[dpdk-dev] [PATCH 15/20] crypto/cnxk: add ZUC and SNOW3G encrypt

[Anoob Joseph <anoobj@marvell.com>]

From: Archana Muniganti <marchana@marvell.com>

Add PDCP opcode which handles ZUC and SNOW3G.

Signed-off-by: Ankur Dwivedi <adwivedi@marvell.com>
Signed-off-by: Anoob Joseph <anoobj@marvell.com>
Signed-off-by: Archana Muniganti <marchana@marvell.com>
Signed-off-by: Tejasree Kondoj <ktejasree@marvell.com>
---
 doc/guides/cryptodevs/features/cn10k.ini |   2 +
 doc/guides/cryptodevs/features/cn9k.ini  |   2 +
 drivers/crypto/cnxk/cnxk_se.h            | 270 ++++++++++++++++++++++++++++++-
 3 files changed, 273 insertions(+), 1 deletion(-)

diff --git a/doc/guides/cryptodevs/features/cn10k.ini b/doc/guides/cryptodevs/features/cn10k.ini
index f097d8e..8f20d07 100644
--- a/doc/guides/cryptodevs/features/cn10k.ini
+++ b/doc/guides/cryptodevs/features/cn10k.ini
@@ -30,6 +30,8 @@ AES CTR (256)  = Y
 AES XTS (128)  = Y
 AES XTS (256)  = Y
 DES CBC        = Y
+SNOW3G UEA2    = Y
+ZUC EEA3       = Y
 
 ;
 ; Supported authentication algorithms of 'cn10k' crypto driver.
diff --git a/doc/guides/cryptodevs/features/cn9k.ini b/doc/guides/cryptodevs/features/cn9k.ini
index 7007d11..fb0c09b 100644
--- a/doc/guides/cryptodevs/features/cn9k.ini
+++ b/doc/guides/cryptodevs/features/cn9k.ini
@@ -30,6 +30,8 @@ AES CTR (256)  = Y
 AES XTS (128)  = Y
 AES XTS (256)  = Y
 DES CBC        = Y
+SNOW3G UEA2    = Y
+ZUC EEA3       = Y
 
 ;
 ; Supported authentication algorithms of 'cn9k' crypto driver.
diff --git a/drivers/crypto/cnxk/cnxk_se.h b/drivers/crypto/cnxk/cnxk_se.h
index 6b2e82d..d24b2f3 100644
--- a/drivers/crypto/cnxk/cnxk_se.h
+++ b/drivers/crypto/cnxk/cnxk_se.h
@@ -1098,6 +1098,270 @@ cpt_dec_hmac_prep(uint32_t flags, uint64_t d_offs, uint64_t d_lens,
 }
 
