X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=lib%2Flibrte_sched%2Frte_sched.c;h=587d5e602dcd26e9afa85079ca4afaeb8fafbc66;hb=244d5130719ca69038c2331e306c7d0934f93b28;hp=15cd297bbdb7beb019a73ee5062e0566a08699be;hpb=d9d15a2eab4bfcbbace9f23ba7ab72e453a1b21f;p=dpdk.git

diff --git a/lib/librte_sched/rte_sched.c b/lib/librte_sched/rte_sched.c
index 15cd297bbd..587d5e602d 100644
--- a/lib/librte_sched/rte_sched.c
+++ b/lib/librte_sched/rte_sched.c
@@ -1,34 +1,5 @@
-/*-
- *   BSD LICENSE
- *
- *   Copyright(c) 2010-2014 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.
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2010-2014 Intel Corporation
  */
 
 #include <stdio.h>
@@ -42,9 +13,10 @@
 #include <rte_prefetch.h>
 #include <rte_branch_prediction.h>
 #include <rte_mbuf.h>
+#include <rte_bitmap.h>
+#include <rte_reciprocal.h>
 
 #include "rte_sched.h"
-#include "rte_bitmap.h"
 #include "rte_sched_common.h"
 #include "rte_approx.h"
 
@@ -53,7 +25,14 @@
 #endif
 
 #ifdef RTE_SCHED_VECTOR
-#include <immintrin.h>
+#include <rte_vect.h>
+
+#ifdef RTE_ARCH_X86
+#define SCHED_VECTOR_SSE4
+#elif defined(RTE_MACHINE_CPUFLAG_NEON)
+#define SCHED_VECTOR_NEON
+#endif
+
 #endif
 
 #define RTE_SCHED_TB_RATE_CONFIG_ERR          (1e-7)
@@ -62,6 +41,11 @@
 #define RTE_SCHED_PIPE_INVALID                UINT32_MAX
 #define RTE_SCHED_BMP_POS_INVALID             UINT32_MAX
 
+/* Scaling for cycles_per_byte calculation
+ * Chosen so that minimum rate is 480 bit/sec
+ */
+#define RTE_SCHED_TIME_SHIFT		      8
+
 struct rte_sched_subport {
 	/* Token bucket (TB) */
 	uint64_t tb_time; /* time of last update */
@@ -215,7 +199,7 @@ struct rte_sched_port {
 	uint64_t time_cpu_cycles;     /* Current CPU time measured in CPU cyles */
 	uint64_t time_cpu_bytes;      /* Current CPU time measured in bytes */
 	uint64_t time;                /* Current NIC TX time measured in bytes */
-	double cycles_per_byte;       /* CPU cycles per byte */
+	struct rte_reciprocal inv_cycles_per_byte; /* CPU cycles per byte */
 
 	/* Scheduling loop detection */
 	uint32_t pipe_loop;
@@ -273,16 +257,79 @@ rte_sched_port_queues_per_port(struct rte_sched_port *port)
 	return RTE_SCHED_QUEUES_PER_PIPE * port->n_pipes_per_subport * port->n_subports_per_port;
 }
 
+static inline struct rte_mbuf **
+rte_sched_port_qbase(struct rte_sched_port *port, uint32_t qindex)
+{
+	uint32_t pindex = qindex >> 4;
+	uint32_t qpos = qindex & 0xF;
+
+	return (port->queue_array + pindex *
+		port->qsize_sum + port->qsize_add[qpos]);
+}
+
+static inline uint16_t
+rte_sched_port_qsize(struct rte_sched_port *port, uint32_t qindex)
+{
+	uint32_t tc = (qindex >> 2) & 0x3;
+
+	return port->qsize[tc];
+}
+
+static int
+pipe_profile_check(struct rte_sched_pipe_params *params,
+	uint32_t rate)
+{
+	uint32_t i;
+
+	/* Pipe parameters */
+	if (params == NULL)
+		return -10;
+
+	/* TB rate: non-zero, not greater than port rate */
+	if (params->tb_rate == 0 ||
+		params->tb_rate > rate)
+		return -11;
+
+	/* TB size: non-zero */
+	if (params->tb_size == 0)
+		return -12;
+
+	/* TC rate: non-zero, less than pipe rate */
+	for (i = 0; i < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; i++) {
+		if (params->tc_rate[i] == 0 ||
+			params->tc_rate[i] > params->tb_rate)
+			return -13;
+	}
+
+	/* TC period: non-zero */
+	if (params->tc_period == 0)
+		return -14;
+
+#ifdef RTE_SCHED_SUBPORT_TC_OV
+	/* TC3 oversubscription weight: non-zero */
+	if (params->tc_ov_weight == 0)
+		return -15;
+#endif
+
+	/* Queue WRR weights: non-zero */
+	for (i = 0; i < RTE_SCHED_QUEUES_PER_PIPE; i++) {
+		if (params->wrr_weights[i] == 0)
+			return -16;
+	}
+
+	return 0;
+}
+
 static int
 rte_sched_port_check_params(struct rte_sched_port_params *params)
 {
-	uint32_t i, j;
+	uint32_t i;
 
