Merge git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6

Pull crypto update from Herbert Xu:
 "Here is the crypto update for 4.1:

  New interfaces:
   - user-space interface for AEAD
   - user-space interface for RNG (i.e., pseudo RNG)

  New hashes:
   - ARMv8 SHA1/256
   - ARMv8 AES
   - ARMv8 GHASH
   - ARM assembler and NEON SHA256
   - MIPS OCTEON SHA1/256/512
   - MIPS img-hash SHA1/256 and MD5
   - Power 8 VMX AES/CBC/CTR/GHASH
   - PPC assembler AES, SHA1/256 and MD5
   - Broadcom IPROC RNG driver

  Cleanups/fixes:
   - prevent internal helper algos from being exposed to user-space
   - merge common code from assembly/C SHA implementations
   - misc fixes"

* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: (169 commits)
  crypto: arm - workaround for building with old binutils
  crypto: arm/sha256 - avoid sha256 code on ARMv7-M
  crypto: x86/sha512_ssse3 - move SHA-384/512 SSSE3 implementation to base layer
  crypto: x86/sha256_ssse3 - move SHA-224/256 SSSE3 implementation to base layer
  crypto: x86/sha1_ssse3 - move SHA-1 SSSE3 implementation to base layer
  crypto: arm64/sha2-ce - move SHA-224/256 ARMv8 implementation to base layer
  crypto: arm64/sha1-ce - move SHA-1 ARMv8 implementation to base layer
  crypto: arm/sha2-ce - move SHA-224/256 ARMv8 implementation to base layer
  crypto: arm/sha256 - move SHA-224/256 ASM/NEON implementation to base layer
  crypto: arm/sha1-ce - move SHA-1 ARMv8 implementation to base layer
  crypto: arm/sha1_neon - move SHA-1 NEON implementation to base layer
  crypto: arm/sha1 - move SHA-1 ARM asm implementation to base layer
  crypto: sha512-generic - move to generic glue implementation
  crypto: sha256-generic - move to generic glue implementation
  crypto: sha1-generic - move to generic glue implementation
  crypto: sha512 - implement base layer for SHA-512
  crypto: sha256 - implement base layer for SHA-256
  crypto: sha1 - implement base layer for SHA-1
  crypto: api - remove instance when test failed
  crypto: api - Move alg ref count init to crypto_check_alg
  ...

22 files changed, 335 insertions, 567 deletions

diff --git a/arch/x86/crypto/aesni-intel_glue.c b/arch/x86/crypto/aesni-intel_glue.c
index 54f60ab..112cefa 100644
--- a/arch/x86/crypto/aesni-intel_glue.c
+++ b/arch/x86/crypto/aesni-intel_glue.c
@@ -797,7 +797,9 @@ static int rfc4106_init(struct crypto_tfm *tfm)
 		PTR_ALIGN((u8 *)crypto_tfm_ctx(tfm), AESNI_ALIGN);
 	struct crypto_aead *cryptd_child;
 	struct aesni_rfc4106_gcm_ctx *child_ctx;
-	cryptd_tfm = cryptd_alloc_aead("__driver-gcm-aes-aesni", 0, 0);
+	cryptd_tfm = cryptd_alloc_aead("__driver-gcm-aes-aesni",
+				       CRYPTO_ALG_INTERNAL,
+				       CRYPTO_ALG_INTERNAL);
 	if (IS_ERR(cryptd_tfm))
 		return PTR_ERR(cryptd_tfm);
 
@@ -890,15 +892,12 @@ out_free_ablkcipher:
 	return ret;
 }
 
-static int rfc4106_set_key(struct crypto_aead *parent, const u8 *key,
-						   unsigned int key_len)
+static int common_rfc4106_set_key(struct crypto_aead *aead, const u8 *key,
+				  unsigned int key_len)
 {
 	int ret = 0;
-	struct crypto_tfm *tfm = crypto_aead_tfm(parent);
-	struct aesni_rfc4106_gcm_ctx *ctx = aesni_rfc4106_gcm_ctx_get(parent);
-	struct crypto_aead *cryptd_child = cryptd_aead_child(ctx->cryptd_tfm);
-	struct aesni_rfc4106_gcm_ctx *child_ctx =
-                                 aesni_rfc4106_gcm_ctx_get(cryptd_child);
+	struct crypto_tfm *tfm = crypto_aead_tfm(aead);
+	struct aesni_rfc4106_gcm_ctx *ctx = aesni_rfc4106_gcm_ctx_get(aead);
 	u8 *new_key_align, *new_key_mem = NULL;
 
 	if (key_len < 4) {
@@ -943,20 +942,31 @@ static int rfc4106_set_key(struct crypto_aead *parent, const u8 *key,
 		goto exit;
 	}
 	ret = rfc4106_set_hash_subkey(ctx->hash_subkey, key, key_len);
-	memcpy(child_ctx, ctx, sizeof(*ctx));
 exit:
 	kfree(new_key_mem);
 	return ret;
 }
 
-/* This is the Integrity Check Value (aka the authentication tag length and can
- * be 8, 12 or 16 bytes long. */
-static int rfc4106_set_authsize(struct crypto_aead *parent,
-				unsigned int authsize)
+static int rfc4106_set_key(struct crypto_aead *parent, const u8 *key,
+			   unsigned int key_len)
 {
 	struct aesni_rfc4106_gcm_ctx *ctx = aesni_rfc4106_gcm_ctx_get(parent);
-	struct crypto_aead *cryptd_child = cryptd_aead_child(ctx->cryptd_tfm);
+	struct crypto_aead *child = cryptd_aead_child(ctx->cryptd_tfm);
+	struct aesni_rfc4106_gcm_ctx *c_ctx = aesni_rfc4106_gcm_ctx_get(child);
+	struct cryptd_aead *cryptd_tfm = ctx->cryptd_tfm;
+	int ret;
 
+	ret = crypto_aead_setkey(child, key, key_len);
+	if (!ret) {
+		memcpy(ctx, c_ctx, sizeof(*ctx));
+		ctx->cryptd_tfm = cryptd_tfm;
+	}
+	return ret;
+}
+
+static int common_rfc4106_set_authsize(struct crypto_aead *aead,
+				       unsigned int authsize)
+{
 	switch (authsize) {
 	case 8:
 	case 12:
@@ -965,51 +975,23 @@ static int rfc4106_set_authsize(struct crypto_aead *parent,
 	default:
 		return -EINVAL;
 	}
-	crypto_aead_crt(parent)->authsize = authsize;
-	crypto_aead_crt(cryptd_child)->authsize = authsize;
+	crypto_aead_crt(aead)->authsize = authsize;
 	return 0;
 }
 
-static int rfc4106_encrypt(struct aead_request *req)
-{
-	int ret;
-	struct crypto_aead *tfm = crypto_aead_reqtfm(req);
-	struct aesni_rfc4106_gcm_ctx *ctx = aesni_rfc4106_gcm_ctx_get(tfm);
-
-	if (!irq_fpu_usable()) {
-		struct aead_request *cryptd_req =
-			(struct aead_request *) aead_request_ctx(req);
-		memcpy(cryptd_req, req, sizeof(*req));
-		aead_request_set_tfm(cryptd_req, &ctx->cryptd_tfm->base);
-		return crypto_aead_encrypt(cryptd_req);
-	} else {
-		struct crypto_aead *cryptd_child = cryptd_aead_child(ctx->cryptd_tfm);
-		kernel_fpu_begin();
-		ret = cryptd_child->base.crt_aead.encrypt(req);
-		kernel_fpu_end();
-		return ret;
-	}
-}
-
-static int rfc4106_decrypt(struct aead_request *req)
+/* This is the Integrity Check Value (aka the authentication tag length and can
+ * be 8, 12 or 16 bytes long. */
+static int rfc4106_set_authsize(struct crypto_aead *parent,
+				unsigned int authsize)
 {
+	struct aesni_rfc4106_gcm_ctx *ctx = aesni_rfc4106_gcm_ctx_get(parent);
+	struct crypto_aead *child = cryptd_aead_child(ctx->cryptd_tfm);
 	int ret;
-	struct crypto_aead *tfm = crypto_aead_reqtfm(req);
-	struct aesni_rfc4106_gcm_ctx *ctx = aesni_rfc4106_gcm_ctx_get(tfm);
 
