crypto: caam/qi2 - add DPAA2-CAAM driver

Add CAAM driver that works using the DPSECI backend, i.e. manages
DPSECI DPAA2 objects sitting on the Management Complex (MC) fsl-mc bus.

Data transfers (crypto requests) are sent/received to/from CAAM crypto
engine via Queue Interface (v2), this being similar to existing caam/qi.
OTOH, configuration/setup (obtaining virtual queue IDs, authorization
etc.) is done by sending commands to the MC f/w.

Note that the CAAM accelerator included in DPAA2 platforms still has
Job Rings. However, the driver being added does not handle access
via this backend. Kconfig & Makefile are updated such that DPAA2-CAAM
(a.k.a. "caam/qi2") driver does not depend on caam/jr or caam/qi
backends - which rely on platform bus support (ctrl.c).

Support for the following aead and authenc algorithms is also added
in this patch:
-aead:
gcm(aes)
rfc4106(gcm(aes))
rfc4543(gcm(aes))
-authenc:
authenc(hmac({md5,sha*}),cbc({aes,des,des3_ede}))
echainiv(authenc(hmac({md5,sha*}),cbc({aes,des,des3_ede})))
authenc(hmac({md5,sha*}),rfc3686(ctr(aes))
seqiv(authenc(hmac({md5,sha*}),rfc3686(ctr(aes)))

Signed-off-by: Horia Geantă <horia.geanta@nxp.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

6 files changed, 3108 insertions, 16 deletions

diff --git a/drivers/crypto/caam/Kconfig b/drivers/crypto/caam/Kconfig
index 1eb852765469..a186b853b05c 100644
--- a/drivers/crypto/caam/Kconfig
+++ b/drivers/crypto/caam/Kconfig
@@ -1,7 +1,11 @@
+config CRYPTO_DEV_FSL_CAAM_COMMON
+	tristate
+
 config CRYPTO_DEV_FSL_CAAM
-	tristate "Freescale CAAM-Multicore driver backend"
+	tristate "Freescale CAAM-Multicore platform driver backend"
 	depends on FSL_SOC || ARCH_MXC || ARCH_LAYERSCAPE
 	select SOC_BUS
+	select CRYPTO_DEV_FSL_CAAM_COMMON
 	help
 	  Enables the driver module for Freescale's Cryptographic Accelerator
 	  and Assurance Module (CAAM), also known as the SEC version 4 (SEC4).
@@ -12,9 +16,16 @@ config CRYPTO_DEV_FSL_CAAM
 	  To compile this driver as a module, choose M here: the module
 	  will be called caam.
 
+if CRYPTO_DEV_FSL_CAAM
+
+config CRYPTO_DEV_FSL_CAAM_DEBUG
+	bool "Enable debug output in CAAM driver"
+	help
+	  Selecting this will enable printing of various debug
+	  information in the CAAM driver.
+
 config CRYPTO_DEV_FSL_CAAM_JR
 	tristate "Freescale CAAM Job Ring driver backend"
-	depends on CRYPTO_DEV_FSL_CAAM
 	default y
 	help
 	  Enables the driver module for Job Rings which are part of
@@ -25,9 +36,10 @@ config CRYPTO_DEV_FSL_CAAM_JR
 	  To compile this driver as a module, choose M here: the module
 	  will be called caam_jr.
 
+if CRYPTO_DEV_FSL_CAAM_JR
+
 config CRYPTO_DEV_FSL_CAAM_RINGSIZE
 	int "Job Ring size"
-	depends on CRYPTO_DEV_FSL_CAAM_JR
 	range 2 9
 	default "9"
 	help
@@ -45,7 +57,6 @@ config CRYPTO_DEV_FSL_CAAM_RINGSIZE
 
 config CRYPTO_DEV_FSL_CAAM_INTC
 	bool "Job Ring interrupt coalescing"
-	depends on CRYPTO_DEV_FSL_CAAM_JR
 	help
 	  Enable the Job Ring's interrupt coalescing feature.
 
@@ -75,7 +86,6 @@ config CRYPTO_DEV_FSL_CAAM_INTC_TIME_THLD
 
 config CRYPTO_DEV_FSL_CAAM_CRYPTO_API
 	tristate "Register algorithm implementations with the Crypto API"
-	depends on CRYPTO_DEV_FSL_CAAM_JR
 	default y
 	select CRYPTO_AEAD
 	select CRYPTO_AUTHENC
@@ -90,7 +100,7 @@ config CRYPTO_DEV_FSL_CAAM_CRYPTO_API
 
 config CRYPTO_DEV_FSL_CAAM_CRYPTO_API_QI
 	tristate "Queue Interface as Crypto API backend"
-	depends on CRYPTO_DEV_FSL_CAAM_JR && FSL_DPAA && NET
+	depends on FSL_DPAA && NET
 	default y
 	select CRYPTO_AUTHENC
 	select CRYPTO_BLKCIPHER
@@ -107,7 +117,6 @@ config CRYPTO_DEV_FSL_CAAM_CRYPTO_API_QI
 
 config CRYPTO_DEV_FSL_CAAM_AHASH_API
 	tristate "Register hash algorithm implementations with Crypto API"
-	depends on CRYPTO_DEV_FSL_CAAM_JR
 	default y
 	select CRYPTO_HASH
 	help
@@ -119,7 +128,6 @@ config CRYPTO_DEV_FSL_CAAM_AHASH_API
 
 config CRYPTO_DEV_FSL_CAAM_PKC_API
         tristate "Register public key cryptography implementations with Crypto API"
-        depends on CRYPTO_DEV_FSL_CAAM_JR
         default y
         select CRYPTO_RSA
         help
@@ -131,7 +139,6 @@ config CRYPTO_DEV_FSL_CAAM_PKC_API
 
 config CRYPTO_DEV_FSL_CAAM_RNG_API
 	tristate "Register caam device for hwrng API"
-	depends on CRYPTO_DEV_FSL_CAAM_JR
 	default y
 	select CRYPTO_RNG
 	select HW_RANDOM
@@ -142,13 +149,25 @@ config CRYPTO_DEV_FSL_CAAM_RNG_API
 	  To compile this as a module, choose M here: the module
 	  will be called caamrng.
 
