug.cgi?id=36615

Austin English  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED

--- Comment #4 from Austin English  ---
I haven't seen this in a while.

[Bug 36615] valgrind shows a definite leak in mshtml/tests/htmldoc.c

[WineHQ Bugzilla]

https://bugs.winehq.org/show_bug.cgi?id=36615

Alexandre Julliard  changed:

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

--- Comment #5 from Alexandre Julliard  ---
Closing bugs fixed in 4.21.