Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20190913072237.GA12381@zn.tnic> <CAHk-=wismo3SQvvKXg8j0W-eC+5Q-ctcYfr1QV3K-i90w5caBA@mail.gmail.com>
 <9dc9f1e6-5d19-167c-793d-2f4a5ebee097@rasmusvillemoes.dk> <20190913104232.GA4190@zn.tnic>
 <20190913163645.GC4190@zn.tnic> <3fc31917-9452-3a10-d11d-056bf2d8b97d@rasmusvillemoes.dk>
 <CAHk-=wjdpJ+VapXfoZE8JRUfvMb8JrVTZe0=TDFYZ-ke+uqBOA@mail.gmail.com>
 <CALCETrX8sR8ELEvUpdHug498dU6+MWSy_SagaRbuZZ9fkztmfw@mail.gmail.com>
 <CAHk-=whZJhU3c-djPcwCPyYh0y1YXKeyBuJZjq3CzW3v_YHgeg@mail.gmail.com> <CALCETrUkBCh8h66pJCJtDGNtvhmVaNuppddsBLkQiHFoNrW-xg@mail.gmail.com>
In-Reply-To: <CALCETrUkBCh8h66pJCJtDGNtvhmVaNuppddsBLkQiHFoNrW-xg@mail.gmail.com>
From:   Linus Torvalds <torvalds@linux-foundation.org>
Date:   Mon, 16 Sep 2019 16:26:20 -0700
Message-ID: <CAHk-=wiw6q1hk7QHP+8M6H9zRDj-yJ8p-KyhQS9gTJA8NopYkA@mail.gmail.com>
Subject: Re: [RFC] Improve memset
To:     Andy Lutomirski <luto@kernel.org>
Cc:     Rasmus Villemoes <linux@rasmusvillemoes.dk>,
        Borislav Petkov <bp@alien8.de>,
        Rasmus Villemoes <mail@rasmusvillemoes.dk>,
        x86-ml <x86@kernel.org>, Josh Poimboeuf <jpoimboe@redhat.com>,
        lkml <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, Sep 16, 2019 at 4:14 PM Andy Lutomirski <luto@kernel.org> wrote:
>
> Well, when I wrote this email, I *thought* it was inlining the
> 'memset' function, but maybe I just can't read gcc's output today.

Not having your compiler, it's also possible that it works for you,
but just doesn't work for me.

> It seems like gcc is maybe smart enough to occasionally optimize
> memset just because it's called 'memset'.  This generates good code:

Yup, that does the rigth thing for me and ignores the definition of
memset() in favor of the built-in one.

However, at least part of this discussion started because of the
reverse problem (turning a couple of assignments into memset), and the
suggestion that we might be able to use -ffreestanding together with

  #define memset __builtin_memset

and then your nice code generation goes away, because the magical
treatment of memset goes away. I get

        one_word:
                xorl    %eax, %eax
                ret

        not_one_word:
                movq    %rsi, %rdx
                xorl    %esi, %esi
                jmp     memset

despite having that "inline void *memset()" definition.

             Linus