-config CRYPTO_DEV_FSL_CAAM_DEBUG
-	bool "Enable debug output in CAAM driver"
-	depends on CRYPTO_DEV_FSL_CAAM
+endif # CRYPTO_DEV_FSL_CAAM_JR
+
+endif # CRYPTO_DEV_FSL_CAAM
+
+config CRYPTO_DEV_FSL_DPAA2_CAAM
+	tristate "QorIQ DPAA2 CAAM (DPSECI) driver"
+	depends on FSL_MC_DPIO
+	select CRYPTO_DEV_FSL_CAAM_COMMON
+	select CRYPTO_AUTHENC
+	select CRYPTO_AEAD
 	help
-	  Selecting this will enable printing of various debug
-	  information in the CAAM driver.
+	  CAAM driver for QorIQ Data Path Acceleration Architecture 2.
+	  It handles DPSECI DPAA2 objects that sit on the Management Complex
+	  (MC) fsl-mc bus.
+
+	  To compile this as a module, choose M here: the module
+	  will be called dpaa2_caam.
 
 config CRYPTO_DEV_FSL_CAAM_CRYPTO_API_DESC
 	def_tristate (CRYPTO_DEV_FSL_CAAM_CRYPTO_API || \
-		      CRYPTO_DEV_FSL_CAAM_CRYPTO_API_QI)
+		      CRYPTO_DEV_FSL_CAAM_CRYPTO_API_QI || \
+		      CRYPTO_DEV_FSL_DPAA2_CAAM)
diff --git a/drivers/crypto/caam/Makefile b/drivers/crypto/caam/Makefile
index cb652ee7dfc8..fd507bab1828 100644
--- a/drivers/crypto/caam/Makefile
+++ b/drivers/crypto/caam/Makefile
@@ -6,6 +6,9 @@ ifeq ($(CONFIG_CRYPTO_DEV_FSL_CAAM_DEBUG), y)
 	ccflags-y := -DDEBUG
 endif
 
+ccflags-y += -DVERSION=\"\"
+
+obj-$(CONFIG_CRYPTO_DEV_FSL_CAAM_COMMON) += error.o
 obj-$(CONFIG_CRYPTO_DEV_FSL_CAAM) += caam.o
 obj-$(CONFIG_CRYPTO_DEV_FSL_CAAM_JR) += caam_jr.o
 obj-$(CONFIG_CRYPTO_DEV_FSL_CAAM_CRYPTO_API) += caamalg.o
@@ -16,9 +19,13 @@ obj-$(CONFIG_CRYPTO_DEV_FSL_CAAM_RNG_API) += caamrng.o
 obj-$(CONFIG_CRYPTO_DEV_FSL_CAAM_PKC_API) += caam_pkc.o
 
