[dpdk-dev] [PATCH v2 12/14] net/avf: enable sse vector Rx Tx func

Jingjing Wu jingjing.wu at intel.com
Fri Nov 24 07:33:24 CET 2017


Signed-off-by: Jingjing Wu <jingjing.wu at intel.com>
---
 config/common_base                    |   1 +
 drivers/net/avf/Makefile              |   1 +
 drivers/net/avf/avf.h                 |   4 +
 drivers/net/avf/avf_ethdev.c          |  11 +
 drivers/net/avf/avf_rxtx.c            | 178 ++++++++-
 drivers/net/avf/avf_rxtx.h            |  34 ++
 drivers/net/avf/avf_rxtx_vec_common.h | 238 ++++++++++++
 drivers/net/avf/avf_rxtx_vec_sse.c    | 680 ++++++++++++++++++++++++++++++++++
 8 files changed, 1136 insertions(+), 11 deletions(-)
 create mode 100644 drivers/net/avf/avf_rxtx_vec_common.h
 create mode 100644 drivers/net/avf/avf_rxtx_vec_sse.c

diff --git a/config/common_base b/config/common_base
index b1f1c1c..cdb8735 100644
--- a/config/common_base
+++ b/config/common_base
@@ -233,6 +233,7 @@ CONFIG_RTE_LIBRTE_AVF_DEBUG_TX=n
 CONFIG_RTE_LIBRTE_AVF_DEBUG_TX_FREE=n
 CONFIG_RTE_LIBRTE_AVF_DEBUG_RX=n
 CONFIG_RTE_LIBRTE_AVF_16BYTE_RX_DESC=n
+CONFIG_RTE_LIBRTE_AVF_INC_VECTOR=y
 
 #
 # Compile burst-oriented Mellanox ConnectX-3 (MLX4) PMD
diff --git a/drivers/net/avf/Makefile b/drivers/net/avf/Makefile
index 6193fa9..ff9d523 100644
--- a/drivers/net/avf/Makefile
+++ b/drivers/net/avf/Makefile
@@ -58,5 +58,6 @@ SRCS-$(CONFIG_RTE_LIBRTE_AVF_PMD) += avf_common.c
 SRCS-$(CONFIG_RTE_LIBRTE_AVF_PMD) += avf_ethdev.c
 SRCS-$(CONFIG_RTE_LIBRTE_AVF_PMD) += avf_vchnl.c
 SRCS-$(CONFIG_RTE_LIBRTE_AVF_PMD) += avf_rxtx.c
+SRCS-$(CONFIG_RTE_LIBRTE_AVF_INC_VECTOR) += avf_rxtx_vec_sse.c
 
 include $(RTE_SDK)/mk/rte.lib.mk
diff --git a/drivers/net/avf/avf.h b/drivers/net/avf/avf.h
index f39bebc..d4c275e 100644
--- a/drivers/net/avf/avf.h
+++ b/drivers/net/avf/avf.h
@@ -146,6 +146,10 @@ struct avf_adapter {
 	struct avf_hw hw;
 	struct rte_eth_dev *eth_dev;
 	struct avf_info vf;
+
+	/* For vector PMD */
+	bool rx_vec_allowed;
+	bool tx_vec_allowed;
 };
 
 /* AVF_DEV_PRIVATE_TO */
diff --git a/drivers/net/avf/avf_ethdev.c b/drivers/net/avf/avf_ethdev.c
index 8f382ff..1e8d9c0 100644
--- a/drivers/net/avf/avf_ethdev.c
+++ b/drivers/net/avf/avf_ethdev.c
@@ -149,6 +149,17 @@ avf_dev_configure(struct rte_eth_dev *dev)
 		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
 	struct rte_eth_conf *dev_conf = &dev->data->dev_conf;
 
+#ifdef RTE_LIBRTE_AVF_INC_VECTOR
+	/* Initialize to TRUE. If any of Rx queues doesn't meet the
+	 * vector Rx/Tx preconditions, it will be reset.
+	 */
+	ad->rx_vec_allowed = true;
+	ad->tx_vec_allowed = true;
+#else
+	ad->rx_vec_allowed = false;
+	ad->tx_vec_allowed = false;
+#endif
+
 	/* Vlan stripping setting */
 	if (dev_conf->rxmode.offloads & DEV_RX_OFFLOAD_VLAN_STRIP)
 		avf_enable_vlan_strip(ad);
diff --git a/drivers/net/avf/avf_rxtx.c b/drivers/net/avf/avf_rxtx.c
index 8e79efd..079d49b 100644
--- a/drivers/net/avf/avf_rxtx.c
+++ b/drivers/net/avf/avf_rxtx.c
@@ -121,6 +121,38 @@ check_tx_thresh(uint16_t nb_desc, uint16_t tx_rs_thresh,
 	return 0;
 }
 
+#ifdef RTE_LIBRTE_AVF_INC_VECTOR
+static inline bool
+check_rx_vec_allow(struct avf_rx_queue *rxq)
+{
+	if (rxq->rx_free_thresh >= AVF_VPMD_RX_MAX_BURST &&
+	    rxq->nb_rx_desc % rxq->rx_free_thresh == 0) {
+		PMD_INIT_LOG(DEBUG, "Vector Rx"
+				    " can be enabled on this rxq.");
+		return TRUE;
+	}
+
+	PMD_INIT_LOG(DEBUG, "Vector Rx"
+			    " cannot be enabled on this rxq.");
+	return FALSE;
+}
+
+static inline bool
+check_tx_vec_allow(struct avf_tx_queue *txq)
+{
+	if (((txq->txq_flags & AVF_SIMPLE_FLAGS) == AVF_SIMPLE_FLAGS) &&
+	    (txq->rs_thresh >= AVF_VPMD_TX_MAX_BURST) &&
+	    (txq->rs_thresh <= AVF_VPMD_TX_MAX_FREE_BUF)) {
+		PMD_INIT_LOG(DEBUG, "Vector tx"
+			     " can be enabled on this txq.");
+		return TRUE;
+	}
+	PMD_INIT_LOG(DEBUG, "Vector Tx"
+			    " cannot be enabled on this txq.");
+	return FALSE;
+}
+#endif
+
 static inline void
 reset_rx_queue(struct avf_rx_queue *rxq)
 {
@@ -254,6 +286,14 @@ release_txq_mbufs(struct avf_tx_queue *txq)
 	}
 }
 
+static const struct avf_rxq_ops def_rxq_ops = {
+	.release_mbufs = release_rxq_mbufs,
+};
+
+static const struct avf_txq_ops def_txq_ops = {
+	.release_mbufs = release_txq_mbufs,
+};
+
 int
 avf_dev_rx_queue_setup(struct rte_eth_dev *dev, uint16_t queue_idx,
 		       uint16_t nb_desc, unsigned int socket_id,
@@ -353,7 +393,12 @@ avf_dev_rx_queue_setup(struct rte_eth_dev *dev, uint16_t queue_idx,
 	rxq->q_set = TRUE;
 	dev->data->rx_queues[queue_idx] = rxq;
 	rxq->qrx_tail = hw->hw_addr + AVF_QRX_TAIL1(rxq->queue_id);
+	rxq->ops = &def_rxq_ops;
 
+#ifdef RTE_LIBRTE_AVF_INC_VECTOR
+	if (check_rx_vec_allow(rxq) == FALSE)
+		ad->rx_vec_allowed = false;
+#endif
 	return 0;
 }
 
