diff options
Diffstat (limited to 'drivers/net/mlx4/mlx4_rxtx.c')
| -rw-r--r-- | drivers/net/mlx4/mlx4_rxtx.c | 566 |
1 files changed, 456 insertions, 110 deletions
diff --git a/drivers/net/mlx4/mlx4_rxtx.c b/drivers/net/mlx4/mlx4_rxtx.c index 8ca8b77c..8c88effc 100644 --- a/drivers/net/mlx4/mlx4_rxtx.c +++ b/drivers/net/mlx4/mlx4_rxtx.c @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: BSD-3-Clause * Copyright 2017 6WIND S.A. - * Copyright 2017 Mellanox + * Copyright 2017 Mellanox Technologies, Ltd */ /** @@ -38,10 +38,29 @@ * DWORD (32 byte) of a TXBB. */ struct pv { - volatile struct mlx4_wqe_data_seg *dseg; + union { + volatile struct mlx4_wqe_data_seg *dseg; + volatile uint32_t *dst; + }; uint32_t val; }; +/** A helper structure for TSO packet handling. */ +struct tso_info { + /** Pointer to the array of saved first DWORD (32 byte) of a TXBB. */ + struct pv *pv; + /** Current entry in the pv array. */ + int pv_counter; + /** Total size of the WQE including padding. */ + uint32_t wqe_size; + /** Size of TSO header to prepend to each packet to send. */ + uint16_t tso_header_size; + /** Total size of the TSO segment in the WQE. */ + uint16_t wqe_tso_seg_size; + /** Raw WQE size in units of 16 Bytes and without padding. */ + uint8_t fence_size; +}; + /** A table to translate Rx completion flags to packet type. */ uint32_t mlx4_ptype_table[0x100] __rte_cache_aligned = { /* @@ -52,49 +71,58 @@ uint32_t mlx4_ptype_table[0x100] __rte_cache_aligned = { * bit[4] - MLX4_CQE_STATUS_TCP * bit[3] - MLX4_CQE_STATUS_IPV4OPT * bit[2] - MLX4_CQE_STATUS_IPV6 - * bit[1] - MLX4_CQE_STATUS_IPV4F + * bit[1] - MLX4_CQE_STATUS_IPF * bit[0] - MLX4_CQE_STATUS_IPV4 * giving a total of up to 256 entries. */ + /* L2 */ [0x00] = RTE_PTYPE_L2_ETHER, + /* L3 */ [0x01] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_L4_NONFRAG, [0x02] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_L4_FRAG, [0x03] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_L4_FRAG, - [0x04] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN, - [0x09] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT, + [0x04] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_NONFRAG, + [0x06] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_FRAG, + [0x08] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT | + RTE_PTYPE_L4_NONFRAG, + [0x09] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT | + RTE_PTYPE_L4_NONFRAG, [0x0a] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT | RTE_PTYPE_L4_FRAG, + [0x0b] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT | + RTE_PTYPE_L4_FRAG, + /* TCP */ [0x11] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_L4_TCP, - [0x12] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_L4_TCP, [0x14] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_L4_TCP, + [0x16] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_FRAG, [0x18] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT | RTE_PTYPE_L4_TCP, [0x19] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT | RTE_PTYPE_L4_TCP, - [0x1a] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT | - RTE_PTYPE_L4_TCP, + /* UDP */ [0x21] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_L4_UDP, - [0x22] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_L4_UDP, [0x24] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_L4_UDP, + [0x26] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_L4_FRAG, [0x28] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT | RTE_PTYPE_L4_UDP, [0x29] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT | RTE_PTYPE_L4_UDP, - [0x2a] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT | - RTE_PTYPE_L4_UDP, /* Tunneled - L3 IPV6 */ [0x80] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN, [0x81] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN, + RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_INNER_L4_NONFRAG, [0x82] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_FRAG, @@ -102,65 +130,58 @@ uint32_t mlx4_ptype_table[0x100] __rte_cache_aligned = { RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_FRAG, [0x84] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN, + RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_INNER_L4_NONFRAG, + [0x86] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_INNER_L4_FRAG, [0x88] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT, + RTE_PTYPE_INNER_L3_IPV4_EXT | + RTE_PTYPE_INNER_L4_NONFRAG, [0x89] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT, + RTE_PTYPE_INNER_L3_IPV4_EXT | + RTE_PTYPE_INNER_L4_NONFRAG, [0x8a] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_FRAG, + RTE_PTYPE_INNER_L3_IPV4_EXT | + RTE_PTYPE_INNER_L4_FRAG, + [0x8b] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_INNER_L3_IPV4_EXT | + RTE_PTYPE_INNER_L4_FRAG, /* Tunneled - L3 IPV6, TCP */ [0x91] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_TCP, - [0x92] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L4_FRAG | - RTE_PTYPE_INNER_L4_TCP, - [0x93] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L4_FRAG | - RTE_PTYPE_INNER_L4_TCP, [0x94] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_TCP, + [0x96] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_INNER_L4_FRAG, [0x98] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT | - RTE_PTYPE_INNER_L4_TCP, + RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_TCP, [0x99] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT | - RTE_PTYPE_INNER_L4_TCP, - [0x9a] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_FRAG | - RTE_PTYPE_INNER_L4_TCP, + RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_TCP, /* Tunneled - L3 IPV6, UDP */ - [0xa1] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L4_UDP, - [0xa2] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + [0xa1] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L4_FRAG | RTE_PTYPE_INNER_L4_UDP, - [0xa3] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L4_FRAG | - RTE_PTYPE_INNER_L4_UDP, - [0xa4] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + [0xa4] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_UDP, - [0xa8] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + [0xa6] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_INNER_L4_FRAG, + [0xa8] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_UDP, - [0xa9] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + [0xa9] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_UDP, - [0xaa] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_FRAG | - RTE_PTYPE_INNER_L4_UDP, /* Tunneled - L3 IPV4 */ [0xc0] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN, [0xc1] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN, + RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_INNER_L4_NONFRAG, [0xc2] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_FRAG, @@ -168,65 +189,54 @@ uint32_t mlx4_ptype_table[0x100] __rte_cache_aligned = { RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_FRAG, [0xc4] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN, + RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_INNER_L4_NONFRAG, + [0xc6] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_INNER_L4_FRAG, [0xc8] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT, + RTE_PTYPE_INNER_L3_IPV4_EXT | + RTE_PTYPE_INNER_L4_NONFRAG, [0xc9] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT, + RTE_PTYPE_INNER_L3_IPV4_EXT | + RTE_PTYPE_INNER_L4_NONFRAG, [0xca] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_FRAG, + [0xcb] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_INNER_L3_IPV4_EXT | + RTE_PTYPE_INNER_L4_FRAG, /* Tunneled - L3 IPV4, TCP */ - [0xd0] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | - RTE_PTYPE_INNER_L4_TCP, [0xd1] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_TCP, - [0xd2] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L4_FRAG | - RTE_PTYPE_INNER_L4_TCP, - [0xd3] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L4_FRAG | - RTE_PTYPE_INNER_L4_TCP, [0xd4] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_TCP, + [0xd6] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_INNER_L4_FRAG, [0xd8] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_TCP, [0xd9] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_TCP, - [0xda] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_FRAG | - RTE_PTYPE_INNER_L4_TCP, /* Tunneled - L3 IPV4, UDP */ - [0xe0] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | - RTE_PTYPE_INNER_L4_UDP, [0xe1] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_UDP, - [0xe2] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L4_FRAG | - RTE_PTYPE_INNER_L4_UDP, - [0xe3] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L4_FRAG | - RTE_PTYPE_INNER_L4_UDP, [0xe4] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_UDP, + [0xe6] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | + RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN | + RTE_PTYPE_INNER_L4_FRAG, [0xe8] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_UDP, + RTE_PTYPE_INNER_L3_IPV4_EXT | + RTE_PTYPE_INNER_L4_UDP, [0xe9] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_UDP, - [0xea] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | - RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_FRAG | + RTE_PTYPE_INNER_L3_IPV4_EXT | RTE_PTYPE_INNER_L4_UDP, }; @@ -263,7 +273,7 @@ mlx4_txq_stamp_freed_wqe(struct mlx4_sq *sq, volatile uint32_t *start, } while (start != (volatile uint32_t *)sq->eob); start = (volatile uint32_t *)sq->buf; /* Flip invalid stamping ownership. */ - stamp ^= RTE_BE32(0x1 << MLX4_SQ_OWNER_BIT); + stamp ^= RTE_BE32(1u << MLX4_SQ_OWNER_BIT); sq->stamp = stamp; if (start == end) return size; @@ -344,24 +354,6 @@ mlx4_txq_complete(struct txq *txq, const unsigned int elts_m, } /** - * Get memory pool (MP) from mbuf. If mbuf is indirect, the pool from which - * the cloned mbuf is allocated is returned instead. - * - * @param buf - * Pointer to mbuf. - * - * @return - * Memory pool where data is located for given mbuf. - */ -static struct rte_mempool * -mlx4_txq_mb2mp(struct rte_mbuf *buf) -{ - if (unlikely(RTE_MBUF_INDIRECT(buf))) - return rte_mbuf_from_indirect(buf)->pool; - return buf->pool; -} - -/** * Write Tx data segment to the SQ. * * @param dseg @@ -378,7 +370,7 @@ mlx4_fill_tx_data_seg(volatile struct mlx4_wqe_data_seg *dseg, uint32_t lkey, uintptr_t addr, rte_be32_t byte_count) { dseg->addr = rte_cpu_to_be_64(addr); - dseg->lkey = rte_cpu_to_be_32(lkey); + dseg->lkey = lkey; #if RTE_CACHE_LINE_SIZE < 64 /* * Need a barrier here before writing the byte_count @@ -395,6 +387,342 @@ mlx4_fill_tx_data_seg(volatile struct mlx4_wqe_data_seg *dseg, } /** + * Obtain and calculate TSO information needed