 caam-objs := ctrl.o
-caam_jr-objs := jr.o key_gen.o error.o
+caam_jr-objs := jr.o key_gen.o
 caam_pkc-y := caampkc.o pkc_desc.o
 ifneq ($(CONFIG_CRYPTO_DEV_FSL_CAAM_CRYPTO_API_QI),)
 	ccflags-y += -DCONFIG_CAAM_QI
 	caam-objs += qi.o
 endif
+
+obj-$(CONFIG_CRYPTO_DEV_FSL_DPAA2_CAAM) += dpaa2_caam.o
+
+dpaa2_caam-y    := caamalg_qi2.o dpseci.o
diff --git a/drivers/crypto/caam/caamalg_qi2.c b/drivers/crypto/caam/caamalg_qi2.c
new file mode 100644
index 000000000000..9d97a51070b1
--- /dev/null
+++ b/drivers/crypto/caam/caamalg_qi2.c
@@ -0,0 +1,2872 @@
+// SPDX-License-Identifier: (GPL-2.0+ OR BSD-3-Clause)
+/*
+ * Copyright 2015-2016 Freescale Semiconductor Inc.
+ * Copyright 2017-2018 NXP
+ */
+
+#include "compat.h"
+#include "regs.h"
+#include "caamalg_qi2.h"
+#include "dpseci_cmd.h"
+#include "desc_constr.h"
+#include "error.h"
+#include "sg_sw_sec4.h"
+#include "sg_sw_qm2.h"
+#include "key_gen.h"
+#include "caamalg_desc.h"
+#include <linux/fsl/mc.h>
+#include <soc/fsl/dpaa2-io.h>
+#include <soc/fsl/dpaa2-fd.h>
+
+#define CAAM_CRA_PRIORITY	2000
+
+/* max key is sum of AES_MAX_KEY_SIZE, max split key size */
+#define CAAM_MAX_KEY_SIZE	(AES_MAX_KEY_SIZE + CTR_RFC3686_NONCE_SIZE + \
+				 SHA512_DIGEST_SIZE * 2)
+
+#ifndef CONFIG_CRYPTO_DEV_FSL_CAAM
+bool caam_little_end;
+EXPORT_SYMBOL(caam_little_end);
+bool caam_imx;
+EXPORT_SYMBOL(caam_imx);
+#endif
+
+/*
+ * This is a a cache of buffers, from which the users of CAAM QI driver
+ * can allocate short buffers. It's speedier than doing kmalloc on the hotpath.
+ * NOTE: A more elegant solution would be to have some headroom in the frames
+ *       being processed. This can be added by the dpaa2-eth driver. This would
+ *       pose a problem for userspace application processing which cannot
+ *       know of this limitation. So for now, this will work.
+ * NOTE: The memcache is SMP-safe. No need to handle spinlocks in-here
+ */
+static struct kmem_cache *qi_cache;
+
+struct caam_alg_entry {
+	struct device *dev;
+	int class1_alg_type;
+	int class2_alg_type;
+	bool rfc3686;
+	bool geniv;
+};
+
+struct caam_aead_alg {
+	struct aead_alg aead;
+	struct caam_alg_entry caam;
+	bool registered;
+};
+
+/**
+ * caam_ctx - per-session context
+ * @flc: Flow Contexts array
+ * @key:  [authentication key], encryption key
+ * @flc_dma: I/O virtual addresses of the Flow Contexts
+ * @key_dma: I/O virtual address of the key
+ * @dir: DMA direction for mapping key and Flow Contexts
+ * @dev: dpseci device
+ * @adata: authentication algorithm details
+ * @cdata: encryption algorithm details
+ * @authsize: authentication tag (a.k.a. ICV / MAC) size
+ */
+struct caam_ctx {
+	struct caam_flc flc[NUM_OP];
+	u8 key[CAAM_MAX_KEY_SIZE];
+	dma_addr_t flc_dma[NUM_OP];
+	dma_addr_t key_dma;
+	enum dma_data_direction dir;
+	struct device *dev;
+	struct alginfo adata;
+	struct alginfo cdata;
+	unsigned int authsize;
+};
+
+static void *dpaa2_caam_iova_to_virt(struct dpaa2_caam_priv *priv,
+				     dma_addr_t iova_addr)
+{
+	phys_addr_t phys_addr;
+
+	phys_addr = priv->domain ? iommu_iova_to_phys(priv->domain, iova_addr) :
+				   iova_addr;
+
+	return phys_to_virt(phys_addr);
+}
+
+/*
+ * qi_cache_zalloc - Allocate buffers from CAAM-QI cache
+ *
+ * Allocate data on the hotpath. Instead of using kzalloc, one can use the
+ * services of the CAAM QI memory cache (backed by kmem_cache). The buffers
+ * will have a size of CAAM_QI_MEMCACHE_SIZE, which should be sufficient for
+ * hosting 16 SG entries.
+ *
+ * @flags - flags that would be used for the equivalent kmalloc(..) call
+ *
+ * Returns a pointer to a retrieved buffer on success or NULL on failure.
+ */
+static inline void *qi_cache_zalloc(gfp_t flags)
+{
+	return kmem_cache_zalloc(qi_cache, flags);
+}
+
+/*
+ * qi_cache_free - Frees buffers allocated from CAAM-QI cache
+ *
+ * @obj - buffer previously allocated by qi_cache_zalloc
+ *
+ * No checking is being done, the call is a passthrough call to
+ * kmem_cache_free(...)
+ */
+static inline void qi_cache_free(void *obj)
+{
+	kmem_cache_free(qi_cache, obj);
+}
+
+static struct caam_request *to_caam_req(struct crypto_async_request *areq)
+{
+	switch (crypto_tfm_alg_type(areq->tfm)) {
+	case CRYPTO_ALG_TYPE_SKCIPHER:
+		return skcipher_request_ctx(skcipher_request_cast(areq));
+	case CRYPTO_ALG_TYPE_AEAD:
+		return aead_request_ctx(container_of(areq, struct aead_request,
+						     base));
+	default:
+		return ERR_PTR(-EINVAL);
+	}
