From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S268278AbUHKWZ0 (ORCPT ); Wed, 11 Aug 2004 18:25:26 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S268280AbUHKWZZ (ORCPT ); Wed, 11 Aug 2004 18:25:25 -0400 Received: from holomorphy.com ([207.189.100.168]:56454 "EHLO holomorphy.com") by vger.kernel.org with ESMTP id S268278AbUHKWYH (ORCPT ); Wed, 11 Aug 2004 18:24:07 -0400 Date: Wed, 11 Aug 2004 15:23:49 -0700 From: William Lee Irwin III To: Linus Torvalds Cc: Pavel Machek , Zwane Mwaikambo , Linux Kernel , Andrew Morton , Matt Mackall Subject: Re: [PATCH][2.6] Completely out of line spinlocks / i386 Message-ID: <20040811222349.GS11200@holomorphy.com> Mail-Followup-To: William Lee Irwin III , Linus Torvalds , Pavel Machek , Zwane Mwaikambo , Linux Kernel , Andrew Morton , Matt Mackall References: <20040811215915.GA21812@elf.ucw.cz> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.6+20040722i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 11 Aug 2004, Pavel Machek wrote: >> Fine, so perhaps we do not want config option? On Wed, Aug 11, 2004 at 03:13:15PM -0700, Linus Torvalds wrote: > The inline spinlocks are _wonderful_ for seeing where the contention is in > a simple profile. > In contrast, in a profile the out-of-lines ones will show "x% was spent on > spinlocks". Which doesn't help much when you want to see where the problem > is. > This was _hugely_ useful, at least for me, for seeing what locks were > problematic. Well, one trick with the kinda-sorta inline spinlocks is that they need additional diagnostics (which are *REALLY* painful to get out of users) to find where the overhead was, hence there were CONFIG_SPINLINE patches to get rid of the lock section bits. -- wli