Re: [g-a-devel] [Libreoffice] Helgrind results for LO

[Michael Meeks <michael meeks novell com>]

On Wed, 2011-04-06 at 11:42 +0100, Caolán McNamara wrote:
> On Mon, 2011-04-04 at 11:39 +0100, Caolán McNamara wrote:
> > I've got visibility enabled in my desktop env and so a goodly chunk of
> > the /memcheck/ warnings are in the gnome a11y stack, something which I've
> > always some dubious feelings about.

	:-) So - I had a quick poke. I am guessing that you're using at-spi2,
which I eventually found in gnome-git, moved from freedesktop moved from
codethink [ I assume ;-] - that could be made more find-able I think.

	I pushed a couple of trivial patches to at-spi2-core and at-spi2-atk to
stop the leaks, should be there in the next stable iteration I hope.

	Traces appended,

	HTH,

		Michael.

==9540== 3 bytes in 1 blocks are definitely lost in loss record 19 of 5,980
==9540==    at 0x4A0640D: malloc (vg_replace_malloc.c:236)
==9540==    by 0x354FE49980: g_malloc (gmem.c:164)
==9540==    by 0x354FE6024D: g_strdup (gstrfuncs.c:102)
==9540==    by 0x14C8C86B: text_wrapper_get_text (atktext.cxx:305)
==9540==    by 0x15AEACFF: ??? (in /usr/lib64/gtk-2.0/modules/libatk-bridge.so)
==9540==    by 0x3550A1EA9A: signal_emit_unlocked_R (gsignal.c:3218)
==9540==    by 0x3550A28971: g_signal_emit_valist (gsignal.c:2983)
==9540==    by 0x3550A28CFC: g_signal_emit_by_name (gsignal.c:3077)
==9540==    by 0x14C892BC: AtkListener::notifyEvent(com::sun::star::accessibility::AccessibleEventObject const&) (atklistener.cxx:435)

and

==9540== 73 bytes in 1 blocks are definitely lost in loss record 3,687 of 5,980
==9540==    at 0x4A0640D: malloc (vg_replace_malloc.c:236)
==9540==    by 0x354FE49980: g_malloc (gmem.c:164)
==9540==    by 0x354FE6024D: g_strdup (gstrfuncs.c:102)
==9540==    by 0x15D17168: atspi_get_a11y_bus (in /usr/lib64/libatspi.so.0.0.1)
==9540==    by 0x15AE7E2C: ??? (in /usr/lib64/gtk-2.0/modules/libatk-bridge.so)
==9540==    by 0x3DB0D6B420: default_display_notify_cb (gtkmodules.c:439)
==9540==    by 0x3550A0E2ED: g_closure_invoke (gclosure.c:767)
==9540==    by 0x3550A1F19F: signal_emit_unlocked_R (gsignal.c:3252)


-- 
 michael meeks novell com  <><, Pseudo Engineer, itinerant idiot