Re: [PATCH 8/9] xen: add deviations for Rule 11.8

[Jan Beulich <jbeulich@xxxxxxxx>]

On 14.12.2023 13:07, Simone Ballarin wrote:
> --- a/automation/eclair_analysis/ECLAIR/deviations.ecl
> +++ b/automation/eclair_analysis/ECLAIR/deviations.ecl
> @@ -292,6 +292,18 @@ constant expressions are required.\""
>  # Series 11
>  #
>  
> +-doc_begin="Violations caused by container_of are due to pointer arithmetic 
> operations
> +with the provided offset. The resulting pointer is then immediately cast 
> back to its
> +original type, which preserves the qualifier. This use can be deemed as safe.
> +Fixing this violation would require to increase code complexity and lower 
> readability."
> +-config=MC3R1.R11.8,reports+={safe,"any_area(any_loc(any_exp(macro(^container_of$))))"}
> +-doc_end
> +
> +-doc_begin="This function is made to explicitly cast an error-valued pointer 
> to a void
> +pointer type to make it clear that's what's going on, so the violation is 
> deliberate."
> +-config=MC3R1.R11.8,reports+={deliberate,"all_area(context(^ERR_CAST\\(.*$))"}
> +-doc_end

While the wording is okay here, ...

> --- a/docs/misra/deviations.rst
> +++ b/docs/misra/deviations.rst
> @@ -248,6 +248,19 @@ Deviations related to MISRA C:2012 Rules:
>         If no bits are set, 0 is returned.
>       - Tagged as `safe` for ECLAIR.
>  
> +   * - R11.8
> +     - Violations caused by container_of are due to pointer arithmetic 
> operations
> +       with the provided offset. The resulting pointer is then immediately 
> cast back to its
> +       original type, which preserves the qualifier. This use can be deemed 
> as safe.
> +       Fixing this violation would require to increase code complexity and 
> lower readability.
> +     - Tagged as `safe` for ECLAIR.
> +    
> +   * - R11.8
> +     - This function is made to explicitly cast an error-valued pointer to a
> +       void pointer type to make it clear that's what's going on, so the
> +       violation is deliberate.
> +     - Tagged ad `deliberate` for ECLAIR.

... "This function" here is entirely unclear. However, seeing ERR_CAST() has
no users, wouldn't we be better off simply removing it? (Otherwise, nit:
"Tagged as ...".)

Jan