-	if (!irq_fpu_usable()) {
-		struct aead_request *cryptd_req =
-			(struct aead_request *) aead_request_ctx(req);
-		memcpy(cryptd_req, req, sizeof(*req));
-		aead_request_set_tfm(cryptd_req, &ctx->cryptd_tfm->base);
-		return crypto_aead_decrypt(cryptd_req);
-	} else {
-		struct crypto_aead *cryptd_child = cryptd_aead_child(ctx->cryptd_tfm);
-		kernel_fpu_begin();
-		ret = cryptd_child->base.crt_aead.decrypt(req);
-		kernel_fpu_end();
-		return ret;
-	}
+	ret = crypto_aead_setauthsize(child, authsize);
+	if (!ret)
+		crypto_aead_crt(parent)->authsize = authsize;
+	return ret;
 }
 
 static int __driver_rfc4106_encrypt(struct aead_request *req)
@@ -1185,6 +1167,78 @@ static int __driver_rfc4106_decrypt(struct aead_request *req)
 	}
 	return retval;
 }
+
+static int rfc4106_encrypt(struct aead_request *req)
+{
+	int ret;
+	struct crypto_aead *tfm = crypto_aead_reqtfm(req);
+	struct aesni_rfc4106_gcm_ctx *ctx = aesni_rfc4106_gcm_ctx_get(tfm);
+
+	if (!irq_fpu_usable()) {
+		struct aead_request *cryptd_req =
+			(struct aead_request *) aead_request_ctx(req);
+
+		memcpy(cryptd_req, req, sizeof(*req));
+		aead_request_set_tfm(cryptd_req, &ctx->cryptd_tfm->base);
+		ret = crypto_aead_encrypt(cryptd_req);
+	} else {
+		kernel_fpu_begin();
+		ret = __driver_rfc4106_encrypt(req);
+		kernel_fpu_end();
+	}
+	return ret;
+}
+
+static int rfc4106_decrypt(struct aead_request *req)
+{
+	int ret;
+	struct crypto_aead *tfm = crypto_aead_reqtfm(req);
+	struct aesni_rfc4106_gcm_ctx *ctx = aesni_rfc4106_gcm_ctx_get(tfm);
+
+	if (!irq_fpu_usable()) {
+		struct aead_request *cryptd_req =
+			(struct aead_request *) aead_request_ctx(req);
+
+		memcpy(cryptd_req, req, sizeof(*req));
+		aead_request_set_tfm(cryptd_req, &ctx->cryptd_tfm->base);
+		ret = crypto_aead_decrypt(cryptd_req);
+	} else {
+		kernel_fpu_begin();
+		ret = __driver_rfc4106_decrypt(req);
+		kernel_fpu_end();
+	}
+	return ret;
+}
+
+static int helper_rfc4106_encrypt(struct aead_request *req)
+{
+	int ret;
+
+	if (unlikely(!irq_fpu_usable())) {
+		WARN_ONCE(1, "__gcm-aes-aesni alg used in invalid context");
+		ret = -EINVAL;
+	} else {
+		kernel_fpu_begin();
+		ret = __driver_rfc4106_encrypt(req);
+		kernel_fpu_end();
+	}
+	return ret;
+}
+
+static int helper_rfc4106_decrypt(struct aead_request *req)
+{
+	int ret;
+
+	if (unlikely(!irq_fpu_usable())) {
+		WARN_ONCE(1, "__gcm-aes-aesni alg used in invalid context");
+		ret = -EINVAL;
+	} else {
+		kernel_fpu_begin();
+		ret = __driver_rfc4106_decrypt(req);
+		kernel_fpu_end();
+	}
+	return ret;
+}
 #endif
 
