Clusters with delegates don't seem to work when used in Bridged Nodes (CON-1900)

[tomasmcguinness]

I am attempting to add a Dishwasher Device Type as a Bridged Node.

Are the delegates for OperationalState and DishwasherMode fully supported?

If I define a Dishwasher endpoint like this

node::config_t node_config;
node_t *node = node::create(&node_config, app_attribute_update_cb, app_identification_cb);
ABORT_APP_ON_FAILURE(node != nullptr, ESP_LOGE(TAG, "Failed to create Matter node"));

static OperationalStateDelegate operational_state_delegate;

dish_washer::config_t dish_washer_config;
dish_washer_config.operational_state.delegate = &operational_state_delegate;
endpoint_t *ep = dish_washer::create(node, &dish_washer_config, ENDPOINT_FLAG_NONE, NULL);

esp_matter::cluster_t *operational_state_cluster = esp_matter::cluster::get(ep, chip::app::Clusters::OperationalState::Id);

esp_matter::cluster::operational_state::attribute::create_countdown_time(operational_state_cluster, 0);

esp_matter::cluster::operational_state::command::create_start(operational_state_cluster);
esp_matter::cluster::operational_state::command::create_stop(operational_state_cluster);
esp_matter::cluster::operational_state::command::create_pause(operational_state_cluster);
esp_matter::cluster::operational_state::command::create_resume(operational_state_cluster);

static DishwasherModeDelegate dishwasher_mode_delegate;

esp_matter::cluster::dish_washer_mode::config_t dish_washer_mode_config;
dish_washer_mode_config.delegate = &dishwasher_mode_delegate;
dish_washer_mode_config.current_mode = ModeNormal;

esp_matter::cluster_t *dishwasher_mode_cluster = esp_matter::cluster::dish_washer_mode::create(ep, &dish_washer_mode_config, CLUSTER_FLAG_SERVER);

esp_matter::cluster::mode_base::attribute::create_supported_modes(dishwasher_mode_cluster, NULL, 0, 0);

esp_matter::cluster::mode_base::command::create_change_to_mode(dishwasher_mode_cluster);

It works perfectly and I can see the device in Home Assistant

If set up a bridge

node::config_t node_config;
node_t *node = node::create(&node_config, app_attribute_update_cb, app_identification_cb);
ABORT_APP_ON_FAILURE(node != nullptr, ESP_LOGE(TAG, "Failed to create Matter node"));

aggregator::config_t aggregator_config;
endpoint_t *aggregator = endpoint::aggregator::create(node, &aggregator_config, ENDPOINT_FLAG_NONE, NULL);
ABORT_APP_ON_FAILURE(aggregator != nullptr, ESP_LOGE(TAG, "Failed to create aggregator endpoint"));

err = app_bridge_initialize(node, create_bridge_devices);
ABORT_APP_ON_FAILURE(err == ESP_OK, ESP_LOGE(TAG, "Failed to resume the bridged endpoints: %d", err));

app_bridge_create_bridged_device(node::get(), aggregator_endpoint_id, ESP_MATTER_DISH_WASHER_DEVICE_TYPE_ID, NULL);

The create_bridge_devices callback uses the same dishwasher code as before, using dish_washer::add instead of dish_washer::create.

I get lots of errors about the OperationalStatus and DishwashMode clusters:

E (6028) data_model: get_val_internal(927): Attribute is not managed by esp matter data model
E (6028) chip[DMG]: Fail to retrieve data, roll back and encode status on clusterId: 0x0000_0060, attributeId: 0x0000_0000err = 501
E (6038) data_model: get_val_internal(927): Attribute is not managed by esp matter data model
E (6048) chip[DMG]: Fail to retrieve data, roll back and encode status on clusterId: 0x0000_0060, attributeId: 0x0000_0001err = 501
E (6058) data_model: get_val_internal(927): Attribute is not managed by esp matter data model
E (6068) chip[DMG]: Fail to retrieve data, roll back and encode status on clusterId: 0x0000_0060, attributeId: 0x0000_0003err = 501
E (6078) data_model: get_val_internal(927): Attribute is not managed by esp matter data model
E (6088) chip[DMG]: Fail to retrieve data, roll back and encode status on clusterId: 0x0000_0060, attributeId: 0x0000_0004err = 501
E (6098) data_model: get_val_internal(927): Attribute is not managed by esp matter data model
E (6108) chip[DMG]: Fail to retrieve data, roll back and encode status on clusterId: 0x0000_0060, attributeId: 0x0000_0005err = 501
E (6118) data_model: get_val_internal(927): Attribute is not managed by esp matter data model
E (6128) chip[DMG]: Fail to retrieve data, roll back and encode status on clusterId: 0x0000_0060, attributeId: 0x0000_0002err = 501
E (6138) data_model: get_val_internal(927): Attribute is not managed by esp matter data model
E (6148) chip[DMG]: Fail to retrieve data, roll back and encode status on clusterId: 0x0000_0059, attributeId: 0x0000_0001err = 501
E (6158) data_model: get_val_internal(927): Attribute is not managed by esp matter data model
E (6168) chip[DMG]: Fail to retrieve data, roll back and encode status on clusterId: 0x0000_0059, attributeId: 0x0000_0000err = 501

Any ideas?

This feels related to the Delegates for both of these attributes, but I can't see what the problem is.

[jadhavrohit924]

@tomasmcguinness Thanks for the report!! We will look into it.

[tomasmcguinness]

I am wondering if the issue is due to the delegate callbacks not being called.

During the normal setup, the init_callback is invoked, which registers the delegate (esp_matter_core.cpp)

if (endpoint::enable_all() != ESP_OK) {
    ESP_LOGE(TAG, "Enable all endpoints failure");
}
// Initialise clusters which have delegate implemented
esp_matter::cluster::delegate_init_callback_common();

[tomasmcguinness]

Confirmed. If I invoke delegate_init_callback_common() after adding an endpoint to the bridge or after restoring the bridged endpoints, my dishwasher works as expected.