Re: [PATCH 0/9] MISRA C 2012 8.1 rule fixes

[Stefano Stabellini <sstabellini@kernel.org>]

On Thu, 23 Jun 2022, Jan Beulich wrote:
> On 23.06.2022 09:37, Roberto Bagnara wrote:
> > Rule 8.1 only applies to C90 code, as all the violating instances are
> > syntax errors in C99 and later versions of the language.  So,
> > the following line does not contain a violation of Rule 8.1:
> > 
> >      unsigned x;
> > 
> > It does contain a violation of Directive 4.6, though, whose correct
> > handling depends on the intention (uint32_t, uin64_t, size_t, ...).

Hi Roberto,

Thank you very much for the quick reply and very clear answer!


> Interesting - this goes straight against a rule we have set in
> ./CODING_STYLE. I'm also puzzled by you including size_t in your list
> of examples, when the spec doesn't. The sole "goal" of the directive
> (which is advisory only anyway) is to be able to determine allocation
> size. size_t size, however, varies as much as short, int, long, etc
> do.

I wouldn't worry about Directive 4.6 for now. We'll talk about it when
we get to it. (Also we already require uint32_t, uint64_t, etc. in all
external interfaces and ABIs which I think is what Dir 4.6 cares about
the most.)

For this series, I suggest to keep the patches because "unsigned int" is
better than "unsigned" from a style perspective, but we need to rephrase
the commit messages because we cannot claim they are fixing Rule 8.1.
Also, thank Jan for spotting the misunderstanding!