 	if (params == NULL)
 		return -1;
 
 	/* socket */
-	if ((params->socket < 0) || (params->socket >= RTE_MAX_NUMA_NODES))
+	if (params->socket < 0)
 		return -3;
 
 	/* rate */
@@ -322,36 +369,11 @@ rte_sched_port_check_params(struct rte_sched_port_params *params)
 
 	for (i = 0; i < params->n_pipe_profiles; i++) {
 		struct rte_sched_pipe_params *p = params->pipe_profiles + i;
+		int status;
 
-		/* TB rate: non-zero, not greater than port rate */
-		if (p->tb_rate == 0 || p->tb_rate > params->rate)
-			return -10;
-
-		/* TB size: non-zero */
-		if (p->tb_size == 0)
-			return -11;
-
-		/* TC rate: non-zero, less than pipe rate */
-		for (j = 0; j < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; j++) {
-			if (p->tc_rate[j] == 0 || p->tc_rate[j] > p->tb_rate)
-				return -12;
-		}
-
-		/* TC period: non-zero */
-		if (p->tc_period == 0)
-			return -13;
-
-#ifdef RTE_SCHED_SUBPORT_TC_OV
-		/* TC3 oversubscription weight: non-zero */
-		if (p->tc_ov_weight == 0)
-			return -14;
-#endif
-
-		/* Queue WRR weights: non-zero */
-		for (j = 0; j < RTE_SCHED_QUEUES_PER_PIPE; j++) {
-			if (p->wrr_weights[j] == 0)
-				return -15;
-		}
+		status = pipe_profile_check(p, params->rate);
+		if (status != 0)
+			return status;
 	}
 
 	return 0;
@@ -434,7 +456,7 @@ rte_sched_port_get_memory_footprint(struct rte_sched_port_params *params)
 	size0 = sizeof(struct rte_sched_port);
 	size1 = rte_sched_port_get_array_base(params, e_RTE_SCHED_PORT_ARRAY_TOTAL);
 
-	return (size0 + size1);
+	return size0 + size1;
 }
 
 static void
@@ -512,69 +534,80 @@ rte_sched_time_ms_to_bytes(uint32_t time_ms, uint32_t rate)
 }
 
 static void
-rte_sched_port_config_pipe_profile_table(struct rte_sched_port *port, struct rte_sched_port_params *params)
+rte_sched_pipe_profile_convert(struct rte_sched_pipe_params *src,
+	struct rte_sched_pipe_profile *dst,
+	uint32_t rate)
 {
-	uint32_t i, j;
+	uint32_t i;
 
-	for (i = 0; i < port->n_pipe_profiles; i++) {
-		struct rte_sched_pipe_params *src = params->pipe_profiles + i;
-		struct rte_sched_pipe_profile *dst = port->pipe_profiles + i;
+	/* Token Bucket */
+	if (src->tb_rate == rate) {
+		dst->tb_credits_per_period = 1;
+		dst->tb_period = 1;
+	} else {
+		double tb_rate = (double) src->tb_rate
+				/ (double) rate;
+		double d = RTE_SCHED_TB_RATE_CONFIG_ERR;
 
-		/* Token Bucket */
-		if (src->tb_rate == params->rate) {
-			dst->tb_credits_per_period = 1;
-			dst->tb_period = 1;
-		} else {
-			double tb_rate = (double) src->tb_rate
-				/ (double) params->rate;
-			double d = RTE_SCHED_TB_RATE_CONFIG_ERR;
-
-			rte_approx(tb_rate, d,
-				   &dst->tb_credits_per_period, &dst->tb_period);
-		}
-		dst->tb_size = src->tb_size;
+		rte_approx(tb_rate, d,
+			&dst->tb_credits_per_period, &dst->tb_period);
+	}
 
-		/* Traffic Classes */
-		dst->tc_period = rte_sched_time_ms_to_bytes(src->tc_period,
-							    params->rate);
+	dst->tb_size = src->tb_size;
 
-		for (j = 0; j < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; j++)
-			dst->tc_credits_per_period[j]
-				= rte_sched_time_ms_to_bytes(src->tc_period,
-							     src->tc_rate[j]);
+	/* Traffic Classes */
+	dst->tc_period = rte_sched_time_ms_to_bytes(src->tc_period,
+						rate);
+
+	for (i = 0; i < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; i++)
+		dst->tc_credits_per_period[i]
+			= rte_sched_time_ms_to_bytes(src->tc_period,
+				src->tc_rate[i]);
 
 #ifdef RTE_SCHED_SUBPORT_TC_OV
-		dst->tc_ov_weight = src->tc_ov_weight;
+	dst->tc_ov_weight = src->tc_ov_weight;
 #endif
 
-		/* WRR */
-		for (j = 0; j < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; j++) {
-			uint32_t wrr_cost[RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS];
-			uint32_t lcd, lcd1, lcd2;
-			uint32_t qindex;
-
-			qindex = j * RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS;
-
-			wrr_cost[0] = src->wrr_weights[qindex];
-			wrr_cost[1] = src->wrr_weights[qindex + 1];
-			wrr_cost[2] = src->wrr_weights[qindex + 2];
-			wrr_cost[3] = src->wrr_weights[qindex + 3];
-
-			lcd1 = rte_get_lcd(wrr_cost[0], wrr_cost[1]);
-			lcd2 = rte_get_lcd(wrr_cost[2], wrr_cost[3]);
-			lcd = rte_get_lcd(lcd1, lcd2);
-
-			wrr_cost[0] = lcd / wrr_cost[0];
-			wrr_cost[1] = lcd / wrr_cost[1];
-			wrr_cost[2] = lcd / wrr_cost[2];
-			wrr_cost[3] = lcd / wrr_cost[3];
-
-			dst->wrr_cost[qindex] = (uint8_t) wrr_cost[0];
-			dst->wrr_cost[qindex + 1] = (uint8_t) wrr_cost[1];
-			dst->wrr_cost[qindex + 2] = (uint8_t) wrr_cost[2];
-			dst->wrr_cost[qindex + 3] = (uint8_t) wrr_cost[3];
-		}
+	/* WRR */
+	for (i = 0; i < RTE_SCHED_TRAFFIC_CLASSES_PER_PIPE; i++) {
+		uint32_t wrr_cost[RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS];
+		uint32_t lcd, lcd1, lcd2;
+		uint32_t qindex;
+
+		qindex = i * RTE_SCHED_QUEUES_PER_TRAFFIC_CLASS;
+
+		wrr_cost[0] = src->wrr_weights[qindex];
+		wrr_cost[1] = src->wrr_weights[qindex + 1];
+		wrr_cost[2] = src->wrr_weights[qindex + 2];
+		wrr_cost[3] = src->wrr_weights[qindex + 3];
+
+		lcd1 = rte_get_lcd(wrr_cost[0], wrr_cost[1]);
+		lcd2 = rte_get_lcd(wrr_cost[2], wrr_cost[3]);
+		lcd = rte_get_lcd(lcd1, lcd2);
+
+		wrr_cost[0] = lcd / wrr_cost[0];
+		wrr_cost[1] = lcd / wrr_cost[1];
+		wrr_cost[2] = lcd / wrr_cost[2];
+		wrr_cost[3] = lcd / wrr_cost[3];
+
+		dst->wrr_cost[qindex] = (uint8_t) wrr_cost[0];
+		dst->wrr_cost[qindex + 1] = (uint8_t) wrr_cost[1];
+		dst->wrr_cost[qindex + 2] = (uint8_t) wrr_cost[2];
+		dst->wrr_cost[qindex + 3] = (uint8_t) wrr_cost[3];
+	}
+}
+
+static void
+rte_sched_port_config_pipe_profile_table(struct rte_sched_port *port,
+	struct rte_sched_port_params *params)
+{
+	uint32_t i;
+
+	for (i = 0; i < port->n_pipe_profiles; i++) {
+		struct rte_sched_pipe_params *src = params->pipe_profiles + i;
+		struct rte_sched_pipe_profile *dst = port->pipe_profiles + i;
 
+		rte_sched_pipe_profile_convert(src, dst, params->rate);
 		rte_sched_port_log_pipe_profile(port, i);
 	}
 