 static struct crypto_alg aesni_algs[] = { {
@@ -1210,7 +1264,7 @@ static struct crypto_alg aesni_algs[] = { {
 	.cra_name		= "__aes-aesni",
 	.cra_driver_name	= "__driver-aes-aesni",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_CIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_CIPHER | CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= AES_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct crypto_aes_ctx) +
 				  AESNI_ALIGN - 1,
@@ -1229,7 +1283,8 @@ static struct crypto_alg aesni_algs[] = { {
 	.cra_name		= "__ecb-aes-aesni",
 	.cra_driver_name	= "__driver-ecb-aes-aesni",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= AES_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct crypto_aes_ctx) +
 				  AESNI_ALIGN - 1,
@@ -1249,7 +1304,8 @@ static struct crypto_alg aesni_algs[] = { {
 	.cra_name		= "__cbc-aes-aesni",
 	.cra_driver_name	= "__driver-cbc-aes-aesni",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= AES_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct crypto_aes_ctx) +
 				  AESNI_ALIGN - 1,
@@ -1313,7 +1369,8 @@ static struct crypto_alg aesni_algs[] = { {
 	.cra_name		= "__ctr-aes-aesni",
 	.cra_driver_name	= "__driver-ctr-aes-aesni",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= 1,
 	.cra_ctxsize		= sizeof(struct crypto_aes_ctx) +
 				  AESNI_ALIGN - 1,
@@ -1357,7 +1414,7 @@ static struct crypto_alg aesni_algs[] = { {
 	.cra_name		= "__gcm-aes-aesni",
 	.cra_driver_name	= "__driver-gcm-aes-aesni",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_AEAD,
+	.cra_flags		= CRYPTO_ALG_TYPE_AEAD | CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= 1,
 	.cra_ctxsize		= sizeof(struct aesni_rfc4106_gcm_ctx) +
 				  AESNI_ALIGN,
@@ -1366,8 +1423,12 @@ static struct crypto_alg aesni_algs[] = { {
 	.cra_module		= THIS_MODULE,
 	.cra_u = {
 		.aead = {
-			.encrypt	= __driver_rfc4106_encrypt,
-			.decrypt	= __driver_rfc4106_decrypt,
+			.setkey		= common_rfc4106_set_key,
+			.setauthsize	= common_rfc4106_set_authsize,
+			.encrypt	= helper_rfc4106_encrypt,
+			.decrypt	= helper_rfc4106_decrypt,
+			.ivsize		= 8,
+			.maxauthsize	= 16,
 		},
 	},
 }, {
@@ -1423,7 +1484,8 @@ static struct crypto_alg aesni_algs[] = { {
 	.cra_name		= "__lrw-aes-aesni",
 	.cra_driver_name	= "__driver-lrw-aes-aesni",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= AES_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct aesni_lrw_ctx),
 	.cra_alignmask		= 0,
@@ -1444,7 +1506,8 @@ static struct crypto_alg aesni_algs[] = { {
 	.cra_name		= "__xts-aes-aesni",
 	.cra_driver_name	= "__driver-xts-aes-aesni",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= AES_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct aesni_xts_ctx),
 	.cra_alignmask		= 0,
diff --git a/arch/x86/crypto/camellia_aesni_avx2_glue.c b/arch/x86/crypto/camellia_aesni_avx2_glue.c
index 9a07faf..baf0ac2 100644
--- a/arch/x86/crypto/camellia_aesni_avx2_glue.c
+++ b/arch/x86/crypto/camellia_aesni_avx2_glue.c
@@ -343,7 +343,8 @@ static struct crypto_alg cmll_algs[10] = { {
 	.cra_name		= "__ecb-camellia-aesni-avx2",
 	.cra_driver_name	= "__driver-ecb-camellia-aesni-avx2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct camellia_ctx),
 	.cra_alignmask		= 0,
@@ -362,7 +363,8 @@ static struct crypto_alg cmll_algs[10] = { {
 	.cra_name		= "__cbc-camellia-aesni-avx2",
 	.cra_driver_name	= "__driver-cbc-camellia-aesni-avx2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct camellia_ctx),
 	.cra_alignmask		= 0,
@@ -381,7 +383,8 @@ static struct crypto_alg cmll_algs[10] = { {
 	.cra_name		= "__ctr-camellia-aesni-avx2",
 	.cra_driver_name	= "__driver-ctr-camellia-aesni-avx2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= 1,
 	.cra_ctxsize		= sizeof(struct camellia_ctx),
 	.cra_alignmask		= 0,
@@ -401,7 +404,8 @@ static struct crypto_alg cmll_algs[10] = { {
 	.cra_name		= "__lrw-camellia-aesni-avx2",
 	.cra_driver_name	= "__driver-lrw-camellia-aesni-avx2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct camellia_lrw_ctx),
 	.cra_alignmask		= 0,
@@ -424,7 +428,8 @@ static struct crypto_alg cmll_algs[10] = { {
 	.cra_name		= "__xts-camellia-aesni-avx2",
 	.cra_driver_name	= "__driver-xts-camellia-aesni-avx2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct camellia_xts_ctx),
 	.cra_alignmask		= 0,
diff --git a/arch/x86/crypto/camellia_aesni_avx_glue.c b/arch/x86/crypto/camellia_aesni_avx_glue.c
index ed38d95..78818a1 100644
--- a/arch/x86/crypto/camellia_aesni_avx_glue.c
+++ b/arch/x86/crypto/camellia_aesni_avx_glue.c
@@ -335,7 +335,8 @@ static struct crypto_alg cmll_algs[10] = { {
 	.cra_name		= "__ecb-camellia-aesni",
 	.cra_driver_name	= "__driver-ecb-camellia-aesni",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct camellia_ctx),
 	.cra_alignmask		= 0,
@@ -354,7 +355,8 @@ static struct crypto_alg cmll_algs[10] = { {
 	.cra_name		= "__cbc-camellia-aesni",
 	.cra_driver_name	= "__driver-cbc-camellia-aesni",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct camellia_ctx),
 	.cra_alignmask		= 0,
@@ -373,7 +375,8 @@ static struct crypto_alg cmll_algs[10] = { {
 	.cra_name		= "__ctr-camellia-aesni",
 	.cra_driver_name	= "__driver-ctr-camellia-aesni",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= 1,
 	.cra_ctxsize		= sizeof(struct camellia_ctx),
 	.cra_alignmask		= 0,
@@ -393,7 +396,8 @@ static struct crypto_alg cmll_algs[10] = { {
 	.cra_name		= "__lrw-camellia-aesni",
 	.cra_driver_name	= "__driver-lrw-camellia-aesni",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct camellia_lrw_ctx),
 	.cra_alignmask		= 0,
@@ -416,7 +420,8 @@ static struct crypto_alg cmll_algs[10] = { {
 	.cra_name		= "__xts-camellia-aesni",
 	.cra_driver_name	= "__driver-xts-camellia-aesni",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct camellia_xts_ctx),
 	.cra_alignmask		= 0,
diff --git a/arch/x86/crypto/cast5_avx_glue.c b/arch/x86/crypto/cast5_avx_glue.c
index 60ada67..236c809 100644
--- a/arch/x86/crypto/cast5_avx_glue.c
+++ b/arch/x86/crypto/cast5_avx_glue.c
@@ -341,7 +341,8 @@ static struct crypto_alg cast5_algs[6] = { {
 	.cra_name		= "__ecb-cast5-avx",
 	.cra_driver_name	= "__driver-ecb-cast5-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAST5_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct cast5_ctx),
 	.cra_alignmask		= 0,
@@ -360,7 +361,8 @@ static struct crypto_alg cast5_algs[6] = { {
 	.cra_name		= "__cbc-cast5-avx",
 	.cra_driver_name	= "__driver-cbc-cast5-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAST5_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct cast5_ctx),
 	.cra_alignmask		= 0,
@@ -379,7 +381,8 @@ static struct crypto_alg cast5_algs[6] = { {
 	.cra_name		= "__ctr-cast5-avx",
 	.cra_driver_name	= "__driver-ctr-cast5-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= 1,
 	.cra_ctxsize		= sizeof(struct cast5_ctx),
 	.cra_alignmask		= 0,
diff --git a/arch/x86/crypto/cast6_avx_glue.c b/arch/x86/crypto/cast6_avx_glue.c
index 0160f68..f448810 100644
--- a/arch/x86/crypto/cast6_avx_glue.c
+++ b/arch/x86/crypto/cast6_avx_glue.c
@@ -372,7 +372,8 @@ static struct crypto_alg cast6_algs[10] = { {
 	.cra_name		= "__ecb-cast6-avx",
 	.cra_driver_name	= "__driver-ecb-cast6-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAST6_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct cast6_ctx),
 	.cra_alignmask		= 0,
@@ -391,7 +392,8 @@ static struct crypto_alg cast6_algs[10] = { {
 	.cra_name		= "__cbc-cast6-avx",
 	.cra_driver_name	= "__driver-cbc-cast6-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAST6_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct cast6_ctx),
 	.cra_alignmask		= 0,
@@ -410,7 +412,8 @@ static struct crypto_alg cast6_algs[10] = { {
 	.cra_name		= "__ctr-cast6-avx",
 	.cra_driver_name	= "__driver-ctr-cast6-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= 1,
 	.cra_ctxsize		= sizeof(struct cast6_ctx),
 	.cra_alignmask		= 0,
@@ -430,7 +433,8 @@ static struct crypto_alg cast6_algs[10] = { {
 	.cra_name		= "__lrw-cast6-avx",
 	.cra_driver_name	= "__driver-lrw-cast6-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAST6_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct cast6_lrw_ctx),
 	.cra_alignmask		= 0,
@@ -453,7 +457,8 @@ static struct crypto_alg cast6_algs[10] = { {
 	.cra_name		= "__xts-cast6-avx",
 	.cra_driver_name	= "__driver-xts-cast6-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= CAST6_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct cast6_xts_ctx),
 	.cra_alignmask		= 0,
diff --git a/arch/x86/crypto/ghash-clmulni-intel_glue.c b/arch/x86/crypto/ghash-clmulni-intel_glue.c
index 8253d85..2079baf 100644
--- a/arch/x86/crypto/ghash-clmulni-intel_glue.c
+++ b/arch/x86/crypto/ghash-clmulni-intel_glue.c
@@ -154,7 +154,8 @@ static struct shash_alg ghash_alg = {
 		.cra_name		= "__ghash",
 		.cra_driver_name	= "__ghash-pclmulqdqni",
 		.cra_priority		= 0,
-		.cra_flags		= CRYPTO_ALG_TYPE_SHASH,
+		.cra_flags		= CRYPTO_ALG_TYPE_SHASH |
+					  CRYPTO_ALG_INTERNAL,
 		.cra_blocksize		= GHASH_BLOCK_SIZE,
 		.cra_ctxsize		= sizeof(struct ghash_ctx),
 		.cra_module		= THIS_MODULE,
@@ -261,7 +262,9 @@ static int ghash_async_init_tfm(struct crypto_tfm *tfm)
 	struct cryptd_ahash *cryptd_tfm;
 	struct ghash_async_ctx *ctx = crypto_tfm_ctx(tfm);
 
-	cryptd_tfm = cryptd_alloc_ahash("__ghash-pclmulqdqni", 0, 0);
+	cryptd_tfm = cryptd_alloc_ahash("__ghash-pclmulqdqni",
+					CRYPTO_ALG_INTERNAL,
+					CRYPTO_ALG_INTERNAL);
 	if (IS_ERR(cryptd_tfm))
 		return PTR_ERR(cryptd_tfm);
 	ctx->cryptd_tfm = cryptd_tfm;
diff --git a/arch/x86/crypto/glue_helper.c b/arch/x86/crypto/glue_helper.c
index 432f1d76..6a85598 100644
--- a/arch/x86/crypto/glue_helper.c
+++ b/arch/x86/crypto/glue_helper.c
@@ -232,7 +232,6 @@ static void glue_ctr_crypt_final_128bit(const common_glue_ctr_func_t fn_ctr,
 
 	le128_to_be128((be128 *)walk->iv, &ctrblk);
 }
-EXPORT_SYMBOL_GPL(glue_ctr_crypt_final_128bit);
 
