[dpdk-dev] [PATCH v2 52/57] net/octeontx2: add Tx multi segment version

[<jerinj@marvell.com>]

From: Nithin Dabilpuram <ndabilpuram@marvell.com>

Add multi segment version of packet Transmit function.

Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
---
 drivers/net/octeontx2/otx2_ethdev.h |  4 ++
 drivers/net/octeontx2/otx2_tx.c     | 58 +++++++++++++++++++++
 drivers/net/octeontx2/otx2_tx.h     | 81 +++++++++++++++++++++++++++++
 3 files changed, 143 insertions(+)

diff --git a/drivers/net/octeontx2/otx2_ethdev.h b/drivers/net/octeontx2/otx2_ethdev.h
index 1f9323fe3..f39fdfa1f 100644
--- a/drivers/net/octeontx2/otx2_ethdev.h
+++ b/drivers/net/octeontx2/otx2_ethdev.h
@@ -89,6 +89,10 @@
 #define NIX_TX_NB_SEG_MAX		9
 #endif
 
+#define NIX_TX_MSEG_SG_DWORDS				\
+	((RTE_ALIGN_MUL_CEIL(NIX_TX_NB_SEG_MAX, 3) / 3)	\
+	 + NIX_TX_NB_SEG_MAX)
+
 /* Apply BP when CQ is 75% full */
 #define NIX_CQ_BP_LEVEL (25 * 256 / 100)
 
diff --git a/drivers/net/octeontx2/otx2_tx.c b/drivers/net/octeontx2/otx2_tx.c
index 16d69b74f..0ac5ea652 100644
--- a/drivers/net/octeontx2/otx2_tx.c
+++ b/drivers/net/octeontx2/otx2_tx.c
@@ -49,6 +49,37 @@ nix_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts,
 	return pkts;
 }
 
+static __rte_always_inline uint16_t
+nix_xmit_pkts_mseg(void *tx_queue, struct rte_mbuf **tx_pkts,
+		   uint16_t pkts, uint64_t *cmd, const uint16_t flags)
+{
+	struct otx2_eth_txq *txq = tx_queue; uint64_t i;
+	const rte_iova_t io_addr = txq->io_addr;
+	void *lmt_addr = txq->lmt_addr;
+	uint16_t segdw;
+
+	NIX_XMIT_FC_OR_RETURN(txq, pkts);
+
+	otx2_lmt_mov(cmd, &txq->cmd[0], otx2_nix_tx_ext_subs(flags));
+
+	/* Lets commit any changes in the packet */
+	rte_cio_wmb();
+
+	for (i = 0; i < pkts; i++) {
+		otx2_nix_xmit_prepare(tx_pkts[i], cmd, flags);
+		segdw = otx2_nix_prepare_mseg(tx_pkts[i], cmd, flags);
+		otx2_nix_xmit_prepare_tstamp(cmd, &txq->cmd[0],
+					     tx_pkts[i]->ol_flags, segdw,
+					     flags);
+		otx2_nix_xmit_mseg_one(cmd, lmt_addr, io_addr, segdw);
+	}
+
+	/* Reduce the cached count */
+	txq->fc_cache_pkts -= pkts;
+
+	return pkts;
+}
+
 #define T(name, f4, f3, f2, f1, f0, sz, flags)				\
 static uint16_t __rte_noinline	__hot					\
 otx2_nix_xmit_pkts_ ## name(void *tx_queue,				\
@@ -62,6 +93,20 @@ otx2_nix_xmit_pkts_ ## name(void *tx_queue,				\
 NIX_TX_FASTPATH_MODES
 #undef T
 
+#define T(name, f4, f3, f2, f1, f0, sz, flags)				\
+static uint16_t __rte_noinline	__hot					\
+otx2_nix_xmit_pkts_mseg_ ## name(void *tx_queue,			\
+			struct rte_mbuf **tx_pkts, uint16_t pkts)	\
+{									\
+	uint64_t cmd[(sz) + NIX_TX_MSEG_SG_DWORDS - 2];			\
+									\
+	return nix_xmit_pkts_mseg(tx_queue, tx_pkts, pkts, cmd,		\
+				  (flags) | NIX_TX_MULTI_SEG_F);	\
+}
+
+NIX_TX_FASTPATH_MODES
+#undef T
+
 static inline void
 pick_tx_func(struct rte_eth_dev *eth_dev,
 	     const eth_tx_burst_t tx_burst[2][2][2][2][2])
@@ -80,15 +125,28 @@ pick_tx_func(struct rte_eth_dev *eth_dev,
 void
 otx2_eth_set_tx_function(struct rte_eth_dev *eth_dev)
 {
+	struct otx2_eth_dev *dev = otx2_eth_pmd_priv(eth_dev);
+
 	const eth_tx_burst_t nix_eth_tx_burst[2][2][2][2][2] = {
 #define T(name, f4, f3, f2, f1, f0, sz, flags)				\
 	[f4][f3][f2][f1][f0] =  otx2_nix_xmit_pkts_ ## name,
 
