On Thu, 2011-11-17 at 21:12 -0500, August Sodora wrote:
> So hooray we save 160 bytes.
Great :-) of course, that is the oosplash wrapper that -should- exit
rather sooner than it does IMHO (I still need to look into that for
3.5). The main interest would be in running valgrind on soffice.bin (I
assume you're following forks ?).
> The second initially makes me think that
> a matching XCloseDisplay is missing but I'm not sure why the call that
> is already there isn't getting called if the first leak was fixed:
>
> ==11027== 124 bytes in 1 blocks are definitely lost in loss record 6 of 8
> ==11027== at 0x4A06031: malloc (vg_replace_malloc.c:236)
> ==11027== by 0x3FA060BD79: get_peer_sock_name (xcb_auth.c:259)
> ==11027== by 0x3FA060BE45: _xcb_get_auth_info (xcb_auth.c:302)
> ==11027== by 0x3FA060B84E: xcb_connect_to_display_with_auth_info
> (xcb_util.c:424)
Of course this could just as easily be a leak in the X server; valgrind
finds them all over the place :-) anyhow, good to have you chasing that.
To debug leaks in soffice.bin you really need to:
export G_SLICE=1
Which makes LibO use the malloc allocator which valgrind can track
instead of it's own internal one. If you have run 'linkoo' then source
ooenv in program/ should set all that good stuff for you :-)
Thanks !
Michael.
--
michael.me...@suse.com <><, Pseudo Engineer, itinerant idiot
_______________________________________________
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice
