linux-kernel - Re: [PATCH v7 2/8] x86/mce: Unify AMD DFR handler with MCA Polling

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20251028152231.GAaQDft32eXtTZBBzg@fat_crate.local>
Date: Tue, 28 Oct 2025 16:22:31 +0100
From: Borislav Petkov <bp@...en8.de>
To: Yazen Ghannam <yazen.ghannam@....com>
Cc: x86@...nel.org, Tony Luck <tony.luck@...el.com>,
	"Rafael J. Wysocki" <rafael@...nel.org>,
	Len Brown <lenb@...nel.org>, linux-kernel@...r.kernel.org,
	linux-edac@...r.kernel.org, Smita.KoralahalliChannabasappa@....com,
	Qiuxu Zhuo <qiuxu.zhuo@...el.com>,
	Nikolay Borisov <nik.borisov@...e.com>,
	Bert Karwatzki <spasswolf@....de>, linux-acpi@...r.kernel.org
Subject: Re: [PATCH v7 2/8] x86/mce: Unify AMD DFR handler with MCA Polling

On Mon, Oct 27, 2025 at 10:11:39AM -0400, Yazen Ghannam wrote:
>  	/*
> -	 * If the MCA_STATUS register has a deferred error, then continue using it as
> -	 * the status register.
> -	 *
> -	 * MCA_DESTAT will be cleared at the end of the handler.
> +	 * If MCA_STATUS happens to have a deferred error, then MCA_DESTAT will
> +	 * be cleared at the end of the handler.
>  	 */
> -	if ((m->status & MCI_STATUS_VAL) && (m->status & MCI_STATUS_DEFERRED))
> +	if (m->status & MCI_STATUS_VAL)
>  		return true;

I'm still confused by those comments - we check VAL but we talk about
deferred...

>  
>  	/*
> -	 * If the MCA_DESTAT register has a deferred error, then use it instead.
> +	 * Use the MCA_DESTAT register if it has a deferred error.

This one...

> The redundant
> +	 * status bit check is to filter out any bogus errors.

... probably only confuses. No need to mention it.

>  	 *
>  	 * MCA_STATUS will not be cleared at the end of the handler.
>  	 */
> @@ -780,7 +777,7 @@ static bool should_log_poll_error(enum mcp_flags flags, struct mce_hw_err *err)
>  	struct mce *m = &err->m;
>  
>  	if (mce_flags.smca)
> -		return smca_should_log_poll_error(flags, err);
> +		return smca_should_log_poll_error(m);
>  
>  	/* If this entry is not valid, ignore it. */
>  	if (!(m->status & MCI_STATUS_VAL))

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette