X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=drivers%2Fnet%2Fmlx5%2Fmlx5_rxtx_vec.h;h=4220b08dd21c6f951ae7e8fc5701d54f0f95c43f;hb=f6d9ab4e769f0f95ecac1b418106e9f8137ca60c;hp=44856bbffec7db1b8460e99145f55dfb0ce6de6a;hpb=8fd92a66c60a7310cf5ab91996b9b09447512a61;p=dpdk.git

diff --git a/drivers/net/mlx5/mlx5_rxtx_vec.h b/drivers/net/mlx5/mlx5_rxtx_vec.h
index 44856bbffe..4220b08dd2 100644
--- a/drivers/net/mlx5/mlx5_rxtx_vec.h
+++ b/drivers/net/mlx5/mlx5_rxtx_vec.h
@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: BSD-3-Clause
  * Copyright 2017 6WIND S.A.
- * Copyright 2017 Mellanox.
+ * Copyright 2017 Mellanox Technologies, Ltd
  */
 
 #ifndef RTE_PMD_MLX5_RXTX_VEC_H_
@@ -22,6 +22,7 @@
 /* HW offload capabilities of vectorized Tx. */
 #define MLX5_VEC_TX_OFFLOAD_CAP \
 	(MLX5_VEC_TX_CKSUM_OFFLOAD_CAP | \
+	 DEV_TX_OFFLOAD_MATCH_METADATA | \
 	 DEV_TX_OFFLOAD_MULTI_SEGS)
 
 /*
@@ -87,21 +88,41 @@ mlx5_rx_replenish_bulk_mbuf(struct mlx5_rxq_data *rxq, uint16_t n)
 	const uint16_t q_mask = q_n - 1;
 	uint16_t elts_idx = rxq->rq_ci & q_mask;
 	struct rte_mbuf **elts = &(*rxq->elts)[elts_idx];
-	volatile struct mlx5_wqe_data_seg *wq = &(*rxq->wqes)[elts_idx];
+	volatile struct mlx5_wqe_data_seg *wq =
+		&((volatile struct mlx5_wqe_data_seg *)rxq->wqes)[elts_idx];
 	unsigned int i;
 
-	assert(n >= MLX5_VPMD_RXQ_RPLNSH_THRESH);
+	assert(n >= MLX5_VPMD_RXQ_RPLNSH_THRESH(q_n));
 	assert(n <= (uint16_t)(q_n - (rxq->rq_ci - rxq->rq_pi)));
-	assert(MLX5_VPMD_RXQ_RPLNSH_THRESH > MLX5_VPMD_DESCS_PER_LOOP);
+	assert(MLX5_VPMD_RXQ_RPLNSH_THRESH(q_n) > MLX5_VPMD_DESCS_PER_LOOP);
 	/* Not to cross queue end. */
 	n = RTE_MIN(n - MLX5_VPMD_DESCS_PER_LOOP, q_n - elts_idx);
 	if (rte_mempool_get_bulk(rxq->mp, (void *)elts, n) < 0) {
 		rxq->stats.rx_nombuf += n;
 		return;
 	}
-	for (i = 0; i < n; ++i)
-		wq[i].addr = rte_cpu_to_be_64((uintptr_t)elts[i]->buf_addr +
+	for (i = 0; i < n; ++i) {
+		void *buf_addr;
+
+		/*
+		 * Load the virtual address for Rx WQE. non-x86 processors
+		 * (mostly RISC such as ARM and Power) are more vulnerable to
+		 * load stall. For x86, reducing the number of instructions
+		 * seems to matter most.
+		 */
+#ifdef RTE_ARCH_X86_64
+		buf_addr = elts[i]->buf_addr;
+		assert(buf_addr == rte_mbuf_buf_addr(elts[i], rxq->mp));
+#else
+		buf_addr = rte_mbuf_buf_addr(elts[i], rxq->mp);
+		assert(buf_addr == elts[i]->buf_addr);
+#endif
+		wq[i].addr = rte_cpu_to_be_64((uintptr_t)buf_addr +
 					      RTE_PKTMBUF_HEADROOM);
+		/* 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))
+			wq[i].lkey = mlx5_rx_mb2mr(rxq, elts[i]);
+	}
 	rxq->rq_ci += n;
 	/* Prevent overflowing into consumed mbufs. */
 	elts_idx = rxq->rq_ci & q_mask;