[dpdk-dev] [PATCH v7 03/14] net/avf: enable queue and device

Wenzhuo Lu wenzhuo.lu at intel.com
Wed Jan 10 14:01:55 CET 2018


From: Jingjing Wu <jingjing.wu at intel.com>

enable device and queue setup ops like:

 - dev_configure
 - dev_start
 - dev_stop
 - dev_close
 - dev_infos_get
 - rx_queue_start
 - rx_queue_stop
 - tx_queue_start
 - tx_queue_stop
 - rx_queue_setup
 - rx_queue_release
 - tx_queue_setup
 - tx_queue_release

Signed-off-by: Jingjing Wu <jingjing.wu at intel.com>
---
 drivers/net/avf/Makefile     |   1 +
 drivers/net/avf/avf.h        |  18 ++
 drivers/net/avf/avf_ethdev.c | 366 +++++++++++++++++++++++++
 drivers/net/avf/avf_rxtx.c   | 616 +++++++++++++++++++++++++++++++++++++++++++
 drivers/net/avf/avf_rxtx.h   | 160 +++++++++++
 drivers/net/avf/avf_vchnl.c  | 359 ++++++++++++++++++++++++-
 6 files changed, 1518 insertions(+), 2 deletions(-)
 create mode 100644 drivers/net/avf/avf_rxtx.c
 create mode 100644 drivers/net/avf/avf_rxtx.h

diff --git a/drivers/net/avf/Makefile b/drivers/net/avf/Makefile
index 2376cfd..e172bf5 100644
--- a/drivers/net/avf/Makefile
+++ b/drivers/net/avf/Makefile
@@ -43,5 +43,6 @@ 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
 
 include $(RTE_SDK)/mk/rte.lib.mk
diff --git a/drivers/net/avf/avf.h b/drivers/net/avf/avf.h
index 4694cc5..22886d4 100644
--- a/drivers/net/avf/avf.h
+++ b/drivers/net/avf/avf.h
@@ -38,6 +38,13 @@
 	VIRTCHNL_VF_OFFLOAD_WB_ON_ITR | \
 	VIRTCHNL_VF_OFFLOAD_RX_POLLING)
 
+#define AVF_RSS_OFFLOAD_ALL ( \
+	ETH_RSS_FRAG_IPV4 |         \
+	ETH_RSS_NONFRAG_IPV4_TCP |  \
+	ETH_RSS_NONFRAG_IPV4_UDP |  \
+	ETH_RSS_NONFRAG_IPV4_SCTP | \
+	ETH_RSS_NONFRAG_IPV4_OTHER)
+
 #define AVF_MISC_VEC_ID                RTE_INTR_VEC_ZERO_OFFSET
 #define AVF_RX_VEC_START               RTE_INTR_VEC_RXTX_OFFSET
 
@@ -184,4 +191,15 @@ struct avf_cmd_info {
 int avf_check_api_version(struct avf_adapter *adapter);
 int avf_get_vf_resource(struct avf_adapter *adapter);
 void avf_handle_virtchnl_msg(struct rte_eth_dev *dev);
+int avf_enable_vlan_strip(struct avf_adapter *adapter);
+int avf_disable_vlan_strip(struct avf_adapter *adapter);
+int avf_switch_queue(struct avf_adapter *adapter, uint16_t qid,
+		     bool rx, bool on);
+int avf_enable_queues(struct avf_adapter *adapter);
+int avf_disable_queues(struct avf_adapter *adapter);
+int avf_configure_rss_lut(struct avf_adapter *adapter);
+int avf_configure_rss_key(struct avf_adapter *adapter);
+int avf_configure_queues(struct avf_adapter *adapter);
+int avf_config_irq_map(struct avf_adapter *adapter);
+void avf_add_del_all_mac_addr(struct avf_adapter *adapter, bool add);
 #endif /* _AVF_ETHDEV_H_ */
diff --git a/drivers/net/avf/avf_ethdev.c b/drivers/net/avf/avf_ethdev.c
index 0ed6e1c..c53f00e 100644
--- a/drivers/net/avf/avf_ethdev.c
+++ b/drivers/net/avf/avf_ethdev.c
@@ -31,6 +31,14 @@
 #include "base/avf_type.h"
 
 #include "avf.h"
+#include "avf_rxtx.h"
+
+static int avf_dev_configure(struct rte_eth_dev *dev);
+static int avf_dev_start(struct rte_eth_dev *dev);
+static void avf_dev_stop(struct rte_eth_dev *dev);
+static void avf_dev_close(struct rte_eth_dev *dev);
+static void avf_dev_info_get(struct rte_eth_dev *dev,
+			     struct rte_eth_dev_info *dev_info);
 
 int avf_logtype_init;
 int avf_logtype_driver;
@@ -40,9 +48,366 @@
 };
 
 static const struct eth_dev_ops avf_eth_dev_ops = {
+	.dev_configure              = avf_dev_configure,
+	.dev_start                  = avf_dev_start,
+	.dev_stop                   = avf_dev_stop,
+	.dev_close                  = avf_dev_close,
+	.dev_infos_get              = avf_dev_info_get,
+	.rx_queue_start             = avf_dev_rx_queue_start,
+	.rx_queue_stop              = avf_dev_rx_queue_stop,
+	.tx_queue_start             = avf_dev_tx_queue_start,
+	.tx_queue_stop              = avf_dev_tx_queue_stop,
+	.rx_queue_setup             = avf_dev_rx_queue_setup,
+	.rx_queue_release           = avf_dev_rx_queue_release,
+	.tx_queue_setup             = avf_dev_tx_queue_setup,
+	.tx_queue_release           = avf_dev_tx_queue_release,
 };
 
 static int
+avf_dev_configure(struct rte_eth_dev *dev)
+{
+	struct avf_adapter *ad =
+		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
+	struct avf_info *vf =  AVF_DEV_PRIVATE_TO_VF(ad);
+	struct rte_eth_conf *dev_conf = &dev->data->dev_conf;
+
+	/* Vlan stripping setting */
+	if (vf->vf_res->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_VLAN) {
+		if (dev_conf->rxmode.offloads & DEV_RX_OFFLOAD_VLAN_STRIP)
+			avf_enable_vlan_strip(ad);
+		else
+			avf_disable_vlan_strip(ad);
+	}
+	return 0;
+}
+
+static int
+avf_init_rss(struct avf_adapter *adapter)
+{
+	struct avf_info *vf =  AVF_DEV_PRIVATE_TO_VF(adapter);
+	struct avf_hw *hw = AVF_DEV_PRIVATE_TO_HW(adapter);
+	struct rte_eth_rss_conf *rss_conf;
+	uint8_t i, j, nb_q;
+	int ret;
+
+	rss_conf = &adapter->eth_dev->data->dev_conf.rx_adv_conf.rss_conf;
+	nb_q = RTE_MIN(adapter->eth_dev->data->nb_rx_queues,
+		       AVF_MAX_NUM_QUEUES);
+
+	if (!(vf->vf_res->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_RSS_PF)) {
+		PMD_DRV_LOG(DEBUG, "RSS is not supported");
+		return -ENOTSUP;
+	}
+	if (adapter->eth_dev->data->dev_conf.rxmode.mq_mode != ETH_MQ_RX_RSS) {
+		PMD_DRV_LOG(WARNING, "RSS is enabled by PF by default");
+		/* set all lut items to default queue */
+		for (i = 0; i < vf->vf_res->rss_lut_size; i++)
+			vf->rss_lut[i] = 0;
+		ret = avf_configure_rss_lut(adapter);
+		return ret;
+	}
+
+	/* In AVF, RSS enablement is set by PF driver. It is not supported
+	 * to set based on rss_conf->rss_hf.
+	 */
+
+	/* configure RSS key */
+	if (!rss_conf->rss_key) {
+		/* Calculate the default hash key */
+		for (i = 0; i <= vf->vf_res->rss_key_size; i++)
+			vf->rss_key[i] = (uint8_t)rte_rand();
+	} else
+		rte_memcpy(vf->rss_key, rss_conf->rss_key,
+			   RTE_MIN(rss_conf->rss_key_len,
+				   vf->vf_res->rss_key_size));
+
+	/* init RSS LUT table */
+	for (i = 0; i < vf->vf_res->rss_lut_size; i++, j++) {
+		if (j >= nb_q)
+			j = 0;
+		vf->rss_lut[i] = j;
+	}
+	/* send virtchnnl ops to configure rss*/
+	ret = avf_configure_rss_lut(adapter);
+	if (ret)
+		return ret;
+	ret = avf_configure_rss_key(adapter);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+static int
+avf_init_rxq(struct rte_eth_dev *dev, struct avf_rx_queue *rxq)
+{
+	struct avf_hw *hw = AVF_DEV_PRIVATE_TO_HW(dev->data->dev_private);
+	struct rte_eth_dev_data *dev_data = dev->data;
+	uint16_t buf_size, max_pkt_len, len;
+
+	buf_size = rte_pktmbuf_data_room_size(rxq->mp) - RTE_PKTMBUF_HEADROOM;
+
+	/* Calculate the maximum packet length allowed */
+	len = rxq->rx_buf_len * AVF_MAX_CHAINED_RX_BUFFERS;
+	max_pkt_len = RTE_MIN(len, dev->data->dev_conf.rxmode.max_rx_pkt_len);
+
+	/* Check if the jumbo frame and maximum packet length are set
+	 * correctly.
+	 */
+	if (dev->data->dev_conf.rxmode.offloads & DEV_RX_OFFLOAD_JUMBO_FRAME) {
+		if (max_pkt_len <= ETHER_MAX_LEN ||
+		    max_pkt_len > AVF_FRAME_SIZE_MAX) {
+			PMD_DRV_LOG(ERR, "maximum packet length must be "
+				    "larger than %u and smaller than %u, "
+				    "as jumbo frame is enabled",
+				    (uint32_t)ETHER_MAX_LEN,
+				    (uint32_t)AVF_FRAME_SIZE_MAX);
+			return -EINVAL;
+		}
+	} else {
+		if (max_pkt_len < ETHER_MIN_LEN ||
+		    max_pkt_len > ETHER_MAX_LEN) {
+			PMD_DRV_LOG(ERR, "maximum packet length must be "
+				    "larger than %u and smaller than %u, "
+				    "as jumbo frame is disabled",
+				    (uint32_t)ETHER_MIN_LEN,
+				    (uint32_t)ETHER_MAX_LEN);
+			return -EINVAL;
+		}
+	}
+
+	rxq->max_pkt_len = max_pkt_len;
+	if ((dev_data->dev_conf.rxmode.offloads & DEV_RX_OFFLOAD_SCATTER) ||
+	    (rxq->max_pkt_len + 2 * AVF_VLAN_TAG_SIZE) > buf_size) {
+		dev_data->scattered_rx = 1;
+	}
+	AVF_PCI_REG_WRITE(rxq->qrx_tail, rxq->nb_rx_desc - 1);
+	AVF_WRITE_FLUSH(hw);
+
+	return 0;
+}
+
+static int
+avf_init_queues(struct rte_eth_dev *dev)
+{
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(dev->data->dev_private);
+	struct avf_rx_queue **rxq =
+		(struct avf_rx_queue **)dev->data->rx_queues;
+	struct avf_tx_queue **txq =
+		(struct avf_tx_queue **)dev->data->tx_queues;
+	int i, ret = AVF_SUCCESS;
+
+	for (i = 0; i < dev->data->nb_rx_queues; i++) {
+		if (!rxq[i] || !rxq[i]->q_set)
+			continue;
+		ret = avf_init_rxq(dev, rxq[i]);
+		if (ret != AVF_SUCCESS)
+			break;
+	}
+	/* TODO: set rx/tx function to vector/scatter/single-segment
+	 * according to parameters
+	 */
+	return ret;
+}
+
+static int
+avf_start_queues(struct rte_eth_dev *dev)
+{
+	struct avf_rx_queue *rxq;
+	struct avf_tx_queue *txq;
+	int i;
+
+	for (i = 0; i < dev->data->nb_tx_queues; i++) {
+		txq = dev->data->tx_queues[i];
+		if (txq->tx_deferred_start)
+			continue;
+		if (avf_dev_tx_queue_start(dev, i) != 0) {
+			PMD_DRV_LOG(ERR, "Fail to start queue %u", i);
+			return -1;
+		}
+	}
+
+	for (i = 0; i < dev->data->nb_rx_queues; i++) {
+		rxq = dev->data->rx_queues[i];
+		if (rxq->rx_deferred_start)
+			continue;
+		if (avf_dev_rx_queue_start(dev, i) != 0) {
+			PMD_DRV_LOG(ERR, "Fail to start queue %u", i);
+			return -1;
+		}
+	}
+
+	return 0;
+}
+
+static int
+avf_dev_start(struct rte_eth_dev *dev)
+{
+	struct avf_adapter *adapter =
+		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(dev->data->dev_private);
+	struct avf_hw *hw = AVF_DEV_PRIVATE_TO_HW(dev->data->dev_private);
+	struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev);
+	struct rte_intr_handle *intr_handle = dev->intr_handle;
+	uint16_t interval;
+	int i;
+
+	PMD_INIT_FUNC_TRACE();
+
+	hw->adapter_stopped = 0;
+
+	vf->max_pkt_len = dev->data->dev_conf.rxmode.max_rx_pkt_len;
+	vf->num_queue_pairs = RTE_MAX(dev->data->nb_rx_queues,
+				      dev->data->nb_tx_queues);
+
+	/* TODO: Rx interrupt */
+
+	if (avf_init_queues(dev) != 0) {
+		PMD_DRV_LOG(ERR, "failed to do Queue init");
+		return -1;
+	}
+
+	if (vf->vf_res->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_RSS_PF) {
+		if (avf_init_rss(adapter) != 0) {
+			PMD_DRV_LOG(ERR, "configure rss failed");
+			goto err_rss;
+		}
+	}
+
+	if (avf_configure_queues(adapter) != 0) {
+		PMD_DRV_LOG(ERR, "configure queues failed");
+		goto err_queue;
+	}
+
+	/* Map interrupt for writeback */
+	vf->nb_msix = 1;
+	if (vf->vf_res->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_WB_ON_ITR) {
+		/* If WB_ON_ITR supports, enable it */
+		vf->msix_base = AVF_RX_VEC_START;
+		AVF_WRITE_REG(hw, AVFINT_DYN_CTLN1(vf->msix_base - 1),
+			      AVFINT_DYN_CTLN1_ITR_INDX_MASK |
+			      AVFINT_DYN_CTLN1_WB_ON_ITR_MASK);
+	} else {
+		/* If no WB_ON_ITR offload flags, need to set interrupt for
+		 * descriptor write back.
+		 */
+		vf->msix_base = AVF_MISC_VEC_ID;
+
+		/* set ITR to max */
+		interval = avf_calc_itr_interval(AVF_QUEUE_ITR_INTERVAL_MAX);
+		AVF_WRITE_REG(hw, AVFINT_DYN_CTL01,
+			      AVFINT_DYN_CTL01_INTENA_MASK |
+			      (AVF_ITR_INDEX_DEFAULT <<
+			       AVFINT_DYN_CTL01_ITR_INDX_SHIFT) |
+			      (interval << AVFINT_DYN_CTL01_INTERVAL_SHIFT));
+	}
+	AVF_WRITE_FLUSH(hw);
+	/* map all queues to the same interrupt */
+	for (i = 0; i < dev->data->nb_rx_queues; i++)
+		vf->rxq_map[0] |= 1 << i;
+	if (avf_config_irq_map(adapter)) {
+		PMD_DRV_LOG(ERR, "config interrupt mapping failed");
+		goto err_queue;
+	}
+
+	/* Set all mac addrs */
+	avf_add_del_all_mac_addr(adapter, TRUE);
+
+	if (avf_start_queues(dev) != 0) {
+		PMD_DRV_LOG(ERR, "enable queues failed");
+		goto err_mac;
+	}
+
+	/* TODO: enable interrupt for RX interrupt */
+	return 0;
+
+err_mac:
+	avf_add_del_all_mac_addr(adapter, FALSE);
+err_queue:
+err_rss:
+	return -1;
+}
+
+static void
+avf_dev_stop(struct rte_eth_dev *dev)
+{
+	struct avf_adapter *adapter =
+		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
+	struct avf_hw *hw = AVF_DEV_PRIVATE_TO_HW(dev);
+	int ret, i;
+
+	PMD_INIT_FUNC_TRACE();
+
+	if (hw->adapter_stopped == 1)
+		return;
+
+	avf_stop_queues(dev);
+
+	/*TODO: Disable the interrupt for Rx*/
+
+	/* TODO: Rx interrupt vector mapping free */
+
+	/* remove all mac addrs */
+	avf_add_del_all_mac_addr(adapter, FALSE);
+	hw->adapter_stopped = 1;
+}
+
+static void
+avf_dev_info_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
+{
+	struct avf_adapter *adapter =
+		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(dev->data->dev_private);
+
+	memset(dev_info, 0, sizeof(*dev_info));
+	dev_info->pci_dev = RTE_ETH_DEV_TO_PCI(dev);
+	dev_info->max_rx_queues = vf->vsi_res->num_queue_pairs;
+	dev_info->max_tx_queues = vf->vsi_res->num_queue_pairs;
+	dev_info->min_rx_bufsize = AVF_BUF_SIZE_MIN;
+	dev_info->max_rx_pktlen = AVF_FRAME_SIZE_MAX;
+	dev_info->hash_key_size = vf->vf_res->rss_key_size;
+	dev_info->reta_size = vf->vf_res->rss_lut_size;
+	dev_info->flow_type_rss_offloads = AVF_RSS_OFFLOAD_ALL;
+	dev_info->max_mac_addrs = AVF_NUM_MACADDR_MAX;
+	dev_info->rx_offload_capa =
+		DEV_RX_OFFLOAD_VLAN_STRIP |
+		DEV_RX_OFFLOAD_IPV4_CKSUM |
+		DEV_RX_OFFLOAD_UDP_CKSUM |
+		DEV_RX_OFFLOAD_TCP_CKSUM;
+	dev_info->tx_offload_capa =
+		DEV_TX_OFFLOAD_VLAN_INSERT |
+		DEV_TX_OFFLOAD_IPV4_CKSUM |
+		DEV_TX_OFFLOAD_UDP_CKSUM |
+		DEV_TX_OFFLOAD_TCP_CKSUM |
+		DEV_TX_OFFLOAD_SCTP_CKSUM |
+		DEV_TX_OFFLOAD_TCP_TSO;
+
+	dev_info->default_rxconf = (struct rte_eth_rxconf) {
+		.rx_free_thresh = AVF_DEFAULT_RX_FREE_THRESH,
+		.rx_drop_en = 0,
+	};
+
+	dev_info->default_txconf = (struct rte_eth_txconf) {
+		.tx_free_thresh = AVF_DEFAULT_TX_FREE_THRESH,
+		.tx_rs_thresh = AVF_DEFAULT_TX_RS_THRESH,
+		.txq_flags = ETH_TXQ_FLAGS_NOMULTSEGS |
+				ETH_TXQ_FLAGS_NOOFFLOADS,
+	};
+
+	dev_info->rx_desc_lim = (struct rte_eth_desc_lim) {
+		.nb_max = AVF_MAX_RING_DESC,
+		.nb_min = AVF_MIN_RING_DESC,
+		.nb_align = AVF_ALIGN_RING_DESC,
+	};
+
+	dev_info->tx_desc_lim = (struct rte_eth_desc_lim) {
+		.nb_max = AVF_MAX_RING_DESC,
+		.nb_min = AVF_MIN_RING_DESC,
+		.nb_align = AVF_ALIGN_RING_DESC,
+	};
+}
+
+static int
 avf_check_vf_reset_done(struct avf_hw *hw)
 {
 	int i, reset;
@@ -250,6 +615,7 @@
 	struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev);
 	struct rte_intr_handle *intr_handle = &pci_dev->intr_handle;
 
