X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=drivers%2Fnet%2Fi40e%2Fi40e_rxtx.c;h=ad06b71e6db52fc40d727d52427cd52d172e6ce0;hb=6c293ffd63b3f61a95d8c492d007160874e7a057;hp=e42ba5dde7f1314ea42fcd5e45298bc43f266013;hpb=3d053e3538995fb03fd497bb5f2afa3f3210f42e;p=dpdk.git diff --git a/drivers/net/i40e/i40e_rxtx.c b/drivers/net/i40e/i40e_rxtx.c index e42ba5dde7..ad06b71e6d 100644 --- a/drivers/net/i40e/i40e_rxtx.c +++ b/drivers/net/i40e/i40e_rxtx.c @@ -61,7 +61,6 @@ #define DEFAULT_TX_RS_THRESH 32 #define DEFAULT_TX_FREE_THRESH 32 -#define I40E_MAX_PKT_TYPE 256 #define I40E_TX_MAX_BURST 32 @@ -75,6 +74,12 @@ #define I40E_TXD_CMD (I40E_TX_DESC_CMD_EOP | I40E_TX_DESC_CMD_RS) +#ifdef RTE_LIBRTE_IEEE1588 +#define I40E_TX_IEEE1588_TMST PKT_TX_IEEE1588_TMST +#else +#define I40E_TX_IEEE1588_TMST 0 +#endif + #define I40E_TX_CKSUM_OFFLOAD_MASK ( \ PKT_TX_IP_CKSUM | \ PKT_TX_L4_MASK | \ @@ -87,7 +92,9 @@ PKT_TX_OUTER_IP_CKSUM | \ PKT_TX_TCP_SEG | \ PKT_TX_QINQ_PKT | \ - PKT_TX_VLAN_PKT) + PKT_TX_VLAN_PKT | \ + PKT_TX_TUNNEL_MASK | \ + I40E_TX_IEEE1588_TMST) #define I40E_TX_OFFLOAD_NOTSUP_MASK \ (PKT_TX_OFFLOAD_MASK ^ I40E_TX_OFFLOAD_MASK) @@ -101,7 +108,7 @@ i40e_rxd_to_vlan_tci(struct rte_mbuf *mb, volatile union i40e_rx_desc *rxdp) { if (rte_le_to_cpu_64(rxdp->wb.qword1.status_error_len) & (1 << I40E_RX_DESC_STATUS_L2TAG1P_SHIFT)) { - mb->ol_flags |= PKT_RX_VLAN_PKT | PKT_RX_VLAN_STRIPPED; + mb->ol_flags |= PKT_RX_VLAN | PKT_RX_VLAN_STRIPPED; mb->vlan_tci = rte_le_to_cpu_16(rxdp->wb.qword0.lo_dword.l2tag1); PMD_RX_LOG(DEBUG, "Descriptor l2tag1: %u", @@ -151,8 +158,11 @@ i40e_rxd_error_to_pkt_flags(uint64_t qword) uint64_t error_bits = (qword >> I40E_RXD_QW1_ERROR_SHIFT); #define I40E_RX_ERR_BITS 0x3f - if (likely((error_bits & I40E_RX_ERR_BITS) == 0)) + if (likely((error_bits & I40E_RX_ERR_BITS) == 0)) { + flags |= (PKT_RX_IP_CKSUM_GOOD | PKT_RX_L4_CKSUM_GOOD); return flags; + } + if (unlikely(error_bits & (1 << I40E_RX_DESC_ERROR_IPE_SHIFT))) flags |= PKT_RX_IP_CKSUM_BAD; else @@ -266,7 +276,7 @@ i40e_parse_tunneling_params(uint64_t ol_flags, *cd_tunneling |= I40E_TXD_CTX_GRE_TUNNELING; break; default: - PMD_TX_LOG(ERR, "Tunnel type not supported\n"); + PMD_TX_LOG(ERR, "Tunnel type not supported"); return; } @@ -447,6 +457,7 @@ i40e_rx_scan_hw_ring(struct i40e_rx_queue *rxq) int32_t s[I40E_LOOK_AHEAD], nb_dd; int32_t i, j, nb_rx = 0; uint64_t pkt_flags; + uint32_t *ptype_tbl = rxq->vsi->adapter->ptype_tbl; rxdp = &rxq->rx_ring[rxq->rx_tail]; rxep = &rxq->sw_ring[rxq->rx_tail]; @@ -495,9 +506,9 @@ i40e_rx_scan_hw_ring(struct i40e_rx_queue *rxq) pkt_flags = i40e_rxd_status_to_pkt_flags(qword1); pkt_flags |= i40e_rxd_error_to_pkt_flags(qword1); mb->packet_type = - i40e_rxd_pkt_type_mapping((uint8_t)((qword1 & - I40E_RXD_QW1_PTYPE_MASK) >> - I40E_RXD_QW1_PTYPE_SHIFT)); + ptype_tbl[(uint8_t)((qword1 & + I40E_RXD_QW1_PTYPE_MASK) >> + I40E_RXD_QW1_PTYPE_SHIFT)]; if (pkt_flags & PKT_RX_RSS_HASH) mb->hash.rss = rte_le_to_cpu_32(\ rxdp[j].wb.qword0.hi_dword.rss); @@ -578,14 +589,14 @@ i40e_rx_alloc_bufs(struct i40e_rx_queue *rxq) mb->nb_segs = 1; mb->port = rxq->port_id; dma_addr = rte_cpu_to_le_64(\ - rte_mbuf_data_dma_addr_default(mb)); + rte_mbuf_data_iova_default(mb)); rxdp[i].read.hdr_addr = 0; rxdp[i].read.pkt_addr = dma_addr; } /* Update rx tail regsiter */ rte_wmb(); - I40E_PCI_REG_WRITE(rxq->qrx_tail, rxq->rx_free_trigger); + I40E_PCI_REG_WRITE_RELAXED(rxq->qrx_tail, rxq->rx_free_trigger); rxq->rx_free_trigger = (uint16_t)(rxq->rx_free_trigger + rxq->rx_free_thresh); @@ -599,6 +610,7 @@ static inline uint16_t rx_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) { struct i40e_rx_queue *rxq = (struct i40e_rx_queue *)rx_queue; + struct rte_eth_dev *dev; uint16_t nb_rx = 0; if (!nb_pkts) @@ -616,9 +628,10 @@ rx_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) if (i40e_rx_alloc_bufs(rxq) != 0) { uint16_t i, j; - PMD_RX_LOG(DEBUG, "Rx mbuf alloc failed for " - "port_id=%u, queue_id=%u", - rxq->port_id, rxq->queue_id); + dev = I40E_VSI_TO_ETH_DEV(rxq->vsi); + dev->data->rx_mbuf_alloc_failed += + rxq->rx_free_thresh; + rxq->rx_nb_avail = 0; rxq->rx_tail = (uint16_t)(rxq->rx_tail - nb_rx); for (i = 0, j = rxq->rx_tail; i < nb_rx; i++, j++) @@ -680,6 +693,7 @@ i40e_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) union i40e_rx_desc rxd; struct i40e_rx_entry *sw_ring; struct i40e_rx_entry *rxe; + struct rte_eth_dev *dev; struct rte_mbuf *rxm; struct rte_mbuf *nmb; uint16_t nb_rx; @@ -689,6 +703,7 @@ i40e_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) uint16_t rx_id, nb_hold; uint64_t dma_addr; uint64_t pkt_flags; + uint32_t *ptype_tbl; nb_rx = 0; nb_hold = 0; @@ -696,6 +711,7 @@ i40e_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) rx_id = rxq->rx_tail; rx_ring = rxq->rx_ring; sw_ring = rxq->sw_ring; + ptype_tbl = rxq->vsi->adapter->ptype_tbl; while (nb_rx < nb_pkts) { rxdp = &rx_ring[rx_id]; @@ -708,10 +724,13 @@ i40e_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) break; nmb = rte_mbuf_raw_alloc(rxq->mp); - if (unlikely(!nmb)) + if (unlikely(!nmb)) { + dev = I40E_VSI_TO_ETH_DEV(rxq->vsi); + dev->data->rx_mbuf_alloc_failed++; break; - rxd = *rxdp; + } + rxd = *rxdp; nb_hold++; rxe = &sw_ring[rx_id]; rx_id++; @@ -733,7 +752,7 @@ i40e_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) rxm = rxe->mbuf; rxe->mbuf = nmb; dma_addr = - rte_cpu_to_le_64(rte_mbuf_data_dma_addr_default(nmb)); + rte_cpu_to_le_64(rte_mbuf_data_iova_default(nmb)); rxdp->read.hdr_addr = 0; rxdp->read.pkt_addr = dma_addr; @@ -752,8 +771,8 @@ i40e_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) pkt_flags = i40e_rxd_status_to_pkt_flags(qword1); pkt_flags |= i40e_rxd_error_to_pkt_flags(qword1); rxm->packet_type = - i40e_rxd_pkt_type_mapping((uint8_t)((qword1 & - I40E_RXD_QW1_PTYPE_MASK) >> I40E_RXD_QW1_PTYPE_SHIFT)); + ptype_tbl[(uint8_t)((qword1 & + I40E_RXD_QW1_PTYPE_MASK) >> I40E_RXD_QW1_PTYPE_SHIFT)]; if (pkt_flags & PKT_RX_RSS_HASH) rxm->hash.rss = rte_le_to_cpu_32(rxd.wb.qword0.hi_dword.rss); @@ -803,10 +822,12 @@ i40e_recv_scattered_pkts(void *rx_queue, struct rte_mbuf *nmb, *rxm; uint16_t rx_id = rxq->rx_tail; uint16_t nb_rx = 0, nb_hold = 0, rx_packet_len; + struct rte_eth_dev *dev; uint32_t rx_status; uint64_t qword1; uint64_t dma_addr; uint64_t pkt_flags; + uint32_t *ptype_tbl = rxq->vsi->adapter->ptype_tbl; while (nb_rx < nb_pkts) { rxdp = &rx_ring[rx_id]; @@ -819,8 +840,12 @@ i40e_recv_scattered_pkts(void *rx_queue, break; nmb = rte_mbuf_raw_alloc(rxq->mp); - if (unlikely(!nmb)) + if (unlikely(!nmb)) { + dev = I40E_VSI_TO_ETH_DEV(rxq->vsi); + dev->data->rx_mbuf_alloc_failed++; break; + } + rxd = *rxdp; nb_hold++; rxe = &sw_ring[rx_id]; @@ -844,7 +869,7 @@ i40e_recv_scattered_pkts(void *rx_queue, rxm = rxe->mbuf; rxe->mbuf = nmb; dma_addr = - rte_cpu_to_le_64(rte_mbuf_data_dma_addr_default(nmb)); + rte_cpu_to_le_64(rte_mbuf_data_iova_default(nmb)); /* Set data buffer address and data length of the mbuf */ rxdp->read.hdr_addr = 0; @@ -914,8 +939,8 @@ i40e_recv_scattered_pkts(void *rx_queue, pkt_flags = i40e_rxd_status_to_pkt_flags(qword1); pkt_flags |= i40e_rxd_error_to_pkt_flags(qword1); first_seg->packet_type = - i40e_rxd_pkt_type_mapping((uint8_t)((qword1 & - I40E_RXD_QW1_PTYPE_MASK) >> I40E_RXD_QW1_PTYPE_SHIFT)); + ptype_tbl[(uint8_t)((qword1 & + I40E_RXD_QW1_PTYPE_MASK) >> I40E_RXD_QW1_PTYPE_SHIFT)]; if (pkt_flags & PKT_RX_RSS_HASH) first_seg->hash.rss = rte_le_to_cpu_32(rxd.wb.qword0.hi_dword.rss); @@ -1023,7 +1048,6 @@ i40e_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts) uint16_t nb_tx; uint32_t td_cmd; uint32_t td_offset; - uint32_t tx_flags; uint32_t td_tag; uint64_t ol_flags; uint16_t nb_used; @@ -1047,7 +1071,6 @@ i40e_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts) td_cmd = 0; td_tag = 0; td_offset = 0; - tx_flags = 0; tx_pkt = *tx_pkts++; RTE_MBUF_PREFETCH_TO_FREE(txe->mbuf); @@ -1094,12 +1117,8 @@ i40e_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts) /* Descriptor based VLAN insertion */ if (ol_flags & (PKT_TX_VLAN_PKT | PKT_TX_QINQ_PKT)) { - tx_flags |= tx_pkt->vlan_tci << - I40E_TX_FLAG_L2TAG1_SHIFT; - tx_flags |= I40E_TX_FLAG_INSERT_VLAN; td_cmd |= I40E_TX_DESC_CMD_IL2TAG1; - td_tag = (tx_flags & I40E_TX_FLAG_L2TAG1_MASK) >> - I40E_TX_FLAG_L2TAG1_SHIFT; + td_tag = tx_pkt->vlan_tci; } /* Always enable CRC offload insertion */ @@ -1183,7 +1202,7 @@ i40e_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts) /* Setup TX Descriptor */ slen = m_seg->data_len; - buf_dma_addr = rte_mbuf_data_dma_addr(m_seg); + buf_dma_addr = rte_mbuf_data_iova(m_seg); PMD_TX_LOG(DEBUG, "mbuf: %p, TDD[%u]:\n" "buf_dma_addr: %#"PRIx64";\n" @@ -1232,13 +1251,13 @@ end_of_tx: (unsigned) txq->port_id, (unsigned) txq->queue_id, (unsigned) tx_id, (unsigned) nb_tx); - I40E_PCI_REG_WRITE(txq->qtx_tail, tx_id); + I40E_PCI_REG_WRITE_RELAXED(txq->qtx_tail, tx_id); txq->tx_tail = tx_id; return nb_tx; } -static inline int __attribute__((always_inline)) +static __rte_always_inline int i40e_tx_free_bufs(struct i40e_tx_queue *txq) { struct i40e_tx_entry *txep; @@ -1282,7 +1301,7 @@ tx4(volatile struct i40e_tx_desc *txdp, struct rte_mbuf **pkts) uint32_t i; for (i = 0; i < 4; i++, txdp++, pkts++) { - dma_addr = rte_mbuf_data_dma_addr(*pkts); + dma_addr = rte_mbuf_data_iova(*pkts); txdp->buffer_addr = rte_cpu_to_le_64(dma_addr); txdp->cmd_type_offset_bsz = i40e_build_ctob((uint32_t)I40E_TD_CMD, 0, @@ -1296,7 +1315,7 @@ tx1(volatile struct i40e_tx_desc *txdp, struct rte_mbuf **pkts) { uint64_t dma_addr; - dma_addr = rte_mbuf_data_dma_addr(*pkts); + dma_addr = rte_mbuf_data_iova(*pkts); txdp->buffer_addr = rte_cpu_to_le_64(dma_addr); txdp->cmd_type_offset_bsz = i40e_build_ctob((uint32_t)I40E_TD_CMD, 0, @@ -1384,7 +1403,7 @@ tx_xmit_pkts(struct i40e_tx_queue *txq, /* Update the tx tail register */ rte_wmb(); - I40E_PCI_REG_WRITE(txq->qtx_tail, txq->tx_tail); + I40E_PCI_REG_WRITE_RELAXED(txq->qtx_tail, txq->tx_tail); return nb_pkts; } @@ -1415,6 +1434,28 @@ i40e_xmit_pkts_simple(void *tx_queue, return