[dpdk-dev] [PATCH v2 08/20] crypto/dpaa_sec: support scatter gather for PDCP

Akhil Goyal akhil.goyal at nxp.com
Mon Sep 30 13:52:25 CEST 2019


This patch add support for chained input or output
mbufs for PDCP operations.

Signed-off-by: Akhil Goyal <akhil.goyal at nxp.com>
---
 drivers/crypto/dpaa_sec/dpaa_sec.c | 122 +++++++++++++++++++++++++++--
 1 file changed, 116 insertions(+), 6 deletions(-)

diff --git a/drivers/crypto/dpaa_sec/dpaa_sec.c b/drivers/crypto/dpaa_sec/dpaa_sec.c
index 3fc4a606f..b5bb87aa6 100644
--- a/drivers/crypto/dpaa_sec/dpaa_sec.c
+++ b/drivers/crypto/dpaa_sec/dpaa_sec.c
@@ -189,12 +189,18 @@ dqrr_out_fq_cb_rx(struct qman_portal *qm __always_unused,
 	if (ctx->op->sess_type == RTE_CRYPTO_OP_SECURITY_SESSION) {
 		struct qm_sg_entry *sg_out;
 		uint32_t len;
+		struct rte_mbuf *mbuf = (ctx->op->sym->m_dst == NULL) ?
+				ctx->op->sym->m_src : ctx->op->sym->m_dst;
 
 		sg_out = &job->sg[0];
 		hw_sg_to_cpu(sg_out);
 		len = sg_out->length;
-		ctx->op->sym->m_src->pkt_len = len;
-		ctx->op->sym->m_src->data_len = len;
+		mbuf->pkt_len = len;
+		while (mbuf->next != NULL) {
+			len -= mbuf->data_len;
+			mbuf = mbuf->next;
+		}
+		mbuf->data_len = len;
 	}
 	dpaa_sec_ops[dpaa_sec_op_nb++] = ctx->op;
 	dpaa_sec_op_ending(ctx);
@@ -260,6 +266,7 @@ static inline int is_auth_cipher(dpaa_sec_session *ses)
 {
 	return ((ses->cipher_alg != RTE_CRYPTO_CIPHER_NULL) &&
 		(ses->auth_alg != RTE_CRYPTO_AUTH_NULL) &&
+		(ses->proto_alg != RTE_SECURITY_PROTOCOL_PDCP) &&
 		(ses->proto_alg != RTE_SECURITY_PROTOCOL_IPSEC));
 }
 
@@ -802,12 +809,18 @@ dpaa_sec_deq(struct dpaa_sec_qp *qp, struct rte_crypto_op **ops, int nb_ops)
 		if (op->sess_type == RTE_CRYPTO_OP_SECURITY_SESSION) {
 			struct qm_sg_entry *sg_out;
 			uint32_t len;
+			struct rte_mbuf *mbuf = (op->sym->m_dst == NULL) ?
+						op->sym->m_src : op->sym->m_dst;
 
 			sg_out = &job->sg[0];
 			hw_sg_to_cpu(sg_out);
 			len = sg_out->length;
-			op->sym->m_src->pkt_len = len;
-			op->sym->m_src->data_len = len;
+			mbuf->pkt_len = len;
+			while (mbuf->next != NULL) {
+				len -= mbuf->data_len;
+				mbuf = mbuf->next;
+			}
+			mbuf->data_len = len;
 		}
 		if (!ctx->fd_status) {
 			op->status = RTE_CRYPTO_OP_STATUS_SUCCESS;
@@ -1636,6 +1649,99 @@ build_proto(struct rte_crypto_op *op, dpaa_sec_session *ses)
 	return cf;
 }
 
