Re: [PATCH v4 04/36] x86/bugs: Restructure rfds mitigation

[Josh Poimboeuf]

On Mon, Mar 10, 2025 at 11:39:51AM -0500, David Kaplan wrote:
>  static void __init rfds_select_mitigation(void)
>  {
>  	if (!boot_cpu_has_bug(X86_BUG_RFDS) || cpu_mitigations_off()) {
>  		rfds_mitigation = RFDS_MITIGATION_OFF;
>  		return;
>  	}
> -	if (rfds_mitigation == RFDS_MITIGATION_OFF)
> -		return;

The removal of this RFDS_MITIGATION_OFF check can cause
verw_mitigation_selected to get wrongly enabled below if it was
RFDS_MITIGATION_OFF to begin with.

I think it's only a bisection issue, that gets re-added later with
"x86/bugs: Add attack vector controls for rfds".

>  
>  	if (rfds_mitigation == RFDS_MITIGATION_AUTO)
>  		rfds_mitigation = RFDS_MITIGATION_VERW;
>  
> -	if (x86_arch_cap_msr & ARCH_CAP_RFDS_CLEAR)
> +	if (rfds_has_ucode())
> +		verw_mitigation_selected = true;
> +}

-- 
Josh