 static unsigned int __glue_ctr_crypt_128bit(const struct common_glue_ctx *gctx,
 					    struct blkcipher_desc *desc,
diff --git a/arch/x86/crypto/serpent_avx2_glue.c b/arch/x86/crypto/serpent_avx2_glue.c
index 437e47a..2f63dc8 100644
--- a/arch/x86/crypto/serpent_avx2_glue.c
+++ b/arch/x86/crypto/serpent_avx2_glue.c
@@ -309,7 +309,8 @@ static struct crypto_alg srp_algs[10] = { {
 	.cra_name		= "__ecb-serpent-avx2",
 	.cra_driver_name	= "__driver-ecb-serpent-avx2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= SERPENT_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct serpent_ctx),
 	.cra_alignmask		= 0,
@@ -329,7 +330,8 @@ static struct crypto_alg srp_algs[10] = { {
 	.cra_name		= "__cbc-serpent-avx2",
 	.cra_driver_name	= "__driver-cbc-serpent-avx2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= SERPENT_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct serpent_ctx),
 	.cra_alignmask		= 0,
@@ -349,7 +351,8 @@ static struct crypto_alg srp_algs[10] = { {
 	.cra_name		= "__ctr-serpent-avx2",
 	.cra_driver_name	= "__driver-ctr-serpent-avx2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= 1,
 	.cra_ctxsize		= sizeof(struct serpent_ctx),
 	.cra_alignmask		= 0,
@@ -370,7 +373,8 @@ static struct crypto_alg srp_algs[10] = { {
 	.cra_name		= "__lrw-serpent-avx2",
 	.cra_driver_name	= "__driver-lrw-serpent-avx2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= SERPENT_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct serpent_lrw_ctx),
 	.cra_alignmask		= 0,
@@ -394,7 +398,8 @@ static struct crypto_alg srp_algs[10] = { {
 	.cra_name		= "__xts-serpent-avx2",
 	.cra_driver_name	= "__driver-xts-serpent-avx2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= SERPENT_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct serpent_xts_ctx),
 	.cra_alignmask		= 0,
diff --git a/arch/x86/crypto/serpent_avx_glue.c b/arch/x86/crypto/serpent_avx_glue.c
index 7e21739..c8d478a 100644
--- a/arch/x86/crypto/serpent_avx_glue.c
+++ b/arch/x86/crypto/serpent_avx_glue.c
@@ -378,7 +378,8 @@ static struct crypto_alg serpent_algs[10] = { {
 	.cra_name		= "__ecb-serpent-avx",
 	.cra_driver_name	= "__driver-ecb-serpent-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= SERPENT_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct serpent_ctx),
 	.cra_alignmask		= 0,
@@ -397,7 +398,8 @@ static struct crypto_alg serpent_algs[10] = { {
 	.cra_name		= "__cbc-serpent-avx",
 	.cra_driver_name	= "__driver-cbc-serpent-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= SERPENT_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct serpent_ctx),
 	.cra_alignmask		= 0,
@@ -416,7 +418,8 @@ static struct crypto_alg serpent_algs[10] = { {
 	.cra_name		= "__ctr-serpent-avx",
 	.cra_driver_name	= "__driver-ctr-serpent-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= 1,
 	.cra_ctxsize		= sizeof(struct serpent_ctx),
 	.cra_alignmask		= 0,
@@ -436,7 +439,8 @@ static struct crypto_alg serpent_algs[10] = { {
 	.cra_name		= "__lrw-serpent-avx",
 	.cra_driver_name	= "__driver-lrw-serpent-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= SERPENT_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct serpent_lrw_ctx),
 	.cra_alignmask		= 0,
@@ -459,7 +463,8 @@ static struct crypto_alg serpent_algs[10] = { {
 	.cra_name		= "__xts-serpent-avx",
 	.cra_driver_name	= "__driver-xts-serpent-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= SERPENT_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct serpent_xts_ctx),
 	.cra_alignmask		= 0,
diff --git a/arch/x86/crypto/serpent_sse2_glue.c b/arch/x86/crypto/serpent_sse2_glue.c
index bf025ad..3643dd5 100644
--- a/arch/x86/crypto/serpent_sse2_glue.c
+++ b/arch/x86/crypto/serpent_sse2_glue.c
@@ -387,7 +387,8 @@ static struct crypto_alg serpent_algs[10] = { {
 	.cra_name		= "__ecb-serpent-sse2",
 	.cra_driver_name	= "__driver-ecb-serpent-sse2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= SERPENT_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct serpent_ctx),
 	.cra_alignmask		= 0,
@@ -406,7 +407,8 @@ static struct crypto_alg serpent_algs[10] = { {
 	.cra_name		= "__cbc-serpent-sse2",
 	.cra_driver_name	= "__driver-cbc-serpent-sse2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= SERPENT_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct serpent_ctx),
 	.cra_alignmask		= 0,
@@ -425,7 +427,8 @@ static struct crypto_alg serpent_algs[10] = { {
 	.cra_name		= "__ctr-serpent-sse2",
 	.cra_driver_name	= "__driver-ctr-serpent-sse2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= 1,
 	.cra_ctxsize		= sizeof(struct serpent_ctx),
 	.cra_alignmask		= 0,
@@ -445,7 +448,8 @@ static struct crypto_alg serpent_algs[10] = { {
 	.cra_name		= "__lrw-serpent-sse2",
 	.cra_driver_name	= "__driver-lrw-serpent-sse2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= SERPENT_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct serpent_lrw_ctx),
 	.cra_alignmask		= 0,
@@ -468,7 +472,8 @@ static struct crypto_alg serpent_algs[10] = { {
 	.cra_name		= "__xts-serpent-sse2",
 	.cra_driver_name	= "__driver-xts-serpent-sse2",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= SERPENT_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct serpent_xts_ctx),
 	.cra_alignmask		= 0,
diff --git a/arch/x86/crypto/sha-mb/sha1_mb.c b/arch/x86/crypto/sha-mb/sha1_mb.c
index fd9f6b0..e510b1c 100644
--- a/arch/x86/crypto/sha-mb/sha1_mb.c
+++ b/arch/x86/crypto/sha-mb/sha1_mb.c
@@ -694,7 +694,8 @@ static struct shash_alg sha1_mb_shash_alg = {
 		 * use ASYNC flag as some buffers in multi-buffer
 		 * algo may not have completed before hashing thread sleep
 		 */
-		.cra_flags	 = CRYPTO_ALG_TYPE_SHASH | CRYPTO_ALG_ASYNC,
+		.cra_flags	 = CRYPTO_ALG_TYPE_SHASH | CRYPTO_ALG_ASYNC |
+				   CRYPTO_ALG_INTERNAL,
 		.cra_blocksize	 = SHA1_BLOCK_SIZE,
 		.cra_module	 = THIS_MODULE,
 		.cra_list	 = LIST_HEAD_INIT(sha1_mb_shash_alg.base.cra_list),
@@ -770,7 +771,9 @@ static int sha1_mb_async_init_tfm(struct crypto_tfm *tfm)
 	struct sha1_mb_ctx *ctx = crypto_tfm_ctx(tfm);
 	struct mcryptd_hash_ctx *mctx;
 
-	mcryptd_tfm = mcryptd_alloc_ahash("__intel_sha1-mb", 0, 0);
+	mcryptd_tfm = mcryptd_alloc_ahash("__intel_sha1-mb",
+					  CRYPTO_ALG_INTERNAL,
+					  CRYPTO_ALG_INTERNAL);
 	if (IS_ERR(mcryptd_tfm))
 		return PTR_ERR(mcryptd_tfm);
 	mctx = crypto_ahash_ctx(&mcryptd_tfm->base);
@@ -828,7 +831,7 @@ static unsigned long sha1_mb_flusher(struct mcryptd_alg_cstate *cstate)
 	while (!list_empty(&cstate->work_list)) {
 		rctx = list_entry(cstate->work_list.next,
 				struct mcryptd_hash_request_ctx, waiter);
-		if time_before(cur_time, rctx->tag.expire)
+		if (time_before(cur_time, rctx->tag.expire))
 			break;
 		kernel_fpu_begin();
 		sha_ctx = (struct sha1_hash_ctx *) sha1_ctx_mgr_flush(cstate->mgr);
diff --git a/arch/x86/crypto/sha-mb/sha1_mb_mgr_init_avx2.c b/arch/x86/crypto/sha-mb/sha1_mb_mgr_init_avx2.c
index 4ca7e16..822acb5 100644
--- a/arch/x86/crypto/sha-mb/sha1_mb_mgr_init_avx2.c
+++ b/arch/x86/crypto/sha-mb/sha1_mb_mgr_init_avx2.c
@@ -56,7 +56,7 @@
 void sha1_mb_mgr_init_avx2(struct sha1_mb_mgr *state)
 {
 	unsigned int j;
-	state->unused_lanes = 0xF76543210;
+	state->unused_lanes = 0xF76543210ULL;
 	for (j = 0; j < 8; j++) {
 		state->lens[j] = 0xFFFFFFFF;
 		state->ldata[j].job_in_lane = NULL;
diff --git a/arch/x86/crypto/sha1_ssse3_glue.c b/arch/x86/crypto/sha1_ssse3_glue.c
index 6c20fe0..33d1b9d 100644
--- a/arch/x86/crypto/sha1_ssse3_glue.c
+++ b/arch/x86/crypto/sha1_ssse3_glue.c
@@ -28,7 +28,7 @@
 #include <linux/cryptohash.h>
 #include <linux/types.h>
 #include <crypto/sha.h>
-#include <asm/byteorder.h>
+#include <crypto/sha1_base.h>
 #include <asm/i387.h>
 #include <asm/xcr.h>
 #include <asm/xsave.h>
@@ -44,132 +44,51 @@ asmlinkage void sha1_transform_avx(u32 *digest, const char *data,
 #define SHA1_AVX2_BLOCK_OPTSIZE	4	/* optimal 4*64 bytes of SHA1 blocks */
 
 asmlinkage void sha1_transform_avx2(u32 *digest, const char *data,
-				unsigned int rounds);
+				    unsigned int rounds);
 #endif
 
