The commit with the same name suppresses some small memory leaks, in order to not hide more serious problems (of which there are none at the moment). Need to investigate these leaks in more detail and come up with proper fixes.
Investigated and fixed: commit d9bd10237ffe6349811c44ef98887b8ba7e4128d Author: Patrick Ohly <patrick.ohly@intel.com> Date: Wed Oct 31 22:30:54 2012 -0700 gee: fix memory leak in GeeCollCXX gee_collection_iterator_get() always transfers ownership. The caller must g_free() (for strings) or unref (for GObjects) the returned value. The revised GeeCollCXX does this automatically, with GObject being the default (relies on the GObject intrusive pointer template classes) and "gchar *" a special case in the template specialization of an utility traits class. The usage is almost as before. However, because of the reliance on SE_GOBJECT_TYPE, that macro must have been used for all types over which is to be iterated. Requires moving some definitions into header files where they can be shared. commit 49d300e3a3015f32a4f467ef402a1f617deeba85 Author: Patrick Ohly <patrick.ohly@intel.com> Date: Wed Oct 31 22:15:00 2012 -0700 glib: fixed memory leak in some GAsyncReady helpers The callback was allocated with new for the duration of the operation and must be deleted. Was done via try/catch in some cases, but not all. The better solution is to use RAII and store the instance in an auto_ptr as soon as that is possible again.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.