From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755845AbXJ3XCq (ORCPT ); Tue, 30 Oct 2007 19:02:46 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752492AbXJ3XCi (ORCPT ); Tue, 30 Oct 2007 19:02:38 -0400 Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:49397 "EHLO sunset.davemloft.net" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1752096AbXJ3XCh (ORCPT ); Tue, 30 Oct 2007 19:02:37 -0400 Date: Tue, 30 Oct 2007 16:02:21 -0700 (PDT) Message-Id: <20071030.160221.27099706.davem@davemloft.net> To: rdreier@cisco.com Cc: jeff@garzik.org, shawnxjin@gmail.com, linux-kernel@vger.kernel.org Subject: Re: Multiple MSI messages support From: David Miller In-Reply-To: References: <4727AED2.6010908@garzik.org> X-Mailer: Mew version 5.2 on Emacs 22.1 / Mule 5.0 (SAKAKI) Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org From: Roland Dreier Date: Tue, 30 Oct 2007 15:51:08 -0700 > In fact at least IBM pSeries boxes seem to be able to use MSI to > generate multiple interrupts from the same device -- BenH can probably > give details. Sparc64 PCI-E controllers can do this as well. > For high-performance devices this is not really true. The InfiniBand > mthca driver can use MSI to get a single interrupt, or MSI-X to get > different interrupts for different types of events. MSI-X allows the > read of the "interrupt cause register" to be avoided, and this ends up > making a measurable performance difference for the fast path. > > Also, obviously, having multiple interrupts means you can bind > different interrupts to different CPUs, which will become very > important if/when things like multi-queue NICs are supported. Right, and this is used heavily in the NIU driver for example.