Message-ID: <554A649C.8070605@zytor.com>
Date: Wed, 06 May 2015 11:59:40 -0700
From: "H. Peter Anvin" <hpa@zytor.com>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.5.0
MIME-Version: 1.0
To: Denys Vlasenko <dvlasenk@redhat.com>, Ingo Molnar <mingo@kernel.org>
CC: Steven Rostedt <rostedt@goodmis.org>, Borislav Petkov <bp@alien8.de>,
        Andy Lutomirski <luto@amacapital.net>,
        Frederic Weisbecker <fweisbec@gmail.com>,
        Alexei Starovoitov <ast@plumgrid.com>, Will Drewry <wad@chromium.org>,
        Kees Cook <keescook@chromium.org>, x86@kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH] x86: Deinline cpuid_eax and friends
References: <1430932057-26574-1-git-send-email-dvlasenk@redhat.com>
In-Reply-To: <1430932057-26574-1-git-send-email-dvlasenk@redhat.com>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 840
Lines: 22

On 05/06/2015 10:07 AM, Denys Vlasenko wrote:
> cpuid_e{a,b,c,d}x() functions compile to 44 bytes of machine code each.
> On x86 allyesconfig build they have 48 callsites.
> Deinlining all four of them shrinks kernel by about 1k:
> 
>    text      data      bss       dec     hex filename
> 82434909 22255384 20627456 125317749 7783275 vmlinux.before
> 82433898 22255384 20627456 125316738 7782e82 vmlinux
> 
> Speed impact: CPUID instruction takes from 50 to 350+ cycles,
> call overhead is negligible in comparison.

How on Earth does it make 44 bytes?  Is this due to paravirt_fail?

	-hpa


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/