+}
+
+static void caam_unmap(struct device *dev, struct scatterlist *src,
+		       struct scatterlist *dst, int src_nents,
+		       int dst_nents, dma_addr_t iv_dma, int ivsize,
+		       dma_addr_t qm_sg_dma, int qm_sg_bytes)
+{
+	if (dst != src) {
+		if (src_nents)
+			dma_unmap_sg(dev, src, src_nents, DMA_TO_DEVICE);
+		dma_unmap_sg(dev, dst, dst_nents, DMA_FROM_DEVICE);
+	} else {
+		dma_unmap_sg(dev, src, src_nents, DMA_BIDIRECTIONAL);
+	}
+
+	if (iv_dma)
+		dma_unmap_single(dev, iv_dma, ivsize, DMA_TO_DEVICE);
+
+	if (qm_sg_bytes)
+		dma_unmap_single(dev, qm_sg_dma, qm_sg_bytes, DMA_TO_DEVICE);
+}
+
+static int aead_set_sh_desc(struct crypto_aead *aead)
+{
+	struct caam_aead_alg *alg = container_of(crypto_aead_alg(aead),
+						 typeof(*alg), aead);
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	unsigned int ivsize = crypto_aead_ivsize(aead);
+	struct device *dev = ctx->dev;
+	struct dpaa2_caam_priv *priv = dev_get_drvdata(dev);
+	struct caam_flc *flc;
+	u32 *desc;
+	u32 ctx1_iv_off = 0;
+	u32 *nonce = NULL;
+	unsigned int data_len[2];
+	u32 inl_mask;
+	const bool ctr_mode = ((ctx->cdata.algtype & OP_ALG_AAI_MASK) ==
+			       OP_ALG_AAI_CTR_MOD128);
+	const bool is_rfc3686 = alg->caam.rfc3686;
+
+	if (!ctx->cdata.keylen || !ctx->authsize)
+		return 0;
+
+	/*
+	 * AES-CTR needs to load IV in CONTEXT1 reg
+	 * at an offset of 128bits (16bytes)
+	 * CONTEXT1[255:128] = IV
+	 */
+	if (ctr_mode)
+		ctx1_iv_off = 16;
+
+	/*
+	 * RFC3686 specific:
+	 *	CONTEXT1[255:128] = {NONCE, IV, COUNTER}
+	 */
+	if (is_rfc3686) {
+		ctx1_iv_off = 16 + CTR_RFC3686_NONCE_SIZE;
+		nonce = (u32 *)((void *)ctx->key + ctx->adata.keylen_pad +
+				ctx->cdata.keylen - CTR_RFC3686_NONCE_SIZE);
+	}
+
+	data_len[0] = ctx->adata.keylen_pad;
+	data_len[1] = ctx->cdata.keylen;
+
+	/* aead_encrypt shared descriptor */
+	if (desc_inline_query((alg->caam.geniv ? DESC_QI_AEAD_GIVENC_LEN :
+						 DESC_QI_AEAD_ENC_LEN) +
+			      (is_rfc3686 ? DESC_AEAD_CTR_RFC3686_LEN : 0),
+			      DESC_JOB_IO_LEN, data_len, &inl_mask,
+			      ARRAY_SIZE(data_len)) < 0)
+		return -EINVAL;
+
+	if (inl_mask & 1)
+		ctx->adata.key_virt = ctx->key;
+	else
+		ctx->adata.key_dma = ctx->key_dma;
+
+	if (inl_mask & 2)
+		ctx->cdata.key_virt = ctx->key + ctx->adata.keylen_pad;
+	else
+		ctx->cdata.key_dma = ctx->key_dma + ctx->adata.keylen_pad;
+
+	ctx->adata.key_inline = !!(inl_mask & 1);
+	ctx->cdata.key_inline = !!(inl_mask & 2);
+
+	flc = &ctx->flc[ENCRYPT];
+	desc = flc->sh_desc;
+
+	if (alg->caam.geniv)
+		cnstr_shdsc_aead_givencap(desc, &ctx->cdata, &ctx->adata,
+					  ivsize, ctx->authsize, is_rfc3686,
+					  nonce, ctx1_iv_off, true,
+					  priv->sec_attr.era);
+	else
+		cnstr_shdsc_aead_encap(desc, &ctx->cdata, &ctx->adata,
+				       ivsize, ctx->authsize, is_rfc3686, nonce,
+				       ctx1_iv_off, true, priv->sec_attr.era);
+
+	flc->flc[1] = cpu_to_caam32(desc_len(desc)); /* SDL */
+	dma_sync_single_for_device(dev, ctx->flc_dma[ENCRYPT],
+				   sizeof(flc->flc) + desc_bytes(desc),
+				   ctx->dir);
+
+	/* aead_decrypt shared descriptor */
+	if (desc_inline_query(DESC_QI_AEAD_DEC_LEN +
+			      (is_rfc3686 ? DESC_AEAD_CTR_RFC3686_LEN : 0),
+			      DESC_JOB_IO_LEN, data_len, &inl_mask,
+			      ARRAY_SIZE(data_len)) < 0)
+		return -EINVAL;
+
+	if (inl_mask & 1)
+		ctx->adata.key_virt = ctx->key;
+	else
+		ctx->adata.key_dma = ctx->key_dma;
+
+	if (inl_mask & 2)
+		ctx->cdata.key_virt = ctx->key + ctx->adata.keylen_pad;
+	else
+		ctx->cdata.key_dma = ctx->key_dma + ctx->adata.keylen_pad;
+
+	ctx->adata.key_inline = !!(inl_mask & 1);
+	ctx->cdata.key_inline = !!(inl_mask & 2);
+
+	flc = &ctx->flc[DECRYPT];
+	desc = flc->sh_desc;
+	cnstr_shdsc_aead_decap(desc, &ctx->cdata, &ctx->adata,
+			       ivsize, ctx->authsize, alg->caam.geniv,
+			       is_rfc3686, nonce, ctx1_iv_off, true,
+			       priv->sec_attr.era);
+	flc->flc[1] = cpu_to_caam32(desc_len(desc)); /* SDL */
+	dma_sync_single_for_device(dev, ctx->flc_dma[DECRYPT],
+				   sizeof(flc->flc) + desc_bytes(desc),
+				   ctx->dir);
+
+	return 0;
+}
+
+static int aead_setauthsize(struct crypto_aead *authenc, unsigned int authsize)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(authenc);
+
+	ctx->authsize = authsize;
+	aead_set_sh_desc(authenc);
+
+	return 0;
+}
+
+static int aead_setkey(struct crypto_aead *aead, const u8 *key,
+		       unsigned int keylen)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *dev = ctx->dev;
+	struct crypto_authenc_keys keys;
+
+	if (crypto_authenc_extractkeys(&keys, key, keylen) != 0)
+		goto badkey;
+
+	dev_dbg(dev, "keylen %d enckeylen %d authkeylen %d\n",
+		keys.authkeylen + keys.enckeylen, keys.enckeylen,
+		keys.authkeylen);
+	print_hex_dump_debug("key in @" __stringify(__LINE__)": ",
+			     DUMP_PREFIX_ADDRESS, 16, 4, key, keylen, 1);
+
+	ctx->adata.keylen = keys.authkeylen;
+	ctx->adata.keylen_pad = split_key_len(ctx->adata.algtype &
+					      OP_ALG_ALGSEL_MASK);
+
+	if (ctx->adata.keylen_pad + keys.enckeylen > CAAM_MAX_KEY_SIZE)
+		goto badkey;
+
+	memcpy(ctx->key, keys.authkey, keys.authkeylen);
+	memcpy(ctx->key + ctx->adata.keylen_pad, keys.enckey, keys.enckeylen);
+	dma_sync_single_for_device(dev, ctx->key_dma, ctx->adata.keylen_pad +
+				   keys.enckeylen, ctx->dir);
+	print_hex_dump_debug("ctx.key@" __stringify(__LINE__)": ",
+			     DUMP_PREFIX_ADDRESS, 16, 4, ctx->key,
+			     ctx->adata.keylen_pad + keys.enckeylen, 1);
+
+	ctx->cdata.keylen = keys.enckeylen;
+
+	memzero_explicit(&keys, sizeof(keys));
+	return aead_set_sh_desc(aead);
+badkey:
+	crypto_aead_set_flags(aead, CRYPTO_TFM_RES_BAD_KEY_LEN);
+	memzero_explicit(&keys, sizeof(keys));
+	return -EINVAL;
+}
+
+static struct aead_edesc *aead_edesc_alloc(struct aead_request *req,
+					   bool encrypt)
+{
+	struct crypto_aead *aead = crypto_aead_reqtfm(req);
+	struct caam_request *req_ctx = aead_request_ctx(req);
+	struct dpaa2_fl_entry *in_fle = &req_ctx->fd_flt[1];
+	struct dpaa2_fl_entry *out_fle = &req_ctx->fd_flt[0];
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct caam_aead_alg *alg = container_of(crypto_aead_alg(aead),
+						 typeof(*alg), aead);
+	struct device *dev = ctx->dev;
+	gfp_t flags = (req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP) ?
+		      GFP_KERNEL : GFP_ATOMIC;
+	int src_nents, mapped_src_nents, dst_nents = 0, mapped_dst_nents = 0;
+	struct aead_edesc *edesc;
+	dma_addr_t qm_sg_dma, iv_dma = 0;
+	int ivsize = 0;
+	unsigned int authsize = ctx->authsize;
+	int qm_sg_index = 0, qm_sg_nents = 0, qm_sg_bytes;
+	int in_len, out_len;
+	struct dpaa2_sg_entry *sg_table;
+
+	/* allocate space for base edesc, link tables and IV */
+	edesc = qi_cache_zalloc(GFP_DMA | flags);
+	if (unlikely(!edesc)) {
+		dev_err(dev, "could not allocate extended descriptor\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	if (unlikely(req->dst != req->src)) {
+		src_nents = sg_nents_for_len(req->src, req->assoclen +
+					     req->cryptlen);
+		if (unlikely(src_nents < 0)) {
+			dev_err(dev, "Insufficient bytes (%d) in src S/G\n",
+				req->assoclen + req->cryptlen);
+			qi_cache_free(edesc);
+			return ERR_PTR(src_nents);
+		}
+
+		dst_nents = sg_nents_for_len(req->dst, req->assoclen +
+					     req->cryptlen +
+					     (encrypt ? authsize :
+							(-authsize)));
+		if (unlikely(dst_nents < 0)) {
+			dev_err(dev, "Insufficient bytes (%d) in dst S/G\n",
+				req->assoclen + req->cryptlen +
+				(encrypt ? authsize : (-authsize)));
+			qi_cache_free(edesc);
+			return ERR_PTR(dst_nents);
+		}
+
+		if (src_nents) {
+			mapped_src_nents = dma_map_sg(dev, req->src, src_nents,
+						      DMA_TO_DEVICE);
+			if (unlikely(!mapped_src_nents)) {
+				dev_err(dev, "unable to map source\n");
+				qi_cache_free(edesc);
+				return ERR_PTR(-ENOMEM);
+			}
+		} else {
+			mapped_src_nents = 0;
+		}
+
+		mapped_dst_nents = dma_map_sg(dev, req->dst, dst_nents,
+					      DMA_FROM_DEVICE);
+		if (unlikely(!mapped_dst_nents)) {
+			dev_err(dev, "unable to map destination\n");
+			dma_unmap_sg(dev, req->src, src_nents, DMA_TO_DEVICE);
+			qi_cache_free(edesc);
+			return ERR_PTR(-ENOMEM);
+		}
+	} else {
+		src_nents = sg_nents_for_len(req->src, req->assoclen +
+					     req->cryptlen +
+						(encrypt ? authsize : 0));
+		if (unlikely(src_nents < 0)) {
+			dev_err(dev, "Insufficient bytes (%d) in src S/G\n",
+				req->assoclen + req->cryptlen +
+				(encrypt ? authsize : 0));
+			qi_cache_free(edesc);
+			return ERR_PTR(src_nents);
+		}
+
+		mapped_src_nents = dma_map_sg(dev, req->src, src_nents,
+					      DMA_BIDIRECTIONAL);
+		if (unlikely(!mapped_src_nents)) {
+			dev_err(dev, "unable to map source\n");
+			qi_cache_free(edesc);
+			return ERR_PTR(-ENOMEM);
+		}
+	}
+
+	if ((alg->caam.rfc3686 && encrypt) || !alg->caam.geniv)
+		ivsize = crypto_aead_ivsize(aead);
+
+	/*
+	 * Create S/G table: req->assoclen, [IV,] req->src [, req->dst].
