MISRA-C-2012

I generate code with Embedded Coder with the objective of MISRA-C-2012 (It is the only 'generation objective' I set in this generation).

It is a complex model that generates more than 30KLOC.

But, when I analyze such code with Polyspace Code Prover I obtain more than 1K rule violations.

Should not be it zero discrepancies?

Is there any other addition configuration in both tools to manage this objective?

Kind regards

Answer 1

[Kausthub]

Hi Luis,

Yes, the aim is for zero discrepancies but the code generated by Embedded Coder is not completelty MISRA C:2012 compliant and has a few violations.

But, you could reduce the number of violations by selecting the right properties from the configuration panel of Embedded Coder. For example you can reduce a lot of MISRA C:2012 violations by setting the casting mode from 'Nomial' to 'Standard Compliance', restricting signed left and right shifts, enabling justifications etc.

Here is an article on how you could reduce MISRA C:2012 violations:

https://www.mathworks.com/help/ecoder/ug/generate-cc-code-with-improved-misra-compliance.html

Hope it helps!