In parse_configuration_models(), it's not an issue whether the calls
succeed or fail. If the entries appear in the model object they are
processed, if not they aren't.
---
The calling sequence is
prov_db_node_set_model_pub()
get_jmodel_obj()
parse_configuration_models()
There seems to be a problem in parse_configuration_models(). In the
loop looking for the model that matches the target_id, if the first
model matches, the model object is returned and the function exits.
If the loop has to iterate, it next calls parse_bindings(). That
function checks if the binding already exists and because it
does, exits false with jmodel set to null.
The consequence is the addition of a second modelId object with the
same model id.
It's not clear why the function exits when it finds the model with the
matching model id and before calling parse_bindings(), etc. I'd think
that the test should go after those calls.
I'm still not sure I completely understand how all this works. It would
also seem that the parse_xxx() functions get called for models up to
the one that matches, but not after.
diff --git a/mesh/prov-db.c b/mesh/prov-db.c
index 04803a5c8..cc3bc8224 100644
--- a/mesh/prov-db.c
+++ b/mesh/prov-db.c
@@ -411,13 +411,10 @@ static bool parse_configuration_models(struct mesh_node *node, int ele_idx,
}
json_object_object_get_ex(jmodel, "bind", &jarray);
- if (jarray && !parse_bindings(node, ele_idx, model_id, jarray))
- return false;
+ parse_bindings(node, ele_idx, model_id, jarray);
json_object_object_get_ex(jmodel, "publish", &jvalue);
-
- if (jvalue && !parse_model_pub(node, ele_idx, model_id, jvalue))
- return false;
+ parse_model_pub(node, ele_idx, model_id, jvalue);
}
return true;
--
To unsubscribe from this list: send the line "unsubscribe linux-bluetooth" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
