[CRYPTO] salsa20: Salsa20 stream cipher

This patch implements the Salsa20 stream cipher using the blkcipher interface.

The core cipher code comes from Daniel Bernstein's submission to eSTREAM:
  http://www.ecrypt.eu.org/stream/svn/viewcvs.cgi/ecrypt/trunk/submissions/salsa20/full/ref/

The test vectors comes from:
  http://www.ecrypt.eu.org/stream/svn/viewcvs.cgi/ecrypt/trunk/submissions/salsa20/full/

It has been tested successfully with "modprobe tcrypt mode=34" on an
UML instance.

Signed-off-by: Tan Swee Heng <thesweeheng@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

crypto/Kconfig

@@ -454,6 +454,18 @@ config CRYPTO_SEED
 	  See also:
 	  <http://www.kisa.or.kr/kisa/seed/jsp/seed_eng.jsp>
 
+config CRYPTO_SALSA20
+	tristate "Salsa20 stream cipher algorithm (EXPERIMENTAL)"
+	depends on EXPERIMENTAL
+	select CRYPTO_BLKCIPHER
+	help
+	  Salsa20 stream cipher algorithm.
+
+	  Salsa20 is a stream cipher submitted to eSTREAM, the ECRYPT
+	  Stream Cipher Project. See <http://www.ecrypt.eu.org/stream/>
+
+	  The Salsa20 stream cipher algorithm is designed by Daniel J.
+	  Bernstein <djb@cr.yp.to>. See <http://cr.yp.to/snuffle.html>
 
 config CRYPTO_DEFLATE
 	tristate "Deflate compression algorithm"

crypto/Makefile

@@ -49,6 +49,7 @@ obj-$(CONFIG_CRYPTO_TEA) += tea.o
 obj-$(CONFIG_CRYPTO_KHAZAD) += khazad.o
 obj-$(CONFIG_CRYPTO_ANUBIS) += anubis.o
 obj-$(CONFIG_CRYPTO_SEED) += seed.o
+obj-$(CONFIG_CRYPTO_SALSA20) += salsa20_generic.o
 obj-$(CONFIG_CRYPTO_DEFLATE) += deflate.o
 obj-$(CONFIG_CRYPTO_MICHAEL_MIC) += michael_mic.o
 obj-$(CONFIG_CRYPTO_CRC32C) += crc32c.o

