From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:49350) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1djtMz-0001Ij-42 for qemu-devel@nongnu.org; Mon, 21 Aug 2017 16:32:22 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1djtMu-0006Iz-7b for qemu-devel@nongnu.org; Mon, 21 Aug 2017 16:32:21 -0400 Received: from mx1.redhat.com ([209.132.183.28]:45664) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1djtMu-0006GV-1L for qemu-devel@nongnu.org; Mon, 21 Aug 2017 16:32:16 -0400 Date: Mon, 21 Aug 2017 23:22:17 +0300 From: "Michael S. Tsirkin" Message-ID: <20170821232052-mutt-send-email-mst@kernel.org> References: <1502940416-42944-1-git-send-email-wei.w.wang@intel.com> <1502940416-42944-4-git-send-email-wei.w.wang@intel.com> <20170818051451-mutt-send-email-mst@kernel.org> <599699AF.1090705@intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <599699AF.1090705@intel.com> Subject: Re: [Qemu-devel] [PATCH v14 3/5] virtio-balloon: VIRTIO_BALLOON_F_SG List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Wei Wang Cc: virtio-dev@lists.oasis-open.org, linux-kernel@vger.kernel.org, qemu-devel@nongnu.org, virtualization@lists.linux-foundation.org, kvm@vger.kernel.org, linux-mm@kvack.org, mhocko@kernel.org, akpm@linux-foundation.org, mawilcox@microsoft.com, david@redhat.com, cornelia.huck@de.ibm.com, mgorman@techsingularity.net, aarcange@redhat.com, amit.shah@redhat.com, pbonzini@redhat.com, willy@infradead.org, liliang.opensource@gmail.com, yang.zhang.wz@gmail.com, quan.xu@aliyun.com On Fri, Aug 18, 2017 at 03:39:27PM +0800, Wei Wang wrote: > On 08/18/2017 10:22 AM, Michael S. Tsirkin wrote: > > +static void send_balloon_page_sg(struct virtio_balloon *vb, > > + struct virtqueue *vq, > > + void *addr, > > + uint32_t size) > > +{ > > + unsigned int len; > > + int ret; > > + > > + do { > > + ret = add_one_sg(vq, addr, size); > > + virtqueue_kick(vq); > > + wait_event(vb->acked, virtqueue_get_buf(vq, &len)); > > + /* > > + * It is uncommon to see the vq is full, because the sg is sent > > + * one by one and the device is able to handle it in time. But > > + * if that happens, we go back to retry after an entry gets > > + * released. > > + */ > > Why send one by one though? Why not batch some s/gs and wait for all > > of them to be completed? If memory if fragmented, waiting every time is > > worse than what we have now (VIRTIO_BALLOON_ARRAY_PFNS_MAX at a time). > > > > OK, I'll do batching in some fashion. > > > Best, > Wei > > btw you need to address the build errors that kbot has found. -- MST