 static __rte_always_inline int
+cpt_zuc_snow3g_enc_prep(uint32_t req_flags, uint64_t d_offs, uint64_t d_lens,
+			struct roc_se_fc_params *params,
+			struct cpt_inst_s *inst)
+{
+	uint32_t size;
+	int32_t inputlen, outputlen;
+	struct roc_se_ctx *se_ctx;
+	uint32_t mac_len = 0;
+	uint8_t pdcp_alg_type, j;
+	uint32_t encr_offset = 0, auth_offset = 0;
+	uint32_t encr_data_len = 0, auth_data_len = 0;
+	int flags, iv_len = 16;
+	uint64_t offset_ctrl;
+	uint64_t *offset_vaddr;
+	uint32_t *iv_s, iv[4];
+	union cpt_inst_w4 cpt_inst_w4;
+
+	se_ctx = params->ctx_buf.vaddr;
+	flags = se_ctx->zsk_flags;
+	mac_len = se_ctx->mac_len;
+	pdcp_alg_type = se_ctx->pdcp_alg_type;
+
+	cpt_inst_w4.s.opcode_major = ROC_SE_MAJOR_OP_ZUC_SNOW3G;
+
+	/* indicates CPTR ctx, operation type, KEY & IV mode from DPTR */
+
+	cpt_inst_w4.s.opcode_minor = ((1 << 7) | (pdcp_alg_type << 5) |
+				      (0 << 4) | (0 << 3) | (flags & 0x7));
+
+	if (flags == 0x1) {
+		/*
+		 * Microcode expects offsets in bytes
+		 * TODO: Rounding off
+		 */
+		auth_data_len = ROC_SE_AUTH_DLEN(d_lens);
+
+		/* EIA3 or UIA2 */
+		auth_offset = ROC_SE_AUTH_OFFSET(d_offs);
+		auth_offset = auth_offset / 8;
+
+		/* consider iv len */
+		auth_offset += iv_len;
+
+		inputlen = auth_offset + (RTE_ALIGN(auth_data_len, 8) / 8);
+		outputlen = mac_len;
+
+		offset_ctrl = rte_cpu_to_be_64((uint64_t)auth_offset);
+
+	} else {
+		/* EEA3 or UEA2 */
+		/*
+		 * Microcode expects offsets in bytes
+		 * TODO: Rounding off
+		 */
+		encr_data_len = ROC_SE_ENCR_DLEN(d_lens);
+
+		encr_offset = ROC_SE_ENCR_OFFSET(d_offs);
+		encr_offset = encr_offset / 8;
+		/* consider iv len */
+		encr_offset += iv_len;
+
+		inputlen = encr_offset + (RTE_ALIGN(encr_data_len, 8) / 8);
+		outputlen = inputlen;
+
+		/* iv offset is 0 */
+		offset_ctrl = rte_cpu_to_be_64((uint64_t)encr_offset << 16);
+	}
+
+	if (unlikely((encr_offset >> 16) || (auth_offset >> 8))) {
+		CPT_LOG_DP_ERR("Offset not supported");
+		CPT_LOG_DP_ERR("enc_offset: %d", encr_offset);
+		CPT_LOG_DP_ERR("auth_offset: %d", auth_offset);
+		return -1;
+	}
+
+	/* IV */
+	iv_s = (flags == 0x1) ? params->auth_iv_buf : params->iv_buf;
+
+	if (pdcp_alg_type == ROC_SE_PDCP_ALG_TYPE_SNOW3G) {
+		/*
+		 * DPDK seems to provide it in form of IV3 IV2 IV1 IV0
+		 * and BigEndian, MC needs it as IV0 IV1 IV2 IV3
+		 */
+
+		for (j = 0; j < 4; j++)
+			iv[j] = iv_s[3 - j];
+	} else {
+		/* ZUC doesn't need a swap */
+		for (j = 0; j < 4; j++)
+			iv[j] = iv_s[j];
+	}
+
+	/*
+	 * GP op header, lengths are expected in bits.
+	 */
+	cpt_inst_w4.s.param1 = encr_data_len;
+	cpt_inst_w4.s.param2 = auth_data_len;
+
+	/*
+	 * In cn9k, cn10k since we have a limitation of
+	 * IV & Offset control word not part of instruction
+	 * and need to be part of Data Buffer, we check if
+	 * head room is there and then only do the Direct mode processing
+	 */
+	if (likely((req_flags & ROC_SE_SINGLE_BUF_INPLACE) &&
+		   (req_flags & ROC_SE_SINGLE_BUF_HEADROOM))) {
+		void *dm_vaddr = params->bufs[0].vaddr;
+
+		/* Use Direct mode */
+
+		offset_vaddr = (uint64_t *)((uint8_t *)dm_vaddr -
+					    ROC_SE_OFF_CTRL_LEN - iv_len);
+
+		/* DPTR */
+		inst->dptr = (uint64_t)offset_vaddr;
+		/* RPTR should just exclude offset control word */
+		inst->rptr = (uint64_t)dm_vaddr - iv_len;
+
+		cpt_inst_w4.s.dlen = inputlen + ROC_SE_OFF_CTRL_LEN;
+
+		if (likely(iv_len)) {
+			uint32_t *iv_d = (uint32_t *)((uint8_t *)offset_vaddr +
+						      ROC_SE_OFF_CTRL_LEN);
+			memcpy(iv_d, iv, 16);
+		}
+
+		*offset_vaddr = offset_ctrl;
+	} else {
+		void *m_vaddr = params->meta_buf.vaddr;
+		uint32_t i, g_size_bytes, s_size_bytes;