+static inline struct dpaa_sec_job *
+build_proto_sg(struct rte_crypto_op *op, dpaa_sec_session *ses)
+{
+	struct rte_crypto_sym_op *sym = op->sym;
+	struct dpaa_sec_job *cf;
+	struct dpaa_sec_op_ctx *ctx;
+	struct qm_sg_entry *sg, *out_sg, *in_sg;
+	struct rte_mbuf *mbuf;
+	uint8_t req_segs;
+	uint32_t in_len = 0, out_len = 0;
+
+	if (sym->m_dst) {
+		mbuf = sym->m_dst;
+	} else {
+		mbuf = sym->m_src;
+	}
+	req_segs = mbuf->nb_segs + sym->m_src->nb_segs + 2;
+	if (req_segs > MAX_SG_ENTRIES) {
+		DPAA_SEC_DP_ERR("Proto: Max sec segs supported is %d",
+				MAX_SG_ENTRIES);
+		return NULL;
+	}
+
+	ctx = dpaa_sec_alloc_ctx(ses);
+	if (!ctx)
+		return NULL;
+	cf = &ctx->job;
+	ctx->op = op;
+	/* output */
+	out_sg = &cf->sg[0];
+	out_sg->extension = 1;
+	qm_sg_entry_set64(out_sg, dpaa_mem_vtop(&cf->sg[2]));
+
+	/* 1st seg */
+	sg = &cf->sg[2];
+	qm_sg_entry_set64(sg, rte_pktmbuf_mtophys(mbuf));
+	sg->offset = 0;
+
+	/* Successive segs */
+	while (mbuf->next) {
+		sg->length = mbuf->data_len;
+		out_len += sg->length;
+		mbuf = mbuf->next;
+		cpu_to_hw_sg(sg);
+		sg++;
+		qm_sg_entry_set64(sg, rte_pktmbuf_mtophys(mbuf));
+		sg->offset = 0;
+	}
+	sg->length = mbuf->buf_len - mbuf->data_off;
+	out_len += sg->length;
+	sg->final = 1;
+	cpu_to_hw_sg(sg);
+
+	out_sg->length = out_len;
+	cpu_to_hw_sg(out_sg);
+
+	/* input */
+	mbuf = sym->m_src;
+	in_sg = &cf->sg[1];
+	in_sg->extension = 1;
+	in_sg->final = 1;
+	in_len = mbuf->data_len;
+
+	sg++;
+	qm_sg_entry_set64(in_sg, dpaa_mem_vtop(sg));
+
+	/* 1st seg */
+	qm_sg_entry_set64(sg, rte_pktmbuf_mtophys(mbuf));
+	sg->length = mbuf->data_len;
+	sg->offset = 0;
+
+	/* Successive segs */
+	mbuf = mbuf->next;
+	while (mbuf) {
+		cpu_to_hw_sg(sg);
+		sg++;
+		qm_sg_entry_set64(sg, rte_pktmbuf_mtophys(mbuf));
+		sg->length = mbuf->data_len;
+		sg->offset = 0;
+		in_len += sg->length;
+		mbuf = mbuf->next;
+	}
+	sg->final = 1;
+	cpu_to_hw_sg(sg);
+
+	in_sg->length = in_len;
+	cpu_to_hw_sg(in_sg);
+
+	sym->m_src->packet_type &= ~RTE_PTYPE_L4_MASK;
+
+	return cf;
+}
+
 static uint16_t
 dpaa_sec_enqueue_burst(void *qp, struct rte_crypto_op **ops,
 		       uint16_t nb_ops)
@@ -1707,7 +1813,9 @@ dpaa_sec_enqueue_burst(void *qp, struct rte_crypto_op **ops,
 
 			auth_only_len = op->sym->auth.data.length -
 						op->sym->cipher.data.length;
-			if (rte_pktmbuf_is_contiguous(op->sym->m_src)) {
+			if (rte_pktmbuf_is_contiguous(op->sym->m_src) &&
+				  ((op->sym->m_dst == NULL) ||
+				   rte_pktmbuf_is_contiguous(op->sym->m_dst))) {
 				if (is_proto_ipsec(ses)) {
 					cf = build_proto(op, ses);
 				} else if (is_proto_pdcp(ses)) {
@@ -1728,7 +1836,9 @@ dpaa_sec_enqueue_burst(void *qp, struct rte_crypto_op **ops,
 					goto send_pkts;
 				}
 			} else {
-				if (is_auth_only(ses)) {
+				if (is_proto_pdcp(ses) || is_proto_ipsec(ses)) {
+					cf = build_proto_sg(op, ses);
+				} else if (is_auth_only(ses)) {
 					cf = build_auth_only_sg(op, ses);
 				} else if (is_cipher_only(ses)) {
 					cf = build_cipher_only_sg(op, ses);
-- 
2.17.1



More information about the dev mailing list