Skip to contentSkip to conversation

camellia: 64-bit optimization and code shrink

PATCH In review Denys Vlasenko 2 participants 8 messages Nov 21, 2007 – Nov 22, 2007 Crypto API

lore
Share & tools
lore

Denys Vlasenko

author
#0 lore raw

cover camellia: 64-bit optimization and code shrink

Patches start from 5 because pathches 1..4 are already in cryptodev-2.6.

camellia5:
adds 64-bit key setup, it is used if BITS_PER_LONG is 64.

camellia6:
unifies encrypt/decrypt routines for different key lengths. This reduces module size by ~25%, with tiny (less than 1%) speed impact.
Also collapses encrypt/decrypt into more readable (visually shorter) form using macros.

camellia7:
Move "key XOR is end of F-function" code part into camellia_setup_tail(), it is sufficiently similar between camellia_setup128 and camellia_setup256.

camellia8:
Analogously to camellia7 patch, move
"absorb kw2 to other subkeys" and "absorb kw4 to other subkeys" code parts into camellia_setup_tail(). This further reduces source and object code size at the cost of two brances in key setup code.

Code sizes:
64-bit:

dec      hex   filename
22786    5902  2.6.23.1.camellia4.t64/crypto/camellia.o
21422    53ae  2.6.23.1.camellia5.t64/crypto/camellia.o
16355    3fe3  2.6.23.1.camellia6.t64/crypto/camellia.o
15813    3dc5  2.6.23.1.camellia7.t64/crypto/camellia.o
15670    3d36  2.6.23.1.camellia8.t64/crypto/camellia.o

32-bit:

29948    74fc  2.6.23.1.camellia4.t/crypto/camellia.o
29457    7311  2.6.23.1.camellia5.t/crypto/camellia.o
21414    53a6  2.6.23.1.camellia6.t/crypto/camellia.o
20518    5026  2.6.23.1.camellia7.t/crypto/camellia.o
18454    4816  2.6.23.1.camellia8.t/crypto/camellia.oamellia.o

These patches apply cleanly to current cryptodev-2.6. They should be applied in order.
Compile and run tested.

Signed-off-by Denys Vlasenko <[email protected]>

--
vda

author
#1 lore raw

PATCH 1/4 camellia: 64-bit optimization

Quoted 6 lines · Denys Vlasenko

Patches start from 5 because pathches 1..4 are already in cryptodev-2.6.

camellia5:
adds 64-bit key setup, it is used if BITS_PER_LONG is 64.

Signed-off-by Denys Vlasenko <[email protected]>
Signature · 1 line
vda
crypto/camellia.c +722 -138
diff -urpN cryptodev-2.6/crypto/camellia.c cryptodev-2.6.camellia5/crypto/camellia.c
@@ -310,6 +310,589 @@ static const u32 camellia_sp4404[256] = 
 #define CAMELLIA_BLOCK_SIZE          16
 #define CAMELLIA_TABLE_BYTE_LEN     272
 
