From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757838AbYG2MCz (ORCPT ); Tue, 29 Jul 2008 08:02:55 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754679AbYG2MCr (ORCPT ); Tue, 29 Jul 2008 08:02:47 -0400 Received: from one.firstfloor.org ([213.235.205.2]:36988 "EHLO one.firstfloor.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755920AbYG2MCq (ORCPT ); Tue, 29 Jul 2008 08:02:46 -0400 Date: Tue, 29 Jul 2008 14:02:37 +0200 From: Andi Kleen To: Jeremy Fitzhardinge Cc: Andi Kleen , Ingo Molnar , Jens Axboe , Linux Kernel Mailing List Subject: Re: x86: Is there still value in having a special tlb flush IPI vector? Message-ID: <20080729120237.GH30344@one.firstfloor.org> References: <488E534F.2030204@goop.org> <488E5455.6010901@goop.org> <20080729021215.GG30344@one.firstfloor.org> <488EB8BE.8050806@goop.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <488EB8BE.8050806@goop.org> User-Agent: Mutt/1.4.2.1i Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Jul 28, 2008 at 11:29:18PM -0700, Jeremy Fitzhardinge wrote: > Andi Kleen wrote: > >>>Now that normal smp_function_call is no longer an enormous bottleneck, > >>> > > > >Hmm? It still uses a global lock at least as of current git tree. > > Yes, but it's only held briefly to put things onto the list. It doesn't > get held over the whole IPI transaction as the old smp_call_function > did, and the tlb flush code still does. RCU is used to manage the list > walk and freeing, so there's no long-held locks there either. If it bounces regularly it will still hurt. -Andi