@@ -365,6 +410,8 @@ avf_dev_tx_queue_setup(struct rte_eth_dev *dev,
 		       const struct rte_eth_txconf *tx_conf)
 {
 	struct avf_hw *hw = AVF_DEV_PRIVATE_TO_HW(dev->data->dev_private);
+	struct avf_adapter *ad =
+		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
 	struct avf_tx_queue *txq;
 	const struct rte_memzone *mz;
 	uint32_t ring_size;
@@ -444,6 +491,12 @@ avf_dev_tx_queue_setup(struct rte_eth_dev *dev,
 	txq->q_set = TRUE;
 	dev->data->tx_queues[queue_idx] = txq;
 	txq->qtx_tail = hw->hw_addr + AVF_QTX_TAIL1(queue_idx);
+	txq->ops = &def_txq_ops;
+
+#ifdef RTE_LIBRTE_AVF_INC_VECTOR
+	if (check_tx_vec_allow(txq) == FALSE)
+		ad->tx_vec_allowed = false;
+#endif
 
 	return 0;
 }
@@ -542,7 +595,7 @@ avf_dev_rx_queue_stop(struct rte_eth_dev *dev, uint16_t rx_queue_id)
 	}
 
 	rxq = dev->data->rx_queues[rx_queue_id];
-	release_rxq_mbufs(rxq);
+	rxq->ops->release_mbufs(rxq);
 	reset_rx_queue(rxq);
 	dev->data->rx_queue_state[rx_queue_id] = RTE_ETH_QUEUE_STATE_STOPPED;
 
@@ -570,7 +623,7 @@ avf_dev_tx_queue_stop(struct rte_eth_dev *dev, uint16_t tx_queue_id)
 	}
 
 	txq = dev->data->tx_queues[tx_queue_id];
-	release_txq_mbufs(txq);
+	txq->ops->release_mbufs(txq);
 	reset_tx_queue(txq);
 	dev->data->tx_queue_state[tx_queue_id] = RTE_ETH_QUEUE_STATE_STOPPED;
 
@@ -585,7 +638,7 @@ avf_dev_rx_queue_release(void *rxq)
 	if (!q)
 		return;
 
-	release_rxq_mbufs(q);
+	q->ops->release_mbufs(q);
 	rte_free(q->sw_ring);
 	rte_memzone_free(q->mz);
 	rte_free(q);
@@ -599,7 +652,7 @@ avf_dev_tx_queue_release(void *txq)
 	if (!q)
 		return;
 
-	release_txq_mbufs(q);
+	q->ops->release_mbufs(q);
 	rte_free(q->sw_ring);
 	rte_memzone_free(q->mz);
 	rte_free(q);
@@ -623,7 +676,7 @@ avf_stop_queues(struct rte_eth_dev *dev)
 		txq = dev->data->tx_queues[i];
 		if (!txq)
 			continue;
-		release_txq_mbufs(txq);
+		txq->ops->release_mbufs(txq);
 		reset_tx_queue(txq);
 		dev->data->tx_queue_state[i] = RTE_ETH_QUEUE_STATE_STOPPED;
 	}
@@ -631,7 +684,7 @@ avf_stop_queues(struct rte_eth_dev *dev)
 		rxq = dev->data->rx_queues[i];
 		if (!rxq)
 			continue;
-		release_rxq_mbufs(rxq);
+		rxq->ops->release_mbufs(rxq);
 		reset_rx_queue(rxq);
 		dev->data->rx_queue_state[i] = RTE_ETH_QUEUE_STATE_STOPPED;
 	}
@@ -1348,6 +1401,28 @@ avf_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
 	return nb_tx;
 }
 
+static uint16_t
+avf_xmit_pkts_vec(void *tx_queue, struct rte_mbuf **tx_pkts,
+		  uint16_t nb_pkts)
+{
+	uint16_t nb_tx = 0;
+	struct avf_tx_queue *txq = (struct avf_tx_queue *)tx_queue;
+
+	while (nb_pkts) {
+		uint16_t ret, num;
+
+		num = (uint16_t)RTE_MIN(nb_pkts, txq->rs_thresh);
+		ret = avf_xmit_fixed_burst_vec(tx_queue, &tx_pkts[nb_tx],
+						num);
+		nb_tx += ret;
+		nb_pkts -= ret;
+		if (ret < num)
+			break;
+	}
+
+	return nb_tx;
+}
+
 /* TX prep functions */
 uint16_t
 avf_prep_pkts(__rte_unused void *tx_queue, struct rte_mbuf **tx_pkts,
@@ -1400,18 +1475,64 @@ avf_prep_pkts(__rte_unused void *tx_queue, struct rte_mbuf **tx_pkts,
 void
 avf_set_rx_function(struct rte_eth_dev *dev)
 {
-	if (dev->data->scattered_rx)
+	struct avf_adapter *adapter =
+		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
+	struct avf_rx_queue *rxq;
+	int i;
+
+	if (adapter->rx_vec_allowed) {
+		if (dev->data->scattered_rx) {
+			PMD_DRV_LOG(DEBUG, "Using Vector Scattered Rx callback"
+				    " (port=%d).", dev->data->port_id);
+			dev->rx_pkt_burst = avf_recv_scattered_pkts_vec;
+		} else {
+			PMD_DRV_LOG(DEBUG, "Using Vector Rx callback"
+				    " (port=%d).", dev->data->port_id);
+			dev->rx_pkt_burst = avf_recv_pkts_vec;
+		}
+		for (i = 0; i < dev->data->nb_rx_queues; i++) {
+			rxq = dev->data->rx_queues[i];
+			if (!rxq)
+				continue;
+			avf_rxq_vec_setup(rxq);
+		}
+	} else if (dev->data->scattered_rx) {
+		PMD_DRV_LOG(DEBUG, "Using a Scattered Rx callback (port=%d).",
+			    dev->data->port_id);
 		dev->rx_pkt_burst = avf_recv_scattered_pkts;
-	else
+	} else {
+		PMD_DRV_LOG(DEBUG, "Using Basic Rx callback (port=%d).",
+			    dev->data->port_id);
 		dev->rx_pkt_burst = avf_recv_pkts;
+	}
 }
 
-/* choose rx function*/
+/* choose tx function*/
 void
 avf_set_tx_function(struct rte_eth_dev *dev)
 {
-	dev->tx_pkt_burst = avf_xmit_pkts;
-	dev->tx_pkt_prepare = avf_prep_pkts;
+	struct avf_adapter *adapter =
+		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
+	struct avf_tx_queue *txq;
+	int i;
+
+	if (adapter->tx_vec_allowed) {
+		PMD_DRV_LOG(DEBUG, "Using Vector Tx callback (port=%d).",
+			    dev->data->port_id);
+		dev->tx_pkt_burst = avf_xmit_pkts_vec;
+		dev->tx_pkt_prepare = NULL;
+		for (i = 0; i < dev->data->nb_tx_queues; i++) {
+			txq = dev->data->tx_queues[i];
+			if (!txq)
+				continue;
+			avf_txq_vec_setup(txq);
+		}
+	} else {
+		PMD_DRV_LOG(DEBUG, "Using Basic Tx callback (port=%d).",
+			    dev->data->port_id);
+		dev->tx_pkt_burst = avf_xmit_pkts;
+		dev->tx_pkt_prepare = avf_prep_pkts;
+	}
 }
 
 void
@@ -1533,3 +1654,38 @@ avf_dev_tx_desc_status(void *tx_queue, uint16_t offset)
 
 	return RTE_ETH_TX_DESC_FULL;
 }
+
+uint16_t __attribute__((weak))
+avf_recv_pkts_vec(void __rte_unused *rx_queue,
+		  struct rte_mbuf __rte_unused **rx_pkts,
+		  uint16_t __rte_unused nb_pkts)
+{
+	return 0;
+}
+
+uint16_t __attribute__((weak))
+avf_recv_scattered_pkts_vec(void __rte_unused *rx_queue,
+	struct rte_mbuf __rte_unused **rx_pkts,
+	uint16_t __rte_unused nb_pkts)
+{
+	return 0;
+}
+
+uint16_t __attribute__((weak))
+avf_xmit_fixed_burst_vec(void __rte_unused * tx_queue,
+			  struct rte_mbuf __rte_unused **tx_pkts,
+			  uint16_t __rte_unused nb_pkts)
+{
+	return 0;
+}
+int __attribute__((weak))
+avf_rxq_vec_setup(struct avf_rx_queue __rte_unused *rxq)
+{
+	return -1;
+}
+
+int __attribute__((weak))
+avf_txq_vec_setup(struct avf_tx_queue __rte_unused *txq)
+{
+	return -1;
+}
diff --git a/drivers/net/avf/avf_rxtx.h b/drivers/net/avf/avf_rxtx.h
index 8e1025c..0246a73 100644
--- a/drivers/net/avf/avf_rxtx.h
+++ b/drivers/net/avf/avf_rxtx.h
@@ -45,6 +45,15 @@
 /* used for Rx Bulk Allocate */
 #define AVF_RX_MAX_BURST         32
 
