[CRYPTO] xts: XTS blockcipher mode implementation without partial blocks

XTS currently considered to be the successor of the LRW mode by the IEEE1619
workgroup. LRW was discarded, because it was not secure if the encyption key
itself is encrypted with LRW.

XTS does not have this problem. The implementation is pretty straightforward,
a new function was added to gf128mul to handle GF(128) elements in ble format.
Four testvectors from the specification
	http://grouper.ieee.org/groups/1619/email/pdf00086.pdf
were added, and they verify on my system.
Signed-off-by: Rik Snel <rsnel@cube.dyndns.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

crypto/Kconfig

@@ -184,6 +184,17 @@ config CRYPTO_LRW
 	  The first 128, 192 or 256 bits in the key are used for AES and the
 	  rest is used to tie each cipher block to its logical position.
 
+config CRYPTO_XTS
+	tristate "XTS support (EXPERIMENTAL)"
+	depends on EXPERIMENTAL
+	select CRYPTO_BLKCIPHER
+	select CRYPTO_MANAGER
+	select CRYPTO_GF128MUL
+	help
+	  XTS: IEEE1619/D16 narrow block cipher use with aes-xts-plain,
+	  key size 256, 384 or 512 bits. This implementation currently
+	  can't handle a sectorsize which is not a multiple of 16 bytes.
+
 config CRYPTO_CRYPTD
 	tristate "Software async crypto daemon"
 	select CRYPTO_ABLKCIPHER

crypto/Makefile

@@ -31,6 +31,7 @@ obj-$(CONFIG_CRYPTO_ECB) += ecb.o
 obj-$(CONFIG_CRYPTO_CBC) += cbc.o
 obj-$(CONFIG_CRYPTO_PCBC) += pcbc.o
 obj-$(CONFIG_CRYPTO_LRW) += lrw.o
+obj-$(CONFIG_CRYPTO_XTS) += xts.o
 obj-$(CONFIG_CRYPTO_CRYPTD) += cryptd.o
 obj-$(CONFIG_CRYPTO_DES) += des.o
 obj-$(CONFIG_CRYPTO_FCRYPT) += fcrypt.o

crypto/gf128mul.c

@@ -142,6 +142,17 @@ static void gf128mul_x_bbe(be128 *r, const be128 *x)
 	r->b = cpu_to_be64((b << 1) ^ _tt);
 }
 
+void gf128mul_x_ble(be128 *r, const be128 *x)
+{
+	u64 a = le64_to_cpu(x->a);
+	u64 b = le64_to_cpu(x->b);
+	u64 _tt = gf128mul_table_bbe[b >> 63];
+
+	r->a = cpu_to_le64((a << 1) ^ _tt);
+	r->b = cpu_to_le64((b << 1) | (a >> 63));
+}
+EXPORT_SYMBOL(gf128mul_x_ble);
+
 static void gf128mul_x8_lle(be128 *x)
 {
 	u64 a = be64_to_cpu(x->a);

crypto/tcrypt.c

@@ -955,6 +955,10 @@ static void do_test(void)
 			    AES_LRW_ENC_TEST_VECTORS);
 		test_cipher("lrw(aes)", DECRYPT, aes_lrw_dec_tv_template,
 			    AES_LRW_DEC_TEST_VECTORS);
+		test_cipher("xts(aes)", ENCRYPT, aes_xts_enc_tv_template,
+			    AES_XTS_ENC_TEST_VECTORS);
+		test_cipher("xts(aes)", DECRYPT, aes_xts_dec_tv_template,
+			    AES_XTS_DEC_TEST_VECTORS);
 
 		//CAST5
 		test_cipher("ecb(cast5)", ENCRYPT, cast5_enc_tv_template,
@@ -1138,6 +1142,10 @@ static void do_test(void)
 			    AES_LRW_ENC_TEST_VECTORS);
 		test_cipher("lrw(aes)", DECRYPT, aes_lrw_dec_tv_template,
 			    AES_LRW_DEC_TEST_VECTORS);
+		test_cipher("xts(aes)", ENCRYPT, aes_xts_enc_tv_template,
+			    AES_XTS_ENC_TEST_VECTORS);
+		test_cipher("xts(aes)", DECRYPT, aes_xts_dec_tv_template,
+			    AES_XTS_DEC_TEST_VECTORS);
 		break;
 
 	case 11:
@@ -1313,6 +1321,10 @@ static void do_test(void)
 				  aes_lrw_speed_template);
 		test_cipher_speed("lrw(aes)", DECRYPT, sec, NULL, 0,
 				  aes_lrw_speed_template);
+		test_cipher_speed("xts(aes)", ENCRYPT, sec, NULL, 0,
+				  aes_xts_speed_template);
+		test_cipher_speed("xts(aes)", DECRYPT, sec, NULL, 0,
+				  aes_xts_speed_template);
 		break;
 
 	case 201:

crypto/xts.c

+/* XTS: as defined in IEEE1619/D16
+ *	http://grouper.ieee.org/groups/1619/email/pdf00086.pdf
+ *	(sector sizes which are not a multiple of 16 bytes are,
+ *	however currently unsupported)
+ *
+ * Copyright (c) 2007 Rik Snel <rsnel@cube.dyndns.org>
+ *
+ * Based om ecb.c
+ * Copyright (c) 2006 Herbert Xu <herbert@gondor.apana.org.au>
+ *
+ * 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 <crypto/algapi.h>
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/scatterlist.h>
+#include <linux/slab.h>
+
+#include <crypto/b128ops.h>
+#include <crypto/gf128mul.h>
+
+struct priv {
+	struct crypto_cipher *child;
+	struct crypto_cipher *tweak;
+};
+
+static int setkey(struct crypto_tfm *parent, const u8 *key,
+		  unsigned int keylen)
+{
+	struct priv *ctx = crypto_tfm_ctx(parent);
+	struct crypto_cipher *child = ctx->tweak;
+	u32 *flags = &parent->crt_flags;
+	int err;
+
+	/* key consists of keys of equal size concatenated, therefore
+	 * the length must be even */
+	if (keylen % 2) {
+		/* tell the user why there was an error */
+		*flags |= CRYPTO_TFM_RES_BAD_KEY_LEN;
+		return -EINVAL;
+	}
+
+	/* we need two cipher instances: one to compute the inital 'tweak'
+	 * by encrypting the IV (usually the 'plain' iv) and the other
+	 * one to encrypt and decrypt the data */
+
+	/* tweak cipher, uses Key2 i.e. the second half of *key */
+	crypto_cipher_clear_flags(child, CRYPTO_TFM_REQ_MASK);
+	crypto_cipher_set_flags(child, crypto_tfm_get_flags(parent) &
+				       CRYPTO_TFM_REQ_MASK);
+	err = crypto_cipher_setkey(child, key + keylen/2, keylen/2);
+	if (err)
+		return err;
+
+	crypto_tfm_set_flags(parent, crypto_cipher_get_flags(child) &
+				     CRYPTO_TFM_RES_MASK);
+
+	child = ctx->child;
+
+	/* data cipher, uses Key1 i.e. the first half of *key */
+	crypto_cipher_clear_flags(child, CRYPTO_TFM_REQ_MASK);
+	crypto_cipher_set_flags(child, crypto_tfm_get_flags(parent) &
+				       CRYPTO_TFM_REQ_MASK);
+	err = crypto_cipher_setkey(child, key, keylen/2);
+	if (err)
+		return err;
+
+	crypto_tfm_set_flags(parent, crypto_cipher_get_flags(child) &
+				     CRYPTO_TFM_RES_MASK);
+
+	return 0;
+}
+
+struct sinfo {
+	be128 t;
+	struct crypto_tfm *tfm;
+	void (*fn)(struct crypto_tfm *, u8 *, const u8 *);
+};
+
+static inline void xts_round(struct sinfo *s, void *dst, const void *src)
+{
+	be128_xor(dst, &s->t, src);		/* PP <- T xor P */
+	s->fn(s->tfm, dst, dst);		/* CC <- E(Key1,PP) */
+	be128_xor(dst, dst, &s->t);		/* C <- T xor CC */
+}
+
+static int crypt(struct blkcipher_desc *d,
+		 struct blkcipher_walk *w, struct priv *ctx,
+		 void (*tw)(struct crypto_tfm *, u8 *, const u8 *),
+		 void (*fn)(struct crypto_tfm *, u8 *, const u8 *))
+{
+	int err;
+	unsigned int avail;
+	const int bs = crypto_cipher_blocksize(ctx->child);
+	struct sinfo s = {
+		.tfm = crypto_cipher_tfm(ctx->child),
+		.fn = fn
+	};
+	be128 *iv;
+	u8 *wsrc;
+	u8 *wdst;
+
+	err = blkcipher_walk_virt(d, w);
+	if (!w->nbytes)
+		return err;
+
+	avail = w->nbytes;
+
+	wsrc = w->src.virt.addr;
+	wdst = w->dst.virt.addr;
+
+	/* calculate first value of T */
+	iv = (be128 *)w->iv;
+	tw(crypto_cipher_tfm(ctx->tweak), (void *)&s.t, w->iv);
+
+	goto first;
+
+	for (;;) {
+		do {
+			gf128mul_x_ble(&s.t, &s.t);
+
+first:
+			xts_round(&s, wdst, wsrc);
+
+			wsrc += bs;
+			wdst += bs;
+		} while ((avail -= bs) >= bs);
+
+		err = blkcipher_walk_done(d, w, avail);
+		if (!w->nbytes)
+			break;
+
+		avail = w->nbytes;
+
+		wsrc = w->src.virt.addr;
+		wdst = w->dst.virt.addr;
+	}
+
+	return err;
+}
+
+static int encrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
+		   struct scatterlist *src, unsigned int nbytes)
+{
+	struct priv *ctx = crypto_blkcipher_ctx(desc->tfm);
+	struct blkcipher_walk w;
+
+	blkcipher_walk_init(&w, dst, src, nbytes);
