crypto: camellia - add AES-NI/AVX/x86_64 assembler implementation of camellia cipher

This patch adds AES-NI/AVX/x86_64 assembler implementation of Camellia block
cipher. Implementation process data in sixteen block chunks, which are
byte-sliced and AES SubBytes is reused for Camellia s-box with help of pre-
and post-filtering.

Patch has been tested with tcrypt and automated filesystem tests.

tcrypt test results:

Intel Core i5-2450M:

camellia-aesni-avx vs camellia-asm-x86_64-2way:
128bit key:                                             (lrw:256bit)    (xts:256bit)
size    ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B     0.98x   0.96x   0.99x   0.96x   0.96x   0.95x   0.95x   0.94x   0.97x   0.98x
64B     0.99x   0.98x   1.00x   0.98x   0.98x   0.99x   0.98x   0.93x   0.99x   0.98x
256B    2.28x   2.28x   1.01x   2.29x   2.25x   2.24x   1.96x   1.97x   1.91x   1.90x
1024B   2.57x   2.56x   1.00x   2.57x   2.51x   2.53x   2.19x   2.17x   2.19x   2.22x
8192B   2.49x   2.49x   1.00x   2.53x   2.48x   2.49x   2.17x   2.17x   2.22x   2.22x

256bit key:                                             (lrw:384bit)    (xts:512bit)
size    ecb-enc ecb-dec cbc-enc cbc-dec ctr-enc ctr-dec lrw-enc lrw-dec xts-enc xts-dec
16B     0.97x   0.98x   0.99x   0.97x   0.97x   0.96x   0.97x   0.98x   0.98x   0.99x
64B     1.00x   1.00x   1.01x   0.99x   0.98x   0.99x   0.99x   0.99x   0.99x   0.99x
256B    2.37x   2.37x   1.01x   2.39x   2.35x   2.33x   2.10x   2.11x   1.99x   2.02x
1024B   2.58x   2.60x   1.00x   2.58x   2.56x   2.56x   2.28x   2.29x   2.28x   2.29x
8192B   2.50x   2.52x   1.00x   2.56x   2.51x   2.51x   2.24x   2.25x   2.26x   2.29x

Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

1 files changed, 558 insertions, 0 deletions

diff --git a/arch/x86/crypto/camellia_aesni_avx_glue.c b/arch/x86/crypto/camellia_aesni_avx_glue.c
new file mode 100644
index 000000000000..96cbb6068fce
--- /dev/null
+++ b/arch/x86/crypto/camellia_aesni_avx_glue.c
@@ -0,0 +1,558 @@
+/*
+ * Glue Code for x86_64/AVX/AES-NI assembler optimized version of Camellia
+ *
+ * Copyright © 2012 Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/crypto.h>
+#include <linux/err.h>
+#include <crypto/algapi.h>
+#include <crypto/ctr.h>
+#include <crypto/lrw.h>
+#include <crypto/xts.h>
+#include <asm/xcr.h>
+#include <asm/xsave.h>
+#include <asm/crypto/camellia.h>
+#include <asm/crypto/ablk_helper.h>
+#include <asm/crypto/glue_helper.h>
+
+#define CAMELLIA_AESNI_PARALLEL_BLOCKS 16
+
+/* 16-way AES-NI parallel cipher functions */
+asmlinkage void camellia_ecb_enc_16way(struct camellia_ctx *ctx, u8 *dst,
+				       const u8 *src);
+asmlinkage void camellia_ecb_dec_16way(struct camellia_ctx *ctx, u8 *dst,
+				       const u8 *src);
+
+asmlinkage void camellia_cbc_dec_16way(struct camellia_ctx *ctx, u8 *dst,
+				       const u8 *src);
+asmlinkage void camellia_ctr_16way(struct camellia_ctx *ctx, u8 *dst,
+				   const u8 *src, le128 *iv);
+