for assembling a TSO WQE. + * + * @param buf + * Pointer to the first packet mbuf. + * @param txq + * Pointer to Tx queue structure. + * @param tinfo + * Pointer to a structure to fill the info with. + * + * @return + * 0 on success, negative value upon error. + */ +static inline int +mlx4_tx_burst_tso_get_params(struct rte_mbuf *buf, + struct txq *txq, + struct tso_info *tinfo) +{ + struct mlx4_sq *sq = &txq->msq; + const uint8_t tunneled = txq->priv->hw_csum_l2tun && + (buf->ol_flags & PKT_TX_TUNNEL_MASK); + + tinfo->tso_header_size = buf->l2_len + buf->l3_len + buf->l4_len; + if (tunneled) + tinfo->tso_header_size += + buf->outer_l2_len + buf->outer_l3_len; + if (unlikely(buf->tso_segsz == 0 || + tinfo->tso_header_size == 0 || + tinfo->tso_header_size > MLX4_MAX_TSO_HEADER || + tinfo->tso_header_size > buf->data_len)) + return -EINVAL; + /* + * Calculate the WQE TSO segment size + * Note: + * 1. An LSO segment must be padded such that the subsequent data + * segment is 16-byte aligned. + * 2. The start address of the TSO segment is always 16 Bytes aligned. + */ + tinfo->wqe_tso_seg_size = RTE_ALIGN(sizeof(struct mlx4_wqe_lso_seg) + + tinfo->tso_header_size, + sizeof(struct mlx4_wqe_data_seg)); + tinfo->fence_size = ((sizeof(struct mlx4_wqe_ctrl_seg) + + tinfo->wqe_tso_seg_size) >> MLX4_SEG_SHIFT) + + buf->nb_segs; + tinfo->wqe_size = + RTE_ALIGN((uint32_t)(tinfo->fence_size << MLX4_SEG_SHIFT), + MLX4_TXBB_SIZE); + /* Validate WQE size and WQE space in the send queue. */ + if (sq->remain_size < tinfo->wqe_size || + tinfo->wqe_size > MLX4_MAX_WQE_SIZE) + return -ENOMEM; + /* Init pv. */ + tinfo->pv = (struct pv *)txq->bounce_buf; + tinfo->pv_counter = 0; + return 0; +} + +/** + * Fill the TSO WQE data segments with info on buffers to transmit . + * + * @param buf + * Pointer to the first packet mbuf. + * @param txq + * Pointer to Tx queue structure. + * @param tinfo + * Pointer to TSO info to use. + * @param dseg + * Pointer to the first data segment in the TSO WQE. + * @param ctrl + * Pointer to the control segment in the TSO WQE. + * + * @return + * 0 on success, negative value upon error. + */ +static inline volatile struct mlx4_wqe_ctrl_seg * +mlx4_tx_burst_fill_tso_dsegs(struct rte_mbuf *buf, + struct txq *txq, + struct tso_info *tinfo, + volatile struct mlx4_wqe_data_seg *dseg, + volatile struct mlx4_wqe_ctrl_seg *ctrl) +{ + uint32_t lkey; + int nb_segs = buf->nb_segs; + int nb_segs_txbb; + struct mlx4_sq *sq = &txq->msq; + struct rte_mbuf *sbuf = buf; + struct pv *pv = tinfo->pv; + int *pv_counter = &tinfo->pv_counter; + volatile struct mlx4_wqe_ctrl_seg *ctrl_next = + (volatile struct mlx4_wqe_ctrl_seg *) + ((volatile uint8_t *)ctrl + tinfo->wqe_size); + uint16_t data_len = sbuf->data_len - tinfo->tso_header_size; + uintptr_t data_addr = rte_pktmbuf_mtod_offset(sbuf, uintptr_t, + tinfo->tso_header_size); + + do { + /* how many dseg entries do we have in the current TXBB ? */ + nb_segs_txbb = (MLX4_TXBB_SIZE - + ((uintptr_t)dseg & (MLX4_TXBB_SIZE - 1))) >> + MLX4_SEG_SHIFT; + switch (nb_segs_txbb) { +#ifndef NDEBUG + default: + /* Should never happen. */ + rte_panic("%p: Invalid number of SGEs(%d) for a TXBB", + (void *)txq, nb_segs_txbb); + /* rte_panic never returns. */ + break; +#endif /* NDEBUG */ + case 4: + /* Memory region key for this memory pool. */ + lkey = mlx4_tx_mb2mr(txq, sbuf); + if (unlikely(lkey == (uint32_t)-1)) + goto err; + dseg->addr = rte_cpu_to_be_64(data_addr); + dseg->lkey = lkey; + /* + * This data segment starts at the beginning of a new + * TXBB, so we need to postpone