Troubleshoot S-Function Checks

To help you identify and troubleshoot issues with your S-function model and source code, S-function checks use S-function analyzer APIs to help you detect potential S-function issues.

Environment Checks

Environment checks inspect your environment for S-function analyzer APIs. This check includes the MEX compiler setup check.

MEX Setup Check

This check indicated the presence of the MEX compiler in the current machine. A default C compiler is included in your MATLAB^® installation. To see the supported list of compilers, see Change Default Compiler. You can change the default compiler by typing mex -setup at the command line.

Source Code Checks

These checks inspect the source code for S-functions. These checks include the MEX compile check and Polyspace^® Code Prover™ check.

MEX Compile Check

The MEX compile check uses MEX compilers to identify compiler warnings and errors in the S-function source code.

Polyspace Code Prover Check

If you have a Polyspace license, this check uses Polyspace Code Prover to conduct code analysis. See Source Code Verification with Polyspace Code Prover (Polyspace Code Prover) for more information.

S-Function MEX File Checks

This set of checks inspects the semantics of S-function MEX-file.

MdlOutputsModifyContinuousStates

This check indicates that the continuous states of the S-function are modified in the mdlOutputs method. Continuous states can only be modified at a major time step and requires the ssSetSolverNeedsReset macro to reset the S-function solver. Changing the S-function states without resetting the solver can lead to unexpected results.

MdlUpdateModifyContinuousStates

This check indicates whether the continuous states of the S-function are modified in the mdlUpdate method. Continuous states can only be modified at a major time step and requires the ssSetSolverNeedsReset macro to reset the S-function solver. Changing the S-function states without resetting the solver can lead to unexpected results.

MinorStepModifyDiscreteStates

This check indicates whether the discrete states of the S-function are modified in its mdlOutputs method at a minor step. Discrete states of an S-function can only be modified at a major step, guarded by sslsMajorTimeStep.

MinorStepModifyModes

This check investigates whether the mode vector of the S-function is modified in the mdlOutputs at a minor step. Mode step of the S-function can only be modified at a major time step, guarded by ssIsMajorTimeStep function.

GlobalStaticAsInvisibleState

This check indicates whether the S-function is using static or global variables to represent internal states. To avoid unexpected behavior that results from having the S-function in multiple S-function blocks, declare the S-function states explicitly using ssSetNumDiscreteStates or using data store memory function APIs.

ContinuousStateWithoutContinuousSampleTime

This check inspects whether the S-function continuous sample time is explicitly declared when it has continuous states. You can use ssSetSampleTime to specify continuous sample time when S-function has continuous states.

CombinedMdlOutputsMdlUpdateWithDiscreteState

This check inspects whether the S-function mdlUpdate and mdlOutputs methods are combined when the S-function has discrete states. When the S-function has discrete states, define mdlUpdate and mdlOutputs methods separately and modify discrete states only in the mdlUpdate method.

DeclareCanBeConditionalExecWithState

This check inspects whether the S-function sets the SS_OPTION_CAN_BE_CALLED_CONDITIONALLY option when it has state-like data or multiple sample times. This option could cause the Simulink^® engine to move the S-functions into a conditionally executed mode, such as a conditionally executed subsystem. Remove this option when you have state-like or multiple sample times.

TestHarnessCreationError

This error code indicates that the S-function analyzer encounters an error when creating test harness for the input S-functions in a Simulink library. Possible reasons could be a missing .mex file or mask parameter definitions.

OperatingPointUnknownWithPWork

This error indicates that the S-function has registered PWork and sets the operating point compliance as OPERATING_POINT_COMPLIANCE_UNKNOWN. This is not supported. To support Simulink saving and restoring operating points for this S-function, use USE_CUSTOM_OPERATING_POINT and provide mdlGetOperatingPoint and mdlSetOperatingPoint methods.

ModelCompileError

The S-function analyzer encounters error when compiling the input model. You can try to recompile the model and check the diagnostic viewer for more information.

Input Parameter Robustness Check

This check inspects the robustness of S-functions by specifying incompatible number and type of parameters with certain random values. This check could cause MATLAB to crash. To fix this issue, follow the reproduction steps when you relaunch S-function checks and check the S-function parameter data types and values.