+	 * Input is not contiguous.
+	 */
+	qm_sg_nents = 1 + !!ivsize + mapped_src_nents +
+		      (mapped_dst_nents > 1 ? mapped_dst_nents : 0);
+	sg_table = &edesc->sgt[0];
+	qm_sg_bytes = qm_sg_nents * sizeof(*sg_table);
+	if (unlikely(offsetof(struct aead_edesc, sgt) + qm_sg_bytes + ivsize >
+		     CAAM_QI_MEMCACHE_SIZE)) {
+		dev_err(dev, "No space for %d S/G entries and/or %dB IV\n",
+			qm_sg_nents, ivsize);
+		caam_unmap(dev, req->src, req->dst, src_nents, dst_nents, 0,
+			   0, 0, 0);
+		qi_cache_free(edesc);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	if (ivsize) {
+		u8 *iv = (u8 *)(sg_table + qm_sg_nents);
+
+		/* Make sure IV is located in a DMAable area */
+		memcpy(iv, req->iv, ivsize);
+
+		iv_dma = dma_map_single(dev, iv, ivsize, DMA_TO_DEVICE);
+		if (dma_mapping_error(dev, iv_dma)) {
+			dev_err(dev, "unable to map IV\n");
+			caam_unmap(dev, req->src, req->dst, src_nents,
+				   dst_nents, 0, 0, 0, 0);
+			qi_cache_free(edesc);
+			return ERR_PTR(-ENOMEM);
+		}
+	}
+
+	edesc->src_nents = src_nents;
+	edesc->dst_nents = dst_nents;
+	edesc->iv_dma = iv_dma;
+
+	edesc->assoclen = cpu_to_caam32(req->assoclen);
+	edesc->assoclen_dma = dma_map_single(dev, &edesc->assoclen, 4,
+					     DMA_TO_DEVICE);
+	if (dma_mapping_error(dev, edesc->assoclen_dma)) {
+		dev_err(dev, "unable to map assoclen\n");
+		caam_unmap(dev, req->src, req->dst, src_nents, dst_nents,
+			   iv_dma, ivsize, 0, 0);
+		qi_cache_free(edesc);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	dma_to_qm_sg_one(sg_table, edesc->assoclen_dma, 4, 0);
+	qm_sg_index++;
+	if (ivsize) {
+		dma_to_qm_sg_one(sg_table + qm_sg_index, iv_dma, ivsize, 0);
+		qm_sg_index++;
+	}
+	sg_to_qm_sg_last(req->src, mapped_src_nents, sg_table + qm_sg_index, 0);
+	qm_sg_index += mapped_src_nents;
+
+	if (mapped_dst_nents > 1)
+		sg_to_qm_sg_last(req->dst, mapped_dst_nents, sg_table +
+				 qm_sg_index, 0);
+
+	qm_sg_dma = dma_map_single(dev, sg_table, qm_sg_bytes, DMA_TO_DEVICE);
+	if (dma_mapping_error(dev, qm_sg_dma)) {
+		dev_err(dev, "unable to map S/G table\n");
+		dma_unmap_single(dev, edesc->assoclen_dma, 4, DMA_TO_DEVICE);
+		caam_unmap(dev, req->src, req->dst, src_nents, dst_nents,
+			   iv_dma, ivsize, 0, 0);
+		qi_cache_free(edesc);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	edesc->qm_sg_dma = qm_sg_dma;
+	edesc->qm_sg_bytes = qm_sg_bytes;
+
+	out_len = req->assoclen + req->cryptlen +
+		  (encrypt ? ctx->authsize : (-ctx->authsize));
+	in_len = 4 + ivsize + req->assoclen + req->cryptlen;
+
+	memset(&req_ctx->fd_flt, 0, sizeof(req_ctx->fd_flt));
+	dpaa2_fl_set_final(in_fle, true);
+	dpaa2_fl_set_format(in_fle, dpaa2_fl_sg);
+	dpaa2_fl_set_addr(in_fle, qm_sg_dma);
+	dpaa2_fl_set_len(in_fle, in_len);
+
+	if (req->dst == req->src) {
+		if (mapped_src_nents == 1) {
+			dpaa2_fl_set_format(out_fle, dpaa2_fl_single);
+			dpaa2_fl_set_addr(out_fle, sg_dma_address(req->src));
+		} else {
+			dpaa2_fl_set_format(out_fle, dpaa2_fl_sg);
+			dpaa2_fl_set_addr(out_fle, qm_sg_dma +
+					  (1 + !!ivsize) * sizeof(*sg_table));
+		}
+	} else if (mapped_dst_nents == 1) {
+		dpaa2_fl_set_format(out_fle, dpaa2_fl_single);
+		dpaa2_fl_set_addr(out_fle, sg_dma_address(req->dst));
+	} else {
+		dpaa2_fl_set_format(out_fle, dpaa2_fl_sg);
+		dpaa2_fl_set_addr(out_fle, qm_sg_dma + qm_sg_index *
+				  sizeof(*sg_table));
+	}
+
+	dpaa2_fl_set_len(out_fle, out_len);
+
+	return edesc;
+}
+
+static int gcm_set_sh_desc(struct crypto_aead *aead)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *dev = ctx->dev;
+	unsigned int ivsize = crypto_aead_ivsize(aead);
+	struct caam_flc *flc;
+	u32 *desc;
+	int rem_bytes = CAAM_DESC_BYTES_MAX - DESC_JOB_IO_LEN -
+			ctx->cdata.keylen;
+
+	if (!ctx->cdata.keylen || !ctx->authsize)
+		return 0;
+
+	/*
+	 * AES GCM encrypt shared descriptor
+	 * Job Descriptor and Shared Descriptor
+	 * must fit into the 64-word Descriptor h/w Buffer
+	 */
+	if (rem_bytes >= DESC_QI_GCM_ENC_LEN) {
+		ctx->cdata.key_inline = true;
+		ctx->cdata.key_virt = ctx->key;
+	} else {
+		ctx->cdata.key_inline = false;
+		ctx->cdata.key_dma = ctx->key_dma;
+	}
+
+	flc = &ctx->flc[ENCRYPT];
+	desc = flc->sh_desc;
+	cnstr_shdsc_gcm_encap(desc, &ctx->cdata, ivsize, ctx->authsize, true);
+	flc->flc[1] = cpu_to_caam32(desc_len(desc)); /* SDL */
+	dma_sync_single_for_device(dev, ctx->flc_dma[ENCRYPT],
+				   sizeof(flc->flc) + desc_bytes(desc),
+				   ctx->dir);
+
+	/*
+	 * Job Descriptor and Shared Descriptors