+static const struct common_glue_ctx camellia_enc = {
+	.num_funcs = 3,
+	.fpu_blocks_limit = CAMELLIA_AESNI_PARALLEL_BLOCKS,
+
+	.funcs = { {
+		.num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS,
+		.fn_u = { .ecb = GLUE_FUNC_CAST(camellia_ecb_enc_16way) }
+	}, {
+		.num_blocks = 2,
+		.fn_u = { .ecb = GLUE_FUNC_CAST(camellia_enc_blk_2way) }
+	}, {
+		.num_blocks = 1,
+		.fn_u = { .ecb = GLUE_FUNC_CAST(camellia_enc_blk) }
+	} }
+};
+
+static const struct common_glue_ctx camellia_ctr = {
+	.num_funcs = 3,
+	.fpu_blocks_limit = CAMELLIA_AESNI_PARALLEL_BLOCKS,
+
+	.funcs = { {
+		.num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS,
+		.fn_u = { .ctr = GLUE_CTR_FUNC_CAST(camellia_ctr_16way) }
+	}, {
+		.num_blocks = 2,
+		.fn_u = { .ctr = GLUE_CTR_FUNC_CAST(camellia_crypt_ctr_2way) }
+	}, {
+		.num_blocks = 1,
+		.fn_u = { .ctr = GLUE_CTR_FUNC_CAST(camellia_crypt_ctr) }
+	} }
+};
+
+static const struct common_glue_ctx camellia_dec = {
+	.num_funcs = 3,
+	.fpu_blocks_limit = CAMELLIA_AESNI_PARALLEL_BLOCKS,
+
+	.funcs = { {
+		.num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS,
+		.fn_u = { .ecb = GLUE_FUNC_CAST(camellia_ecb_dec_16way) }
+	}, {
+		.num_blocks = 2,
+		.fn_u = { .ecb = GLUE_FUNC_CAST(camellia_dec_blk_2way) }
+	}, {
+		.num_blocks = 1,
+		.fn_u = { .ecb = GLUE_FUNC_CAST(camellia_dec_blk) }
+	} }
+};
+
+static const struct common_glue_ctx camellia_dec_cbc = {
+	.num_funcs = 3,
+	.fpu_blocks_limit = CAMELLIA_AESNI_PARALLEL_BLOCKS,
+
+	.funcs = { {
+		.num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS,
+		.fn_u = { .cbc = GLUE_CBC_FUNC_CAST(camellia_cbc_dec_16way) }
+	}, {
+		.num_blocks = 2,
+		.fn_u = { .cbc = GLUE_CBC_FUNC_CAST(camellia_decrypt_cbc_2way) }
+	}, {
+		.num_blocks = 1,
+		.fn_u = { .cbc = GLUE_CBC_FUNC_CAST(camellia_dec_blk) }
+	} }
+};
+
+static int ecb_encrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
+		       struct scatterlist *src, unsigned int nbytes)
+{
+	return glue_ecb_crypt_128bit(&camellia_enc, desc, dst, src, nbytes);
+}
+
+static int ecb_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
+		       struct scatterlist *src, unsigned int nbytes)
+{
+	return glue_ecb_crypt_128bit(&camellia_dec, desc, dst, src, nbytes);
+}
+
+static int cbc_encrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
+		       struct scatterlist *src, unsigned int nbytes)
+{
+	return glue_cbc_encrypt_128bit(GLUE_FUNC_CAST(camellia_enc_blk), desc,
+				       dst, src, nbytes);
+}
+
+static int cbc_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
+		       struct scatterlist *src, unsigned int nbytes)
+{
+	return glue_cbc_decrypt_128bit(&camellia_dec_cbc, desc, dst, src,
+				       nbytes);
+}
+
+static int ctr_crypt(struct blkcipher_desc *desc, struct scatterlist *dst,
+		     struct scatterlist *src, unsigned int nbytes)
+{
+	return glue_ctr_crypt_128bit(&camellia_ctr, desc, dst, src, nbytes);
+}
+
+static inline bool camellia_fpu_begin(bool fpu_enabled, unsigned int nbytes)
