Recommended Model Configuration Parameters for Polyspace Analysis
Before analyzing the generated code, check that your model configuration is optimized
for a Polyspace® analysis. Set the following parameters to the recommended value. If you do
not use the recommended value for SystemTargetFile, you get an error.
For other parameters, if you do not use the recommended value, you get a
warning.
| Grouping | Command-Line | Name and Location in Configuration | Rationale for Recommendation |
|---|---|---|---|
| Code Generation | Name: Value: An Embedded Coder® Target Language Compiler (TLC) file. For example | Location: Code Generation Name: System target file Value: Embedded Coder target file | Polyspace supports targets based on
|
Name: Value:
| Location: Code Generation > Interface Name: MAT-file logging Value:
| Disabling MAT file logging enhances precision of Polyspace analysis | |
Name: Value:
| Location: Code Generation > Templates Name: Generate an example main program Value:
| Polyspace removes the example main before analysis if an example main is generated. | |
Name: Value:
| Location: Code Generation > Comments Name: Include comments Value:
| Including comments allow navigating from the generated code to the corresponding parts of the model. Additionally, Embedded Coder can insert comments that justify come coding rule violations automatically. | |
| Optimization | Name: Value:
| Location: Optimization Name: Default parameter behavior Value:
| The value Inlined enhances precision of
Polyspace analysis. If you want to specify external constraints
on the parameters, set the value to
Tunable. |
Name: Value:
| Location: Optimization Name: Use memset to initialize floats and doubles to 0.0 Value:
| Using memset() to initialize floats and
doubles to 0.0 can cause additional violations
and orange checks in a Polyspace analysis. | |
Name: Value:
| Location: Optimization Name: Remove root level I/O zero initialization Value:
| Not selecting this configuration parameter can result in false positive violations or more orange checks. | |
| Solver | Name: Value:
| Location: Solver Name: Type Value:
| Using fixed step solvers enhances the precision of Polyspace analysis. |
Name: Value:
| Location: Solver Name: Solver Value:
| Using discrete solvers enhances the precision of Polyspace analysis. |
In addition to setting the preceding model configuration parameters, you can use the Code generation Advisor to check the compatibility of your model with a Polyspace analysis of the generated code. See Analyze Code and Perform Software-in-the-Loop Testing (Simulink Check).
