Re: [XEN PATCH] xen: Add SAF deviations for MISRA C:2012 Rule 7.1

[Nicola Vetrini <nicola.vetrini@bugseng.com>]

On 10/10/2023 09:29, Luca Fancellu wrote:
>> 
>>>> /* SAF-x-safe */
>>>> if ( modrm_mod == MASK_EXTR(instr_modrm, 0300) && (modrm_reg & 7) ==
>>>> MASK_EXTR(instr_modrm, 0070) && (modrm_rm & 7)  == 
>>>> MASK_EXTR(instr_modrm,
>>>> 0007) )
>>>> 
>>>> Any other ideas?
>>> 
>>> Could we have a number in the comment to indicate the number of lines 
>>> the
>>> comment applies to?
>> 
>> Luca can confirm that what I am about to write is correct; my
>> understanding is that ECLAIR supports it, but cppcheck does not. Which
>> means for cppcheck we would have to translate the SAF tag with
>> xen_analyize to:
>> 
>> /* cppcheck tag */
>> line1
>> /* cppcheck tag */
>> line2
>> /* cppcheck tag */
>> line3
>> 
>> and that would end up changing the line numbers in the source files so
>> the cppcheck report wouldn't match with the original line numbers any
>> longer
> 
> Yes, but it’s not only Cppcheck, it’s also Coverity that supports only
> the above notation.
> 
> For Cppcheck we could do something, but for Coverity we can’t.
> 
> Anyway, Stefano or Nicola, I would like to understand where Eclair
> reports the violation
> in the case of #define, does it report at the usage or at the 
> definition?
> 
> Cheers,
> Luca

The report is at the usage site, but ECLAIR can be configured to deviate 
based on a comment
at the macro definition, or also just the macro name for reports of that 
rule (e.g. Q[1-3]):

#define M(a, b) (b)
/* -E> safe MC3R1.R7.1 1 blabla */
#define Q1(s) M(s, 0300)
/* -E> safe MC3R1.R7.1 1 blabla */
#define Q2(s) M(s, 0070)
/* -E> safe MC3R1.R7.1 1 blabla */
#define Q3(s) M(s, 0007)

void f(void) {
   int x = 1;
   int y = 2;
   if ( (x & 2) == Q1(y) &&
        (x & 3) == Q2(y) &&
        (x & 7) == Q3(y) )
   {
      y = y + 1;
   }
}

-- 
Nicola Vetrini, BSc
Software Engineer, BUGSENG srl (https://bugseng.com)