Re: [PATCH] Return ENXIO instead of EPERM when speculation control is unimplemented

[Thomas Gleixner]

Anthony,

Anthony Steinhauser <asteinhauser@xxxxxxxxxx> writes:
>  		return "";
>  
>  	switch (spectre_v2_user) {
> -	case SPECTRE_V2_USER_NONE:
> +	case SPECTRE_V2_USER_UNAVAILABLE:
> +		return ", STIBP: unavailable";

Shouldn't this for correctness differentiate between the case where the
STIBP mitigation feature is not available and the case where STIBP is
not used because SMT is not possible?

Thanks,

        tglx