Re: [PATCH v2] coverity: physmem: use simple assertions instead of modelling

[Paolo Bonzini]

On Thu, Oct 5, 2023 at 4:04 PM Vladimir Sementsov-Ogievskiy
<vsementsov@yandex-team.ru> wrote:
> +            /*
> +             * Assure Coverity (and ourselves) that we are not going to 
> OVERRUN
> +             * the buffer by following ldn_he_p().
> +             */
> +            assert((l == 1 && len >= 1) ||
> +                   (l == 2 && len >= 2) ||
> +                   (l == 4 && len >= 4) ||
> +                   (l == 8 && len >= 8));

I'll queue it shortly, but perhaps you can try if assert(l <= len) is enough?

Alternatively I can try applying the patch on top of the tree that we
test with, and see how things go.

Paolo

>              val = ldn_he_p(buf, l);
>              result |= memory_region_dispatch_write(mr, addr1, val,
>                                                     size_memop(l), attrs);
> @@ -2784,6 +2793,15 @@ MemTxResult flatview_read_continue(FlatView *fv, 
> hwaddr addr,
>              l = memory_access_size(mr, l, addr1);
>              result |= memory_region_dispatch_read(mr, addr1, &val,
>                                                    size_memop(l), attrs);
> +
> +            /*
> +             * Assure Coverity (and ourselves) that we are not going to 
> OVERRUN
> +             * the buffer by following stn_he_p().
> +             */
> +            assert((l == 1 && len >= 1) ||
> +                   (l == 2 && len >= 2) ||
> +                   (l == 4 && len >= 4) ||
> +                   (l == 8 && len >= 8));
>              stn_he_p(buf, l, val);
>          } else {
>              /* RAM case */
> --
> 2.34.1
>