Re: [PATCH v3] xen: address violations of MISRA C:2012 Rule 11.8.

[Simone Ballarin <simone.ballarin@xxxxxxxxxxx>]

On 04/12/23 17:54, Jan Beulich wrote:
> On 04.12.2023 17:32, Simone Ballarin wrote:
> > From: Maria Celeste Cesario <maria.celeste.cesario@xxxxxxxxxxx>
> >
> > Remove or amend casts to comply with Rule 11.8.
> >
> > Fix violations by adding missing const qualifier in cast.
> > Fix violations by removing unnecessary cast.
> > Change type of operands from char* to uintptr_t: uintptr_t is
> > the appropriate type for memory address operations.
> >
> > No functional changes.
> >
> > Signed-off-by: Maria Celeste Cesario  <maria.celeste.cesario@xxxxxxxxxxx>
> > Signed-off-by: Simone Ballarin  <simone.ballarin@xxxxxxxxxxx>
> I consider it good practice to at least briefly say what the rule is
> about, so it is clear why certain changes need doing.
>
> > --- a/xen/arch/arm/include/asm/atomic.h
> > +++ b/xen/arch/arm/include/asm/atomic.h
> > @@ -154,7 +154,7 @@ static always_inline void write_atomic_size(volatile void 
> > *p,
> >    */
> >   static inline int atomic_read(const atomic_t *v)
> >   {
> > -    return *(volatile int *)&v->counter;
> > +    return *(const volatile int *)&v->counter;
> >   }
> What about PPC's identical code?
If the ARM part is accepted, I will include this change in an upcoming 
series for Rule 11.8 (which I'm working on), otherwise in v4.
> > --- a/xen/arch/x86/include/asm/regs.h
> > +++ b/xen/arch/x86/include/asm/regs.h
> > @@ -6,7 +6,7 @@
> >   
> >   #define guest_mode(r)                                                         \
> >   ({                                                                            
> > \
> > -    unsigned long diff = (char *)guest_cpu_user_regs() - (char *)(r);         \
> > +    unsigned long diff = (uintptr_t)guest_cpu_user_regs() - (uintptr_t)(r);   \
> >       /* Frame pointer must point into current CPU stack. */                    
> > \
> >       ASSERT(diff < STACK_SIZE);                                                
> > \
> >       /* If not a guest frame, it must be a hypervisor frame. */                
> > \
> This part
> Acked-by: Jan Beulich <jbeulich@xxxxxxxx>
>
> Jan
--
Simone Ballarin, M.Sc.

Field Application Engineer, BUGSENG (https://bugseng.com)