[PATCH v2 2/3] net/iavf: enable Rx timestamp on Flex Descriptor

[Simei Su <simei.su@intel.com>]

Dump Rx timestamp value into dynamic mbuf field by flex descriptor.
This feature is turned on by dev config "enable-rx-timestamp".
Currently, it's only supported under scalar path.

Signed-off-by: Simei Su <simei.su@intel.com>
---
 doc/guides/nics/features/iavf.ini       |  1 +
 doc/guides/rel_notes/release_22_07.rst  |  1 +
 drivers/net/iavf/iavf.h                 |  5 ++
 drivers/net/iavf/iavf_ethdev.c          | 26 +++++++++++
 drivers/net/iavf/iavf_rxtx.c            | 58 +++++++++++++++++++++++
 drivers/net/iavf/iavf_rxtx.h            | 22 +++++++++
 drivers/net/iavf/iavf_rxtx_vec_common.h |  3 ++
 drivers/net/iavf/iavf_vchnl.c           | 83 ++++++++++++++++++++++++++++-----
 8 files changed, 188 insertions(+), 11 deletions(-)

diff --git a/doc/guides/nics/features/iavf.ini b/doc/guides/nics/features/iavf.ini
index 01f5142..5a0d9d8 100644
--- a/doc/guides/nics/features/iavf.ini
+++ b/doc/guides/nics/features/iavf.ini
@@ -24,6 +24,7 @@ CRC offload          = Y
 VLAN offload         = Y
 L3 checksum offload  = P
 L4 checksum offload  = P
+Timestamp offload    = P
 Packet type parsing  = Y
 Rx descriptor status = Y
 Tx descriptor status = Y
diff --git a/doc/guides/rel_notes/release_22_07.rst b/doc/guides/rel_notes/release_22_07.rst
index f1b4057..567f23d 100644
--- a/doc/guides/rel_notes/release_22_07.rst
+++ b/doc/guides/rel_notes/release_22_07.rst
@@ -59,6 +59,7 @@ New Features
 
   * Added Tx QoS queue rate limitation support.
   * Added quanta size configuration support.
+  * Added ``DEV_RX_OFFLOAD_TIMESTAMP`` support.
 
 Removed Items
 -------------
diff --git a/drivers/net/iavf/iavf.h b/drivers/net/iavf/iavf.h
index c0a4a47..3255c93 100644
--- a/drivers/net/iavf/iavf.h
+++ b/drivers/net/iavf/iavf.h
@@ -268,6 +268,8 @@ struct iavf_info {
 	struct iavf_tm_conf tm_conf;
 
 	struct rte_eth_dev *eth_dev;
+
+	uint32_t ptp_caps;
 };
 
 #define IAVF_MAX_PKT_TYPE 1024
@@ -312,6 +314,7 @@ struct iavf_adapter {
 	bool stopped;
 	uint16_t fdir_ref_cnt;
 	struct iavf_devargs devargs;
+	uint64_t phc_time;
 };
 
 /* IAVF_DEV_PRIVATE_TO */
@@ -476,4 +479,6 @@ int iavf_ipsec_crypto_request(struct iavf_adapter *adapter,
 		uint8_t *msg, size_t msg_len,
 		uint8_t *resp_msg, size_t resp_msg_len);
 extern const struct rte_tm_ops iavf_tm_ops;
+int iavf_get_ptp_cap(struct iavf_adapter *adapter);
+int iavf_get_phc_time(struct iavf_adapter *adapter);
 #endif /* _IAVF_ETHDEV_H_ */
diff --git a/drivers/net/iavf/iavf_ethdev.c b/drivers/net/iavf/iavf_ethdev.c
index 7d093bd..89e4240 100644
--- a/drivers/net/iavf/iavf_ethdev.c
+++ b/drivers/net/iavf/iavf_ethdev.c
@@ -36,6 +36,9 @@
 #define IAVF_PROTO_XTR_ARG         "proto_xtr"
 #define IAVF_QUANTA_SIZE_ARG       "quanta_size"
 
+uint64_t iavf_timestamp_dynflag;
+int iavf_timestamp_dynfield_offset = -1;
+
 static const char * const iavf_valid_args[] = {
 	IAVF_PROTO_XTR_ARG,
 	IAVF_QUANTA_SIZE_ARG,
@@ -687,6 +690,7 @@ iavf_init_rxq(struct rte_eth_dev *dev, struct iavf_rx_queue *rxq)
 	struct rte_eth_dev_data *dev_data = dev->data;
 	uint16_t buf_size, max_pkt_len;
 	uint32_t frame_size = dev->data->mtu + IAVF_ETH_OVERHEAD;
+	enum iavf_status err;
 
 	buf_size = rte_pktmbuf_data_room_size(rxq->mp) - RTE_PKTMBUF_HEADROOM;
 
@@ -705,6 +709,18 @@ iavf_init_rxq(struct rte_eth_dev *dev, struct iavf_rx_queue *rxq)
 		return -EINVAL;
 	}
 
+	if (rxq->offloads & RTE_ETH_RX_OFFLOAD_TIMESTAMP) {
+		/* Register mbuf field and flag for Rx timestamp */
+		err = rte_mbuf_dyn_rx_timestamp_register(
+			&iavf_timestamp_dynfield_offset,
+			&iavf_timestamp_dynflag);
+		if (err) {
+			PMD_DRV_LOG(ERR,
+				    "Cannot register mbuf field/flag for timestamp");
+			return -EINVAL;
+		}
+	}
+
 	rxq->max_pkt_len = max_pkt_len;
 	if ((dev_data->dev_conf.rxmode.offloads & RTE_ETH_RX_OFFLOAD_SCATTER) ||
 	    rxq->max_pkt_len > buf_size) {
@@ -947,6 +963,13 @@ iavf_dev_start(struct rte_eth_dev *dev)
 			return -1;
 		}
 
+	if (vf->vf_res->vf_cap_flags & VIRTCHNL_VF_CAP_PTP) {
+		if (iavf_get_ptp_cap(adapter)) {
+			PMD_INIT_LOG(ERR, "Failed to get ptp capability");
+			return -1;
+		}
+	}
+
 	if (iavf_init_queues(dev) != 0) {
 		PMD_DRV_LOG(ERR, "failed to do Queue init");
 		return -1;
@@ -1092,6 +1115,9 @@ iavf_dev_info_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
 	if (vf->vf_res->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_CRC)
 		dev_info->rx_offload_capa |= RTE_ETH_RX_OFFLOAD_KEEP_CRC;
 
+	if (vf->vf_res->vf_cap_flags & VIRTCHNL_VF_CAP_PTP)
+		dev_info->rx_offload_capa |= RTE_ETH_RX_OFFLOAD_TIMESTAMP;
+
 	if (iavf_ipsec_crypto_supported(adapter)) {
 		dev_info->rx_offload_capa |= RTE_ETH_RX_OFFLOAD_SECURITY;
 		dev_info->tx_offload_capa |= RTE_ETH_TX_OFFLOAD_SECURITY;
diff --git a/drivers/net/iavf/iavf_rxtx.c b/drivers/net/iavf/iavf_rxtx.c
index c21f818..2d3bafd 100644
--- a/drivers/net/iavf/iavf_rxtx.c
+++ b/drivers/net/iavf/iavf_rxtx.c
@@ -1422,6 +1422,11 @@ iavf_recv_pkts_flex_rxd(void *rx_queue,
 	uint64_t dma_addr;
 	uint64_t pkt_flags;
 	const uint32_t *ptype_tbl;
+	struct iavf_adapter *ad = rxq->vsi->adapter;
+	uint64_t ts_ns;
+
+	if (rxq->offloads & RTE_ETH_RX_OFFLOAD_TIMESTAMP)
+		rxq->hw_register_set = 1;
 
 	nb_rx = 0;
 	nb_hold = 0;
@@ -1491,6 +1496,21 @@ iavf_recv_pkts_flex_rxd(void *rx_queue,
 				&rxq->stats.ipsec_crypto);
 		rxd_to_pkt_fields_ops[rxq->rxdid](rxq, rxm, &rxd);
 		pkt_flags = iavf_flex_rxd_error_to_pkt_flags(rx_stat_err0);
+
+		if (iavf_timestamp_dynflag > 0) {
+			if (rxq->hw_register_set)
+				iavf_get_phc_time(ad);
+
+			rxq->hw_register_set = 0;
+			ts_ns = iavf_tstamp_convert_32b_64b(ad->phc_time,
+				rte_le_to_cpu_32(rxd.wb.flex_ts.ts_high));
+
+			*RTE_MBUF_DYNFIELD(rxm,
+				iavf_timestamp_dynfield_offset,
+				rte_mbuf_timestamp_t *) = ts_ns;
+			rxm->ol_flags |= iavf_timestamp_dynflag;
+		}
+
 		rxm->ol_flags |= pkt_flags;
 
