Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S934254AbZDBBzt (ORCPT ); Wed, 1 Apr 2009 21:55:49 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1756272AbZDBBzk (ORCPT ); Wed, 1 Apr 2009 21:55:40 -0400 Received: from hera.kernel.org ([140.211.167.34]:41331 "EHLO hera.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755764AbZDBBzj (ORCPT ); Wed, 1 Apr 2009 21:55:39 -0400 Message-ID: <49D41ABC.7070603@kernel.org> Date: Thu, 02 Apr 2009 10:54:04 +0900 From: Tejun Heo User-Agent: Thunderbird 2.0.0.19 (X11/20081227) MIME-Version: 1.0 To: Martin Schwidefsky CC: David Miller , mingo@elte.hu, rusty@rustcorp.com.au, tglx@linutronix.de, x86@kernel.org, linux-kernel@vger.kernel.org, hpa@zytor.com, lethal@linux-sh.org, rmk@arm.linux.org.uk, starvik@axis.com, ralf@linux-mips.org, cooloney@kernel.org, kyle@mcmartin.ca, matthew@wil.cx, grundler@parisc-linux.org, takata@linux-m32r.org, benh@kernel.crashing.org, rth@twiddle.net, ink@jurassic.park.msu.ru, heiko.carstens@de.ibm.com Subject: Re: [GIT RFC] percpu: use dynamic percpu allocator as the default percpu allocator References: <20090401101054.0a4b187d@skybase> <49D3231D.2040403@kernel.org> <20090401103257.12c2517e@skybase> <20090401.013702.66171229.davem@davemloft.net> <20090401104737.22c8c66c@skybase> <49D32AE5.70709@kernel.org> <20090401110846.0536df11@skybase> In-Reply-To: <20090401110846.0536df11@skybase> X-Enigmail-Version: 0.95.7 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.0 (hera.kernel.org [127.0.0.1]); Thu, 02 Apr 2009 01:53:26 +0000 (UTC) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1048 Lines: 30 Hello, Martin Schwidefsky wrote: >> Also, there is no guarantee that the offset from dynamic allocator >> will fall in the same 4G. There is reserve mechanism for static ones >> for archs which need it but for dynamic ones, the offset can be any >> value. > > If we map the modules with a 4GB distance to the main kernel image then > we don't have to worry about the offsets anymore. The compiler could > just use LARL to get the address of the static per-cpu variables and > SHIFT_PERCPU_PTR could be a RELOC_HIDE. Ah... okay. Great. No more GOTENT trick for static ones and for dynamic ones the compiler will generate indirection. > It just that the remapping we need to do is sooo icky .. Fingers and toes crossed. It would improve general percpu access performance too. Thanks. -- tejun -- 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/