+NIX_TX_FASTPATH_MODES
+#undef T
+	};
+
+	const eth_tx_burst_t nix_eth_tx_burst_mseg[2][2][2][2][2] = {
+#define T(name, f4, f3, f2, f1, f0, sz, flags)				\
+	[f4][f3][f2][f1][f0] =  otx2_nix_xmit_pkts_mseg_ ## name,
+
 NIX_TX_FASTPATH_MODES
 #undef T
 	};
 
 	pick_tx_func(eth_dev, nix_eth_tx_burst);
 
+	if (dev->tx_offloads & DEV_TX_OFFLOAD_MULTI_SEGS)
+		pick_tx_func(eth_dev, nix_eth_tx_burst_mseg);
+
 	rte_mb();
 }
diff --git a/drivers/net/octeontx2/otx2_tx.h b/drivers/net/octeontx2/otx2_tx.h
index db4c1f70f..b75a220ea 100644
--- a/drivers/net/octeontx2/otx2_tx.h
+++ b/drivers/net/octeontx2/otx2_tx.h
@@ -212,6 +212,87 @@ otx2_nix_xmit_one(uint64_t *cmd, void *lmt_addr,
 	} while (lmt_status == 0);
 }
 
+static __rte_always_inline uint16_t
+otx2_nix_prepare_mseg(struct rte_mbuf *m, uint64_t *cmd, const uint16_t flags)
+{
+	struct nix_send_hdr_s *send_hdr;
+	union nix_send_sg_s *sg;
+	struct rte_mbuf *m_next;
+	uint64_t *slist, sg_u;
+	uint64_t nb_segs;
+	uint64_t segdw;
+	uint8_t off, i;
+
+	send_hdr = (struct nix_send_hdr_s *)cmd;
+	send_hdr->w0.total = m->pkt_len;
+	send_hdr->w0.aura = npa_lf_aura_handle_to_aura(m->pool->pool_id);
+
+	if (flags & NIX_TX_NEED_EXT_HDR)
+		off = 2;
+	else
+		off = 0;
+
+	sg = (union nix_send_sg_s *)&cmd[2 + off];
+	sg_u = sg->u;
+	slist = &cmd[3 + off];
+
+	i = 0;
+	nb_segs = m->nb_segs;
+
+	/* Fill mbuf segments */
+	do {
+		m_next = m->next;
+		sg_u = sg_u | ((uint64_t)m->data_len << (i << 4));
+		*slist = rte_mbuf_data_iova(m);
+		/* Set invert df if reference count > 1 */
+		if (flags & NIX_TX_OFFLOAD_MBUF_NOFF_F)
+			sg_u |=
+			((uint64_t)(rte_pktmbuf_prefree_seg(m) == NULL) <<
+			 (i + 55));
+		/* Mark mempool object as "put" since it is freed by NIX */
+		if (!(sg_u & (1ULL << (i + 55)))) {
+			m->next = NULL;
+			__mempool_check_cookies(m->pool, (void **)&m, 1, 0);
+		}
+		slist++;
+		i++;
+		nb_segs--;
+		if (i > 2 && nb_segs) {
+			i = 0;
+			/* Next SG subdesc */
+			*(uint64_t *)slist = sg_u & 0xFC00000000000000;
+			sg->u = sg_u;
+			sg->segs = 3;
+			sg = (union nix_send_sg_s *)slist;
+			sg_u = sg->u;
+			slist++;
+		}
+		m = m_next;
+	} while (nb_segs);
+
+	sg->u = sg_u;
+	sg->segs = i;
+	segdw = (uint64_t *)slist - (uint64_t *)&cmd[2 + off];
+	/* Roundup extra dwords to multiple of 2 */
+	segdw = (segdw >> 1) + (segdw & 0x1);
+	/* Default dwords */
+	segdw += (off >> 1) + 1 + !!(flags & NIX_TX_OFFLOAD_TSTAMP_F);
+	send_hdr->w0.sizem1 = segdw - 1;
+
+	return segdw;
+}
+
+static __rte_always_inline void
+otx2_nix_xmit_mseg_one(uint64_t *cmd, void *lmt_addr,
+		       rte_iova_t io_addr, uint16_t segdw)
+{
+	uint64_t lmt_status;
+
+	do {
+		otx2_lmt_mov_seg(lmt_addr, (const void *)cmd, segdw);
+		lmt_status = otx2_lmt_submit(io_addr);
+	} while (lmt_status == 0);
+}
 
 #define L3L4CSUM_F   NIX_TX_OFFLOAD_L3_L4_CSUM_F
 #define OL3OL4CSUM_F NIX_TX_OFFLOAD_OL3_OL4_CSUM_F
-- 
2.21.0