Changes on top of http://dpdk.org/dev/patchwork/patch/24184/

0001-Custom-changes-on-top-of-l3fwd-patch-v2.patch

From 1735d691e5b5d7fb3f0d2bcd6d1fbb9f7712d7af Mon Sep 17 00:00:00 2001
From: Ashwin Sekhar T K <ashwin.sekhar@caviumnetworks.com>
Date: Wed, 10 May 2017 14:24:03 +0000
Subject: [PATCH] Custom changes on top of l3fwd patch v2

---
 examples/l3fwd/l3fwd_em_hlm.h        |  2 -
 examples/l3fwd/l3fwd_em_sequential.h |  2 -
 examples/l3fwd/l3fwd_lpm_neon.h      | 87 +++++++++++++++++++++++-------------
 examples/l3fwd/l3fwd_neon.h          |  2 +-
 4 files changed, 57 insertions(+), 36 deletions(-)

diff --git a/examples/l3fwd/l3fwd_em_hlm.h b/examples/l3fwd/l3fwd_em_hlm.h
index 10a9c95..098b396 100644
--- a/examples/l3fwd/l3fwd_em_hlm.h
+++ b/examples/l3fwd/l3fwd_em_hlm.h
@@ -170,7 +170,6 @@ l3fwd_em_send_packets(int nb_rx, struct rte_mbuf **pkts_burst,
 	int32_t n = RTE_ALIGN_FLOOR(nb_rx, EM_HASH_LOOKUP_COUNT);
 
 	for (j = 0; j < EM_HASH_LOOKUP_COUNT && j < nb_rx; j++) {
-		rte_prefetch0(pkts_burst[j]);
 		rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[j],
 					       struct ether_hdr *) + 1);
 	}
@@ -189,7 +188,6 @@ l3fwd_em_send_packets(int nb_rx, struct rte_mbuf **pkts_burst,
 
 		for (i = 0, pos = j + EM_HASH_LOOKUP_COUNT;
 		     i < EM_HASH_LOOKUP_COUNT && pos < nb_rx; i++, pos++) {
-			rte_prefetch0(pkts_burst[pos]);
 			rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[pos],
 						       struct ether_hdr *) + 1);
 		}
diff --git a/examples/l3fwd/l3fwd_em_sequential.h b/examples/l3fwd/l3fwd_em_sequential.h
index 63c5c12..c7d477d 100644
--- a/examples/l3fwd/l3fwd_em_sequential.h
+++ b/examples/l3fwd/l3fwd_em_sequential.h
@@ -109,14 +109,12 @@ l3fwd_em_send_packets(int nb_rx, struct rte_mbuf **pkts_burst,
 	uint16_t dst_port[MAX_PKT_BURST];
 
 	if (nb_rx > 0) {
-		rte_prefetch0(pkts_burst[0]);
 		rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[0],
 					       struct ether_hdr *) + 1);
 	}
 
 	for (i = 1, j = 0; j < nb_rx; i++, j++) {
 		if (i < nb_rx) {
-			rte_prefetch0(pkts_burst[i]);
 			rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[i],
 						       struct ether_hdr *) + 1);
 		}
diff --git a/examples/l3fwd/l3fwd_lpm_neon.h b/examples/l3fwd/l3fwd_lpm_neon.h
index 2f047b3..375101b 100644
--- a/examples/l3fwd/l3fwd_lpm_neon.h
+++ b/examples/l3fwd/l3fwd_lpm_neon.h
@@ -117,46 +117,71 @@ static inline void
 l3fwd_lpm_send_packets(int nb_rx, struct rte_mbuf **pkts_burst,
 			uint8_t portid, struct lcore_conf *qconf)
 {
-	int32_t i, j, pos;
+	int32_t j = 0, i = 0;
 	uint16_t dst_port[MAX_PKT_BURST];
-	int32x4_t dip[MAX_PKT_BURST / FWDSTEP];
-	uint32_t ipv4_flag[MAX_PKT_BURST / FWDSTEP];
+	int32x4_t dip;
+	uint32_t ipv4_flag;
 	const int32_t k = RTE_ALIGN_FLOOR(nb_rx, FWDSTEP);
+	const int32_t m = nb_rx % FWDSTEP;
 