@@ -592,7 +625,7 @@ struct rte_sched_port *
 rte_sched_port_config(struct rte_sched_port_params *params)
 {
 	struct rte_sched_port *port = NULL;
-	uint32_t mem_size, bmp_mem_size, n_queues_per_port, i;
+	uint32_t mem_size, bmp_mem_size, n_queues_per_port, i, cycles_per_byte;
 
 	/* Check user parameters. Determine the amount of memory to allocate */
 	mem_size = rte_sched_port_get_memory_footprint(params);
@@ -600,7 +633,8 @@ rte_sched_port_config(struct rte_sched_port_params *params)
 		return NULL;
 
 	/* Allocate memory to store the data structures */
-	port = rte_zmalloc("qos_params", mem_size, RTE_CACHE_LINE_SIZE);
+	port = rte_zmalloc_socket("qos_params", mem_size, RTE_CACHE_LINE_SIZE,
+		params->socket);
 	if (port == NULL)
 		return NULL;
 
@@ -643,7 +677,10 @@ rte_sched_port_config(struct rte_sched_port_params *params)
 	port->time_cpu_cycles = rte_get_tsc_cycles();
 	port->time_cpu_bytes = 0;
 	port->time = 0;
-	port->cycles_per_byte = ((double) rte_get_tsc_hz()) / ((double) params->rate);
+
+	cycles_per_byte = (rte_get_tsc_hz() << RTE_SCHED_TIME_SHIFT)
+		/ params->rate;
+	port->inv_cycles_per_byte = rte_reciprocal_value(cycles_per_byte);
 
 	/* Scheduling loop detection */
 	port->pipe_loop = RTE_SCHED_PIPE_INVALID;
@@ -702,10 +739,27 @@ rte_sched_port_config(struct rte_sched_port_params *params)
 void
 rte_sched_port_free(struct rte_sched_port *port)
 {
+	uint32_t qindex;
+	uint32_t n_queues_per_port;
+
 	/* Check user parameters */
 	if (port == NULL)
 		return;
 
+	n_queues_per_port = rte_sched_port_queues_per_port(port);
+
+	/* Free enqueued mbufs */
+	for (qindex = 0; qindex < n_queues_per_port; qindex++) {
+		struct rte_mbuf **mbufs = rte_sched_port_qbase(port, qindex);
+		uint16_t qsize = rte_sched_port_qsize(port, qindex);
+		struct rte_sched_queue *queue = port->queue + qindex;
+		uint16_t qr = queue->qr & (qsize - 1);
+		uint16_t qw = queue->qw & (qsize - 1);
+
+		for (; qr != qw; qr = (qr + 1) & (qsize - 1))
+			rte_pktmbuf_free(mbufs[qr]);
+	}
+
 	rte_bitmap_free(port->bmp);
 	rte_free(port);
 }
@@ -910,6 +964,48 @@ rte_sched_pipe_config(struct rte_sched_port *port,
 	return 0;
 }
 
+int __rte_experimental
+rte_sched_port_pipe_profile_add(struct rte_sched_port *port,
+	struct rte_sched_pipe_params *params,
+	uint32_t *pipe_profile_id)
+{
+	struct rte_sched_pipe_profile *pp;
+	uint32_t i;
+	int status;
+
+	/* Port */
+	if (port == NULL)
+		return -1;
+
+	/* Pipe profiles not exceeds the max limit */
+	if (port->n_pipe_profiles >= RTE_SCHED_PIPE_PROFILES_PER_PORT)
+		return -2;
+
+	/* Pipe params */
+	status = pipe_profile_check(params, port->rate);
+	if (status != 0)
+		return status;
+
+	pp = &port->pipe_profiles[port->n_pipe_profiles];
+	rte_sched_pipe_profile_convert(params, pp, port->rate);
+
+	/* Pipe profile not exists */
+	for (i = 0; i < port->n_pipe_profiles; i++)
+		if (memcmp(port->pipe_profiles + i, pp, sizeof(*pp)) == 0)
+			return -3;
+
+	/* Pipe profile commit */
+	*pipe_profile_id = port->n_pipe_profiles;
+	port->n_pipe_profiles++;
+
+	if (port->pipe_tc3_rate_max < params->tc_rate[3])
+		port->pipe_tc3_rate_max = params->tc_rate[3];
+
+	rte_sched_port_log_pipe_profile(port, *pipe_profile_id);
+
+	return 0;
+}
+
 void
 rte_sched_port_pkt_write(struct rte_mbuf *pkt,
 			 uint32_t subport, uint32_t pipe, uint32_t traffic_class,
@@ -969,7 +1065,7 @@ rte_sched_subport_read_stats(struct rte_sched_port *port,
 	memcpy(stats, &s->stats, sizeof(struct rte_sched_subport_stats));
 	memset(&s->stats, 0, sizeof(struct rte_sched_subport_stats));
 
-	/* Subport TC ovesubscription status */
+	/* Subport TC oversubscription status */
 	*tc_ov = s->tc_ov;
 
 	return 0;
@@ -1016,23 +1112,6 @@ rte_sched_port_qindex(struct rte_sched_port *port, uint32_t subport, uint32_t pi
 	return result;
 }
 
-static inline struct rte_mbuf **
-rte_sched_port_qbase(struct rte_sched_port *port, uint32_t qindex)
-{
-	uint32_t pindex = qindex >> 4;
-	uint32_t qpos = qindex & 0xF;
-
-	return (port->queue_array + pindex * port->qsize_sum + port->qsize_add[qpos]);
-}
-
-static inline uint16_t
-rte_sched_port_qsize(struct rte_sched_port *port, uint32_t qindex)
-{
-	uint32_t tc = (qindex >> 2) & 0x3;
-
-	return port->qsize[tc];
-}
-
 #ifdef RTE_SCHED_DEBUG
 
 static inline int
@@ -1040,17 +1119,7 @@ rte_sched_port_queue_is_empty(struct rte_sched_port *port, uint32_t qindex)
 {
 	struct rte_sched_queue *queue = port->queue + qindex;
 
-	return (queue->qr == queue->qw);
-}
-
-static inline int
-rte_sched_port_queue_is_full(struct rte_sched_port *port, uint32_t qindex)
-{
-	struct rte_sched_queue *queue = port->queue + qindex;
-	uint16_t qsize = rte_sched_port_qsize(port, qindex);
-	uint16_t qlen = queue->qw - queue->qr;
-
-	return (qlen >= qsize);
+	return queue->qr == queue->qw;
 }
 
 #endif /* RTE_SCHED_DEBUG */
@@ -1068,8 +1137,17 @@ rte_sched_port_update_subport_stats(struct rte_sched_port *port, uint32_t qindex
 	s->stats.n_bytes_tc[tc_index] += pkt_len;
 }
 
+#ifdef RTE_SCHED_RED
+static inline void
+rte_sched_port_update_subport_stats_on_drop(struct rte_sched_port *port,
+						uint32_t qindex,
+						struct rte_mbuf *pkt, uint32_t red)
+#else
 static inline void
-rte_sched_port_update_subport_stats_on_drop(struct rte_sched_port *port, uint32_t qindex, struct rte_mbuf *pkt)
+rte_sched_port_update_subport_stats_on_drop(struct rte_sched_port *port,
+						uint32_t qindex,
+						struct rte_mbuf *pkt, __rte_unused uint32_t red)
+#endif
 {
 	struct rte_sched_subport *s = port->subport + (qindex / rte_sched_port_queues_per_subport(port));
 	uint32_t tc_index = (qindex >> 2) & 0x3;
@@ -1077,6 +1155,9 @@ rte_sched_port_update_subport_stats_on_drop(struct rte_sched_port *port, uint32_
 
 	s->stats.n_pkts_tc_dropped[tc_index] += 1;
 	s->stats.n_bytes_tc_dropped[tc_index] += pkt_len;
+#ifdef RTE_SCHED_RED
+	s->stats.n_pkts_red_dropped[tc_index] += red;
+#endif
 }
 
 static inline void
@@ -1089,14 +1170,26 @@ rte_sched_port_update_queue_stats(struct rte_sched_port *port, uint32_t qindex,
 	qe->stats.n_bytes += pkt_len;
 }
 
+#ifdef RTE_SCHED_RED
 static inline void
-rte_sched_port_update_queue_stats_on_drop(struct rte_sched_port *port, uint32_t qindex, struct rte_mbuf *pkt)
+rte_sched_port_update_queue_stats_on_drop(struct rte_sched_port *port,
+						uint32_t qindex,
+						struct rte_mbuf *pkt, uint32_t red)
+#else
+static inline void
+rte_sched_port_update_queue_stats_on_drop(struct rte_sched_port *port,
+						uint32_t qindex,
+						struct rte_mbuf *pkt, __rte_unused uint32_t red)
+#endif
 {
 	struct rte_sched_queue_extra *qe = port->queue_extra + qindex;
 	uint32_t pkt_len = pkt->pkt_len;
 
 	qe->stats.n_pkts_dropped += 1;
 	qe->stats.n_bytes_dropped += pkt_len;
+#ifdef RTE_SCHED_RED
+	qe->stats.n_pkts_red_dropped += red;
+#endif
 }
 