+{
+	return glue_fpu_begin(CAMELLIA_BLOCK_SIZE,
+			      CAMELLIA_AESNI_PARALLEL_BLOCKS, NULL, fpu_enabled,
+			      nbytes);
+}
+
+static inline void camellia_fpu_end(bool fpu_enabled)
+{
+	glue_fpu_end(fpu_enabled);
+}
+
+static int camellia_setkey(struct crypto_tfm *tfm, const u8 *in_key,
+			   unsigned int key_len)
+{
+	return __camellia_setkey(crypto_tfm_ctx(tfm), in_key, key_len,
+				 &tfm->crt_flags);
+}
+
+struct crypt_priv {
+	struct camellia_ctx *ctx;
+	bool fpu_enabled;
+};
+
+static void encrypt_callback(void *priv, u8 *srcdst, unsigned int nbytes)
+{
+	const unsigned int bsize = CAMELLIA_BLOCK_SIZE;
+	struct crypt_priv *ctx = priv;
+	int i;
+
+	ctx->fpu_enabled = camellia_fpu_begin(ctx->fpu_enabled, nbytes);
+
+	if (nbytes >= CAMELLIA_AESNI_PARALLEL_BLOCKS * bsize) {
+		camellia_ecb_enc_16way(ctx->ctx, srcdst, srcdst);
+		srcdst += bsize * CAMELLIA_AESNI_PARALLEL_BLOCKS;
+		nbytes -= bsize * CAMELLIA_AESNI_PARALLEL_BLOCKS;
+	}
+
+	while (nbytes >= CAMELLIA_PARALLEL_BLOCKS * bsize) {
+		camellia_enc_blk_2way(ctx->ctx, srcdst, srcdst);
+		srcdst += bsize * CAMELLIA_PARALLEL_BLOCKS;
+		nbytes -= bsize * CAMELLIA_PARALLEL_BLOCKS;
+	}
+
+	for (i = 0; i < nbytes / bsize; i++, srcdst += bsize)
+		camellia_enc_blk(ctx->ctx, srcdst, srcdst);
+}
+
+static void decrypt_callback(void *priv, u8 *srcdst, unsigned int nbytes)
+{
+	const unsigned int bsize = CAMELLIA_BLOCK_SIZE;
+	struct crypt_priv *ctx = priv;
+	int i;
+
+	ctx->fpu_enabled = camellia_fpu_begin(ctx->fpu_enabled, nbytes);
+
+	if (nbytes >= CAMELLIA_AESNI_PARALLEL_BLOCKS * bsize) {
+		camellia_ecb_dec_16way(ctx->ctx, srcdst, srcdst);
+		srcdst += bsize * CAMELLIA_AESNI_PARALLEL_BLOCKS;
+		nbytes -= bsize * CAMELLIA_AESNI_PARALLEL_BLOCKS;
+	}
+
+	while (nbytes >= CAMELLIA_PARALLEL_BLOCKS * bsize) {
+		camellia_dec_blk_2way(ctx->ctx, srcdst, srcdst);
+		srcdst += bsize * CAMELLIA_PARALLEL_BLOCKS;
+		nbytes -= bsize * CAMELLIA_PARALLEL_BLOCKS;
+	}
+
+	for (i = 0; i < nbytes / bsize; i++, srcdst += bsize)
+		camellia_dec_blk(ctx->ctx, srcdst, srcdst);
+}
+
+static int lrw_encrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
+		       struct scatterlist *src, unsigned int nbytes)
+{
+	struct camellia_lrw_ctx *ctx = crypto_blkcipher_ctx(desc->tfm);
+	be128 buf[CAMELLIA_AESNI_PARALLEL_BLOCKS];
+	struct crypt_priv crypt_ctx = {
+		.ctx = &ctx->camellia_ctx,
+		.fpu_enabled = false,
+	};
+	struct lrw_crypt_req req = {
+		.tbuf = buf,
+		.tbuflen = sizeof(buf),
+
+		.table_ctx = &ctx->lrw_table,
+		.crypt_ctx = &crypt_ctx,
+		.crypt_fn = encrypt_callback,
+	};
+	int ret;
+
+	desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
+	ret = lrw_crypt(desc, dst, src, nbytes, &req);
+	camellia_fpu_end(crypt_ctx.fpu_enabled);
+
+	return ret;
+}
+
+static int lrw_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
+		       struct scatterlist *src, unsigned int nbytes)
+{
+	struct camellia_lrw_ctx *ctx = crypto_blkcipher_ctx(desc->tfm);
+	be128 buf[CAMELLIA_AESNI_PARALLEL_BLOCKS];
