X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=drivers%2Fnet%2Fmlx5%2Fmlx5_rxtx.c;h=88402ef5eb0a470b387ec0b65b3322a1f55671c0;hb=1be514fbcea9e8964296b46c91dbb56715503ae7;hp=b55db4fa74551c07e0f809fc7deedcc58966392a;hpb=7593cf1d3500dfce7c9faacfc01f5797556b6a80;p=dpdk.git diff --git a/drivers/net/mlx5/mlx5_rxtx.c b/drivers/net/mlx5/mlx5_rxtx.c index b55db4fa74..88402ef5eb 100644 --- a/drivers/net/mlx5/mlx5_rxtx.c +++ b/drivers/net/mlx5/mlx5_rxtx.c @@ -7,17 +7,6 @@ #include #include -/* Verbs header. */ -/* ISO C doesn't support unnamed structs/unions, disabling -pedantic. */ -#ifdef PEDANTIC -#pragma GCC diagnostic ignored "-Wpedantic" -#endif -#include -#include -#ifdef PEDANTIC -#pragma GCC diagnostic error "-Wpedantic" -#endif - #include #include #include @@ -27,15 +16,15 @@ #include #include -#include #include #include +#include "mlx5_autoconf.h" #include "mlx5_defs.h" #include "mlx5.h" +#include "mlx5_mr.h" #include "mlx5_utils.h" #include "mlx5_rxtx.h" -#include "mlx5_autoconf.h" /* TX burst subroutines return codes. */ enum mlx5_txcmp_code { @@ -65,6 +54,7 @@ enum mlx5_txcmp_code { #define MLX5_TXOFF_CONFIG_METADATA (1u << 6) /* Flow metadata. */ #define MLX5_TXOFF_CONFIG_EMPW (1u << 8) /* Enhanced MPW supported.*/ #define MLX5_TXOFF_CONFIG_MPW (1u << 9) /* Legacy MPW supported.*/ +#define MLX5_TXOFF_CONFIG_TXPP (1u << 10) /* Scheduling on timestamp.*/ /* The most common offloads groups. */ #define MLX5_TXOFF_CONFIG_NONE 0 @@ -103,22 +93,18 @@ static __rte_always_inline void rxq_cq_to_mbuf(struct mlx5_rxq_data *rxq, struct rte_mbuf *pkt, volatile struct mlx5_cqe *cqe, uint32_t rss_hash_res); -static __rte_always_inline void -mprq_buf_replace(struct mlx5_rxq_data *rxq, uint16_t rq_idx, - const unsigned int strd_n); - static int mlx5_queue_state_modify(struct rte_eth_dev *dev, struct mlx5_mp_arg_queue_state_modify *sm); static inline void -mlx5_lro_update_tcp_hdr(struct rte_tcp_hdr *restrict tcp, - volatile struct mlx5_cqe *restrict cqe, +mlx5_lro_update_tcp_hdr(struct rte_tcp_hdr *__rte_restrict tcp, + volatile struct mlx5_cqe *__rte_restrict cqe, uint32_t phcsum); static inline void -mlx5_lro_update_hdr(uint8_t *restrict padd, - volatile struct mlx5_cqe *restrict cqe, +mlx5_lro_update_hdr(uint8_t *__rte_restrict padd, + volatile struct mlx5_cqe *__rte_restrict cqe, uint32_t len); uint32_t mlx5_ptype_table[] __rte_cache_aligned = { @@ -373,7 +359,7 @@ mlx5_set_swp_types_table(void) * Software Parser flags are set by pointer. */ static __rte_always_inline uint32_t -txq_mbuf_to_swp(struct mlx5_txq_local *restrict loc, +txq_mbuf_to_swp(struct mlx5_txq_local *__rte_restrict loc, uint8_t *swp_flags, unsigned int olx) { @@ -473,19 +459,11 @@ rx_queue_count(struct mlx5_rxq_data *rxq) { struct rxq_zip *zip = &rxq->zip; volatile struct mlx5_cqe *cqe; + unsigned int cq_ci = rxq->cq_ci; const unsigned int cqe_n = (1 << rxq->cqe_n); const unsigned int cqe_cnt = cqe_n - 1; - unsigned int cq_ci; - unsigned int used; + unsigned int used = 0; - /* if we are processing a compressed cqe */ - if (zip->ai) { - used = zip->cqe_cnt - zip->ca; - cq_ci = zip->cq_ci; - } else { - used = 0; - cq_ci = rxq->cq_ci; - } cqe = &(*rxq->cqes)[cq_ci & cqe_cnt]; while (check_cqe(cqe, cqe_n, cq_ci) != MLX5_CQE_STATUS_HW_OWN) { int8_t op_own; @@ -493,14 +471,17 @@ rx_queue_count(struct mlx5_rxq_data *rxq) op_own = cqe->op_own; if (MLX5_CQE_FORMAT(op_own) == MLX5_COMPRESSED) - n = rte_be_to_cpu_32(cqe->byte_cnt); + if (unlikely(zip->ai)) + n = zip->cqe_cnt - zip->ai; + else + n = rte_be_to_cpu_32(cqe->byte_cnt); else n = 1; cq_ci += n; used += n; cqe = &(*rxq->cqes)[cq_ci & cqe_cnt]; } - used = RTE_MIN(used, (1U << rxq->elts_n) - 1); + used = RTE_MIN(used, cqe_n); return used; } @@ -523,11 +504,12 @@ mlx5_rx_descriptor_status(void *rx_queue, uint16_t offset) container_of(rxq, struct mlx5_rxq_ctrl, rxq); struct rte_eth_dev *dev = ETH_DEV(rxq_ctrl->priv); - if (dev->rx_pkt_burst != mlx5_rx_burst) { + if (dev->rx_pkt_burst == NULL || + dev->rx_pkt_burst == removed_rx_burst) { rte_errno = ENOTSUP; return -rte_errno; } - if (offset >= (1 << rxq->elts_n)) { + if (offset >= (1 << rxq->cqe_n)) { rte_errno = EINVAL; return -rte_errno; } @@ -598,7 +580,14 @@ mlx5_rx_burst_mode_get(struct rte_eth_dev *dev, struct rte_eth_burst_mode *mode) { eth_rx_burst_t pkt_burst = dev->rx_pkt_burst; + struct mlx5_priv *priv = dev->data->dev_private; + struct mlx5_rxq_data *rxq; + rxq = (*priv->rxqs)[rx_queue_id]; + if (!rxq) { + rte_errno = EINVAL; + return -rte_errno; + } if (pkt_burst == mlx5_rx_burst) { snprintf(mode->info, sizeof(mode->info), "%s", "Scalar"); } else if (pkt_burst == mlx5_rx_burst_mprq) { @@ -612,6 +601,16 @@ mlx5_rx_burst_mode_get(struct rte_eth_dev *dev, snprintf(mode->info, sizeof(mode->info), "%s", "Vector AltiVec"); #else return -EINVAL; +#endif + } else if (pkt_burst == mlx5_rx_burst_mprq_vec) { +#if defined RTE_ARCH_X86_64 + snprintf(mode->info, sizeof(mode->info), "%s", "MPRQ Vector SSE"); +#elif defined RTE_ARCH_ARM64 + snprintf(mode->info, sizeof(mode->info), "%s", "MPRQ Vector Neon"); +#elif defined RTE_ARCH_PPC_64 + snprintf(mode->info, sizeof(mode->info), "%s", "MPRQ Vector AltiVec"); +#else + return -EINVAL; #endif } else { return -EINVAL; @@ -638,7 +637,8 @@ mlx5_rx_queue_count(struct rte_eth_dev *dev, uint16_t rx_queue_id) struct mlx5_priv *priv = dev->data->dev_private; struct mlx5_rxq_data *rxq; - if (dev->rx_pkt_burst != mlx5_rx_burst) { + if (dev->rx_pkt_burst == NULL || + dev->rx_pkt_burst == removed_rx_burst) { rte_errno = ENOTSUP; return -rte_errno; } @@ -746,7 +746,7 @@ check_err_cqe_seen(volatile struct mlx5_err_cqe *err_cqe) * the error completion entry is handled successfully. */ static int -mlx5_tx_error_cqe_handle(struct mlx5_txq_data *restrict txq, +mlx5_tx_error_cqe_handle(struct mlx5_txq_data *__rte_restrict txq, volatile struct mlx5_err_cqe *err_cqe) { if (err_cqe->syndrome != MLX5_CQE_SYNDROME_WR_FLUSH_ERR) { @@ -879,9 +879,11 @@ mlx5_rxq_initialize(struct mlx5_rxq_data *rxq) rxq->zip = (struct rxq_zip){ .ai = 0, }; + rxq->elts_ci = mlx5_rxq_mprq_enabled(rxq) ? + (wqe_n >> rxq->sges_n) * (1 << rxq->strd_num_n) : 0; /* Update doorbell counter. */ rxq->rq_ci = wqe_n >> rxq->sges_n; - rte_cio_wmb(); + rte_io_wmb(); *rxq->rq_db = rte_cpu_to_be_32(rxq->rq_ci); } @@ -909,30 +911,7 @@ mlx5_queue_state_modify_primary(struct rte_eth_dev *dev, struct mlx5_rxq_ctrl *rxq_ctrl = container_of(rxq, struct mlx5_rxq_ctrl, rxq); - if (rxq_ctrl->obj->type == MLX5_RXQ_OBJ_TYPE_IBV) { - struct ibv_wq_attr mod = { - .attr_mask = IBV_WQ_ATTR_STATE, - .wq_state = sm->state, - }; - - ret = mlx5_glue->modify_wq(rxq_ctrl->obj->wq, &mod); - } else { /* rxq_ctrl->obj->type == MLX5_RXQ_OBJ_TYPE_DEVX_RQ. */ - struct mlx5_devx_modify_rq_attr rq_attr; - - memset(&rq_attr, 0, sizeof(rq_attr)); - if (sm->state == IBV_WQS_RESET) { - rq_attr.rq_state = MLX5_RQC_STATE_ERR; - rq_attr.state = MLX5_RQC_STATE_RST; - } else if (sm->state == IBV_WQS_RDY) { - rq_attr.rq_state = MLX5_RQC_STATE_RST; - rq_attr.state = MLX5_RQC_STATE_RDY; - } else if (sm->state == IBV_WQS_ERR) { - rq_attr.rq_state = MLX5_RQC_STATE_RDY; - rq_attr.state = MLX5_RQC_STATE_ERR; - } - ret = mlx5_devx_cmd_modify_rq(rxq_ctrl->obj->rq, - &rq_attr); - } + ret = priv->obj_ops.rxq_obj_modify(rxq_ctrl->obj, sm->state); if (ret) { DRV_LOG(ERR, "Cannot change Rx WQ state to %u - %s", sm->state, strerror(errno)); @@ -943,44 +922,12 @@ mlx5_queue_state_modify_primary(struct rte_eth_dev *dev, struct mlx5_txq_data *txq = (*priv->txqs)[sm->queue_id]; struct mlx5_txq_ctrl *txq_ctrl = container_of(txq, struct mlx5_txq_ctrl, txq); - struct ibv_qp_attr mod = { - .qp_state = IBV_QPS_RESET, - .port_num = (uint8_t)priv->ibv_port, - }; - struct ibv_qp *qp = txq_ctrl->obj->qp; - ret = mlx5_glue->modify_qp(qp, &mod, IBV_QP_STATE); - if (ret) { - DRV_LOG(ERR, "Cannot change the Tx QP state to RESET " - "%s", strerror(errno)); - rte_errno = errno; - return ret; - } - mod.qp_state = IBV_QPS_INIT; - ret = mlx5_glue->modify_qp(qp, &mod, - (IBV_QP_STATE | IBV_QP_PORT)); - if (ret) { - DRV_LOG(ERR, "Cannot change Tx QP state to INIT %s", - strerror(errno)); - rte_errno = errno; - return ret; - } - mod.qp_state = IBV_QPS_RTR; - ret = mlx5_glue->modify_qp(qp, &mod, IBV_QP_STATE); - if (ret) { - DRV_LOG(ERR, "Cannot change Tx QP state to RTR %s", - strerror(errno)); - rte_errno = errno; - return ret; - } - mod.qp_state = IBV_QPS_RTS; - ret = mlx5_glue->modify_qp(qp, &mod, IBV_QP_STATE); - if (ret) { - DRV_LOG(ERR, "Cannot change Tx QP state to RTS %s", - strerror(errno)); - rte_errno = errno; + ret = priv->obj_ops.txq_obj_modify(txq_ctrl->obj, + MLX5_TXQ_MOD_ERR2RDY, + (uint8_t)priv->dev_port); + if (ret) return ret; - } } return 0; } @@ -1000,6 +947,7 @@ static int mlx5_queue_state_modify(struct rte_eth_dev *dev, struct mlx5_mp_arg_queue_state_modify *sm) { + struct mlx5_priv *priv = dev->data->dev_private; int ret = 0; switch (rte_eal_process_type()) { @@ -1007,7 +955,7 @@ mlx5_queue_state_modify(struct rte_eth_dev *dev, ret = mlx5_queue_state_modify_primary(dev, sm); break; case RTE_PROC_SECONDARY: - ret = mlx5_mp_req_queue_state_modify(dev, sm); + ret = mlx5_mp_req_queue_state_modify(&priv->mp_id, sm); break; default: break; @@ -1036,7 +984,8 @@ mlx5_rx_err_handle(struct mlx5_rxq_data *rxq, uint8_t vec) { const uint16_t cqe_n = 1 << rxq->cqe_n; const uint16_t cqe_mask = cqe_n - 1; - const unsigned int wqe_n = 1 << rxq->elts_n; + const uint16_t wqe_n = 1 << rxq->elts_n; + const uint16_t strd_n = 1 << rxq->strd_num_n; struct mlx5_rxq_ctrl *rxq_ctrl = container_of(rxq, struct mlx5_rxq_ctrl, rxq); union { @@ -1084,15 +1033,15 @@ mlx5_rx_err_handle(struct mlx5_rxq_data *rxq, uint8_t vec) case MLX5_RXQ_ERR_STATE_NEED_READY: ret = check_cqe(u.cqe, cqe_n, rxq->cq_ci); if (ret == MLX5_CQE_STATUS_HW_OWN) { - rte_cio_wmb(); + rte_io_wmb(); *rxq->cq_db = rte_cpu_to_be_32(rxq->cq_ci); - rte_cio_wmb(); + rte_io_wmb(); /* * The RQ consumer index must be zeroed while moving * from RESET state to RDY state. */ *rxq->rq_db = rte_cpu_to_be_32(0); - rte_cio_wmb(); + rte_io_wmb(); sm.is_wq = 1; sm.queue_id = rxq->idx; sm.state = IBV_WQS_RDY; @@ -1100,21 +1049,27 @@ mlx5_rx_err_handle(struct mlx5_rxq_data *rxq, uint8_t vec) &sm)) return -1; if (vec) { - const uint16_t q_mask = wqe_n - 1; - uint16_t elt_idx; + const uint32_t elts_n = + mlx5_rxq_mprq_enabled(rxq) ? + wqe_n * strd_n : wqe_n; + const uint32_t e_mask = elts_n - 1; + uint32_t elts_ci = + mlx5_rxq_mprq_enabled(rxq) ? + rxq->elts_ci : rxq->rq_ci; + uint32_t elt_idx; struct rte_mbuf **elt; int i; - unsigned int n = wqe_n - (rxq->rq_ci - + unsigned int n = elts_n - (elts_ci - rxq->rq_pi); for (i = 0; i < (int)n; ++i) { - elt_idx = (rxq->rq_ci + i) & q_mask; + elt_idx = (elts_ci + i) & e_mask; elt = &(*rxq->elts)[elt_idx]; *elt = rte_mbuf_raw_alloc(rxq->mp); if (!*elt) { for (i--; i >= 0; --i) { - elt_idx = (rxq->rq_ci + - i) & q_mask; + elt_idx = (elts_ci + + i) & elts_n; elt = &(*rxq->elts) [elt_idx]; rte_pktmbuf_free_seg @@ -1123,7 +1078,7 @@ mlx5_rx_err_handle(struct mlx5_rxq_data *rxq, uint8_t vec) return -1; } } - for (i = 0; i < (int)wqe_n; ++i) { + for (i = 0; i < (int)elts_n; ++i) { elt = &(*rxq->elts)[i]; DATA_LEN(*elt) = (uint16_t)((*elt)->buf_len - @@ -1131,7 +1086,7 @@ mlx5_rx_err_handle(struct mlx5_rxq_data *rxq, uint8_t vec) } /* Padding with a fake mbuf for vec Rx. */ for (i = 0; i < MLX5_VPMD_DESCS_PER_LOOP; ++i) - (*rxq->elts)[wqe_n + i] = + (*rxq->elts)[elts_n + i] = &rxq->fake_mbuf; } mlx5_rxq_initialize(rxq); @@ -1337,9 +1292,10 @@ rxq_cq_to_mbuf(struct mlx5_rxq_data *rxq, struct rte_mbuf *pkt, pkt->hash.fdir.hi = mlx5_flow_mark_get(mark); } } - if (rte_flow_dynf_metadata_avail() && cqe->flow_table_metadata) { - pkt->ol_flags |= PKT_RX_DYNF_METADATA; - *RTE_FLOW_DYNF_METADATA(pkt) = cqe->flow_table_metadata; + if (rxq->dynf_meta && cqe->flow_table_metadata) { + pkt->ol_flags |= rxq->flow_meta_mask; + *RTE_MBUF_DYNFIELD(pkt, rxq->flow_meta_offset, uint32_t *) = + cqe->flow_table_metadata; } if (rxq->csum) pkt->ol_flags |= rxq_cq_to_ol_flags(cqe); @@ -1349,8 +1305,12 @@ rxq_cq_to_mbuf(struct mlx5_rxq_data *rxq, struct rte_mbuf *pkt, pkt->vlan_tci = rte_be_to_cpu_16(cqe->vlan_info); } if (rxq->hw_timestamp) { - pkt->timestamp = rte_be_to_cpu_64(cqe->timestamp); - pkt->ol_flags |= PKT_RX_TIMESTAMP; + uint64_t ts = rte_be_to_cpu_64(cqe->timestamp); + + if (rxq->rt_timestamp) + ts = mlx5_txpp_convert_rx_ts(rxq->sh, ts); + mlx5_timestamp_set(pkt, rxq->timestamp_offset, ts); + pkt->ol_flags |= rxq->timestamp_rx_flag; } } @@ -1396,7 +1356,8 @@ mlx5_rx_burst(void *dpdk_rxq, struct rte_mbuf **pkts, uint16_t pkts_n) rte_prefetch0(seg); rte_prefetch0(cqe); rte_prefetch0(wqe); - rep = rte_mbuf_raw_alloc(rxq->mp); + /* Allocate the buf from the same pool. */ + rep = rte_mbuf_raw_alloc(seg->pool); if (unlikely(rep == NULL)) { ++rxq->stats.rx_nombuf; if (!pkt) { @@ -1481,9 +1442,9 @@ mlx5_rx_burst(void *dpdk_rxq, struct rte_mbuf **pkts, uint16_t pkts_n) return 0; /* Update the consumer index. */ rxq->rq_ci = rq_ci >> sges_n; - rte_cio_wmb(); + rte_io_wmb(); *rxq->cq_db = rte_cpu_to_be_32(rxq->cq_ci); - rte_cio_wmb(); + rte_io_wmb(); *rxq->rq_db = rte_cpu_to_be_32(rxq->rq_ci); #ifdef MLX5_PMD_SOFT_COUNTERS /* Increment packets counter. */ @@ -1505,8 +1466,8 @@ mlx5_rx_burst(void *dpdk_rxq, struct rte_mbuf **pkts, uint16_t pkts_n) * The L3 pseudo-header checksum. */ static inline void -mlx5_lro_update_tcp_hdr(struct rte_tcp_hdr *restrict tcp, - volatile struct mlx5_cqe *restrict cqe, +mlx5_lro_update_tcp_hdr(struct rte_tcp_hdr *__rte_restrict tcp, + volatile struct mlx5_cqe *__rte_restrict cqe, uint32_t phcsum) { uint8_t l4_type = (rte_be_to_cpu_16(cqe->hdr_type_etc) & @@ -1526,7 +1487,7 @@ mlx5_lro_update_tcp_hdr(struct rte_tcp_hdr *restrict tcp, if (cqe->lro_tcppsh_abort_dupack & MLX5_CQE_LRO_PUSH_MASK) tcp->tcp_flags |= RTE_TCP_PSH_FLAG; tcp->cksum = 0; - csum += rte_raw_cksum(tcp, (tcp->data_off & 0xF) * 4); + csum += rte_raw_cksum(tcp, (tcp->data_off >> 4) * 4); csum = ((csum & 0xffff0000) >> 16) + (csum & 0xffff); csum = (~csum) & 0xffff; if (csum == 0) @@ -1547,8 +1508,8 @@ mlx5_lro_update_tcp_hdr(struct rte_tcp_hdr *restrict tcp, * The packet length. */ static inline void -mlx5_lro_update_hdr(uint8_t *restrict padd, - volatile struct mlx5_cqe *restrict cqe, +mlx5_lro_update_hdr(uint8_t *__rte_restrict padd, + volatile struct mlx5_cqe *__rte_restrict cqe, uint32_t len) { union { @@ -1592,10 +1553,11 @@ mlx5_mprq_buf_free_cb(void *addr __rte_unused, void *opaque) { struct mlx5_mprq_buf *buf = opaque; - if (rte_atomic16_read(&buf->refcnt) == 1) { + if (__atomic_load_n(&buf->refcnt, __ATOMIC_RELAXED) == 1) { rte_mempool_put(buf->mp, buf); - } else if (rte_atomic16_add_return(&buf->refcnt, -1) == 0) { - rte_atomic16_set(&buf->refcnt, 1); + } else if (unlikely(__atomic_sub_fetch(&buf->refcnt, 1, + __ATOMIC_RELAXED) == 0)) { + __atomic_store_n(&buf->refcnt, 1, __ATOMIC_RELAXED); rte_mempool_put(buf->mp, buf); } } @@ -1606,31 +1568,6 @@ mlx5_mprq_buf_free(struct mlx5_mprq_buf *buf) mlx5_mprq_buf_free_cb(NULL, buf); } -static inline void -mprq_buf_replace(struct mlx5_rxq_data *rxq, uint16_t rq_idx, - const unsigned int strd_n) -{ - struct mlx5_mprq_buf *rep = rxq->mprq_repl; - volatile struct mlx5_wqe_data_seg *wqe = - &((volatile struct mlx5_wqe_mprq *)rxq->wqes)[rq_idx].dseg; - void *addr; - - MLX5_ASSERT(rep != NULL); - /* Replace MPRQ buf. */ - (*rxq->mprq_bufs)[rq_idx] = rep; - /* Replace WQE. */ - addr = mlx5_mprq_buf_addr(rep, strd_n); - wqe->addr = rte_cpu_to_be_64((uintptr_t)addr); - /* If there's only one MR, no need to replace LKey in WQE. */ - if (unlikely(mlx5_mr_btree_len(&rxq->mr_ctrl.cache_bh) > 1)) - wqe->lkey = mlx5_rx_addr2mr(rxq, (uintptr_t)addr); - /* Stash a mbuf for next replacement. */ - if (likely(!rte_mempool_get(rxq->mprq_mp, (void **)&rep))) - rxq->mprq_repl = rep; - else - rxq->mprq_repl = NULL; -} - /** * DPDK callback for RX with Multi-Packet RQ support. * @@ -1648,51 +1585,29 @@ uint16_t mlx5_rx_burst_mprq(void *dpdk_rxq, struct rte_mbuf **pkts, uint16_t pkts_n) { struct mlx5_rxq_data *rxq = dpdk_rxq; - const unsigned int strd_n = 1 << rxq->strd_num_n; - const unsigned int strd_sz = 1 << rxq->strd_sz_n; - const unsigned int strd_shift = - MLX5_MPRQ_STRIDE_SHIFT_BYTE * rxq->strd_shift_en; - const unsigned int cq_mask = (1 << rxq->cqe_n) - 1; - const unsigned int wq_mask = (1 << rxq->elts_n) - 1; + const uint32_t strd_n = 1 << rxq->strd_num_n; + const uint32_t cq_mask = (1 << rxq->cqe_n) - 1; + const uint32_t wq_mask = (1 << rxq->elts_n) - 1; volatile struct mlx5_cqe *cqe = &(*rxq->cqes)[rxq->cq_ci & cq_mask]; unsigned int i = 0; uint32_t rq_ci = rxq->rq_ci; uint16_t consumed_strd = rxq->consumed_strd; - uint16_t headroom_sz = rxq->strd_headroom_en * RTE_PKTMBUF_HEADROOM; struct mlx5_mprq_buf *buf = (*rxq->mprq_bufs)[rq_ci & wq_mask]; while (i < pkts_n) { struct rte_mbuf *pkt; - void *addr; int ret; - unsigned int len; + uint32_t len; uint16_t strd_cnt; uint16_t strd_idx; - uint32_t offset; uint32_t byte_cnt; volatile struct mlx5_mini_cqe8 *mcqe = NULL; uint32_t rss_hash_res = 0; - uint8_t lro_num_seg; + enum mlx5_rqx_code rxq_code; if (consumed_strd == strd_n) { - /* Replace WQE only if the buffer is still in use. */ - if (rte_atomic16_read(&buf->refcnt) > 1) { - mprq_buf_replace(rxq, rq_ci & wq_mask, strd_n); - /* Release the old buffer. */ - mlx5_mprq_buf_free(buf); - } else if (unlikely(rxq->mprq_repl == NULL)) { - struct mlx5_mprq_buf *rep; - - /* - * Currently, the MPRQ mempool is out of buffer - * and doing memcpy regardless of the size of Rx - * packet. Retry allocation to get back to - * normal. - */ - if (!rte_mempool_get(rxq->mprq_mp, - (void **)&rep)) - rxq->mprq_repl = rep; - } + /* Replace WQE if the buffer is still in use. */ + mprq_buf_replace(rxq, rq_ci & wq_mask); /* Advance to the next WQE. */ consumed_strd = 0; ++rq_ci; @@ -1719,18 +1634,6 @@ mlx5_rx_burst_mprq(void *dpdk_rxq, struct rte_mbuf **pkts, uint16_t pkts_n) MLX5_ASSERT(strd_idx < strd_n); MLX5_ASSERT(!((rte_be_to_cpu_16(cqe->wqe_id) ^ rq_ci) & wq_mask)); - lro_num_seg = cqe->lro_num_seg; - /* - * Currently configured to receive a packet per a stride. But if - * MTU is adjusted through kernel interface, device could - * consume multiple strides without raising an error. In this - * case, the packet should be dropped because it is bigger than - * the max_rx_pkt_len. - */ - if (unlikely(!lro_num_seg && strd_cnt > 1)) { - ++rxq->stats.idropped; - continue; - } pkt = rte_pktmbuf_alloc(rxq->mp); if (unlikely(pkt == NULL)) { ++rxq->stats.rx_nombuf; @@ -1740,91 +1643,25 @@ mlx5_rx_burst_mprq(void *dpdk_rxq, struct rte_mbuf **pkts, uint16_t pkts_n) MLX5_ASSERT((int)len >= (rxq->crc_present << 2)); if (rxq->crc_present) len -= RTE_ETHER_CRC_LEN; - offset = strd_idx * strd_sz + strd_shift; - addr = RTE_PTR_ADD(mlx5_mprq_buf_addr(buf, strd_n), offset); - /* - * Memcpy packets to the target mbuf if: - * - The size of packet is smaller than mprq_max_memcpy_len. - * - Out of buffer in the Mempool for Multi-Packet RQ. - */ - if (len <= rxq->mprq_max_memcpy_len || rxq->mprq_repl == NULL) { - /* - * When memcpy'ing packet due to out-of-buffer, the - * packet must be smaller than the target mbuf. - */ - if (unlikely(rte_pktmbuf_tailroom(pkt) < len)) { - rte_pktmbuf_free_seg(pkt); + rxq_code = mprq_buf_to_pkt(rxq, pkt, len, buf, + strd_idx, strd_cnt); + if (unlikely(rxq_code != MLX5_RXQ_CODE_EXIT)) { + rte_pktmbuf_free_seg(pkt); + if (rxq_code == MLX5_RXQ_CODE_DROPPED) { ++rxq->stats.idropped; continue; } - rte_memcpy(rte_pktmbuf_mtod(pkt, void *), addr, len); - DATA_LEN(pkt) = len; - } else { - rte_iova_t buf_iova; - struct rte_mbuf_ext_shared_info *shinfo; - uint16_t buf_len = strd_cnt * strd_sz; - void *buf_addr; - - /* Increment the refcnt of the whole chunk. */ - rte_atomic16_add_return(&buf->refcnt, 1); - MLX5_ASSERT((uint16_t)rte_atomic16_read(&buf->refcnt) <= - strd_n + 1); - buf_addr = RTE_PTR_SUB(addr, headroom_sz); - /* - * MLX5 device doesn't use iova but it is necessary in a - * case where the Rx packet is transmitted via a - * different PMD. - */ - buf_iova = rte_mempool_virt2iova(buf) + - RTE_PTR_DIFF(buf_addr, buf); - shinfo = &buf->shinfos[strd_idx]; - rte_mbuf_ext_refcnt_set(shinfo, 1); - /* - * EXT_ATTACHED_MBUF will be set to pkt->ol_flags when - * attaching the stride to mbuf and more offload flags - * will be added below by calling rxq_cq_to_mbuf(). - * Other fields will be overwritten. - */ - rte_pktmbuf_attach_extbuf(pkt, buf_addr, buf_iova, - buf_len, shinfo); - /* Set mbuf head-room. */ - pkt->data_off = headroom_sz; - MLX5_ASSERT(pkt->ol_flags == EXT_ATTACHED_MBUF); - /* - * Prevent potential overflow due to MTU change through - * kernel interface. - */ - if (unlikely(rte_pktmbuf_tailroom(pkt) < len)) { - rte_pktmbuf_free_seg(pkt); - ++rxq->stats.idropped; - continue; - } - DATA_LEN(pkt) = len; - /* - * LRO packet may consume all the stride memory, in this - * case packet head-room space is not guaranteed so must - * to add an empty mbuf for the head-room. - */ - if (!rxq->strd_headroom_en) { - struct rte_mbuf *headroom_mbuf = - rte_pktmbuf_alloc(rxq->mp); - - if (unlikely(headroom_mbuf == NULL)) { - rte_pktmbuf_free_seg(pkt); - ++rxq->stats.rx_nombuf; - break; - } - PORT(pkt) = rxq->port_id; - NEXT(headroom_mbuf) = pkt; - pkt = headroom_mbuf; - NB_SEGS(pkt) = 2; + if (rxq_code == MLX5_RXQ_CODE_NOMBUF) { + ++rxq->stats.rx_nombuf; + break; } } rxq_cq_to_mbuf(rxq, pkt, cqe, rss_hash_res); - if (lro_num_seg > 1) { - mlx5_lro_update_hdr(addr, cqe, len); + if (cqe->lro_num_seg > 1) { + mlx5_lro_update_hdr(rte_pktmbuf_mtod(pkt, uint8_t *), + cqe, len); pkt->ol_flags |= PKT_RX_LRO; - pkt->tso_segsz = strd_sz; + pkt->tso_segsz = len / cqe->lro_num_seg; } PKT_LEN(pkt) = len; PORT(pkt) = rxq->port_id; @@ -1838,11 +1675,11 @@ mlx5_rx_burst_mprq(void *dpdk_rxq, struct rte_mbuf **pkts, uint16_t pkts_n) } /* Update the consumer indexes. */ rxq->consumed_strd = consumed_strd; - rte_cio_wmb(); + rte_io_wmb(); *rxq->cq_db = rte_cpu_to_be_32(rxq->cq_ci); if (rq_ci != rxq->rq_ci) { rxq->rq_ci = rq_ci; - rte_cio_wmb(); + rte_io_wmb(); *rxq->rq_db = rte_cpu_to_be_32(rxq->rq_ci); } #ifdef MLX5_PMD_SOFT_COUNTERS @@ -1917,6 +1754,14 @@ mlx5_rx_burst_vec(void *dpdk_txq __rte_unused, return 0; } +__rte_weak uint16_t +mlx5_rx_burst_mprq_vec(void *dpdk_txq __rte_unused, + struct rte_mbuf **pkts __rte_unused, + uint16_t pkts_n __rte_unused) +{ + return 0; +} + __rte_weak int mlx5_rxq_check_vec_support(struct mlx5_rxq_data *rxq __rte_unused) { @@ -1941,7 +1786,7 @@ mlx5_check_vec_rx_support(struct rte_eth_dev *dev __rte_unused) * compile time and may be used for optimization. */ static __rte_always_inline void -mlx5_tx_free_mbuf(struct rte_mbuf **restrict pkts, +mlx5_tx_free_mbuf(struct rte_mbuf **__rte_restrict pkts, unsigned int pkts_n, unsigned int olx __rte_unused) { @@ -2046,7 +1891,7 @@ mlx5_tx_free_mbuf(struct rte_mbuf **restrict pkts, * compile time and may be used for optimization. */ static __rte_always_inline void -mlx5_tx_free_elts(struct mlx5_txq_data *restrict txq, +mlx5_tx_free_elts(struct mlx5_txq_data *__rte_restrict txq, uint16_t tail, unsigned int olx __rte_unused) { @@ -2087,8 +1932,8 @@ mlx5_tx_free_elts(struct mlx5_txq_data *restrict txq, * compile time and may be used for optimization. */ static __rte_always_inline void -mlx5_tx_copy_elts(struct mlx5_txq_data *restrict txq, - struct rte_mbuf **restrict pkts, +mlx5_tx_copy_elts(struct mlx5_txq_data *__rte_restrict txq, + struct rte_mbuf **__rte_restrict pkts, unsigned int pkts_n, unsigned int olx __rte_unused) { @@ -2124,7 +1969,7 @@ mlx5_tx_copy_elts(struct mlx5_txq_data *restrict txq, * compile time and may be used for optimization. */ static __rte_always_inline void -mlx5_tx_comp_flush(struct mlx5_txq_data *restrict txq, +mlx5_tx_comp_flush(struct mlx5_txq_data *__rte_restrict txq, volatile struct mlx5_cqe *last_cqe, unsigned int olx __rte_unused) { @@ -2155,12 +2000,12 @@ mlx5_tx_comp_flush(struct mlx5_txq_data *restrict txq, * routine smaller, simple and faster - from experiments. */ static void -mlx5_tx_handle_completion(struct mlx5_txq_data *restrict txq, +mlx5_tx_handle_completion(struct mlx5_txq_data *__rte_restrict txq, unsigned int olx __rte_unused) { unsigned int count = MLX5_TX_COMP_MAX_CQE; volatile struct mlx5_cqe *last_cqe = NULL; - uint16_t ci = txq->cq_ci; + bool ring_doorbell = false; int ret; static_assert(MLX5_CQE_STATUS_HW_OWN < 0, "Must be negative value"); @@ -2168,8 +2013,8 @@ mlx5_tx_handle_completion(struct mlx5_txq_data *restrict txq, do { volatile struct mlx5_cqe *cqe; - cqe = &txq->cqes[ci & txq->cqe_m]; - ret = check_cqe(cqe, txq->cqe_s, ci); + cqe = &txq->cqes[txq->cq_ci & txq->cqe_m]; + ret = check_cqe(cqe, txq->cqe_s, txq->cq_ci); if (unlikely(ret != MLX5_CQE_STATUS_SW_OWN)) { if (likely(ret != MLX5_CQE_STATUS_ERR)) { /* No new CQEs in completion queue. */ @@ -2183,7 +2028,6 @@ mlx5_tx_handle_completion(struct mlx5_txq_data *restrict txq, * here, before we might perform SQ reset. */ rte_wmb(); - txq->cq_ci = ci; ret = mlx5_tx_error_cqe_handle (txq, (volatile struct mlx5_err_cqe *)cqe); if (unlikely(ret < 0)) { @@ -2199,16 +2043,18 @@ mlx5_tx_handle_completion(struct mlx5_txq_data *restrict txq, * MLX5_CQE_SYNDROME_WR_FLUSH_ERR status. * The send queue is supposed to be empty. */ - ++ci; - txq->cq_pi = ci; + ring_doorbell = true; + ++txq->cq_ci; + txq->cq_pi = txq->cq_ci; last_cqe = NULL; continue; } /* Normal transmit completion. */ - MLX5_ASSERT(ci != txq->cq_pi); - MLX5_ASSERT((txq->fcqs[ci & txq->cqe_m] >> 16) == + MLX5_ASSERT(txq->cq_ci != txq->cq_pi); + MLX5_ASSERT((txq->fcqs[txq->cq_ci & txq->cqe_m] >> 16) == cqe->wqe_counter); - ++ci; + ring_doorbell = true; + ++txq->cq_ci; last_cqe = cqe; /* * We have to restrict the amount of processed CQEs @@ -2221,14 +2067,10 @@ mlx5_tx_handle_completion(struct mlx5_txq_data *restrict txq, if (likely(--count == 0)) break; } while (true); - if (likely(ci != txq->cq_ci)) { - /* - * Update completion queue consuming index - * and ring doorbell to notify hardware. - */ + if (likely(ring_doorbell)) { + /* Ring doorbell to notify hardware. */ rte_compiler_barrier(); - txq->cq_ci = ci; - *txq->cq_db = rte_cpu_to_be_32(ci); + *txq->cq_db = rte_cpu_to_be_32(txq->cq_ci); mlx5_tx_comp_flush(txq, last_cqe, olx); } } @@ -2247,8 +2089,8 @@ mlx5_tx_handle_completion(struct mlx5_txq_data *restrict txq, * compile time and may be used for optimization. */ static __rte_always_inline void -mlx5_tx_request_completion(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc, +mlx5_tx_request_completion(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc, unsigned int olx) { uint16_t head = txq->elts_head; @@ -2262,6 +2104,7 @@ mlx5_tx_request_completion(struct mlx5_txq_data *restrict txq, (uint16_t)(txq->wqe_ci - txq->wqe_comp) >= txq->wqe_thres)) { volatile struct mlx5_wqe *last = loc->wqe_last; + MLX5_ASSERT(last); txq->elts_comp = head; if (MLX5_TXOFF_CONFIG(INLINE)) txq->wqe_comp = txq->wqe_ci; @@ -2294,7 +2137,7 @@ mlx5_tx_request_completion(struct mlx5_txq_data *restrict txq, int mlx5_tx_descriptor_status(void *tx_queue, uint16_t offset) { - struct mlx5_txq_data *restrict txq = tx_queue; + struct mlx5_txq_data *__rte_restrict txq = tx_queue; uint16_t used; mlx5_tx_handle_completion(txq, 0); @@ -2325,14 +2168,14 @@ mlx5_tx_descriptor_status(void *tx_queue, uint16_t offset) * compile time and may be used for optimization. */ static __rte_always_inline void -mlx5_tx_cseg_init(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc __rte_unused, - struct mlx5_wqe *restrict wqe, +mlx5_tx_cseg_init(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc __rte_unused, + struct mlx5_wqe *__rte_restrict wqe, unsigned int ds, unsigned int opcode, unsigned int olx __rte_unused) { - struct mlx5_wqe_cseg *restrict cs = &wqe->cseg; + struct mlx5_wqe_cseg *__rte_restrict cs = &wqe->cseg; /* For legacy MPW replace the EMPW by TSO with modifier. */ if (MLX5_TXOFF_CONFIG(MPW) && opcode == MLX5_OPCODE_ENHANCED_MPSW) @@ -2344,6 +2187,37 @@ mlx5_tx_cseg_init(struct mlx5_txq_data *restrict txq, cs->misc = RTE_BE32(0); } +/** + * Build the Synchronize Queue Segment with specified completion index. + * + * @param txq + * Pointer to TX queue structure. + * @param loc + * Pointer to burst routine local context. + * @param wqe + * Pointer to WQE to fill with built Control Segment. + * @param wci + * Completion index in Clock Queue to wait. + * @param olx + * Configured Tx offloads mask. It is fully defined at + * compile time and may be used for optimization. + */ +static __rte_always_inline void +mlx5_tx_wseg_init(struct mlx5_txq_data *restrict txq, + struct mlx5_txq_local *restrict loc __rte_unused, + struct mlx5_wqe *restrict wqe, + unsigned int wci, + unsigned int olx __rte_unused) +{ + struct mlx5_wqe_qseg *qs; + + qs = RTE_PTR_ADD(wqe, MLX5_WSEG_SIZE); + qs->max_index = rte_cpu_to_be_32(wci); + qs->qpn_cqn = rte_cpu_to_be_32(txq->sh->txpp.clock_queue.cq->id); + qs->reserved0 = RTE_BE32(0); + qs->reserved1 = RTE_BE32(0); +} + /** * Build the Ethernet Segment without inlined data. * Supports Software Parser, Checksums and VLAN @@ -2360,12 +2234,12 @@ mlx5_tx_cseg_init(struct mlx5_txq_data *restrict txq, * compile time and may be used for optimization. */ static __rte_always_inline void -mlx5_tx_eseg_none(struct mlx5_txq_data *restrict txq __rte_unused, - struct mlx5_txq_local *restrict loc, - struct mlx5_wqe *restrict wqe, +mlx5_tx_eseg_none(struct mlx5_txq_data *__rte_restrict txq __rte_unused, + struct mlx5_txq_local *__rte_restrict loc, + struct mlx5_wqe *__rte_restrict wqe, unsigned int olx) { - struct mlx5_wqe_eseg *restrict es = &wqe->eseg; + struct mlx5_wqe_eseg *__rte_restrict es = &wqe->eseg; uint32_t csum; /* @@ -2418,13 +2292,13 @@ mlx5_tx_eseg_none(struct mlx5_txq_data *restrict txq __rte_unused, * compile time and may be used for optimization. */ static __rte_always_inline void -mlx5_tx_eseg_dmin(struct mlx5_txq_data *restrict txq __rte_unused, - struct mlx5_txq_local *restrict loc, - struct mlx5_wqe *restrict wqe, +mlx5_tx_eseg_dmin(struct mlx5_txq_data *__rte_restrict txq __rte_unused, + struct mlx5_txq_local *__rte_restrict loc, + struct mlx5_wqe *__rte_restrict wqe, unsigned int vlan, unsigned int olx) { - struct mlx5_wqe_eseg *restrict es = &wqe->eseg; + struct mlx5_wqe_eseg *__rte_restrict es = &wqe->eseg; uint32_t csum; uint8_t *psrc, *pdst; @@ -2502,15 +2376,15 @@ mlx5_tx_eseg_dmin(struct mlx5_txq_data *restrict txq __rte_unused, * Pointer to the next Data Segment (aligned and wrapped around). */ static __rte_always_inline struct mlx5_wqe_dseg * -mlx5_tx_eseg_data(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc, - struct mlx5_wqe *restrict wqe, +mlx5_tx_eseg_data(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc, + struct mlx5_wqe *__rte_restrict wqe, unsigned int vlan, unsigned int inlen, unsigned int tso, unsigned int olx) { - struct mlx5_wqe_eseg *restrict es = &wqe->eseg; + struct mlx5_wqe_eseg *__rte_restrict es = &wqe->eseg; uint32_t csum; uint8_t *psrc, *pdst; unsigned int part; @@ -2628,7 +2502,7 @@ mlx5_tx_eseg_data(struct mlx5_txq_data *restrict txq, */ static __rte_always_inline unsigned int mlx5_tx_mseg_memcpy(uint8_t *pdst, - struct mlx5_txq_local *restrict loc, + struct mlx5_txq_local *__rte_restrict loc, unsigned int len, unsigned int must, unsigned int olx __rte_unused) @@ -2725,15 +2599,15 @@ mlx5_tx_mseg_memcpy(uint8_t *pdst, * wrapping check on its own). */ static __rte_always_inline struct mlx5_wqe_dseg * -mlx5_tx_eseg_mdat(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc, - struct mlx5_wqe *restrict wqe, +mlx5_tx_eseg_mdat(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc, + struct mlx5_wqe *__rte_restrict wqe, unsigned int vlan, unsigned int inlen, unsigned int tso, unsigned int olx) { - struct mlx5_wqe_eseg *restrict es = &wqe->eseg; + struct mlx5_wqe_eseg *__rte_restrict es = &wqe->eseg; uint32_t csum; uint8_t *pdst; unsigned int part, tlen = 0; @@ -2829,9 +2703,9 @@ mlx5_tx_eseg_mdat(struct mlx5_txq_data *restrict txq, * compile time and may be used for optimization. */ static __rte_always_inline void -mlx5_tx_dseg_ptr(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc, - struct mlx5_wqe_dseg *restrict dseg, +mlx5_tx_dseg_ptr(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc, + struct mlx5_wqe_dseg *__rte_restrict dseg, uint8_t *buf, unsigned int len, unsigned int olx __rte_unused) @@ -2863,9 +2737,9 @@ mlx5_tx_dseg_ptr(struct mlx5_txq_data *restrict txq, * compile time and may be used for optimization. */ static __rte_always_inline void -mlx5_tx_dseg_iptr(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc, - struct mlx5_wqe_dseg *restrict dseg, +mlx5_tx_dseg_iptr(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc, + struct mlx5_wqe_dseg *__rte_restrict dseg, uint8_t *buf, unsigned int len, unsigned int olx __rte_unused) @@ -2939,9 +2813,9 @@ mlx5_tx_dseg_iptr(struct mlx5_txq_data *restrict txq, * last packet in the eMPW session. */ static __rte_always_inline struct mlx5_wqe_dseg * -mlx5_tx_dseg_empw(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc __rte_unused, - struct mlx5_wqe_dseg *restrict dseg, +mlx5_tx_dseg_empw(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc __rte_unused, + struct mlx5_wqe_dseg *__rte_restrict dseg, uint8_t *buf, unsigned int len, unsigned int olx __rte_unused) @@ -3002,9 +2876,9 @@ mlx5_tx_dseg_empw(struct mlx5_txq_data *restrict txq, * Ring buffer wraparound check is needed. */ static __rte_always_inline struct mlx5_wqe_dseg * -mlx5_tx_dseg_vlan(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc __rte_unused, - struct mlx5_wqe_dseg *restrict dseg, +mlx5_tx_dseg_vlan(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc __rte_unused, + struct mlx5_wqe_dseg *__rte_restrict dseg, uint8_t *buf, unsigned int len, unsigned int olx __rte_unused) @@ -3090,15 +2964,15 @@ mlx5_tx_dseg_vlan(struct mlx5_txq_data *restrict txq, * Actual size of built WQE in segments. */ static __rte_always_inline unsigned int -mlx5_tx_mseg_build(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc, - struct mlx5_wqe *restrict wqe, +mlx5_tx_mseg_build(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc, + struct mlx5_wqe *__rte_restrict wqe, unsigned int vlan, unsigned int inlen, unsigned int tso, unsigned int olx __rte_unused) { - struct mlx5_wqe_dseg *restrict dseg; + struct mlx5_wqe_dseg *__rte_restrict dseg; unsigned int ds; MLX5_ASSERT((rte_pktmbuf_pkt_len(loc->mbuf) + vlan) >= inlen); @@ -3181,6 +3055,59 @@ dseg_done: return ds; } +/** + * The routine checks timestamp flag in the current packet, + * and push WAIT WQE into the queue if scheduling is required. + * + * @param txq + * Pointer to TX queue structure. + * @param loc + * Pointer to burst routine local context. + * @param olx + * Configured Tx offloads mask. It is fully defined at + * compile time and may be used for optimization. + * + * @return + * MLX5_TXCMP_CODE_EXIT - sending is done or impossible. + * MLX5_TXCMP_CODE_SINGLE - continue processing with the packet. + * MLX5_TXCMP_CODE_MULTI - the WAIT inserted, continue processing. + * Local context variables partially updated. + */ +static __rte_always_inline enum mlx5_txcmp_code +mlx5_tx_schedule_send(struct mlx5_txq_data *restrict txq, + struct mlx5_txq_local *restrict loc, + unsigned int olx) +{ + if (MLX5_TXOFF_CONFIG(TXPP) && + loc->mbuf->ol_flags & txq->ts_mask) { + struct mlx5_wqe *wqe; + uint64_t ts; + int32_t wci; + + /* + * Estimate the required space quickly and roughly. + * We would like to ensure the packet can be pushed + * to the queue and we won't get the orphan WAIT WQE. + */ + if (loc->wqe_free <= MLX5_WQE_SIZE_MAX / MLX5_WQE_SIZE || + loc->elts_free < NB_SEGS(loc->mbuf)) + return MLX5_TXCMP_CODE_EXIT; + /* Convert the timestamp into completion to wait. */ + ts = *RTE_MBUF_DYNFIELD(loc->mbuf, txq->ts_offset, uint64_t *); + wci = mlx5_txpp_convert_tx_ts(txq->sh, ts); + if (unlikely(wci < 0)) + return MLX5_TXCMP_CODE_SINGLE; + /* Build the WAIT WQE with specified completion. */ + wqe = txq->wqes + (txq->wqe_ci & txq->wqe_m); + mlx5_tx_cseg_init(txq, loc, wqe, 2, MLX5_OPCODE_WAIT, olx); + mlx5_tx_wseg_init(txq, loc, wqe, wci, olx); + ++txq->wqe_ci; + --loc->wqe_free; + return MLX5_TXCMP_CODE_MULTI; + } + return MLX5_TXCMP_CODE_SINGLE; +} + /** * Tx one packet function for multi-segment TSO. Supports all * types of Tx offloads, uses MLX5_OPCODE_TSO to build WQEs, @@ -3203,13 +3130,23 @@ dseg_done: * Local context variables partially updated. */ static __rte_always_inline enum mlx5_txcmp_code -mlx5_tx_packet_multi_tso(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc, +mlx5_tx_packet_multi_tso(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc, unsigned int olx) { - struct mlx5_wqe *restrict wqe; + struct mlx5_wqe *__rte_restrict wqe; unsigned int ds, dlen, inlen, ntcp, vlan = 0; + if (MLX5_TXOFF_CONFIG(TXPP)) { + enum mlx5_txcmp_code wret; + + /* Generate WAIT for scheduling if requested. */ + wret = mlx5_tx_schedule_send(txq, loc, olx); + if (wret == MLX5_TXCMP_CODE_EXIT) + return MLX5_TXCMP_CODE_EXIT; + if (wret == MLX5_TXCMP_CODE_ERROR) + return MLX5_TXCMP_CODE_ERROR; + } /* * Calculate data length to be inlined to estimate * the required space in WQE ring buffer. @@ -3292,15 +3229,25 @@ mlx5_tx_packet_multi_tso(struct mlx5_txq_data *restrict txq, * Local context variables partially updated. */ static __rte_always_inline enum mlx5_txcmp_code -mlx5_tx_packet_multi_send(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc, +mlx5_tx_packet_multi_send(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc, unsigned int olx) { - struct mlx5_wqe_dseg *restrict dseg; - struct mlx5_wqe *restrict wqe; + struct mlx5_wqe_dseg *__rte_restrict dseg; + struct mlx5_wqe *__rte_restrict wqe; unsigned int ds, nseg; MLX5_ASSERT(NB_SEGS(loc->mbuf) > 1); + if (MLX5_TXOFF_CONFIG(TXPP)) { + enum mlx5_txcmp_code wret; + + /* Generate WAIT for scheduling if requested. */ + wret = mlx5_tx_schedule_send(txq, loc, olx); + if (wret == MLX5_TXCMP_CODE_EXIT) + return MLX5_TXCMP_CODE_EXIT; + if (wret == MLX5_TXCMP_CODE_ERROR) + return MLX5_TXCMP_CODE_ERROR; + } /* * No inline at all, it means the CPU cycles saving * is prioritized at configuration, we should not @@ -3400,15 +3347,25 @@ mlx5_tx_packet_multi_send(struct mlx5_txq_data *restrict txq, * Local context variables partially updated. */ static __rte_always_inline enum mlx5_txcmp_code -mlx5_tx_packet_multi_inline(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc, +mlx5_tx_packet_multi_inline(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc, unsigned int olx) { - struct mlx5_wqe *restrict wqe; + struct mlx5_wqe *__rte_restrict wqe; unsigned int ds, inlen, dlen, vlan = 0; MLX5_ASSERT(MLX5_TXOFF_CONFIG(INLINE)); MLX5_ASSERT(NB_SEGS(loc->mbuf) > 1); + if (MLX5_TXOFF_CONFIG(TXPP)) { + enum mlx5_txcmp_code wret; + + /* Generate WAIT for scheduling if requested. */ + wret = mlx5_tx_schedule_send(txq, loc, olx); + if (wret == MLX5_TXCMP_CODE_EXIT) + return MLX5_TXCMP_CODE_EXIT; + if (wret == MLX5_TXCMP_CODE_ERROR) + return MLX5_TXCMP_CODE_ERROR; + } /* * First calculate data length to be inlined * to estimate the required space for WQE. @@ -3565,10 +3522,10 @@ do_align: * Local context variables updated. */ static __rte_always_inline enum mlx5_txcmp_code -mlx5_tx_burst_mseg(struct mlx5_txq_data *restrict txq, - struct rte_mbuf **restrict pkts, +mlx5_tx_burst_mseg(struct mlx5_txq_data *__rte_restrict txq, + struct rte_mbuf **__rte_restrict pkts, unsigned int pkts_n, - struct mlx5_txq_local *restrict loc, + struct mlx5_txq_local *__rte_restrict loc, unsigned int olx) { MLX5_ASSERT(loc->elts_free && loc->wqe_free); @@ -3654,10 +3611,10 @@ mlx5_tx_burst_mseg(struct mlx5_txq_data *restrict txq, * Local context variables updated. */ static __rte_always_inline enum mlx5_txcmp_code -mlx5_tx_burst_tso(struct mlx5_txq_data *restrict txq, - struct rte_mbuf **restrict pkts, +mlx5_tx_burst_tso(struct mlx5_txq_data *__rte_restrict txq, + struct rte_mbuf **__rte_restrict pkts, unsigned int pkts_n, - struct mlx5_txq_local *restrict loc, + struct mlx5_txq_local *__rte_restrict loc, unsigned int olx) { MLX5_ASSERT(loc->elts_free && loc->wqe_free); @@ -3665,12 +3622,22 @@ mlx5_tx_burst_tso(struct mlx5_txq_data *restrict txq, pkts += loc->pkts_sent + 1; pkts_n -= loc->pkts_sent; for (;;) { - struct mlx5_wqe_dseg *restrict dseg; - struct mlx5_wqe *restrict wqe; + struct mlx5_wqe_dseg *__rte_restrict dseg; + struct mlx5_wqe *__rte_restrict wqe; unsigned int ds, dlen, hlen, ntcp, vlan = 0; uint8_t *dptr; MLX5_ASSERT(NB_SEGS(loc->mbuf) == 1); + if (MLX5_TXOFF_CONFIG(TXPP)) { + enum mlx5_txcmp_code wret; + + /* Generate WAIT for scheduling if requested. */ + wret = mlx5_tx_schedule_send(txq, loc, olx); + if (wret == MLX5_TXCMP_CODE_EXIT) + return MLX5_TXCMP_CODE_EXIT; + if (wret == MLX5_TXCMP_CODE_ERROR) + return MLX5_TXCMP_CODE_ERROR; + } dlen = rte_pktmbuf_data_len(loc->mbuf); if (MLX5_TXOFF_CONFIG(VLAN) && loc->mbuf->ol_flags & PKT_TX_VLAN_PKT) { @@ -3778,8 +3745,8 @@ mlx5_tx_burst_tso(struct mlx5_txq_data *restrict txq, * MLX5_TXCMP_CODE_EMPW - single-segment packet, use MPW. */ static __rte_always_inline enum mlx5_txcmp_code -mlx5_tx_able_to_empw(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc, +mlx5_tx_able_to_empw(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc, unsigned int olx, bool newp) { @@ -3833,9 +3800,9 @@ mlx5_tx_able_to_empw(struct mlx5_txq_data *restrict txq, * false - no match, eMPW should be restarted. */ static __rte_always_inline bool -mlx5_tx_match_empw(struct mlx5_txq_data *restrict txq __rte_unused, - struct mlx5_wqe_eseg *restrict es, - struct mlx5_txq_local *restrict loc, +mlx5_tx_match_empw(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_wqe_eseg *__rte_restrict es, + struct mlx5_txq_local *__rte_restrict loc, uint32_t dlen, unsigned int olx) { @@ -3862,6 +3829,10 @@ mlx5_tx_match_empw(struct mlx5_txq_data *restrict txq __rte_unused, /* There must be no VLAN packets in eMPW loop. */ if (MLX5_TXOFF_CONFIG(VLAN)) MLX5_ASSERT(!(loc->mbuf->ol_flags & PKT_TX_VLAN_PKT)); + /* Check if the scheduling is requested. */ + if (MLX5_TXOFF_CONFIG(TXPP) && + loc->mbuf->ol_flags & txq->ts_mask) + return false; return true; } @@ -3887,8 +3858,8 @@ mlx5_tx_match_empw(struct mlx5_txq_data *restrict txq __rte_unused, * false - no match, eMPW should be restarted. */ static __rte_always_inline void -mlx5_tx_sdone_empw(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc, +mlx5_tx_sdone_empw(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc, unsigned int ds, unsigned int slen, unsigned int olx __rte_unused) @@ -3921,6 +3892,8 @@ mlx5_tx_sdone_empw(struct mlx5_txq_data *restrict txq, * Total size of descriptor/data in bytes. * @param slen * Accumulated statistics, data bytes sent. + * @param wqem + * The base WQE for the eMPW/MPW descriptor. * @param olx * Configured Tx offloads mask. It is fully defined at * compile time and may be used for optimization. @@ -3930,13 +3903,14 @@ mlx5_tx_sdone_empw(struct mlx5_txq_data *restrict txq, * false - no match, eMPW should be restarted. */ static __rte_always_inline void -mlx5_tx_idone_empw(struct mlx5_txq_data *restrict txq, - struct mlx5_txq_local *restrict loc, +mlx5_tx_idone_empw(struct mlx5_txq_data *__rte_restrict txq, + struct mlx5_txq_local *__rte_restrict loc, unsigned int len, unsigned int slen, + struct mlx5_wqe *__rte_restrict wqem, unsigned int olx __rte_unused) { - struct mlx5_wqe_dseg *dseg = &loc->wqe_last->dseg[0]; + struct mlx5_wqe_dseg *dseg = &wqem->dseg[0]; MLX5_ASSERT(MLX5_TXOFF_CONFIG(INLINE)); #ifdef MLX5_PMD_SOFT_COUNTERS @@ -3963,9 +3937,10 @@ mlx5_tx_idone_empw(struct mlx5_txq_data *restrict txq, MLX5_ASSERT((len % MLX5_WSEG_SIZE) == 0); len = len / MLX5_WSEG_SIZE + 2; } - loc->wqe_last->cseg.sq_ds = rte_cpu_to_be_32(txq->qp_num_8s | len); + wqem->cseg.sq_ds = rte_cpu_to_be_32(txq->qp_num_8s | len); txq->wqe_ci += (len + 3) / 4; loc->wqe_free -= (len + 3) / 4; + loc->wqe_last = wqem; } /** @@ -4016,10 +3991,10 @@ mlx5_tx_idone_empw(struct mlx5_txq_data *restrict txq, * No VLAN insertion is supported. */ static __rte_always_inline enum mlx5_txcmp_code -mlx5_tx_burst_empw_simple(struct mlx5_txq_data *restrict txq, - struct rte_mbuf **restrict pkts, +mlx5_tx_burst_empw_simple(struct mlx5_txq_data *__rte_restrict txq, + struct rte_mbuf **__rte_restrict pkts, unsigned int pkts_n, - struct mlx5_txq_local *restrict loc, + struct mlx5_txq_local *__rte_restrict loc, unsigned int olx) { /* @@ -4035,14 +4010,24 @@ mlx5_tx_burst_empw_simple(struct mlx5_txq_data *restrict txq, pkts += loc->pkts_sent + 1; pkts_n -= loc->pkts_sent; for (;;) { - struct mlx5_wqe_dseg *restrict dseg; - struct mlx5_wqe_eseg *restrict eseg; + struct mlx5_wqe_dseg *__rte_restrict dseg; + struct mlx5_wqe_eseg *__rte_restrict eseg; enum mlx5_txcmp_code ret; unsigned int part, loop; unsigned int slen = 0; next_empw: MLX5_ASSERT(NB_SEGS(loc->mbuf) == 1); + if (MLX5_TXOFF_CONFIG(TXPP)) { + enum mlx5_txcmp_code wret; + + /* Generate WAIT for scheduling if requested. */ + wret = mlx5_tx_schedule_send(txq, loc, olx); + if (wret == MLX5_TXCMP_CODE_EXIT) + return MLX5_TXCMP_CODE_EXIT; + if (wret == MLX5_TXCMP_CODE_ERROR) + return MLX5_TXCMP_CODE_ERROR; + } part = RTE_MIN(pkts_n, MLX5_TXOFF_CONFIG(MPW) ? MLX5_MPW_MAX_PACKETS : MLX5_EMPW_MAX_PACKETS); @@ -4138,6 +4123,7 @@ next_empw: * - metadata value * - software parser settings * - packets length (legacy MPW only) + * - scheduling is not required */ if (!mlx5_tx_match_empw(txq, eseg, loc, dlen, olx)) { MLX5_ASSERT(loop); @@ -4182,10 +4168,10 @@ next_empw: * with inlining, optionally supports VLAN insertion. */ static __rte_always_inline enum mlx5_txcmp_code -mlx5_tx_burst_empw_inline(struct mlx5_txq_data *restrict txq, - struct rte_mbuf **restrict pkts, +mlx5_tx_burst_empw_inline(struct mlx5_txq_data *__rte_restrict txq, + struct rte_mbuf **__rte_restrict pkts, unsigned int pkts_n, - struct mlx5_txq_local *restrict loc, + struct mlx5_txq_local *__rte_restrict loc, unsigned int olx) { /* @@ -4201,13 +4187,23 @@ mlx5_tx_burst_empw_inline(struct mlx5_txq_data *restrict txq, pkts += loc->pkts_sent + 1; pkts_n -= loc->pkts_sent; for (;;) { - struct mlx5_wqe_dseg *restrict dseg; - struct mlx5_wqe_eseg *restrict eseg; + struct mlx5_wqe_dseg *__rte_restrict dseg; + struct mlx5_wqe *__rte_restrict wqem; enum mlx5_txcmp_code ret; unsigned int room, part, nlim; unsigned int slen = 0; MLX5_ASSERT(NB_SEGS(loc->mbuf) == 1); + if (MLX5_TXOFF_CONFIG(TXPP)) { + enum mlx5_txcmp_code wret; + + /* Generate WAIT for scheduling if requested. */ + wret = mlx5_tx_schedule_send(txq, loc, olx); + if (wret == MLX5_TXCMP_CODE_EXIT) + return MLX5_TXCMP_CODE_EXIT; + if (wret == MLX5_TXCMP_CODE_ERROR) + return MLX5_TXCMP_CODE_ERROR; + } /* * Limits the amount of packets in one WQE * to improve CQE latency generation. @@ -4221,22 +4217,21 @@ mlx5_tx_burst_empw_inline(struct mlx5_txq_data *restrict txq, return MLX5_TXCMP_CODE_EXIT; if (likely(pkts_n > 1)) rte_prefetch0(*pkts); - loc->wqe_last = txq->wqes + (txq->wqe_ci & txq->wqe_m); + wqem = txq->wqes + (txq->wqe_ci & txq->wqe_m); /* * Build eMPW title WQEBB: * - Control Segment, eMPW opcode, zero DS * - Ethernet Segment, no inline */ - mlx5_tx_cseg_init(txq, loc, loc->wqe_last, 0, + mlx5_tx_cseg_init(txq, loc, wqem, 0, MLX5_OPCODE_ENHANCED_MPSW, olx); - mlx5_tx_eseg_none(txq, loc, loc->wqe_last, + mlx5_tx_eseg_none(txq, loc, wqem, olx & ~MLX5_TXOFF_CONFIG_VLAN); - eseg = &loc->wqe_last->eseg; - dseg = &loc->wqe_last->dseg[0]; + dseg = &wqem->dseg[0]; /* Store the packet length for legacy MPW. */ if (MLX5_TXOFF_CONFIG(MPW)) - eseg->mss = rte_cpu_to_be_16 - (rte_pktmbuf_data_len(loc->mbuf)); + wqem->eseg.mss = rte_cpu_to_be_16 + (rte_pktmbuf_data_len(loc->mbuf)); room = RTE_MIN(MLX5_WQE_SIZE_MAX / MLX5_WQE_SIZE, loc->wqe_free) * MLX5_WQE_SIZE - MLX5_WQE_CSEG_SIZE - @@ -4273,7 +4268,8 @@ mlx5_tx_burst_empw_inline(struct mlx5_txq_data *restrict txq, * We have some successfully built * packet Data Segments to send. */ - mlx5_tx_idone_empw(txq, loc, part, slen, olx); + mlx5_tx_idone_empw(txq, loc, part, + slen, wqem, olx); return MLX5_TXCMP_CODE_ERROR; } /* Inline or not inline - that's the Question. */ @@ -4281,6 +4277,8 @@ mlx5_tx_burst_empw_inline(struct mlx5_txq_data *restrict txq, loc->mbuf->ol_flags & PKT_TX_DYNF_NOINLINE) goto pointer_empw; if (MLX5_TXOFF_CONFIG(MPW)) { + if (dlen > txq->inlen_send) + goto pointer_empw; tlen = dlen; if (part == room) { /* Open new inline MPW session. */ @@ -4293,7 +4291,7 @@ mlx5_tx_burst_empw_inline(struct mlx5_txq_data *restrict txq, * No pointer and inline descriptor * intermix for legacy MPW sessions. */ - if (loc->wqe_last->dseg[0].bcount) + if (wqem->dseg[0].bcount) break; } } else { @@ -4342,7 +4340,7 @@ pointer_empw: */ if (MLX5_TXOFF_CONFIG(MPW) && part != room && - loc->wqe_last->dseg[0].bcount == RTE_BE32(0)) + wqem->dseg[0].bcount == RTE_BE32(0)) break; /* * Not inlinable VLAN packets are @@ -4372,7 +4370,8 @@ next_mbuf: * continue build descriptors. */ part -= room; - mlx5_tx_idone_empw(txq, loc, part, slen, olx); + mlx5_tx_idone_empw(txq, loc, part, + slen, wqem, olx); return MLX5_TXCMP_CODE_EXIT; } loc->mbuf = *pkts++; @@ -4386,7 +4385,8 @@ next_mbuf: */ if (ret == MLX5_TXCMP_CODE_MULTI) { part -= room; - mlx5_tx_idone_empw(txq, loc, part, slen, olx); + mlx5_tx_idone_empw(txq, loc, part, + slen, wqem, olx); if (unlikely(!loc->elts_free || !loc->wqe_free)) return MLX5_TXCMP_CODE_EXIT; @@ -4395,7 +4395,8 @@ next_mbuf: MLX5_ASSERT(NB_SEGS(loc->mbuf) == 1); if (ret == MLX5_TXCMP_CODE_TSO) { part -= room; - mlx5_tx_idone_empw(txq, loc, part, slen, olx); + mlx5_tx_idone_empw(txq, loc, part, + slen, wqem, olx); if (unlikely(!loc->elts_free || !loc->wqe_free)) return MLX5_TXCMP_CODE_EXIT; @@ -4403,7 +4404,8 @@ next_mbuf: } if (ret == MLX5_TXCMP_CODE_SINGLE) { part -= room; - mlx5_tx_idone_empw(txq, loc, part, slen, olx); + mlx5_tx_idone_empw(txq, loc, part, + slen, wqem, olx); if (unlikely(!loc->elts_free || !loc->wqe_free)) return MLX5_TXCMP_CODE_EXIT; @@ -4412,7 +4414,8 @@ next_mbuf: if (ret != MLX5_TXCMP_CODE_EMPW) { MLX5_ASSERT(false); part -= room; - mlx5_tx_idone_empw(txq, loc, part, slen, olx); + mlx5_tx_idone_empw(txq, loc, part, + slen, wqem, olx); return MLX5_TXCMP_CODE_ERROR; } /* Check if we have minimal room left. */ @@ -4426,8 +4429,10 @@ next_mbuf: * - metadata value * - software parser settings * - packets length (legacy MPW only) + * - scheduling is not required */ - if (!mlx5_tx_match_empw(txq, eseg, loc, dlen, olx)) + if (!mlx5_tx_match_empw(txq, &wqem->eseg, + loc, dlen, olx)) break; /* Packet attributes match, continue the same eMPW. */ if ((uintptr_t)dseg >= (uintptr_t)txq->wqes_end) @@ -4441,7 +4446,7 @@ next_mbuf: part -= room; if (unlikely(!part)) return MLX5_TXCMP_CODE_EXIT; - mlx5_tx_idone_empw(txq, loc, part, slen, olx); + mlx5_tx_idone_empw(txq, loc, part, slen, wqem, olx); if (unlikely(!loc->elts_free || !loc->wqe_free)) return MLX5_TXCMP_CODE_EXIT; @@ -4455,10 +4460,10 @@ next_mbuf: * Data inlining and VLAN insertion are supported. */ static __rte_always_inline enum mlx5_txcmp_code -mlx5_tx_burst_single_send(struct mlx5_txq_data *restrict txq, - struct rte_mbuf **restrict pkts, +mlx5_tx_burst_single_send(struct mlx5_txq_data *__rte_restrict txq, + struct rte_mbuf **__rte_restrict pkts, unsigned int pkts_n, - struct mlx5_txq_local *restrict loc, + struct mlx5_txq_local *__rte_restrict loc, unsigned int olx) { /* @@ -4470,10 +4475,20 @@ mlx5_tx_burst_single_send(struct mlx5_txq_data *restrict txq, pkts += loc->pkts_sent + 1; pkts_n -= loc->pkts_sent; for (;;) { - struct mlx5_wqe *restrict wqe; + struct mlx5_wqe *__rte_restrict wqe; enum mlx5_txcmp_code ret; MLX5_ASSERT(NB_SEGS(loc->mbuf) == 1); + if (MLX5_TXOFF_CONFIG(TXPP)) { + enum mlx5_txcmp_code wret; + + /* Generate WAIT for scheduling if requested. */ + wret = mlx5_tx_schedule_send(txq, loc, olx); + if (wret == MLX5_TXCMP_CODE_EXIT) + return MLX5_TXCMP_CODE_EXIT; + if (wret == MLX5_TXCMP_CODE_ERROR) + return MLX5_TXCMP_CODE_ERROR; + } if (MLX5_TXOFF_CONFIG(INLINE)) { unsigned int inlen, vlan = 0; @@ -4514,6 +4529,8 @@ mlx5_tx_burst_single_send(struct mlx5_txq_data *restrict txq, txq->inlen_mode) || (MLX5_TXOFF_CONFIG(MPW) && txq->inlen_mode)) { + if (inlen <= txq->inlen_send) + goto single_inline; /* * The hardware requires the * minimal inline data header. @@ -4530,6 +4547,7 @@ mlx5_tx_burst_single_send(struct mlx5_txq_data *restrict txq, } goto single_no_inline; } +single_inline: /* * Completely inlined packet data WQE: * - Control Segment, SEND opcode @@ -4568,7 +4586,7 @@ mlx5_tx_burst_single_send(struct mlx5_txq_data *restrict txq, * not contain inlined data for eMPW due to * segment shared for all packets. */ - struct mlx5_wqe_dseg *restrict dseg; + struct mlx5_wqe_dseg *__rte_restrict dseg; unsigned int ds; uint8_t *dptr; @@ -4731,10 +4749,10 @@ single_no_inline: } static __rte_always_inline enum mlx5_txcmp_code -mlx5_tx_burst_single(struct mlx5_txq_data *restrict txq, - struct rte_mbuf **restrict pkts, +mlx5_tx_burst_single(struct mlx5_txq_data *__rte_restrict txq, + struct rte_mbuf **__rte_restrict pkts, unsigned int pkts_n, - struct mlx5_txq_local *restrict loc, + struct mlx5_txq_local *__rte_restrict loc, unsigned int olx) { enum mlx5_txcmp_code ret; @@ -4785,8 +4803,8 @@ ordinary_send: * Number of packets successfully transmitted (<= pkts_n). */ static __rte_always_inline uint16_t -mlx5_tx_burst_tmpl(struct mlx5_txq_data *restrict txq, - struct rte_mbuf **restrict pkts, +mlx5_tx_burst_tmpl(struct mlx5_txq_data *__rte_restrict txq, + struct rte_mbuf **__rte_restrict pkts, uint16_t pkts_n, unsigned int olx) { @@ -4816,7 +4834,7 @@ send_loop: /* * Calculate the number of available resources - elts and WQEs. * There are two possible different scenarios: - * - no data inlining into WQEs, one WQEBB may contains upto + * - no data inlining into WQEs, one WQEBB may contains up to * four packets, in this case elts become scarce resource * - data inlining into WQEs, one packet may require multiple * WQEBBs, the WQEs become the limiting factor. @@ -5198,9 +5216,48 @@ MLX5_TXOFF_DECL(iv, MLX5_TXOFF_CONFIG_INLINE | MLX5_TXOFF_CONFIG_VLAN | MLX5_TXOFF_CONFIG_METADATA) +/* Generate routines with timestamp scheduling. */ +MLX5_TXOFF_DECL(full_ts_nompw, + MLX5_TXOFF_CONFIG_FULL | MLX5_TXOFF_CONFIG_TXPP) + +MLX5_TXOFF_DECL(full_ts_nompwi, + MLX5_TXOFF_CONFIG_MULTI | MLX5_TXOFF_CONFIG_TSO | + MLX5_TXOFF_CONFIG_SWP | MLX5_TXOFF_CONFIG_CSUM | + MLX5_TXOFF_CONFIG_VLAN | MLX5_TXOFF_CONFIG_METADATA | + MLX5_TXOFF_CONFIG_TXPP) + +MLX5_TXOFF_DECL(full_ts, + MLX5_TXOFF_CONFIG_FULL | MLX5_TXOFF_CONFIG_TXPP | + MLX5_TXOFF_CONFIG_EMPW) + +MLX5_TXOFF_DECL(full_ts_noi, + MLX5_TXOFF_CONFIG_MULTI | MLX5_TXOFF_CONFIG_TSO | + MLX5_TXOFF_CONFIG_SWP | MLX5_TXOFF_CONFIG_CSUM | + MLX5_TXOFF_CONFIG_VLAN | MLX5_TXOFF_CONFIG_METADATA | + MLX5_TXOFF_CONFIG_TXPP | MLX5_TXOFF_CONFIG_EMPW) + +MLX5_TXOFF_DECL(none_ts, + MLX5_TXOFF_CONFIG_NONE | MLX5_TXOFF_CONFIG_TXPP | + MLX5_TXOFF_CONFIG_EMPW) + +MLX5_TXOFF_DECL(mdi_ts, + MLX5_TXOFF_CONFIG_INLINE | MLX5_TXOFF_CONFIG_METADATA | + MLX5_TXOFF_CONFIG_TXPP | MLX5_TXOFF_CONFIG_EMPW) + +MLX5_TXOFF_DECL(mti_ts, + MLX5_TXOFF_CONFIG_MULTI | MLX5_TXOFF_CONFIG_TSO | + MLX5_TXOFF_CONFIG_INLINE | MLX5_TXOFF_CONFIG_METADATA | + MLX5_TXOFF_CONFIG_TXPP | MLX5_TXOFF_CONFIG_EMPW) + +MLX5_TXOFF_DECL(mtiv_ts, + MLX5_TXOFF_CONFIG_MULTI | MLX5_TXOFF_CONFIG_TSO | + MLX5_TXOFF_CONFIG_INLINE | MLX5_TXOFF_CONFIG_VLAN | + MLX5_TXOFF_CONFIG_METADATA | MLX5_TXOFF_CONFIG_TXPP | + MLX5_TXOFF_CONFIG_EMPW) + /* * Generate routines with Legacy Multi-Packet Write support. - * This mode is supported by ConnectX-4LX only and imposes + * This mode is supported by ConnectX-4 Lx only and imposes * offload limitations, not supported: * - ACL/Flows (metadata are becoming meaningless) * - WQE Inline headers @@ -5302,6 +5359,44 @@ MLX5_TXOFF_INFO(iv_empw, MLX5_TXOFF_CONFIG_INLINE | MLX5_TXOFF_CONFIG_VLAN | MLX5_TXOFF_CONFIG_METADATA | MLX5_TXOFF_CONFIG_EMPW) +MLX5_TXOFF_INFO(full_ts_nompw, + MLX5_TXOFF_CONFIG_FULL | MLX5_TXOFF_CONFIG_TXPP) + +MLX5_TXOFF_INFO(full_ts_nompwi, + MLX5_TXOFF_CONFIG_MULTI | MLX5_TXOFF_CONFIG_TSO | + MLX5_TXOFF_CONFIG_SWP | MLX5_TXOFF_CONFIG_CSUM | + MLX5_TXOFF_CONFIG_VLAN | MLX5_TXOFF_CONFIG_METADATA | + MLX5_TXOFF_CONFIG_TXPP) + +MLX5_TXOFF_INFO(full_ts, + MLX5_TXOFF_CONFIG_FULL | MLX5_TXOFF_CONFIG_TXPP | + MLX5_TXOFF_CONFIG_EMPW) + +MLX5_TXOFF_INFO(full_ts_noi, + MLX5_TXOFF_CONFIG_MULTI | MLX5_TXOFF_CONFIG_TSO | + MLX5_TXOFF_CONFIG_SWP | MLX5_TXOFF_CONFIG_CSUM | + MLX5_TXOFF_CONFIG_VLAN | MLX5_TXOFF_CONFIG_METADATA | + MLX5_TXOFF_CONFIG_TXPP | MLX5_TXOFF_CONFIG_EMPW) + +MLX5_TXOFF_INFO(none_ts, + MLX5_TXOFF_CONFIG_NONE | MLX5_TXOFF_CONFIG_TXPP | + MLX5_TXOFF_CONFIG_EMPW) + +MLX5_TXOFF_INFO(mdi_ts, + MLX5_TXOFF_CONFIG_INLINE | MLX5_TXOFF_CONFIG_METADATA | + MLX5_TXOFF_CONFIG_TXPP | MLX5_TXOFF_CONFIG_EMPW) + +MLX5_TXOFF_INFO(mti_ts, + MLX5_TXOFF_CONFIG_MULTI | MLX5_TXOFF_CONFIG_TSO | + MLX5_TXOFF_CONFIG_INLINE | MLX5_TXOFF_CONFIG_METADATA | + MLX5_TXOFF_CONFIG_TXPP | MLX5_TXOFF_CONFIG_EMPW) + +MLX5_TXOFF_INFO(mtiv_ts, + MLX5_TXOFF_CONFIG_MULTI | MLX5_TXOFF_CONFIG_TSO | + MLX5_TXOFF_CONFIG_INLINE | MLX5_TXOFF_CONFIG_VLAN | + MLX5_TXOFF_CONFIG_METADATA | MLX5_TXOFF_CONFIG_TXPP | + MLX5_TXOFF_CONFIG_EMPW) + MLX5_TXOFF_INFO(full, MLX5_TXOFF_CONFIG_MULTI | MLX5_TXOFF_CONFIG_TSO | MLX5_TXOFF_CONFIG_SWP | MLX5_TXOFF_CONFIG_CSUM | @@ -5448,6 +5543,14 @@ mlx5_select_tx_function(struct rte_eth_dev *dev) /* We should support VLAN insertion. */ olx |= MLX5_TXOFF_CONFIG_VLAN; } + if (tx_offloads & DEV_TX_OFFLOAD_SEND_ON_TIMESTAMP && + rte_mbuf_dynflag_lookup + (RTE_MBUF_DYNFLAG_TX_TIMESTAMP_NAME, NULL) >= 0 && + rte_mbuf_dynfield_lookup + (RTE_MBUF_DYNFIELD_TIMESTAMP_NAME, NULL) >= 0) { + /* Offload configured, dynamic entities registered. */ + olx |= MLX5_TXOFF_CONFIG_TXPP; + } if (priv->txqs_n && (*priv->txqs)[0]) { struct mlx5_txq_data *txd = (*priv->txqs)[0]; @@ -5508,12 +5611,18 @@ mlx5_select_tx_function(struct rte_eth_dev *dev) /* Does not meet requested offloads at all. */ continue; } + if ((olx ^ tmp) & MLX5_TXOFF_CONFIG_MPW) + /* Do not enable legacy MPW if not configured. */ + continue; if ((olx ^ tmp) & MLX5_TXOFF_CONFIG_EMPW) /* Do not enable eMPW if not configured. */ continue; if ((olx ^ tmp) & MLX5_TXOFF_CONFIG_INLINE) /* Do not enable inlining if not configured. */ continue; + if ((olx ^ tmp) & MLX5_TXOFF_CONFIG_TXPP) + /* Do not enable scheduling if not configured. */ + continue; /* * Some routine meets the requirements. * Check whether it has minimal amount @@ -5558,6 +5667,8 @@ mlx5_select_tx_function(struct rte_eth_dev *dev) DRV_LOG(DEBUG, "\tVLANI (VLAN insertion)"); if (txoff_func[m].olx & MLX5_TXOFF_CONFIG_METADATA) DRV_LOG(DEBUG, "\tMETAD (tx Flow metadata)"); + if (txoff_func[m].olx & MLX5_TXOFF_CONFIG_TXPP) + DRV_LOG(DEBUG, "\tMETAD (tx Scheduling)"); if (txoff_func[m].olx & MLX5_TXOFF_CONFIG_EMPW) { if (txoff_func[m].olx & MLX5_TXOFF_CONFIG_MPW) DRV_LOG(DEBUG, "\tMPW (Legacy MPW)"); @@ -5632,7 +5743,7 @@ mlx5_tx_burst_mode_get(struct rte_eth_dev *dev, if (pkt_burst == txoff_func[i].func) { olx = txoff_func[i].olx; snprintf(mode->info, sizeof(mode->info), - "%s%s%s%s%s%s%s%s", + "%s%s%s%s%s%s%s%s%s", (olx & MLX5_TXOFF_CONFIG_EMPW) ? ((olx & MLX5_TXOFF_CONFIG_MPW) ? "Legacy MPW" : "Enhanced MPW") : "No MPW", @@ -5649,7 +5760,9 @@ mlx5_tx_burst_mode_get(struct rte_eth_dev *dev, (olx & MLX5_TXOFF_CONFIG_VLAN) ? " + VLAN" : "", (olx & MLX5_TXOFF_CONFIG_METADATA) ? - " + METADATA" : ""); + " + METADATA" : "", + (olx & MLX5_TXOFF_CONFIG_TXPP) ? + " + TXPP" : ""); return 0; } }