+/* used for Vector PMD */
+#define AVF_VPMD_RX_MAX_BURST    32
+#define AVF_VPMD_TX_MAX_BURST    32
+#define AVF_VPMD_DESCS_PER_LOOP  4
+#define AVF_VPMD_TX_MAX_FREE_BUF 64
+
+#define AVF_SIMPLE_FLAGS ((uint32_t)ETH_TXQ_FLAGS_NOMULTSEGS | \
+			  ETH_TXQ_FLAGS_NOOFFLOADS)
+
 #define DEFAULT_TX_RS_THRESH     32
 #define DEFAULT_TX_FREE_THRESH   32
 
@@ -74,6 +83,14 @@
 #define avf_rx_desc avf_32byte_rx_desc
 #endif
 
+struct avf_rxq_ops {
+	void (*release_mbufs)(struct avf_rx_queue *rxq);
+};
+
+struct avf_txq_ops {
+	void (*release_mbufs)(struct avf_tx_queue *txq);
+};
+
 /* Structure associated with each Rx queue. */
 struct avf_rx_queue {
 	struct rte_mempool *mp;       /* mbuf pool to populate Rx ring */
@@ -90,6 +107,11 @@ struct avf_rx_queue {
 	struct rte_mbuf *pkt_last_seg;  /* last segment of current packet */
 	struct rte_mbuf fake_mbuf;      /* dummy mbuf */
 
+	/* used for VPMD */
+	uint16_t rxrearm_nb;       /* number of remaining to be re-armed */
+	uint16_t rxrearm_start;    /* the idx we start the re-arming from */
+	uint64_t mbuf_initializer; /* value to init mbufs */
+
 	uint8_t port_id;        /* device port ID */
 	uint8_t crc_len;        /* 0 if CRC stripped, 4 otherwise */
 	uint16_t queue_id;      /* Rx queue index */
@@ -99,6 +121,7 @@ struct avf_rx_queue {
 
 	bool q_set;             /* if rx queue has been configured */
 	bool rx_deferred_start; /* don't start this queue in dev start */
+	const struct avf_rxq_ops *ops;
 };
 
 struct avf_tx_entry {
@@ -130,6 +153,7 @@ struct avf_tx_queue {
 
 	bool q_set;                    /* if rx queue has been configured */
 	bool tx_deferred_start;        /* don't start this queue in dev start */
+	const struct avf_txq_ops *ops;
 };
 
 /* Offload features */
@@ -183,6 +207,16 @@ uint32_t avf_dev_rxq_count(struct rte_eth_dev *dev, uint16_t queue_id);
 int avf_dev_rx_desc_status(void *rx_queue, uint16_t offset);
 int avf_dev_tx_desc_status(void *tx_queue, uint16_t offset);
 