+	struct crypt_priv crypt_ctx = {
+		.ctx = &ctx->camellia_ctx,
+		.fpu_enabled = false,
+	};
+	struct lrw_crypt_req req = {
+		.tbuf = buf,
+		.tbuflen = sizeof(buf),
+
+		.table_ctx = &ctx->lrw_table,
+		.crypt_ctx = &crypt_ctx,
+		.crypt_fn = decrypt_callback,
+	};
+	int ret;
+
+	desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
+	ret = lrw_crypt(desc, dst, src, nbytes, &req);
+	camellia_fpu_end(crypt_ctx.fpu_enabled);
+
+	return ret;
+}
+
+static int xts_encrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
+		       struct scatterlist *src, unsigned int nbytes)
+{
+	struct camellia_xts_ctx *ctx = crypto_blkcipher_ctx(desc->tfm);
+	be128 buf[CAMELLIA_AESNI_PARALLEL_BLOCKS];
+	struct crypt_priv crypt_ctx = {
+		.ctx = &ctx->crypt_ctx,
+		.fpu_enabled = false,
+	};
+	struct xts_crypt_req req = {
+		.tbuf = buf,
+		.tbuflen = sizeof(buf),
+
+		.tweak_ctx = &ctx->tweak_ctx,
+		.tweak_fn = XTS_TWEAK_CAST(camellia_enc_blk),
+		.crypt_ctx = &crypt_ctx,
+		.crypt_fn = encrypt_callback,
+	};
+	int ret;
+
+	desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
+	ret = xts_crypt(desc, dst, src, nbytes, &req);
+	camellia_fpu_end(crypt_ctx.fpu_enabled);
+
+	return ret;
+}
+
+static int xts_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
+		       struct scatterlist *src, unsigned int nbytes)
+{
+	struct camellia_xts_ctx *ctx = crypto_blkcipher_ctx(desc->tfm);
+	be128 buf[CAMELLIA_AESNI_PARALLEL_BLOCKS];
+	struct crypt_priv crypt_ctx = {
+		.ctx = &ctx->crypt_ctx,
+		.fpu_enabled = false,
+	};
+	struct xts_crypt_req req = {
+		.tbuf = buf,
+		.tbuflen = sizeof(buf),
+
+		.tweak_ctx = &ctx->tweak_ctx,
+		.tweak_fn = XTS_TWEAK_CAST(camellia_enc_blk),
+		.crypt_ctx = &crypt_ctx,
+		.crypt_fn = decrypt_callback,
+	};
+	int ret;
+
+	desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
+	ret = xts_crypt(desc, dst, src, nbytes, &req);
+	camellia_fpu_end(crypt_ctx.fpu_enabled);
+
+	return ret;
+}
+
+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_blocksize		= CAMELLIA_BLOCK_SIZE,
+	.cra_ctxsize		= sizeof(struct camellia_ctx),
+	.cra_alignmask		= 0,
+	.cra_type		= &crypto_blkcipher_type,
+	.cra_module		= THIS_MODULE,
+	.cra_u = {
+		.blkcipher = {
+			.min_keysize	= CAMELLIA_MIN_KEY_SIZE,
+			.max_keysize	= CAMELLIA_MAX_KEY_SIZE,
+			.setkey		= camellia_setkey,
+			.encrypt	= ecb_encrypt,
+			.decrypt	= ecb_decrypt,
+		},
+	},
+}, {
+	.cra_name		= "__cbc-camellia-aesni",
+	.cra_driver_name	= "__driver-cbc-camellia-aesni",
+	.cra_priority		= 0,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
+	.cra_ctxsize		= sizeof(struct camellia_ctx),
+	.cra_alignmask		= 0,
+	.cra_type		= &crypto_blkcipher_type,
+	.cra_module		= THIS_MODULE,
+	.cra_u = {
+		.blkcipher = {
+			.min_keysize	= CAMELLIA_MIN_KEY_SIZE,
+			.max_keysize	= CAMELLIA_MAX_KEY_SIZE,
+			.setkey		= camellia_setkey,
+			.encrypt	= cbc_encrypt,
+			.decrypt	= cbc_decrypt,
+		},
+	},
+}, {
+	.cra_name		= "__ctr-camellia-aesni",
