From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: [PATCH 0/9] pci-assign: 64bit MMIO + better MSI-X table support Date: Tue, 31 Jan 2012 14:52:34 +0200 Message-ID: <4F27E412.9070103@redhat.com> References: <20120128142104.25681.93072.stgit@bling.home> <4F266CDB.4080504@siemens.com> <1327931042.6937.6.camel@bling.home> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Cc: Jan Kiszka , kvm@vger.kernel.org, shashidhar.patil@gmail.com To: Alex Williamson Return-path: Received: from mx1.redhat.com ([209.132.183.28]:58963 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752243Ab2AaMwj (ORCPT ); Tue, 31 Jan 2012 07:52:39 -0500 In-Reply-To: <1327931042.6937.6.camel@bling.home> Sender: kvm-owner@vger.kernel.org List-ID: On 01/30/2012 03:44 PM, Alex Williamson wrote: > On Mon, 2012-01-30 at 11:11 +0100, Jan Kiszka wrote: > > On 2012-01-28 15:21, Alex Williamson wrote: > > > Patch 1 & 2 here are independent of the rest, but I include them > > > here to avoid conflicts. The first patch enables exposing MMIO > > > BARs as their native width to the guest. I added a config option > > > for this with the default to use the existing behavior as I > > > suspect we may have some latent issues there. Patch 2 is just > > > some trivial debug build warning fixes. > > > > > > The rest of the patches work on improving MSI-X table support. > > > Particularly, vectors can now be updated by the guest after > > > MSI-X is enabled to support things like irqbalance for SMP > > > affinity tuning. We also now update MSI-X configuration as > > > new vectors are unmasked, which enables assignment of MSI-X > > > devices on FreeBSD. I was able to assign and use an 82576 > > > (PF & VF) on a FreeBSD 9.0 guest with this series. Hopefully > > > Shashidhar can report whether this improves the behavior he > > > as seeing with an 82599. > > > > > > I wasn't able to get masking to work reliably, so I left that > > > as is for now. Perhaps someone has suggestions on getting that > > > to work. Thanks, > > > > Unless it's urging, let's focus on getting this implemented via the > > MSI/MSI-X core, not widely duplicated in device-assignment. > > I disagree. This isn't making the code duplication worse and it solves > at least two use cases that are currently broken. This won't make it > any more difficult to eventually move to msix.c, if it does, the core > needs more work. Thanks, > I agree (with Alex), but maybe I missed something? Patch 9 does call kvm directly instead of going through msi services, but I don't think this should hold the patches. -- error compiling committee.c: too many arguments to function