Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.31 as permitted sender) client-ip=23.128.96.31;
MIME-Version: 1.0
References: <20231010164234.140750-1-ubizjak@gmail.com> <CAHk-=whYWhZN52SJN-Th9x2L2V-vHtAXUgiy_nSJ3+vQU6ak4Q@mail.gmail.com>
 <CAFULd4ZqH3FeG8_mjDvUAU9QiGB36wDu3MzUtadgAgoVuQ9QRg@mail.gmail.com>
 <CAHk-=wiALZxieQQmvv5sW15HYB_YwC3d_ma9sdp7Zb4Fb4uK2w@mail.gmail.com>
 <CAFULd4Y8NSArDqH=VMy8F97eNosUUGxrBMEyHH=MytjUBSCmjg@mail.gmail.com>
 <CAHk-=whMr8V_q3dq4iS0dpx4Nssu+aYWz+mA36p2ykA+OXTjXA@mail.gmail.com>
 <CAFULd4Z-F0mFov5zhLya=OfPnQx=oOHkyLqs_UVTbQaWjk7AMA@mail.gmail.com> <CAFULd4bPPRmdwUCehKhx=cSVAK8r+6+7cEL24WdUNCPO3Y5Rbw@mail.gmail.com>
In-Reply-To: <CAFULd4bPPRmdwUCehKhx=cSVAK8r+6+7cEL24WdUNCPO3Y5Rbw@mail.gmail.com>
From:   Linus Torvalds <torvalds@linux-foundation.org>
Date:   Wed, 11 Oct 2023 12:40:09 -0700
Message-ID: <CAHk-=wiSn8fY4xF-=Bz=qPcbmT_Yzzb4jvvtMBuNQksk94Y=dw@mail.gmail.com>
Subject: Re: [PATCH v2 -tip] x86/percpu: Use C for arch_raw_cpu_ptr()
To:     Uros Bizjak <ubizjak@gmail.com>
Cc:     x86@kernel.org, linux-kernel@vger.kernel.org,
        Nadav Amit <namit@vmware.com>,
        Andy Lutomirski <luto@kernel.org>,
        Brian Gerst <brgerst@gmail.com>,
        Denys Vlasenko <dvlasenk@redhat.com>,
        "H . Peter Anvin" <hpa@zytor.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Josh Poimboeuf <jpoimboe@redhat.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Wed, 11 Oct 2023 at 00:45, Uros Bizjak <ubizjak@gmail.com> wrote:
>
> In fact, rdgsbase is available only for 64-bit targets.

Not even all 64-bit targets. That's why I did that ALTERNATIVE() thing
with X86_FEATURE_FSGSBASE, which uses the kernel instruction
re-writing.

So that suggested asm of mine defaults to loading the value from
memory through %gs, but with X86_FEATURE_FSGSBASE it gets rewritten to
use rdgsbase.

And again - I'm not sure it's any faster, but it's _potentially_
certainly better in that it doesn't use the cache-line.

           Linus