From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sheng Yang Subject: Re: [PATCH 5/5] KVM: assigned dev: MSI-X mask support Date: Mon, 8 Nov 2010 13:41:40 +0800 Message-ID: <201011081341.40232.sheng@linux.intel.com> References: <1288851321-3964-1-git-send-email-sheng@linux.intel.com> <201011051853.51093.sheng@linux.intel.com> <20101106002451.GA410@amt.cnet> Mime-Version: 1.0 Content-Type: Text/Plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Cc: "Michael S. Tsirkin" , Avi Kivity , kvm@vger.kernel.org To: Marcelo Tosatti Return-path: Received: from mga03.intel.com ([143.182.124.21]:15213 "EHLO mga03.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1749667Ab0KHFlU (ORCPT ); Mon, 8 Nov 2010 00:41:20 -0500 In-Reply-To: <20101106002451.GA410@amt.cnet> Sender: kvm-owner@vger.kernel.org List-ID: On Saturday 06 November 2010 08:24:51 Marcelo Tosatti wrote: > On Fri, Nov 05, 2010 at 06:53:50PM +0800, Sheng Yang wrote: > > On Friday 05 November 2010 16:43:33 Michael S. Tsirkin wrote: > > > On Fri, Nov 05, 2010 at 10:44:19AM +0800, Sheng Yang wrote: > > > > > > +}; > > > > > > + > > > > > > +This ioctl would enable in-kernel MSI-X emulation, which would > > > > > > handle MSI-X +mask bit in the kernel. > > > > > > > > > > What happens on repeated calls when it's already enabled? > > > > > How does one disable after it's enabled? > > > > > > > > Suppose this should only be called once. But again would update the > > > > MMIO base. > > > > > > So maybe add this all to documentation. > > > > > > > It > > > > would be disabled along with device deassignment. > > > > > > So what are you going to do when guest enables and later disables MSIX? > > > Disable assignment completely? > > > > This device goes with PCI resources allocation, rather than IRQ > > assignment. > > What about hot plug, for example? I can't see how this can function > without unregistering regions. The register time is when PCI resources allocation, and the unregister time is when device was deassigned. I suppose hot plug path would also notify kernel to deassign the device? -- regards Yang, Sheng