+	 * must all fit into the 64-word Descriptor h/w Buffer
+	 */
+	if (rem_bytes >= DESC_QI_GCM_DEC_LEN) {
+		ctx->cdata.key_inline = true;
+		ctx->cdata.key_virt = ctx->key;
+	} else {
+		ctx->cdata.key_inline = false;
+		ctx->cdata.key_dma = ctx->key_dma;
+	}
+
+	flc = &ctx->flc[DECRYPT];
+	desc = flc->sh_desc;
+	cnstr_shdsc_gcm_decap(desc, &ctx->cdata, ivsize, ctx->authsize, true);
+	flc->flc[1] = cpu_to_caam32(desc_len(desc)); /* SDL */
+	dma_sync_single_for_device(dev, ctx->flc_dma[DECRYPT],
+				   sizeof(flc->flc) + desc_bytes(desc),
+				   ctx->dir);
+
+	return 0;
+}
+
+static int gcm_setauthsize(struct crypto_aead *authenc, unsigned int authsize)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(authenc);
+
+	ctx->authsize = authsize;
+	gcm_set_sh_desc(authenc);
+
+	return 0;
+}
+
+static int gcm_setkey(struct crypto_aead *aead,
+		      const u8 *key, unsigned int keylen)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *dev = ctx->dev;
+
+	print_hex_dump_debug("key in @" __stringify(__LINE__)": ",
+			     DUMP_PREFIX_ADDRESS, 16, 4, key, keylen, 1);
+
+	memcpy(ctx->key, key, keylen);
+	dma_sync_single_for_device(dev, ctx->key_dma, keylen, ctx->dir);
+	ctx->cdata.keylen = keylen;
+
+	return gcm_set_sh_desc(aead);
+}
+
+static int rfc4106_set_sh_desc(struct crypto_aead *aead)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *dev = ctx->dev;
+	unsigned int ivsize = crypto_aead_ivsize(aead);
+	struct caam_flc *flc;
+	u32 *desc;
+	int rem_bytes = CAAM_DESC_BYTES_MAX - DESC_JOB_IO_LEN -
+			ctx->cdata.keylen;
+
+	if (!ctx->cdata.keylen || !ctx->authsize)
+		return 0;
+
+	ctx->cdata.key_virt = ctx->key;
+
+	/*
+	 * RFC4106 encrypt shared descriptor
+	 * Job Descriptor and Shared Descriptor
+	 * must fit into the 64-word Descriptor h/w Buffer
+	 */
+	if (rem_bytes >= DESC_QI_RFC4106_ENC_LEN) {
+		ctx->cdata.key_inline = true;
+	} else {
+		ctx->cdata.key_inline = false;
+		ctx->cdata.key_dma = ctx->key_dma;
+	}
+
+	flc = &ctx->flc[ENCRYPT];
+	desc = flc->sh_desc;
+	cnstr_shdsc_rfc4106_encap(desc, &ctx->cdata, ivsize, ctx->authsize,
+				  true);
+	flc->flc[1] = cpu_to_caam32(desc_len(desc)); /* SDL */
+	dma_sync_single_for_device(dev, ctx->flc_dma[ENCRYPT],
+				   sizeof(flc->flc) + desc_bytes(desc),
+				   ctx->dir);
+
+	/*
+	 * Job Descriptor and Shared Descriptors
+	 * must all fit into the 64-word Descriptor h/w Buffer
+	 */
+	if (rem_bytes >= DESC_QI_RFC4106_DEC_LEN) {
+		ctx->cdata.key_inline = true;
+	} else {
+		ctx->cdata.key_inline = false;
+		ctx->cdata.key_dma = ctx->key_dma;
+	}
+
+	flc = &ctx->flc[DECRYPT];
+	desc = flc->sh_desc;
+	cnstr_shdsc_rfc4106_decap(desc, &ctx->cdata, ivsize, ctx->authsize,
+				  true);
+	flc->flc[1] = cpu_to_caam32(desc_len(desc)); /* SDL */
+	dma_sync_single_for_device(dev, ctx->flc_dma[DECRYPT],
+				   sizeof(flc->flc) + desc_bytes(desc),
+				   ctx->dir);
+
+	return 0;
+}
+
+static int rfc4106_setauthsize(struct crypto_aead *authenc,
+			       unsigned int authsize)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(authenc);
+
+	ctx->authsize = authsize;
+	rfc4106_set_sh_desc(authenc);
+
+	return 0;
+}
+
+static int rfc4106_setkey(struct crypto_aead *aead,
+			  const u8 *key, unsigned int keylen)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *dev = ctx->dev;
+
+	if (keylen < 4)
+		return -EINVAL;
+
+	print_hex_dump_debug("key in @" __stringify(__LINE__)": ",
+			     DUMP_PREFIX_ADDRESS, 16, 4, key, keylen, 1);
+
+	memcpy(ctx->key, key, keylen);
+	/*
+	 * The last four bytes of the key material are used as the salt value
+	 * in the nonce. Update the AES key length.
+	 */
+	ctx->cdata.keylen = keylen - 4;
+	dma_sync_single_for_device(dev, ctx->key_dma, ctx->cdata.keylen,
+				   ctx->dir);
+
+	return rfc4106_set_sh_desc(aead);
+}
+
+static int rfc4543_set_sh_desc(struct crypto_aead *aead)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *dev = ctx->dev;
+	unsigned int ivsize = crypto_aead_ivsize(aead);
+	struct caam_flc *flc;
+	u32 *desc;
+	int rem_bytes = CAAM_DESC_BYTES_MAX - DESC_JOB_IO_LEN -
+			ctx->cdata.keylen;
+
+	if (!ctx->cdata.keylen || !ctx->authsize)
+		return 0;
+
+	ctx->cdata.key_virt = ctx->key;
+
+	/*
+	 * RFC4543 encrypt shared descriptor
+	 * Job Descriptor and Shared Descriptor
+	 * must fit into the 64-word Descriptor h/w Buffer
+	 */
+	if (rem_bytes >= DESC_QI_RFC4543_ENC_LEN) {
+		ctx->cdata.key_inline = true;
+	} else {
+		ctx->cdata.key_inline = false;
