From mboxrd@z Thu Jan 1 00:00:00 1970 From: Paolo Bonzini Subject: Re: [RFC PATCH 3/4] KVM: x86: Add EOI exit bitmap inference Date: Thu, 14 May 2015 11:20:08 +0200 Message-ID: <555468C8.8050303@redhat.com> References: <1431481652-27268-1-git-send-email-srutherford@google.com> <1431481652-27268-3-git-send-email-srutherford@google.com> <5553028F.1010204@redhat.com> <20150513222406.GA24121@google.com> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Cc: kvm@vger.kernel.org, ahonig@google.com, Alex Williamson To: Steve Rutherford Return-path: Received: from mx1.redhat.com ([209.132.183.28]:60113 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752660AbbENJUO (ORCPT ); Thu, 14 May 2015 05:20:14 -0400 In-Reply-To: <20150513222406.GA24121@google.com> Sender: kvm-owner@vger.kernel.org List-ID: On 14/05/2015 00:24, Steve Rutherford wrote: > Seems reasonable. > > While we're on the topic of scanning the IOAPIC, should this also > scan the IOAPIC when (un)registering irq ack notifiers? [Which is > currently done for the in-kernel IOAPIC.] Would irq_ack_notifiers be used at all with this patch set? Resampling of IOAPIC level-triggered interrupts would be implemented in userspace. For the same reason, assigned devices using legacy device assignment probably would not be able to use INTX (so this feature should depend on !KVM_DEVICE_ASSIGNMENT). Add the emulated i8254 and the last user of irq_ack_notifiers goes away. Alex, how does VFIO do INTX resampling if you're using TCG or -machine kernel_irqchip=off? (Context: this series keeps the local APIC emulation in the kernel, thus including MSI, but moves the IOAPIC emualtion to userspace). Paolo