 		rx_pkts[nb_rx++] = rxm;
@@ -1519,11 +1539,16 @@ iavf_recv_scattered_pkts_flex_rxd(void *rx_queue, struct rte_mbuf **rx_pkts,
 	uint16_t rx_stat_err0;
 	uint64_t dma_addr;
 	uint64_t pkt_flags;
+	struct iavf_adapter *ad = rxq->vsi->adapter;
+	uint64_t ts_ns;
 
 	volatile union iavf_rx_desc *rx_ring = rxq->rx_ring;
 	volatile union iavf_rx_flex_desc *rxdp;
 	const uint32_t *ptype_tbl = rxq->vsi->adapter->ptype_tbl;
 
+	if (rxq->offloads & RTE_ETH_RX_OFFLOAD_TIMESTAMP)
+		rxq->hw_register_set = 1;
+
 	while (nb_rx < nb_pkts) {
 		rxdp = (volatile union iavf_rx_flex_desc *)&rx_ring[rx_id];
 		rx_stat_err0 = rte_le_to_cpu_16(rxdp->wb.status_error0);
@@ -1636,6 +1661,20 @@ iavf_recv_scattered_pkts_flex_rxd(void *rx_queue, struct rte_mbuf **rx_pkts,
 		rxd_to_pkt_fields_ops[rxq->rxdid](rxq, first_seg, &rxd);
 		pkt_flags = iavf_flex_rxd_error_to_pkt_flags(rx_stat_err0);
 
+		if (iavf_timestamp_dynflag > 0) {
+			if (rxq->hw_register_set)
+				iavf_get_phc_time(ad);
+
+			rxq->hw_register_set = 0;
+			ts_ns = iavf_tstamp_convert_32b_64b(ad->phc_time,
+				rte_le_to_cpu_32(rxd.wb.flex_ts.ts_high));
+
+			*RTE_MBUF_DYNFIELD(first_seg,
+				iavf_timestamp_dynfield_offset,
+				rte_mbuf_timestamp_t *) = ts_ns;
+			first_seg->ol_flags |= iavf_timestamp_dynflag;
+		}
+
 		first_seg->ol_flags |= pkt_flags;
 
 		/* Prefetch data of first segment, if configured to do so. */
@@ -1831,6 +1870,8 @@ iavf_rx_scan_hw_ring_flex_rxd(struct iavf_rx_queue *rxq,
 	int32_t nb_staged = 0;
 	uint64_t pkt_flags;
 	const uint32_t *ptype_tbl = rxq->vsi->adapter->ptype_tbl;
+	struct iavf_adapter *ad = rxq->vsi->adapter;
+	uint64_t ts_ns;
 
 	rxdp = (volatile union iavf_rx_flex_desc *)&rxq->rx_ring[rxq->rx_tail];
 	rxep = &rxq->sw_ring[rxq->rx_tail];
@@ -1841,6 +1882,9 @@ iavf_rx_scan_hw_ring_flex_rxd(struct iavf_rx_queue *rxq,
 	if (!(stat_err0 & (1 << IAVF_RX_FLEX_DESC_STATUS0_DD_S)))
 		return 0;
 
+	if (rxq->offloads & RTE_ETH_RX_OFFLOAD_TIMESTAMP)
+		rxq->hw_register_set = 1;
+
 	/* Scan LOOK_AHEAD descriptors at a time to determine which
 	 * descriptors reference packets that are ready to be received.
 	 */
@@ -1897,6 +1941,20 @@ iavf_rx_scan_hw_ring_flex_rxd(struct iavf_rx_queue *rxq,
 			stat_err0 = rte_le_to_cpu_16(rxdp[j].wb.status_error0);
 			pkt_flags = iavf_flex_rxd_error_to_pkt_flags(stat_err0);
 
+			if (iavf_timestamp_dynflag > 0) {
+				if (rxq->hw_register_set)
+					iavf_get_phc_time(ad);
+
+				rxq->hw_register_set = 0;
+				ts_ns = iavf_tstamp_convert_32b_64b(ad->phc_time,
+					rte_le_to_cpu_32(rxdp[j].wb.flex_ts.ts_high));
+
+				*RTE_MBUF_DYNFIELD(mb,
+					iavf_timestamp_dynfield_offset,
+					rte_mbuf_timestamp_t *) = ts_ns;
+				mb->ol_flags |= iavf_timestamp_dynflag;
+			}
+
 			mb->ol_flags |= pkt_flags;
 