 #endif /* RTE_SCHED_COLLECT_STATS */
@@ -1144,27 +1237,6 @@ rte_sched_port_set_queue_empty_timestamp(struct rte_sched_port *port, uint32_t q
 
 #ifdef RTE_SCHED_DEBUG
 
-static inline int
-debug_pipe_is_empty(struct rte_sched_port *port, uint32_t pindex)
-{
-	uint32_t qindex, i;
-
-	qindex = pindex << 4;
-
-	for (i = 0; i < 16; i++) {
-		uint32_t queue_empty = rte_sched_port_queue_is_empty(port, qindex + i);
-		uint32_t bmp_bit_clear = (rte_bitmap_get(port->bmp, qindex + i) == 0);
-
-		if (queue_empty != bmp_bit_clear)
-			rte_panic("Queue status mismatch for queue %u of pipe %u\n", i, pindex);
-
-		if (!queue_empty)
-			return 0;
-	}
-
-	return 1;
-}
-
 static inline void
 debug_check_queue_slab(struct rte_sched_port *port, uint32_t bmp_pos,
 		       uint64_t bmp_slab)
@@ -1248,8 +1320,10 @@ rte_sched_port_enqueue_qwa(struct rte_sched_port *port, uint32_t qindex,
 		     (qlen >= qsize))) {
 		rte_pktmbuf_free(pkt);
 #ifdef RTE_SCHED_COLLECT_STATS
-		rte_sched_port_update_subport_stats_on_drop(port, qindex, pkt);
-		rte_sched_port_update_queue_stats_on_drop(port, qindex, pkt);
+		rte_sched_port_update_subport_stats_on_drop(port, qindex, pkt,
+							    qlen < qsize);
+		rte_sched_port_update_queue_stats_on_drop(port, qindex, pkt,
+							  qlen < qsize);
 #endif
 		return 0;
 	}
@@ -1686,7 +1760,7 @@ grinder_schedule(struct rte_sched_port *port, uint32_t pos)
 	return 1;
 }
 
-#ifdef RTE_SCHED_VECTOR
+#ifdef SCHED_VECTOR_SSE4
 
 static inline int
 grinder_pipe_exists(struct rte_sched_port *port, uint32_t base_pipe)
@@ -1705,6 +1779,26 @@ grinder_pipe_exists(struct rte_sched_port *port, uint32_t base_pipe)
 	return 1;
 }
 
+#elif defined(SCHED_VECTOR_NEON)
+
+static inline int
+grinder_pipe_exists(struct rte_sched_port *port, uint32_t base_pipe)
+{
+	uint32x4_t index, pipes;
+	uint32_t *pos = (uint32_t *)port->grinder_base_bmp_pos;
+
+	index = vmovq_n_u32(base_pipe);
+	pipes = vld1q_u32(pos);
+	if (!vminvq_u32(veorq_u32(pipes, index)))
+		return 1;
+
+	pipes = vld1q_u32(pos + 4);
+	if (!vminvq_u32(veorq_u32(pipes, index)))
+		return 1;
+
+	return 0;
+}
+
 #else
 
 static inline int
@@ -2095,11 +2189,15 @@ rte_sched_port_time_resync(struct rte_sched_port *port)
 {
 	uint64_t cycles = rte_get_tsc_cycles();
 	uint64_t cycles_diff = cycles - port->time_cpu_cycles;
-	double bytes_diff = ((double) cycles_diff) / port->cycles_per_byte;
+	uint64_t bytes_diff;
+
+	/* Compute elapsed time in bytes */
+	bytes_diff = rte_reciprocal_divide(cycles_diff << RTE_SCHED_TIME_SHIFT,
+					   port->inv_cycles_per_byte);
 
 	/* Advance port time */
 	port->time_cpu_cycles = cycles;
-	port->time_cpu_bytes += (uint64_t) bytes_diff;
+	port->time_cpu_bytes += bytes_diff;
 	if (port->time < port->time_cpu_bytes)
 		port->time = port->time_cpu_bytes;