MIME-Version: 1.0
References: <20230221184908.2349578-1-kpsingh@kernel.org> <20230221184908.2349578-2-kpsingh@kernel.org>
 <Y/d9qajJnR/ZcHvB@zn.tnic>
In-Reply-To: <Y/d9qajJnR/ZcHvB@zn.tnic>
From:   KP Singh <kpsingh@kernel.org>
Date:   Sat, 25 Feb 2023 20:42:35 -0500
Message-ID: <CACYkzJ5=Ru1MvMWKHmfuHO-+nVPSgM8n3JSeCfFTvZgSFcmOmA@mail.gmail.com>
Subject: Re: [PATCH v2 2/2] Documentation/hw-vuln: Document the interaction
 between IBRS and STIBP
To:     Borislav Petkov <bp@alien8.de>
Cc:     linux-kernel@vger.kernel.org, pjt@google.com, evn@google.com,
        jpoimboe@kernel.org, tglx@linutronix.de, x86@kernel.org,
        hpa@zytor.com, peterz@infradead.org,
        pawan.kumar.gupta@linux.intel.com, kim.phillips@amd.com,
        alexandre.chartre@oracle.com, daniel.sneddon@linux.intel.com,
        corbet@lwn.net, bp@suse.de, linyujun809@huawei.com,
        jmattson@google.com
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Thu, Feb 23, 2023 at 9:52 AM Borislav Petkov <bp@alien8.de> wrote:
>
> On Tue, Feb 21, 2023 at 07:49:08PM +0100, KP Singh wrote:
> > ... Consequently, STIBP needs to be explicitly
> > +   enabled to guard against cross-thread attacks in userspace.
>
> needs?
>
> That sounds like something the user needs to do. But we do it by
> default. Let's rephrase:
>
> "Systems which support enhanced IBRS (eIBRS) enable IBRS protections once at
> boot and they're automatically protected against Spectre v2 variant
> attacks, including cross-thread branch target injections on SMT systems
> (STIBP). IOW, eIBRS enables STIBP too.
>
> Legacy IBRS systems clear the IBRS bit on exit to userspace and
> therefore explicitly enable STIBP for that."

+   Systems which support enhanced IBRS (eIBRS) enable IBRS protection once at
+   boot, by setting the IBRS bit, and they're automatically protected against
+   Spectre v2 variant attacks, including cross-thread branch target injections
+   on SMT systems (STIBP). In other words, eIBRS enables STIBP too.
+
+   Legacy IBRS systems clear the IBRS bit on exit to userspace and
+   therefore explicitly enable STIBP for that


I did add one phrase, we really need to stress on the IBRS bit here.
Had we been enabling KERNEL_IBRS accidentally with eIBRS, it would
still mess things up as the bit being set is important.

This is why my original patch felt "obtuse" as it focused on
KERNEL_IBRS instead of IBRS or eIBRS :).


>
> Simple.
>
> --
> Regards/Gruss,
>     Boris.
>
> https://people.kernel.org/tglx/notes-about-netiquette