Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Thu, 14 Oct 2021 00:14:51 +0200
From:   Peter Zijlstra <peterz@infradead.org>
To:     Josh Poimboeuf <jpoimboe@redhat.com>
Cc:     x86@kernel.org, andrew.cooper3@citrix.com,
        linux-kernel@vger.kernel.org, alexei.starovoitov@gmail.com,
        ndesaulniers@google.com
Subject: Re: [PATCH 4/9] x86/alternative: Implement .retpoline_sites support
Message-ID: <YWdaW0niLW4HPM0W@hirez.programming.kicks-ass.net>
References: <20211013122217.304265366@infradead.org>
 <20211013123645.002402102@infradead.org>
 <20211013211118.apsj66c6b6abubeo@treble>
 <YWdTDz8Qqs3jjCf1@hirez.programming.kicks-ass.net>
 <20211013220520.xzsp5q2amuwy6n6o@treble>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20211013220520.xzsp5q2amuwy6n6o@treble>
Precedence: bulk

On Wed, Oct 13, 2021 at 03:05:20PM -0700, Josh Poimboeuf wrote:
> On Wed, Oct 13, 2021 at 11:43:43PM +0200, Peter Zijlstra wrote:
> > On Wed, Oct 13, 2021 at 02:11:18PM -0700, Josh Poimboeuf wrote:
> > > On Wed, Oct 13, 2021 at 02:22:21PM +0200, Peter Zijlstra wrote:
> > > > +#ifdef CONFIG_X86_64
> > > > +
> > > > +/*
> > > > + * CALL/JMP *%\reg
> > > > + */
> > > > +static int emit_indirect(int op, int reg, u8 *bytes)
> > > 
> > > X86_64 is already equivalent to STACK_VALIDATION these days, but might
> > > as well clarify here where the retpoline_sites dependency comes from by
> > > changing this to '#ifdef CONFIG_STACK_VALIDATION'.
> > 
> > Yeah, I was contemplating having x86_64 unconditionally select that.
> > Maybe we should.
> 
> As far as I can tell, it already does that:
> 
>         select HAVE_STACK_VALIDATION            if X86_64
>         select HAVE_STATIC_CALL_INLINE          if HAVE_STACK_VALIDATION
>         select STACK_VALIDATION                 if HAVE_STACK_VALIDATION && (HAVE_STATIC_CALL_INLINE || RETPOLINE)

Oh right, I thought there was still a possible hole in there, but I
guess that's pretty solid. I suppose we should just remove the && ...
from the last line.