-static asmlinkage void (*sha1_transform_asm)(u32 *, const char *, unsigned int);
-
-
-static int sha1_ssse3_init(struct shash_desc *desc)
-{
-	struct sha1_state *sctx = shash_desc_ctx(desc);
-
-	*sctx = (struct sha1_state){
-		.state = { SHA1_H0, SHA1_H1, SHA1_H2, SHA1_H3, SHA1_H4 },
-	};
-
-	return 0;
-}
-
-static int __sha1_ssse3_update(struct shash_desc *desc, const u8 *data,
-			       unsigned int len, unsigned int partial)
-{
-	struct sha1_state *sctx = shash_desc_ctx(desc);
-	unsigned int done = 0;
-
-	sctx->count += len;
-
-	if (partial) {
-		done = SHA1_BLOCK_SIZE - partial;
-		memcpy(sctx->buffer + partial, data, done);
-		sha1_transform_asm(sctx->state, sctx->buffer, 1);
-	}
-
-	if (len - done >= SHA1_BLOCK_SIZE) {
-		const unsigned int rounds = (len - done) / SHA1_BLOCK_SIZE;
-
-		sha1_transform_asm(sctx->state, data + done, rounds);
-		done += rounds * SHA1_BLOCK_SIZE;
-	}
-
-	memcpy(sctx->buffer, data + done, len - done);
-
-	return 0;
-}
+static void (*sha1_transform_asm)(u32 *, const char *, unsigned int);
 
 static int sha1_ssse3_update(struct shash_desc *desc, const u8 *data,
 			     unsigned int len)
 {
 	struct sha1_state *sctx = shash_desc_ctx(desc);
-	unsigned int partial = sctx->count % SHA1_BLOCK_SIZE;
-	int res;
 
-	/* Handle the fast case right here */
-	if (partial + len < SHA1_BLOCK_SIZE) {
-		sctx->count += len;
-		memcpy(sctx->buffer + partial, data, len);
+	if (!irq_fpu_usable() ||
+	    (sctx->count % SHA1_BLOCK_SIZE) + len < SHA1_BLOCK_SIZE)
+		return crypto_sha1_update(desc, data, len);
 
-		return 0;
-	}
+	/* make sure casting to sha1_block_fn() is safe */
+	BUILD_BUG_ON(offsetof(struct sha1_state, state) != 0);
 
-	if (!irq_fpu_usable()) {
-		res = crypto_sha1_update(desc, data, len);
-	} else {
-		kernel_fpu_begin();
-		res = __sha1_ssse3_update(desc, data, len, partial);
-		kernel_fpu_end();
-	}
-
-	return res;
-}
-
-
-/* Add padding and return the message digest. */
-static int sha1_ssse3_final(struct shash_desc *desc, u8 *out)
-{
-	struct sha1_state *sctx = shash_desc_ctx(desc);
-	unsigned int i, index, padlen;
-	__be32 *dst = (__be32 *)out;
-	__be64 bits;
-	static const u8 padding[SHA1_BLOCK_SIZE] = { 0x80, };
-
-	bits = cpu_to_be64(sctx->count << 3);
-
-	/* Pad out to 56 mod 64 and append length */
-	index = sctx->count % SHA1_BLOCK_SIZE;
-	padlen = (index < 56) ? (56 - index) : ((SHA1_BLOCK_SIZE+56) - index);
-	if (!irq_fpu_usable()) {
-		crypto_sha1_update(desc, padding, padlen);
-		crypto_sha1_update(desc, (const u8 *)&bits, sizeof(bits));
-	} else {
-		kernel_fpu_begin();
-		/* We need to fill a whole block for __sha1_ssse3_update() */
-		if (padlen <= 56) {
-			sctx->count += padlen;
-			memcpy(sctx->buffer + index, padding, padlen);
-		} else {
-			__sha1_ssse3_update(desc, padding, padlen, index);
-		}
-		__sha1_ssse3_update(desc, (const u8 *)&bits, sizeof(bits), 56);
-		kernel_fpu_end();
-	}
-
-	/* Store state in digest */
-	for (i = 0; i < 5; i++)
-		dst[i] = cpu_to_be32(sctx->state[i]);
-
-	/* Wipe context */
-	memset(sctx, 0, sizeof(*sctx));
+	kernel_fpu_begin();
+	sha1_base_do_update(desc, data, len,
+			    (sha1_block_fn *)sha1_transform_asm);
+	kernel_fpu_end();
 
 	return 0;
 }
 
-static int sha1_ssse3_export(struct shash_desc *desc, void *out)
+static int sha1_ssse3_finup(struct shash_desc *desc, const u8 *data,
+			      unsigned int len, u8 *out)
 {
-	struct sha1_state *sctx = shash_desc_ctx(desc);
+	if (!irq_fpu_usable())
+		return crypto_sha1_finup(desc, data, len, out);
 
-	memcpy(out, sctx, sizeof(*sctx));
+	kernel_fpu_begin();
+	if (len)
+		sha1_base_do_update(desc, data, len,
+				    (sha1_block_fn *)sha1_transform_asm);
+	sha1_base_do_finalize(desc, (sha1_block_fn *)sha1_transform_asm);
+	kernel_fpu_end();
 
-	return 0;
+	return sha1_base_finish(desc, out);
 }
 
-static int sha1_ssse3_import(struct shash_desc *desc, const void *in)
+/* Add padding and return the message digest. */
+static int sha1_ssse3_final(struct shash_desc *desc, u8 *out)
 {
-	struct sha1_state *sctx = shash_desc_ctx(desc);
-
-	memcpy(sctx, in, sizeof(*sctx));
-
-	return 0;
+	return sha1_ssse3_finup(desc, NULL, 0, out);
 }
 
 #ifdef CONFIG_AS_AVX2
@@ -186,13 +105,11 @@ static void sha1_apply_transform_avx2(u32 *digest, const char *data,
 
 static struct shash_alg alg = {
 	.digestsize	=	SHA1_DIGEST_SIZE,
-	.init		=	sha1_ssse3_init,
+	.init		=	sha1_base_init,
 	.update		=	sha1_ssse3_update,
 	.final		=	sha1_ssse3_final,
-	.export		=	sha1_ssse3_export,
-	.import		=	sha1_ssse3_import,
+	.finup		=	sha1_ssse3_finup,
 	.descsize	=	sizeof(struct sha1_state),
-	.statesize	=	sizeof(struct sha1_state),
 	.base		=	{
 		.cra_name	=	"sha1",
 		.cra_driver_name=	"sha1-ssse3",
diff --git a/arch/x86/crypto/sha256-avx-asm.S b/arch/x86/crypto/sha256-avx-asm.S
index 642f156..92b3b5d 100644
--- a/arch/x86/crypto/sha256-avx-asm.S
+++ b/arch/x86/crypto/sha256-avx-asm.S
@@ -96,10 +96,10 @@ SHUF_DC00 = %xmm12      # shuffle xDxC -> DC00
 BYTE_FLIP_MASK = %xmm13
 
 NUM_BLKS = %rdx   # 3rd arg
-CTX = %rsi        # 2nd arg
-INP = %rdi        # 1st arg
+INP = %rsi        # 2nd arg
+CTX = %rdi        # 1st arg
 
-SRND = %rdi       # clobbers INP
+SRND = %rsi       # clobbers INP
 c = %ecx
 d = %r8d
 e = %edx
@@ -342,8 +342,8 @@ a = TMP_
 
 ########################################################################
 ## void sha256_transform_avx(void *input_data, UINT32 digest[8], UINT64 num_blks)
-## arg 1 : pointer to input data
-## arg 2 : pointer to digest
+## arg 1 : pointer to digest
+## arg 2 : pointer to input data
 ## arg 3 : Num blocks
 ########################################################################
 .text
diff --git a/arch/x86/crypto/sha256-avx2-asm.S b/arch/x86/crypto/sha256-avx2-asm.S
index 9e86944..570ec5e 100644
--- a/arch/x86/crypto/sha256-avx2-asm.S
+++ b/arch/x86/crypto/sha256-avx2-asm.S
@@ -91,12 +91,12 @@ BYTE_FLIP_MASK = %ymm13
 X_BYTE_FLIP_MASK = %xmm13 # XMM version of BYTE_FLIP_MASK
 
 NUM_BLKS = %rdx	# 3rd arg
-CTX	= %rsi  # 2nd arg
-INP	= %rdi	# 1st arg
+INP	= %rsi  # 2nd arg
+CTX	= %rdi	# 1st arg
 c	= %ecx
 d	= %r8d
 e       = %edx	# clobbers NUM_BLKS
-y3	= %edi	# clobbers INP
+y3	= %esi	# clobbers INP
 
 
 TBL	= %rbp
@@ -523,8 +523,8 @@ STACK_SIZE	= _RSP      + _RSP_SIZE
 
 ########################################################################
 ## void sha256_transform_rorx(void *input_data, UINT32 digest[8], UINT64 num_blks)
-## arg 1 : pointer to input data
-## arg 2 : pointer to digest
+## arg 1 : pointer to digest
+## arg 2 : pointer to input data
 ## arg 3 : Num blocks
 ########################################################################
 .text
diff --git a/arch/x86/crypto/sha256-ssse3-asm.S b/arch/x86/crypto/sha256-ssse3-asm.S
index f833b74..2cedc44 100644
--- a/arch/x86/crypto/sha256-ssse3-asm.S
+++ b/arch/x86/crypto/sha256-ssse3-asm.S
@@ -88,10 +88,10 @@ SHUF_DC00 = %xmm11      # shuffle xDxC -> DC00
 BYTE_FLIP_MASK = %xmm12
 
