Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <Yv7aRJ/SvVhSdnSB@decadent.org.uk> <Yv9OGVc+WpoDAB0X@worktop.programming.kicks-ass.net>
 <Yv9tj9vbQ9nNlXoY@worktop.programming.kicks-ass.net> <9395338630e3313c1bf0393ae507925d1f9af870.camel@decadent.org.uk>
 <Yv9+8vR4QH6j6J/5@worktop.programming.kicks-ass.net>
In-Reply-To: <Yv9+8vR4QH6j6J/5@worktop.programming.kicks-ass.net>
Reply-To: martin-eric.racine@iki.fi
From:   =?UTF-8?Q?Martin=2D=C3=89ric_Racine?= <martin-eric.racine@iki.fi>
Date:   Tue, 30 Aug 2022 14:42:04 +0300
Message-ID: <CAPZXPQeYh_BrZzinsvCjHvd=szAsOXUmkVYS1tJC5vwamx+Wow@mail.gmail.com>
Subject: Re: [PATCH] x86/speculation: Avoid LFENCE in FILL_RETURN_BUFFER on
 CPUs that lack it
To:     Peter Zijlstra <peterz@infradead.org>
Cc:     Ben Hutchings <ben@decadent.org.uk>, x86@kernel.org,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        1017425@bugs.debian.org, stable@vger.kernel.org,
        regressions@lists.linux.dev,
        Daniel Sneddon <daniel.sneddon@linux.intel.com>,
        Pawan Gupta <pawan.kumar.gupta@linux.intel.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Greetings,

On Fri, Aug 19, 2022 at 3:15 PM Peter Zijlstra <peterz@infradead.org> wrote=
:
>
> On Fri, Aug 19, 2022 at 01:38:27PM +0200, Ben Hutchings wrote:
>
> > So that puts the whole __FILL_RETURN_BUFFER inside an alternative, and
> > we can't have nested alternatives.  That's unfortunate.
>
> Well, both alternatives end with the LFENCE instruction, so I could pull
> it out and do two consequtive ALTs, but unrolling the loop for i386 is
> a better solution in that the sequence, while larger, removes the need
> for the LFENCE.

Have we reached a definitive conclusion on to how to fix this?

Martin-=C3=89ric