+uint16_t avf_recv_pkts_vec(void *rx_queue, struct rte_mbuf **rx_pkts,
+			   uint16_t nb_pkts);
+uint16_t avf_recv_scattered_pkts_vec(void *rx_queue,
+				     struct rte_mbuf **rx_pkts,
+				     uint16_t nb_pkts);
+uint16_t avf_xmit_fixed_burst_vec(void *tx_queue, struct rte_mbuf **tx_pkts,
+				  uint16_t nb_pkts);
+int avf_rxq_vec_setup(struct avf_rx_queue *rxq);
+int avf_txq_vec_setup(struct avf_tx_queue *txq);
+
 static inline
 void avf_dump_rx_descriptor(struct avf_rx_queue *rxq,
 			    const void *desc,
diff --git a/drivers/net/avf/avf_rxtx_vec_common.h b/drivers/net/avf/avf_rxtx_vec_common.h
new file mode 100644
index 0000000..726b68e
--- /dev/null
+++ b/drivers/net/avf/avf_rxtx_vec_common.h
@@ -0,0 +1,238 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright(c) 2017 Intel Corporation. All rights reserved.
+ *   All rights reserved.
+ *
+ *   Redistribution and use in source and binary forms, with or without
+ *   modification, are permitted provided that the following conditions
+ *   are met:
+ *
+ *     * Redistributions of source code must retain the above copyright
+ *       notice, this list of conditions and the following disclaimer.
+ *     * Redistributions in binary form must reproduce the above copyright
+ *       notice, this list of conditions and the following disclaimer in
+ *       the documentation and/or other materials provided with the
+ *       distribution.
+ *     * Neither the name of Intel Corporation nor the names of its
+ *       contributors may be used to endorse or promote products derived
+ *       from this software without specific prior written permission.
+ *
+ *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef _AVF_RXTX_VEC_COMMON_H_
+#define _AVF_RXTX_VEC_COMMON_H_
+#include <stdint.h>
+#include <rte_ethdev.h>
+#include <rte_malloc.h>
+
+#include "avf.h"
+#include "avf_rxtx.h"
+
+static inline uint16_t
+reassemble_packets(struct avf_rx_queue *rxq, struct rte_mbuf **rx_bufs,
+		   uint16_t nb_bufs, uint8_t *split_flags)
+{
+	struct rte_mbuf *pkts[AVF_VPMD_RX_MAX_BURST];
+	struct rte_mbuf *start = rxq->pkt_first_seg;
+	struct rte_mbuf *end =  rxq->pkt_last_seg;
+	unsigned pkt_idx, buf_idx;
+
+	for (buf_idx = 0, pkt_idx = 0; buf_idx < nb_bufs; buf_idx++) {
+		if (end != NULL) {
+			/* processing a split packet */
+			end->next = rx_bufs[buf_idx];
+			rx_bufs[buf_idx]->data_len += rxq->crc_len;
+
+			start->nb_segs++;
+			start->pkt_len += rx_bufs[buf_idx]->data_len;
+			end = end->next;
+
+			if (!split_flags[buf_idx]) {
+				/* it's the last packet of the set */
+				start->hash = end->hash;
+				start->ol_flags = end->ol_flags;
+				/* we need to strip crc for the whole packet */
+				start->pkt_len -= rxq->crc_len;
+				if (end->data_len > rxq->crc_len)
+					end->data_len -= rxq->crc_len;
+				else {
+					/* free up last mbuf */
+					struct rte_mbuf *secondlast = start;
+
+					start->nb_segs--;
+					while (secondlast->next != end)
+						secondlast = secondlast->next;
+					secondlast->data_len -= (rxq->crc_len -
+							end->data_len);
+					secondlast->next = NULL;
+					rte_pktmbuf_free_seg(end);
+				}
+				pkts[pkt_idx++] = start;
+				start = end = NULL;
+			}
+		} else {
+			/* not processing a split packet */
+			if (!split_flags[buf_idx]) {
+				/* not a split packet, save and skip */
+				pkts[pkt_idx++] = rx_bufs[buf_idx];
+				continue;
+			}
+			end = start = rx_bufs[buf_idx];
+			rx_bufs[buf_idx]->data_len += rxq->crc_len;
+			rx_bufs[buf_idx]->pkt_len += rxq->crc_len;
+		}
+	}
+
+	/* save the partial packet for next time */
+	rxq->pkt_first_seg = start;
+	rxq->pkt_last_seg = end;
+	memcpy(rx_bufs, pkts, pkt_idx * (sizeof(*pkts)));
+	return pkt_idx;
+}
+
+static __rte_always_inline int
+avf_tx_free_bufs(struct avf_tx_queue *txq)
+{
+	struct avf_tx_entry *txep;
+	uint32_t n;
+	uint32_t i;
+	int nb_free = 0;
+	struct rte_mbuf *m, *free[AVF_VPMD_TX_MAX_FREE_BUF];
+
+	/* check DD bits on threshold descriptor */
+	if ((txq->tx_ring[txq->next_dd].cmd_type_offset_bsz &
+			rte_cpu_to_le_64(AVF_TXD_QW1_DTYPE_MASK)) !=
+			rte_cpu_to_le_64(AVF_TX_DESC_DTYPE_DESC_DONE))
+		return 0;
+
+	n = txq->rs_thresh;
+
+	 /* first buffer to free from S/W ring is at index
+	  * tx_next_dd - (tx_rs_thresh-1)
+	  */
+	txep = &txq->sw_ring[txq->next_dd - (n - 1)];
+	m = rte_pktmbuf_prefree_seg(txep[0].mbuf);
+	if (likely(m != NULL)) {
+		free[0] = m;
+		nb_free = 1;
+		for (i = 1; i < n; i++) {
+			m = rte_pktmbuf_prefree_seg(txep[i].mbuf);
+			if (likely(m != NULL)) {
+				if (likely(m->pool == free[0]->pool)) {
+					free[nb_free++] = m;
+				} else {
+					rte_mempool_put_bulk(free[0]->pool,
+							     (void *)free,
+							     nb_free);
+					free[0] = m;
+					nb_free = 1;
+				}
+			}
+		}
+		rte_mempool_put_bulk(free[0]->pool, (void **)free, nb_free);
+	} else {
+		for (i = 1; i < n; i++) {
+			m = rte_pktmbuf_prefree_seg(txep[i].mbuf);
+			if (m != NULL)
+				rte_mempool_put(m->pool, m);
+		}
+	}
+
+	/* buffers were freed, update counters */
+	txq->nb_free = (uint16_t)(txq->nb_free + txq->rs_thresh);
+	txq->next_dd = (uint16_t)(txq->next_dd + txq->rs_thresh);
+	if (txq->next_dd >= txq->nb_tx_desc)
+		txq->next_dd = (uint16_t)(txq->rs_thresh - 1);
+
+	return txq->rs_thresh;
+}
+
+static __rte_always_inline void
+tx_backlog_entry(struct avf_tx_entry *txep,
+		 struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
+{
+	int i;
+
+	for (i = 0; i < (int)nb_pkts; ++i)
+		txep[i].mbuf = tx_pkts[i];
+}
+
+static inline void
+_avf_rx_queue_release_mbufs_vec(struct avf_rx_queue *rxq)
+{
+	const unsigned mask = rxq->nb_rx_desc - 1;
+	unsigned i;
+
+	if (!rxq->sw_ring || rxq->rxrearm_nb >= rxq->nb_rx_desc)
+		return;
+
+	/* free all mbufs that are valid in the ring */
+	if (rxq->rxrearm_nb == 0) {
+		for (i = 0; i < rxq->nb_rx_desc; i++) {
+			if (rxq->sw_ring[i] != NULL)
+				rte_pktmbuf_free_seg(rxq->sw_ring[i]);
+		}
+	} else {
+		for (i = rxq->rx_tail;
+		     i != rxq->rxrearm_start;
+		     i = (i + 1) & mask) {
+			if (rxq->sw_ring[i] != NULL)
+				rte_pktmbuf_free_seg(rxq->sw_ring[i]);
+		}
+	}
+
+	rxq->rxrearm_nb = rxq->nb_rx_desc;
+
+	/* set all entries to NULL */
+	memset(rxq->sw_ring, 0, sizeof(rxq->sw_ring[0]) * rxq->nb_rx_desc);
+}
+
+static inline void
+_avf_tx_queue_release_mbufs_vec(struct avf_tx_queue *txq)
+{
+	unsigned i;
+	const uint16_t max_desc = (uint16_t)(txq->nb_tx_desc - 1);
+
+	if (!txq->sw_ring || txq->nb_free == max_desc)
+		return;
+
+	i = txq->next_dd - txq->rs_thresh + 1;
+	if (txq->tx_tail < i) {
+		for (; i < txq->nb_tx_desc; i++) {
+			rte_pktmbuf_free_seg(txq->sw_ring[i].mbuf);
+			txq->sw_ring[i].mbuf = NULL;
+		}
+		i = 0;
+	}
+}
+
+static inline int
+avf_rxq_vec_setup_default(struct avf_rx_queue *rxq)
+{
+	uintptr_t p;
+	struct rte_mbuf mb_def = { .buf_addr = 0 }; /* zeroed mbuf */
+
+	mb_def.nb_segs = 1;
+	mb_def.data_off = RTE_PKTMBUF_HEADROOM;
+	mb_def.port = rxq->port_id;
+	rte_mbuf_refcnt_set(&mb_def, 1);
+
+	/* prevent compiler reordering: rearm_data covers previous fields */
+	rte_compiler_barrier();
+	p = (uintptr_t)&mb_def.rearm_data;
+	rxq->mbuf_initializer = *(uint64_t *)p;
+	return 0;
+}
+#endif
diff --git a/drivers/net/avf/avf_rxtx_vec_sse.c b/drivers/net/avf/avf_rxtx_vec_sse.c
new file mode 100644
index 0000000..dce55a1
--- /dev/null
+++ b/drivers/net/avf/avf_rxtx_vec_sse.c
@@ -0,0 +1,680 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright(c) 2017 Intel Corporation. All rights reserved.
+ *   All rights reserved.
+ *
+ *   Redistribution and use in source and binary forms, with or without
+ *   modification, are permitted provided that the following conditions
+ *   are met:
+ *
+ *     * Redistributions of source code must retain the above copyright
+ *       notice, this list of conditions and the following disclaimer.
+ *     * Redistributions in binary form must reproduce the above copyright
+ *       notice, this list of conditions and the following disclaimer in
+ *       the documentation and/or other materials provided with the
+ *       distribution.
+ *     * Neither the name of Intel Corporation nor the names of its
+ *       contributors may be used to endorse or promote products derived
+ *       from this software without specific prior written permission.
+ *
+ *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <stdint.h>
+#include <rte_ethdev.h>
+#include <rte_malloc.h>
+
+#include "base/avf_prototype.h"
+#include "base/avf_type.h"
+#include "avf.h"
+#include "avf_rxtx.h"
+#include "avf_rxtx_vec_common.h"
+
+#include <tmmintrin.h>
+
+#ifndef __INTEL_COMPILER
+#pragma GCC diagnostic ignored "-Wcast-qual"
+#endif
+
+static inline void
+avf_rxq_rearm(struct avf_rx_queue *rxq)
+{
+	int i;
+	uint16_t rx_id;
+	volatile union avf_rx_desc *rxdp;
+	struct rte_mbuf **rxp = &rxq->sw_ring[rxq->rxrearm_start];
+	struct rte_mbuf *mb0, *mb1;
+	__m128i hdr_room = _mm_set_epi64x(RTE_PKTMBUF_HEADROOM,
+			RTE_PKTMBUF_HEADROOM);
+	__m128i dma_addr0, dma_addr1;
+
+	rxdp = rxq->rx_ring + rxq->rxrearm_start;
+
+	/* Pull 'n' more MBUFs into the software ring */
+	if (rte_mempool_get_bulk(rxq->mp, (void *)rxp,
+				 rxq->rx_free_thresh) < 0) {
+		if (rxq->rxrearm_nb + rxq->rx_free_thresh >= rxq->nb_rx_desc) {
+			dma_addr0 = _mm_setzero_si128();
+			for (i = 0; i < AVF_VPMD_DESCS_PER_LOOP; i++) {
+				rxp[i] = &rxq->fake_mbuf;
+				_mm_store_si128((__m128i *)&rxdp[i].read,
+						dma_addr0);
+			}
+		}
+		rte_eth_devices[rxq->port_id].data->rx_mbuf_alloc_failed +=
+			rxq->rx_free_thresh;
+		return;
+	}
+
+	/* Initialize the mbufs in vector, process 2 mbufs in one loop */
+	for (i = 0; i < rxq->rx_free_thresh; i += 2, rxp += 2) {
+		__m128i vaddr0, vaddr1;
+
+		mb0 = rxp[0];
+		mb1 = rxp[1];
+
+		/* load buf_addr(lo 64bit) and buf_iova(hi 64bit) */
+		RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, buf_iova) !=
+				offsetof(struct rte_mbuf, buf_addr) + 8);
+		vaddr0 = _mm_loadu_si128((__m128i *)&mb0->buf_addr);
+		vaddr1 = _mm_loadu_si128((__m128i *)&mb1->buf_addr);
+
+		/* convert pa to dma_addr hdr/data */
+		dma_addr0 = _mm_unpackhi_epi64(vaddr0, vaddr0);
+		dma_addr1 = _mm_unpackhi_epi64(vaddr1, vaddr1);
+
+		/* add headroom to pa values */
+		dma_addr0 = _mm_add_epi64(dma_addr0, hdr_room);
+		dma_addr1 = _mm_add_epi64(dma_addr1, hdr_room);
+
+		/* flush desc with pa dma_addr */
+		_mm_store_si128((__m128i *)&rxdp++->read, dma_addr0);
+		_mm_store_si128((__m128i *)&rxdp++->read, dma_addr1);
+	}
+
+	rxq->rxrearm_start += rxq->rx_free_thresh;
+	if (rxq->rxrearm_start >= rxq->nb_rx_desc)
+		rxq->rxrearm_start = 0;
+
+	rxq->rxrearm_nb -= rxq->rx_free_thresh;
+
+	rx_id = (uint16_t)((rxq->rxrearm_start == 0) ?
+			   (rxq->nb_rx_desc - 1) : (rxq->rxrearm_start - 1));
+
+	PMD_RX_LOG(DEBUG, "port_id=%u queue_id=%u rx_tail=%u "
+		   "rearm_start=%u rearm_nb=%u",
+		   rxq->port_id, rxq->queue_id,
+		   rx_id, rxq->rxrearm_start, rxq->rxrearm_nb);
+
+	/* Update the tail pointer on the NIC */
+	AVF_PCI_REG_WRITE(rxq->qrx_tail, rx_id);
+}
+
+static inline void
+desc_to_olflags_v(struct avf_rx_queue *rxq, __m128i descs[4],
+		  struct rte_mbuf **rx_pkts)
+{
+	const __m128i mbuf_init = _mm_set_epi64x(0, rxq->mbuf_initializer);
+	__m128i rearm0, rearm1, rearm2, rearm3;
+
+	__m128i vlan0, vlan1, rss, l3_l4e;
+
+	/* mask everything except RSS, flow director and VLAN flags
+	 * bit2 is for VLAN tag, bit11 for flow director indication
+	 * bit13:12 for RSS indication.
+	 */
+	const __m128i rss_vlan_msk = _mm_set_epi32(
+			0x1c03804, 0x1c03804, 0x1c03804, 0x1c03804);
+
+	const __m128i cksum_mask = _mm_set_epi32(
+			PKT_RX_IP_CKSUM_GOOD | PKT_RX_IP_CKSUM_BAD |
+			PKT_RX_L4_CKSUM_GOOD | PKT_RX_L4_CKSUM_BAD |
+			PKT_RX_EIP_CKSUM_BAD,
+			PKT_RX_IP_CKSUM_GOOD | PKT_RX_IP_CKSUM_BAD |
+			PKT_RX_L4_CKSUM_GOOD | PKT_RX_L4_CKSUM_BAD |
+			PKT_RX_EIP_CKSUM_BAD,
+			PKT_RX_IP_CKSUM_GOOD | PKT_RX_IP_CKSUM_BAD |
+			PKT_RX_L4_CKSUM_GOOD | PKT_RX_L4_CKSUM_BAD |
+			PKT_RX_EIP_CKSUM_BAD,
+			PKT_RX_IP_CKSUM_GOOD | PKT_RX_IP_CKSUM_BAD |
