linux-kernel - Re: [PATCH V2 1/1] x86/tdx: Route safe halt execution via tdx_safe

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: <CAGtprH-grdiE2AaagnYNUQC4ytSMxErYoh_Xyg2Nwmto33Yyyw@mail.gmail.com>
Date: Mon, 3 Feb 2025 14:08:41 -0800
From: Vishal Annapurve <vannapurve@...gle.com>
To: Dave Hansen <dave.hansen@...el.com>
Cc: "Kirill A. Shutemov" <kirill@...temov.name>, x86@...nel.org, linux-kernel@...r.kernel.org, 
	pbonzini@...hat.com, seanjc@...gle.com, erdemaktas@...gle.com, 
	ackerleytng@...gle.com, jxgao@...gle.com, sagis@...gle.com, oupton@...gle.com, 
	pgonda@...gle.com, dave.hansen@...ux.intel.com, linux-coco@...ts.linux.dev, 
	chao.p.peng@...ux.intel.com, isaku.yamahata@...il.com, stable@...r.kernel.org
Subject: Re: [PATCH V2 1/1] x86/tdx: Route safe halt execution via tdx_safe_halt()

On Mon, Feb 3, 2025 at 1:19 PM Dave Hansen <dave.hansen@...el.com> wrote:
>
> On 2/3/25 12:09, Kirill A. Shutemov wrote:
> ...
> > But Sean's proposal with HLT check before enabling interrupts looks better
> > to me.
>
> "Sean's proposal" being this:
>
>         https://lore.kernel.org/all/Z5l6L3Hen9_Y3SGC@google.com/
>
> ?
Yes.

>
> Is that just intended to quietly fix up a hlt-induced #VE? I'm not sure
> that's a good idea. The TDVMCALL is slow, but the #VE is also presumably
> quite slow. This is (presumably) getting called in an idle path which is
> actually one of the most performance-sensitive things we have in the kernel.
>
> Or am I missing the point of Sean's proposal?

I think you have captured the intent correctly.

>
> I don't mind having the #VE handler warn about the situation if we end
> up there accidentally.
>
> I'd much rather have a kernel configured in a way that we are pretty
> sure there's no path to even call hlt.

+1.