 			/* Put up to nb_pkts directly into buffers */
diff --git a/drivers/net/iavf/iavf_rxtx.h b/drivers/net/iavf/iavf_rxtx.h
index bf8aebb..37453c4 100644
--- a/drivers/net/iavf/iavf_rxtx.h
+++ b/drivers/net/iavf/iavf_rxtx.h
@@ -72,6 +72,9 @@
 #define IAVF_TX_OFFLOAD_NOTSUP_MASK \
 		(RTE_MBUF_F_TX_OFFLOAD_MASK ^ IAVF_TX_OFFLOAD_MASK)
 
+extern uint64_t iavf_timestamp_dynflag;
+extern int iavf_timestamp_dynfield_offset;
+
 /**
  * Rx Flex Descriptors
  * These descriptors are used instead of the legacy version descriptors
@@ -219,6 +222,7 @@ struct iavf_rx_queue {
 		/* flexible descriptor metadata extraction offload flag */
 	struct iavf_rx_queue_stats stats;
 	uint64_t offloads;
+	uint32_t hw_register_set;
 };
 
 struct iavf_tx_entry {
@@ -778,6 +782,24 @@ void iavf_fdir_rx_proc_enable(struct iavf_adapter *ad, bool on)
 	}
 }
 
+static inline
+uint64_t iavf_tstamp_convert_32b_64b(uint64_t time, uint32_t in_timestamp)
+{
+	const uint64_t mask = 0xFFFFFFFF;
+	uint32_t delta;
+	uint64_t ns;
+
+	delta = (in_timestamp - (uint32_t)(time & mask));
+	if (delta > (mask / 2)) {
+		delta = ((uint32_t)(time & mask) - in_timestamp);
+		ns = time - delta;
+	} else {
+		ns = time + delta;
+	}
+
+	return ns;
+}
+
 #ifdef RTE_LIBRTE_IAVF_DEBUG_DUMP_DESC
 #define IAVF_DUMP_RX_DESC(rxq, desc, rx_id) \
 	iavf_dump_rx_descriptor(rxq, desc, rx_id)
diff --git a/drivers/net/iavf/iavf_rxtx_vec_common.h b/drivers/net/iavf/iavf_rxtx_vec_common.h
index 1fd37b7..a59cb2c 100644
--- a/drivers/net/iavf/iavf_rxtx_vec_common.h
+++ b/drivers/net/iavf/iavf_rxtx_vec_common.h
@@ -231,6 +231,9 @@ iavf_rx_vec_queue_default(struct iavf_rx_queue *rxq)
 	if (rxq->proto_xtr != IAVF_PROTO_XTR_NONE)
 		return -1;
 
+	if (rxq->offloads & RTE_ETH_RX_OFFLOAD_TIMESTAMP)
+		return -1;
+
 	if (rxq->offloads & IAVF_RX_VECTOR_OFFLOAD)
 		return IAVF_VECTOR_OFFLOAD_PATH;
 
diff --git a/drivers/net/iavf/iavf_vchnl.c b/drivers/net/iavf/iavf_vchnl.c
index f9452d1..b654433 100644
--- a/drivers/net/iavf/iavf_vchnl.c
+++ b/drivers/net/iavf/iavf_vchnl.c
@@ -502,7 +502,8 @@ iavf_get_vf_resource(struct iavf_adapter *adapter)
 		VIRTCHNL_VF_OFFLOAD_VLAN_V2 |
 		VIRTCHNL_VF_LARGE_NUM_QPAIRS |
 		VIRTCHNL_VF_OFFLOAD_QOS |
-		VIRTCHNL_VF_OFFLOAD_INLINE_IPSEC_CRYPTO;
+		VIRTCHNL_VF_OFFLOAD_INLINE_IPSEC_CRYPTO |
+		VIRTCHNL_VF_CAP_PTP;
 