+	avf_dev_stop(dev);
 	avf_shutdown_adminq(hw);
 	/* disable uio intr before callback unregister */
 	rte_intr_disable(intr_handle);
diff --git a/drivers/net/avf/avf_rxtx.c b/drivers/net/avf/avf_rxtx.c
new file mode 100644
index 0000000..2d4fb4c
--- /dev/null
+++ b/drivers/net/avf/avf_rxtx.c
@@ -0,0 +1,616 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2017 Intel Corporation
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <errno.h>
+#include <stdint.h>
+#include <stdarg.h>
+#include <unistd.h>
+#include <inttypes.h>
+#include <sys/queue.h>
+
+#include <rte_string_fns.h>
+#include <rte_memzone.h>
+#include <rte_mbuf.h>
+#include <rte_malloc.h>
+#include <rte_ether.h>
+#include <rte_ethdev.h>
+#include <rte_tcp.h>
+#include <rte_sctp.h>
+#include <rte_udp.h>
+#include <rte_ip.h>
+#include <rte_net.h>
+
+#include "avf_log.h"
+#include "base/avf_prototype.h"
+#include "base/avf_type.h"
+#include "avf.h"
+#include "avf_rxtx.h"
+
+static inline int
+check_rx_thresh(uint16_t nb_desc, uint16_t thresh)
+{
+	/* The following constraints must be satisfied:
+	 *   thresh >= AVF_RX_MAX_BURST
+	 *   thresh < rxq->nb_rx_desc
+	 *   (rxq->nb_rx_desc % thresh) == 0
+	 */
+	if (thresh < AVF_RX_MAX_BURST ||
+	    thresh >= nb_desc ||
+	    (nb_desc % thresh != 0)) {
+		PMD_INIT_LOG(ERR, "rx_free_thresh (%u) must be less than %u, "
+			     "greater than or equal to %u, "
+			     "and a divisor of %u",
+			     thresh, nb_desc, AVF_RX_MAX_BURST, nb_desc);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static inline int
+check_tx_thresh(uint16_t nb_desc, uint16_t tx_rs_thresh,
+		uint16_t tx_free_thresh)
+{
+	/* TX descriptors will have their RS bit set after tx_rs_thresh
+	 * descriptors have been used. The TX descriptor ring will be cleaned
+	 * after tx_free_thresh descriptors are used or if the number of
+	 * descriptors required to transmit a packet is greater than the
+	 * number of free TX descriptors.
+	 *
+	 * The following constraints must be satisfied:
+	 *  - tx_rs_thresh must be less than the size of the ring minus 2.
+	 *  - tx_free_thresh must be less than the size of the ring minus 3.
+	 *  - tx_rs_thresh must be less than or equal to tx_free_thresh.
+	 *  - tx_rs_thresh must be a divisor of the ring size.
+	 *
+	 * One descriptor in the TX ring is used as a sentinel to avoid a H/W
+	 * race condition, hence the maximum threshold constraints. When set
+	 * to zero use default values.
+	 */
+	if (tx_rs_thresh >= (nb_desc - 2)) {
+		PMD_INIT_LOG(ERR, "tx_rs_thresh (%u) must be less than the "
+			     "number of TX descriptors (%u) minus 2",
+			     tx_rs_thresh, nb_desc);
+		return -EINVAL;
+	}
+	if (tx_free_thresh >= (nb_desc - 3)) {
+		PMD_INIT_LOG(ERR, "tx_free_thresh (%u) must be less than the "
+			     "number of TX descriptors (%u) minus 3.",
+			     tx_free_thresh, nb_desc);
+		return -EINVAL;
+	}
+	if (tx_rs_thresh > tx_free_thresh) {
+		PMD_INIT_LOG(ERR, "tx_rs_thresh (%u) must be less than or "
+			     "equal to tx_free_thresh (%u).",
+			     tx_rs_thresh, tx_free_thresh);
+		return -EINVAL;
+	}
+	if ((nb_desc % tx_rs_thresh) != 0) {
+		PMD_INIT_LOG(ERR, "tx_rs_thresh (%u) must be a divisor of the "
+			     "number of TX descriptors (%u).",
+			     tx_rs_thresh, nb_desc);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static inline void
+reset_rx_queue(struct avf_rx_queue *rxq)
+{
+	uint16_t len, i;
+
+	if (!rxq)
+		return;
+
+	len = rxq->nb_rx_desc + AVF_RX_MAX_BURST;
+
+	for (i = 0; i < len * sizeof(union avf_rx_desc); i++)
+		((volatile char *)rxq->rx_ring)[i] = 0;
+
+	memset(&rxq->fake_mbuf, 0x0, sizeof(rxq->fake_mbuf));
+
+	for (i = 0; i < AVF_RX_MAX_BURST; i++)
+		rxq->sw_ring[rxq->nb_rx_desc + i] = &rxq->fake_mbuf;
+
+	rxq->rx_tail = 0;
+	rxq->nb_rx_hold = 0;
+	rxq->pkt_first_seg = NULL;
+	rxq->pkt_last_seg = NULL;
+}
+
+static inline void
+reset_tx_queue(struct avf_tx_queue *txq)
+{
+	struct avf_tx_entry *txe;
+	uint16_t i, prev, size;
+
+	if (!txq) {
+		PMD_DRV_LOG(DEBUG, "Pointer to txq is NULL");
+		return;
+	}
+
+	txe = txq->sw_ring;
+	size = sizeof(struct avf_tx_desc) * txq->nb_tx_desc;
+	for (i = 0; i < size; i++)
+		((volatile char *)txq->tx_ring)[i] = 0;
+
+	prev = (uint16_t)(txq->nb_tx_desc - 1);
+	for (i = 0; i < txq->nb_tx_desc; i++) {
+		txq->tx_ring[i].cmd_type_offset_bsz =
+			rte_cpu_to_le_64(AVF_TX_DESC_DTYPE_DESC_DONE);
+		txe[i].mbuf =  NULL;
+		txe[i].last_id = i;
+		txe[prev].next_id = i;
+		prev = i;
+	}
+
+	txq->tx_tail = 0;
+	txq->nb_used = 0;
+
+	txq->last_desc_cleaned = txq->nb_tx_desc - 1;
+	txq->nb_free = txq->nb_tx_desc - 1;
+
+	txq->next_dd = txq->rs_thresh - 1;
+	txq->next_rs = txq->rs_thresh - 1;
+}
+
+static int
+alloc_rxq_mbufs(struct avf_rx_queue *rxq)
+{
+	volatile union avf_rx_desc *rxd;
+	struct rte_mbuf *mbuf = NULL;
+	uint64_t dma_addr;
+	uint16_t i;
+
+	for (i = 0; i < rxq->nb_rx_desc; i++) {
+		mbuf = rte_mbuf_raw_alloc(rxq->mp);
+		if (unlikely(!mbuf)) {
+			PMD_DRV_LOG(ERR, "Failed to allocate mbuf for RX");
+			return -ENOMEM;
+		}
+
+		rte_mbuf_refcnt_set(mbuf, 1);
+		mbuf->next = NULL;
+		mbuf->data_off = RTE_PKTMBUF_HEADROOM;
+		mbuf->nb_segs = 1;
+		mbuf->port = rxq->port_id;
+
+		dma_addr =
+			rte_cpu_to_le_64(rte_mbuf_data_iova_default(mbuf));
+
+		rxd = &rxq->rx_ring[i];
+		rxd->read.pkt_addr = dma_addr;
+		rxd->read.hdr_addr = 0;
+#ifndef RTE_LIBRTE_AVF_16BYTE_RX_DESC
+		rxd->read.rsvd1 = 0;
+		rxd->read.rsvd2 = 0;
+#endif
+
+		rxq->sw_ring[i] = mbuf;
+	}
+
+	return 0;
+}
+
+static inline void
+release_rxq_mbufs(struct avf_rx_queue *rxq)
+{
+	struct rte_mbuf *mbuf;
