X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=drivers%2Fnet%2Fice%2Fice_rxtx.c;h=3f6e7359844b8214627957520fcf6711836b5820;hb=05b405d581486651305551a9f7295f40388d95db;hp=93a0ac6918e50314c035cae8f584c51abe815b99;hpb=7a340b0b4e03ba5d308af5698d35f59969f75583;p=dpdk.git diff --git a/drivers/net/ice/ice_rxtx.c b/drivers/net/ice/ice_rxtx.c index 93a0ac6918..3f6e735984 100644 --- a/drivers/net/ice/ice_rxtx.c +++ b/drivers/net/ice/ice_rxtx.c @@ -2,11 +2,13 @@ * Copyright(c) 2018 Intel Corporation */ -#include +#include #include +#include #include "rte_pmd_ice.h" #include "ice_rxtx.h" +#include "ice_rxtx_vec_common.h" #define ICE_TX_CKSUM_OFFLOAD_MASK ( \ PKT_TX_IP_CKSUM | \ @@ -25,6 +27,32 @@ uint64_t rte_net_ice_dynflag_proto_xtr_ipv6_flow_mask; uint64_t rte_net_ice_dynflag_proto_xtr_tcp_mask; uint64_t rte_net_ice_dynflag_proto_xtr_ip_offset_mask; +int +ice_get_monitor_addr(void *rx_queue, struct rte_power_monitor_cond *pmc) +{ + volatile union ice_rx_flex_desc *rxdp; + struct ice_rx_queue *rxq = rx_queue; + uint16_t desc; + + desc = rxq->rx_tail; + rxdp = &rxq->rx_ring[desc]; + /* watch for changes in status bit */ + pmc->addr = &rxdp->wb.status_error0; + + /* + * we expect the DD bit to be set to 1 if this descriptor was already + * written to. + */ + pmc->val = rte_cpu_to_le_16(1 << ICE_RX_FLEX_DESC_STATUS0_DD_S); + pmc->mask = rte_cpu_to_le_16(1 << ICE_RX_FLEX_DESC_STATUS0_DD_S); + + /* register is 16-bit */ + pmc->size = sizeof(uint16_t); + + return 0; +} + + static inline uint8_t ice_proto_xtr_type_to_rxdid(uint8_t xtr_type) { @@ -42,6 +70,28 @@ ice_proto_xtr_type_to_rxdid(uint8_t xtr_type) rxdid_map[xtr_type] : ICE_RXDID_COMMS_OVS; } +static inline void +ice_rxd_to_pkt_fields_by_comms_generic(__rte_unused struct ice_rx_queue *rxq, + struct rte_mbuf *mb, + volatile union ice_rx_flex_desc *rxdp) +{ + volatile struct ice_32b_rx_flex_desc_comms *desc = + (volatile struct ice_32b_rx_flex_desc_comms *)rxdp; + uint16_t stat_err = rte_le_to_cpu_16(desc->status_error0); + + if (likely(stat_err & (1 << ICE_RX_FLEX_DESC_STATUS0_RSS_VALID_S))) { + mb->ol_flags |= PKT_RX_RSS_HASH; + mb->hash.rss = rte_le_to_cpu_32(desc->rss_hash); + } + +#ifndef RTE_LIBRTE_ICE_16BYTE_RX_DESC + if (desc->flow_id != 0xFFFFFFFF) { + mb->ol_flags |= PKT_RX_FDIR | PKT_RX_FDIR_ID; + mb->hash.fdir.hi = rte_le_to_cpu_32(desc->flow_id); + } +#endif +} + static inline void ice_rxd_to_pkt_fields_by_comms_ovs(__rte_unused struct ice_rx_queue *rxq, struct rte_mbuf *mb, @@ -147,7 +197,7 @@ ice_rxd_to_pkt_fields_by_comms_aux_v2(struct ice_rx_queue *rxq, #endif } -static void +void ice_select_rxd_to_pkt_fields_handler(struct ice_rx_queue *rxq, uint32_t rxdid) { switch (rxdid) { @@ -181,6 +231,10 @@ ice_select_rxd_to_pkt_fields_handler(struct ice_rx_queue *rxq, uint32_t rxdid) rxq->rxd_to_pkt_fields = ice_rxd_to_pkt_fields_by_comms_aux_v2; break; + case ICE_RXDID_COMMS_GENERIC: + rxq->rxd_to_pkt_fields = ice_rxd_to_pkt_fields_by_comms_generic; + break; + case ICE_RXDID_COMMS_OVS: rxq->rxd_to_pkt_fields = ice_rxd_to_pkt_fields_by_comms_ovs; break; @@ -201,11 +255,11 @@ ice_program_hw_rx_queue(struct ice_rx_queue *rxq) struct ice_vsi *vsi = rxq->vsi; struct ice_hw *hw = ICE_VSI_TO_HW(vsi); struct ice_pf *pf = ICE_VSI_TO_PF(vsi); - struct rte_eth_dev *dev = ICE_VSI_TO_ETH_DEV(rxq->vsi); + struct rte_eth_dev_data *dev_data = rxq->vsi->adapter->pf.dev_data; struct ice_rlan_ctx rx_ctx; enum ice_status err; - uint16_t buf_size, len; - struct rte_eth_rxmode *rxmode = &dev->data->dev_conf.rxmode; + uint16_t buf_size; + struct rte_eth_rxmode *rxmode = &dev_data->dev_conf.rxmode; uint32_t rxdid = ICE_RXDID_COMMS_OVS; uint32_t regval; @@ -214,28 +268,28 @@ ice_program_hw_rx_queue(struct ice_rx_queue *rxq) RTE_PKTMBUF_HEADROOM); rxq->rx_hdr_len = 0; rxq->rx_buf_len = RTE_ALIGN(buf_size, (1 << ICE_RLAN_CTX_DBUF_S)); - len = ICE_SUPPORT_CHAIN_NUM * rxq->rx_buf_len; - rxq->max_pkt_len = RTE_MIN(len, - dev->data->dev_conf.rxmode.max_rx_pkt_len); + rxq->max_pkt_len = RTE_MIN((uint32_t) + ICE_SUPPORT_CHAIN_NUM * rxq->rx_buf_len, + dev_data->dev_conf.rxmode.max_rx_pkt_len); if (rxmode->offloads & DEV_RX_OFFLOAD_JUMBO_FRAME) { - if (rxq->max_pkt_len <= RTE_ETHER_MAX_LEN || + if (rxq->max_pkt_len <= ICE_ETH_MAX_LEN || rxq->max_pkt_len > ICE_FRAME_SIZE_MAX) { PMD_DRV_LOG(ERR, "maximum packet length must " "be larger than %u and smaller than %u," "as jumbo frame is enabled", - (uint32_t)RTE_ETHER_MAX_LEN, + (uint32_t)ICE_ETH_MAX_LEN, (uint32_t)ICE_FRAME_SIZE_MAX); return -EINVAL; } } else { if (rxq->max_pkt_len < RTE_ETHER_MIN_LEN || - rxq->max_pkt_len > RTE_ETHER_MAX_LEN) { + rxq->max_pkt_len > ICE_ETH_MAX_LEN) { PMD_DRV_LOG(ERR, "maximum packet length must be " "larger than %u and smaller than %u, " "as jumbo frame is disabled", (uint32_t)RTE_ETHER_MIN_LEN, - (uint32_t)RTE_ETHER_MAX_LEN); + (uint32_t)ICE_ETH_MAX_LEN); return -EINVAL; } } @@ -312,7 +366,7 @@ ice_program_hw_rx_queue(struct ice_rx_queue *rxq) /* Check if scattered RX needs to be used. */ if (rxq->max_pkt_len > buf_size) - dev->data->scattered_rx = 1; + dev_data->scattered_rx = 1; rxq->qrx_tail = hw->hw_addr + QRX_TAIL(rxq->reg_idx); @@ -674,7 +728,7 @@ ice_fdir_program_hw_rx_queue(struct ice_rx_queue *rxq) rx_ctx.hbuf = rxq->rx_hdr_len >> ICE_RLAN_CTX_HBUF_S; rx_ctx.dtype = 0; /* No Header Split mode */ rx_ctx.dsize = 1; /* 32B descriptors */ - rx_ctx.rxmax = RTE_ETHER_MAX_LEN; + rx_ctx.rxmax = ICE_ETH_MAX_LEN; /* TPH: Transaction Layer Packet (TLP) processing hints */ rx_ctx.tphrdesc_ena = 1; rx_ctx.tphwdesc_ena = 1; @@ -1005,6 +1059,7 @@ ice_rx_queue_setup(struct rte_eth_dev *dev, uint32_t ring_size; uint16_t len; int use_def_burst_func = 1; + uint64_t offloads; if (nb_desc % ICE_ALIGN_RING_DESC != 0 || nb_desc > ICE_MAX_RING_DESC || @@ -1014,6 +1069,8 @@ ice_rx_queue_setup(struct rte_eth_dev *dev, return -EINVAL; } + offloads = rx_conf->offloads | dev->data->dev_conf.rxmode.offloads; + /* Free memory if needed */ if (dev->data->rx_queues[queue_idx]) { ice_rx_queue_release(dev->data->rx_queues[queue_idx]); @@ -1034,6 +1091,7 @@ ice_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; + rxq->offloads = offloads; rxq->reg_idx = vsi->base_queue + queue_idx; rxq->port_id = dev->data->port_id; @@ -1422,7 +1480,12 @@ ice_rxd_error_to_pkt_flags(uint16_t stat_err0) flags |= PKT_RX_L4_CKSUM_GOOD; if (unlikely(stat_err0 & (1 << ICE_RX_FLEX_DESC_STATUS0_XSUM_EIPE_S))) - flags |= PKT_RX_EIP_CKSUM_BAD; + flags |= PKT_RX_OUTER_IP_CKSUM_BAD; + + if (unlikely(stat_err0 & (1 << ICE_RX_FLEX_DESC_STATUS0_XSUM_EUDPE_S))) + flags |= PKT_RX_OUTER_L4_CKSUM_BAD; + else + flags |= PKT_RX_OUTER_L4_CKSUM_GOOD; return flags; } @@ -1612,7 +1675,6 @@ rx_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) { struct ice_rx_queue *rxq = (struct ice_rx_queue *)rx_queue; uint16_t nb_rx = 0; - struct rte_eth_dev *dev; if (!nb_pkts) return 0; @@ -1629,8 +1691,7 @@ rx_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, uint16_t nb_pkts) if (ice_rx_alloc_bufs(rxq) != 0) { uint16_t i, j; - dev = ICE_VSI_TO_ETH_DEV(rxq->vsi); - dev->data->rx_mbuf_alloc_failed += + rxq->vsi->adapter->pf.dev_data->rx_mbuf_alloc_failed += rxq->rx_free_thresh; PMD_RX_LOG(DEBUG, "Rx mbuf alloc failed for " "port_id=%u, queue_id=%u", @@ -1703,7 +1764,6 @@ ice_recv_scattered_pkts(void *rx_queue, uint64_t dma_addr; uint64_t pkt_flags; uint32_t *ptype_tbl = rxq->vsi->adapter->ptype_tbl; - struct rte_eth_dev *dev; while (nb_rx < nb_pkts) { rxdp = &rx_ring[rx_id]; @@ -1716,8 +1776,7 @@ ice_recv_scattered_pkts(void *rx_queue, /* allocate mbuf */ nmb = rte_mbuf_raw_alloc(rxq->mp); if (unlikely(!nmb)) { - dev = ICE_VSI_TO_ETH_DEV(rxq->vsi); - dev->data->rx_mbuf_alloc_failed++; + rxq->vsi->adapter->pf.dev_data->rx_mbuf_alloc_failed++; break; } rxd = *rxdp; /* copy descriptor in ring to temp variable*/ @@ -1841,7 +1900,7 @@ ice_recv_scattered_pkts(void *rx_queue, rx_id = (uint16_t)(rx_id == 0 ? (rxq->nb_rx_desc - 1) : (rx_id - 1)); /* write TAIL register */ - ICE_PCI_REG_WRITE(rxq->qrx_tail, rx_id); + ICE_PCI_REG_WC_WRITE(rxq->qrx_tail, rx_id); nb_hold = 0; } rxq->nb_rx_hold = nb_hold; @@ -1929,8 +1988,16 @@ ice_dev_supported_ptypes_get(struct rte_eth_dev *dev) #ifdef RTE_ARCH_X86 if (dev->rx_pkt_burst == ice_recv_pkts_vec || dev->rx_pkt_burst == ice_recv_scattered_pkts_vec || +#ifdef CC_AVX512_SUPPORT + dev->rx_pkt_burst == ice_recv_pkts_vec_avx512 || + dev->rx_pkt_burst == ice_recv_pkts_vec_avx512_offload || + dev->rx_pkt_burst == ice_recv_scattered_pkts_vec_avx512 || + dev->rx_pkt_burst == ice_recv_scattered_pkts_vec_avx512_offload || +#endif dev->rx_pkt_burst == ice_recv_pkts_vec_avx2 || - dev->rx_pkt_burst == ice_recv_scattered_pkts_vec_avx2) + dev->rx_pkt_burst == ice_recv_pkts_vec_avx2_offload || + dev->rx_pkt_burst == ice_recv_scattered_pkts_vec_avx2 || + dev->rx_pkt_burst == ice_recv_scattered_pkts_vec_avx2_offload) return ptypes; #endif @@ -2035,7 +2102,7 @@ ice_fdir_setup_tx_resources(struct ice_pf *pf) return -EINVAL; } - dev = pf->adapter->eth_dev; + dev = &rte_eth_devices[pf->adapter->pf.dev_data->port_id]; /* Allocate the TX queue data structure. */ txq = rte_zmalloc_socket("ice fdir tx queue", @@ -2093,7 +2160,7 @@ ice_fdir_setup_rx_resources(struct ice_pf *pf) return -EINVAL; } - dev = pf->adapter->eth_dev; + dev = &rte_eth_devices[pf->adapter->pf.dev_data->port_id]; /* Allocate the RX queue data structure. */ rxq = rte_zmalloc_socket("ice fdir rx queue", @@ -2162,7 +2229,6 @@ ice_recv_pkts(void *rx_queue, uint64_t dma_addr; uint64_t pkt_flags; uint32_t *ptype_tbl = rxq->vsi->adapter->ptype_tbl; - struct rte_eth_dev *dev; while (nb_rx < nb_pkts) { rxdp = &rx_ring[rx_id]; @@ -2175,8 +2241,7 @@ ice_recv_pkts(void *rx_queue, /* allocate mbuf */ nmb = rte_mbuf_raw_alloc(rxq->mp); if (unlikely(!nmb)) { - dev = ICE_VSI_TO_ETH_DEV(rxq->vsi); - dev->data->rx_mbuf_alloc_failed++; + rxq->vsi->adapter->pf.dev_data->rx_mbuf_alloc_failed++; break; } rxd = *rxdp; /* copy descriptor in ring to temp variable*/ @@ -2231,7 +2296,7 @@ ice_recv_pkts(void *rx_queue, rx_id = (uint16_t)(rx_id == 0 ? (rxq->nb_rx_desc - 1) : (rx_id - 1)); /* write TAIL register */ - ICE_PCI_REG_WRITE(rxq->qrx_tail, rx_id); + ICE_PCI_REG_WC_WRITE(rxq->qrx_tail, rx_id); nb_hold = 0; } rxq->nb_rx_hold = nb_hold; @@ -2288,8 +2353,11 @@ ice_parse_tunneling_params(uint64_t ol_flags, *cd_tunneling |= (tx_offload.l2_len >> 1) << ICE_TXD_CTX_QW0_NATLEN_S; - if ((ol_flags & PKT_TX_OUTER_UDP_CKSUM) && - (ol_flags & PKT_TX_OUTER_IP_CKSUM) && + /** + * Calculate the tunneling UDP checksum. + * Shall be set only if L4TUNT = 01b and EIPT is not zero + */ + if (!(*cd_tunneling & ICE_TX_CTX_EIPT_NONE) && (*cd_tunneling & ICE_TXD_CTX_UDP_TUNNELING)) *cd_tunneling |= ICE_TXD_CTX_QW0_L4T_CS_M; } @@ -2371,11 +2439,11 @@ ice_xmit_cleanup(struct ice_tx_queue *txq) desc_to_clean_to = sw_ring[desc_to_clean_to].last_id; if (!(txd[desc_to_clean_to].cmd_type_offset_bsz & rte_cpu_to_le_64(ICE_TX_DESC_DTYPE_DESC_DONE))) { - PMD_TX_FREE_LOG(DEBUG, "TX descriptor %4u is not done " - "(port=%d queue=%d) value=0x%"PRIx64"\n", - desc_to_clean_to, - txq->port_id, txq->queue_id, - txd[desc_to_clean_to].cmd_type_offset_bsz); + PMD_TX_LOG(DEBUG, "TX descriptor %4u is not done " + "(port=%d queue=%d) value=0x%"PRIx64"\n", + desc_to_clean_to, + txq->port_id, txq->queue_id, + txd[desc_to_clean_to].cmd_type_offset_bsz); /* Failed to clean any descriptors */ return -1; } @@ -2666,10 +2734,10 @@ ice_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts) /* set RS bit on the last descriptor of one packet */ if (txq->nb_tx_used >= txq->tx_rs_thresh) { - PMD_TX_FREE_LOG(DEBUG, - "Setting RS bit on TXD id=" - "%4u (port=%d queue=%d)", - tx_last, txq->port_id, txq->queue_id); + PMD_TX_LOG(DEBUG, + "Setting RS bit on TXD id=" + "%4u (port=%d queue=%d)", + tx_last, txq->port_id, txq->queue_id); td_cmd |= ICE_TX_DESC_CMD_RS; @@ -2946,7 +3014,7 @@ tx_xmit_pkts(struct ice_tx_queue *txq, txq->tx_tail = 0; /* Update the tx tail register */ - ICE_PCI_REG_WRITE(txq->qtx_tail, txq->tx_tail); + ICE_PCI_REG_WC_WRITE(txq->qtx_tail, txq->tx_tail); return nb_pkts; } @@ -2986,10 +3054,14 @@ ice_set_rx_function(struct rte_eth_dev *dev) #ifdef RTE_ARCH_X86 struct ice_rx_queue *rxq; int i; - bool use_avx2 = false; + int rx_check_ret = -1; if (rte_eal_process_type() == RTE_PROC_PRIMARY) { - if (!ice_rx_vec_dev_check(dev) && ad->rx_bulk_alloc_allowed) { + ad->rx_use_avx512 = false; + ad->rx_use_avx2 = false; + rx_check_ret = ice_rx_vec_dev_check(dev); + if (rx_check_ret >= 0 && ad->rx_bulk_alloc_allowed && + rte_vect_get_max_simd_bitwidth() >= RTE_VECT_SIMD_128) { ad->rx_vec_allowed = true; for (i = 0; i < dev->data->nb_rx_queues; i++) { rxq = dev->data->rx_queues[i]; @@ -2999,9 +3071,20 @@ ice_set_rx_function(struct rte_eth_dev *dev) } } - if (rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX2) == 1 || - rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX512F) == 1) - use_avx2 = true; + if (rte_vect_get_max_simd_bitwidth() >= RTE_VECT_SIMD_512 && + rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX512F) == 1 && + rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX512BW) == 1) +#ifdef CC_AVX512_SUPPORT + ad->rx_use_avx512 = true; +#else + PMD_DRV_LOG(NOTICE, + "AVX512 is not supported in build env"); +#endif + if (!ad->rx_use_avx512 && + (rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX2) == 1 || + rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX512F) == 1) && + rte_vect_get_max_simd_bitwidth() >= RTE_VECT_SIMD_256) + ad->rx_use_avx2 = true; } else { ad->rx_vec_allowed = false; @@ -3010,20 +3093,79 @@ ice_set_rx_function(struct rte_eth_dev *dev) if (ad->rx_vec_allowed) { if (dev->data->scattered_rx) { - PMD_DRV_LOG(DEBUG, - "Using %sVector Scattered Rx (port %d).", - use_avx2 ? "avx2 " : "", + if (ad->rx_use_avx512) { +#ifdef CC_AVX512_SUPPORT + if (rx_check_ret == ICE_VECTOR_OFFLOAD_PATH) { + PMD_DRV_LOG(NOTICE, + "Using AVX512 OFFLOAD Vector Scattered Rx (port %d).", + dev->data->port_id); + dev->rx_pkt_burst = + ice_recv_scattered_pkts_vec_avx512_offload; + } else { + PMD_DRV_LOG(NOTICE, + "Using AVX512 Vector Scattered Rx (port %d).", + dev->data->port_id); + dev->rx_pkt_burst = + ice_recv_scattered_pkts_vec_avx512; + } +#endif + } else if (ad->rx_use_avx2) { + if (rx_check_ret == ICE_VECTOR_OFFLOAD_PATH) { + PMD_DRV_LOG(NOTICE, + "Using AVX2 OFFLOAD Vector Scattered Rx (port %d).", + dev->data->port_id); + dev->rx_pkt_burst = + ice_recv_scattered_pkts_vec_avx2_offload; + } else { + PMD_DRV_LOG(NOTICE, + "Using AVX2 Vector Scattered Rx (port %d).", + dev->data->port_id); + dev->rx_pkt_burst = + ice_recv_scattered_pkts_vec_avx2; + } + } else { + PMD_DRV_LOG(DEBUG, + "Using Vector Scattered Rx (port %d).", dev->data->port_id); - dev->rx_pkt_burst = use_avx2 ? - ice_recv_scattered_pkts_vec_avx2 : - ice_recv_scattered_pkts_vec; + dev->rx_pkt_burst = ice_recv_scattered_pkts_vec; + } } else { - PMD_DRV_LOG(DEBUG, "Using %sVector Rx (port %d).", - use_avx2 ? "avx2 " : "", + if (ad->rx_use_avx512) { +#ifdef CC_AVX512_SUPPORT + if (rx_check_ret == ICE_VECTOR_OFFLOAD_PATH) { + PMD_DRV_LOG(NOTICE, + "Using AVX512 OFFLOAD Vector Rx (port %d).", + dev->data->port_id); + dev->rx_pkt_burst = + ice_recv_pkts_vec_avx512_offload; + } else { + PMD_DRV_LOG(NOTICE, + "Using AVX512 Vector Rx (port %d).", + dev->data->port_id); + dev->rx_pkt_burst = + ice_recv_pkts_vec_avx512; + } +#endif + } else if (ad->rx_use_avx2) { + if (rx_check_ret == ICE_VECTOR_OFFLOAD_PATH) { + PMD_DRV_LOG(NOTICE, + "Using