From mboxrd@z Thu Jan 1 00:00:00 1970 From: Paolo Bonzini Subject: Re: [PATCH] KVM: x86: Add lowest-priority support for vt-d posted-interrupts Date: Wed, 25 Nov 2015 15:38:01 +0100 Message-ID: <5655C7C9.1010008@redhat.com> References: <1447037208-75615-1-git-send-email-feng.wu@intel.com> <20151116190314.GA12245@potion.brq.redhat.com> <20151124143154.GB13925@potion.brq.redhat.com> <20151125141238.GC13925@potion.brq.redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: "kvm@vger.kernel.org" , "linux-kernel@vger.kernel.org" To: =?UTF-8?Q?Radim_Krcm=c3=a1r?= , "Wu, Feng" Return-path: Received: from mx1.redhat.com ([209.132.183.28]:42689 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751939AbbKYOiG (ORCPT ); Wed, 25 Nov 2015 09:38:06 -0500 In-Reply-To: <20151125141238.GC13925@potion.brq.redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On 25/11/2015 15:12, Radim Krcm=C3=A1r wrote: > I think it's ok to pick any algorithm we like. It's unlikely that > software would recognize and take advantage of the hardware algorithm > without adding a special treatment for KVM. > (I'd vote for the simple pick-first-APIC lowest priority algorithm ..= =2E > I don't see much point in complicating lowest priority when it doesn= 't > deliver to lowest priority CPU anyway.) Vector hashing is an improvement for the common case where all vectors are set to all CPUs. Sure you can get an unlucky assignment, but it's still better than pick-first-APIC. Paolo