linux-kernel - Re: [PATCH] x86, mem: move memmove to out of line assembler

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 23 Sep 2022 11:05:59 -0700
From:   Linus Torvalds <torvalds@...ux-foundation.org>
To:     Nick Desaulniers <ndesaulniers@...gle.com>
Cc:     Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
        Dave Hansen <dave.hansen@...ux.intel.com>, x86@...nel.org,
        "H. Peter Anvin" <hpa@...or.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Kees Cook <keescook@...omium.org>,
        linux-kernel@...r.kernel.org, llvm@...ts.linux.dev,
        Andy Lutomirski <luto@...nel.org>
Subject: Re: [PATCH] x86, mem: move memmove to out of line assembler

On Fri, Sep 23, 2022 at 10:55 AM Nick Desaulniers
<ndesaulniers@...gle.com> wrote:
>
> We could remove __HAVE_ARCH_MEMMOVE from
> arch/x86/include/asm/string_32.h for ARCH=i386 then rip this
> arch-specific definition of memmove out.
>
> Might performance regressions be a concern with that approach?

memmove() isn't particularly common, but it does happen for some paths
that can be hot - the usual case of moving parts of an array around. I
see filesystems and networking paths doing that.

The generic memmove() is a horrendous byte-at-a-time thing and only
good for bring-up of new architectures. That's not an option.

But I'm looking at that x86-64 memcpy_orig, and I think it looks
fairly good as a template for doing the same on x86-32. And we could
get rid of the duplication on the x86-64 side.

That said, your patch looks fine too, as a "minimal changes" thing.

                Linus