+	.cra_driver_name	= "__driver-ctr-camellia-aesni",
+	.cra_priority		= 0,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_blocksize		= 1,
+	.cra_ctxsize		= sizeof(struct camellia_ctx),
+	.cra_alignmask		= 0,
+	.cra_type		= &crypto_blkcipher_type,
+	.cra_module		= THIS_MODULE,
+	.cra_u = {
+		.blkcipher = {
+			.min_keysize	= CAMELLIA_MIN_KEY_SIZE,
+			.max_keysize	= CAMELLIA_MAX_KEY_SIZE,
+			.ivsize		= CAMELLIA_BLOCK_SIZE,
+			.setkey		= camellia_setkey,
+			.encrypt	= ctr_crypt,
+			.decrypt	= ctr_crypt,
+		},
+	},
+}, {
+	.cra_name		= "__lrw-camellia-aesni",
+	.cra_driver_name	= "__driver-lrw-camellia-aesni",
+	.cra_priority		= 0,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
+	.cra_ctxsize		= sizeof(struct camellia_lrw_ctx),
+	.cra_alignmask		= 0,
+	.cra_type		= &crypto_blkcipher_type,
+	.cra_module		= THIS_MODULE,
+	.cra_exit		= lrw_camellia_exit_tfm,
+	.cra_u = {
+		.blkcipher = {
+			.min_keysize	= CAMELLIA_MIN_KEY_SIZE +
+					  CAMELLIA_BLOCK_SIZE,
+			.max_keysize	= CAMELLIA_MAX_KEY_SIZE +
+					  CAMELLIA_BLOCK_SIZE,
+			.ivsize		= CAMELLIA_BLOCK_SIZE,
+			.setkey		= lrw_camellia_setkey,
+			.encrypt	= lrw_encrypt,
+			.decrypt	= lrw_decrypt,
+		},
+	},
+}, {
+	.cra_name		= "__xts-camellia-aesni",
+	.cra_driver_name	= "__driver-xts-camellia-aesni",
+	.cra_priority		= 0,
+	.cra_flags		= CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
+	.cra_ctxsize		= sizeof(struct camellia_xts_ctx),
+	.cra_alignmask		= 0,
+	.cra_type		= &crypto_blkcipher_type,
+	.cra_module		= THIS_MODULE,
+	.cra_u = {
+		.blkcipher = {
+			.min_keysize	= CAMELLIA_MIN_KEY_SIZE * 2,
+			.max_keysize	= CAMELLIA_MAX_KEY_SIZE * 2,
+			.ivsize		= CAMELLIA_BLOCK_SIZE,
+			.setkey		= xts_camellia_setkey,
+			.encrypt	= xts_encrypt,
+			.decrypt	= xts_decrypt,
+		},
+	},
+}, {
+	.cra_name		= "ecb(camellia)",
+	.cra_driver_name	= "ecb-camellia-aesni",
+	.cra_priority		= 400,
+	.cra_flags		= CRYPTO_ALG_TYPE_ABLKCIPHER | CRYPTO_ALG_ASYNC,
+	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
+	.cra_ctxsize		= sizeof(struct async_helper_ctx),
+	.cra_alignmask		= 0,
+	.cra_type		= &crypto_ablkcipher_type,
+	.cra_module		= THIS_MODULE,
+	.cra_init		= ablk_init,
+	.cra_exit		= ablk_exit,
+	.cra_u = {
+		.ablkcipher = {
+			.min_keysize	= CAMELLIA_MIN_KEY_SIZE,
+			.max_keysize	= CAMELLIA_MAX_KEY_SIZE,
+			.setkey		= ablk_set_key,
+			.encrypt	= ablk_encrypt,
+			.decrypt	= ablk_decrypt,
+		},
+	},
+}, {
+	.cra_name		= "cbc(camellia)",
+	.cra_driver_name	= "cbc-camellia-aesni",
+	.cra_priority		= 400,
+	.cra_flags		= CRYPTO_ALG_TYPE_ABLKCIPHER | CRYPTO_ALG_ASYNC,
+	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
+	.cra_ctxsize		= sizeof(struct async_helper_ctx),
+	.cra_alignmask		= 0,
+	.cra_type		= &crypto_ablkcipher_type,
+	.cra_module		= THIS_MODULE,
+	.cra_init		= ablk_init,
+	.cra_exit		= ablk_exit,
+	.cra_u = {
+		.ablkcipher = {
+			.min_keysize	= CAMELLIA_MIN_KEY_SIZE,
+			.max_keysize	= CAMELLIA_MAX_KEY_SIZE,
+			.ivsize		= CAMELLIA_BLOCK_SIZE,