nb_tx; } +static uint16_t +i40e_xmit_pkts_vec(void *tx_queue, struct rte_mbuf **tx_pkts, + uint16_t nb_pkts) +{ + uint16_t nb_tx = 0; + struct i40e_tx_queue *txq = (struct i40e_tx_queue *)tx_queue; + + while (nb_pkts) { + uint16_t ret, num; + + num = (uint16_t)RTE_MIN(nb_pkts, txq->tx_rs_thresh); + ret = i40e_xmit_fixed_burst_vec(tx_queue, &tx_pkts[nb_tx], + num); + nb_tx += ret; + nb_pkts -= ret; + if (ret < num) + break; + } + + return nb_tx; +} + /********************************************************************* * * TX prep functions @@ -1432,13 +1473,10 @@ i40e_prep_pkts(__rte_unused void *tx_queue, struct rte_mbuf **tx_pkts, m = tx_pkts[i]; ol_flags = m->ol_flags; - /** - * m->nb_segs is uint8_t, so nb_segs is always less than - * I40E_TX_MAX_SEG. - * We check only a condition for nb_segs > I40E_TX_MAX_MTU_SEG. - */ + /* Check for m->nb_segs to not exceed the limits. */ if (!(ol_flags & PKT_TX_TCP_SEG)) { - if (m->nb_segs > I40E_TX_MAX_MTU_SEG) { + if (m->nb_segs > I40E_TX_MAX_SEG || + m->nb_segs > I40E_TX_MAX_MTU_SEG) { rte_errno = -EINVAL; return i; } @@ -1567,7 +1605,7 @@ i40e_dev_rx_queue_stop(struct rte_eth_dev *dev, uint16_t rx_queue_id) rxq = dev->data->rx_queues[rx_queue_id]; /* - * rx_queue_id is queue id aplication refers to, while + * rx_queue_id is queue id application refers to, while * rxq->reg_idx is the real queue index. */ err = i40e_switch_rx_queue(hw, rxq->reg_idx, FALSE); @@ -1598,7 +1636,7 @@ i40e_dev_tx_queue_start(struct rte_eth_dev *dev, uint16_t tx_queue_id) txq = dev->data->tx_queues[tx_queue_id]; /* - * tx_queue_id is queue id aplication refers to, while + * tx_queue_id is queue id application refers to, while * rxq->reg_idx is the real queue index. */ err = i40e_switch_tx_queue(hw, txq->reg_idx, TRUE); @@ -1623,7 +1661,7 @@ i40e_dev_tx_queue_stop(struct rte_eth_dev *dev, uint16_t tx_queue_id) txq = dev->data->tx_queues[tx_queue_id]; /* - * tx_queue_id is queue id aplication refers to, while + * tx_queue_id is queue id application refers to, while * txq->reg_idx is the real queue index. */ err = i40e_switch_tx_queue(hw, txq->reg_idx, FALSE); @@ -1693,36 +1731,42 @@ i40e_dev_rx_queue_setup(struct rte_eth_dev *dev, const struct rte_eth_rxconf *rx_conf, struct rte_mempool *mp) { - struct i40e_vsi *vsi; struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); - struct i40e_pf *pf = I40E_DEV_PRIVATE_TO_PF(dev->data->dev_private); struct i40e_adapter *ad = I40E_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private); + struct i40e_vsi *vsi; + struct i40e_pf *pf = NULL; + struct i40e_vf *vf = NULL; struct i40e_rx_queue *rxq; const struct rte_memzone *rz; uint32_t ring_size; uint16_t len, i; - uint16_t base, bsf, tc_mapping; - int use_def_burst_func = 1; + uint16_t reg_idx, base, bsf, tc_mapping; + int q_offset, use_def_burst_func = 1; if (hw->mac.type == I40E_MAC_VF || hw->mac.type == I40E_MAC_X722_VF) { - struct i40e_vf *vf = - I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); + vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); vsi = &vf->vsi; - } else + if (!vsi) + return -EINVAL; + reg_idx = queue_idx; + } else { + pf = I40E_DEV_PRIVATE_TO_PF(dev->data->dev_private); vsi = i40e_pf_get_vsi_by_qindex(pf, queue_idx); - - if (vsi == NULL) { - PMD_DRV_LOG(ERR, "VSI not available or queue " - "index exceeds the