-	for (j = 0; j < FWDSTEP && j < nb_rx; j++) {
-		rte_prefetch0(pkts_burst[j]);
-		rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[j],
-					       struct ether_hdr *) + 1);
-	}
+	if (k) {
+		for (i = 0; i < FWDSTEP; i++) {
+			rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[i],
+					       struct ether_hdr *));
+		}
 
-	for (j = 0; j != k; j += FWDSTEP) {
-		for (i = 0, pos = j + FWDSTEP; i < FWDSTEP && pos < nb_rx;
-		     i++, pos++) {
-			rte_prefetch0(pkts_burst[pos]);
-			rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[pos],
-						       struct ether_hdr *) + 1);
+		for (j = 0; j != k - FWDSTEP; j += FWDSTEP) {
+			for (i = 0; i < FWDSTEP; i++) {
+				rte_prefetch0(rte_pktmbuf_mtod(
+						pkts_burst[j + i + FWDSTEP],
+					       struct ether_hdr *));
+			}
+			processx4_step1(&pkts_burst[j], &dip, &ipv4_flag);
+			processx4_step2(qconf, dip,
+				ipv4_flag, portid, &pkts_burst[j],
+				&dst_port[j]);
 		}
-		processx4_step1(&pkts_burst[j], &dip[j / FWDSTEP],
-				&ipv4_flag[j / FWDSTEP]);
 
-		processx4_step2(qconf, dip[j / FWDSTEP],
-				ipv4_flag[j / FWDSTEP], portid, &pkts_burst[j],
+		processx4_step1(&pkts_burst[j], &dip, &ipv4_flag);
+		processx4_step2(qconf, dip, ipv4_flag, portid, &pkts_burst[j],
 				&dst_port[j]);
+		j += FWDSTEP;
 	}
 
-	/* Classify last up to 3 packets one by one */
-	switch (nb_rx % FWDSTEP) {
-	case 3:
-		dst_port[j] = lpm_get_dst_port(qconf, pkts_burst[j], portid);
-		j++;
-		/* fallthrough */
-	case 2:
-		dst_port[j] = lpm_get_dst_port(qconf, pkts_burst[j], portid);
-		j++;
-		/* fallthrough */
-	case 1:
-		dst_port[j] = lpm_get_dst_port(qconf, pkts_burst[j], portid);
-		j++;
+	if (m) {
+		/* Prefetch last up to 3 packets one by one */
+		switch (m) {
+		case 3:
+			rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[j], void *));
+			j++;
+			/* fallthrough */
+		case 2:
+			rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[j], void *));
+			j++;
+			/* fallthrough */
+		case 1:
+			rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[j], void *));
+			j++;
+		}
+
+		j -= m;
+		/* Classify last up to 3 packets one by one */
+		switch (m) {
+		case 3:
+			dst_port[j] = lpm_get_dst_port(qconf, pkts_burst[j],
+					portid);
+			j++;
+			/* fallthrough */
+		case 2:
+			dst_port[j] = lpm_get_dst_port(qconf, pkts_burst[j],
+					portid);
+			j++;
+			/* fallthrough */
+		case 1:
+			dst_port[j] = lpm_get_dst_port(qconf, pkts_burst[j],
+					portid);
+			j++;
+		}
 	}
 
 	send_packets_multi(qconf, pkts_burst, dst_port, nb_rx);
diff --git a/examples/l3fwd/l3fwd_neon.h b/examples/l3fwd/l3fwd_neon.h
index 75c8976..fe351db 100644
--- a/examples/l3fwd/l3fwd_neon.h
+++ b/examples/l3fwd/l3fwd_neon.h
@@ -145,7 +145,7 @@ process_packet(struct rte_mbuf *pkt, uint16_t *dst_port)
 	rfc1812_process((struct ipv4_hdr *)(eth_hdr + 1), dst_port,
 			pkt->packet_type);
 
-	ve = vsetq_lane_u32(vgetq_lane_u32(te, 3), ve, 3);
+	ve = vcopyq_laneq_u32(ve, 3, te, 3);
 	vst1q_u32((uint32_t *)eth_hdr, ve);
 }
 
-- 
2.1.4