[dpdk-dev] [PATCH] crypto/scheduler: add packet size based scheduling mode

[Fan Zhang]

This patch adds packet size based distribution scheduling mode
to cryptodev scheduler PMD. In this mode, there are 2 cryptodev
slaves (slave 0 and slave 1), for all enqueued crypto ops
with packet payload over 127 bytes will be distributed to slave 0,
the rest will goes to slave 1.

Signed-off-by: Fan Zhang <roy.fan.zhang at intel.com>
---
 drivers/crypto/scheduler/Makefile                  |   1 +
 drivers/crypto/scheduler/rte_cryptodev_scheduler.c |   7 +
 drivers/crypto/scheduler/rte_cryptodev_scheduler.h |   2 +
 drivers/crypto/scheduler/scheduler_packet_size.c   | 487 +++++++++++++++++++++
 4 files changed, 497 insertions(+)
 create mode 100644 drivers/crypto/scheduler/scheduler_packet_size.c

diff --git a/drivers/crypto/scheduler/Makefile b/drivers/crypto/scheduler/Makefile
index 0cce6f2..7b4afb4 100644
--- a/drivers/crypto/scheduler/Makefile
+++ b/drivers/crypto/scheduler/Makefile
@@ -54,6 +54,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_PMD_CRYPTO_SCHEDULER) += scheduler_pmd.c
 SRCS-$(CONFIG_RTE_LIBRTE_PMD_CRYPTO_SCHEDULER) += scheduler_pmd_ops.c
 SRCS-$(CONFIG_RTE_LIBRTE_PMD_CRYPTO_SCHEDULER) += rte_cryptodev_scheduler.c
 SRCS-$(CONFIG_RTE_LIBRTE_PMD_CRYPTO_SCHEDULER) += scheduler_roundrobin.c
+SRCS-$(CONFIG_RTE_LIBRTE_PMD_CRYPTO_SCHEDULER) += scheduler_packet_size.c
 
 # library dependencies
 DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_CRYPTO_SCHEDULER) += lib/librte_cryptodev
diff --git a/drivers/crypto/scheduler/rte_cryptodev_scheduler.c b/drivers/crypto/scheduler/rte_cryptodev_scheduler.c
index 11e8143..6eea05a 100644
--- a/drivers/crypto/scheduler/rte_cryptodev_scheduler.c
+++ b/drivers/crypto/scheduler/rte_cryptodev_scheduler.c
@@ -336,6 +336,13 @@ rte_crpytodev_scheduler_mode_set(uint8_t scheduler_id,
 			return -1;
 		}
 		break;
+	case CDEV_SCHED_MODE_PKT_SIZE_DISTRBUTE:
+		if (rte_cryptodev_scheduler_load_user_scheduler(scheduler_id,
+				packet_size_based_scheduler) < 0) {
+			CS_LOG_ERR("Failed to load scheduler");
+			return -1;
+		}
+		break;
 	default:
 		CS_LOG_ERR("Not yet supported");
 		return -ENOTSUP;
diff --git a/drivers/crypto/scheduler/rte_cryptodev_scheduler.h b/drivers/crypto/scheduler/rte_cryptodev_scheduler.h
index 7ef44e7..471e389 100644
--- a/drivers/crypto/scheduler/rte_cryptodev_scheduler.h
+++ b/drivers/crypto/scheduler/rte_cryptodev_scheduler.h
@@ -47,6 +47,7 @@ enum rte_cryptodev_scheduler_mode {
 	CDEV_SCHED_MODE_NOT_SET = 0,
 	CDEV_SCHED_MODE_USERDEFINED,
 	CDEV_SCHED_MODE_ROUNDROBIN,
+	CDEV_SCHED_MODE_PKT_SIZE_DISTRBUTE,
 
 	CDEV_SCHED_MODE_COUNT /* number of modes */
 };
@@ -158,6 +159,7 @@ struct rte_cryptodev_scheduler {
 };
 
 extern struct rte_cryptodev_scheduler *roundrobin_scheduler;
+extern struct rte_cryptodev_scheduler *packet_size_based_scheduler;
 
 #ifdef __cplusplus
 }
diff --git a/drivers/crypto/scheduler/scheduler_packet_size.c b/drivers/crypto/scheduler/scheduler_packet_size.c
new file mode 100644
index 0000000..0185557
--- /dev/null
+++ b/drivers/crypto/scheduler/scheduler_packet_size.c
@@ -0,0 +1,487 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright(c) 2017 Intel Corporation. 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.