 NUM_BLKS = %rdx   # 3rd arg
-CTX = %rsi        # 2nd arg
-INP = %rdi        # 1st arg
+INP = %rsi        # 2nd arg
+CTX = %rdi        # 1st arg
 
-SRND = %rdi       # clobbers INP
+SRND = %rsi       # clobbers INP
 c = %ecx
 d = %r8d
 e = %edx
@@ -348,8 +348,8 @@ a = TMP_
 
 ########################################################################
 ## void sha256_transform_ssse3(void *input_data, UINT32 digest[8], UINT64 num_blks)
-## arg 1 : pointer to input data
-## arg 2 : pointer to digest
+## arg 1 : pointer to digest
+## arg 2 : pointer to input data
 ## arg 3 : Num blocks
 ########################################################################
 .text
diff --git a/arch/x86/crypto/sha256_ssse3_glue.c b/arch/x86/crypto/sha256_ssse3_glue.c
index 8fad72f..ccc3388 100644
--- a/arch/x86/crypto/sha256_ssse3_glue.c
+++ b/arch/x86/crypto/sha256_ssse3_glue.c
@@ -36,195 +36,74 @@
 #include <linux/cryptohash.h>
 #include <linux/types.h>
 #include <crypto/sha.h>
-#include <asm/byteorder.h>
+#include <crypto/sha256_base.h>
 #include <asm/i387.h>
 #include <asm/xcr.h>
 #include <asm/xsave.h>
 #include <linux/string.h>
 
-asmlinkage void sha256_transform_ssse3(const char *data, u32 *digest,
-				     u64 rounds);
+asmlinkage void sha256_transform_ssse3(u32 *digest, const char *data,
+				       u64 rounds);
 #ifdef CONFIG_AS_AVX
-asmlinkage void sha256_transform_avx(const char *data, u32 *digest,
+asmlinkage void sha256_transform_avx(u32 *digest, const char *data,
 				     u64 rounds);
 #endif
 #ifdef CONFIG_AS_AVX2
-asmlinkage void sha256_transform_rorx(const char *data, u32 *digest,
-				     u64 rounds);
+asmlinkage void sha256_transform_rorx(u32 *digest, const char *data,
+				      u64 rounds);
 #endif
 
-static asmlinkage void (*sha256_transform_asm)(const char *, u32 *, u64);
-
-
-static int sha256_ssse3_init(struct shash_desc *desc)
-{
-	struct sha256_state *sctx = shash_desc_ctx(desc);
-
-	sctx->state[0] = SHA256_H0;
-	sctx->state[1] = SHA256_H1;
-	sctx->state[2] = SHA256_H2;
-	sctx->state[3] = SHA256_H3;
-	sctx->state[4] = SHA256_H4;
-	sctx->state[5] = SHA256_H5;
-	sctx->state[6] = SHA256_H6;
-	sctx->state[7] = SHA256_H7;
-	sctx->count = 0;
-
-	return 0;
-}
-
-static int __sha256_ssse3_update(struct shash_desc *desc, const u8 *data,
-			       unsigned int len, unsigned int partial)
-{
-	struct sha256_state *sctx = shash_desc_ctx(desc);
-	unsigned int done = 0;
-
-	sctx->count += len;
-
-	if (partial) {
-		done = SHA256_BLOCK_SIZE - partial;
-		memcpy(sctx->buf + partial, data, done);
-		sha256_transform_asm(sctx->buf, sctx->state, 1);
-	}
-
-	if (len - done >= SHA256_BLOCK_SIZE) {
-		const unsigned int rounds = (len - done) / SHA256_BLOCK_SIZE;
-
-		sha256_transform_asm(data + done, sctx->state, (u64) rounds);
-
-		done += rounds * SHA256_BLOCK_SIZE;
-	}
-
-	memcpy(sctx->buf, data + done, len - done);
-
-	return 0;
-}
+static void (*sha256_transform_asm)(u32 *, const char *, u64);
 
 static int sha256_ssse3_update(struct shash_desc *desc, const u8 *data,
 			     unsigned int len)
 {
 	struct sha256_state *sctx = shash_desc_ctx(desc);
-	unsigned int partial = sctx->count % SHA256_BLOCK_SIZE;
-	int res;
 
-	/* Handle the fast case right here */
-	if (partial + len < SHA256_BLOCK_SIZE) {
-		sctx->count += len;
-		memcpy(sctx->buf + partial, data, len);
+	if (!irq_fpu_usable() ||
+	    (sctx->count % SHA256_BLOCK_SIZE) + len < SHA256_BLOCK_SIZE)
+		return crypto_sha256_update(desc, data, len);
 
-		return 0;
-	}
-
-	if (!irq_fpu_usable()) {
-		res = crypto_sha256_update(desc, data, len);
-	} else {
-		kernel_fpu_begin();
-		res = __sha256_ssse3_update(desc, data, len, partial);
-		kernel_fpu_end();
-	}
-
-	return res;
-}
+	/* make sure casting to sha256_block_fn() is safe */
+	BUILD_BUG_ON(offsetof(struct sha256_state, state) != 0);
 
-
-/* Add padding and return the message digest. */
-static int sha256_ssse3_final(struct shash_desc *desc, u8 *out)
-{
-	struct sha256_state *sctx = shash_desc_ctx(desc);
-	unsigned int i, index, padlen;
-	__be32 *dst = (__be32 *)out;
-	__be64 bits;
-	static const u8 padding[SHA256_BLOCK_SIZE] = { 0x80, };
-
-	bits = cpu_to_be64(sctx->count << 3);
-
-	/* Pad out to 56 mod 64 and append length */
-	index = sctx->count % SHA256_BLOCK_SIZE;
-	padlen = (index < 56) ? (56 - index) : ((SHA256_BLOCK_SIZE+56)-index);
-
-	if (!irq_fpu_usable()) {
-		crypto_sha256_update(desc, padding, padlen);
-		crypto_sha256_update(desc, (const u8 *)&bits, sizeof(bits));
-	} else {
-		kernel_fpu_begin();
-		/* We need to fill a whole block for __sha256_ssse3_update() */
-		if (padlen <= 56) {
-			sctx->count += padlen;
-			memcpy(sctx->buf + index, padding, padlen);
-		} else {
-			__sha256_ssse3_update(desc, padding, padlen, index);
-		}
-		__sha256_ssse3_update(desc, (const u8 *)&bits,
-					sizeof(bits), 56);
-		kernel_fpu_end();
-	}
-
-	/* Store state in digest */
-	for (i = 0; i < 8; i++)
-		dst[i] = cpu_to_be32(sctx->state[i]);
-
-	/* Wipe context */
-	memset(sctx, 0, sizeof(*sctx));
+	kernel_fpu_begin();
+	sha256_base_do_update(desc, data, len,
+			      (sha256_block_fn *)sha256_transform_asm);
+	kernel_fpu_end();
 
 	return 0;
 }
 
-static int sha256_ssse3_export(struct shash_desc *desc, void *out)
+static int sha256_ssse3_finup(struct shash_desc *desc, const u8 *data,
+			      unsigned int len, u8 *out)
 {
-	struct sha256_state *sctx = shash_desc_ctx(desc);
+	if (!irq_fpu_usable())
+		return crypto_sha256_finup(desc, data, len, out);
 
-	memcpy(out, sctx, sizeof(*sctx));
+	kernel_fpu_begin();
+	if (len)
+		sha256_base_do_update(desc, data, len,
+				      (sha256_block_fn *)sha256_transform_asm);
+	sha256_base_do_finalize(desc, (sha256_block_fn *)sha256_transform_asm);
+	kernel_fpu_end();
 
-	return 0;
+	return sha256_base_finish(desc, out);
 }
 
-static int sha256_ssse3_import(struct shash_desc *desc, const void *in)
-{
-	struct sha256_state *sctx = shash_desc_ctx(desc);
-
-	memcpy(sctx, in, sizeof(*sctx));
-
-	return 0;
-}
-
-static int sha224_ssse3_init(struct shash_desc *desc)
-{
-	struct sha256_state *sctx = shash_desc_ctx(desc);
-
-	sctx->state[0] = SHA224_H0;
-	sctx->state[1] = SHA224_H1;
-	sctx->state[2] = SHA224_H2;
-	sctx->state[3] = SHA224_H3;
-	sctx->state[4] = SHA224_H4;
-	sctx->state[5] = SHA224_H5;
-	sctx->state[6] = SHA224_H6;
-	sctx->state[7] = SHA224_H7;
-	sctx->count = 0;
-
-	return 0;
-}
-
-static int sha224_ssse3_final(struct shash_desc *desc, u8 *hash)
+/* Add padding and return the message digest. */
+static int sha256_ssse3_final(struct shash_desc *desc, u8 *out)
 {
-	u8 D[SHA256_DIGEST_SIZE];
-
-	sha256_ssse3_final(desc, D);
-
-	memcpy(hash, D, SHA224_DIGEST_SIZE);
-	memzero_explicit(D, SHA256_DIGEST_SIZE);
-
-	return 0;
+	return sha256_ssse3_finup(desc, NULL, 0, out);
 }
 
 static struct shash_alg algs[] = { {
 	.digestsize	=	SHA256_DIGEST_SIZE,
-	.init		=	sha256_ssse3_init,
+	.init		=	sha256_base_init,
 	.update		=	sha256_ssse3_update,
 	.final		=	sha256_ssse3_final,
-	.export		=	sha256_ssse3_export,
-	.import		=	sha256_ssse3_import,
+	.finup		=	sha256_ssse3_finup,
 	.descsize	=	sizeof(struct sha256_state),
-	.statesize	=	sizeof(struct sha256_state),
 	.base		=	{
 		.cra_name	=	"sha256",
 		.cra_driver_name =	"sha256-ssse3",
@@ -235,13 +114,11 @@ static struct shash_alg algs[] = { {
 	}
 }, {
 	.digestsize	=	SHA224_DIGEST_SIZE,
-	.init		=	sha224_ssse3_init,
+	.init		=	sha224_base_init,
 	.update		=	sha256_ssse3_update,
-	.final		=	sha224_ssse3_final,
-	.export		=	sha256_ssse3_export,
-	.import		=	sha256_ssse3_import,
+	.final		=	sha256_ssse3_final,
+	.finup		=	sha256_ssse3_finup,
 	.descsize	=	sizeof(struct sha256_state),
-	.statesize	=	sizeof(struct sha256_state),
 	.base		=	{
 		.cra_name	=	"sha224",
 		.cra_driver_name =	"sha224-ssse3",
diff --git a/arch/x86/crypto/sha512-avx-asm.S b/arch/x86/crypto/sha512-avx-asm.S
index 974dde9..565274d 100644
--- a/arch/x86/crypto/sha512-avx-asm.S
+++ b/arch/x86/crypto/sha512-avx-asm.S
@@ -54,9 +54,9 @@
 
 # Virtual Registers
 # ARG1
-msg	= %rdi
+digest	= %rdi
 # ARG2
-digest	= %rsi
+msg	= %rsi
 # ARG3
 msglen	= %rdx
 T1	= %rcx
@@ -271,7 +271,7 @@ frame_size = frame_GPRSAVE + GPRSAVE_SIZE
 .endm
 
 ########################################################################
-# void sha512_transform_avx(const void* M, void* D, u64 L)
+# void sha512_transform_avx(void* D, const void* M, u64 L)
 # Purpose: Updates the SHA512 digest stored at D with the message stored in M.
 # The size of the message pointed to by M must be an integer multiple of SHA512
 # message blocks.
diff --git a/arch/x86/crypto/sha512-avx2-asm.S b/arch/x86/crypto/sha512-avx2-asm.S
index 568b961..a4771dc 100644
--- a/arch/x86/crypto/sha512-avx2-asm.S
+++ b/arch/x86/crypto/sha512-avx2-asm.S
@@ -70,9 +70,9 @@ XFER  = YTMP0
 BYTE_FLIP_MASK  = %ymm9
 
 # 1st arg
-INP         = %rdi
+CTX         = %rdi
 # 2nd arg
-CTX         = %rsi
+INP         = %rsi
 # 3rd arg
 NUM_BLKS    = %rdx
 
@@ -562,7 +562,7 @@ frame_size = frame_GPRSAVE + GPRSAVE_SIZE
 .endm
 
 ########################################################################
-# void sha512_transform_rorx(const void* M, void* D, uint64_t L)#
+# void sha512_transform_rorx(void* D, const void* M, uint64_t L)#
 # Purpose: Updates the SHA512 digest stored at D with the message stored in M.
 # The size of the message pointed to by M must be an integer multiple of SHA512
 #   message blocks.
diff --git a/arch/x86/crypto/sha512-ssse3-asm.S b/arch/x86/crypto/sha512-ssse3-asm.S
index fb56855..e610e29 100644
--- a/arch/x86/crypto/sha512-ssse3-asm.S
+++ b/arch/x86/crypto/sha512-ssse3-asm.S
@@ -53,9 +53,9 @@
 
 # Virtual Registers
 # ARG1
-msg =		%rdi
+digest =	%rdi
 # ARG2
-digest =	%rsi
+msg =		%rsi
 # ARG3
 msglen =	%rdx
 T1 =		%rcx
@@ -269,7 +269,7 @@ frame_size = frame_GPRSAVE + GPRSAVE_SIZE
 .endm
 
 ########################################################################
-# void sha512_transform_ssse3(const void* M, void* D, u64 L)#
+# void sha512_transform_ssse3(void* D, const void* M, u64 L)#
 # Purpose: Updates the SHA512 digest stored at D with the message stored in M.
 # The size of the message pointed to by M must be an integer multiple of SHA512
 #   message blocks.
diff --git a/arch/x86/crypto/sha512_ssse3_glue.c b/arch/x86/crypto/sha512_ssse3_glue.c
index 0b6af26..d9fa4c1 100644
--- a/arch/x86/crypto/sha512_ssse3_glue.c
+++ b/arch/x86/crypto/sha512_ssse3_glue.c
@@ -34,205 +34,75 @@
 #include <linux/cryptohash.h>
 #include <linux/types.h>
 #include <crypto/sha.h>
-#include <asm/byteorder.h>
+#include <crypto/sha512_base.h>
 #include <asm/i387.h>
 #include <asm/xcr.h>
 #include <asm/xsave.h>
 
 #include <linux/string.h>
 
-asmlinkage void sha512_transform_ssse3(const char *data, u64 *digest,
-				     u64 rounds);
+asmlinkage void sha512_transform_ssse3(u64 *digest, const char *data,
+				       u64 rounds);
 #ifdef CONFIG_AS_AVX
-asmlinkage void sha512_transform_avx(const char *data, u64 *digest,
+asmlinkage void sha512_transform_avx(u64 *digest, const char *data,
 				     u64 rounds);
 #endif
 #ifdef CONFIG_AS_AVX2
-asmlinkage void sha512_transform_rorx(const char *data, u64 *digest,
-				     u64 rounds);
+asmlinkage void sha512_transform_rorx(u64 *digest, const char *data,
+				      u64 rounds);
 #endif
 
-static asmlinkage void (*sha512_transform_asm)(const char *, u64 *, u64);
-
-
-static int sha512_ssse3_init(struct shash_desc *desc)
-{
-	struct sha512_state *sctx = shash_desc_ctx(desc);
-
-	sctx->state[0] = SHA512_H0;
-	sctx->state[1] = SHA512_H1;
-	sctx->state[2] = SHA512_H2;
-	sctx->state[3] = SHA512_H3;
-	sctx->state[4] = SHA512_H4;
-	sctx->state[5] = SHA512_H5;
-	sctx->state[6] = SHA512_H6;
-	sctx->state[7] = SHA512_H7;
-	sctx->count[0] = sctx->count[1] = 0;
-
-	return 0;
-}
+static void (*sha512_transform_asm)(u64 *, const char *, u64);
 
-static int __sha512_ssse3_update(struct shash_desc *desc, const u8 *data,
-			       unsigned int len, unsigned int partial)
+static int sha512_ssse3_update(struct shash_desc *desc, const u8 *data,
+			       unsigned int len)
 {
 	struct sha512_state *sctx = shash_desc_ctx(desc);
-	unsigned int done = 0;
-
-	sctx->count[0] += len;
-	if (sctx->count[0] < len)
-		sctx->count[1]++;
 
-	if (partial) {
-		done = SHA512_BLOCK_SIZE - partial;
-		memcpy(sctx->buf + partial, data, done);
-		sha512_transform_asm(sctx->buf, sctx->state, 1);
-	}
-
-	if (len - done >= SHA512_BLOCK_SIZE) {
-		const unsigned int rounds = (len - done) / SHA512_BLOCK_SIZE;
+	if (!irq_fpu_usable() ||
+	    (sctx->count[0] % SHA512_BLOCK_SIZE) + len < SHA512_BLOCK_SIZE)
+		return crypto_sha512_update(desc, data, len);
 
-		sha512_transform_asm(data + done, sctx->state, (u64) rounds);
-
-		done += rounds * SHA512_BLOCK_SIZE;
-	}
+	/* make sure casting to sha512_block_fn() is safe */
+	BUILD_BUG_ON(offsetof(struct sha512_state, state) != 0);
 
