From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265501AbUGHJAs (ORCPT ); Thu, 8 Jul 2004 05:00:48 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265487AbUGHJAs (ORCPT ); Thu, 8 Jul 2004 05:00:48 -0400 Received: from hermine.aitel.hist.no ([158.38.50.15]:9234 "HELO hermine.aitel.hist.no") by vger.kernel.org with SMTP id S265422AbUGHJAr (ORCPT ); Thu, 8 Jul 2004 05:00:47 -0400 Message-ID: <40ED0DDC.20602@hist.no> Date: Thu, 08 Jul 2004 11:03:24 +0200 From: Helge Hafting User-Agent: Mozilla Thunderbird 0.6 (X11/20040605) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Roland Dreier CC: Davide Rossetti , linux-kernel@vger.kernel.org Subject: Re: MSI to memory? References: <200407011215.59723.bjorn.helgaas@hp.com> <20040701115339.A4265@unix-os.sc.intel.com> <40EBED33.3050707@roma1.infn.it> <40EBF07B.8040003@hist.no> <52vfgzoisd.fsf@topspin.com> In-Reply-To: <52vfgzoisd.fsf@topspin.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Roland Dreier wrote: > Helge> Won't that put a bad load on the bus? Someone else might > Helge> need it: * Another cpu in a smp system * Any device doing > Helge> bus-master transfers, even in a UP system > >Actually with MSI, the PCI device writes directly to a host address. >In the proposed usage in this mail thread, the address is in host >memory, so there's no bus load to poll the memory. Presumably the >memory will be pulled into cache for the duration of the poll loop, so >there's not even any memory bandwidth consumed. (Of course this only >works on an architecture where PCI DMA is cache coherent) > > I see. Still, this should only be used when we expect a short wait only, similiar to situations when we use a spinlock. Otherwise, the cpu could be put to better use by scheduling. Helge Hafting