From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Thu, 6 Dec 2001 02:21:57 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Thu, 6 Dec 2001 02:21:47 -0500 Received: from zok.sgi.com ([204.94.215.101]:25480 "EHLO zok.sgi.com") by vger.kernel.org with ESMTP id ; Thu, 6 Dec 2001 02:21:38 -0500 X-Mailer: exmh version 2.2 06/23/2000 with nmh-1.0.4 From: Keith Owens To: Rusty Russell Cc: linux-kernel@vger.kernel.org, torvalds@transmeta.com Subject: Re: [PATCH] 2.5.1-pre5: per-cpu areas In-Reply-To: Your message of "Thu, 06 Dec 2001 09:09:35 +1100." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 06 Dec 2001 18:21:25 +1100 Message-ID: <12352.1007623285@kao2.melbourne.sgi.com> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Thu, 06 Dec 2001 09:09:35 +1100, Rusty Russell wrote: >The following patch implements convenient per-cpu areas: Did you look at PERCPU_ADDR in ia64? Much (all?) of the per cpu data is in struct cpuinfo_ia64 which is at the same virtual address on all cpus but with different physical addresses on each cpu. Let the mmu do the work. S390 does a similar trick, using the Prefixed Save Area (PSA) which is virtual 0 but different physical addresses on each cpu.