AVX2 OFFLOAD Vector Rx (port %d).", + dev->data->port_id); + dev->rx_pkt_burst = + ice_recv_pkts_vec_avx2_offload; + } else { + PMD_DRV_LOG(NOTICE, + "Using AVX2 Vector Rx (port %d).", + dev->data->port_id); + dev->rx_pkt_burst = + ice_recv_pkts_vec_avx2; + } + } else { + PMD_DRV_LOG(DEBUG, + "Using Vector Rx (port %d).", dev->data->port_id); - dev->rx_pkt_burst = use_avx2 ? - ice_recv_pkts_vec_avx2 : - ice_recv_pkts_vec; + dev->rx_pkt_burst = ice_recv_pkts_vec; + } } return; } @@ -3060,8 +3202,16 @@ static const struct { { ice_recv_pkts_bulk_alloc, "Scalar Bulk Alloc" }, { ice_recv_pkts, "Scalar" }, #ifdef RTE_ARCH_X86 +#ifdef CC_AVX512_SUPPORT + { ice_recv_scattered_pkts_vec_avx512, "Vector AVX512 Scattered" }, + { ice_recv_scattered_pkts_vec_avx512_offload, "Offload Vector AVX512 Scattered" }, + { ice_recv_pkts_vec_avx512, "Vector AVX512" }, + { ice_recv_pkts_vec_avx512_offload, "Offload Vector AVX512" }, +#endif { ice_recv_scattered_pkts_vec_avx2, "Vector AVX2 Scattered" }, + { ice_recv_scattered_pkts_vec_avx2_offload, "Offload Vector AVX2 Scattered" }, { ice_recv_pkts_vec_avx2, "Vector AVX2" }, + { ice_recv_pkts_vec_avx2_offload, "Offload Vector AVX2" }, { ice_recv_scattered_pkts_vec, "Vector SSE Scattered" }, { ice_recv_pkts_vec, "Vector SSE" }, #endif @@ -3140,7 +3290,7 @@ ice_prep_pkts(__rte_unused void *tx_queue, struct rte_mbuf **tx_pkts, return i; } -#ifdef RTE_LIBRTE_ETHDEV_DEBUG +#ifdef RTE_ETHDEV_DEBUG_TX ret = rte_validate_tx_offload(m); if (ret != 0) { rte_errno = -ret; @@ -3164,36 +3314,84 @@ ice_set_tx_function(struct rte_eth_dev *dev) #ifdef RTE_ARCH_X86 struct ice_tx_queue *txq; int i; - bool use_avx2 = false; + int tx_check_ret = -1; if (rte_eal_process_type() == RTE_PROC_PRIMARY) { - if (!ice_tx_vec_dev_check(dev)) { + ad->tx_use_avx2 = false; + ad->tx_use_avx512 = false; + tx_check_ret = ice_tx_vec_dev_check(dev); + if (tx_check_ret >= 0 && + rte_vect_get_max_simd_bitwidth() >= RTE_VECT_SIMD_128) { ad->tx_vec_allowed = true; - for (i = 0; i < dev->data->nb_tx_queues; i++) { - txq = dev->data->tx_queues[i]; - if (txq && ice_txq_vec_setup(txq)) { - ad->tx_vec_allowed = false; - break; + + if (rte_vect_get_max_simd_bitwidth() >= RTE_VECT_SIMD_512 && + rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX512F) == 1 && + rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX512BW) == 1) +#ifdef CC_AVX512_SUPPORT + ad->tx_use_avx512 = true; +#else + PMD_DRV_LOG(NOTICE, + "AVX512 is not supported in build env"); +#endif + if (!ad->tx_use_avx512 && + (rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX2) == 1 || + rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX512F) == 1) && + rte_vect_get_max_simd_bitwidth() >= RTE_VECT_SIMD_256) + ad->tx_use_avx2 = true; + + if (!ad->tx_use_avx2 && !ad->tx_use_avx512 && + tx_check_ret == ICE_VECTOR_OFFLOAD_PATH) + ad->tx_vec_allowed = false; + + if (ad->tx_vec_allowed) { + for (i = 0; i < dev->data->nb_tx_queues; i++) { + txq = dev->data->tx_queues[i]; + if (txq && ice_txq_vec_setup(txq)) { + ad->tx_vec_allowed = false; + break; + } } } - - if (rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX2) == 1 || - rte_cpu_get_flag_enabled(RTE_CPUFLAG_AVX512F) == 1) - use_avx2 = true; - } else { ad->tx_vec_allowed = false; } } if (ad->tx_vec_allowed) { - PMD_DRV_LOG(DEBUG, "Using %sVector Tx (port %d).", - use_avx2 ? "avx2 " : "", - dev->data->port_id); - dev->tx_pkt_burst = use_avx2 ? - ice_xmit_pkts_vec_avx2 : - ice_xmit_pkts_vec; dev->tx_pkt_prepare = NULL; + if (ad->tx_use_avx512) { +#ifdef CC_AVX512_SUPPORT + if (tx_check_ret == ICE_VECTOR_OFFLOAD_PATH) { + PMD_DRV_LOG(NOTICE, + "Using AVX512 OFFLOAD Vector Tx (port %d).", + dev->data->port_id); + dev->tx_pkt_burst = + ice_xmit_pkts_vec_avx512_offload; + dev->tx_pkt_prepare = ice_prep_pkts; + } else { + PMD_DRV_LOG(NOTICE, + "Using AVX512 Vector Tx (port %d).", + dev->data->port_id); + dev->tx_pkt_burst = ice_xmit_pkts_vec_avx512; + } +#endif + } else { + if (tx_check_ret == ICE_VECTOR_OFFLOAD_PATH) { + PMD_DRV_LOG(NOTICE, + "Using AVX2 OFFLOAD Vector Tx (port %d).", + dev->data->port_id); + dev->tx_pkt_burst = + ice_xmit_pkts_vec_avx2_offload; + dev->tx_pkt_prepare = ice_prep_pkts; + } else { + PMD_DRV_LOG(DEBUG, "Using %sVector Tx (port %d).", + ad->tx_use_avx2 ? "avx2 " : "", + dev->data->port_id); + dev->tx_pkt_burst = ad->tx_use_avx2 ? + ice_xmit_pkts_vec_avx2 : + ice_xmit_pkts_vec; + } + } return; } @@ -3217,6 +3415,10 @@ static const struct { { ice_xmit_pkts_simple, "Scalar Simple" }, { ice_xmit_pkts, "Scalar" }, #ifdef RTE_ARCH_X86 +#ifdef CC_AVX512_SUPPORT + { ice_xmit_pkts_vec_avx512, "Vector AVX512" }, + { ice_xmit_pkts_vec_avx512_offload, "Offload Vector AVX512" }, +#endif { ice_xmit_pkts_vec_avx2, "Vector AVX2" }, { ice_xmit_pkts_vec, "Vector SSE" }, #endif @@ -3779,6 +3981,50 @@ ice_get_default_pkt_type(uint16_t ptype) RTE_PTYPE_TUNNEL_GTPU | RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_ICMP, + + /* IPv4 --> UDP ECPRI */ + [372] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [373] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [374] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [375] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [376] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [377] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [378] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [379] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [380] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [381] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + + /* IPV6 --> UDP ECPRI */ + [382] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [383] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [384] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [385] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [386] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [387] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [388] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [389] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [390] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, + [391] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_UDP, /* All others reserved */ };