maximum"); - return I40E_ERR_PARAM; + if (!vsi) + return -EINVAL; + q_offset = i40e_get_queue_offset_by_qindex(pf, queue_idx); + if (q_offset < 0) + return -EINVAL; + reg_idx = vsi->base_queue + q_offset; } + if (nb_desc % I40E_ALIGN_RING_DESC != 0 || - (nb_desc > I40E_MAX_RING_DESC) || - (nb_desc < I40E_MIN_RING_DESC)) { + (nb_desc > I40E_MAX_RING_DESC) || + (nb_desc < I40E_MIN_RING_DESC)) { PMD_DRV_LOG(ERR, "Number (%u) of receive descriptors is " "invalid", nb_desc); - return I40E_ERR_PARAM; + return -EINVAL; } /* Free memory if needed */ @@ -1745,12 +1789,7 @@ i40e_dev_rx_queue_setup(struct rte_eth_dev *dev, rxq->nb_rx_desc = nb_desc; rxq->rx_free_thresh = rx_conf->rx_free_thresh; rxq->queue_id = queue_idx; - if (hw->mac.type == I40E_MAC_VF || hw->mac.type == I40E_MAC_X722_VF) - rxq->reg_idx = queue_idx; - else /* PF device */ - rxq->reg_idx = vsi->base_queue + - i40e_get_queue_offset_by_qindex(pf, queue_idx); - + rxq->reg_idx = reg_idx; rxq->port_id = dev->data->port_id; rxq->crc_len = (uint8_t) ((dev->data->dev_conf.rxmode.hw_strip_crc) ? 0 : ETHER_CRC_LEN); @@ -1761,13 +1800,11 @@ i40e_dev_rx_queue_setup(struct rte_eth_dev *dev, /* Allocate the maximun number of RX ring hardware descriptor. */ len = I40E_MAX_RING_DESC; -#ifdef RTE_LIBRTE_I40E_RX_ALLOW_BULK_ALLOC /** * Allocating a little more memory because vectorized/bulk_alloc Rx * functions doesn't check boundaries each time. */ len += RTE_PMD_I40E_RX_MAX_BURST; -#endif ring_size = RTE_ALIGN(len * sizeof(union i40e_rx_desc), I40E_DMA_MEM_ALIGN); @@ -1783,14 +1820,10 @@ i40e_dev_rx_queue_setup(struct rte_eth_dev *dev, /* Zero all the descriptors in the ring. */ memset(rz->addr, 0, ring_size); - rxq->rx_ring_phys_addr = rte_mem_phy2mch(rz->memseg_id, rz->phys_addr); + rxq->rx_ring_phys_addr = rz->iova; rxq->rx_ring = (union i40e_rx_desc *)rz->addr; -#ifdef RTE_LIBRTE_I40E_RX_ALLOW_BULK_ALLOC len = (uint16_t)(nb_desc + RTE_PMD_I40E_RX_MAX_BURST); -#else - len = nb_desc; -#endif /* Allocate the software ring. */ rxq->sw_ring = @@ -1865,11 +1898,6 @@ i40e_dev_rx_queue_count(struct rte_eth_dev *dev, uint16_t rx_queue_id) struct i40e_rx_queue *rxq; uint16_t desc = 0; - if (unlikely(rx_queue_id >= dev->data->nb_rx_queues)) { - PMD_DRV_LOG(ERR, "Invalid RX queue id %u", rx_queue_id); - return 0; - } - rxq = dev->data->rx_queues[rx_queue_id]; rxdp = &(rxq->rx_ring[rxq->rx_tail]); while ((desc < rxq->nb_rx_desc) && @@ -1900,7 +1928,7 @@ i40e_dev_rx_descriptor_done(void *rx_queue, uint16_t offset) int ret; if (unlikely(offset >= rxq->nb_rx_desc)) { - PMD_DRV_LOG(ERR, "Invalid RX queue id %u", offset); + PMD_DRV_LOG(ERR, "Invalid RX descriptor id %u", offset); return 0; } @@ -1917,6 +1945,64 @@ i40e_dev_rx_descriptor_done(void *rx_queue, uint16_t offset) return ret; } +int +i40e_dev_rx_descriptor_status(void *rx_queue, uint16_t offset) +{ + struct i40e_rx_queue *rxq = rx_queue; + volatile uint64_t *status; + uint64_t mask; + uint32_t desc; + + if (unlikely(offset >= rxq->nb_rx_desc)) + return -EINVAL; + + if (offset >= rxq->nb_rx_desc - rxq->nb_rx_hold) + return RTE_ETH_RX_DESC_UNAVAIL; + + desc = rxq->rx_tail + offset; + if (desc >= rxq->nb_rx_desc) + desc -= rxq->nb_rx_desc; + + status = &rxq->rx_ring[desc].wb.qword1.status_error_len; + mask = rte_le_to_cpu_64((1ULL << I40E_RX_DESC_STATUS_DD_SHIFT) + << I40E_RXD_QW1_STATUS_SHIFT); + if (*status & mask) + return RTE_ETH_RX_DESC_DONE; + + return RTE_ETH_RX_DESC_AVAIL; +} + +int +i40e_dev_tx_descriptor_status(void *tx_queue, uint16_t offset) +{ + struct i40e_tx_queue *txq = tx_queue; + volatile uint64_t *status; + uint64_t mask, expect; + uint32_t desc; + + if (unlikely(offset >= txq->nb_tx_desc)) + return -EINVAL; + + desc = txq->tx_tail + offset; + /* go to next desc that has the RS bit */ + desc = ((desc + txq->tx_rs_thresh - 1) / txq->tx_rs_thresh) * + txq->tx_rs_thresh; + if (desc >= txq->nb_tx_desc) { + desc -= txq->nb_tx_desc; + if (desc >= txq->nb_tx_desc) + desc -= txq->nb_tx_desc; + } + + status = &txq->tx_ring[desc].cmd_type_offset_bsz; + mask = rte_le_to_cpu_64(I40E_TXD_QW1_DTYPE_MASK); + expect = rte_cpu_to_le_64( + I40E_TX_DESC_DTYPE_DESC_DONE << I40E_TXD_QW1_DTYPE_SHIFT); + if ((*status & mask) == expect) + return RTE_ETH_TX_DESC_DONE; + + return RTE_ETH_TX_DESC_FULL; +} + int i40e_dev_tx_queue_setup(struct rte_eth_dev *dev, uint16_t queue_idx, @@ -1924,34 +2010,40 @@ i40e_dev_tx_queue_setup(struct rte_eth_dev *dev, unsigned int socket_id, const struct rte_eth_txconf *tx_conf) { - struct i40e_vsi *vsi; struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); - struct i40e_pf *pf = I40E_DEV_PRIVATE_TO_PF(dev->data->dev_private); + struct i40e_vsi *vsi; + struct i40e_pf *pf = NULL; + struct i40e_vf *vf = NULL; struct i40e_tx_queue *txq; const struct rte_memzone *tz; uint32_t ring_size; uint16_t tx_rs_thresh, tx_free_thresh; - uint16_t i, base, bsf, tc_mapping; + uint16_t reg_idx, i, base, bsf, tc_mapping; + int q_offset; if (hw->mac.type == I40E_MAC_VF || hw->mac.type == I40E_MAC_X722_VF) { - struct i40e_vf *vf = - I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); + vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); vsi = &vf->vsi; - } else + if (!vsi) + return -EINVAL; + reg_idx = queue_idx; + } else { + pf = I40E_DEV_PRIVATE_TO_PF(dev->data->dev_private); vsi = i40e_pf_get_vsi_by_qindex(pf, queue_idx); - - if (vsi == NULL) { - PMD_DRV_LOG(ERR, "VSI is NULL, or queue index (%u) " - "exceeds the maximum", queue_idx); - return I40E_ERR_PARAM; + if (!vsi) + return -EINVAL; + q_offset = i40e_get_queue_offset_by_qindex(pf, queue_idx); + if (q_offset < 0) + return -EINVAL; + reg_idx = vsi->base_queue + q_offset; } if (nb_desc % I40E_ALIGN_RING_DESC != 0 || - (nb_desc > I40E_MAX_RING_DESC) || - (nb_desc < I40E_MIN_RING_DESC)) { + (nb_desc > I40E_MAX_RING_DESC) || + (nb_desc < I40E_MIN_RING_DESC)) { PMD_DRV_LOG(ERR, "Number (%u) of transmit descriptors is " "invalid", nb_desc); - return I40E_ERR_PARAM; + return -EINVAL; } /** @@ -2060,18 +2152,13 @@ i40e_dev_tx_queue_setup(struct rte_eth_dev *dev, txq->hthresh = tx_conf->tx_thresh.hthresh; txq->wthresh = tx_conf->tx_thresh.wthresh; txq->queue_id = queue_idx; - if (hw->mac.type == I40E_MAC_VF || hw->mac.type == I40E_MAC_X722_VF) - txq->reg_idx = queue_idx; - else /* PF device */ - txq->reg_idx = vsi->base_queue + - i40e_get_queue_offset_by_qindex(pf, queue_idx); - + txq->reg_idx = reg_idx; txq->port_id = dev->data->port_id; txq->txq_flags = tx_conf->txq_flags; txq->vsi = vsi; txq->tx_deferred_start = tx_conf->tx_deferred_start; - txq->tx_ring_phys_addr = rte_mem_phy2mch(tz->memseg_id, tz->phys_addr); + txq->tx_ring_phys_addr = tz->iova; txq->tx_ring = (struct i40e_tx_desc *)tz->addr; /* Allocate software ring */ @@ -2133,12 +2220,8 @@ i40e_memzone_reserve(const char *name, uint32_t len, int socket_id) if (mz) return mz; - if (rte_xen_dom0_supported()) - mz = rte_memzone_reserve_bounded(name, len, - socket_id, 0, I40E_RING_BASE_ALIGN, RTE_PGSIZE_2M); - else - mz = rte_memzone_reserve_aligned(name, len, - socket_id, 0, I40E_RING_BASE_ALIGN); + mz = rte_memzone_reserve_aligned(name, len, + socket_id, 0, I40E_RING_BASE_ALIGN); return mz; } @@ -2198,11 +2281,11 @@ i40e_reset_rx_queue(struct i40e_rx_queue *rxq) for (i = 0; i < len * sizeof(union i40e_rx_desc); i++) ((volatile char *)rxq->rx_ring)[i] = 0; -#ifdef RTE_LIBRTE_I40E_RX_ALLOW_BULK_ALLOC memset(&rxq->fake_mbuf, 0x0, sizeof(rxq->fake_mbuf)); for (i = 0; i < RTE_PMD_I40E_RX_MAX_BURST; ++i) rxq->sw_ring[rxq->nb_rx_desc + i].mbuf = &rxq->fake_mbuf; +#ifdef RTE_LIBRTE_I40E_RX_ALLOW_BULK_ALLOC rxq->rx_nb_avail = 0; rxq->rx_next_avail = 0; rxq->rx_free_trigger = (uint16_t)(rxq->rx_free_thresh - 1); @@ -2219,18 +2302,40 @@ i40e_reset_rx_queue(struct i40e_rx_queue *rxq) void i40e_tx_queue_release_mbufs(struct i40e_tx_queue *txq) { + struct rte_eth_dev *dev; uint16_t i; + dev = &rte_eth_devices[txq->port_id]; + if (!txq || !txq->sw_ring) { PMD_DRV_LOG(DEBUG, "Pointer to rxq or sw_ring is NULL"); return; } - for (i = 0; i < txq->nb_tx_desc; i++) { - if (txq->sw_ring[i].mbuf) { + /** + * vPMD tx will not set sw_ring's mbuf to NULL after free, + * so need to free remains more carefully. + */ + if (dev->tx_pkt_burst == i40e_xmit_pkts_vec) { + i = txq->tx_next_dd - txq->tx_rs_thresh + 1; + if (txq->tx_tail < i) { + for (; i < txq->nb_tx_desc; i++) { + rte_pktmbuf_free_seg(txq->sw_ring[i].mbuf); + txq->sw_ring[i].mbuf = NULL; + } + i = 0; + } + for (; i < txq->tx_tail; i++) { rte_pktmbuf_free_seg(txq->sw_ring[i].mbuf); txq->sw_ring[i].mbuf = NULL; } + } else { + for (i = 0; i < txq->nb_tx_desc; i++) { + if (txq->sw_ring[i].mbuf) { + rte_pktmbuf_free_seg(txq->sw_ring[i].mbuf); + txq->sw_ring[i].mbuf = NULL; + } + } } } @@ -2343,7 +2448,7 @@ i40e_alloc_rx_queue_mbufs(struct i40e_rx_queue *rxq) mbuf->port = rxq->port_id; dma_addr = - rte_cpu_to_le_64(rte_mbuf_data_dma_addr_default(mbuf)); + rte_cpu_to_le_64(rte_mbuf_data_iova_default(mbuf)); rxd = &rxq->rx_ring[i]; rxd->read.pkt_addr = dma_addr; @@ -2386,7 +2491,7 @@ i40e_rx_queue_config(struct i40e_rx_queue *rxq) case I40E_FLAG_HEADER_SPLIT_DISABLED: default: rxq->rx_hdr_len = 0; - rxq->rx_buf_len = RTE_ALIGN(buf_size, + rxq->rx_buf_len = RTE_ALIGN_FLOOR(buf_size, (1 << I40E_RXQ_CTX_DBUFF_SHIFT)); rxq->hs_mode = i40e_header_split_none; break; @@ -2587,7 +2692,7 @@ i40e_fdir_setup_tx_resources(struct i40e_pf *pf) txq->reg_idx = pf->fdir.fdir_vsi->base_queue; txq->vsi = pf->fdir.fdir_vsi; - txq->tx_ring_phys_addr = rte_mem_phy2mch(tz->memseg_id, tz->phys_addr); + txq->tx_ring_phys_addr = tz->iova; txq->tx_ring = (struct i40e_tx_desc *)tz->addr; /* * don't need to allocate software ring and reset for the fdir @@ -2643,7 +2748,7 @@ i40e_fdir_setup_rx_resources(struct i40e_pf *pf) rxq->reg_idx = pf->fdir.fdir_vsi->base_queue; rxq->vsi = pf->fdir.fdir_vsi; - rxq->rx_ring_phys_addr = rte_mem_phy2mch(rz->memseg_id, rz->phys_addr); + rxq->rx_ring_phys_addr = rz->iova; rxq->rx_ring = (union i40e_rx_desc *)rz->addr; /* @@ -2842,6 +2947,75 @@ i40e_set_tx_function(struct