+	uint16_t i;
+
+	if (!rxq->sw_ring)
+		return;
+
+	for (i = 0; i < rxq->nb_rx_desc; i++) {
+		if (rxq->sw_ring[i]) {
+			rte_pktmbuf_free_seg(rxq->sw_ring[i]);
+			rxq->sw_ring[i] = NULL;
+		}
+	}
+}
+
+static inline void
+release_txq_mbufs(struct avf_tx_queue *txq)
+{
+	uint16_t i;
+
+	if (!txq || !txq->sw_ring) {
+		PMD_DRV_LOG(DEBUG, "Pointer to rxq or sw_ring is NULL");
+		return;
+	}
+
+	for (i = 0; i < txq->nb_tx_desc; i++) {
+		if (txq->sw_ring[i].mbuf) {
+			rte_pktmbuf_free_seg(txq->sw_ring[i].mbuf);
+			txq->sw_ring[i].mbuf = NULL;
+		}
+	}
+}
+
+int
+avf_dev_rx_queue_setup(struct rte_eth_dev *dev, uint16_t queue_idx,
+		       uint16_t nb_desc, unsigned int socket_id,
+		       const struct rte_eth_rxconf *rx_conf,
+		       struct rte_mempool *mp)
+{
+	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_rx_queue *rxq;
+	const struct rte_memzone *mz;
+	uint32_t ring_size;
+	uint16_t len, i;
+	uint16_t rx_free_thresh;
+	uint16_t base, bsf, tc_mapping;
+
+	PMD_INIT_FUNC_TRACE();
+
+	if (nb_desc % AVF_ALIGN_RING_DESC != 0 ||
+	    nb_desc > AVF_MAX_RING_DESC ||
+	    nb_desc < AVF_MIN_RING_DESC) {
+		PMD_INIT_LOG(ERR, "Number (%u) of receive descriptors is "
+			     "invalid", nb_desc);
+		return -EINVAL;
+	}
+
+	/* Check free threshold */
+	rx_free_thresh = (rx_conf->rx_free_thresh == 0) ?
+			 AVF_DEFAULT_RX_FREE_THRESH :
+			 rx_conf->rx_free_thresh;
+	if (check_rx_thresh(nb_desc, rx_free_thresh) != 0)
+		return -EINVAL;
+
+	/* Free memory if needed */
+	if (dev->data->rx_queues[queue_idx]) {
+		avf_dev_rx_queue_release(dev->data->rx_queues[queue_idx]);
+		dev->data->rx_queues[queue_idx] = NULL;
+	}
+
+	/* Allocate the rx queue data structure */
+	rxq = rte_zmalloc_socket("avf rxq",
+				 sizeof(struct avf_rx_queue),
+				 RTE_CACHE_LINE_SIZE,
+				 socket_id);
+	if (!rxq) {
+		PMD_INIT_LOG(ERR, "Failed to allocate memory for "
+			     "rx queue data structure");
+		return -ENOMEM;
+	}
+
+	rxq->mp = mp;
+	rxq->nb_rx_desc = nb_desc;
+	rxq->rx_free_thresh = rx_free_thresh;
+	rxq->queue_id = queue_idx;
+	rxq->port_id = dev->data->port_id;
+	rxq->crc_len = 0; /* crc stripping by default */
+	rxq->rx_deferred_start = rx_conf->rx_deferred_start;
+	rxq->rx_hdr_len = 0;
+
+	len = rte_pktmbuf_data_room_size(rxq->mp) - RTE_PKTMBUF_HEADROOM;
+	rxq->rx_buf_len = RTE_ALIGN(len, (1 << AVF_RXQ_CTX_DBUFF_SHIFT));
+
+	/* Allocate the software ring. */
+	len = nb_desc + AVF_RX_MAX_BURST;
+	rxq->sw_ring =
+		rte_zmalloc_socket("avf rx sw ring",
+				   sizeof(struct rte_mbuf *) * len,
+				   RTE_CACHE_LINE_SIZE,
+				   socket_id);
+	if (!rxq->sw_ring) {
+		PMD_INIT_LOG(ERR, "Failed to allocate memory for SW ring");
+		rte_free(rxq);
+		return -ENOMEM;
+	}
+
+	/* Allocate the maximun number of RX ring hardware descriptor with
+	 * a liitle more to support bulk allocate.
+	 */
+	len = AVF_MAX_RING_DESC + AVF_RX_MAX_BURST;
+	ring_size = RTE_ALIGN(len * sizeof(union avf_rx_desc),
+			      AVF_DMA_MEM_ALIGN);
+	mz = rte_eth_dma_zone_reserve(dev, "rx_ring", queue_idx,
+				      ring_size, AVF_RING_BASE_ALIGN,
+				      socket_id);
+	if (!mz) {
+		PMD_INIT_LOG(ERR, "Failed to reserve DMA memory for RX");
+		rte_free(rxq->sw_ring);
+		rte_free(rxq);
+		return -ENOMEM;
+	}
+	/* Zero all the descriptors in the ring. */
+	memset(mz->addr, 0, ring_size);
+	rxq->rx_ring_phys_addr = mz->iova;
+	rxq->rx_ring = (union avf_rx_desc *)mz->addr;
+
+	rxq->mz = mz;
+	reset_rx_queue(rxq);
+	rxq->q_set = TRUE;
+	dev->data->rx_queues[queue_idx] = rxq;
+	rxq->qrx_tail = hw->hw_addr + AVF_QRX_TAIL1(rxq->queue_id);
+
+	return 0;
+}
+
+int
+avf_dev_tx_queue_setup(struct rte_eth_dev *dev,
+		       uint16_t queue_idx,
+		       uint16_t nb_desc,
+		       unsigned int socket_id,
+		       const struct rte_eth_txconf *tx_conf)
+{
+	struct avf_hw *hw = AVF_DEV_PRIVATE_TO_HW(dev->data->dev_private);
+	struct avf_tx_queue *txq;
+	const struct rte_memzone *mz;
+	uint32_t ring_size;
+	uint16_t tx_rs_thresh, tx_free_thresh;
+	uint16_t i, base, bsf, tc_mapping;
+
+	PMD_INIT_FUNC_TRACE();
+
+	if (nb_desc % AVF_ALIGN_RING_DESC != 0 ||
+	    nb_desc > AVF_MAX_RING_DESC ||
+	    nb_desc < AVF_MIN_RING_DESC) {
+		PMD_INIT_LOG(ERR, "Number (%u) of transmit descriptors is "
+			    "invalid", nb_desc);
+		return -EINVAL;
+	}
+
+	tx_rs_thresh = (uint16_t)((tx_conf->tx_rs_thresh) ?
+		tx_conf->tx_rs_thresh : DEFAULT_TX_RS_THRESH);
+	tx_free_thresh = (uint16_t)((tx_conf->tx_free_thresh) ?
+		tx_conf->tx_free_thresh : DEFAULT_TX_FREE_THRESH);
+	check_tx_thresh(nb_desc, tx_rs_thresh, tx_rs_thresh);
+
+	/* Free memory if needed. */
+	if (dev->data->tx_queues[queue_idx]) {
+		avf_dev_tx_queue_release(dev->data->tx_queues[queue_idx]);
+		dev->data->tx_queues[queue_idx] = NULL;
+	}
+
+	/* Allocate the TX queue data structure. */
+	txq = rte_zmalloc_socket("avf txq",
+				 sizeof(struct avf_tx_queue),
+				 RTE_CACHE_LINE_SIZE,
+				 socket_id);
+	if (!txq) {
+		PMD_INIT_LOG(ERR, "Failed to allocate memory for "
+			     "tx queue structure");
+		return -ENOMEM;
+	}
+
+	txq->nb_tx_desc = nb_desc;
+	txq->rs_thresh = tx_rs_thresh;
+	txq->free_thresh = tx_free_thresh;
+	txq->queue_id = queue_idx;
+	txq->port_id = dev->data->port_id;
+	txq->txq_flags = tx_conf->txq_flags;
+	txq->tx_deferred_start = tx_conf->tx_deferred_start;
+
+	/* Allocate software ring */
+	txq->sw_ring =
+		rte_zmalloc_socket("avf tx sw ring",
+				   sizeof(struct avf_tx_entry) * nb_desc,
+				   RTE_CACHE_LINE_SIZE,
+				   socket_id);
+	if (!txq->sw_ring) {
+		PMD_INIT_LOG(ERR, "Failed to allocate memory for SW TX ring");
+		rte_free(txq);