+			PKT_RX_L4_CKSUM_GOOD | PKT_RX_L4_CKSUM_BAD |
+			PKT_RX_EIP_CKSUM_BAD);
+
+	/* map rss and vlan type to rss hash and vlan flag */
+	const __m128i vlan_flags = _mm_set_epi8(0, 0, 0, 0,
+			0, 0, 0, 0,
+			0, 0, 0, PKT_RX_VLAN | PKT_RX_VLAN_STRIPPED,
+			0, 0, 0, 0);
+
+	const __m128i rss_flags = _mm_set_epi8(0, 0, 0, 0,
+			0, 0, 0, 0,
+			PKT_RX_RSS_HASH | PKT_RX_FDIR, PKT_RX_RSS_HASH, 0, 0,
+			0, 0, PKT_RX_FDIR, 0);
+
+	const __m128i l3_l4e_flags = _mm_set_epi8(0, 0, 0, 0, 0, 0, 0, 0,
+			/* shift right 1 bit to make sure it not exceed 255 */
+			(PKT_RX_EIP_CKSUM_BAD | PKT_RX_L4_CKSUM_BAD |
+			 PKT_RX_IP_CKSUM_BAD) >> 1,
+			(PKT_RX_IP_CKSUM_GOOD | PKT_RX_EIP_CKSUM_BAD |
+			 PKT_RX_L4_CKSUM_BAD) >> 1,
+			(PKT_RX_EIP_CKSUM_BAD | PKT_RX_IP_CKSUM_BAD) >> 1,
+			(PKT_RX_IP_CKSUM_GOOD | PKT_RX_EIP_CKSUM_BAD) >> 1,
+			(PKT_RX_L4_CKSUM_BAD | PKT_RX_IP_CKSUM_BAD) >> 1,
+			(PKT_RX_IP_CKSUM_GOOD | PKT_RX_L4_CKSUM_BAD) >> 1,
+			PKT_RX_IP_CKSUM_BAD >> 1,
+			(PKT_RX_IP_CKSUM_GOOD | PKT_RX_L4_CKSUM_GOOD) >> 1);
+
+	vlan0 = _mm_unpackhi_epi32(descs[0], descs[1]);
+	vlan1 = _mm_unpackhi_epi32(descs[2], descs[3]);
+	vlan0 = _mm_unpacklo_epi64(vlan0, vlan1);
+
+	vlan1 = _mm_and_si128(vlan0, rss_vlan_msk);
+	vlan0 = _mm_shuffle_epi8(vlan_flags, vlan1);
+
+	rss = _mm_srli_epi32(vlan1, 11);
+	rss = _mm_shuffle_epi8(rss_flags, rss);
+
+	l3_l4e = _mm_srli_epi32(vlan1, 22);
+	l3_l4e = _mm_shuffle_epi8(l3_l4e_flags, l3_l4e);
+	/* then we shift left 1 bit */
+	l3_l4e = _mm_slli_epi32(l3_l4e, 1);
+	/* we need to mask out the reduntant bits */
+	l3_l4e = _mm_and_si128(l3_l4e, cksum_mask);
+
+	vlan0 = _mm_or_si128(vlan0, rss);
+	vlan0 = _mm_or_si128(vlan0, l3_l4e);
+
+	/* At this point, we have the 4 sets of flags in the low 16-bits
+	 * of each 32-bit value in vlan0.
+	 * We want to extract these, and merge them with the mbuf init data
+	 * so we can do a single 16-byte write to the mbuf to set the flags
+	 * and all the other initialization fields. Extracting the
+	 * appropriate flags means that we have to do a shift and blend for
+	 * each mbuf before we do the write.
+	 */
+	rearm0 = _mm_blend_epi16(mbuf_init, _mm_slli_si128(vlan0, 8), 0x10);
+	rearm1 = _mm_blend_epi16(mbuf_init, _mm_slli_si128(vlan0, 4), 0x10);
+	rearm2 = _mm_blend_epi16(mbuf_init, vlan0, 0x10);
+	rearm3 = _mm_blend_epi16(mbuf_init, _mm_srli_si128(vlan0, 4), 0x10);
+
+	/* write the rearm data and the olflags in one write */
+	RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, ol_flags) !=
+			offsetof(struct rte_mbuf, rearm_data) + 8);
+	RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, rearm_data) !=
+			RTE_ALIGN(offsetof(struct rte_mbuf, rearm_data), 16));
+	_mm_store_si128((__m128i *)&rx_pkts[0]->rearm_data, rearm0);
+	_mm_store_si128((__m128i *)&rx_pkts[1]->rearm_data, rearm1);
+	_mm_store_si128((__m128i *)&rx_pkts[2]->rearm_data, rearm2);
+	_mm_store_si128((__m128i *)&rx_pkts[3]->rearm_data, rearm3);
+}
+
+#define PKTLEN_SHIFT     10
+
+static inline void
+desc_to_ptype_v(__m128i descs[4], struct rte_mbuf **rx_pkts)
+{
+	__m128i ptype0 = _mm_unpackhi_epi64(descs[0], descs[1]);
+	__m128i ptype1 = _mm_unpackhi_epi64(descs[2], descs[3]);
+	static const uint32_t type_table[UINT8_MAX + 1] __rte_cache_aligned = {
+		/* [0] reserved */
+		[1] = RTE_PTYPE_L2_ETHER,
+		/* [2] - [21] reserved */
+		[22] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+			RTE_PTYPE_L4_FRAG,
+		[23] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+			RTE_PTYPE_L4_NONFRAG,
+		[24] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+			RTE_PTYPE_L4_UDP,
+		/* [25] reserved */
+		[26] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+			RTE_PTYPE_L4_TCP,
+		[27] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+			RTE_PTYPE_L4_SCTP,
+		[28] = RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN |
+			RTE_PTYPE_L4_ICMP,
+		/* All others reserved */
+	};
+
+	ptype0 = _mm_srli_epi64(ptype0, 30);
+	ptype1 = _mm_srli_epi64(ptype1, 30);
+
+	rx_pkts[0]->packet_type = type_table[_mm_extract_epi8(ptype0, 0)];
+	rx_pkts[1]->packet_type = type_table[_mm_extract_epi8(ptype0, 8)];
+	rx_pkts[2]->packet_type = type_table[_mm_extract_epi8(ptype1, 0)];
+	rx_pkts[3]->packet_type = type_table[_mm_extract_epi8(ptype1, 8)];
+}
+
+/* Notice:
+ * - nb_pkts < AVF_VPMD_DESCS_PER_LOOP, just return no packet
+ * - nb_pkts > AVF_VPMD_RX_MAX_BURST, only scan AVF_VPMD_RX_MAX_BURST
+ *   numbers of DD bits
+ */
+static inline uint16_t
+_recv_raw_pkts_vec(struct avf_rx_queue *rxq, struct rte_mbuf **rx_pkts,
+		   uint16_t nb_pkts, uint8_t *split_packet)
+{
+	volatile union avf_rx_desc *rxdp;
+	struct rte_mbuf **sw_ring;
+	uint16_t nb_pkts_recd;
+	int pos;
+	uint64_t var;
+	__m128i shuf_msk;
+
+	__m128i crc_adjust = _mm_set_epi16(
+				0, 0, 0,    /* ignore non-length fields */
+				-rxq->crc_len, /* sub crc on data_len */
+				0,          /* ignore high-16bits of pkt_len */
+				-rxq->crc_len, /* sub crc on pkt_len */
+				0, 0            /* ignore pkt_type field */
+			);
+	/* compile-time check the above crc_adjust layout is correct.
+	 * NOTE: the first field (lowest address) is given last in set_epi16
+	 * call above.
+	 */
+	RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, pkt_len) !=
+			offsetof(struct rte_mbuf, rx_descriptor_fields1) + 4);
+	RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, data_len) !=
+			offsetof(struct rte_mbuf, rx_descriptor_fields1) + 8);
+	__m128i dd_check, eop_check;
+
+	/* nb_pkts shall be less equal than AVF_VPMD_RX_MAX_BURST */
+	nb_pkts = RTE_MIN(nb_pkts, AVF_VPMD_RX_MAX_BURST);
+
+	/* nb_pkts has to be floor-aligned to AVF_VPMD_DESCS_PER_LOOP */
+	nb_pkts = RTE_ALIGN_FLOOR(nb_pkts, AVF_VPMD_DESCS_PER_LOOP);
+
+	/* Just the act of getting into the function from the application is
+	 * going to cost about 7 cycles
+	 */
+	rxdp = rxq->rx_ring + rxq->rx_tail;
+
+	rte_prefetch0(rxdp);
+
+	/* See if we need to rearm the RX queue - gives the prefetch a bit
+	 * of time to act
+	 */
+	if (rxq->rxrearm_nb > rxq->rx_free_thresh)
