From mboxrd@z Thu Jan 1 00:00:00 1970 From: Paolo Bonzini Subject: Re: [kvmarm] [RFC PATCH 0/3] KVM: ARM: Get rid of hardcoded VGIC addresses Date: Thu, 25 Oct 2012 18:14:34 +0200 Message-ID: <5089656A.9010704@redhat.com> References: <1350173065-35350-1-git-send-email-c.dall@virtualopensystems.com> <507F172D.2030802@suse.de> <20121017221022.GA4333@bloggs.ozlabs.ibm.com> <1350518331.4678.114.camel@pasglop> <508675FC.7090901@siemens.com> <20121024005017.GA17834@bloggs.ozlabs.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: Jan Kiszka , Christoffer Dall , Benjamin Herrenschmidt , Alexander Graf , kvmarm@lists.cs.columbia.edu, kvm@vger.kernel.org, kvm-ppc , Peter Maydell To: Paul Mackerras Return-path: Received: from mail-wi0-f172.google.com ([209.85.212.172]:38882 "EHLO mail-wi0-f172.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S935783Ab2JYQOk (ORCPT ); Thu, 25 Oct 2012 12:14:40 -0400 In-Reply-To: <20121024005017.GA17834@bloggs.ozlabs.ibm.com> Sender: kvm-owner@vger.kernel.org List-ID: Il 24/10/2012 02:50, Paul Mackerras ha scritto: > So we could indeed use the existing KVM_CREATE_IRQCHIP to tell KVM to > create a presentation controller per vcpu. But then how do we tell > KVM how many source controllers we want and how many interrupts each > source controller should handle? The source controllers are not tied > to any particular vcpu, and a source controller could potentially have > 100s of interrupts or more (particularly with MSIs). Configuration of > each source fits into 64 bits, so if we tried to use KVM_SET_IRQCHIP > for configuring a source controller we'd be limited to 64 interrupts > per source controller. As Jan said, there's more than a few similarities between the x86 and PPC model. Taking inspiration from the x86 API, configuring the source controller seems to be more of a task for KVM_SET_GSI_ROUTING (a very x86-centric name, I must admit). Paolo