+		return -ENOMEM;
+	}
+
+	/* Allocate TX hardware ring descriptors. */
+	ring_size = sizeof(struct avf_tx_desc) * AVF_MAX_RING_DESC;
+	ring_size = RTE_ALIGN(ring_size, AVF_DMA_MEM_ALIGN);
+	mz = rte_eth_dma_zone_reserve(dev, "tx_ring", queue_idx,
+				      ring_size, AVF_RING_BASE_ALIGN,
+				      socket_id);
+	if (!mz) {
+		PMD_INIT_LOG(ERR, "Failed to reserve DMA memory for TX");
+		rte_free(txq->sw_ring);
+		rte_free(txq);
+		return -ENOMEM;
+	}
+	txq->tx_ring_phys_addr = mz->iova;
+	txq->tx_ring = (struct avf_tx_desc *)mz->addr;
+
+	txq->mz = mz;
+	reset_tx_queue(txq);
+	txq->q_set = TRUE;
+	dev->data->tx_queues[queue_idx] = txq;
+	txq->qtx_tail = hw->hw_addr + AVF_QTX_TAIL1(queue_idx);
+
+	return 0;
+}
+
+int
+avf_dev_rx_queue_start(struct rte_eth_dev *dev, uint16_t rx_queue_id)
+{
+	struct avf_adapter *adapter =
+		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
+	struct avf_hw *hw = AVF_DEV_PRIVATE_TO_HW(dev->data->dev_private);
+	struct avf_rx_queue *rxq;
+	int err = 0;
+
+	PMD_DRV_FUNC_TRACE();
+
+	if (rx_queue_id >= dev->data->nb_rx_queues)
+		return -EINVAL;
+
+	rxq = dev->data->rx_queues[rx_queue_id];
+
+	err = alloc_rxq_mbufs(rxq);
+	if (err) {
+		PMD_DRV_LOG(ERR, "Failed to allocate RX queue mbuf");
+		return err;
+	}
+
+	rte_wmb();
+
+	/* Init the RX tail register. */
+	AVF_PCI_REG_WRITE(rxq->qrx_tail, rxq->nb_rx_desc - 1);
+	AVF_WRITE_FLUSH(hw);
+
+	/* Ready to switch the queue on */
+	err = avf_switch_queue(adapter, rx_queue_id, TRUE, TRUE);
+	if (err)
+		PMD_DRV_LOG(ERR, "Failed to switch RX queue %u on",
+			    rx_queue_id);
+	else
+		dev->data->rx_queue_state[rx_queue_id] =
+			RTE_ETH_QUEUE_STATE_STARTED;
+
+	return err;
+}
+
+int
+avf_dev_tx_queue_start(struct rte_eth_dev *dev, uint16_t tx_queue_id)
+{
+	struct avf_adapter *adapter =
+		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
+	struct avf_hw *hw = AVF_DEV_PRIVATE_TO_HW(dev->data->dev_private);
+	struct avf_tx_queue *txq;
+	int err = 0;
+
+	PMD_DRV_FUNC_TRACE();
+
+	if (tx_queue_id >= dev->data->nb_tx_queues)
+		return -EINVAL;
+
+	txq = dev->data->tx_queues[tx_queue_id];
+
+	/* Init the RX tail register. */
+	AVF_PCI_REG_WRITE(txq->qtx_tail, 0);
+	AVF_WRITE_FLUSH(hw);
+
+	/* Ready to switch the queue on */
+	err = avf_switch_queue(adapter, tx_queue_id, FALSE, TRUE);
+
+	if (err)
+		PMD_DRV_LOG(ERR, "Failed to switch TX queue %u on",
+			    tx_queue_id);
+	else
+		dev->data->tx_queue_state[tx_queue_id] =
+			RTE_ETH_QUEUE_STATE_STARTED;
+
+	return err;
+}
+
+int
+avf_dev_rx_queue_stop(struct rte_eth_dev *dev, uint16_t rx_queue_id)
+{
+	struct avf_adapter *adapter =
+		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
+	struct avf_rx_queue *rxq;
+	int err;
+
+	PMD_DRV_FUNC_TRACE();
+
+	if (rx_queue_id >= dev->data->nb_rx_queues)
+		return -EINVAL;
+
+	err = avf_switch_queue(adapter, rx_queue_id, TRUE, FALSE);
+	if (err) {
+		PMD_DRV_LOG(ERR, "Failed to switch RX queue %u off",
+			    rx_queue_id);
+		return err;
+	}
+
+	rxq = dev->data->rx_queues[rx_queue_id];
+	release_rxq_mbufs(rxq);
+	reset_rx_queue(rxq);
+	dev->data->rx_queue_state[rx_queue_id] = RTE_ETH_QUEUE_STATE_STOPPED;
+
+	return 0;
+}
+
+int
+avf_dev_tx_queue_stop(struct rte_eth_dev *dev, uint16_t tx_queue_id)
+{
+	struct avf_adapter *adapter =
+		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
+	struct avf_tx_queue *txq;
+	int err;
+
+	PMD_DRV_FUNC_TRACE();
+
+	if (tx_queue_id >= dev->data->nb_tx_queues)
+		return -EINVAL;
+
+	err = avf_switch_queue(adapter, tx_queue_id, FALSE, FALSE);
+	if (err) {
+		PMD_DRV_LOG(ERR, "Failed to switch TX queue %u off",
+			    tx_queue_id);
+		return err;
+	}
+
+	txq = dev->data->tx_queues[tx_queue_id];
+	release_txq_mbufs(txq);
+	reset_tx_queue(txq);
+	dev->data->tx_queue_state[tx_queue_id] = RTE_ETH_QUEUE_STATE_STOPPED;
+
+	return 0;
+}
+
+void
+avf_dev_rx_queue_release(void *rxq)
+{
+	struct avf_rx_queue *q = (struct avf_rx_queue *)rxq;
+
+	if (!q)
+		return;
+
+	release_rxq_mbufs(q);
+	rte_free(q->sw_ring);
+	rte_memzone_free(q->mz);
+	rte_free(q);
+}
+
+void
+avf_dev_tx_queue_release(void *txq)
+{
+	struct avf_tx_queue *q = (struct avf_tx_queue *)txq;
+
+	if (!q)
+		return;
+
+	release_txq_mbufs(q);
+	rte_free(q->sw_ring);
+	rte_memzone_free(q->mz);
+	rte_free(q);
+}
+
+void
+avf_stop_queues(struct rte_eth_dev *dev)
+{
+	struct avf_adapter *adapter =
+		AVF_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private);
+	struct avf_rx_queue *rxq;
+	struct avf_tx_queue *txq;
+	int ret, i;
+
+	/* Stop All queues */
+	ret = avf_disable_queues(adapter);
+	if (ret)
+		PMD_DRV_LOG(WARNING, "Fail to stop queues");
+
+	for (i = 0; i < dev->data->nb_tx_queues; i++) {
+		txq = dev->data->tx_queues[i];
+		if (!txq)
+			continue;
+		release_txq_mbufs(txq);
+		reset_tx_queue(txq);
+		dev->data->tx_queue_state[i] = RTE_ETH_QUEUE_STATE_STOPPED;
+	}
+	for (i = 0; i < dev->data->nb_rx_queues; i++) {
+		rxq = dev->data->rx_queues[i];
+		if (!rxq)
+			continue;
+		release_rxq_mbufs(rxq);
+		reset_rx_queue(rxq);
+		dev->data->rx_queue_state[i] = RTE_ETH_QUEUE_STATE_STOPPED;
+	}
+}
diff --git a/drivers/net/avf/avf_rxtx.h b/drivers/net/avf/avf_rxtx.h
new file mode 100644
index 0000000..e227cd1
--- /dev/null
+++ b/drivers/net/avf/avf_rxtx.h
@@ -0,0 +1,160 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2017 Intel Corporation
+ */
+
+#ifndef _AVF_RXTX_H_
+#define _AVF_RXTX_H_
+
+/* In QLEN must be whole number of 32 descriptors. */
+#define AVF_ALIGN_RING_DESC      32
+#define AVF_MIN_RING_DESC        64
+#define AVF_MAX_RING_DESC        4096
+#define AVF_DMA_MEM_ALIGN        4096
