Re: [XEN PATCH v2 04/13] xen/arm: fix violations of MISRA C:2012 Rule 7.2

[Julien Grall <julien@xxxxxxx>]

Hi,

On 05/07/2023 17:27, Luca Fancellu wrote:
> > diff --git a/xen/arch/arm/include/asm/arm64/efibind.h 
> > b/xen/arch/arm/include/asm/arm64/efibind.h
> > index 8b43bb8495..0e890b666d 100644
> > --- a/xen/arch/arm/include/asm/arm64/efibind.h
> > +++ b/xen/arch/arm/include/asm/arm64/efibind.h
> > @@ -22,12 +22,12 @@ Revision History
> > #pragma pack()
> > #endif
> >
> > -#define EFIERR(a)           (0x8000000000000000 | a)
> > -#define EFI_ERROR_MASK      0x8000000000000000
> > -#define EFIERR_OEM(a)       (0xc000000000000000 | a)
> > +#define EFIERR(a)           (0x8000000000000000U | a)
> > +#define EFI_ERROR_MASK      0x8000000000000000U
> > +#define EFIERR_OEM(a)       (0xc000000000000000U | a)
> >
> > -#define BAD_POINTER         0xFBFBFBFBFBFBFBFB
> > -#define MAX_ADDRESS         0xFFFFFFFFFFFFFFFF
> > +#define BAD_POINTER         0xFBFBFBFBFBFBFBFBU
> > +#define MAX_ADDRESS         0xFFFFFFFFFFFFFFFFU
> Should these be UL? (From EFIERR to MAX_ADDRESS)
Even though this arm64 only, it would be best to use ULL.

[...]

> > diff --git a/xen/arch/arm/vgic-v3.c b/xen/arch/arm/vgic-v3.c
> > index 9c1fdcc3f8..f3e9eb2f92 100644
> > --- a/xen/arch/arm/vgic-v3.c
> > +++ b/xen/arch/arm/vgic-v3.c
> > @@ -354,7 +354,7 @@ read_reserved:
> >      return 1;
> >
> > read_unknown:
> > -    *r = vreg_reg64_extract(0xdeadbeafdeadbeaf, info);
> > +    *r = vreg_reg64_extract(0xdeadbeafdeadbeafU, info);
> Should this be UL?
The code is common between arm32 and arm64. So this wants to be ULL (UL 
is 32-bit on arm32).
Cheers,

--
Julien Grall