Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755932AbYGJVK4 (ORCPT ); Thu, 10 Jul 2008 17:10:56 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751017AbYGJVKs (ORCPT ); Thu, 10 Jul 2008 17:10:48 -0400 Received: from terminus.zytor.com ([198.137.202.10]:35652 "EHLO terminus.zytor.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750885AbYGJVKr (ORCPT ); Thu, 10 Jul 2008 17:10:47 -0400 Message-ID: <4876791F.40603@zytor.com> Date: Thu, 10 Jul 2008 17:03:27 -0400 From: "H. Peter Anvin" User-Agent: Thunderbird 2.0.0.14 (X11/20080501) MIME-Version: 1.0 To: Jeremy Fitzhardinge CC: Christoph Lameter , "Eric W. Biederman" , Ingo Molnar , Mike Travis , Andrew Morton , Jack Steiner , linux-kernel@vger.kernel.org, Arjan van de Ven Subject: Re: [RFC 00/15] x86_64: Optimize percpu accesses References: <20080709165129.292635000@polaris-admin.engr.sgi.com> <20080709200757.GD14009@elte.hu> <48751B57.8030605@goop.org> <48751CF9.4020901@linux-foundation.org> <4875209D.8010603@goop.org> <48752CCD.30507@linux-foundation.org> <48753C99.5050408@goop.org> <487555A8.2050007@zytor.com> <487556A5.5090907@goop.org> <4876194E.4080205@linux-foundation.org> <48761C06.3020003@zytor.com> <48762A3B.5050104@linux-foundation.org> <48762DD2.5090802@zytor.com> <487637A1.4080403@linux-foundation.org> <487639ED.7000502@zytor.com> <48763CA6.9030802@linux-foundation.org> <487647EF.5010609@goop.org> <48764A01.1070402@linux-foundation.org> <48764C7C.5010309@goop.org> <48767692.4080504@linux-foundation.org> <487677F0.4000404@goop.org> In-Reply-To: <487677F0.4000404@goop.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 936 Lines: 23 Jeremy Fitzhardinge wrote: > > Percpu on i386 hasn't been a point of discussion. It works fine, and > has been working fine for a long time. The same mechanism would work > fine on x86-64. Its only "issue" is that it doesn't support the broken > gcc abi for stack-protector. > > The problem is all zero-based percpu on x86-64. > Well, x86-64 has *two* issues: limited range of offsets (regardless of if we do RIP-relative or not), and the stack-protector ABI. I'm still trying to reproduce Mike's setup, but I suspect it can be switched to RIP-relative for the fixed-offset (static) stuff; for the dynamic stuff it's all via pointers anyway so the offsets don't matter. -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/