rte_eth_dev *dev) } } +void __attribute__((cold)) +i40e_set_default_ptype_table(struct rte_eth_dev *dev) +{ + struct i40e_adapter *ad = + I40E_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private); + int i; + + for (i = 0; i < I40E_MAX_PKT_TYPE; i++) + ad->ptype_tbl[i] = i40e_get_default_pkt_type(i); +} + +void __attribute__((cold)) +i40e_set_default_pctype_table(struct rte_eth_dev *dev) +{ + struct i40e_adapter *ad = + I40E_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private); + struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); + int i; + + for (i = 0; i < I40E_FLOW_TYPE_MAX; i++) + ad->pctypes_tbl[i] = 0ULL; + ad->flow_types_mask = 0ULL; + ad->pctypes_mask = 0ULL; + + ad->pctypes_tbl[RTE_ETH_FLOW_FRAG_IPV4] = + (1ULL << I40E_FILTER_PCTYPE_FRAG_IPV4); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV4_UDP] = + (1ULL << I40E_FILTER_PCTYPE_NONF_IPV4_UDP); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV4_TCP] = + (1ULL << I40E_FILTER_PCTYPE_NONF_IPV4_TCP); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV4_SCTP] = + (1ULL << I40E_FILTER_PCTYPE_NONF_IPV4_SCTP); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV4_OTHER] = + (1ULL << I40E_FILTER_PCTYPE_NONF_IPV4_OTHER); + ad->pctypes_tbl[RTE_ETH_FLOW_FRAG_IPV6] = + (1ULL << I40E_FILTER_PCTYPE_FRAG_IPV6); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV6_UDP] = + (1ULL << I40E_FILTER_PCTYPE_NONF_IPV6_UDP); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV6_TCP] = + (1ULL << I40E_FILTER_PCTYPE_NONF_IPV6_TCP); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV6_SCTP] = + (1ULL << I40E_FILTER_PCTYPE_NONF_IPV6_SCTP); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV6_OTHER] = + (1ULL << I40E_FILTER_PCTYPE_NONF_IPV6_OTHER); + ad->pctypes_tbl[RTE_ETH_FLOW_L2_PAYLOAD] = + (1ULL << I40E_FILTER_PCTYPE_L2_PAYLOAD); + + if (hw->mac.type == I40E_MAC_X722) { + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV4_UDP] |= + (1ULL << I40E_FILTER_PCTYPE_NONF_UNICAST_IPV4_UDP); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV4_UDP] |= + (1ULL << I40E_FILTER_PCTYPE_NONF_MULTICAST_IPV4_UDP); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV4_TCP] |= + (1ULL << I40E_FILTER_PCTYPE_NONF_IPV4_TCP_SYN_NO_ACK); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV6_UDP] |= + (1ULL << I40E_FILTER_PCTYPE_NONF_UNICAST_IPV6_UDP); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV6_UDP] |= + (1ULL << I40E_FILTER_PCTYPE_NONF_MULTICAST_IPV6_UDP); + ad->pctypes_tbl[RTE_ETH_FLOW_NONFRAG_IPV6_TCP] |= + (1ULL << I40E_FILTER_PCTYPE_NONF_IPV6_TCP_SYN_NO_ACK); + } + + for (i = 0; i < I40E_FLOW_TYPE_MAX; i++) { + if (ad->pctypes_tbl[i]) + ad->flow_types_mask |= (1ULL << i); + ad->pctypes_mask |= ad->pctypes_tbl[i]; + } +} + /* Stubs needed for linkage when CONFIG_RTE_I40E_INC_VECTOR is set to 'n' */ int __attribute__((weak)) i40e_rx_vec_dev_conf_condition_check(struct rte_eth_dev __rte_unused *dev) @@ -2886,9 +3060,9 @@ i40e_rx_queue_release_mbufs_vec(struct i40e_rx_queue __rte_unused*rxq) } uint16_t __attribute__((weak)) -i40e_xmit_pkts_vec(void __rte_unused *tx_queue, - struct rte_mbuf __rte_unused **tx_pkts, - uint16_t __rte_unused nb_pkts) +i40e_xmit_fixed_burst_vec(void __rte_unused * tx_queue, + struct rte_mbuf __rte_unused **tx_pkts, + uint16_t __rte_unused nb_pkts) { return 0; }