 	args.in_args = (uint8_t *)&caps;
 	args.in_args_size = sizeof(caps);
@@ -1047,16 +1048,21 @@ iavf_configure_queues(struct iavf_adapter *adapter,
 		vc_qp->rxq.crc_disable = rxq[i]->crc_len != 0 ? 1 : 0;
 #ifndef RTE_LIBRTE_IAVF_16BYTE_RX_DESC
 		if (vf->vf_res->vf_cap_flags &
-		    VIRTCHNL_VF_OFFLOAD_RX_FLEX_DESC &&
-		    vf->supported_rxdid & BIT(rxq[i]->rxdid)) {
-			vc_qp->rxq.rxdid = rxq[i]->rxdid;
-			PMD_DRV_LOG(NOTICE, "request RXDID[%d] in Queue[%d]",
-				    vc_qp->rxq.rxdid, i);
-		} else {
-			PMD_DRV_LOG(NOTICE, "RXDID[%d] is not supported, "
-				    "request default RXDID[%d] in Queue[%d]",
-				    rxq[i]->rxdid, IAVF_RXDID_LEGACY_1, i);
-			vc_qp->rxq.rxdid = IAVF_RXDID_LEGACY_1;
+		    VIRTCHNL_VF_OFFLOAD_RX_FLEX_DESC) {
+			if (vf->supported_rxdid & BIT(rxq[i]->rxdid)) {
+				vc_qp->rxq.rxdid = rxq[i]->rxdid;
+				PMD_DRV_LOG(NOTICE, "request RXDID[%d] in Queue[%d]",
+					    vc_qp->rxq.rxdid, i);
+			} else {
+				PMD_DRV_LOG(NOTICE, "RXDID[%d] is not supported, "
+					    "request default RXDID[%d] in Queue[%d]",
+					    rxq[i]->rxdid, IAVF_RXDID_LEGACY_1, i);
+				vc_qp->rxq.rxdid = IAVF_RXDID_LEGACY_1;
+			}
+
+			if (vf->vf_res->vf_cap_flags & VIRTCHNL_VF_CAP_PTP &&
+			    vf->ptp_caps & VIRTCHNL_1588_PTP_CAP_RX_TSTAMP)
+				vc_qp->rxq.flags |= VIRTCHNL_PTP_RX_TSTAMP;
 		}
 #else
 		if (vf->vf_res->vf_cap_flags &
@@ -1859,3 +1865,58 @@ iavf_set_vf_quanta_size(struct iavf_adapter *adapter, u16 start_queue_id, u16 nu
 
 	return 0;
 }
+
+int
+iavf_get_ptp_cap(struct iavf_adapter *adapter)
+{
+	struct iavf_info *vf = IAVF_DEV_PRIVATE_TO_VF(adapter);
+	struct virtchnl_ptp_caps ptp_caps;
+	struct iavf_cmd_info args;
+	int err;
+
+	ptp_caps.caps = VIRTCHNL_1588_PTP_CAP_RX_TSTAMP |
+			VIRTCHNL_1588_PTP_CAP_READ_PHC;
+
+	args.ops = VIRTCHNL_OP_1588_PTP_GET_CAPS;
+	args.in_args = (uint8_t *)&ptp_caps;
+	args.in_args_size = sizeof(ptp_caps);
+	args.out_buffer = vf->aq_resp;
+	args.out_size = IAVF_AQ_BUF_SZ;
+
+	err = iavf_execute_vf_cmd(adapter, &args, 0);
+	if (err) {
+		PMD_DRV_LOG(ERR,
+			    "Failed to execute command of OP_1588_PTP_GET_CAPS");
+		return err;
+	}
+
+	vf->ptp_caps = ((struct virtchnl_ptp_caps *)args.out_buffer)->caps;
+
+	return 0;
+}
+
+int
+iavf_get_phc_time(struct iavf_adapter *adapter)
+{
+	struct iavf_info *vf = IAVF_DEV_PRIVATE_TO_VF(adapter);
+	struct virtchnl_phc_time phc_time;
+	struct iavf_cmd_info args;
+	int err;
+
+	args.ops = VIRTCHNL_OP_1588_PTP_GET_TIME;
+	args.in_args = (uint8_t *)&phc_time;
+	args.in_args_size = sizeof(phc_time);
+	args.out_buffer = vf->aq_resp;
+	args.out_size = IAVF_AQ_BUF_SZ;
+
+	err = iavf_execute_vf_cmd(adapter, &args, 0);
+	if (err) {
+		PMD_DRV_LOG(ERR,
+			    "Failed to execute command of VIRTCHNL_OP_1588_PTP_GET_TIME");
+		return err;
+	}
+
+	adapter->phc_time = ((struct virtchnl_phc_time *)args.out_buffer)->time;
+
+	return 0;
+}
-- 
2.9.5