net/nfp: add TSO support

This patch implements NFP PMD support for TSO but it also requires
a firmware advertising the capability.

Signed-off-by: Alejandro Lucero <alejandro.lucero@netronome.com>

diff --git a/doc/guides/nics/features/nfp.ini b/doc/guides/nics/features/nfp.ini
index 3004a67..a1281d2 100644 (file)
--- a/doc/guides/nics/features/nfp.ini
+++ b/doc/guides/nics/features/nfp.ini
@@ -12,6 +12,7 @@ Queue start/stop     = Y
 MTU update           = Y
 Jumbo frame          = Y
 Promiscuous mode     = Y
+TSO                  = Y
 RSS hash             = Y
 RSS key update       = Y
 RSS reta update      = Y

diff --git a/drivers/net/nfp/nfp_net.c b/drivers/net/nfp/nfp_net.c
index 43e9c2a..31e72e2 100644 (file)
--- a/drivers/net/nfp/nfp_net.c
+++ b/drivers/net/nfp/nfp_net.c
@@ -1152,6 +1152,9 @@ nfp_net_infos_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
        dev_info->speed_capa = ETH_SPEED_NUM_1G | ETH_LINK_SPEED_10G |
                               ETH_SPEED_NUM_25G | ETH_SPEED_NUM_40G |
                               ETH_SPEED_NUM_50G | ETH_LINK_SPEED_100G;
+
+       if (hw->cap & NFP_NET_CFG_CTRL_LSO)
+               dev_info->tx_offload_capa |= DEV_TX_OFFLOAD_TCP_TSO;
 }
 
 static const uint32_t *
@@ -1653,6 +1656,27 @@ nfp_net_tx_queue_setup(struct rte_eth_dev *dev, uint16_t queue_idx,
        return 0;
 }
 
+/* nfp_net_tx_tso - Set TX descriptor for TSO */
+static inline void
+nfp_net_tx_tso(struct nfp_net_txq *txq, struct nfp_net_tx_desc *txd,
+              struct rte_mbuf *mb)
+{
+       uint64_t ol_flags;
+       struct nfp_net_hw *hw = txq->hw;
+
+       if (!(hw->cap & NFP_NET_CFG_CTRL_LSO))
+               return;
+
+       ol_flags = mb->ol_flags;
+
+       if (!(ol_flags & PKT_TX_TCP_SEG))
+               return;
+
+       txd->l4_offset = mb->l2_len + mb->l3_len + mb->l4_len;
+       txd->lso = rte_cpu_to_le_16(mb->tso_segsz);
+       txd->flags |= PCIE_DESC_TX_LSO;
+}
+
 /* nfp_net_tx_cksum - Set TX CSUM offload flags in TX descriptor */
 static inline void
 nfp_net_tx_cksum(struct nfp_net_txq *txq, struct nfp_net_tx_desc *txd,
@@ -2021,7 +2045,7 @@ nfp_net_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
 {
        struct nfp_net_txq *txq;
        struct nfp_net_hw *hw;
-       struct nfp_net_tx_desc *txds;
+       struct nfp_net_tx_desc *txds, txd;
        struct rte_mbuf *pkt;
        uint64_t dma_addr;
        int pkt_size, dma_size;
@@ -2070,19 +2094,17 @@ nfp_net_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
 
                /*
                 * Checksum and VLAN flags just in the first descriptor for a
-                * multisegment packet
+                * multisegment packet, but TSO info needs to be in all of them.
                 */
-               nfp_net_tx_cksum(txq, txds, pkt);
+               nfp_net_tx_tso(txq, &txd, pkt);
+               nfp_net_tx_cksum(txq, &txd, pkt);
 
                if ((pkt->ol_flags & PKT_TX_VLAN_PKT) &&
                    (hw->cap & NFP_NET_CFG_CTRL_TXVLAN)) {
-                       txds->flags |= PCIE_DESC_TX_VLAN;
-                       txds->vlan = pkt->vlan_tci;
+                       txd.flags |= PCIE_DESC_TX_VLAN;
+                       txd.vlan = pkt->vlan_tci;
                }
 
-               if (pkt->ol_flags & PKT_TX_TCP_SEG)
-                       rte_panic("TSO is not supported\n");
-
                /*
                 * mbuf data_len is the data in one segment and pkt_len data
                 * in the whole packet. When the packet is just one segment,
@@ -2100,6 +2122,8 @@ nfp_net_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
                *lmbuf = pkt;
 
                while (pkt_size) {
+                       /* Copying TSO, VLAN and cksum info */
+                       *txds = txd;
                        dma_size = pkt->data_len;
                        dma_addr = rte_mbuf_data_dma_addr(pkt);
                        PMD_TX_LOG(DEBUG, "Working with mbuf at dma address:"