+ */
+
+#include <rte_cryptodev.h>
+#include <rte_malloc.h>
+
+#include "rte_cryptodev_scheduler_operations.h"
+#include "scheduler_pmd_private.h"
+
+#define NB_PKT_SIZE_SLAVES			(2)
+#define PKT_SIZE_THRESHOLD			(0xff80)
+
+struct psd_scheduler_qp_ctx {
+	struct scheduler_slave slaves[NB_PKT_SIZE_SLAVES];
+	uint8_t last_deq_index;
+	uint16_t threshold;
+} __rte_cache_aligned;
+
+static uint16_t
+schedule_enqueue(void *qp, struct rte_crypto_op **ops, uint16_t nb_ops)
+{
+	struct psd_scheduler_qp_ctx *qp_ctx =
+			((struct scheduler_qp_ctx *)qp)->private_qp_ctx;
+	struct rte_crypto_op *sched_ops[NB_PKT_SIZE_SLAVES][nb_ops];
+	struct scheduler_slave *slave;
+	uint16_t nb_sched_ops[NB_PKT_SIZE_SLAVES] = {0};
+	uint16_t i, j, processed_ops, total_ops = 0;
+	struct rte_cryptodev_sym_session *sessions[NB_PKT_SIZE_SLAVES][nb_ops];
+	struct scheduler_session *sess0, *sess1, *sess2, *sess3;
+	struct rte_mbuf *mbuf0, *mbuf1, *mbuf2, *mbuf3;
+	uint8_t index0, index1, index2, index3;
+
+	if (unlikely(nb_ops == 0))
+		return 0;
+
+	for (i = 0; i < nb_ops && i < 4; i++) {
+		rte_prefetch0(ops[i]->sym->session);
+		rte_prefetch0(ops[i]->sym->m_src);
+	}
+
+	for (i = 0; (i < (nb_ops - 8)) && (nb_ops > 8); i += 4) {
+		rte_prefetch0(ops[i + 4]->sym->session);
+		rte_prefetch0(ops[i + 5]->sym->session);
+		rte_prefetch0(ops[i + 6]->sym->session);
+		rte_prefetch0(ops[i + 7]->sym->session);
+		rte_prefetch0(ops[i + 4]->sym->m_src);
+		rte_prefetch0(ops[i + 5]->sym->m_src);
+		rte_prefetch0(ops[i + 6]->sym->m_src);
+		rte_prefetch0(ops[i + 7]->sym->m_src);
+
+		sess0 = (struct scheduler_session *)
+				ops[i]->sym->session->_private;
+		sess1 = (struct scheduler_session *)
+				ops[i+1]->sym->session->_private;
+		sess2 = (struct scheduler_session *)
+				ops[i+2]->sym->session->_private;
+		sess3 = (struct scheduler_session *)
+				ops[i+3]->sym->session->_private;
+
+		mbuf0 = ops[i]->sym->m_src;
+		mbuf1 = ops[i + 1]->sym->m_src;
+		mbuf2 = ops[i + 2]->sym->m_src;
+		mbuf3 = ops[i + 3]->sym->m_src;
+
+		index0 = mbuf0->data_len & qp_ctx->threshold ? 0 : 1;
+		index1 = mbuf1->data_len & qp_ctx->threshold ? 0 : 1;
+		index2 = mbuf2->data_len & qp_ctx->threshold ? 0 : 1;
+		index3 = mbuf3->data_len & qp_ctx->threshold ? 0 : 1;
+
+		sched_ops[index0][nb_sched_ops[index0]] = ops[i];
+		sessions[index0][nb_sched_ops[index0]++] =
+				ops[i]->sym->session;
+		sched_ops[index1][nb_sched_ops[index1]] = ops[i + 1];
+		sessions[index1][nb_sched_ops[index1]++] =
+				ops[i + 1]->sym->session;
+		sched_ops[index2][nb_sched_ops[index2]] = ops[i + 2];
+		sessions[index2][nb_sched_ops[index2]++] =
+				ops[i + 2]->sym->session;
+		sched_ops[index3][nb_sched_ops[index3]] = ops[i + 3];
+		sessions[index3][nb_sched_ops[index3]++] =
+				ops[i + 3]->sym->session;
+
+		ops[i]->sym->session = sess0->sessions[index0];
+		ops[i + 1]->sym->session = sess1->sessions[index1];
+		ops[i + 2]->sym->session = sess2->sessions[index2];
+		ops[i + 3]->sym->session = sess3->sessions[index3];
+	}
+
+	for (; i < nb_ops; i++) {
+		sess0 = (struct scheduler_session *)
+				ops[i]->sym->session->_private;
+		mbuf0 = ops[i]->sym->m_src;
+		index0 = mbuf0->data_len & qp_ctx->threshold ? 0 : 1;
+		sched_ops[index0][nb_sched_ops[index0]] = ops[i];
+		sessions[index0][nb_sched_ops[index0]++] =
+				ops[i]->sym->session;
+		ops[i]->sym->session = sess0->sessions[index0];
+	}
+
+	for (i = 0; i < NB_PKT_SIZE_SLAVES; i++) {
+		slave = &qp_ctx->slaves[i];
+		processed_ops = rte_cryptodev_enqueue_burst(slave->dev_id,
+				slave->qp_id, sched_ops[i], nb_sched_ops[i]);
+		slave->nb_inflight_cops += processed_ops;
+
+		if (unlikely(processed_ops < nb_sched_ops[i]))
+			for (j = processed_ops; j < nb_sched_ops[i]; j++)
+				sched_ops[i][j]->sym->session = sessions[i][j];
+		total_ops += processed_ops;
+	}
+
+	return total_ops;
+}
+
+static uint16_t
+schedule_enqueue_ordering(void *qp, struct rte_crypto_op **ops,
+		uint16_t nb_ops)
+{
+	struct scheduler_qp_ctx *qp_ctx = qp;
+	struct psd_scheduler_qp_ctx *ps_qp_ctx = qp_ctx->private_qp_ctx;
+	struct rte_crypto_op *sched_ops[NB_PKT_SIZE_SLAVES][nb_ops];
+	struct scheduler_slave *slave;
+	uint16_t nb_sched_ops[NB_PKT_SIZE_SLAVES] = {0};
+	uint16_t i, j, processed_ops, total_ops = 0;
+	struct rte_cryptodev_sym_session *sessions[NB_PKT_SIZE_SLAVES][nb_ops];
+	struct scheduler_session *sess0, *sess1, *sess2, *sess3;
+	struct rte_mbuf *mbuf0, *mbuf1, *mbuf2, *mbuf3;
+	uint8_t index0, index1, index2, index3;
+
+	if (unlikely(nb_ops == 0))
+		return 0;
+
+	for (i = 0; i < nb_ops && i < 4; i++) {
+		rte_prefetch0(ops[i]->sym->session);
+		rte_prefetch0(ops[i]->sym->m_src);
+	}
+
+	for (i = 0; (i < (nb_ops - 8)) && (nb_ops > 8); i += 4) {
+		rte_prefetch0(ops[i + 4]->sym->session);
+		rte_prefetch0(ops[i + 5]->sym->session);
+		rte_prefetch0(ops[i + 6]->sym->session);
+		rte_prefetch0(ops[i + 7]->sym->session);
+		rte_prefetch0(ops[i + 4]->sym->m_src);
+		rte_prefetch0(ops[i + 5]->sym->m_src);
+		rte_prefetch0(ops[i + 6]->sym->m_src);
+		rte_prefetch0(ops[i + 7]->sym->m_src);
+
+		sess0 = (struct scheduler_session *)
+				ops[i]->sym->session->_private;
+		sess1 = (struct scheduler_session *)
+				ops[i+1]->sym->session->_private;
+		sess2 = (struct scheduler_session *)
+				ops[i+2]->sym->session->_private;
+		sess3 = (struct scheduler_session *)
+				ops[i+3]->sym->session->_private;
+
+		mbuf0 = ops[i]->sym->m_src;
+		mbuf1 = ops[i + 1]->sym->m_src;
+		mbuf2 = ops[i + 2]->sym->m_src;
+		mbuf3 = ops[i + 3]->sym->m_src;
+
+		index0 = mbuf0->data_len & ps_qp_ctx->threshold ? 0 : 1;
+		index1 = mbuf1->data_len & ps_qp_ctx->threshold ? 0 : 1;
+		index2 = mbuf2->data_len & ps_qp_ctx->threshold ? 0 : 1;
+		index3 = mbuf3->data_len & ps_qp_ctx->threshold ? 0 : 1;
