core: Check out asynchronously

This can be a large performance win in certain circumstances:

 * Cold buffer cache (we don't block the whole process)
 * Requiring a copy instead of hardlink

tests/ostree-valgrind.supp

@@ -189,3 +189,11 @@
    fun:_g_io_module_get_default
    ...
 }
+
+{
+   _dl_allocate_tls
+   Memcheck:Leak
+   ...
+   fun:_dl_allocate_tls
+   ...
+}