+/* Base address of the HW descriptor ring should be 128B aligned. */
+#define AVF_RING_BASE_ALIGN      128
+
+/* used for Rx Bulk Allocate */
+#define AVF_RX_MAX_BURST         32
+
+#define DEFAULT_TX_RS_THRESH     32
+#define DEFAULT_TX_FREE_THRESH   32
+
+/* HW desc structure, both 16-byte and 32-byte types are supported */
+#ifdef RTE_LIBRTE_AVF_16BYTE_RX_DESC
+#define avf_rx_desc avf_16byte_rx_desc
+#else
+#define avf_rx_desc avf_32byte_rx_desc
+#endif
+
+/* Structure associated with each Rx queue. */
+struct avf_rx_queue {
+	struct rte_mempool *mp;       /* mbuf pool to populate Rx ring */
+	const struct rte_memzone *mz; /* memzone for Rx ring */
+	volatile union avf_rx_desc *rx_ring; /* Rx ring virtual address */
+	uint64_t rx_ring_phys_addr;   /* Rx ring DMA address */
+	struct rte_mbuf **sw_ring;     /* address of SW ring */
+	uint16_t nb_rx_desc;          /* ring length */
+	uint16_t rx_tail;             /* current value of tail */
+	volatile uint8_t *qrx_tail;   /* register address of tail */
+	uint16_t rx_free_thresh;      /* max free RX desc to hold */
+	uint16_t nb_rx_hold;          /* number of held free RX desc */
+	struct rte_mbuf *pkt_first_seg; /* first segment of current packet */
+	struct rte_mbuf *pkt_last_seg;  /* last segment of current packet */
+	struct rte_mbuf fake_mbuf;      /* dummy mbuf */
+
+	uint16_t port_id;       /* device port ID */
+	uint8_t crc_len;        /* 0 if CRC stripped, 4 otherwise */
+	uint16_t queue_id;      /* Rx queue index */
+	uint16_t rx_buf_len;    /* The packet buffer size */
+	uint16_t rx_hdr_len;    /* The header buffer size */
+	uint16_t max_pkt_len;   /* Maximum packet length */
+
+	bool q_set;             /* if rx queue has been configured */
+	bool rx_deferred_start; /* don't start this queue in dev start */
+};
+
+struct avf_tx_entry {
+	struct rte_mbuf *mbuf;
+	uint16_t next_id;
+	uint16_t last_id;
+};
+
+/* Structure associated with each TX queue. */
+struct avf_tx_queue {
+	const struct rte_memzone *mz;  /* memzone for Tx ring */
+	volatile struct avf_tx_desc *tx_ring; /* Tx ring virtual address */
+	uint64_t tx_ring_phys_addr;    /* Tx ring DMA address */
+	struct avf_tx_entry *sw_ring;  /* address array of SW ring */
+	uint16_t nb_tx_desc;           /* ring length */
+	uint16_t tx_tail;              /* current value of tail */
+	volatile uint8_t *qtx_tail;    /* register address of tail */
+	/* number of used desc since RS bit set */
+	uint16_t nb_used;
+	uint16_t nb_free;
+	uint16_t last_desc_cleaned;    /* last desc have been cleaned*/
+	uint16_t free_thresh;
+	uint16_t rs_thresh;
+
+	uint16_t port_id;
+	uint16_t queue_id;
+	uint32_t txq_flags;
+	uint16_t next_dd;              /* next to set RS, for VPMD */
+	uint16_t next_rs;              /* next to check DD,  for VPMD */
+
+	bool q_set;                    /* if rx queue has been configured */
+	bool tx_deferred_start;        /* don't start this queue in dev start */
+};
+
+int avf_dev_rx_queue_setup(struct rte_eth_dev *dev,
+			   uint16_t queue_idx,
+			   uint16_t nb_desc,
+			   unsigned int socket_id,
+			   const struct rte_eth_rxconf *rx_conf,
+			   struct rte_mempool *mp);
+
+int avf_dev_rx_queue_start(struct rte_eth_dev *dev, uint16_t rx_queue_id);
+int avf_dev_rx_queue_stop(struct rte_eth_dev *dev, uint16_t rx_queue_id);
+void avf_dev_rx_queue_release(void *rxq);
+
+int avf_dev_tx_queue_setup(struct rte_eth_dev *dev,
+			   uint16_t queue_idx,
+			   uint16_t nb_desc,
+			   unsigned int socket_id,
+			   const struct rte_eth_txconf *tx_conf);
+int avf_dev_tx_queue_start(struct rte_eth_dev *dev, uint16_t tx_queue_id);
+int avf_dev_tx_queue_stop(struct rte_eth_dev *dev, uint16_t tx_queue_id);
+void avf_dev_tx_queue_release(void *txq);
+void avf_stop_queues(struct rte_eth_dev *dev);
+
+static inline
+void avf_dump_rx_descriptor(struct avf_rx_queue *rxq,
+			    const void *desc,
+			    uint16_t rx_id)
+{
+#ifdef RTE_LIBRTE_AVF_16BYTE_RX_DESC
+	const union avf_16byte_rx_desc *rx_desc = desc;
+
+	printf("Queue %d Rx_desc %d: QW0: 0x%016"PRIx64" QW1: 0x%016"PRIx64"\n",
+	       rxq->queue_id, rx_id, rx_desc->read.pkt_addr,
+	       rx_desc->read.hdr_addr);
+#else
+	const union avf_32byte_rx_desc *rx_desc = desc;
+
+	printf("Queue %d Rx_desc %d: QW0: 0x%016"PRIx64" QW1: 0x%016"PRIx64
+	       " QW2: 0x%016"PRIx64" QW3: 0x%016"PRIx64"\n", rxq->queue_id,
+	       rx_id, rx_desc->read.pkt_addr, rx_desc->read.hdr_addr,
+	       rx_desc->read.rsvd1, rx_desc->read.rsvd2);
+#endif
+}
+
+/* All the descriptors are 16 bytes, so just use one of them
+ * to print the qwords
+ */
+static inline
+void avf_dump_tx_descriptor(const struct avf_tx_queue *txq,
+			    const void *desc, uint16_t tx_id)
+{
+	char *name;
+	const struct avf_tx_desc *tx_desc = desc;
+	enum avf_tx_desc_dtype_value type;
+
+	type = (enum avf_tx_desc_dtype_value)rte_le_to_cpu_64(
+		tx_desc->cmd_type_offset_bsz &
+		rte_cpu_to_le_64(AVF_TXD_QW1_DTYPE_MASK));
+	switch (type) {
+	case AVF_TX_DESC_DTYPE_DATA:
+		name = "Tx_data_desc";
+		break;
+	case AVF_TX_DESC_DTYPE_CONTEXT:
+		name = "Tx_context_desc";
+		break;
+	default:
+		name = "unknown_desc";
+		break;
+	}
+
+	printf("Queue %d %s %d: QW0: 0x%016"PRIx64" QW1: 0x%016"PRIx64"\n",
+	       txq->queue_id, name, tx_id, tx_desc->buffer_addr,
+	       tx_desc->cmd_type_offset_bsz);
+}
+#endif /* _AVF_RXTX_H_ */
diff --git a/drivers/net/avf/avf_vchnl.c b/drivers/net/avf/avf_vchnl.c
index ebbee31..55a425a 100644
--- a/drivers/net/avf/avf_vchnl.c
+++ b/drivers/net/avf/avf_vchnl.c
@@ -25,6 +25,7 @@
 #include "base/avf_type.h"
 
