From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jan Kiszka Subject: Re: [PATCH 0/9] pci-assign: 64bit MMIO + better MSI-X table support Date: Tue, 31 Jan 2012 13:56:15 +0100 Message-ID: <4F27E4EF.5000204@siemens.com> References: <20120128142104.25681.93072.stgit@bling.home> <4F266CDB.4080504@siemens.com> <1327931042.6937.6.camel@bling.home> <4F27E412.9070103@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Cc: Alex Williamson , "kvm@vger.kernel.org" , "shashidhar.patil@gmail.com" To: Avi Kivity Return-path: Received: from david.siemens.de ([192.35.17.14]:23338 "EHLO david.siemens.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752243Ab2AaM4W (ORCPT ); Tue, 31 Jan 2012 07:56:22 -0500 In-Reply-To: <4F27E412.9070103@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On 2012-01-31 13:52, Avi Kivity wrote: > 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. If this solves the issues, ok. That all needs quite some refactoring anyway (including the kvm core services). Jan -- Siemens AG, Corporate Technology, CT T DE IT 1 Corporate Competence Center Embedded Linux