-	memcpy(sctx->buf, data + done, len - done);
+	kernel_fpu_begin();
+	sha512_base_do_update(desc, data, len,
+			      (sha512_block_fn *)sha512_transform_asm);
+	kernel_fpu_end();
 
 	return 0;
 }
 
-static int sha512_ssse3_update(struct shash_desc *desc, const u8 *data,
-			     unsigned int len)
+static int sha512_ssse3_finup(struct shash_desc *desc, const u8 *data,
+			      unsigned int len, u8 *out)
 {
-	struct sha512_state *sctx = shash_desc_ctx(desc);
-	unsigned int partial = sctx->count[0] % SHA512_BLOCK_SIZE;
-	int res;
-
-	/* Handle the fast case right here */
-	if (partial + len < SHA512_BLOCK_SIZE) {
-		sctx->count[0] += len;
-		if (sctx->count[0] < len)
-			sctx->count[1]++;
-		memcpy(sctx->buf + partial, data, len);
-
-		return 0;
-	}
+	if (!irq_fpu_usable())
+		return crypto_sha512_finup(desc, data, len, out);
 
-	if (!irq_fpu_usable()) {
-		res = crypto_sha512_update(desc, data, len);
-	} else {
-		kernel_fpu_begin();
-		res = __sha512_ssse3_update(desc, data, len, partial);
-		kernel_fpu_end();
-	}
+	kernel_fpu_begin();
+	if (len)
+		sha512_base_do_update(desc, data, len,
+				      (sha512_block_fn *)sha512_transform_asm);
+	sha512_base_do_finalize(desc, (sha512_block_fn *)sha512_transform_asm);
+	kernel_fpu_end();
 
-	return res;
+	return sha512_base_finish(desc, out);
 }
 
-
 /* Add padding and return the message digest. */
 static int sha512_ssse3_final(struct shash_desc *desc, u8 *out)
 {
-	struct sha512_state *sctx = shash_desc_ctx(desc);
-	unsigned int i, index, padlen;
-	__be64 *dst = (__be64 *)out;
-	__be64 bits[2];
-	static const u8 padding[SHA512_BLOCK_SIZE] = { 0x80, };
-
-	/* save number of bits */
-	bits[1] = cpu_to_be64(sctx->count[0] << 3);
-	bits[0] = cpu_to_be64(sctx->count[1] << 3 | sctx->count[0] >> 61);
-
-	/* Pad out to 112 mod 128 and append length */
-	index = sctx->count[0] & 0x7f;
-	padlen = (index < 112) ? (112 - index) : ((128+112) - index);
-
-	if (!irq_fpu_usable()) {
-		crypto_sha512_update(desc, padding, padlen);
-		crypto_sha512_update(desc, (const u8 *)&bits, sizeof(bits));
-	} else {
-		kernel_fpu_begin();
-		/* We need to fill a whole block for __sha512_ssse3_update() */
-		if (padlen <= 112) {
-			sctx->count[0] += padlen;
-			if (sctx->count[0] < padlen)
-				sctx->count[1]++;
-			memcpy(sctx->buf + index, padding, padlen);
-		} else {
-			__sha512_ssse3_update(desc, padding, padlen, index);
-		}
-		__sha512_ssse3_update(desc, (const u8 *)&bits,
-					sizeof(bits), 112);
-		kernel_fpu_end();
-	}
-
-	/* Store state in digest */
-	for (i = 0; i < 8; i++)
-		dst[i] = cpu_to_be64(sctx->state[i]);
-
-	/* Wipe context */
-	memset(sctx, 0, sizeof(*sctx));
-
-	return 0;
-}
-
-static int sha512_ssse3_export(struct shash_desc *desc, void *out)
-{
-	struct sha512_state *sctx = shash_desc_ctx(desc);
-
-	memcpy(out, sctx, sizeof(*sctx));
-
-	return 0;
-}
-
-static int sha512_ssse3_import(struct shash_desc *desc, const void *in)
-{
-	struct sha512_state *sctx = shash_desc_ctx(desc);
-
-	memcpy(sctx, in, sizeof(*sctx));
-
-	return 0;
-}
-
-static int sha384_ssse3_init(struct shash_desc *desc)
-{
-	struct sha512_state *sctx = shash_desc_ctx(desc);
-
-	sctx->state[0] = SHA384_H0;
-	sctx->state[1] = SHA384_H1;
-	sctx->state[2] = SHA384_H2;
-	sctx->state[3] = SHA384_H3;
-	sctx->state[4] = SHA384_H4;
-	sctx->state[5] = SHA384_H5;
-	sctx->state[6] = SHA384_H6;
-	sctx->state[7] = SHA384_H7;
-
-	sctx->count[0] = sctx->count[1] = 0;
-
-	return 0;
-}
-
-static int sha384_ssse3_final(struct shash_desc *desc, u8 *hash)
-{
-	u8 D[SHA512_DIGEST_SIZE];
-
-	sha512_ssse3_final(desc, D);
-
-	memcpy(hash, D, SHA384_DIGEST_SIZE);
-	memzero_explicit(D, SHA512_DIGEST_SIZE);
-
-	return 0;
+	return sha512_ssse3_finup(desc, NULL, 0, out);
 }
 
 static struct shash_alg algs[] = { {
 	.digestsize	=	SHA512_DIGEST_SIZE,
-	.init		=	sha512_ssse3_init,
+	.init		=	sha512_base_init,
 	.update		=	sha512_ssse3_update,
 	.final		=	sha512_ssse3_final,
-	.export		=	sha512_ssse3_export,
-	.import		=	sha512_ssse3_import,
+	.finup		=	sha512_ssse3_finup,
 	.descsize	=	sizeof(struct sha512_state),
-	.statesize	=	sizeof(struct sha512_state),
 	.base		=	{
 		.cra_name	=	"sha512",
 		.cra_driver_name =	"sha512-ssse3",
@@ -243,13 +113,11 @@ static struct shash_alg algs[] = { {
 	}
 },  {
 	.digestsize	=	SHA384_DIGEST_SIZE,
-	.init		=	sha384_ssse3_init,
+	.init		=	sha384_base_init,
 	.update		=	sha512_ssse3_update,
-	.final		=	sha384_ssse3_final,
-	.export		=	sha512_ssse3_export,
-	.import		=	sha512_ssse3_import,
+	.final		=	sha512_ssse3_final,
+	.finup		=	sha512_ssse3_finup,
 	.descsize	=	sizeof(struct sha512_state),
-	.statesize	=	sizeof(struct sha512_state),
 	.base		=	{
 		.cra_name	=	"sha384",
 		.cra_driver_name =	"sha384-ssse3",
diff --git a/arch/x86/crypto/twofish_avx_glue.c b/arch/x86/crypto/twofish_avx_glue.c
index 1ac531e..b5e2d56 100644
--- a/arch/x86/crypto/twofish_avx_glue.c
+++ b/arch/x86/crypto/twofish_avx_glue.c
@@ -340,7 +340,8 @@ static struct crypto_alg twofish_algs[10] = { {
 	.cra_name		= "__ecb-twofish-avx",
 	.cra_driver_name	= "__driver-ecb-twofish-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= TF_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct twofish_ctx),
 	.cra_alignmask		= 0,
@@ -359,7 +360,8 @@ static struct crypto_alg twofish_algs[10] = { {
 	.cra_name		= "__cbc-twofish-avx",
 	.cra_driver_name	= "__driver-cbc-twofish-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= TF_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct twofish_ctx),
 	.cra_alignmask		= 0,
@@ -378,7 +380,8 @@ static struct crypto_alg twofish_algs[10] = { {
 	.cra_name		= "__ctr-twofish-avx",
 	.cra_driver_name	= "__driver-ctr-twofish-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= 1,
 	.cra_ctxsize		= sizeof(struct twofish_ctx),
 	.cra_alignmask		= 0,
@@ -398,7 +401,8 @@ static struct crypto_alg twofish_algs[10] = { {
 	.cra_name		= "__lrw-twofish-avx",
 	.cra_driver_name	= "__driver-lrw-twofish-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= TF_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct twofish_lrw_ctx),
 	.cra_alignmask		= 0,
@@ -421,7 +425,8 @@ static struct crypto_alg twofish_algs[10] = { {
 	.cra_name		= "__xts-twofish-avx",
 	.cra_driver_name	= "__driver-xts-twofish-avx",
 	.cra_priority		= 0,
-	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER |
+				  CRYPTO_ALG_INTERNAL,
 	.cra_blocksize		= TF_BLOCK_SIZE,
 	.cra_ctxsize		= sizeof(struct twofish_xts_ctx),
 	.cra_alignmask		= 0,