+		ctx->cdata.key_dma = ctx->key_dma;
+	}
+
+	flc = &ctx->flc[ENCRYPT];
+	desc = flc->sh_desc;
+	cnstr_shdsc_rfc4543_encap(desc, &ctx->cdata, ivsize, ctx->authsize,
+				  true);
+	flc->flc[1] = cpu_to_caam32(desc_len(desc)); /* SDL */
+	dma_sync_single_for_device(dev, ctx->flc_dma[ENCRYPT],
+				   sizeof(flc->flc) + desc_bytes(desc),
+				   ctx->dir);
+
+	/*
+	 * Job Descriptor and Shared Descriptors
+	 * must all fit into the 64-word Descriptor h/w Buffer
+	 */
+	if (rem_bytes >= DESC_QI_RFC4543_DEC_LEN) {
+		ctx->cdata.key_inline = true;
+	} else {
+		ctx->cdata.key_inline = false;
+		ctx->cdata.key_dma = ctx->key_dma;
+	}
+
+	flc = &ctx->flc[DECRYPT];
+	desc = flc->sh_desc;
+	cnstr_shdsc_rfc4543_decap(desc, &ctx->cdata, ivsize, ctx->authsize,
+				  true);
+	flc->flc[1] = cpu_to_caam32(desc_len(desc)); /* SDL */
+	dma_sync_single_for_device(dev, ctx->flc_dma[DECRYPT],
+				   sizeof(flc->flc) + desc_bytes(desc),
+				   ctx->dir);
+
+	return 0;
+}
+
+static int rfc4543_setauthsize(struct crypto_aead *authenc,
+			       unsigned int authsize)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(authenc);
+
+	ctx->authsize = authsize;
+	rfc4543_set_sh_desc(authenc);
+
+	return 0;
+}
+
+static int rfc4543_setkey(struct crypto_aead *aead,
+			  const u8 *key, unsigned int keylen)
+{
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct device *dev = ctx->dev;
+
+	if (keylen < 4)
+		return -EINVAL;
+
+	print_hex_dump_debug("key in @" __stringify(__LINE__)": ",
+			     DUMP_PREFIX_ADDRESS, 16, 4, key, keylen, 1);
+
+	memcpy(ctx->key, key, keylen);
+	/*
+	 * The last four bytes of the key material are used as the salt value
+	 * in the nonce. Update the AES key length.
+	 */
+	ctx->cdata.keylen = keylen - 4;
+	dma_sync_single_for_device(dev, ctx->key_dma, ctx->cdata.keylen,
+				   ctx->dir);
+
+	return rfc4543_set_sh_desc(aead);
+}
+
+static void aead_unmap(struct device *dev, struct aead_edesc *edesc,
+		       struct aead_request *req)
+{
+	struct crypto_aead *aead = crypto_aead_reqtfm(req);
+	int ivsize = crypto_aead_ivsize(aead);
+
+	caam_unmap(dev, req->src, req->dst, edesc->src_nents, edesc->dst_nents,
+		   edesc->iv_dma, ivsize, edesc->qm_sg_dma, edesc->qm_sg_bytes);
+	dma_unmap_single(dev, edesc->assoclen_dma, 4, DMA_TO_DEVICE);
+}
+
+static void aead_encrypt_done(void *cbk_ctx, u32 status)
+{
+	struct crypto_async_request *areq = cbk_ctx;
+	struct aead_request *req = container_of(areq, struct aead_request,
+						base);
+	struct caam_request *req_ctx = to_caam_req(areq);
+	struct aead_edesc *edesc = req_ctx->edesc;
+	struct crypto_aead *aead = crypto_aead_reqtfm(req);
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	int ecode = 0;
+
+	dev_dbg(ctx->dev, "%s %d: err 0x%x\n", __func__, __LINE__, status);
+
+	if (unlikely(status)) {
+		caam_qi2_strstatus(ctx->dev, status);
+		ecode = -EIO;
+	}
+
+	aead_unmap(ctx->dev, edesc, req);
+	qi_cache_free(edesc);
+	aead_request_complete(req, ecode);
+}
+
+static void aead_decrypt_done(void *cbk_ctx, u32 status)
+{
+	struct crypto_async_request *areq = cbk_ctx;
+	struct aead_request *req = container_of(areq, struct aead_request,
+						base);
+	struct caam_request *req_ctx = to_caam_req(areq);
+	struct aead_edesc *edesc = req_ctx->edesc;
+	struct crypto_aead *aead = crypto_aead_reqtfm(req);
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	int ecode = 0;
+
+	dev_dbg(ctx->dev, "%s %d: err 0x%x\n", __func__, __LINE__, status);
+
+	if (unlikely(status)) {
+		caam_qi2_strstatus(ctx->dev, status);
+		/*
+		 * verify hw auth check passed else return -EBADMSG
+		 */
+		if ((status & JRSTA_CCBERR_ERRID_MASK) ==
+		     JRSTA_CCBERR_ERRID_ICVCHK)
+			ecode = -EBADMSG;
+		else
+			ecode = -EIO;
+	}
+
+	aead_unmap(ctx->dev, edesc, req);
+	qi_cache_free(edesc);
+	aead_request_complete(req, ecode);
+}
+
+static int aead_encrypt(struct aead_request *req)
+{
+	struct aead_edesc *edesc;
+	struct crypto_aead *aead = crypto_aead_reqtfm(req);
+	struct caam_ctx *ctx = crypto_aead_ctx(aead);
+	struct caam_request *caam_req = aead_request_ctx(req);
+	int ret;
+
+	/* allocate extended descriptor */
+	edesc = aead_edesc_alloc(req, true);
+	if (IS_ERR(edesc))
+		return PTR_ERR(edesc);
+
+	caam_req->flc = &ctx->flc[ENCRYPT];
+	caam_req->flc_dma = ctx->flc_dma[ENCRYPT];
+	caam_req->cbk = aead_encrypt_done;
+	caam_req->ctx = &req->base;
+	caam_req->edesc = edesc;
+	ret = dpaa2_caam_enqueue(ctx->dev, caam_req)