Re: [Xen-devel] [PATCH v2 1/4] x86emul: always fill x86_insn_modrm()'s outputs

[Andrew Cooper <andrew.cooper3@xxxxxxxxxx>]

On 18/04/17 11:29, Jan Beulich wrote:
> The function is rather unlikely to be called for insns which don't have
> ModRM bytes, and hence addressing Coverity's recurring complaint of
> callers potentially consuming uninitialized data when they know that
> certain opcodes have ModRM bytes can be suppressed this way without
> unduly adding overhead to fast paths.
>
> Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>

Reviewed-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>

>
> --- a/xen/arch/x86/x86_emulate/x86_emulate.c
> +++ b/xen/arch/x86/x86_emulate/x86_emulate.c
> @@ -8017,8 +8017,14 @@ x86_insn_modrm(const struct x86_emulate_
>  {
>      check_state(state);
>  
> -    if ( state->modrm_mod > 3 )
> +    if ( unlikely(state->modrm_mod > 3) )
> +    {
> +        if ( rm )
> +            *rm = ~0U;
> +        if ( reg )
> +            *reg = ~0U;
>          return -EINVAL;
> +    }
>  
>      if ( rm )
>          *rm = state->modrm_rm;
>
>
>


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel