diff options
Diffstat (limited to 'drivers/net/ethernet/mellanox/mlx4/en_tx.c')
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx4/en_tx.c | 1139 |
1 files changed, 792 insertions, 347 deletions
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_tx.c b/drivers/net/ethernet/mellanox/mlx4/en_tx.c index 7c492382da09..87f35bcbeff8 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_tx.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_tx.c @@ -37,120 +37,150 @@ #include <linux/mlx4/qp.h> #include <linux/skbuff.h> #include <linux/if_vlan.h> +#include <linux/prefetch.h> #include <linux/vmalloc.h> #include <linux/tcp.h> -#include <linux/moduleparam.h> +#include <linux/ip.h> +#include <linux/ipv6.h> +#include <linux/indirect_call_wrapper.h> +#include <net/ipv6.h> +#include <net/page_pool/helpers.h> #include "mlx4_en.h" -enum { - MAX_INLINE = 104, /* 128 - 16 - 4 - 4 */ - MAX_BF = 256, -}; - -static int inline_thold __read_mostly = MAX_INLINE; - -module_param_named(inline_thold, inline_thold, int, 0444); -MODULE_PARM_DESC(inline_thold, "threshold for using inline data"); - int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv, - struct mlx4_en_tx_ring *ring, int qpn, u32 size, - u16 stride) + struct mlx4_en_tx_ring **pring, u32 size, + u16 stride, int node, int queue_index) { struct mlx4_en_dev *mdev = priv->mdev; + struct mlx4_en_tx_ring *ring; int tmp; int err; + ring = kzalloc_node(sizeof(*ring), GFP_KERNEL, node); + if (!ring) { + en_err(priv, "Failed allocating TX ring\n"); + return -ENOMEM; + } + ring->size = size; ring->size_mask = size - 1; - ring->stride = stride; - - inline_thold = min(inline_thold, MAX_INLINE); + ring->sp_stride = stride; + ring->full_size = ring->size - HEADROOM - MLX4_MAX_DESC_TXBBS; tmp = size * sizeof(struct mlx4_en_tx_info); - ring->tx_info = vmalloc(tmp); - if (!ring->tx_info) - return -ENOMEM; + ring->tx_info = kvmalloc_node(tmp, GFP_KERNEL, node); + if (!ring->tx_info) { + err = -ENOMEM; + goto err_ring; + } en_dbg(DRV, priv, "Allocated tx_info ring at addr:%p size:%d\n", ring->tx_info, tmp); - ring->bounce_buf = kmalloc(MAX_DESC_SIZE, GFP_KERNEL); + ring->bounce_buf = kmalloc_node(MLX4_TX_BOUNCE_BUFFER_SIZE, + GFP_KERNEL, node); if (!ring->bounce_buf) { - err = -ENOMEM; - goto err_tx; + ring->bounce_buf = kmalloc(MLX4_TX_BOUNCE_BUFFER_SIZE, + GFP_KERNEL); + if (!ring->bounce_buf) { + err = -ENOMEM; + goto err_info; + } } - ring->buf_size = ALIGN(size * ring->stride, MLX4_EN_PAGE_SIZE); + ring->buf_size = ALIGN(size * ring->sp_stride, MLX4_EN_PAGE_SIZE); - err = mlx4_alloc_hwq_res(mdev->dev, &ring->wqres, ring->buf_size, - 2 * PAGE_SIZE); + /* Allocate HW buffers on provided NUMA node */ + set_dev_node(&mdev->dev->persist->pdev->dev, node); + err = mlx4_alloc_hwq_res(mdev->dev, &ring->sp_wqres, ring->buf_size); + set_dev_node(&mdev->dev->persist->pdev->dev, mdev->dev->numa_node); if (err) { en_err(priv, "Failed allocating hwq resources\n"); goto err_bounce; } - err = mlx4_en_map_buffer(&ring->wqres.buf); + ring->buf = ring->sp_wqres.buf.direct.buf; + + en_dbg(DRV, priv, "Allocated TX ring (addr:%p) - buf:%p size:%d buf_size:%d dma:%llx\n", + ring, ring->buf, ring->size, ring->buf_size, + (unsigned long long) ring->sp_wqres.buf.direct.map); + + err = mlx4_qp_reserve_range(mdev->dev, 1, 1, &ring->qpn, + MLX4_RESERVE_ETH_BF_QP, + MLX4_RES_USAGE_DRIVER); if (err) { - en_err(priv, "Failed to map TX buffer\n"); + en_err(priv, "failed reserving qp for TX ring\n"); goto err_hwq_res; } - ring->buf = ring->wqres.buf.direct.buf; - - en_dbg(DRV, priv, "Allocated TX ring (addr:%p) - buf:%p size:%d " - "buf_size:%d dma:%llx\n", ring, ring->buf, ring->size, - ring->buf_size, (unsigned long long) ring->wqres.buf.direct.map); - - ring->qpn = qpn; - err = mlx4_qp_alloc(mdev->dev, ring->qpn, &ring->qp); + err = mlx4_qp_alloc(mdev->dev, ring->qpn, &ring->sp_qp); if (err) { en_err(priv, "Failed allocating qp %d\n", ring->qpn); - goto err_map; + goto err_reserve; } - ring->qp.event = mlx4_en_sqp_event; + ring->sp_qp.event = mlx4_en_sqp_event; - err = mlx4_bf_alloc(mdev->dev, &ring->bf); + err = mlx4_bf_alloc(mdev->dev, &ring->bf, node); if (err) { - en_dbg(DRV, priv, "working without blueflame (%d)", err); + en_dbg(DRV, priv, "working without blueflame (%d)\n", err); ring->bf.uar = &mdev->priv_uar; ring->bf.uar->map = mdev->uar_map; ring->bf_enabled = false; - } else - ring->bf_enabled = true; + ring->bf_alloced = false; + priv->pflags &= ~MLX4_EN_PRIV_FLAGS_BLUEFLAME; + } else { + ring->bf_alloced = true; + ring->bf_enabled = !!(priv->pflags & + MLX4_EN_PRIV_FLAGS_BLUEFLAME); + } + ring->doorbell_address = ring->bf.uar->map + MLX4_SEND_DOORBELL; ring->hwtstamp_tx_type = priv->hwtstamp_config.tx_type; + ring->queue_index = queue_index; + + if (queue_index < priv->num_tx_rings_p_up) + cpumask_set_cpu(cpumask_local_spread(queue_index, + priv->mdev->dev->numa_node), + &ring->sp_affinity_mask); + *pring = ring; return 0; -err_map: - mlx4_en_unmap_buffer(&ring->wqres.buf); +err_reserve: + mlx4_qp_release_range(mdev->dev, ring->qpn, 1); err_hwq_res: - mlx4_free_hwq_res(mdev->dev, &ring->wqres, ring->buf_size); + mlx4_free_hwq_res(mdev->dev, &ring->sp_wqres, ring->buf_size); err_bounce: kfree(ring->bounce_buf); ring->bounce_buf = NULL; -err_tx: - vfree(ring->tx_info); +err_info: + kvfree(ring->tx_info); ring->tx_info = NULL; +err_ring: + kfree(ring); + *pring = NULL; return err; } void mlx4_en_destroy_tx_ring(struct mlx4_en_priv *priv, - struct mlx4_en_tx_ring *ring) + struct mlx4_en_tx_ring **pring) { struct mlx4_en_dev *mdev = priv->mdev; + struct mlx4_en_tx_ring *ring = *pring; en_dbg(DRV, priv, "Destroying tx ring, qpn: %d\n", ring->qpn); - if (ring->bf_enabled) + if (ring->bf_alloced) mlx4_bf_free(mdev->dev, &ring->bf); - mlx4_qp_remove(mdev->dev, &ring->qp); - mlx4_qp_free(mdev->dev, &ring->qp); - mlx4_en_unmap_buffer(&ring->wqres.buf); - mlx4_free_hwq_res(mdev->dev, &ring->wqres, ring->buf_size); + mlx4_qp_remove(mdev->dev, &ring->sp_qp); + mlx4_qp_free(mdev->dev, &ring->sp_qp); + mlx4_qp_release_range(priv->mdev->dev, ring->qpn, 1); + mlx4_free_hwq_res(mdev->dev, &ring->sp_wqres, ring->buf_size); kfree(ring->bounce_buf); ring->bounce_buf = NULL; - vfree(ring->tx_info); + kvfree(ring->tx_info); ring->tx_info = NULL; + kfree(ring); + *pring = NULL; } int mlx4_en_activate_tx_ring(struct mlx4_en_priv *priv, @@ -160,24 +190,30 @@ int mlx4_en_activate_tx_ring(struct mlx4_en_priv *priv, struct mlx4_en_dev *mdev = priv->mdev; int err; - ring->cqn = cq; + ring->sp_cqn = cq; ring->prod = 0; ring->cons = 0xffffffff; ring->last_nr_txbb = 1; - ring->poll_cnt = 0; memset(ring->tx_info, 0, ring->size * sizeof(struct mlx4_en_tx_info)); memset(ring->buf, 0, ring->buf_size); + ring->free_tx_desc = mlx4_en_free_tx_desc; - ring->qp_state = MLX4_QP_STATE_RST; - ring->doorbell_qpn = ring->qp.qpn << 8; + ring->sp_qp_state = MLX4_QP_STATE_RST; + ring->doorbell_qpn = cpu_to_be32(ring->sp_qp.qpn << 8); + ring->mr_key = cpu_to_be32(mdev->mr.key); - mlx4_en_fill_qp_context(priv, ring->size, ring->stride, 1, 0, ring->qpn, - ring->cqn, user_prio, &ring->context); - if (ring->bf_enabled) - ring->context.usr_page = cpu_to_be32(ring->bf.uar->index); + mlx4_en_fill_qp_context(priv, ring->size, ring->sp_stride, 1, 0, ring->qpn, + ring->sp_cqn, user_prio, &ring->sp_context); + if (ring->bf_alloced) + ring->sp_context.usr_page = + cpu_to_be32(mlx4_to_hw_uar_index(mdev->dev, + ring->bf.uar->index)); - err = mlx4_qp_to_ready(mdev->dev, &ring->wqres.mtt, &ring->context, - &ring->qp, &ring->qp_state); + err = mlx4_qp_to_ready(mdev->dev, &ring->sp_wqres.mtt, &ring->sp_context, + &ring->sp_qp, &ring->sp_qp_state); + if (!cpumask_empty(&ring->sp_affinity_mask)) + netif_set_xps_queue(priv->dev, &ring->sp_affinity_mask, + ring->queue_index); return err; } @@ -187,97 +223,141 @@ void mlx4_en_deactivate_tx_ring(struct mlx4_en_priv *priv, { struct mlx4_en_dev *mdev = priv->mdev; - mlx4_qp_modify(mdev->dev, NULL, ring->qp_state, - MLX4_QP_STATE_RST, NULL, 0, 0, &ring->qp); + mlx4_qp_modify(mdev->dev, NULL, ring->sp_qp_state, + MLX4_QP_STATE_RST, NULL, 0, 0, &ring->sp_qp); } +static inline bool mlx4_en_is_tx_ring_full(struct mlx4_en_tx_ring *ring) +{ + u32 used = READ_ONCE(ring->prod) - READ_ONCE(ring->cons); + + return used > ring->full_size; +} -static u32 mlx4_en_free_tx_desc(struct mlx4_en_priv *priv, - struct mlx4_en_tx_ring *ring, - int index, u8 owner, u64 timestamp) +static void mlx4_en_stamp_wqe(struct mlx4_en_priv *priv, + struct mlx4_en_tx_ring *ring, int index, + u8 owner) { - struct mlx4_en_dev *mdev = priv->mdev; + __be32 stamp = cpu_to_be32(STAMP_VAL | (!!owner << STAMP_SHIFT)); + struct mlx4_en_tx_desc *tx_desc = ring->buf + (index << LOG_TXBB_SIZE); struct mlx4_en_tx_info *tx_info = &ring->tx_info[index]; - struct mlx4_en_tx_desc *tx_desc = ring->buf + index * TXBB_SIZE; - struct mlx4_wqe_data_seg *data = (void *) tx_desc + tx_info->data_offset; - struct sk_buff *skb = tx_info->skb; - struct skb_frag_struct *frag; void *end = ring->buf + ring->buf_size; - int frags = skb_shinfo(skb)->nr_frags; - int i; __be32 *ptr = (__be32 *)tx_desc; - __be32 stamp = cpu_to_be32(STAMP_VAL | (!!owner << STAMP_SHIFT)); - struct skb_shared_hwtstamps hwts; - - if (timestamp) { - mlx4_en_fill_hwtstamps(mdev, &hwts, timestamp); - skb_tstamp_tx(skb, &hwts); - } + int i; /* Optimize the common case when there are no wraparounds */ - if (likely((void *) tx_desc + tx_info->nr_txbb * TXBB_SIZE <= end)) { - if (!tx_info->inl) { - if (tx_info->linear) { - dma_unmap_single(priv->ddev, - (dma_addr_t) be64_to_cpu(data->addr), - be32_to_cpu(data->byte_count), - PCI_DMA_TODEVICE); - ++data; - } - - for (i = 0; i < frags; i++) { - frag = &skb_shinfo(skb)->frags[i]; - dma_unmap_page(priv->ddev, - (dma_addr_t) be64_to_cpu(data[i].addr), - skb_frag_size(frag), PCI_DMA_TODEVICE); - } - } + if (likely((void *)tx_desc + + (tx_info->nr_txbb << LOG_TXBB_SIZE) <= end)) { /* Stamp the freed descriptor */ - for (i = 0; i < tx_info->nr_txbb * TXBB_SIZE; i += STAMP_STRIDE) { + for (i = 0; i < tx_info->nr_txbb << LOG_TXBB_SIZE; + i += STAMP_STRIDE) { *ptr = stamp; ptr += STAMP_DWORDS; } - } else { - if (!tx_info->inl) { - if ((void *) data >= end) { - data = ring->buf + ((void *)data - end); + /* Stamp the freed descriptor */ + for (i = 0; i < tx_info->nr_txbb << LOG_TXBB_SIZE; + i += STAMP_STRIDE) { + *ptr = stamp; + ptr += STAMP_DWORDS; + if ((void *)ptr >= end) { + ptr = ring->buf; + stamp ^= cpu_to_be32(0x80000000); } + } + } +} + +INDIRECT_CALLABLE_DECLARE(u32 mlx4_en_free_tx_desc(struct mlx4_en_priv *priv, + struct mlx4_en_tx_ring *ring, + int index, u64 timestamp, + int napi_mode)); + +u32 mlx4_en_free_tx_desc(struct mlx4_en_priv *priv, + struct mlx4_en_tx_ring *ring, + int index, u64 timestamp, + int napi_mode) +{ + struct mlx4_en_tx_info *tx_info = &ring->tx_info[index]; + struct mlx4_en_tx_desc *tx_desc = ring->buf + (index << LOG_TXBB_SIZE); + struct mlx4_wqe_data_seg *data = (void *) tx_desc + tx_info->data_offset; + void *end = ring->buf + ring->buf_size; + struct sk_buff *skb = tx_info->skb; + int nr_maps = tx_info->nr_maps; + int i; + + /* We do not touch skb here, so prefetch skb->users location + * to speedup consume_skb() + */ + prefetchw(&skb->users); + + if (unlikely(timestamp)) { + struct skb_shared_hwtstamps hwts; - if (tx_info->linear) { - dma_unmap_single(priv->ddev, - (dma_addr_t) be64_to_cpu(data->addr), - be32_to_cpu(data->byte_count), - PCI_DMA_TODEVICE); - ++data; + mlx4_en_fill_hwtstamps(priv->mdev, &hwts, timestamp); + skb_tstamp_tx(skb, &hwts); + } + + if (!tx_info->inl) { + if (tx_info->linear) + dma_unmap_single(priv->ddev, + tx_info->map0_dma, + tx_info->map0_byte_count, + DMA_TO_DEVICE); + else + dma_unmap_page(priv->ddev, + tx_info->map0_dma, + tx_info->map0_byte_count, + DMA_TO_DEVICE); + /* Optimize the common case when there are no wraparounds */ + if (likely((void *)tx_desc + + (tx_info->nr_txbb << LOG_TXBB_SIZE) <= end)) { + for (i = 1; i < nr_maps; i++) { + data++; + dma_unmap_page(priv->ddev, + (dma_addr_t)be64_to_cpu(data->addr), + be32_to_cpu(data->byte_count), + DMA_TO_DEVICE); } + } else { + if ((void *)data >= end) + data = ring->buf + ((void *)data - end); - for (i = 0; i < frags; i++) { + for (i = 1; i < nr_maps; i++) { + data++; /* Check for wraparound before unmapping */ if ((void *) data >= end) data = ring->buf; - frag = &skb_shinfo(skb)->frags[i]; dma_unmap_page(priv->ddev, - (dma_addr_t) be64_to_cpu(data->addr), - skb_frag_size(frag), PCI_DMA_TODEVICE); - ++data; - } - } - /* Stamp the freed descriptor */ - for (i = 0; i < tx_info->nr_txbb * TXBB_SIZE; i += STAMP_STRIDE) { - *ptr = stamp; - ptr += STAMP_DWORDS; - if ((void *) ptr >= end) { - ptr = ring->buf; - stamp ^= cpu_to_be32(0x80000000); + (dma_addr_t)be64_to_cpu(data->addr), + be32_to_cpu(data->byte_count), + DMA_TO_DEVICE); } } - } - dev_kfree_skb_any(skb); + napi_consume_skb(skb, napi_mode); + return tx_info->nr_txbb; } +INDIRECT_CALLABLE_DECLARE(u32 mlx4_en_recycle_tx_desc(struct mlx4_en_priv *priv, + struct mlx4_en_tx_ring *ring, + int index, u64 timestamp, + int napi_mode)); + +u32 mlx4_en_recycle_tx_desc(struct mlx4_en_priv *priv, + struct mlx4_en_tx_ring *ring, + int index, u64 timestamp, + int napi_mode) +{ + struct mlx4_en_tx_info *tx_info = &ring->tx_info[index]; + struct page_pool *pool = ring->recycle_ring->pp; + + /* Note that napi_mode = 0 means ndo_close() path, not budget = 0 */ + page_pool_put_full_page(pool, tx_info->page, !!napi_mode); + + return tx_info->nr_txbb; +} int mlx4_en_free_tx_buf(struct net_device *dev, struct mlx4_en_tx_ring *ring) { @@ -296,14 +376,15 @@ int mlx4_en_free_tx_buf(struct net_device *dev, struct mlx4_en_tx_ring *ring) } while (ring->cons != ring->prod) { - ring->last_nr_txbb = mlx4_en_free_tx_desc(priv, ring, + ring->last_nr_txbb = ring->free_tx_desc(priv, ring, ring->cons & ring->size_mask, - !!(ring->cons & ring->size), 0); + 0, 0 /* Non-NAPI caller */); ring->cons += ring->last_nr_txbb; cnt++; } - netdev_tx_reset_queue(ring->tx_queue); + if (ring->tx_queue) + netdev_tx_reset_queue(ring->tx_queue); if (cnt) en_dbg(DRV, priv, "Freed %d uncompleted tx descriptors\n", cnt); @@ -311,15 +392,45 @@ int mlx4_en_free_tx_buf(struct net_device *dev, struct mlx4_en_tx_ring *ring) return cnt; } -static void mlx4_en_process_tx_cq(struct net_device *dev, struct mlx4_en_cq *cq) +static void mlx4_en_handle_err_cqe(struct mlx4_en_priv *priv, struct mlx4_err_cqe *err_cqe, + u16 cqe_index, struct mlx4_en_tx_ring *ring) +{ + struct mlx4_en_dev *mdev = priv->mdev; + struct mlx4_en_tx_info *tx_info; + struct mlx4_en_tx_desc *tx_desc; + u16 wqe_index; + int desc_size; + + en_err(priv, "CQE error - cqn 0x%x, ci 0x%x, vendor syndrome: 0x%x syndrome: 0x%x\n", + ring->sp_cqn, cqe_index, err_cqe->vendor_err_syndrome, err_cqe->syndrome); + print_hex_dump(KERN_WARNING, "", DUMP_PREFIX_OFFSET, 16, 1, err_cqe, sizeof(*err_cqe), + false); + + wqe_index = be16_to_cpu(err_cqe->wqe_index) & ring->size_mask; + tx_info = &ring->tx_info[wqe_index]; + desc_size = tx_info->nr_txbb << LOG_TXBB_SIZE; + en_err(priv, "Related WQE - qpn 0x%x, wqe index 0x%x, wqe size 0x%x\n", ring->qpn, + wqe_index, desc_size); + tx_desc = ring->buf + (wqe_index << LOG_TXBB_SIZE); + print_hex_dump(KERN_WARNING, "", DUMP_PREFIX_OFFSET, 16, 1, tx_desc, desc_size, false); + + if (test_and_set_bit(MLX4_EN_STATE_FLAG_RESTARTING, &priv->state)) + return; + + en_err(priv, "Scheduling port restart\n"); + queue_work(mdev->workqueue, &priv->restart_task); +} + +int mlx4_en_process_tx_cq(struct net_device *dev, + struct mlx4_en_cq *cq, int napi_budget) { struct mlx4_en_priv *priv = netdev_priv(dev); struct mlx4_cq *mcq = &cq->mcq; - struct mlx4_en_tx_ring *ring = &priv->tx_ring[cq->ring]; + struct mlx4_en_tx_ring *ring = priv->tx_ring[cq->type][cq->ring]; struct mlx4_cqe *cqe; - u16 index; - u16 new_index, ring_index; + u16 index, ring_index, stamp_index; u32 txbbs_skipped = 0; + u32 txbbs_stamp = 0; u32 cons_index = mcq->cons_index; int size = cq->size; u32 size_mask = ring->size_mask; @@ -327,48 +438,75 @@ static void mlx4_en_process_tx_cq(struct net_device *dev, struct mlx4_en_cq *cq) u32 packets = 0; u32 bytes = 0; int factor = priv->cqe_factor; - u64 timestamp = 0; + int done = 0; + int budget = priv->tx_work_limit; + u32 last_nr_txbb; + u32 ring_cons; - if (!priv->port_up) - return; + if (unlikely(!priv->port_up)) + return 0; + if (unlikely(!napi_budget) && cq->type == TX_XDP) + return 0; + + netdev_txq_bql_complete_prefetchw(ring->tx_queue); index = cons_index & size_mask; - cqe = &buf[(index << factor) + factor]; - ring_index = ring->cons & size_mask; + cqe = mlx4_en_get_cqe(buf, index, priv->cqe_size) + factor; + last_nr_txbb = READ_ONCE(ring->last_nr_txbb); + ring_cons = READ_ONCE(ring->cons); + ring_index = ring_cons & size_mask; + stamp_index = ring_index; /* Process all completed CQEs */ while (XNOR(cqe->owner_sr_opcode & MLX4_CQE_OWNER_MASK, - cons_index & size)) { + cons_index & size) && (done < budget)) { + u16 new_index; + /* * make sure we read the CQE after we read the * ownership bit */ - rmb(); + dma_rmb(); + + if (unlikely((cqe->owner_sr_opcode & MLX4_CQE_OPCODE_MASK) == + MLX4_CQE_OPCODE_ERROR)) + if (!test_and_set_bit(MLX4_EN_TX_RING_STATE_RECOVERING, &ring->state)) + mlx4_en_handle_err_cqe(priv, (struct mlx4_err_cqe *)cqe, index, + ring); /* Skip over last polled CQE */ new_index = be16_to_cpu(cqe->wqe_index) & size_mask; do { - txbbs_skipped += ring->last_nr_txbb; - ring_index = (ring_index + ring->last_nr_txbb) & size_mask; - if (ring->tx_info[ring_index].ts_requested) + u64 timestamp = 0; + + txbbs_skipped += last_nr_txbb; + ring_index = (ring_index + last_nr_txbb) & size_mask; + + if (unlikely(ring->tx_info[ring_index].ts_requested)) timestamp = mlx4_en_get_cqe_ts(cqe); /* free next descriptor */ - ring->last_nr_txbb = mlx4_en_free_tx_desc( + last_nr_txbb = INDIRECT_CALL_2(ring->free_tx_desc, + mlx4_en_free_tx_desc, + mlx4_en_recycle_tx_desc, priv, ring, ring_index, - !!((ring->cons + txbbs_skipped) & - ring->size), timestamp); + timestamp, napi_budget); + + mlx4_en_stamp_wqe(priv, ring, stamp_index, + !!((ring_cons + txbbs_stamp) & + ring->size)); + stamp_index = ring_index; + txbbs_stamp = txbbs_skipped; packets++; bytes += ring->tx_info[ring_index].nr_bytes; - } while (ring_index != new_index); + } while ((++done < budget) && (ring_index != new_index)); ++cons_index; index = cons_index & size_mask; - cqe = &buf[(index << factor) + factor]; + cqe = mlx4_en_get_cqe(buf, index, priv->cqe_size) + factor; } - /* * To prevent CQ overflow we first update CQ consumer and only then * the ring consumer. @@ -376,17 +514,25 @@ static void mlx4_en_process_tx_cq(struct net_device *dev, struct mlx4_en_cq *cq) mcq->cons_index = cons_index; mlx4_cq_set_ci(mcq); wmb(); - ring->cons += txbbs_skipped; + + /* we want to dirty this cache line once */ + WRITE_ONCE(ring->last_nr_txbb, last_nr_txbb); + WRITE_ONCE(ring->cons, ring_cons + txbbs_skipped); + + if (cq->type == TX_XDP) + return done; + netdev_tx_completed_queue(ring->tx_queue, packets, bytes); - /* - * Wakeup Tx queue if this stopped, and at least 1 packet - * was completed + /* Wakeup Tx queue if this stopped, and ring is not full. */ - if (netif_tx_queue_stopped(ring->tx_queue) && txbbs_skipped > 0) { + if (netif_tx_queue_stopped(ring->tx_queue) && + !mlx4_en_is_tx_ring_full(ring)) { netif_tx_wake_queue(ring->tx_queue); - priv->port_stats.wake_queue++; + ring->wake_queue++; } + + return done; } void mlx4_en_tx_irq(struct mlx4_cq *mcq) @@ -394,17 +540,36 @@ void mlx4_en_tx_irq(struct mlx4_cq *mcq) struct mlx4_en_cq *cq = container_of(mcq, struct mlx4_en_cq, mcq); struct mlx4_en_priv *priv = netdev_priv(cq->dev); - mlx4_en_process_tx_cq(cq->dev, cq); - mlx4_en_arm_cq(priv, cq); + if (likely(priv->port_up)) + napi_schedule_irqoff(&cq->napi); + else + mlx4_en_arm_cq(priv, cq); } +/* TX CQ polling - called by NAPI */ +int mlx4_en_poll_tx_cq(struct napi_struct *napi, int budget) +{ + struct mlx4_en_cq *cq = container_of(napi, struct mlx4_en_cq, napi); + struct net_device *dev = cq->dev; + struct mlx4_en_priv *priv = netdev_priv(dev); + int work_done; + + work_done = mlx4_en_process_tx_cq(dev, cq, budget); + if (work_done >= budget) + return budget; + + if (napi_complete_done(napi, work_done)) + mlx4_en_arm_cq(priv, cq); + + return 0; +} static struct mlx4_en_tx_desc *mlx4_en_bounce_to_desc(struct mlx4_en_priv *priv, struct mlx4_en_tx_ring *ring, u32 index, unsigned int desc_size) { - u32 copy = (ring->size - index) * TXBB_SIZE; + u32 copy = (ring->size - index) << LOG_TXBB_SIZE; int i; for (i = desc_size - copy - 4; i >= 0; i -= 4) { @@ -419,38 +584,43 @@ static struct mlx4_en_tx_desc *mlx4_en_bounce_to_desc(struct mlx4_en_priv *priv, if ((i & (TXBB_SIZE - 1)) == 0) wmb(); - *((u32 *) (ring->buf + index * TXBB_SIZE + i)) = + *((u32 *)(ring->buf + (index << LOG_TXBB_SIZE) + i)) = *((u32 *) (ring->bounce_buf + i)); } /* Return real descriptor location */ - return ring->buf + index * TXBB_SIZE; + return ring->buf + (index << LOG_TXBB_SIZE); } -static int is_inline(struct sk_buff *skb, void **pfrag) +/* Decide if skb can be inlined in tx descriptor to avoid dma mapping + * + * It seems strange we do not simply use skb_copy_bits(). + * This would allow to inline all skbs iff skb->len <= inline_thold + * + * Note that caller already checked skb was not a gso packet + */ +static bool is_inline(int inline_thold, const struct sk_buff *skb, + const struct skb_shared_info *shinfo, + void **pfrag) { void *ptr; - if (inline_thold && !skb_is_gso(skb) && skb->len <= inline_thold) { - if (skb_shinfo(skb)->nr_frags == 1) { - ptr = skb_frag_address_safe(&skb_shinfo(skb)->frags[0]); - if (unlikely(!ptr)) - return 0; - - if (pfrag) - *pfrag = ptr; + if (skb->len > inline_thold || !inline_thold) + return false; - return 1; - } else if (unlikely(skb_shinfo(skb)->nr_frags)) - return 0; - else - return 1; + if (shinfo->nr_frags == 1) { + ptr = skb_frag_address_safe(&shinfo->frags[0]); + if (unlikely(!ptr)) + return false; + *pfrag = ptr; + return true; } - - return 0; + if (shinfo->nr_frags) + return false; + return true; } -static int inline_size(struct sk_buff *skb) +static int inline_size(const struct sk_buff *skb) { if (skb->len + CTRL_SIZE + sizeof(struct mlx4_wqe_inline_seg) <= MLX4_INLINE_ALIGN) @@ -461,16 +631,33 @@ static int inline_size(struct sk_buff *skb) sizeof(struct mlx4_wqe_inline_seg), 16); } -static int get_real_size(struct sk_buff *skb, struct net_device *dev, - int *lso_header_size) +static int get_real_size(const struct sk_buff *skb, + const struct skb_shared_info *shinfo, + struct net_device *dev, + int *lso_header_size, + bool *inline_ok, + void **pfrag, + int *hopbyhop) { struct mlx4_en_priv *priv = netdev_priv(dev); int real_size; - if (skb_is_gso(skb)) { - *lso_header_size = skb_transport_offset(skb) + tcp_hdrlen(skb); - real_size = CTRL_SIZE + skb_shinfo(skb)->nr_frags * DS_SIZE + - ALIGN(*lso_header_size + 4, DS_SIZE); + if (shinfo->gso_size) { + *inline_ok = false; + *hopbyhop = 0; + if (skb->encapsulation) { + *lso_header_size = skb_inner_tcp_all_headers(skb); + } else { + /* Detects large IPV6 TCP packets and prepares for removal of + * HBH header that has been pushed by ip6_xmit(), + * mainly so that tcpdump can dissect them. + */ + if (ipv6_has_hopopt_jumbo(skb)) + *hopbyhop = sizeof(struct hop_jumbo_hdr); + *lso_header_size = skb_tcp_all_headers(skb); + } + real_size = CTRL_SIZE + shinfo->nr_frags * DS_SIZE + + ALIGN(*lso_header_size - *hopbyhop + 4, DS_SIZE); if (unlikely(*lso_header_size != skb_headlen(skb))) { /* We add a segment for the skb linear buffer only if * it contains data */ @@ -484,145 +671,262 @@ static int get_real_size(struct sk_buff *skb, struct net_device *dev, } } else { *lso_header_size = 0; - if (!is_inline(skb, NULL)) - real_size = CTRL_SIZE + (skb_shinfo(skb)->nr_frags + 1) * DS_SIZE; - else + *inline_ok = is_inline(priv->prof->inline_thold, skb, + shinfo, pfrag); + + if (*inline_ok) real_size = inline_size(skb); + else + real_size = CTRL_SIZE + + (shinfo->nr_frags + 1) * DS_SIZE; } return real_size; } -static void build_inline_wqe(struct mlx4_en_tx_desc *tx_desc, struct sk_buff *skb, - int real_size, u16 *vlan_tag, int tx_ind, void *fragptr) +static void build_inline_wqe(struct mlx4_en_tx_desc *tx_desc, + const struct sk_buff *skb, + const struct skb_shared_info *shinfo, + void *fragptr) { struct mlx4_wqe_inline_seg *inl = &tx_desc->inl; - int spc = MLX4_INLINE_ALIGN - CTRL_SIZE - sizeof *inl; + int spc = MLX4_INLINE_ALIGN - CTRL_SIZE - sizeof(*inl); + unsigned int hlen = skb_headlen(skb); if (skb->len <= spc) { - inl->byte_count = cpu_to_be32(1 << 31 | skb->len); - skb_copy_from_linear_data(skb, inl + 1, skb_headlen(skb)); - if (skb_shinfo(skb)->nr_frags) - memcpy(((void *)(inl + 1)) + skb_headlen(skb), fragptr, - skb_frag_size(&skb_shinfo(skb)->frags[0])); + if (likely(skb->len >= MIN_PKT_LEN)) { + inl->byte_count = cpu_to_be32(1 << 31 | skb->len); + } else { + inl->byte_count = cpu_to_be32(1 << 31 | MIN_PKT_LEN); + memset(inl->data + skb->len, 0, + MIN_PKT_LEN - skb->len); + } + skb_copy_from_linear_data(skb, inl->data, hlen); + if (shinfo->nr_frags) + memcpy(inl->data + hlen, fragptr, + skb_frag_size(&shinfo->frags[0])); } else { inl->byte_count = cpu_to_be32(1 << 31 | spc); - if (skb_headlen(skb) <= spc) { - skb_copy_from_linear_data(skb, inl + 1, skb_headlen(skb)); - if (skb_headlen(skb) < spc) { - memcpy(((void *)(inl + 1)) + skb_headlen(skb), - fragptr, spc - skb_headlen(skb)); - fragptr += spc - skb_headlen(skb); + if (hlen <= spc) { + skb_copy_from_linear_data(skb, inl->data, hlen); + if (hlen < spc) { + memcpy(inl->data + hlen, + fragptr, spc - hlen); + fragptr += spc - hlen; } - inl = (void *) (inl + 1) + spc; - memcpy(((void *)(inl + 1)), fragptr, skb->len - spc); + inl = (void *)inl->data + spc; + memcpy(inl->data, fragptr, skb->len - spc); } else { - skb_copy_from_linear_data(skb, inl + 1, spc); - inl = (void *) (inl + 1) + spc; - skb_copy_from_linear_data_offset(skb, spc, inl + 1, - skb_headlen(skb) - spc); - if (skb_shinfo(skb)->nr_frags) - memcpy(((void *)(inl + 1)) + skb_headlen(skb) - spc, - fragptr, skb_frag_size(&skb_shinfo(skb)->frags[0])); + skb_copy_from_linear_data(skb, inl->data, spc); + inl = (void *)inl->data + spc; + skb_copy_from_linear_data_offset(skb, spc, inl->data, + hlen - spc); + if (shinfo->nr_frags) + memcpy(inl->data + hlen - spc, + fragptr, + skb_frag_size(&shinfo->frags[0])); } - wmb(); + dma_wmb(); inl->byte_count = cpu_to_be32(1 << 31 | (skb->len - spc)); } } -u16 mlx4_en_select_queue(struct net_device *dev, struct sk_buff *skb) +u16 mlx4_en_select_queue(struct net_device *dev, struct sk_buff *skb, + struct net_device *sb_dev) { struct mlx4_en_priv *priv = netdev_priv(dev); u16 rings_p_up = priv->num_tx_rings_p_up; - u8 up = 0; - if (dev->num_tc) - return skb_tx_hash(dev, skb); + if (netdev_get_num_tc(dev)) + return netdev_pick_tx(dev, skb, NULL); - if (vlan_tx_tag_present(skb)) - up = vlan_tx_tag_get(skb) >> VLAN_PRIO_SHIFT; - - return __netdev_pick_tx(dev, skb) % rings_p_up + up * rings_p_up; + return netdev_pick_tx(dev, skb, NULL) % rings_p_up; } -static void mlx4_bf_copy(void __iomem *dst, unsigned long *src, unsigned bytecnt) +static void mlx4_bf_copy(void __iomem *dst, const void *src, + unsigned int bytecnt) { __iowrite64_copy(dst, src, bytecnt / 8); } +void mlx4_en_xmit_doorbell(struct mlx4_en_tx_ring *ring) +{ + wmb(); + /* Since there is no iowrite*_native() that writes the + * value as is, without byteswapping - using the one + * the doesn't do byteswapping in the relevant arch + * endianness. + */ +#if defined(__LITTLE_ENDIAN) + iowrite32( +#else + iowrite32be( +#endif + (__force u32)ring->doorbell_qpn, ring->doorbell_address); +} + +static void mlx4_en_tx_write_desc(struct mlx4_en_tx_ring *ring, + struct mlx4_en_tx_desc *tx_desc, + union mlx4_wqe_qpn_vlan qpn_vlan, + int desc_size, int bf_index, + __be32 op_own, bool bf_ok, + bool send_doorbell) +{ + tx_desc->ctrl.qpn_vlan = qpn_vlan; + + if (bf_ok) { + op_own |= htonl((bf_index & 0xffff) << 8); + /* Ensure new descriptor hits memory + * before setting ownership of this descriptor to HW + */ + dma_wmb(); + tx_desc->ctrl.owner_opcode = op_own; + + wmb(); + + mlx4_bf_copy(ring->bf.reg + ring->bf.offset, &tx_desc->ctrl, + desc_size); + + wmb(); + + ring->bf.offset ^= ring->bf.buf_size; + } else { + /* Ensure new descriptor hits memory + * before setting ownership of this descriptor to HW + */ + dma_wmb(); + tx_desc->ctrl.owner_opcode = op_own; + if (send_doorbell) + mlx4_en_xmit_doorbell(ring); + else + ring->xmit_more++; + } +} + +static bool mlx4_en_build_dma_wqe(struct mlx4_en_priv *priv, + struct skb_shared_info *shinfo, + struct mlx4_wqe_data_seg *data, + struct sk_buff *skb, + int lso_header_size, + __be32 mr_key, + struct mlx4_en_tx_info *tx_info) +{ + struct device *ddev = priv->ddev; + dma_addr_t dma = 0; + u32 byte_count = 0; + int i_frag; + + /* Map fragments if any */ + for (i_frag = shinfo->nr_frags - 1; i_frag >= 0; i_frag--) { + const skb_frag_t *frag = &shinfo->frags[i_frag]; + byte_count = skb_frag_size(frag); + dma = skb_frag_dma_map(ddev, frag, + 0, byte_count, + DMA_TO_DEVICE); + if (dma_mapping_error(ddev, dma)) + goto tx_drop_unmap; + + data->addr = cpu_to_be64(dma); + data->lkey = mr_key; + dma_wmb(); + data->byte_count = cpu_to_be32(byte_count); + --data; + } + + /* Map linear part if needed */ + if (tx_info->linear) { + byte_count = skb_headlen(skb) - lso_header_size; + + dma = dma_map_single(ddev, skb->data + + lso_header_size, byte_count, + DMA_TO_DEVICE); + if (dma_mapping_error(ddev, dma)) + goto tx_drop_unmap; + + data->addr = cpu_to_be64(dma); + data->lkey = mr_key; + dma_wmb(); + data->byte_count = cpu_to_be32(byte_count); + } + /* tx completion can avoid cache line miss for common cases */ + tx_info->map0_dma = dma; + tx_info->map0_byte_count = byte_count; + + return true; + +tx_drop_unmap: + en_err(priv, "DMA mapping error\n"); + + while (++i_frag < shinfo->nr_frags) { + ++data; + dma_unmap_page(ddev, (dma_addr_t)be64_to_cpu(data->addr), + be32_to_cpu(data->byte_count), + DMA_TO_DEVICE); + } + + return false; +} + netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev) { + struct skb_shared_info *shinfo = skb_shinfo(skb); struct mlx4_en_priv *priv = netdev_priv(dev); - struct mlx4_en_dev *mdev = priv->mdev; + union mlx4_wqe_qpn_vlan qpn_vlan = {}; struct mlx4_en_tx_ring *ring; struct mlx4_en_tx_desc *tx_desc; struct mlx4_wqe_data_seg *data; - struct skb_frag_struct *frag; struct mlx4_en_tx_info *tx_info; - struct ethhdr *ethh; - int tx_ind = 0; + u32 __maybe_unused ring_cons; + int tx_ind; int nr_txbb; int desc_size; int real_size; - dma_addr_t dma; u32 index, bf_index; + struct ipv6hdr *h6; __be32 op_own; - u16 vlan_tag = 0; - int i; int lso_header_size; - void *fragptr; + void *fragptr = NULL; bool bounce = false; + bool send_doorbell; + bool stop_queue; + bool inline_ok; + u8 data_offset; + int hopbyhop; + bool bf_ok; - if (!priv->port_up) + tx_ind = skb_get_queue_mapping(skb); + ring = priv->tx_ring[TX][tx_ind]; + + if (unlikely(!priv->port_up)) goto tx_drop; - real_size = get_real_size(skb, dev, &lso_header_size); + real_size = get_real_size(skb, shinfo, dev, &lso_header_size, + &inline_ok, &fragptr, &hopbyhop); if (unlikely(!real_size)) - goto tx_drop; + goto tx_drop_count; /* Align descriptor to TXBB size */ desc_size = ALIGN(real_size, TXBB_SIZE); - nr_txbb = desc_size / TXBB_SIZE; - if (unlikely(nr_txbb > MAX_DESC_TXBBS)) { - if (netif_msg_tx_err(priv)) - en_warn(priv, "Oversized header or SG list\n"); - goto tx_drop; - } - - tx_ind = skb->queue_mapping; - ring = &priv->tx_ring[tx_ind]; - if (vlan_tx_tag_present(skb)) - vlan_tag = vlan_tx_tag_get(skb); - - /* Check available TXBBs And 2K spare for prefetch */ - if (unlikely(((int)(ring->prod - ring->cons)) > - ring->size - HEADROOM - MAX_DESC_TXBBS)) { - /* every full Tx ring stops queue */ - netif_tx_stop_queue(ring->tx_queue); - priv->port_stats.queue_stopped++; - - /* If queue was emptied after the if, and before the - * stop_queue - need to wake the queue, or else it will remain - * stopped forever. - * Need a memory barrier to make sure ring->cons was not - * updated before queue was stopped. - */ - wmb(); - - if (unlikely(((int)(ring->prod - ring->cons)) <= - ring->size - HEADROOM - MAX_DESC_TXBBS)) { - netif_tx_wake_queue(ring->tx_queue); - priv->port_stats.wake_queue++; - } else { - return NETDEV_TX_BUSY; - } + nr_txbb = desc_size >> LOG_TXBB_SIZE; + + bf_ok = ring->bf_enabled; + if (skb_vlan_tag_present(skb)) { + u16 vlan_proto; + + qpn_vlan.vlan_tag = cpu_to_be16(skb_vlan_tag_get(skb)); + vlan_proto = be16_to_cpu(skb->vlan_proto); + if (vlan_proto == ETH_P_8021AD) + qpn_vlan.ins_vlan = MLX4_WQE_CTRL_INS_SVLAN; + else if (vlan_proto == ETH_P_8021Q) + qpn_vlan.ins_vlan = MLX4_WQE_CTRL_INS_CVLAN; + else + qpn_vlan.ins_vlan = 0; + bf_ok = false; } - /* Track current inflight packets for performance analysis */ - AVG_PERF_COUNTER(priv->pstats.inflight_avg, - (u32) (ring->prod - ring->cons - 1)); + netdev_txq_bql_enqueue_prefetchw(ring->tx_queue); /* Packet is good - grab an index and transmit it */ index = ring->prod & ring->size_mask; @@ -631,10 +935,16 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev) /* See if we have enough space for whole descriptor TXBB for setting * SW ownership on next descriptor; if not, use a bounce buffer. */ if (likely(index + nr_txbb <= ring->size)) - tx_desc = ring->buf + index * TXBB_SIZE; + tx_desc = ring->buf + (index << LOG_TXBB_SIZE); else { + if (unlikely(nr_txbb > MLX4_MAX_DESC_TXBBS)) { + if (netif_msg_tx_err(priv)) + en_warn(priv, "Oversized header or SG list\n"); + goto tx_drop_count; + } tx_desc = (struct mlx4_en_tx_desc *) ring->bounce_buf; bounce = true; + bf_ok = false; } /* Save skb in tx_info ring */ @@ -642,30 +952,58 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev) tx_info->skb = skb; tx_info->nr_txbb = nr_txbb; + if (!lso_header_size) { + data = &tx_desc->data; + data_offset = offsetof(struct mlx4_en_tx_desc, data); + } else { + int lso_align = ALIGN(lso_header_size - hopbyhop + 4, DS_SIZE); + + data = (void *)&tx_desc->lso + lso_align; + data_offset = offsetof(struct mlx4_en_tx_desc, lso) + lso_align; + } + + /* valid only for none inline segments */ + tx_info->data_offset = data_offset; + + tx_info->inl = inline_ok; + + tx_info->linear = lso_header_size < skb_headlen(skb) && !inline_ok; + + tx_info->nr_maps = shinfo->nr_frags + tx_info->linear; + data += tx_info->nr_maps - 1; + + if (!tx_info->inl) + if (!mlx4_en_build_dma_wqe(priv, shinfo, data, skb, + lso_header_size, ring->mr_key, + tx_info)) + goto tx_drop_count; + /* * For timestamping add flag to skb_shinfo and * set flag for further reference */ - if (ring->hwtstamp_tx_type == HWTSTAMP_TX_ON && - skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP) { - skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS; + tx_info->ts_requested = 0; + if (unlikely(ring->hwtstamp_tx_type == HWTSTAMP_TX_ON && + shinfo->tx_flags & SKBTX_HW_TSTAMP)) { + shinfo->tx_flags |= SKBTX_IN_PROGRESS; tx_info->ts_requested = 1; } - /* Prepare ctrl segement apart opcode+ownership, which depends on + /* Prepare ctrl segment apart opcode+ownership, which depends on * whether LSO is used */ - tx_desc->ctrl.vlan_tag = cpu_to_be16(vlan_tag); - tx_desc->ctrl.ins_vlan = MLX4_WQE_CTRL_INS_VLAN * - !!vlan_tx_tag_present(skb); - tx_desc->ctrl.fence_size = (real_size / 16) & 0x3f; tx_desc->ctrl.srcrb_flags = priv->ctrl_flags; if (likely(skb->ip_summed == CHECKSUM_PARTIAL)) { - tx_desc->ctrl.srcrb_flags |= cpu_to_be32(MLX4_WQE_CTRL_IP_CSUM | - MLX4_WQE_CTRL_TCP_UDP_CSUM); + if (!skb->encapsulation) + tx_desc->ctrl.srcrb_flags |= cpu_to_be32(MLX4_WQE_CTRL_IP_CSUM | + MLX4_WQE_CTRL_TCP_UDP_CSUM); + else + tx_desc->ctrl.srcrb_flags |= cpu_to_be32(MLX4_WQE_CTRL_IP_CSUM); ring->tx_csum++; } if (priv->flags & MLX4_EN_FLAG_ENABLE_HW_LOOPBACK) { + struct ethhdr *ethh; + /* Copy dst mac address to wqe. This allows loopback in eSwitch, * so that VFs and PF can communicate with each other */ @@ -676,24 +1014,41 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev) /* Handle LSO (TSO) packets */ if (lso_header_size) { + int i; + /* Mark opcode as LSO */ op_own = cpu_to_be32(MLX4_OPCODE_LSO | (1 << 6)) | ((ring->prod & ring->size) ? cpu_to_be32(MLX4_EN_BIT_DESC_OWN) : 0); + lso_header_size -= hopbyhop; /* Fill in the LSO prefix */ tx_desc->lso.mss_hdr_size = cpu_to_be32( - skb_shinfo(skb)->gso_size << 16 | lso_header_size); - - /* Copy headers; - * note that we already verified that it is linear */ - memcpy(tx_desc->lso.header, skb->data, lso_header_size); - data = ((void *) &tx_desc->lso + - ALIGN(lso_header_size + 4, DS_SIZE)); + shinfo->gso_size << 16 | lso_header_size); + + + if (unlikely(hopbyhop)) { + /* remove the HBH header. + * Layout: [Ethernet header][IPv6 header][HBH][TCP header] + */ + memcpy(tx_desc->lso.header, skb->data, ETH_HLEN + sizeof(*h6)); + h6 = (struct ipv6hdr *)((char *)tx_desc->lso.header + ETH_HLEN); + h6->nexthdr = IPPROTO_TCP; + /* Copy the TCP header after the IPv6 one */ + memcpy(h6 + 1, + skb->data + ETH_HLEN + sizeof(*h6) + + sizeof(struct hop_jumbo_hdr), + tcp_hdrlen(skb)); + /* Leave ipv6 payload_len set to 0, as LSO v2 specs request. */ + } else { + /* Copy headers; + * note that we already verified that it is linear + */ + memcpy(tx_desc->lso.header, skb->data, lso_header_size); + } + ring->tso_packets++; - priv->port_stats.tso_packets++; - i = ((skb->len - lso_header_size) / skb_shinfo(skb)->gso_size) + - !!((skb->len - lso_header_size) % skb_shinfo(skb)->gso_size); + i = shinfo->gso_segs; tx_info->nr_bytes = skb->len + (i - 1) * lso_header_size; ring->packets += i; } else { @@ -701,89 +1056,179 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev) op_own = cpu_to_be32(MLX4_OPCODE_SEND) | ((ring->prod & ring->size) ? cpu_to_be32(MLX4_EN_BIT_DESC_OWN) : 0); - data = &tx_desc->data; tx_info->nr_bytes = max_t(unsigned int, skb->len, ETH_ZLEN); ring->packets++; - } ring->bytes += tx_info->nr_bytes; - netdev_tx_sent_queue(ring->tx_queue, tx_info->nr_bytes); - AVG_PERF_COUNTER(priv->pstats.tx_pktsz_avg, skb->len); + if (tx_info->inl) + build_inline_wqe(tx_desc, skb, shinfo, fragptr); - /* valid only for none inline segments */ - tx_info->data_offset = (void *) data - (void *) tx_desc; - - tx_info->linear = (lso_header_size < skb_headlen(skb) && !is_inline(skb, NULL)) ? 1 : 0; - data += skb_shinfo(skb)->nr_frags + tx_info->linear - 1; - - if (!is_inline(skb, &fragptr)) { - /* Map fragments */ - for (i = skb_shinfo(skb)->nr_frags - 1; i >= 0; i--) { - frag = &skb_shinfo(skb)->frags[i]; - dma = skb_frag_dma_map(priv->ddev, frag, - 0, skb_frag_size(frag), - DMA_TO_DEVICE); - data->addr = cpu_to_be64(dma); - data->lkey = cpu_to_be32(mdev->mr.key); - wmb(); - data->byte_count = cpu_to_be32(skb_frag_size(frag)); - --data; - } + if (skb->encapsulation) { + union { + struct iphdr *v4; + struct ipv6hdr *v6; + unsigned char *hdr; + } ip; + u8 proto; - /* Map linear part */ - if (tx_info->linear) { - dma = dma_map_single(priv->ddev, skb->data + lso_header_size, - skb_headlen(skb) - lso_header_size, PCI_DMA_TODEVICE); - data->addr = cpu_to_be64(dma); - data->lkey = cpu_to_be32(mdev->mr.key); - wmb(); - data->byte_count = cpu_to_be32(skb_headlen(skb) - lso_header_size); - } - tx_info->inl = 0; - } else { - build_inline_wqe(tx_desc, skb, real_size, &vlan_tag, tx_ind, fragptr); - tx_info->inl = 1; + ip.hdr = skb_inner_network_header(skb); + proto = (ip.v4->version == 4) ? ip.v4->protocol : + ip.v6->nexthdr; + + if (proto == IPPROTO_TCP || proto == IPPROTO_UDP) + op_own |= cpu_to_be32(MLX4_WQE_CTRL_IIP | MLX4_WQE_CTRL_ILP); + else + op_own |= cpu_to_be32(MLX4_WQE_CTRL_IIP); } - ring->prod += nr_txbb; + WRITE_ONCE(ring->prod, ring->prod + nr_txbb); /* If we used a bounce buffer then copy descriptor back into place */ - if (bounce) + if (unlikely(bounce)) tx_desc = mlx4_en_bounce_to_desc(priv, ring, index, desc_size); skb_tx_timestamp(skb); - if (ring->bf_enabled && desc_size <= MAX_BF && !bounce && !vlan_tx_tag_present(skb)) { - *(__be32 *) (&tx_desc->ctrl.vlan_tag) |= cpu_to_be32(ring->doorbell_qpn); - op_own |= htonl((bf_index & 0xffff) << 8); - /* Ensure new descirptor hits memory - * before setting ownership of this descriptor to HW */ - wmb(); - tx_desc->ctrl.owner_opcode = op_own; + /* Check available TXBBs And 2K spare for prefetch */ + stop_queue = mlx4_en_is_tx_ring_full(ring); + if (unlikely(stop_queue)) { + netif_tx_stop_queue(ring->tx_queue); + ring->queue_stopped++; + } - wmb(); + send_doorbell = __netdev_tx_sent_queue(ring->tx_queue, + tx_info->nr_bytes, + netdev_xmit_more()); - mlx4_bf_copy(ring->bf.reg + ring->bf.offset, (unsigned long *) &tx_desc->ctrl, - desc_size); + real_size = (real_size / 16) & 0x3f; - wmb(); + bf_ok &= desc_size <= MAX_BF && send_doorbell; - ring->bf.offset ^= ring->bf.buf_size; - } else { - /* Ensure new descirptor hits memory - * before setting ownership of this descriptor to HW */ - wmb(); - tx_desc->ctrl.owner_opcode = op_own; - wmb(); - iowrite32be(ring->doorbell_qpn, ring->bf.uar->map + MLX4_SEND_DOORBELL); - } + if (bf_ok) + qpn_vlan.bf_qpn = ring->doorbell_qpn | cpu_to_be32(real_size); + else + qpn_vlan.fence_size = real_size; + + mlx4_en_tx_write_desc(ring, tx_desc, qpn_vlan, desc_size, bf_index, + op_own, bf_ok, send_doorbell); + + if (unlikely(stop_queue)) { + /* If queue was emptied after the if (stop_queue) , and before + * the netif_tx_stop_queue() - need to wake the queue, + * or else it will remain stopped forever. + * Need a memory barrier to make sure ring->cons was not + * updated before queue was stopped. + */ + smp_rmb(); + if (unlikely(!mlx4_en_is_tx_ring_full(ring))) { + netif_tx_wake_queue(ring->tx_queue); + ring->wake_queue++; + } + } return NETDEV_TX_OK; +tx_drop_count: + ring->tx_dropped++; tx_drop: dev_kfree_skb_any(skb); - priv->stats.tx_dropped++; return NETDEV_TX_OK; } +#define MLX4_EN_XDP_TX_NRTXBB 1 +#define MLX4_EN_XDP_TX_REAL_SZ (((CTRL_SIZE + MLX4_EN_XDP_TX_NRTXBB * DS_SIZE) \ + / 16) & 0x3f) + +void mlx4_en_init_tx_xdp_ring_descs(struct mlx4_en_priv *priv, + struct mlx4_en_tx_ring *ring) +{ + int i; + + for (i = 0; i < ring->size; i++) { + struct mlx4_en_tx_info *tx_info = &ring->tx_info[i]; + struct mlx4_en_tx_desc *tx_desc = ring->buf + + (i << LOG_TXBB_SIZE); + + tx_info->map0_byte_count = PAGE_SIZE; + tx_info->nr_txbb = MLX4_EN_XDP_TX_NRTXBB; + tx_info->data_offset = offsetof(struct mlx4_en_tx_desc, data); + tx_info->ts_requested = 0; + tx_info->nr_maps = 1; + tx_info->linear = 1; + tx_info->inl = 0; + + tx_desc->data.lkey = ring->mr_key; + tx_desc->ctrl.qpn_vlan.fence_size = MLX4_EN_XDP_TX_REAL_SZ; + tx_desc->ctrl.srcrb_flags = priv->ctrl_flags; + } +} + +netdev_tx_t mlx4_en_xmit_frame(struct mlx4_en_rx_ring *rx_ring, + struct mlx4_en_rx_alloc *frame, + struct mlx4_en_priv *priv, unsigned int length, + int tx_ind, bool *doorbell_pending) +{ + struct mlx4_en_tx_desc *tx_desc; + struct mlx4_en_tx_info *tx_info; + struct mlx4_wqe_data_seg *data; + struct mlx4_en_tx_ring *ring; + dma_addr_t dma; + __be32 op_own; + int index; + + if (unlikely(!priv->port_up)) + goto tx_drop; + + ring = priv->tx_ring[TX_XDP][tx_ind]; + + if (unlikely(mlx4_en_is_tx_ring_full(ring))) + goto tx_drop_count; + + index = ring->prod & ring->size_mask; + tx_info = &ring->tx_info[index]; + + tx_desc = ring->buf + (index << LOG_TXBB_SIZE); + data = &tx_desc->data; + + dma = page_pool_get_dma_addr(frame->page); + + tx_info->page = frame->page; + frame->page = NULL; + tx_info->map0_dma = dma; + tx_info->nr_bytes = max_t(unsigned int, length, ETH_ZLEN); + + dma_sync_single_range_for_device(priv->ddev, dma, frame->page_offset, + length, DMA_TO_DEVICE); + + data->addr = cpu_to_be64(dma + frame->page_offset); + dma_wmb(); + data->byte_count = cpu_to_be32(length); + + /* tx completion can avoid cache line miss for common cases */ + + op_own = cpu_to_be32(MLX4_OPCODE_SEND) | + ((ring->prod & ring->size) ? + cpu_to_be32(MLX4_EN_BIT_DESC_OWN) : 0); + + rx_ring->xdp_tx++; + + WRITE_ONCE(ring->prod, ring->prod + MLX4_EN_XDP_TX_NRTXBB); + + /* Ensure new descriptor hits memory + * before setting ownership of this descriptor to HW + */ + dma_wmb(); + tx_desc->ctrl.owner_opcode = op_own; + ring->xmit_more++; + + *doorbell_pending = true; + + return NETDEV_TX_OK; + +tx_drop_count: + rx_ring->xdp_tx_full++; + *doorbell_pending = true; +tx_drop: + return NETDEV_TX_BUSY; +} |