The when clause in an Error flow declaration does not prevent error state propagation.

[fletchwsf]

The when clause in an Error flow declaration does not prevent error state propagation. In the following code snippet, the out going propagation supports two error states, with a when clause that should restrict propagation to specific error conditions. Specifically, when the component transitions to critical failure mode it should 'transmit' {badData} and in non-critical failure mode it should transmit {noData}. Instead of the expected behaviour, the Fault Impact Assessment Report shows 6 entries associated with the computer instances. Two entries for critical and non-critical mode are correct. The detected fault lists two entries, one entry for {badData} and one entry for {noData}. This condition should only propagate {noData}.

The full code listing is available from the gitHub link beow, see package named 'computer'.

error propagations
fault from the power supplyflightControlDataOut: out propagation {badData,noData};
flows
ers1: error source flightControlDataOut{noData} when NonCriticalModeFailure;
ers2: error source flightControlDataOut{badData} when CriticalModeFailure;

https://github.com/fletch787/Models.git
See branch Bug Repot 2014-07-20

E.7.2 Error Flow Declarations
error_flow ::=
error_source | error_sink | error_path
error_source ::=
defining_error_source_identifier : error source
outgoing_error_propagation_point_reference [ error_type_set ]
[ when fault_source ] ;
fault_source ::= error_behavior_state | error_type_set_reference