its byte_count writing + * for later. + */ + pv[*pv_counter].dseg = dseg; + /* + * Zero length segment is treated as inline segment + * with zero data. + */ + pv[(*pv_counter)++].val = + rte_cpu_to_be_32(data_len ? + data_len : + 0x80000000); + if (--nb_segs == 0) + return ctrl_next; + /* Prepare next buf info */ + sbuf = sbuf->next; + dseg++; + data_len = sbuf->data_len; + data_addr = rte_pktmbuf_mtod(sbuf, uintptr_t); + /* fallthrough */ + case 3: + lkey = mlx4_tx_mb2mr(txq, sbuf); + if (unlikely(lkey == (uint32_t)-1)) + goto err; + mlx4_fill_tx_data_seg(dseg, lkey, data_addr, + rte_cpu_to_be_32(data_len ? + data_len : + 0x80000000)); + if (--nb_segs == 0) + return ctrl_next; + /* Prepare next buf info */ + sbuf = sbuf->next; + dseg++; + data_len = sbuf->data_len; + data_addr = rte_pktmbuf_mtod(sbuf, uintptr_t); + /* fallthrough */ + case 2: + lkey = mlx4_tx_mb2mr(txq, sbuf); + if (unlikely(lkey == (uint32_t)-1)) + goto err; + mlx4_fill_tx_data_seg(dseg, lkey, data_addr, + rte_cpu_to_be_32(data_len ? + data_len : + 0x80000000)); + if (--nb_segs == 0) + return ctrl_next; + /* Prepare next buf info */ + sbuf = sbuf->next; + dseg++; + data_len = sbuf->data_len; + data_addr = rte_pktmbuf_mtod(sbuf, uintptr_t); + /* fallthrough */ + case 1: + lkey = mlx4_tx_mb2mr(txq, sbuf); + if (unlikely(lkey == (uint32_t)-1)) + goto err; + mlx4_fill_tx_data_seg(dseg, lkey, data_addr, + rte_cpu_to_be_32(data_len ? + data_len : + 0x80000000)); + if (--nb_segs == 0) + return ctrl_next; + /* Prepare next buf info */ + sbuf = sbuf->next; + dseg++; + data_len = sbuf->data_len; + data_addr = rte_pktmbuf_mtod(sbuf, uintptr_t); + /* fallthrough */ + } + /* Wrap dseg if it points at the end of the queue. */ + if ((volatile uint8_t *)dseg >= sq->eob) + dseg = (volatile struct mlx4_wqe_data_seg *) + ((volatile uint8_t *)dseg - sq->size); + } while (true); +err: + return NULL; +} + +/** + * Fill the packet's l2, l3 and l4 headers to the WQE. + * + * This will be used as the header for each TSO segment that is transmitted. + * + * @param buf + * Pointer to the first packet mbuf. + * @param txq + * Pointer to Tx queue structure. + * @param tinfo + * Pointer to TSO info to use. + * @param ctrl + * Pointer to the control segment in the TSO WQE. + * + * @return + * 0 on success, negative value upon error. + */ +static inline volatile struct mlx4_wqe_data_seg * +mlx4_tx_burst_fill_tso_hdr(struct rte_mbuf *buf, + struct txq *txq, + struct tso_info *tinfo, + volatile struct mlx4_wqe_ctrl_seg *ctrl) +{ + volatile struct mlx4_wqe_lso_seg *tseg = + (volatile struct mlx4_wqe_lso_seg *)(ctrl + 1); + struct mlx4_sq *sq = &txq->msq; + struct pv *pv = tinfo->pv; + int *pv_counter = &tinfo->pv_counter; + int remain_size = tinfo->tso_header_size; + char *from = rte_pktmbuf_mtod(buf, char *); + uint16_t txbb_avail_space; + /* Union to overcome volatile constraints when copying TSO header. */ + union { + volatile uint8_t *vto; + uint8_t *to; + } thdr = { .vto = (volatile uint8_t *)tseg->header, }; + + /* + * TSO data always starts at offset 20 from the beginning of the TXBB + * (16 byte ctrl + 4byte TSO desc). Since each TXBB is 64Byte aligned + * we can write the first 44 TSO header bytes without worry for TxQ + * wrapping or overwriting the first TXBB 32bit word. + */ + txbb_avail_space = MLX4_TXBB_SIZE - + (sizeof(struct mlx4_wqe_ctrl_seg) + + sizeof(struct mlx4_wqe_lso_seg)); + while (remain_size >= (int)(txbb_avail_space + sizeof(uint32_t))) { + /* Copy to end of txbb. */ + rte_memcpy(thdr.to, from, txbb_avail_space); + from += txbb_avail_space; + thdr.to += txbb_avail_space; + /* New TXBB, Check for TxQ wrap. */ + if (thdr.to >= sq->eob) + thdr.vto = sq->buf; + /* New TXBB, stash the first 32bits for later use. */ + pv[*pv_counter].dst = (volatile uint32_t *)thdr.to; + pv[(*pv_counter)++].val = *(uint32_t *)from, + from += sizeof(uint32_t); + thdr.to += sizeof(uint32_t); + remain_size -= txbb_avail_space + sizeof(uint32_t); + /* Avail space in new TXBB is TXBB size - 4 */ + txbb_avail_space = MLX4_TXBB_SIZE - sizeof(uint32_t); + } + if (remain_size > txbb_avail_space) { + rte_memcpy(thdr.to, from, txbb_avail_space); + from += txbb_avail_space; + thdr.to += txbb_avail_space; + remain_size -= txbb_avail_space; + /* New TXBB, Check for TxQ wrap. */ + if (thdr.to >= sq->eob) + thdr.vto = sq->buf; + pv[*pv_counter].dst = (volatile uint32_t *)thdr.to; + rte_memcpy(&pv[*pv_counter].val, from, remain_size); + (*pv_counter)++; + } else if (remain_size) { + rte_memcpy(thdr.to, from, remain_size); + } + tseg->mss_hdr_size = rte_cpu_to_be_32((buf->tso_segsz << 16) | + tinfo->tso_header_size); + /* Calculate data segment location */ + return (volatile struct mlx4_wqe_data_seg *) + ((uintptr_t)tseg + tinfo->wqe_tso_seg_size); +} + +/** + * Write data segments and header for TSO uni/multi segment packet. + * + * @param buf + * Pointer to the first packet mbuf. + * @param txq + * Pointer to Tx queue structure. + * @param ctrl + * Pointer to the WQE control segment. + * + * @return + * Pointer to the next WQE control segment on success, NULL otherwise. + */ +static volatile struct mlx4_wqe_ctrl_seg * +mlx4_tx_burst_tso(struct rte_mbuf *buf, struct txq *txq, + volatile struct mlx4_wqe_ctrl_seg *ctrl) +{ + volatile struct mlx4_wqe_data_seg *dseg; + volatile struct mlx4_wqe_ctrl_seg *ctrl_next; + struct mlx4_sq *sq = &txq->msq; + struct tso_info tinfo; + struct pv *pv; + int pv_counter; + int ret; + + ret = mlx4_tx_burst_tso_get_params(buf, txq, &tinfo); + if (unlikely(ret)) + goto error; + dseg = mlx4_tx_burst_fill_tso_hdr(buf, txq, &tinfo, ctrl); + if (unlikely(dseg == NULL)) + goto error; + if ((uintptr_t)dseg >= (uintptr_t)sq->eob) + dseg = (volatile struct mlx4_wqe_data_seg *) + ((uintptr_t)dseg - sq->size); + ctrl_next = mlx4_tx_burst_fill_tso_dsegs(buf, txq, &tinfo, dseg, ctrl); + if (unlikely(ctrl_next == NULL)) + goto error; + /* Write the first DWORD of each TXBB save earlier. */ + if (likely(tinfo.pv_counter)) { + pv = tinfo.pv; + pv_counter = tinfo.pv_counter; + /* Need a barrier here before writing the first TXBB word. */ + rte_io_wmb(); + do { + --pv_counter; + *pv[pv_counter].dst = pv[pv_counter].val; + } while (pv_counter > 0); + } + ctrl->fence_size = tinfo.fence_size; + sq->remain_size -= tinfo.wqe_size; + return ctrl_next; +error: + txq->stats.odropped++; + return NULL; +} + +/** * Write data segments of multi-segment packet. * * @param buf @@ -437,7 +765,7 @@ mlx4_tx_burst_segs(struct rte_mbuf *buf, struct txq *txq, goto txbb_tail_segs; txbb_head_seg: /* Memory region key (big endian) for this memory pool. */ - lkey = mlx4_txq_mp2mr(txq, mlx4_txq_mb2mp(sbuf)); + lkey = mlx4_tx_mb2mr(txq, sbuf); if (unlikely(lkey == (uint32_t)-1)) { DEBUG("%p: unable to get MP <-> MR association", (void *)txq); @@ -449,7 +777,7 @@ txbb_head_seg: dseg = (volatile struct mlx4_wqe_data_seg *) sq->buf; dseg->addr = rte_cpu_to_be_64(rte_pktmbuf_mtod(sbuf, uintptr_t)); - dseg->lkey = rte_cpu_to_be_32(lkey); + dseg->lkey = lkey; /* * This data segment starts at the beginning of a new * TXBB, so we need to postpone its byte_count writing @@ -469,7 +797,7 @@ txbb_tail_segs: /* Jump to default if there are more than two segments remaining. */ switch (nb_segs) { default: - lkey = mlx4_txq_mp2mr(txq, mlx4_txq_mb2mp(sbuf)); + lkey = mlx4_tx_mb2mr(txq, sbuf); if (unlikely(lkey == (uint32_t)-1)) { DEBUG("%p: unable to get MP <-> MR association", (void *)txq); @@ -485,7 +813,7 @@ txbb_tail_segs: nb_segs--; /* fallthrough */ case 2: - lkey = mlx4_txq_mp2mr(txq, mlx4_txq_mb2mp(sbuf)); + lkey = mlx4_tx_mb2mr(txq, sbuf); if (unlikely(lkey == (uint32_t)-1)) { DEBUG("%p: unable to get MP <-> MR association", (void *)txq); @@ -501,7 +829,7 @@ txbb_tail_segs: nb_segs--; /* fallthrough */ case 1: - lkey = mlx4_txq_mp2mr(txq, mlx4_txq_mb2mp(sbuf)); + lkey = mlx4_tx_mb2mr(txq, sbuf); if (unlikely(lkey == (uint32_t)-1)) { DEBUG("%p: unable to get MP <-> MR association", (void *)txq); @@ -587,6 +915,7 @@ mlx4_tx_burst(void *dpdk_txq, struct rte_mbuf **pkts, uint16_t pkts_n) uint16_t flags16[2]; } srcrb; uint32_t lkey; + bool tso = txq->priv->tso && (buf->ol_flags & PKT_TX_TCP_SEG); /* Clean up old buffer. */ if (likely(elt->buf != NULL)) { @@ -605,13 +934,22 @@ mlx4_tx_burst(void *dpdk_txq, struct rte_mbuf **pkts, uint16_t pkts_n) } while (tmp != NULL); } RTE_MBUF_PREFETCH_TO_FREE(elt_next->buf); - if (buf->nb_segs == 1) { + if (tso) { + /* Change opcode to TSO */ + owner_opcode &= ~MLX4_OPCODE_CONFIG_CMD; + owner_opcode |= MLX4_OPCODE_LSO | MLX4_WQE_CTRL_RR; + ctrl_next = mlx4_tx_burst_tso(buf, txq, ctrl); + if (!ctrl_next) { + elt->buf = NULL; + break; + } + } else if (buf->nb_segs == 1) { /* Validate WQE space in the send queue. */ if (sq->remain_size < MLX4_TXBB_SIZE) { elt->buf = NULL; break; } - lkey = mlx4_txq_mp2mr(txq, mlx4_txq_mb2mp(buf)); + lkey = mlx4_tx_mb2mr(txq, buf); if (unlikely(lkey == (uint32_t)-1)) { /* MR does not exist. */ DEBUG("%p: unable to get MP <-> MR association", @@ -639,7 +977,7 @@ mlx4_tx_burst(void *dpdk_txq, struct rte_mbuf **pkts, uint16_t pkts_n) ctrl_next = (volatile struct mlx4_wqe_ctrl_seg *) ((volatile uint8_t *)ctrl_next - sq->size); /* Flip HW valid ownership. */ - sq->owner_opcode ^= 0x1 << MLX4_SQ_OWNER_BIT; + sq->owner_opcode ^= 1u << MLX4_SQ_OWNER_BIT; } /* * For raw Ethernet, the SOLICIT flag is used to indicate @@ -746,11 +1084,13 @@ rxq_cq_to_pkt_type(volatile struct mlx4_cqe *cqe, * bit[4] - MLX4_CQE_STATUS_TCP * bit[3] - MLX4_CQE_STATUS_IPV4OPT * bit[2] - MLX4_CQE_STATUS_IPV6 - * bit[1] - MLX4_CQE_STATUS_IPV4F + * bit[1] - MLX4_CQE_STATUS_IPF * bit[0] - MLX4_CQE_STATUS_IPV4 * giving a total of up to 256 entries. */ idx |= ((status & MLX4_CQE_STATUS_PTYPE_MASK) >> 22); + if (status & MLX4_CQE_STATUS_IPV6) + idx |= ((status & MLX4_CQE_STATUS_IPV6F) >> 11); return mlx4_ptype_table[idx]; } @@ -934,11 +1274,14 @@ mlx4_rx_burst(void *dpdk_rxq, struct rte_mbuf **pkts, uint16_t pkts_n) goto skip; } pkt = seg; + assert(len >= (rxq->crc_present << 2)); /* Update packet information. */ pkt->packet_type = rxq_cq_to_pkt_type(cqe, rxq->l2tun_offload); pkt->ol_flags = PKT_RX_RSS_HASH; pkt->hash.rss = cqe->immed_rss_invalid; + if (rxq->crc_present) + len -= ETHER_CRC_LEN; pkt->pkt_len = len; if (rxq->csum | rxq->csum_l2tun) { uint32_t flags = @@ -963,6 +1306,9 @@ mlx4_rx_burst(void *dpdk_rxq, struct rte_mbuf **pkts, uint16_t pkts_n) * changes. */ scat->addr = rte_cpu_to_be_64(rte_pktmbuf_mtod(rep, uintptr_t)); + /* If there's only one MR, no need to replace LKey in WQE. */ + if (unlikely(mlx4_mr_btree_len(&rxq->mr_ctrl.cache_bh) > 1)) + scat->lkey = mlx4_rx_mb2mr(rxq, rep); if (len > seg->data_len) { len -= seg->data_len; ++pkt->nb_segs; |