 #include "avf.h"
+#include "avf_rxtx.h"
 
 #define MAX_TRY_TIMES 200
 #define ASQ_DELAY_MS  10
@@ -196,6 +197,48 @@
 	}
 }
 
+int
+avf_enable_vlan_strip(struct avf_adapter *adapter)
+{
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(adapter);
+	struct avf_cmd_info args;
+	int ret;
+
+	memset(&args, 0, sizeof(args));
+	args.ops = VIRTCHNL_OP_ENABLE_VLAN_STRIPPING;
+	args.in_args = NULL;
+	args.in_args_size = 0;
+	args.out_buffer = vf->aq_resp;
+	args.out_size = AVF_AQ_BUF_SZ;
+	ret = avf_execute_vf_cmd(adapter, &args);
+	if (ret)
+		PMD_DRV_LOG(ERR, "Failed to execute command of"
+			    " OP_ENABLE_VLAN_STRIPPING");
+
+	return ret;
+}
+
+int
+avf_disable_vlan_strip(struct avf_adapter *adapter)
+{
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(adapter);
+	struct avf_cmd_info args;
+	int ret;
+
+	memset(&args, 0, sizeof(args));
+	args.ops = VIRTCHNL_OP_DISABLE_VLAN_STRIPPING;
+	args.in_args = NULL;
+	args.in_args_size = 0;
+	args.out_buffer = vf->aq_resp;
+	args.out_size = AVF_AQ_BUF_SZ;
+	ret = avf_execute_vf_cmd(adapter, &args);
+	if (ret)
+		PMD_DRV_LOG(ERR, "Failed to execute command of"
+			    " OP_DISABLE_VLAN_STRIPPING");
+
+	return ret;
+}
+
 #define VIRTCHNL_VERSION_MAJOR_START 1
 #define VIRTCHNL_VERSION_MINOR_START 1
 
@@ -274,8 +317,8 @@
 	err = avf_execute_vf_cmd(adapter, &args);
 
 	if (err) {
-		PMD_DRV_LOG(ERR, "Failed to execute command of "
-				 "OP_GET_VF_RESOURCE");
+		PMD_DRV_LOG(ERR,
+			    "Failed to execute command of OP_GET_VF_RESOURCE");
 		return -1;
 	}
 