+		avf_rxq_rearm(rxq);
+
+	/* Before we start moving massive data around, check to see if
+	 * there is actually a packet available
+	 */
+	if (!(rxdp->wb.qword1.status_error_len &
+	      rte_cpu_to_le_32(1 << AVF_RX_DESC_STATUS_DD_SHIFT)))
+		return 0;
+
+	/* 4 packets DD mask */
+	dd_check = _mm_set_epi64x(0x0000000100000001LL, 0x0000000100000001LL);
+
+	/* 4 packets EOP mask */
+	eop_check = _mm_set_epi64x(0x0000000200000002LL, 0x0000000200000002LL);
+
+	/* mask to shuffle from desc. to mbuf */
+	shuf_msk = _mm_set_epi8(
+		7, 6, 5, 4,  /* octet 4~7, 32bits rss */
+		3, 2,        /* octet 2~3, low 16 bits vlan_macip */
+		15, 14,      /* octet 15~14, 16 bits data_len */
+		0xFF, 0xFF,  /* skip high 16 bits pkt_len, zero out */
+		15, 14,      /* octet 15~14, low 16 bits pkt_len */
+		0xFF, 0xFF, 0xFF, 0xFF /* pkt_type set as unknown */
+		);
+	/* Compile-time verify the shuffle mask
+	 * NOTE: some field positions already verified above, but duplicated
+	 * here for completeness in case of future modifications.
+	 */
+	RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, pkt_len) !=
+			offsetof(struct rte_mbuf, rx_descriptor_fields1) + 4);
+	RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, data_len) !=
+			offsetof(struct rte_mbuf, rx_descriptor_fields1) + 8);
+	RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, vlan_tci) !=
+			offsetof(struct rte_mbuf, rx_descriptor_fields1) + 10);
+	RTE_BUILD_BUG_ON(offsetof(struct rte_mbuf, hash) !=
+			offsetof(struct rte_mbuf, rx_descriptor_fields1) + 12);
+
+	/* Cache is empty -> need to scan the buffer rings, but first move
+	 * the next 'n' mbufs into the cache
+	 */
+	sw_ring = &rxq->sw_ring[rxq->rx_tail];
+
+	/* A. load 4 packet in one loop
+	 * [A*. mask out 4 unused dirty field in desc]
+	 * B. copy 4 mbuf point from swring to rx_pkts
+	 * C. calc the number of DD bits among the 4 packets
+	 * [C*. extract the end-of-packet bit, if requested]
+	 * D. fill info. from desc to mbuf
+	 */
+
+	for (pos = 0, nb_pkts_recd = 0; pos < nb_pkts;
+	     pos += AVF_VPMD_DESCS_PER_LOOP,
+	     rxdp += AVF_VPMD_DESCS_PER_LOOP) {
+		__m128i descs[AVF_VPMD_DESCS_PER_LOOP];
+		__m128i pkt_mb1, pkt_mb2, pkt_mb3, pkt_mb4;
+		__m128i zero, staterr, sterr_tmp1, sterr_tmp2;
+		/* 2 64 bit or 4 32 bit mbuf pointers in one XMM reg. */
+		__m128i mbp1;
+#if defined(RTE_ARCH_X86_64)
+		__m128i mbp2;
+#endif
+
+		/* B.1 load 2 (64 bit) or 4 (32 bit) mbuf points */
+		mbp1 = _mm_loadu_si128((__m128i *)&sw_ring[pos]);
+		/* Read desc statuses backwards to avoid race condition */
+		/* A.1 load 4 pkts desc */
+		descs[3] = _mm_loadu_si128((__m128i *)(rxdp + 3));
+		rte_compiler_barrier();
+
+		/* B.2 copy 2 64 bit or 4 32 bit mbuf point into rx_pkts */
+		_mm_storeu_si128((__m128i *)&rx_pkts[pos], mbp1);
+
+#if defined(RTE_ARCH_X86_64)
+		/* B.1 load 2 64 bit mbuf points */
+		mbp2 = _mm_loadu_si128((__m128i *)&sw_ring[pos + 2]);
+#endif
+
+		descs[2] = _mm_loadu_si128((__m128i *)(rxdp + 2));
+		rte_compiler_barrier();
+		/* B.1 load 2 mbuf point */
+		descs[1] = _mm_loadu_si128((__m128i *)(rxdp + 1));
+		rte_compiler_barrier();
+		descs[0] = _mm_loadu_si128((__m128i *)(rxdp));
+
+#if defined(RTE_ARCH_X86_64)
+		/* B.2 copy 2 mbuf point into rx_pkts  */
+		_mm_storeu_si128((__m128i *)&rx_pkts[pos + 2], mbp2);
+#endif
+
+		if (split_packet) {
+			rte_mbuf_prefetch_part2(rx_pkts[pos]);
+			rte_mbuf_prefetch_part2(rx_pkts[pos + 1]);
+			rte_mbuf_prefetch_part2(rx_pkts[pos + 2]);
+			rte_mbuf_prefetch_part2(rx_pkts[pos + 3]);
+		}
+
+		/* avoid compiler reorder optimization */
+		rte_compiler_barrier();
+
+		/* pkt 3,4 shift the pktlen field to be 16-bit aligned*/
+		const __m128i len3 = _mm_slli_epi32(descs[3], PKTLEN_SHIFT);
+		const __m128i len2 = _mm_slli_epi32(descs[2], PKTLEN_SHIFT);
+
+		/* merge the now-aligned packet length fields back in */
+		descs[3] = _mm_blend_epi16(descs[3], len3, 0x80);
+		descs[2] = _mm_blend_epi16(descs[2], len2, 0x80);
+
+		/* D.1 pkt 3,4 convert format from desc to pktmbuf */
+		pkt_mb4 = _mm_shuffle_epi8(descs[3], shuf_msk);
+		pkt_mb3 = _mm_shuffle_epi8(descs[2], shuf_msk);
+
+		/* C.1 4=>2 status err info only */
+		sterr_tmp2 = _mm_unpackhi_epi32(descs[3], descs[2]);
+		sterr_tmp1 = _mm_unpackhi_epi32(descs[1], descs[0]);
+
+		desc_to_olflags_v(rxq, descs, &rx_pkts[pos]);
+
+		/* D.2 pkt 3,4 set in_port/nb_seg and remove crc */
+		pkt_mb4 = _mm_add_epi16(pkt_mb4, crc_adjust);
+		pkt_mb3 = _mm_add_epi16(pkt_mb3, crc_adjust);
+
+		/* pkt 1,2 shift the pktlen field to be 16-bit aligned*/
+		const __m128i len1 = _mm_slli_epi32(descs[1], PKTLEN_SHIFT);
+		const __m128i len0 = _mm_slli_epi32(descs[0], PKTLEN_SHIFT);
+
+		/* merge the now-aligned packet length fields back in */
+		descs[1] = _mm_blend_epi16(descs[1], len1, 0x80);
+		descs[0] = _mm_blend_epi16(descs[0], len0, 0x80);
+
+		/* D.1 pkt 1,2 convert format from desc to pktmbuf */
+		pkt_mb2 = _mm_shuffle_epi8(descs[1], shuf_msk);
+		pkt_mb1 = _mm_shuffle_epi8(descs[0], shuf_msk);
+
+		/* C.2 get 4 pkts status err value  */
+		zero = _mm_xor_si128(dd_check, dd_check);
+		staterr = _mm_unpacklo_epi32(sterr_tmp1, sterr_tmp2);
+
+		/* D.3 copy final 3,4 data to rx_pkts */
+		_mm_storeu_si128((void *)&rx_pkts[pos + 3]->rx_descriptor_fields1,
+				 pkt_mb4);
+		_mm_storeu_si128((void *)&rx_pkts[pos + 2]->rx_descriptor_fields1,
+				 pkt_mb3);
+
+		/* D.2 pkt 1,2 remove crc */
+		pkt_mb2 = _mm_add_epi16(pkt_mb2, crc_adjust);
+		pkt_mb1 = _mm_add_epi16(pkt_mb1, crc_adjust);
+
+		/* C* extract and record EOP bit */
+		if (split_packet) {
+			__m128i eop_shuf_mask = _mm_set_epi8(
+					0xFF, 0xFF, 0xFF, 0xFF,
+					0xFF, 0xFF, 0xFF, 0xFF,
+					0xFF, 0xFF, 0xFF, 0xFF,
+					0x04, 0x0C, 0x00, 0x08
+					);
+
+			/* and with mask to extract bits, flipping 1-0 */
+			__m128i eop_bits = _mm_andnot_si128(staterr, eop_check);
+			/* the staterr values are not in order, as the count
+			 * count of dd bits doesn't care. However, for end of
