From patchwork Wed Mar 20 11:48:21 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Rahul Lakkireddy X-Patchwork-Id: 51390 X-Patchwork-Delegate: ferruh.yigit@amd.com Return-Path: X-Original-To: patchwork@dpdk.org Delivered-To: patchwork@dpdk.org Received: from [92.243.14.124] (localhost [127.0.0.1]) by dpdk.org (Postfix) with ESMTP id D096B4C91; Wed, 20 Mar 2019 12:52:25 +0100 (CET) Received: from stargate.chelsio.com (stargate.chelsio.com [12.32.117.8]) by dpdk.org (Postfix) with ESMTP id 369172C60; Wed, 20 Mar 2019 12:52:24 +0100 (CET) Received: from localhost (scalar.blr.asicdesigners.com [10.193.185.94]) by stargate.chelsio.com (8.13.8/8.13.8) with ESMTP id x2KBqK07031219; Wed, 20 Mar 2019 04:52:20 -0700 From: Rahul Lakkireddy To: dev@dpdk.org Cc: vishal@chelsio.com, nirranjan@chelsio.com, indranil@chelsio.com, stable@dpdk.org Date: Wed, 20 Mar 2019 17:18:21 +0530 Message-Id: <1553082501-20447-1-git-send-email-rahul.lakkireddy@chelsio.com> X-Mailer: git-send-email 2.5.3 Subject: [dpdk-dev] [PATCH] net/cxgbe: fix missing csum ol_flags and ptype X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" From: Vishal Kulkarni Checksum good offload flags are not being set and some of the packet type flags are missing on received packets. So, rework Rx path to set proper ol_flags and packet_type in mbufs. Fixes: 78fc1a716ae8 ("cxgbe: improve Rx performance") Cc: stable@dpdk.org Signed-off-by: Vishal Kulkarni Signed-off-by: Rahul Lakkireddy --- drivers/net/cxgbe/sge.c | 78 +++++++++++++++++++++++++---------------- 1 file changed, 48 insertions(+), 30 deletions(-) diff --git a/drivers/net/cxgbe/sge.c b/drivers/net/cxgbe/sge.c index 673c4fbb3..8c26ef2ef 100644 --- a/drivers/net/cxgbe/sge.c +++ b/drivers/net/cxgbe/sge.c @@ -1605,6 +1605,52 @@ static inline void rspq_next(struct sge_rspq *q) } } +static inline void cxgbe_set_mbuf_info(struct rte_mbuf *pkt, uint32_t ptype, + uint64_t ol_flags) +{ + pkt->packet_type |= ptype; + pkt->ol_flags |= ol_flags; +} + +static inline void cxgbe_fill_mbuf_info(struct adapter *adap, + const struct cpl_rx_pkt *cpl, + struct rte_mbuf *pkt) +{ + bool csum_ok; + u16 err_vec; + + if (adap->params.tp.rx_pkt_encap) + err_vec = G_T6_COMPR_RXERR_VEC(ntohs(cpl->err_vec)); + else + err_vec = ntohs(cpl->err_vec); + + csum_ok = cpl->csum_calc && !err_vec; + + if (cpl->vlan_ex) + cxgbe_set_mbuf_info(pkt, RTE_PTYPE_L2_ETHER_VLAN, + PKT_RX_VLAN | PKT_RX_VLAN_STRIPPED); + else + cxgbe_set_mbuf_info(pkt, RTE_PTYPE_L2_ETHER, 0); + + if (cpl->l2info & htonl(F_RXF_IP)) + cxgbe_set_mbuf_info(pkt, RTE_PTYPE_L3_IPV4, + csum_ok ? PKT_RX_IP_CKSUM_GOOD : + PKT_RX_IP_CKSUM_BAD); + else if (cpl->l2info & htonl(F_RXF_IP6)) + cxgbe_set_mbuf_info(pkt, RTE_PTYPE_L3_IPV6, + csum_ok ? PKT_RX_IP_CKSUM_GOOD : + PKT_RX_IP_CKSUM_BAD); + + if (cpl->l2info & htonl(F_RXF_TCP)) + cxgbe_set_mbuf_info(pkt, RTE_PTYPE_L4_TCP, + csum_ok ? PKT_RX_L4_CKSUM_GOOD : + PKT_RX_L4_CKSUM_BAD); + else if (cpl->l2info & htonl(F_RXF_UDP)) + cxgbe_set_mbuf_info(pkt, RTE_PTYPE_L4_UDP, + csum_ok ? PKT_RX_L4_CKSUM_GOOD : + PKT_RX_L4_CKSUM_BAD); +} + /** * process_responses - process responses from an SGE response queue * @q: the ingress queue to process @@ -1656,8 +1702,6 @@ static int process_responses(struct sge_rspq *q, int budget, (const void *)&q->cur_desc[1]; struct rte_mbuf *pkt, *npkt; u32 len, bufsz; - bool csum_ok; - u16 err_vec; rc = (const struct rsp_ctrl *) ((const char *)q->cur_desc + @@ -1674,16 +1718,6 @@ static int process_responses(struct sge_rspq *q, int budget, len = G_RSPD_LEN(len); pkt->pkt_len = len; - /* Compressed error vector is enabled for - * T6 only - */ - if (q->adapter->params.tp.rx_pkt_encap) - err_vec = G_T6_COMPR_RXERR_VEC( - ntohs(cpl->err_vec)); - else - err_vec = ntohs(cpl->err_vec); - csum_ok = cpl->csum_calc && !err_vec; - /* Chain mbufs into len if necessary */ while (len) { struct rte_mbuf *new_pkt = rsd->buf; @@ -1701,20 +1735,7 @@ static int process_responses(struct sge_rspq *q, int budget, npkt->next = NULL; pkt->nb_segs--; - if (cpl->l2info & htonl(F_RXF_IP)) { - pkt->packet_type = RTE_PTYPE_L3_IPV4; - if (unlikely(!csum_ok)) - pkt->ol_flags |= - PKT_RX_IP_CKSUM_BAD; - - if ((cpl->l2info & - htonl(F_RXF_UDP | F_RXF_TCP)) && - !csum_ok) - pkt->ol_flags |= - PKT_RX_L4_CKSUM_BAD; - } else if (cpl->l2info & htonl(F_RXF_IP6)) { - pkt->packet_type = RTE_PTYPE_L3_IPV6; - } + cxgbe_fill_mbuf_info(q->adapter, cpl, pkt); if (!rss_hdr->filter_tid && rss_hdr->hash_type) { @@ -1723,11 +1744,8 @@ static int process_responses(struct sge_rspq *q, int budget, ntohl(rss_hdr->hash_val); } - if (cpl->vlan_ex) { - pkt->ol_flags |= PKT_RX_VLAN | - PKT_RX_VLAN_STRIPPED; + if (cpl->vlan_ex) pkt->vlan_tci = ntohs(cpl->vlan); - } rte_pktmbuf_adj(pkt, s->pktshift); rxq->stats.pkts++;