From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH v2 net-next] virtio: Fix affinity for #VCPUs != #queue pairs Date: Wed, 08 Feb 2017 13:37:44 -0500 (EST) Message-ID: <20170208.133744.558863839333233338.davem@davemloft.net> References: <20170207181506.36668-1-serebrin@google.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: borntraeger@de.ibm.com, netdev@vger.kernel.org, mst@redhat.com, jasowang@redhat.com, willemb@google.com, venkateshs@google.com, jonolson@google.com, willemdebruijn.kernel@gmail.com, rick.jones2@hpe.com, jmattson@google.com, linux-s390@vger.kernel.org, linux-arch@vger.kernel.org To: serebrin@google.com Return-path: Received: from shards.monkeyblade.net ([184.105.139.130]:56462 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751940AbdBHSp4 (ORCPT ); Wed, 8 Feb 2017 13:45:56 -0500 In-Reply-To: <20170207181506.36668-1-serebrin@google.com> Sender: netdev-owner@vger.kernel.org List-ID: From: Ben Serebrin Date: Tue, 7 Feb 2017 10:15:06 -0800 > If the number of virtio queue pairs is not equal to the > number of VCPUs, the virtio guest driver doesn't assign > any CPU affinity for the queue interrupts or the xps > aggregation interrupt. (In contrast, the driver does assign > both if the counts of VCPUs and queues are equal, which is a good > default behavior.) ... Can I get some reviews, please?