[Bug 36601] New: valgrind shows a definite leak in wininet/tests/internet.c

https://bugs.winehq.org/show_bug.cgi?id=36601

            Bug ID: 36601
           Summary: valgrind shows a definite leak in
                    wininet/tests/internet.c
           Product: Wine
           Version: 1.7.19
          Hardware: x86
                OS: Linux
            Status: NEW
          Keywords: download, source, testcase
          Severity: normal
          Priority: P2
         Component: wininet
          Assignee: wine-bugs(a)winehq.org
          Reporter: austinenglish(a)gmail.com

==19689== 2 bytes in 1 blocks are definitely lost in loss record 8 of 555
==19689==    at 0x7BC4C6DF: notify_alloc (heap.c:255)
==19689==    by 0x7BC50F23: RtlAllocateHeap (heap.c:1716)
==19689==    by 0x4E59F79: heap_alloc (internet.h:116)
==19689==    by 0x4E5D203: InternetGetConnectedStateExA (internet.c:1292)
==19689==    by 0x4CD3B0D: test_InternetGetConnectedStateExA (internet.c:1514)
==19689==    by 0x4CD452F: func_internet (internet.c:1616)
==19689==    by 0x4CE1B13: run_test (test.h:584)
==19689==    by 0x4CE1F02: main (test.h:654)
==19689== 

and a possible leak:

==19689== 232 bytes in 1 blocks are possibly lost in loss record 427 of 555
==19689==    at 0x7BC4C6DF: notify_alloc (heap.c:255)
==19689==    by 0x7BC50F23: RtlAllocateHeap (heap.c:1716)
==19689==    by 0x4E48FF5: heap_alloc_zero (internet.h:121)
==19689==    by 0x4E49608: get_server (http.c:301)
==19689==    by 0x4E51CD4: HTTP_HttpOpenRequestW (http.c:3318)
==19689==    by 0x4E5226F: HttpOpenRequestW (http.c:3431)
==19689==    by 0x4E4C9E2: HttpOpenRequestA (http.c:1482)
==19689==    by 0x4CD2D81: test_InternetErrorDlg (internet.c:1354)
==19689==    by 0x4CD455C: func_internet (internet.c:1625)
==19689==    by 0x4CE1B13: run_test (test.h:584)
==19689==    by 0x4CE1F02: main (test.h:654)
==19689==

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.

[Bug 36601] valgrind shows a definite leak in wininet/tests/internet.c

https://bugs.winehq.org/show_bug.cgi?id=36601

Austin English <austinenglish(a)gmail.com> changed:

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

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.

[Bug 36601] valgrind shows a definite leak in wininet/tests/internet.c

http://bugs.winehq.org/show_bug.cgi?id=36601

Jacek Caban <jacek(a)codeweavers.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jacek(a)codeweavers.com

--- Comment #1 from Jacek Caban <jacek(a)codeweavers.com> ---
I sent a patch for the first leak:

http://source.winehq.org/patches/data/105716

The second one is not a real one. We free this on DLL unload, but not on
process unload. That's by design. I believe that many other DLLs have similar
cases. Maybe we should always pretend regular unload when ran on Valgrind?

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.

[Bug 36601] valgrind shows a definite leak in wininet/tests/internet.c

https://bugs.winehq.org/show_bug.cgi?id=36601

Austin English <austinenglish(a)gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Fixed by SHA1|                            |e1dbfb762c5991bba130bceb5f9
                   |                            |ddfe35ff92def
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED

--- Comment #2 from Austin English <austinenglish(a)gmail.com> ---
(In reply to Jacek Caban from comment #1)
> I sent a patch for the first leak:
>
> http://source.winehq.org/patches/data/105716

https://source.winehq.org/git/wine.git/commitdiff/e1dbfb762c5991bba130bceb5f...
> The second one is not a real one. We free this on DLL unload, but not on
> process unload. That's by design. I believe that many other DLLs have
> similar cases. Maybe we should always pretend regular unload when ran on
> Valgrind?

I've marked it as intentional, thanks.

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.

[Bug 36601] valgrind shows a definite leak in wininet/tests/internet.c

https://bugs.winehq.org/show_bug.cgi?id=36601

Alexandre Julliard <julliard(a)winehq.org> changed:

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

--- Comment #3 from Alexandre Julliard <julliard(a)winehq.org> ---
Closing bugs fixed in 1.7.23.

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.