From patchwork Wed May 23 16:47:16 2018 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Lam, Tiago" X-Patchwork-Id: 919193 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (mailfrom) smtp.mailfrom=openvswitch.org (client-ip=140.211.169.12; helo=mail.linuxfoundation.org; envelope-from=ovs-dev-bounces@openvswitch.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=intel.com Received: from mail.linuxfoundation.org (mail.linuxfoundation.org [140.211.169.12]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 40rdk740Jcz9s0q for ; Thu, 24 May 2018 02:48:15 +1000 (AEST) Received: from mail.linux-foundation.org (localhost [127.0.0.1]) by mail.linuxfoundation.org (Postfix) with ESMTP id DF24BD83; Wed, 23 May 2018 16:47:48 +0000 (UTC) X-Original-To: ovs-dev@openvswitch.org Delivered-To: ovs-dev@mail.linuxfoundation.org Received: from smtp1.linuxfoundation.org (smtp1.linux-foundation.org [172.17.192.35]) by mail.linuxfoundation.org (Postfix) with ESMTPS id 0BF79D7A for ; Wed, 23 May 2018 16:47:47 +0000 (UTC) X-Greylist: domain auto-whitelisted by SQLgrey-1.7.6 Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by smtp1.linuxfoundation.org (Postfix) with ESMTPS id 50CD7284 for ; Wed, 23 May 2018 16:47:46 +0000 (UTC) X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga004.jf.intel.com ([10.7.209.38]) by orsmga102.jf.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 23 May 2018 09:47:46 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.49,433,1520924400"; d="scan'208";a="201845950" Received: from silpixa00399125.ir.intel.com ([10.237.223.34]) by orsmga004.jf.intel.com with ESMTP; 23 May 2018 09:47:45 -0700 From: Tiago Lam To: ovs-dev@openvswitch.org Date: Wed, 23 May 2018 17:47:16 +0100 Message-Id: <1527094048-105084-2-git-send-email-tiago.lam@intel.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1527094048-105084-1-git-send-email-tiago.lam@intel.com> References: <1527094048-105084-1-git-send-email-tiago.lam@intel.com> X-Spam-Status: No, score=-6.9 required=5.0 tests=BAYES_00,RCVD_IN_DNSWL_HI autolearn=ham version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on smtp1.linux-foundation.org Subject: [ovs-dev] [RFC v7 01/13] netdev-dpdk: fix mbuf sizing X-BeenThere: ovs-dev@openvswitch.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: ovs-dev-bounces@openvswitch.org Errors-To: ovs-dev-bounces@openvswitch.org From: Mark Kavanagh There are numerous factors that must be considered when calculating the size of an mbuf: - the data portion of the mbuf must be sized in accordance With Rx buffer alignment (typically 1024B). So, for example, in order to successfully receive and capture a 1500B packet, mbufs with a data portion of size 2048B must be used. - in OvS, the elements that comprise an mbuf are: * the dp packet, which includes a struct rte mbuf (704B) * RTE_PKTMBUF_HEADROOM (128B) * packet data (aligned to 1k, as previously described) * RTE_PKTMBUF_TAILROOM (typically 0) Some PMDs require that the total mbuf size (i.e. the total sum of all of the above-listed components' lengths) is cache-aligned. To satisfy this requirement, it may be necessary to round up the total mbuf size with respect to cacheline size. In doing so, it's possible that the dp_packet's data portion is inadvertently increased in size, such that it no longer adheres to Rx buffer alignment. Consequently, the following property of the mbuf no longer holds true: mbuf.data_len == mbuf.buf_len - mbuf.data_off This creates a problem in the case of multi-segment mbufs, where that assumption is assumed to be true for all but the final segment in an mbuf chain. Resolve this issue by adjusting the size of the mbuf's private data portion, as opposed to the packet data portion when aligning mbuf size to cachelines. Fixes: 4be4d22 ("netdev-dpdk: clean up mbuf initialization") Fixes: 31b88c9 ("netdev-dpdk: round up mbuf_size to cache_line_size") CC: Santosh Shukla Signed-off-by: Mark Kavanagh --- lib/netdev-dpdk.c | 46 ++++++++++++++++++++++++++++++---------------- 1 file changed, 30 insertions(+), 16 deletions(-) diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c index 87152a7..356a967 100644 --- a/lib/netdev-dpdk.c +++ b/lib/netdev-dpdk.c @@ -82,12 +82,6 @@ static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(5, 20); + (2 * VLAN_HEADER_LEN)) #define MTU_TO_FRAME_LEN(mtu) ((mtu) + ETHER_HDR_LEN + ETHER_CRC_LEN) #define MTU_TO_MAX_FRAME_LEN(mtu) ((mtu) + ETHER_HDR_MAX_LEN) -#define FRAME_LEN_TO_MTU(frame_len) ((frame_len) \ - - ETHER_HDR_LEN - ETHER_CRC_LEN) -#define MBUF_SIZE(mtu) ROUND_UP((MTU_TO_MAX_FRAME_LEN(mtu) \ - + sizeof(struct dp_packet) \ - + RTE_PKTMBUF_HEADROOM), \ - RTE_CACHE_LINE_SIZE) #define NETDEV_DPDK_MBUF_ALIGN 1024 #define NETDEV_DPDK_MAX_PKT_LEN 9728 @@ -491,7 +485,7 @@ is_dpdk_class(const struct netdev_class *class) * behaviour, which reduces performance. To prevent this, use a buffer size * that is closest to 'mtu', but which satisfies the aforementioned criteria. */ -static uint32_t +static uint16_t dpdk_buf_size(int mtu) { return ROUND_UP((MTU_TO_MAX_FRAME_LEN(mtu) + RTE_PKTMBUF_HEADROOM), @@ -582,7 +576,7 @@ dpdk_mp_do_not_free(struct rte_mempool *mp) OVS_REQUIRES(dpdk_mp_mutex) * - a new mempool was just created; * - a matching mempool already exists. */ static struct rte_mempool * -dpdk_mp_create(struct netdev_dpdk *dev, int mtu) +dpdk_mp_create(struct netdev_dpdk *dev, uint16_t mbuf_pkt_data_len) { char mp_name[RTE_MEMPOOL_NAMESIZE]; const char *netdev_name = netdev_get_name(&dev->up); @@ -590,6 +584,7 @@ dpdk_mp_create(struct netdev_dpdk *dev, int mtu) uint32_t n_mbufs; uint32_t hash = hash_string(netdev_name, 0); struct rte_mempool *mp = NULL; + uint16_t mbuf_size, aligned_mbuf_size, mbuf_priv_data_len; /* * XXX: rough estimation of number of mbufs required for this port: @@ -609,12 +604,13 @@ dpdk_mp_create(struct netdev_dpdk *dev, int mtu) * longer than RTE_MEMPOOL_NAMESIZE. */ int ret = snprintf(mp_name, RTE_MEMPOOL_NAMESIZE, "ovs%08x%02d%05d%07u", - hash, socket_id, mtu, n_mbufs); + hash, socket_id, mbuf_pkt_data_len, n_mbufs); if (ret < 0 || ret >= RTE_MEMPOOL_NAMESIZE) { VLOG_DBG("snprintf returned %d. " "Failed to generate a mempool name for \"%s\". " - "Hash:0x%x, socket_id: %d, mtu:%d, mbufs:%u.", - ret, netdev_name, hash, socket_id, mtu, n_mbufs); + "Hash:0x%x, socket_id: %d, pkt data room:%d, mbufs:%u.", + ret, netdev_name, hash, socket_id, mbuf_pkt_data_len, + n_mbufs); break; } @@ -623,13 +619,31 @@ dpdk_mp_create(struct netdev_dpdk *dev, int mtu) netdev_name, n_mbufs, socket_id, dev->requested_n_rxq, dev->requested_n_txq); + mbuf_priv_data_len = sizeof(struct dp_packet) - + sizeof(struct rte_mbuf); + /* The size of the entire mbuf. */ + mbuf_size = sizeof (struct dp_packet) + + mbuf_pkt_data_len + RTE_PKTMBUF_HEADROOM; + /* mbuf size, rounded up to cacheline size. */ + aligned_mbuf_size = ROUND_UP(mbuf_size, RTE_CACHE_LINE_SIZE); + /* If there is a size discrepancy, add padding to mbuf_priv_data_len. + * This maintains mbuf size cache alignment, while also honoring RX + * buffer alignment in the data portion of the mbuf. If this adjustment + * is not made, there is a possiblity later on that for an element of + * the mempool, buf, buf->data_len < (buf->buf_len - buf->data_off). + * This is problematic in the case of multi-segment mbufs, particularly + * when an mbuf segment needs to be resized (when [push|popp]ing a VLAN + * header, for example. + */ + mbuf_priv_data_len += (aligned_mbuf_size - mbuf_size); + mp = rte_pktmbuf_pool_create(mp_name, n_mbufs, MP_CACHE_SZ, - sizeof (struct dp_packet) - sizeof (struct rte_mbuf), - MBUF_SIZE(mtu) - sizeof(struct dp_packet), socket_id); + mbuf_priv_data_len, + mbuf_pkt_data_len + RTE_PKTMBUF_HEADROOM, socket_id); if (mp) { VLOG_DBG("Allocated \"%s\" mempool with %u mbufs", - mp_name, n_mbufs); + mp_name, n_mbufs); /* rte_pktmbuf_pool_create has done some initialization of the * rte_mbuf part of each dp_packet. Some OvS specific fields * of the packet still need to be initialized by @@ -690,13 +704,13 @@ static int netdev_dpdk_mempool_configure(struct netdev_dpdk *dev) OVS_REQUIRES(dev->mutex) { - uint32_t buf_size = dpdk_buf_size(dev->requested_mtu); + uint16_t buf_size = dpdk_buf_size(dev->requested_mtu); struct rte_mempool *mp; int ret = 0; dpdk_mp_sweep(); - mp = dpdk_mp_create(dev, FRAME_LEN_TO_MTU(buf_size)); + mp = dpdk_mp_create(dev, buf_size); if (!mp) { VLOG_ERR("Failed to create memory pool for netdev " "%s, with MTU %d on socket %d: %s\n",