+			 * packet tracking, we do care, so shuffle. This also
+			 * compresses the 32-bit values to 8-bit
+			 */
+			eop_bits = _mm_shuffle_epi8(eop_bits, eop_shuf_mask);
+			/* store the resulting 32-bit value */
+			*(int *)split_packet = _mm_cvtsi128_si32(eop_bits);
+			split_packet += AVF_VPMD_DESCS_PER_LOOP;
+		}
+
+		/* C.3 calc available number of desc */
+		staterr = _mm_and_si128(staterr, dd_check);
+		staterr = _mm_packs_epi32(staterr, zero);
+
+		/* D.3 copy final 1,2 data to rx_pkts */
+		_mm_storeu_si128((void *)&rx_pkts[pos+1]->rx_descriptor_fields1,
+				 pkt_mb2);
+		_mm_storeu_si128((void *)&rx_pkts[pos]->rx_descriptor_fields1,
+				 pkt_mb1);
+		desc_to_ptype_v(descs, &rx_pkts[pos]);
+		/* C.4 calc avaialbe number of desc */
+		var = __builtin_popcountll(_mm_cvtsi128_si64(staterr));
+		nb_pkts_recd += var;
+		if (likely(var != AVF_VPMD_DESCS_PER_LOOP))
+			break;
+	}
+
+	/* Update our internal tail pointer */
+	rxq->rx_tail = (uint16_t)(rxq->rx_tail + nb_pkts_recd);
+	rxq->rx_tail = (uint16_t)(rxq->rx_tail & (rxq->nb_rx_desc - 1));
+	rxq->rxrearm_nb = (uint16_t)(rxq->rxrearm_nb + nb_pkts_recd);
+
+	return nb_pkts_recd;
+}
+
+/* Notice:
+ * - nb_pkts < AVF_DESCS_PER_LOOP, just return no packet
+ * - nb_pkts > AVF_VPMD_RX_MAX_BURST, only scan AVF_VPMD_RX_MAX_BURST
+ *   numbers of DD bits
+ */
+uint16_t
+avf_recv_pkts_vec(void *rx_queue, struct rte_mbuf **rx_pkts,
+		  uint16_t nb_pkts)
+{
+	return _recv_raw_pkts_vec(rx_queue, rx_pkts, nb_pkts, NULL);
+}
+
+/* vPMD receive routine that reassembles scattered packets
+ * Notice:
+ * - nb_pkts < AVF_VPMD_DESCS_PER_LOOP, just return no packet
+ * - nb_pkts > VPMD_RX_MAX_BURST, only scan AVF_VPMD_RX_MAX_BURST
+ *   numbers of DD bits
+ */
+uint16_t
+avf_recv_scattered_pkts_vec(void *rx_queue, struct rte_mbuf **rx_pkts,
+			    uint16_t nb_pkts)
+{
+	struct avf_rx_queue *rxq = rx_queue;
+	uint8_t split_flags[AVF_VPMD_RX_MAX_BURST] = {0};
+
+	/* get some new buffers */
+	uint16_t nb_bufs = _recv_raw_pkts_vec(rxq, rx_pkts, nb_pkts,
+					      split_flags);
+	if (nb_bufs == 0)
+		return 0;
+
+	/* happy day case, full burst + no packets to be joined */
+	const uint64_t *split_fl64 = (uint64_t *)split_flags;
+
+	if (rxq->pkt_first_seg == NULL &&
+	    split_fl64[0] == 0 && split_fl64[1] == 0 &&
+	    split_fl64[2] == 0 && split_fl64[3] == 0)
+		return nb_bufs;
+
+	/* reassemble any packets that need reassembly*/
+	unsigned i = 0;
+
+	if (rxq->pkt_first_seg == NULL) {
+		/* find the first split flag, and only reassemble then*/
+		while (i < nb_bufs && !split_flags[i])
+			i++;
+		if (i == nb_bufs)
+			return nb_bufs;
+	}
+	return i + reassemble_packets(rxq, &rx_pkts[i], nb_bufs - i,
+		&split_flags[i]);
+}
+
+static inline void
+vtx1(volatile struct avf_tx_desc *txdp, struct rte_mbuf *pkt, uint64_t flags)
+{
+	uint64_t high_qw =
+			(AVF_TX_DESC_DTYPE_DATA |
+			((uint64_t)flags  << AVF_TXD_QW1_CMD_SHIFT) |
+			((uint64_t)pkt->data_len << AVF_TXD_QW1_TX_BUF_SZ_SHIFT));
+
+	__m128i descriptor = _mm_set_epi64x(high_qw,
+					    pkt->buf_iova + pkt->data_off);
+	_mm_store_si128((__m128i *)txdp, descriptor);
+}
+
+static inline void
+vtx(volatile struct avf_tx_desc *txdp, struct rte_mbuf **pkt,
+    uint16_t nb_pkts,  uint64_t flags)
+{
+	int i;
+
+	for (i = 0; i < nb_pkts; ++i, ++txdp, ++pkt)
+		vtx1(txdp, *pkt, flags);
+}
+
+uint16_t
+avf_xmit_fixed_burst_vec(void *tx_queue, struct rte_mbuf **tx_pkts,
+			 uint16_t nb_pkts)
+{
+	struct avf_tx_queue *txq = (struct avf_tx_queue *)tx_queue;
+	volatile struct avf_tx_desc *txdp;
+	struct avf_tx_entry *txep;
+	uint16_t n, nb_commit, tx_id;
+	uint64_t flags = AVF_TX_DESC_CMD_EOP | 0x04;  /* bit 2 must be set */
+	uint64_t rs = AVF_TX_DESC_CMD_RS | flags;
+	int i;
+
+	/* cross rx_thresh boundary is not allowed */
+	nb_pkts = RTE_MIN(nb_pkts, txq->rs_thresh);
+
+	if (txq->nb_free < txq->free_thresh)
+		avf_tx_free_bufs(txq);
+
+	nb_commit = nb_pkts = (uint16_t)RTE_MIN(txq->nb_free, nb_pkts);
+	if (unlikely(nb_pkts == 0))
+		return 0;
+
+	tx_id = txq->tx_tail;
+	txdp = &txq->tx_ring[tx_id];
+	txep = &txq->sw_ring[tx_id];
+
+	txq->nb_free = (uint16_t)(txq->nb_free - nb_pkts);
+
+	n = (uint16_t)(txq->nb_tx_desc - tx_id);
+	if (nb_commit >= n) {
+		tx_backlog_entry(txep, tx_pkts, n);
+
+		for (i = 0; i < n - 1; ++i, ++tx_pkts, ++txdp)
+			vtx1(txdp, *tx_pkts, flags);
+
+		vtx1(txdp, *tx_pkts++, rs);
+
+		nb_commit = (uint16_t)(nb_commit - n);
+
+		tx_id = 0;
+		txq->next_rs = (uint16_t)(txq->rs_thresh - 1);
+
+		/* avoid reach the end of ring */
+		txdp = &txq->tx_ring[tx_id];
+		txep = &txq->sw_ring[tx_id];
+	}
+
+	tx_backlog_entry(txep, tx_pkts, nb_commit);
+
+	vtx(txdp, tx_pkts, nb_commit, flags);
+
+	tx_id = (uint16_t)(tx_id + nb_commit);
+	if (tx_id > txq->next_rs) {
+		txq->tx_ring[txq->next_rs].cmd_type_offset_bsz |=
+			rte_cpu_to_le_64(((uint64_t)AVF_TX_DESC_CMD_RS) <<
+					 AVF_TXD_QW1_CMD_SHIFT);
+		txq->next_rs =
+			(uint16_t)(txq->next_rs + txq->rs_thresh);
+	}
+
+	txq->tx_tail = tx_id;
+
+	PMD_TX_LOG(DEBUG, "port_id=%u queue_id=%u tx_tail=%u nb_pkts=%u",
+		   txq->port_id, txq->queue_id, tx_id, nb_pkts);
+
+	AVF_PCI_REG_WRITE(txq->qtx_tail, txq->tx_tail);
+
+	return nb_pkts;
+}
+
+void __attribute__((cold))
+avf_rx_queue_release_mbufs_sse(struct avf_rx_queue *rxq)
+{
+	_avf_rx_queue_release_mbufs_vec(rxq);
+}
+
+static void __attribute__((cold))
+avf_tx_queue_release_mbufs_sse(struct avf_tx_queue *txq)
+{
+	_avf_tx_queue_release_mbufs_vec(txq);
+}
+
+static const struct avf_rxq_ops sse_vec_rxq_ops = {
+	.release_mbufs = avf_rx_queue_release_mbufs_sse,
+};
+
+static const struct avf_txq_ops sse_vec_txq_ops = {
+	.release_mbufs = avf_tx_queue_release_mbufs_sse,
+};
+
+int __attribute__((cold))
+avf_txq_vec_setup(struct avf_tx_queue *txq)
+{
+	txq->ops = &sse_vec_txq_ops;
+	return 0;
+}
+
+int __attribute__((cold))
+avf_rxq_vec_setup(struct avf_rx_queue *rxq)
+{
+	rxq->ops = &sse_vec_rxq_ops;
+	return avf_rxq_vec_setup_default(rxq);
+}
-- 
2.4.11



More information about the dev mailing list