// SPDX-License-Identifier: GPL-2.0-or-later
/*
 * Glue Code for x86_64/AVX2 assembler optimized version of Serpent
 *
 * Copyright © 2012-2013 Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/crypto.h>
#include <linux/err.h>
#include <crypto/algapi.h>
#include <crypto/internal/simd.h>
#include <crypto/serpent.h>
#include <crypto/xts.h>
#include <asm/crypto/glue_helper.h>
#include <asm/crypto/serpent-avx.h>

#define SERPENT_AVX2_PARALLEL_BLOCKS 16

/* 16-way AVX2 parallel cipher functions */
asmlinkage void serpent_ecb_enc_16way(const void *ctx, u8 *dst, const u8 *src);
asmlinkage void serpent_ecb_dec_16way(const void *ctx, u8 *dst, const u8 *src);
asmlinkage void serpent_cbc_dec_16way(const void *ctx, u8 *dst, const u8 *src);

asmlinkage void serpent_ctr_16way(const void *ctx, u8 *dst, const u8 *src,
				  le128 *iv);
asmlinkage void serpent_xts_enc_16way(const void *ctx, u8 *dst, const u8 *src,
				      le128 *iv);
asmlinkage void serpent_xts_dec_16way(const void *ctx, u8 *dst, const u8 *src,
				      le128 *iv);

static int serpent_setkey_skcipher(struct crypto_skcipher *tfm,
				   const u8 *key, unsigned int keylen)
{
	return __serpent_setkey(crypto_skcipher_ctx(tfm), key, keylen);
}

static const struct common_glue_ctx serpent_enc = {
	.num_funcs = 3,
	.fpu_blocks_limit = 8,

	.funcs = { {
		.num_blocks = 16,
		.fn_u = { .ecb = serpent_ecb_enc_16way }
	}, {
		.num_blocks = 8,
		.fn_u = { .ecb = serpent_ecb_enc_8way_avx }
	}, {
		.num_blocks = 1,
		.fn_u = { .ecb = __serpent_encrypt }
	} }
};

static const struct common_glue_ctx serpent_ctr = {
	.num_funcs = 3,
	.fpu_blocks_limit = 8,

	.funcs = { {
		.num_blocks = 16,
		.fn_u = { .ctr = serpent_ctr_16way }
	},  {
		.num_blocks = 8,
		.fn_u = { .ctr = serpent_ctr_8way_avx }
	}, {
		.num_blocks = 1,
		.fn_u = { .ctr = __serpent_crypt_ctr }
	} }
};

static const struct common_glue_ctx serpent_enc_xts = {
	.num_funcs = 3,
	.fpu_blocks_limit = 8,

	.funcs = { {
		.num_blocks = 16,
		.fn_u = { .xts = serpent_xts_enc_16way }
	}, {
		.num_blocks = 8,
		.fn_u = { .xts = serpent_xts_enc_8way_avx }
	}, {
		.num_blocks = 1,
		.fn_u = { .xts = serpent_xts_enc }
	} }
};

static const struct common_glue_ctx serpent_dec = {
	.num_funcs = 3,
	.fpu_blocks_limit = 8,

	.funcs = { {
		.num_blocks = 16,
		.fn_u = { .ecb = serpent_ecb_dec_16way }
	}, {
		.num_blocks = 8,
		.fn_u = { .ecb = serpent_ecb_dec_8way_avx }
	}, {
		.num_blocks = 1,
		.fn_u = { .ecb = __serpent_decrypt }
	} }
};

static const struct common_glue_ctx serpent_dec_cbc = {
	.num_funcs = 3,
	.fpu_blocks_limit = 8,

	.funcs = { {
		.num_blocks = 16,
		.fn_u = { .cbc = serpent_cbc_dec_16way }
	}, {
		.num_blocks = 8,
		.fn_u = { .cbc = serpent_cbc_dec_8way_avx }
	}, {
		.num_blocks = 1,
		.fn_u = { .cbc = __serpent_decrypt }
	} }
};

static const struct common_glue_ctx serpent_dec_xts = {
	.num_funcs = 3,
	.fpu_blocks_limit = 8,

	.funcs = { {
		.num_blocks = 16,
		.fn_u = { .xts = serpent_xts_dec_16way }
	}, {
		.num_blocks = 8,
		.fn_u = { .xts = serpent_xts_dec_8way_avx }
	}, {
		.num_blocks = 1,
		.fn_u = { .xts = serpent_xts_dec }
	} }
};

static int ecb_encrypt(struct skcipher_request *req)
{
	return glue_ecb_req_128bit(&serpent_enc, req);
}

static int ecb_decrypt(struct skcipher_request *req)
{
	return glue_ecb_req_128bit(&serpent_dec, req);
}

static int cbc_encrypt(struct skcipher_request *req)
{
	return glue_cbc_encrypt_req_128bit(__serpent_encrypt, req);
}

static int cbc_decrypt(struct skcipher_request *req)
{
	return glue_cbc_decrypt_req_128bit(&serpent_dec_cbc, req);
}

static int ctr_crypt(struct skcipher_request *req)
{
	return glue_ctr_req_128bit(&serpent_ctr, req);
}