+			.setkey		= ablk_set_key,
+			.encrypt	= __ablk_encrypt,
+			.decrypt	= ablk_decrypt,
+		},
+	},
+}, {
+	.cra_name		= "ctr(camellia)",
+	.cra_driver_name	= "ctr-camellia-aesni",
+	.cra_priority		= 400,
+	.cra_flags		= CRYPTO_ALG_TYPE_ABLKCIPHER | CRYPTO_ALG_ASYNC,
+	.cra_blocksize		= 1,
+	.cra_ctxsize		= sizeof(struct async_helper_ctx),
+	.cra_alignmask		= 0,
+	.cra_type		= &crypto_ablkcipher_type,
+	.cra_module		= THIS_MODULE,
+	.cra_init		= ablk_init,
+	.cra_exit		= ablk_exit,
+	.cra_u = {
+		.ablkcipher = {
+			.min_keysize	= CAMELLIA_MIN_KEY_SIZE,
+			.max_keysize	= CAMELLIA_MAX_KEY_SIZE,
+			.ivsize		= CAMELLIA_BLOCK_SIZE,
+			.setkey		= ablk_set_key,
+			.encrypt	= ablk_encrypt,
+			.decrypt	= ablk_encrypt,
+			.geniv		= "chainiv",
+		},
+	},
+}, {
+	.cra_name		= "lrw(camellia)",
+	.cra_driver_name	= "lrw-camellia-aesni",
+	.cra_priority		= 400,
+	.cra_flags		= CRYPTO_ALG_TYPE_ABLKCIPHER | CRYPTO_ALG_ASYNC,
+	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
+	.cra_ctxsize		= sizeof(struct async_helper_ctx),
+	.cra_alignmask		= 0,
+	.cra_type		= &crypto_ablkcipher_type,
+	.cra_module		= THIS_MODULE,
+	.cra_init		= ablk_init,
+	.cra_exit		= ablk_exit,
+	.cra_u = {
+		.ablkcipher = {
+			.min_keysize	= CAMELLIA_MIN_KEY_SIZE +
+					  CAMELLIA_BLOCK_SIZE,
+			.max_keysize	= CAMELLIA_MAX_KEY_SIZE +
+					  CAMELLIA_BLOCK_SIZE,
+			.ivsize		= CAMELLIA_BLOCK_SIZE,
+			.setkey		= ablk_set_key,
+			.encrypt	= ablk_encrypt,
+			.decrypt	= ablk_decrypt,
+		},
+	},
+}, {
+	.cra_name		= "xts(camellia)",
+	.cra_driver_name	= "xts-camellia-aesni",
+	.cra_priority		= 400,
+	.cra_flags		= CRYPTO_ALG_TYPE_ABLKCIPHER | CRYPTO_ALG_ASYNC,
+	.cra_blocksize		= CAMELLIA_BLOCK_SIZE,
+	.cra_ctxsize		= sizeof(struct async_helper_ctx),
+	.cra_alignmask		= 0,
+	.cra_type		= &crypto_ablkcipher_type,
+	.cra_module		= THIS_MODULE,
+	.cra_init		= ablk_init,
+	.cra_exit		= ablk_exit,
+	.cra_u = {
+		.ablkcipher = {
+			.min_keysize	= CAMELLIA_MIN_KEY_SIZE * 2,
+			.max_keysize	= CAMELLIA_MAX_KEY_SIZE * 2,
+			.ivsize		= CAMELLIA_BLOCK_SIZE,
+			.setkey		= ablk_set_key,
+			.encrypt	= ablk_encrypt,
+			.decrypt	= ablk_decrypt,
+		},
+	},
+} };
+
+static int __init camellia_aesni_init(void)
+{
+	u64 xcr0;
+
+	if (!cpu_has_avx || !cpu_has_aes || !cpu_has_osxsave) {
+		pr_info("AVX or AES-NI instructions are not detected.\n");
+		return -ENODEV;
+	}
+
+	xcr0 = xgetbv(XCR_XFEATURE_ENABLED_MASK);
+	if ((xcr0 & (XSTATE_SSE | XSTATE_YMM)) != (XSTATE_SSE | XSTATE_YMM)) {
+		pr_info("AVX detected but unusable.\n");
+		return -ENODEV;
+	}
+
+	return crypto_register_algs(cmll_algs, ARRAY_SIZE(cmll_algs));
+}
+
+static void __exit camellia_aesni_fini(void)
+{
+	crypto_unregister_algs(cmll_algs, ARRAY_SIZE(cmll_algs));
+}
+
+module_init(camellia_aesni_init);
+module_exit(camellia_aesni_fini);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Camellia Cipher Algorithm, AES-NI/AVX optimized");
+MODULE_ALIAS("camellia");
+MODULE_ALIAS("camellia-asm");