+	return crypt(desc, &w, ctx, crypto_cipher_alg(ctx->tweak)->cia_encrypt,
+		     crypto_cipher_alg(ctx->child)->cia_encrypt);
+}
+
+static int decrypt(struct blkcipher_desc *desc, struct scatterlist *dst,
+		   struct scatterlist *src, unsigned int nbytes)
+{
+	struct priv *ctx = crypto_blkcipher_ctx(desc->tfm);
+	struct blkcipher_walk w;
+
+	blkcipher_walk_init(&w, dst, src, nbytes);
+	return crypt(desc, &w, ctx, crypto_cipher_alg(ctx->tweak)->cia_encrypt,
+		     crypto_cipher_alg(ctx->child)->cia_decrypt);
+}
+
+static int init_tfm(struct crypto_tfm *tfm)
+{
+	struct crypto_cipher *cipher;
+	struct crypto_instance *inst = (void *)tfm->__crt_alg;
+	struct crypto_spawn *spawn = crypto_instance_ctx(inst);
+	struct priv *ctx = crypto_tfm_ctx(tfm);
+	u32 *flags = &tfm->crt_flags;
+
+	cipher = crypto_spawn_cipher(spawn);
+	if (IS_ERR(cipher))
+		return PTR_ERR(cipher);
+
+	if (crypto_cipher_blocksize(cipher) != 16) {
+		*flags |= CRYPTO_TFM_RES_BAD_BLOCK_LEN;
+		crypto_free_cipher(cipher);
+		return -EINVAL;
+	}
+
+	ctx->child = cipher;
+
+	cipher = crypto_spawn_cipher(spawn);
+	if (IS_ERR(cipher)) {
+		crypto_free_cipher(ctx->child);
+		return PTR_ERR(cipher);
+	}
+
+	/* this check isn't really needed, leave it here just in case */
+	if (crypto_cipher_blocksize(cipher) != 16) {
+		crypto_free_cipher(cipher);
+		crypto_free_cipher(ctx->child);
+		*flags |= CRYPTO_TFM_RES_BAD_BLOCK_LEN;
+		return -EINVAL;
+	}
+
+	ctx->tweak = cipher;
+
+	return 0;
+}
+
+static void exit_tfm(struct crypto_tfm *tfm)
+{
+	struct priv *ctx = crypto_tfm_ctx(tfm);
+	crypto_free_cipher(ctx->child);
+	crypto_free_cipher(ctx->tweak);
+}
+
+static struct crypto_instance *alloc(struct rtattr **tb)
+{
+	struct crypto_instance *inst;
+	struct crypto_alg *alg;
+	int err;
+
+	err = crypto_check_attr_type(tb, CRYPTO_ALG_TYPE_BLKCIPHER);
+	if (err)
+		return ERR_PTR(err);
+
+	alg = crypto_get_attr_alg(tb, CRYPTO_ALG_TYPE_CIPHER,
+				  CRYPTO_ALG_TYPE_MASK);
+	if (IS_ERR(alg))
+		return ERR_PTR(PTR_ERR(alg));
+
+	inst = crypto_alloc_instance("xts", alg);
+	if (IS_ERR(inst))
+		goto out_put_alg;
+
+	inst->alg.cra_flags = CRYPTO_ALG_TYPE_BLKCIPHER;
+	inst->alg.cra_priority = alg->cra_priority;
+	inst->alg.cra_blocksize = alg->cra_blocksize;
+
+	if (alg->cra_alignmask < 7)
+		inst->alg.cra_alignmask = 7;
+	else
+		inst->alg.cra_alignmask = alg->cra_alignmask;
+
+	inst->alg.cra_type = &crypto_blkcipher_type;
+
+	inst->alg.cra_blkcipher.ivsize = alg->cra_blocksize;
+	inst->alg.cra_blkcipher.min_keysize =
+		2 * alg->cra_cipher.cia_min_keysize;
+	inst->alg.cra_blkcipher.max_keysize =
+		2 * alg->cra_cipher.cia_max_keysize;
+
+	inst->alg.cra_ctxsize = sizeof(struct priv);
+
+	inst->alg.cra_init = init_tfm;
+	inst->alg.cra_exit = exit_tfm;
+
+	inst->alg.cra_blkcipher.setkey = setkey;
+	inst->alg.cra_blkcipher.encrypt = encrypt;
+	inst->alg.cra_blkcipher.decrypt = decrypt;
+
+out_put_alg:
+	crypto_mod_put(alg);
+	return inst;
+}
+
+static void free(struct crypto_instance *inst)
+{
+	crypto_drop_spawn(crypto_instance_ctx(inst));
+	kfree(inst);
+}
+
+static struct crypto_template crypto_tmpl = {
+	.name = "xts",
+	.alloc = alloc,
+	.free = free,
+	.module = THIS_MODULE,
+};
+
+static int __init crypto_module_init(void)
+{
+	return crypto_register_template(&crypto_tmpl);
+}
+
+static void __exit crypto_module_exit(void)
+{
+	crypto_unregister_template(&crypto_tmpl);
+}
+
+module_init(crypto_module_init);
+module_exit(crypto_module_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("XTS block cipher mode");

include/crypto/gf128mul.h

@@ -161,6 +161,8 @@ void gf128mul_lle(be128 *a, const be128 *b);
 
 void gf128mul_bbe(be128 *a, const be128 *b);
 
+/* multiply by x in ble format, needed by XTS */
+void gf128mul_x_ble(be128 *a, const be128 *b);
 
 /* 4k table optimization */