static int xts_encrypt(struct skcipher_request *req)
{
	struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req);
	struct serpent_xts_ctx *ctx = crypto_skcipher_ctx(tfm);

	return glue_xts_req_128bit(&serpent_enc_xts, req,
				   __serpent_encrypt, &ctx->tweak_ctx,
				   &ctx->crypt_ctx, false);
}

static int xts_decrypt(struct skcipher_request *req)
{
	struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req);
	struct serpent_xts_ctx *ctx = crypto_skcipher_ctx(tfm);

	return glue_xts_req_128bit(&serpent_dec_xts, req,
				   __serpent_encrypt, &ctx->tweak_ctx,
				   &ctx->crypt_ctx, true);
}

static struct skcipher_alg serpent_algs[] = {
	{
		.base.cra_name		= "__ecb(serpent)",
		.base.cra_driver_name	= "__ecb-serpent-avx2",
		.base.cra_priority	= 600,
		.base.cra_flags		= CRYPTO_ALG_INTERNAL,
		.base.cra_blocksize	= SERPENT_BLOCK_SIZE,
		.base.cra_ctxsize	= sizeof(struct serpent_ctx),
		.base.cra_module	= THIS_MODULE,
		.min_keysize		= SERPENT_MIN_KEY_SIZE,
		.max_keysize		= SERPENT_MAX_KEY_SIZE,
		.setkey			= serpent_setkey_skcipher,
		.encrypt		= ecb_encrypt,
		.decrypt		= ecb_decrypt,
	}, {
		.base.cra_name		= "__cbc(serpent)",
		.base.cra_driver_name	= "__cbc-serpent-avx2",
		.base.cra_priority	= 600,
		.base.cra_flags		= CRYPTO_ALG_INTERNAL,
		.base.cra_blocksize	= SERPENT_BLOCK_SIZE,
		.base.cra_ctxsize	= sizeof(struct serpent_ctx),
		.base.cra_module	= THIS_MODULE,
		.min_keysize		= SERPENT_MIN_KEY_SIZE,
		.max_keysize		= SERPENT_MAX_KEY_SIZE,
		.ivsize			= SERPENT_BLOCK_SIZE,
		.setkey			= serpent_setkey_skcipher,
		.encrypt		= cbc_encrypt,
		.decrypt		= cbc_decrypt,
	}, {
		.base.cra_name		= "__ctr(serpent)",
		.base.cra_driver_name	= "__ctr-serpent-avx2",
		.base.cra_priority	= 600,
		.base.cra_flags		= CRYPTO_ALG_INTERNAL,
		.base.cra_blocksize	= 1,
		.base.cra_ctxsize	= sizeof(struct serpent_ctx),
		.base.cra_module	= THIS_MODULE,
		.min_keysize		= SERPENT_MIN_KEY_SIZE,
		.max_keysize		= SERPENT_MAX_KEY_SIZE,
		.ivsize			= SERPENT_BLOCK_SIZE,
		.chunksize		= SERPENT_BLOCK_SIZE,
		.setkey			= serpent_setkey_skcipher,
		.encrypt		= ctr_crypt,
		.decrypt		= ctr_crypt,
	}, {
		.base.cra_name		= "__xts(serpent)",
		.base.cra_driver_name	= "__xts-serpent-avx2",
		.base.cra_priority	= 600,
		.base.cra_flags		= CRYPTO_ALG_INTERNAL,
		.base.cra_blocksize	= SERPENT_BLOCK_SIZE,
		.base.cra_ctxsize	= sizeof(struct serpent_xts_ctx),
		.base.cra_module	= THIS_MODULE,
		.min_keysize		= 2 * SERPENT_MIN_KEY_SIZE,
		.max_keysize		= 2 * SERPENT_MAX_KEY_SIZE,
		.ivsize			= SERPENT_BLOCK_SIZE,
		.setkey			= xts_serpent_setkey,
		.encrypt		= xts_encrypt,
		.decrypt		= xts_decrypt,
	},
};

static struct simd_skcipher_alg *serpent_simd_algs[ARRAY_SIZE(serpent_algs)];

static int __init init(void)
{
	const char *feature_name;

	if (!boot_cpu_has(X86_FEATURE_AVX2) || !boot_cpu_has(X86_FEATURE_OSXSAVE)) {
		pr_info("AVX2 instructions are not detected.\n");
		return -ENODEV;
	}
	if (!cpu_has_xfeatures(XFEATURE_MASK_SSE | XFEATURE_MASK_YMM,
				&feature_name)) {
		pr_info("CPU feature '%s' is not supported.\n", feature_name);
		return -ENODEV;
	}

	return simd_register_skciphers_compat(serpent_algs,
					      ARRAY_SIZE(serpent_algs),
					      serpent_simd_algs);
}

static void __exit fini(void)
{
	simd_unregister_skciphers(serpent_algs, ARRAY_SIZE(serpent_algs),
				  serpent_simd_algs);
}

module_init(init);
module_exit(fini);

MODULE_LICENSE("GPL");
MODULE_DESCRIPTION("Serpent Cipher Algorithm, AVX2 optimized");
MODULE_ALIAS_CRYPTO("serpent");
MODULE_ALIAS_CRYPTO("serpent-asm");