Possible False Positive Misra rule 8.4

[vix]

I enabled Misra C 2012 extension through the GUI of CppCheck 2.12.
I cannot use CppCheck 2.13 for this issue

This is my code:
my_file.h

#ifndef MY_FILE_H__
#define MY_FILE_H__
int foo(int a);
#endif

my_file.c

#include "my_file.h"

int foo(int a)
{
    return a;
}

But I ge the violation of misra rule 8.4.
I think this should be a Misra 2012 rule 8.4 (not Misra 2004), but here I can see Misra 2012 rules
https://www.ibm.com/docs/en/devops-test-embedded/8.2.0?topic=review-code-misra-2012-rules
And it should be

A prototype for the global function '%name%' should be declared before defining the function

But as far as I understand, the prototype is in the .h file.
So I cannot understand the reason.

Another tool, that supports Misra 2004 only, notifies me about violation of rule 8.1

Functions shall have prototype and prototype shall be visible at both the definition and call

Are these false positve?
Or is there a way to fix this?

[vix]

Is it possible that this happens because the path of the folder including my_file.h has not been added to cppcheck include paths?
And so the addon doesn't find the file and this produces this behavior?

[Daniel Marjamäki]

yes if you didn't add the include path for my_file.h explicitly then it is very likely.

[Daniel Marjamäki]

In the normal cppcheck analysis I normally don't want that missing declarations leads to false positives. I consider it the recommended and usual case that users do not provide all the include paths for all headers.

In misra checking we have a bit more strict checking because it's used in safety. I think that:
* false positives are not "wanted" but it's better than false negatives. I consider a false positive to be "safe" as it will give you a hint you need to solve something.
* false negatives are not safe. If a misra checker does not know if there is a problem or not because of missing configuration, my intention is that a "misra-config" message will be written.