+
+		sched_ops[index0][nb_sched_ops[index0]] = ops[i];
+		sessions[index0][nb_sched_ops[index0]++] =
+				ops[i]->sym->session;
+		mbuf0->seqn = qp_ctx->seqn++;
+		sched_ops[index1][nb_sched_ops[index1]] = ops[i + 1];
+		sessions[index1][nb_sched_ops[index1]++] =
+				ops[i + 1]->sym->session;
+		mbuf1->seqn = qp_ctx->seqn++;
+		sched_ops[index2][nb_sched_ops[index2]] = ops[i + 2];
+		sessions[index2][nb_sched_ops[index2]++] =
+				ops[i + 2]->sym->session;
+		mbuf2->seqn = qp_ctx->seqn++;
+		sched_ops[index3][nb_sched_ops[index3]] = ops[i + 3];
+		sessions[index3][nb_sched_ops[index3]++] =
+				ops[i + 3]->sym->session;
+		mbuf3->seqn = qp_ctx->seqn++;
+
+		ops[i]->sym->session = sess0->sessions[index0];
+		ops[i + 1]->sym->session = sess1->sessions[index1];
+		ops[i + 2]->sym->session = sess2->sessions[index2];
+		ops[i + 3]->sym->session = sess3->sessions[index3];
+	}
+
+	for (; i < nb_ops; i++) {
+		sess0 = (struct scheduler_session *)
+				ops[i]->sym->session->_private;
+		mbuf0 = ops[i]->sym->m_src;
+		index0 = mbuf0->data_len & ps_qp_ctx->threshold ? 0 : 1;
+		sched_ops[index0][nb_sched_ops[index0]] = ops[i];
+		sessions[index0][nb_sched_ops[index0]++] =
+				ops[i]->sym->session;
+		ops[i]->sym->session = sess0->sessions[index0];
+	}
+
+	for (i = 0; i < NB_PKT_SIZE_SLAVES; i++) {
+		slave = &ps_qp_ctx->slaves[i];
+		processed_ops = rte_cryptodev_enqueue_burst(slave->dev_id,
+				slave->qp_id, sched_ops[i], nb_sched_ops[i]);
+		slave->nb_inflight_cops += processed_ops;
+
+		if (unlikely(processed_ops < nb_sched_ops[i]))
+			for (j = processed_ops; j < nb_sched_ops[i]; j++)
+				sched_ops[i][j]->sym->session = sessions[i][j];
+		total_ops += processed_ops;
+	}
+
+	if (total_ops < nb_ops)
+		qp_ctx->seqn -= nb_ops - total_ops;
+
+	return total_ops;
+}
+
+static uint16_t
+schedule_dequeue(void *qp, struct rte_crypto_op **ops, uint16_t nb_ops)
+{
+	struct psd_scheduler_qp_ctx *qp_ctx =
+			((struct scheduler_qp_ctx *)qp)->private_qp_ctx;
+	struct scheduler_slave *slave;
+	uint16_t nb_deq_ops = 0, nb_2nd_deq_ops = 0;
+
+	/* rotating between the 2 slaves */
+	slave = &qp_ctx->slaves[qp_ctx->last_deq_index];
+	if (slave->nb_inflight_cops) {
+		nb_deq_ops = rte_cryptodev_dequeue_burst(slave->dev_id,
+			slave->qp_id, ops, nb_ops);
+		slave->nb_inflight_cops -= nb_deq_ops;
+	}
+
+	qp_ctx->last_deq_index = ~qp_ctx->last_deq_index & 0x01;
+
+	if (nb_deq_ops < nb_ops) {
+		slave = &qp_ctx->slaves[qp_ctx->last_deq_index];
+
+		if (slave->nb_inflight_cops) {
+			nb_2nd_deq_ops = rte_cryptodev_dequeue_burst(
+					slave->dev_id, slave->qp_id,
+					&ops[nb_deq_ops],
+					nb_ops - nb_deq_ops);
+
+			slave->nb_inflight_cops -= nb_2nd_deq_ops;
+			nb_deq_ops += nb_2nd_deq_ops;
+		}
+	}
+
+	return nb_deq_ops;
+}
+
+static uint16_t
+schedule_dequeue_ordering(void *qp, struct rte_crypto_op **ops,
+		uint16_t nb_ops)
+{
+	struct scheduler_qp_ctx *qp_ctx = (struct scheduler_qp_ctx *)qp;
+	struct psd_scheduler_qp_ctx *ps_qp_ctx = qp_ctx->private_qp_ctx;
+	struct scheduler_slave *slave;
+	struct rte_reorder_buffer *reorder_buff = qp_ctx->reorder_buf;
+	struct rte_mbuf *mbuf0, *mbuf1, *mbuf2, *mbuf3;
+	uint16_t nb_deq_ops = 0, nb_2nd_deq_ops = 0, nb_drained_mbufs;
+	const uint16_t nb_op_ops = nb_ops;
+	struct rte_crypto_op *op_ops[nb_op_ops];
+	struct rte_mbuf *reorder_mbufs[nb_op_ops];
+	uint16_t i;
+
+	/* check which slave has more inflight ops */
+	slave = &ps_qp_ctx->slaves[ps_qp_ctx->last_deq_index];
+	if (!slave->nb_inflight_cops) {
+		nb_deq_ops = rte_cryptodev_dequeue_burst(slave->dev_id,
+			slave->qp_id, ops, nb_ops);
+		slave->nb_inflight_cops -= nb_deq_ops;
+	}
+
+	ps_qp_ctx->last_deq_index = ~ps_qp_ctx->last_deq_index & 0x01;
+
+	if (nb_deq_ops < nb_ops) {
+		slave = &ps_qp_ctx->slaves[ps_qp_ctx->last_deq_index];
+
+		if (slave->nb_inflight_cops) {
+			nb_2nd_deq_ops = rte_cryptodev_dequeue_burst(
+					slave->dev_id, slave->qp_id,
+					&ops[nb_deq_ops],
+					nb_ops - nb_deq_ops);
+
+			slave->nb_inflight_cops -= nb_2nd_deq_ops;
+			nb_deq_ops += nb_2nd_deq_ops;
+		}
+	}
+
+	for (i = 0; i < nb_deq_ops && i < 4; i++)
+		rte_prefetch0(op_ops[i]->sym->m_src);
+
+	for (i = 0; (i < (nb_deq_ops - 8)) && (nb_deq_ops > 8); i += 4) {
+		mbuf0 = op_ops[i]->sym->m_src;
+		mbuf1 = op_ops[i + 1]->sym->m_src;
+		mbuf2 = op_ops[i + 2]->sym->m_src;
+		mbuf3 = op_ops[i + 3]->sym->m_src;
+
+		mbuf0->userdata = op_ops[i];
+		mbuf1->userdata = op_ops[i + 1];
+		mbuf2->userdata = op_ops[i + 2];
+		mbuf3->userdata = op_ops[i + 3];
+
+		rte_reorder_insert(reorder_buff, mbuf0);
+		rte_reorder_insert(reorder_buff, mbuf1);
+		rte_reorder_insert(reorder_buff, mbuf2);
+		rte_reorder_insert(reorder_buff, mbuf3);
+
+		rte_prefetch0(op_ops[i + 4]->sym->m_src);
+		rte_prefetch0(op_ops[i + 5]->sym->m_src);
+		rte_prefetch0(op_ops[i + 6]->sym->m_src);
+		rte_prefetch0(op_ops[i + 7]->sym->m_src);
+	}
+
+	for (; i < nb_deq_ops; i++) {
+		mbuf0 = op_ops[i]->sym->m_src;
+		mbuf0->userdata = op_ops[i];
+		rte_reorder_insert(reorder_buff, mbuf0);
+	}
+
+	nb_drained_mbufs = rte_reorder_drain(reorder_buff, reorder_mbufs,
+			nb_ops);
+	for (i = 0; i < nb_drained_mbufs && i < 4; i++)
+		rte_prefetch0(reorder_mbufs[i]);
+
+	for (i = 0; (i < (nb_drained_mbufs - 8)) && (nb_drained_mbufs > 8);
+			i += 4) {
+		ops[i] = *(struct rte_crypto_op **)reorder_mbufs[i]->userdata;
+		ops[i + 1] = *(struct rte_crypto_op **)
+			reorder_mbufs[i + 1]->userdata;
+		ops[i + 2] = *(struct rte_crypto_op **)
+			reorder_mbufs[i + 2]->userdata;
+		ops[i + 3] = *(struct rte_crypto_op **)
+			reorder_mbufs[i + 3]->userdata;
+
+		reorder_mbufs[i]->userdata = NULL;
+		reorder_mbufs[i + 1]->userdata = NULL;
+		reorder_mbufs[i + 2]->userdata = NULL;
+		reorder_mbufs[i + 3]->userdata = NULL;
+
+		rte_prefetch0(reorder_mbufs[i + 4]);
+		rte_prefetch0(reorder_mbufs[i + 5]);
+		rte_prefetch0(reorder_mbufs[i + 6]);
+		rte_prefetch0(reorder_mbufs[i + 7]);
+	}
+
+	for (; i < nb_drained_mbufs; i++) {
+		ops[i] = *(struct rte_crypto_op **)
+			reorder_mbufs[i]->userdata;
+		reorder_mbufs[i]->userdata = NULL;
+	}
+
+	return nb_drained_mbufs;
+}
+
+static int
+slave_attach(__rte_unused struct rte_cryptodev *dev,
+		__rte_unused uint8_t slave_id)
+{
+	return 0;
+}
+
+static int
+slave_detach(__rte_unused struct rte_cryptodev *dev,
+		__rte_unused uint8_t slave_id)
+{
+	return 0;
+}
+
+static int
+scheduler_start(struct rte_cryptodev *dev)
+{
+	struct scheduler_ctx *sched_ctx = dev->data->dev_private;
+	uint16_t i;
+
+	/* for packet size based scheduler, nb_slaves have to >= 2 */
+	if (sched_ctx->nb_slaves < NB_PKT_SIZE_SLAVES)
+		return -1;
+
+	for (i = 0; i < dev->data->nb_queue_pairs; i++) {
+		struct scheduler_qp_ctx *qp_ctx = dev->data->queue_pairs[i];
+		struct psd_scheduler_qp_ctx *ps_qp_ctx =
+				qp_ctx->private_qp_ctx;
+		uint32_t j;
+
+		memset(ps_qp_ctx->slaves, 0, NB_PKT_SIZE_SLAVES *
+				sizeof(struct scheduler_slave));
+		for (j = 0; j < NB_PKT_SIZE_SLAVES; j++) {
+			ps_qp_ctx->slaves[j].dev_id =
+					sched_ctx->slaves[j].dev_id;
+			ps_qp_ctx->slaves[j].qp_id = i;
+		}
+
+		ps_qp_ctx->threshold = PKT_SIZE_THRESHOLD;
+	}
+
+	if (sched_ctx->reordering_enabled) {
+		dev->enqueue_burst = &schedule_enqueue_ordering;
+		dev->dequeue_burst = &schedule_dequeue_ordering;
+	} else {
+		dev->enqueue_burst = &schedule_enqueue;
+		dev->dequeue_burst = &schedule_dequeue;
+	}
+
+	return 0;
+}
+
+static int
+scheduler_stop(__rte_unused struct rte_cryptodev *dev)
+{
+	return 0;
+}
+
+static int
+scheduler_config_qp(struct rte_cryptodev *dev, uint16_t qp_id)
+{
+	struct scheduler_qp_ctx *qp_ctx = dev->data->queue_pairs[qp_id];
+	struct psd_scheduler_qp_ctx *ps_qp_ctx;
+
+	ps_qp_ctx = rte_zmalloc_socket(NULL, sizeof(*ps_qp_ctx), 0,
+			rte_socket_id());
+	if (!ps_qp_ctx) {
+		CS_LOG_ERR("failed allocate memory for private queue pair");
+		return -ENOMEM;
+	}
+
+	qp_ctx->private_qp_ctx = (void *)ps_qp_ctx;
+
+	return 0;
+}
+
+static int
+scheduler_create_private_ctx(__rte_unused struct rte_cryptodev *dev)
+{
+	return 0;
+}
+
+struct rte_cryptodev_scheduler_ops scheduler_ps_ops = {
+	slave_attach,
+	slave_detach,
+	scheduler_start,
+	scheduler_stop,
+	scheduler_config_qp,
+	scheduler_create_private_ctx
+};
+
+struct rte_cryptodev_scheduler psd_scheduler = {
+		.name = "packet-size-based-scheduler",
+		.description = "scheduler which will distribute crypto op "
+				"burst based on the packet size",
+		.mode = CDEV_SCHED_MODE_PKT_SIZE_DISTRBUTE,
+		.ops = &scheduler_ps_ops
+};
+
+struct rte_cryptodev_scheduler *packet_size_based_scheduler = &psd_scheduler;
-- 
2.7.4