From mboxrd@z Thu Jan 1 00:00:00 1970 From: Maxime Coquelin Subject: Re: [PATCH v2 6/8] net/virtio: support IN_ORDER Rx and Tx Date: Wed, 27 Jun 2018 10:18:15 +0200 Message-ID: References: <20180625151710.29437-1-yong.liu@intel.com> <20180625151710.29437-7-yong.liu@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit Cc: zhihong.wang@intel.com, dev@dpdk.org To: Marvin Liu , tiwei.bie@intel.com Return-path: Received: from mx1.redhat.com (mx3-rdu2.redhat.com [66.187.233.73]) by dpdk.org (Postfix) with ESMTP id 05E881BE88 for ; Wed, 27 Jun 2018 10:18:20 +0200 (CEST) In-Reply-To: <20180625151710.29437-7-yong.liu@intel.com> Content-Language: en-US List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" On 06/25/2018 05:17 PM, Marvin Liu wrote: > IN_ORDER Rx function can support merge-able feature. Descriptors My understanding of the code is that IN_ORDER Rx function only supports mergeable feature (and it seems to be confirmed by patch 7). So maybe change the text to: "IN_ORDER Rx function requires merge-able feature." > allocation and free will be done in bulk. > > Virtio dequeue logic: > dequeue_burst_rx(burst mbufs) > for (each mbuf b) { > if (b need merge) { > merge remained mbufs > add merged mbuf to return mbufs list > } else { > add mbuf to return mbufs list > } > } > if (last mbuf c need merge) { > dequeue_burst_rx(required mbufs) > merge last mbuf c > } > refill_avail_ring_bulk() > update_avail_ring() > return mbufs list > > IN_ORDER Tx function can support offloading features. Packets can't be s/can support/supports/ > transmitted by IN_ORDER Tx will be handled by normal Tx. Can you clarify why some packets can't be transmitted by IN_ORDER Tx? > Virtio enqueue logic: > xmit_cleanup(used descs) > for (each xmit mbuf b) { > if (b can inorder xmit) { > add mbuf b to inorder burst list > continue > } else { > xmit inorder burst list > xmit mbuf b with normal xmit > } > } > if (inorder burst list not empty) { > xmit inorder burst list > } > update_avail_ring() > > Signed-off-by: Marvin Liu > Other than above clarification needed, the code looks good to me. Thanks, Maxime