From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:38449) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1f9DXU-0003lT-7F for qemu-devel@nongnu.org; Thu, 19 Apr 2018 13:40:11 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1f9DXR-0005QD-1J for qemu-devel@nongnu.org; Thu, 19 Apr 2018 13:40:08 -0400 Received: from mx3-rdu2.redhat.com ([66.187.233.73]:55624 helo=mx1.redhat.com) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1f9DXQ-0005Ps-S1 for qemu-devel@nongnu.org; Thu, 19 Apr 2018 13:40:04 -0400 Date: Thu, 19 Apr 2018 20:39:58 +0300 From: "Michael S. Tsirkin" Message-ID: <20180419203852-mutt-send-email-mst@kernel.org> References: <20180412151232.17506-7-tiwei.bie@intel.com> <20180418192154-mutt-send-email-mst@kernel.org> <20180419111439.i6gfhnept6wy7uzp@debian> <20180419182035-mutt-send-email-mst@kernel.org> <28ab1675-15ae-dab4-fd71-7632666f57fc@redhat.com> <20180419200049-mutt-send-email-mst@kernel.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Subject: Re: [Qemu-devel] [virtio-dev] RE: [PATCH v3 6/6] vhost-user: support registering external host notifiers List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Paolo Bonzini Cc: "Liang, Cunming" , "Bie, Tiwei" , "jasowang@redhat.com" , "alex.williamson@redhat.com" , "stefanha@redhat.com" , "qemu-devel@nongnu.org" , "virtio-dev@lists.oasis-open.org" , "Daly, Dan" , "Tan, Jianfeng" , "Wang, Zhihong" , "Wang, Xiao W" On Thu, Apr 19, 2018 at 07:35:57PM +0200, Paolo Bonzini wrote: > On 19/04/2018 19:27, Michael S. Tsirkin wrote: > > > > That CONFIG_SMP here is clearly wrong but I don't really know what > > to set it to. Also, we probably should switch virtio_wmb to dma_XX > > barriers. > > > > That's actually easy. Will try to do. > > Should it be dma_wmb() before updating the indices, and wmb() before > writing the "kick virtqueue" register? > > Paolo if anything kick virtqueue should do wmb internally within virtio pci. No one uses strong barriers with virtio pci right now so we don't do it. -- MST