@@ -302,3 +345,315 @@
 
 	return 0;
 }
+
+int
+avf_enable_queues(struct avf_adapter *adapter)
+{
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(adapter);
+	struct virtchnl_queue_select queue_select;
+	struct avf_cmd_info args;
+	int err;
+
+	memset(&queue_select, 0, sizeof(queue_select));
+	queue_select.vsi_id = vf->vsi_res->vsi_id;
+
+	queue_select.rx_queues = BIT(adapter->eth_dev->data->nb_rx_queues) - 1;
+	queue_select.tx_queues = BIT(adapter->eth_dev->data->nb_tx_queues) - 1;
+
+	args.ops = VIRTCHNL_OP_ENABLE_QUEUES;
+	args.in_args = (u8 *)&queue_select;
+	args.in_args_size = sizeof(queue_select);
+	args.out_buffer = vf->aq_resp;
+	args.out_size = AVF_AQ_BUF_SZ;
+	err = avf_execute_vf_cmd(adapter, &args);
+	if (err) {
+		PMD_DRV_LOG(ERR,
+			    "Failed to execute command of OP_ENABLE_QUEUES");
+		return err;
+	}
+	return 0;
+}
+
+int
+avf_disable_queues(struct avf_adapter *adapter)
+{
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(adapter);
+	struct virtchnl_queue_select queue_select;
+	struct avf_cmd_info args;
+	int err;
+
+	memset(&queue_select, 0, sizeof(queue_select));
+	queue_select.vsi_id = vf->vsi_res->vsi_id;
+
+	queue_select.rx_queues = BIT(adapter->eth_dev->data->nb_rx_queues) - 1;
+	queue_select.tx_queues = BIT(adapter->eth_dev->data->nb_tx_queues) - 1;
+
+	args.ops = VIRTCHNL_OP_DISABLE_QUEUES;
+	args.in_args = (u8 *)&queue_select;
+	args.in_args_size = sizeof(queue_select);
+	args.out_buffer = vf->aq_resp;
+	args.out_size = AVF_AQ_BUF_SZ;
+	err = avf_execute_vf_cmd(adapter, &args);
+	if (err) {
+		PMD_DRV_LOG(ERR,
+			    "Failed to execute command of OP_DISABLE_QUEUES");
+		return err;
+	}
+	return 0;
+}
+
+int
+avf_switch_queue(struct avf_adapter *adapter, uint16_t qid,
+		 bool rx, bool on)
+{
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(adapter);
+	struct virtchnl_queue_select queue_select;
+	struct avf_cmd_info args;
+	int err;
+
+	memset(&queue_select, 0, sizeof(queue_select));
+	queue_select.vsi_id = vf->vsi_res->vsi_id;
+	if (rx)
+		queue_select.rx_queues |= 1 << qid;
+	else
+		queue_select.tx_queues |= 1 << qid;
+
+	if (on)
+		args.ops = VIRTCHNL_OP_ENABLE_QUEUES;
+	else
+		args.ops = VIRTCHNL_OP_DISABLE_QUEUES;
+	args.in_args = (u8 *)&queue_select;
+	args.in_args_size = sizeof(queue_select);
+	args.out_buffer = vf->aq_resp;
+	args.out_size = AVF_AQ_BUF_SZ;
+	err = avf_execute_vf_cmd(adapter, &args);
+	if (err)
+		PMD_DRV_LOG(ERR, "Failed to execute command of %s",
+			    on ? "OP_ENABLE_QUEUES" : "OP_DISABLE_QUEUES");
+	return err;
+}
+
+int
+avf_configure_rss_lut(struct avf_adapter *adapter)
+{
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(adapter);
+	struct virtchnl_rss_lut *rss_lut;
+	struct avf_cmd_info args;
+	int len, err = 0;
+
+	len = sizeof(*rss_lut) + vf->vf_res->rss_lut_size - 1;
+	rss_lut = rte_zmalloc("rss_lut", len, 0);
+	if (!rss_lut)
+		return -ENOMEM;
+
+	rss_lut->vsi_id = vf->vsi_res->vsi_id;
+	rss_lut->lut_entries = vf->vf_res->rss_lut_size;
+	rte_memcpy(rss_lut->lut, vf->rss_lut, vf->vf_res->rss_lut_size);
+
+	args.ops = VIRTCHNL_OP_CONFIG_RSS_LUT;
+	args.in_args = (u8 *)rss_lut;
+	args.in_args_size = len;
+	args.out_buffer = vf->aq_resp;
+	args.out_size = AVF_AQ_BUF_SZ;
+
+	err = avf_execute_vf_cmd(adapter, &args);
+	if (err)
+		PMD_DRV_LOG(ERR,
+			    "Failed to execute command of OP_CONFIG_RSS_LUT");
+
+	rte_free(rss_lut);
+	return err;
+}
+
+int
+avf_configure_rss_key(struct avf_adapter *adapter)
+{
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(adapter);
+	struct virtchnl_rss_key *rss_key;
+	struct avf_cmd_info args;
+	int len, err = 0;
+
+	len = sizeof(*rss_key) + vf->vf_res->rss_key_size - 1;
+	rss_key = rte_zmalloc("rss_key", len, 0);
+	if (!rss_key)
+		return -ENOMEM;
+
+	rss_key->vsi_id = vf->vsi_res->vsi_id;
+	rss_key->key_len = vf->vf_res->rss_key_size;
+	rte_memcpy(rss_key->key, vf->rss_key, vf->vf_res->rss_key_size);
+
+	args.ops = VIRTCHNL_OP_CONFIG_RSS_KEY;
+	args.in_args = (u8 *)rss_key;
+	args.in_args_size = len;
+	args.out_buffer = vf->aq_resp;
+	args.out_size = AVF_AQ_BUF_SZ;
+
+	err = avf_execute_vf_cmd(adapter, &args);
+	if (err)
+		PMD_DRV_LOG(ERR,
+			    "Failed to execute command of OP_CONFIG_RSS_KEY");
+
+	rte_free(rss_key);
+	return err;
+}
+
+int
+avf_configure_queues(struct avf_adapter *adapter)
+{
+	struct avf_rx_queue **rxq =
+		(struct avf_rx_queue **)adapter->eth_dev->data->rx_queues;
+	struct avf_tx_queue **txq =
+		(struct avf_tx_queue **)adapter->eth_dev->data->tx_queues;
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(adapter);
+	struct virtchnl_vsi_queue_config_info *vc_config;
+	struct virtchnl_queue_pair_info *vc_qp;
+	struct avf_cmd_info args;
+	uint16_t i, size;
+	int err;
+
+	size = sizeof(*vc_config) +
+	       sizeof(vc_config->qpair[0]) * vf->num_queue_pairs;
+	vc_config = rte_zmalloc("cfg_queue", size, 0);
+	if (!vc_config)
+		return -ENOMEM;
+
+	vc_config->vsi_id = vf->vsi_res->vsi_id;
+	vc_config->num_queue_pairs = vf->num_queue_pairs;
+
+	for (i = 0, vc_qp = vc_config->qpair;
+	     i < vf->num_queue_pairs;
+	     i++, vc_qp++) {
+		vc_qp->txq.vsi_id = vf->vsi_res->vsi_id;
+		vc_qp->txq.queue_id = i;
+		/* Virtchnnl configure queues by pairs */
+		if (i < adapter->eth_dev->data->nb_tx_queues) {
+			vc_qp->txq.ring_len = txq[i]->nb_tx_desc;
+			vc_qp->txq.dma_ring_addr = txq[i]->tx_ring_phys_addr;
+		}
+		vc_qp->rxq.vsi_id = vf->vsi_res->vsi_id;
+		vc_qp->rxq.queue_id = i;
+		vc_qp->rxq.max_pkt_size = vf->max_pkt_len;
+		/* Virtchnnl configure queues by pairs */
+		if (i < adapter->eth_dev->data->nb_rx_queues) {
+			vc_qp->rxq.ring_len = rxq[i]->nb_rx_desc;
+			vc_qp->rxq.dma_ring_addr = rxq[i]->rx_ring_phys_addr;
+			vc_qp->rxq.databuffer_size = rxq[i]->rx_buf_len;
+		}
+	}
+
+	memset(&args, 0, sizeof(args));
+	args.ops = VIRTCHNL_OP_CONFIG_VSI_QUEUES;
+	args.in_args = (uint8_t *)vc_config;
+	args.in_args_size = size;
+	args.out_buffer = vf->aq_resp;
+	args.out_size = AVF_AQ_BUF_SZ;
+
+	err = avf_execute_vf_cmd(adapter, &args);
+	if (err)
+		PMD_DRV_LOG(ERR, "Failed to execute command of"
+			    " VIRTCHNL_OP_CONFIG_VSI_QUEUES");
+
+	rte_free(vc_config);
+	return err;
+}
+
+int
+avf_config_irq_map(struct avf_adapter *adapter)
+{
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(adapter);
+	struct virtchnl_irq_map_info *map_info;
+	struct virtchnl_vector_map *vecmap;
+	struct avf_cmd_info args;
+	uint32_t vector_id;
+	int len, i, err;
+
+	len = sizeof(struct virtchnl_irq_map_info) +
+	      sizeof(struct virtchnl_vector_map) * vf->nb_msix;
+
+	map_info = rte_zmalloc("map_info", len, 0);
+	if (!map_info)
+		return -ENOMEM;
+
+	map_info->num_vectors = vf->nb_msix;
+	for (i = 0; i < vf->nb_msix; i++) {
+		vecmap = &map_info->vecmap[i];
+		vecmap->vsi_id = vf->vsi_res->vsi_id;
+		vecmap->rxitr_idx = AVF_ITR_INDEX_DEFAULT;
+		vecmap->vector_id = vf->msix_base + i;
+		vecmap->txq_map = 0;
+		vecmap->rxq_map = vf->rxq_map[vf->msix_base + i];
+	}
+
+	args.ops = VIRTCHNL_OP_CONFIG_IRQ_MAP;
+	args.in_args = (u8 *)map_info;
+	args.in_args_size = len;
+	args.out_buffer = vf->aq_resp;
+	args.out_size = AVF_AQ_BUF_SZ;
+	err = avf_execute_vf_cmd(adapter, &args);
+	if (err)
+		PMD_DRV_LOG(ERR, "fail to execute command OP_CONFIG_IRQ_MAP");
+
+	rte_free(map_info);
+	return err;
+}
+
+void
+avf_add_del_all_mac_addr(struct avf_adapter *adapter, bool add)
+{
+	struct virtchnl_ether_addr_list *list;
+	struct avf_info *vf = AVF_DEV_PRIVATE_TO_VF(adapter);
+	struct ether_addr *addr;
+	struct avf_cmd_info args;
+	int len, err, i, j;
+	int next_begin = 0;
+	int begin = 0;
+
+	do {
+		j = 0;
+		len = sizeof(struct virtchnl_ether_addr_list);
+		for (i = begin; i < AVF_NUM_MACADDR_MAX; i++, next_begin++) {
+			addr = &adapter->eth_dev->data->mac_addrs[i];
+			if (is_zero_ether_addr(addr))
+				continue;
+			len += sizeof(struct virtchnl_ether_addr);
+			if (len >= AVF_AQ_BUF_SZ) {
+				next_begin = i + 1;
+				break;
+			}
+		}
+
+		list = rte_zmalloc("avf_del_mac_buffer", len, 0);
+		if (!list) {
+			PMD_DRV_LOG(ERR, "fail to allocate memory");
+			return;
+		}
+
+		for (i = begin; i < next_begin; i++) {
+			addr = &adapter->eth_dev->data->mac_addrs[i];
+			if (is_zero_ether_addr(addr))
+				continue;
+			rte_memcpy(list->list[j].addr, addr->addr_bytes,
+				   sizeof(addr->addr_bytes));
+			PMD_DRV_LOG(DEBUG, "add/rm mac:%x:%x:%x:%x:%x:%x",
+				    addr->addr_bytes[0], addr->addr_bytes[1],
+				    addr->addr_bytes[2], addr->addr_bytes[3],
+				    addr->addr_bytes[4], addr->addr_bytes[5]);
+			j++;
+		}
+		list->vsi_id = vf->vsi_res->vsi_id;
+		list->num_elements = j;
+		args.ops = add ? VIRTCHNL_OP_ADD_ETH_ADDR :
+			   VIRTCHNL_OP_DEL_ETH_ADDR;
+		args.in_args = (uint8_t *)list;
+		args.in_args_size = len;
+		args.out_buffer = vf->aq_resp;
+		args.out_size = AVF_AQ_BUF_SZ;
+		err = avf_execute_vf_cmd(adapter, &args);
+		if (err)
+			PMD_DRV_LOG(ERR, "fail to execute command %s",
+				    add ? "OP_ADD_ETHER_ADDRESS" :
+				    "OP_DEL_ETHER_ADDRESS");
+		rte_free(list);
+		begin = next_begin;
+	} while (begin < AVF_NUM_MACADDR_MAX);
+}
-- 
1.9.3



More information about the dev mailing list