crypto/salsa20_generic.c

@@ -0,0 +1,243 @@
+/*
+ * Salsa20: Salsa20 stream cipher algorithm
+ *
+ * Copyright (c) 2007 Tan Swee Heng <thesweeheng@gmail.com>
+ *
+ * Derived from:
+ * - salsa20.c: Public domain C code by Daniel J. Bernstein <djb@cr.yp.to>
+ *
+ * Salsa20 is a stream cipher candidate in eSTREAM, the ECRYPT Stream
+ * Cipher Project. It is designed by Daniel J. Bernstein <djb@cr.yp.to>.
+ * More information about eSTREAM and Salsa20 can be found here:
+ *   http://www.ecrypt.eu.org/stream/
+ *   http://cr.yp.to/snuffle.html
+ *
+ * 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/init.h>
+#include <linux/module.h>
+#include <linux/errno.h>
+#include <linux/crypto.h>
+#include <linux/types.h>
+#include <crypto/algapi.h>
+#include <asm/byteorder.h>
+
+#define SALSA20_IV_SIZE        8U
+#define SALSA20_MIN_KEY_SIZE  16U
+#define SALSA20_MAX_KEY_SIZE  32U
+
+/*
+ * Start of code taken from D. J. Bernstein's reference implementation.
+ * With some modifications and optimizations made to suit our needs.
+ */
+
+/*
+salsa20-ref.c version 20051118
+D. J. Bernstein
+Public domain.
+*/
+
+#define ROTATE(v,n) (((v) << (n)) | ((v) >> (32 - (n))))
+#define XOR(v,w) ((v) ^ (w))
+#define PLUS(v,w) (((v) + (w)))
+#define PLUSONE(v) (PLUS((v),1))
+#define U32TO8_LITTLE(p, v) \
+	{ (p)[0] = (v >>  0) & 0xff; (p)[1] = (v >>  8) & 0xff; \
+	  (p)[2] = (v >> 16) & 0xff; (p)[3] = (v >> 24) & 0xff; }
+#define U8TO32_LITTLE(p)   \
+	(((u32)((p)[0])      ) | ((u32)((p)[1]) <<  8) | \
+	 ((u32)((p)[2]) << 16) | ((u32)((p)[3]) << 24)   )
+
+struct salsa20_ctx
+{
+	u32 input[16];
+};
+
+static void salsa20_wordtobyte(u8 output[64], const u32 input[16])
+{
+	u32 x[16];
+	int i;
+
+	memcpy(x, input, sizeof(x));
+	for (i = 20; i > 0; i -= 2) {
+		x[ 4] = XOR(x[ 4],ROTATE(PLUS(x[ 0],x[12]), 7));
+		x[ 8] = XOR(x[ 8],ROTATE(PLUS(x[ 4],x[ 0]), 9));
+		x[12] = XOR(x[12],ROTATE(PLUS(x[ 8],x[ 4]),13));
+		x[ 0] = XOR(x[ 0],ROTATE(PLUS(x[12],x[ 8]),18));
+		x[ 9] = XOR(x[ 9],ROTATE(PLUS(x[ 5],x[ 1]), 7));
+		x[13] = XOR(x[13],ROTATE(PLUS(x[ 9],x[ 5]), 9));
+		x[ 1] = XOR(x[ 1],ROTATE(PLUS(x[13],x[ 9]),13));
+		x[ 5] = XOR(x[ 5],ROTATE(PLUS(x[ 1],x[13]),18));
+		x[14] = XOR(x[14],ROTATE(PLUS(x[10],x[ 6]), 7));
+		x[ 2] = XOR(x[ 2],ROTATE(PLUS(x[14],x[10]), 9));
+		x[ 6] = XOR(x[ 6],ROTATE(PLUS(x[ 2],x[14]),13));
+		x[10] = XOR(x[10],ROTATE(PLUS(x[ 6],x[ 2]),18));
+		x[ 3] = XOR(x[ 3],ROTATE(PLUS(x[15],x[11]), 7));
+		x[ 7] = XOR(x[ 7],ROTATE(PLUS(x[ 3],x[15]), 9));
+		x[11] = XOR(x[11],ROTATE(PLUS(x[ 7],x[ 3]),13));
+		x[15] = XOR(x[15],ROTATE(PLUS(x[11],x[ 7]),18));
+		x[ 1] = XOR(x[ 1],ROTATE(PLUS(x[ 0],x[ 3]), 7));
+		x[ 2] = XOR(x[ 2],ROTATE(PLUS(x[ 1],x[ 0]), 9));
+		x[ 3] = XOR(x[ 3],ROTATE(PLUS(x[ 2],x[ 1]),13));
+		x[ 0] = XOR(x[ 0],ROTATE(PLUS(x[ 3],x[ 2]),18));
+		x[ 6] = XOR(x[ 6],ROTATE(PLUS(x[ 5],x[ 4]), 7));
+		x[ 7] = XOR(x[ 7],ROTATE(PLUS(x[ 6],x[ 5]), 9));
+		x[ 4] = XOR(x[ 4],ROTATE(PLUS(x[ 7],x[ 6]),13));
+		x[ 5] = XOR(x[ 5],ROTATE(PLUS(x[ 4],x[ 7]),18));
+		x[11] = XOR(x[11],ROTATE(PLUS(x[10],x[ 9]), 7));
+		x[ 8] = XOR(x[ 8],ROTATE(PLUS(x[11],x[10]), 9));
+		x[ 9] = XOR(x[ 9],ROTATE(PLUS(x[ 8],x[11]),13));
+		x[10] = XOR(x[10],ROTATE(PLUS(x[ 9],x[ 8]),18));
+		x[12] = XOR(x[12],ROTATE(PLUS(x[15],x[14]), 7));
+		x[13] = XOR(x[13],ROTATE(PLUS(x[12],x[15]), 9));
+		x[14] = XOR(x[14],ROTATE(PLUS(x[13],x[12]),13));
+		x[15] = XOR(x[15],ROTATE(PLUS(x[14],x[13]),18));
+	}
+	for (i = 0; i < 16; ++i)
+		x[i] = PLUS(x[i],input[i]);
+	for (i = 0; i < 16; ++i)
+		U32TO8_LITTLE(output + 4 * i,x[i]);
+}
+
+static const char sigma[16] = "expand 32-byte k";
+static const char tau[16] = "expand 16-byte k";
+
+static void salsa20_keysetup(struct salsa20_ctx *ctx, const u8 *k, u32 kbytes)
+{
+	const char *constants;
+
+	ctx->input[1] = U8TO32_LITTLE(k + 0);
+	ctx->input[2] = U8TO32_LITTLE(k + 4);
+	ctx->input[3] = U8TO32_LITTLE(k + 8);
+	ctx->input[4] = U8TO32_LITTLE(k + 12);