+		struct roc_se_sglist_comp *gather_comp;
+		struct roc_se_sglist_comp *scatter_comp;
+		uint8_t *in_buffer;
+		uint32_t *iv_d;
+
+		/* save space for iv */
+		offset_vaddr = m_vaddr;
+
+		m_vaddr = (uint8_t *)m_vaddr + ROC_SE_OFF_CTRL_LEN + iv_len;
+
+		cpt_inst_w4.s.opcode_major |= (uint64_t)ROC_SE_DMA_MODE;
+
+		/* DPTR has SG list */
+		in_buffer = m_vaddr;
+
+		((uint16_t *)in_buffer)[0] = 0;
+		((uint16_t *)in_buffer)[1] = 0;
+
+		/* TODO Add error check if space will be sufficient */
+		gather_comp =
+			(struct roc_se_sglist_comp *)((uint8_t *)m_vaddr + 8);
+
+		/*
+		 * Input Gather List
+		 */
+		i = 0;
+
+		/* Offset control word followed by iv */
+
+		i = fill_sg_comp(gather_comp, i, (uint64_t)offset_vaddr,
+				 ROC_SE_OFF_CTRL_LEN + iv_len);
+
+		/* iv offset is 0 */
+		*offset_vaddr = offset_ctrl;
+
+		iv_d = (uint32_t *)((uint8_t *)offset_vaddr +
+				    ROC_SE_OFF_CTRL_LEN);
+		memcpy(iv_d, iv, 16);
+
+		/* input data */
+		size = inputlen - iv_len;
+		if (size) {
+			i = fill_sg_comp_from_iov(gather_comp, i,
+						  params->src_iov, 0, &size,
+						  NULL, 0);
+			if (unlikely(size)) {
+				CPT_LOG_DP_ERR("Insufficient buffer space,"
+					       " size %d needed",
+					       size);
+				return -1;
+			}
+		}
+		((uint16_t *)in_buffer)[2] = rte_cpu_to_be_16(i);
+		g_size_bytes =
+			((i + 3) / 4) * sizeof(struct roc_se_sglist_comp);
+
+		/*
+		 * Output Scatter List
+		 */
+
+		i = 0;
+		scatter_comp =
+			(struct roc_se_sglist_comp *)((uint8_t *)gather_comp +
+						      g_size_bytes);
+
+		if (flags == 0x1) {
+			/* IV in SLIST only for EEA3 & UEA2 */
+			iv_len = 0;
+		}
+
+		if (iv_len) {
+			i = fill_sg_comp(scatter_comp, i,
+					 (uint64_t)offset_vaddr +
+						 ROC_SE_OFF_CTRL_LEN,
+					 iv_len);
+		}
+
+		/* Add output data */
+		if (req_flags & ROC_SE_VALID_MAC_BUF) {
+			size = outputlen - iv_len - mac_len;
+			if (size) {
+				i = fill_sg_comp_from_iov(scatter_comp, i,
+							  params->dst_iov, 0,
+							  &size, NULL, 0);
+
+				if (unlikely(size)) {
+					CPT_LOG_DP_ERR(
+						"Insufficient buffer space,"
+						" size %d needed",
+						size);
+					return -1;
+				}
+			}
+
+			/* mac data */
+			if (mac_len) {
+				i = fill_sg_comp_from_buf(scatter_comp, i,
+							  &params->mac_buf);
+			}
+		} else {
+			/* Output including mac */
+			size = outputlen - iv_len;
+			if (size) {
+				i = fill_sg_comp_from_iov(scatter_comp, i,
+							  params->dst_iov, 0,
+							  &size, NULL, 0);
+
+				if (unlikely(size)) {
+					CPT_LOG_DP_ERR(
+						"Insufficient buffer space,"
+						" size %d needed",
+						size);
+					return -1;
+				}
+			}
+		}
+		((uint16_t *)in_buffer)[3] = rte_cpu_to_be_16(i);
+		s_size_bytes =
+			((i + 3) / 4) * sizeof(struct roc_se_sglist_comp);
+
+		size = g_size_bytes + s_size_bytes + ROC_SE_SG_LIST_HDR_SIZE;
+
+		/* This is DPTR len in case of SG mode */
+		cpt_inst_w4.s.dlen = size;
+
+		inst->dptr = (uint64_t)in_buffer;
+	}
+
+	inst->w4.u64 = cpt_inst_w4.u64;
+
+	return 0;
+}
+
+static __rte_always_inline int
 cpt_fc_dec_hmac_prep(uint32_t flags, uint64_t d_offs, uint64_t d_lens,
 		     struct roc_se_fc_params *fc_params,
 		     struct cpt_inst_s *inst)
@@ -1124,8 +1388,12 @@ cpt_fc_enc_hmac_prep(uint32_t flags, uint64_t d_offs, uint64_t d_lens,
 
 	fc_type = ctx->fc_type;
 
-	if (likely(fc_type == ROC_SE_FC_GEN))
+	if (likely(fc_type == ROC_SE_FC_GEN)) {
 		ret = cpt_enc_hmac_prep(flags, d_offs, d_lens, fc_params, inst);
+	} else if (fc_type == ROC_SE_PDCP) {
+		ret = cpt_zuc_snow3g_enc_prep(flags, d_offs, d_lens, fc_params,
+					      inst);
+	}
 
 	return ret;
 }
-- 
2.7.4