+	if (kbytes == 32) { /* recommended */
+		k += 16;
+		constants = sigma;
+	} else { /* kbytes == 16 */
+		constants = tau;
+	}
+	ctx->input[11] = U8TO32_LITTLE(k + 0);
+	ctx->input[12] = U8TO32_LITTLE(k + 4);
+	ctx->input[13] = U8TO32_LITTLE(k + 8);
+	ctx->input[14] = U8TO32_LITTLE(k + 12);
+	ctx->input[0] = U8TO32_LITTLE(constants + 0);
+	ctx->input[5] = U8TO32_LITTLE(constants + 4);
+	ctx->input[10] = U8TO32_LITTLE(constants + 8);
+	ctx->input[15] = U8TO32_LITTLE(constants + 12);
+}
+
+static void salsa20_ivsetup(struct salsa20_ctx *ctx, const u8 *iv)
+{
+	ctx->input[6] = U8TO32_LITTLE(iv + 0);
+	ctx->input[7] = U8TO32_LITTLE(iv + 4);
+	ctx->input[8] = 0;
+	ctx->input[9] = 0;
+}
+
+static void salsa20_encrypt_bytes(struct salsa20_ctx *ctx, u8 *dst,
+				  const u8 *src, unsigned int bytes)
+{
+	u8 buf[64];
+	int i;
+
+	if (dst != src)
+		memcpy(dst, src, bytes);
+
+	while (bytes) {
+		salsa20_wordtobyte(buf, ctx->input);
+
+		ctx->input[8] = PLUSONE(ctx->input[8]);
+		if (!ctx->input[8])
+			ctx->input[9] = PLUSONE(ctx->input[9]);
+
+		if (bytes <= 64) {
+			for (i = 0; i < bytes/4; ++i)
+				((u32*)dst)[i] ^= ((u32*)buf)[i];
+			for (i = bytes - bytes % 4; i < bytes; ++i)
+				dst[i] ^= buf[i];
+			return;
+		}
+
+		for (i = 0; i < 64/4; ++i)
+			((u32*)dst)[i] ^= ((u32*)buf)[i];
+		bytes -= 64;
+		dst += 64;
+	}
+}
+
+/*
+ * End of code taken from D. J. Bernstein's reference implementation.
+ */
+
+static int setkey(struct crypto_tfm *tfm, const u8 *key,
+		  unsigned int keysize)
+{
+	struct salsa20_ctx *ctx = crypto_tfm_ctx(tfm);
+	salsa20_keysetup(ctx, key, keysize);
+	return 0;
+}
+
+static int encrypt(struct blkcipher_desc *desc,
+		   struct scatterlist *dst, struct scatterlist *src,
+		   unsigned int nbytes)
+{
+	struct blkcipher_walk walk;
+	struct crypto_blkcipher *tfm = desc->tfm;
+	struct salsa20_ctx *ctx = crypto_blkcipher_ctx(tfm);
+	int err;
+
+	blkcipher_walk_init(&walk, dst, src, nbytes);
+	err = blkcipher_walk_virt(desc, &walk);
+
+	salsa20_ivsetup(ctx, walk.iv);
+	salsa20_encrypt_bytes(ctx, walk.dst.virt.addr,
+			      walk.src.virt.addr, nbytes);
+
+	err = blkcipher_walk_done(desc, &walk, 0);
+	return err;
+}
+
+static struct crypto_alg alg = {
+	.cra_name           =   "salsa20",
+	.cra_driver_name    =   "salsa20-generic",
+	.cra_priority       =   100,
+	.cra_flags          =   CRYPTO_ALG_TYPE_BLKCIPHER,
+	.cra_type           =   &crypto_blkcipher_type,
+	.cra_blocksize      =   1,
+	.cra_ctxsize        =   sizeof(struct salsa20_ctx),
+	.cra_alignmask      =	3,
+	.cra_module         =   THIS_MODULE,
+	.cra_list           =   LIST_HEAD_INIT(alg.cra_list),
+	.cra_u              =   {
+		.blkcipher = {
+			.setkey         =   setkey,
+			.encrypt        =   encrypt,
+			.decrypt        =   encrypt,
+			.min_keysize    =   SALSA20_MIN_KEY_SIZE,
+			.max_keysize    =   SALSA20_MAX_KEY_SIZE,
+			.ivsize         =   SALSA20_IV_SIZE,
+		}
+	}
+};
+
+static int __init init(void)
+{
+	return crypto_register_alg(&alg);
+}
+
+static void __exit fini(void)
+{
+	crypto_unregister_alg(&alg);
+}
+
+module_init(init);
+module_exit(fini);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION ("Salsa20 stream cipher algorithm");
+MODULE_ALIAS("salsa20");

crypto/tcrypt.c

@@ -80,7 +80,7 @@ static char *check[] = {
 	"cast6", "arc4", "michael_mic", "deflate", "crc32c", "tea", "xtea",
 	"arc4", "michael_mic", "deflate", "crc32c", "tea", "xtea",
 	"khazad", "wp512", "wp384", "wp256", "tnepres", "xeta",  "fcrypt",
-	"camellia", "seed", NULL
+	"camellia", "seed", "salsa20", NULL
 };
 
 static void hexdump(unsigned char *buf, unsigned int len)
@@ -1309,6 +1309,12 @@ static void do_test(void)
 		test_hash("sha224", sha224_tv_template, SHA224_TEST_VECTORS);
 		break;
 
+	case 34:
+		test_cipher("salsa20", ENCRYPT,
+			    salsa20_stream_enc_tv_template,
+			    SALSA20_STREAM_ENC_TEST_VECTORS);
+		break;
+
 	case 100:
 		test_hash("hmac(md5)", hmac_md5_tv_template,
 			  HMAC_MD5_TEST_VECTORS);