+/*
+ * NB: L and R below stand for 'left' and 'right' as in written numbers.
+ * That is, in (xxxL,xxxR) pair xxxL holds most significant digits,
+ * _not_ least significant ones!
+ */
+
+
+
+#if BITS_PER_LONG >= 64
+
+/*
+ * Key setup implementation with mostly 64-bit ops
+ */
+
+/* key constants */
+
+#define CAMELLIA_SIGMA1 (0xA09E667F3BCC908B)
+#define CAMELLIA_SIGMA2 (0xB67AE8584CAA73B2)
+#define CAMELLIA_SIGMA3 (0xC6EF372FE94F82BE)
+#define CAMELLIA_SIGMA4 (0x54FF53A5F1D36F1C)
+#define CAMELLIA_SIGMA5 (0x10E527FADE682D1D)
+#define CAMELLIA_SIGMA6 (0xB05688C2B3E6C1FD)
+
+/*
+ *  macros
+ */
+#define GETU64(v, pt) \
+    do { \
+	/* latest breed of gcc is clever enough to use move */ \
+	memcpy(&(v), (pt), 8); \
+	(v) = be64_to_cpu(v); \
+    } while(0)
+
+/* rotation right shift 1byte */
+#define ROR8(x) (((x) >> 8) + ((x) << (sizeof(x)*8 - 8)))
+/* rotation left shift 1bit */
+#define ROL1(x) (((x) << 1) + ((x) >> (sizeof(x)*8 - 1)))
+/* rotation left shift 1byte */
+#define ROL8(x) (((x) << 8) + ((x) >> (sizeof(x)*8 - 8)))
+
+#define ROLDQ(l, r, w, bits)				\
+    do {						\
+	w = l;						\
+	l = (l << bits) + (r >> (64 - bits));		\
+	r = (r << bits) + (w >> (64 - bits));		\
+    } while(0)
+
+#define CAMELLIA_F(x, k, y, i)					\
+    do {							\
+	u32 yl, yr;						\
+	i = x ^ k;						\
+	yl = camellia_sp1110[(u8)i]				\
+	   ^ camellia_sp0222[(u8)(i >> 24)]			\
+	   ^ camellia_sp3033[(u8)(i >> 16)]			\
+	   ^ camellia_sp4404[(u8)(i >> 8)];			\
+	yr = camellia_sp1110[    (i >> 56)]			\
+	   ^ camellia_sp0222[(u8)(i >> 48)]			\
+	   ^ camellia_sp3033[(u8)(i >> 40)]			\
+	   ^ camellia_sp4404[(u8)(i >> 32)];			\
+	yl ^= yr;						\
+	yr = ROR8(yr);						\
+	yr ^= yl;						\
+	y = ((u64)yl << 32) + yr;				\
+    } while(0)
+
+#define SUBKEY(INDEX) (subkey[(INDEX)])
+
+#ifdef __BIG_ENDIAN
+#define SUBKEY_L(INDEX) (((u32*)subkey)[(INDEX)*2])
+#define SUBKEY_R(INDEX) (((u32*)subkey)[(INDEX)*2 + 1])
+#else
+#define SUBKEY_L(INDEX) (((u32*)subkey)[(INDEX)*2 + 1])
+#define SUBKEY_R(INDEX) (((u32*)subkey)[(INDEX)*2])
+#endif
+
+static void camellia_setup_tail(u64 *subkey, int max)
+{
+	u32 dw;
+	int i = 2;
+	do {
+		dw = SUBKEY_L(i + 0) ^ SUBKEY_R(i + 0); dw = ROL8(dw);/* round 1 */
+		SUBKEY_R(i + 0) = SUBKEY_L(i + 0) ^ dw; SUBKEY_L(i + 0) = dw;
+		dw = SUBKEY_L(i + 1) ^ SUBKEY_R(i + 1); dw = ROL8(dw);/* round 2 */
+		SUBKEY_R(i + 1) = SUBKEY_L(i + 1) ^ dw; SUBKEY_L(i + 1) = dw;
+		dw = SUBKEY_L(i + 2) ^ SUBKEY_R(i + 2); dw = ROL8(dw);/* round 3 */
+		SUBKEY_R(i + 2) = SUBKEY_L(i + 2) ^ dw; SUBKEY_L(i + 2) = dw;
+		dw = SUBKEY_L(i + 3) ^ SUBKEY_R(i + 3); dw = ROL8(dw);/* round 4 */
+		SUBKEY_R(i + 3) = SUBKEY_L(i + 3) ^ dw; SUBKEY_L(i + 3) = dw;
+		dw = SUBKEY_L(i + 4) ^ SUBKEY_R(i + 4); dw = ROL8(dw);/* round 5 */
+		SUBKEY_R(i + 4) = SUBKEY_L(i + 4) ^ dw; SUBKEY_L(i + 4) = dw;
+		dw = SUBKEY_L(i + 5) ^ SUBKEY_R(i + 5); dw = ROL8(dw);/* round 6 */
+		SUBKEY_R(i + 5) = SUBKEY_L(i + 5) ^ dw; SUBKEY_L(i + 5) = dw;
+		i += 8;
+	} while (i < max);
+}
+
+#ifdef __BIG_ENDIAN
+#define subL(INDEX) (((u32*)sub)[(INDEX)*2])
+#define subR(INDEX) (((u32*)sub)[(INDEX)*2 + 1])
+#else
+#define subL(INDEX) (((u32*)sub)[(INDEX)*2 + 1])
+#define subR(INDEX) (((u32*)sub)[(INDEX)*2])
+#endif
+
+static void camellia_setup128(const unsigned char *key, u64 *subkey)
+{
+	u64 kl, kr;
+	u64 i, t, w;
+	u64 kw4;
+	u32 dw;
+	u64 sub[26];
+
+	/**
+	 *  k == kl || kr (|| is concatination)
+	 */
+	GETU64(kl, key     );
+	GETU64(kr, key +  8);
+
+	/**
+	 * generate KL dependent subkeys
+	 */
+	/* kw1 */
+	sub[0] = kl;
+	/* kw2 */
+	sub[1] = kr;
+	/* rotation left shift 15bit */
+	ROLDQ(kl, kr, w, 15);
+	/* k3 */
+	sub[4] = kl;
+	/* k4 */
+	sub[5] = kr;
+	/* rotation left shift 15+30bit */
+	ROLDQ(kl, kr, w, 30);
+	/* k7 */
+	sub[10] = kl;
+	/* k8 */
+	sub[11] = kr;
+	/* rotation left shift 15+30+15bit */
+	ROLDQ(kl, kr, w, 15);
+	/* k10 */
+	sub[13] = kr;
+	/* rotation left shift 15+30+15+17 bit */
+	ROLDQ(kl, kr, w, 17);
+	/* kl3 */
+	sub[16] = kl;
+	/* kl4 */
+	sub[17] = kr;
+	/* rotation left shift 15+30+15+17+17 bit */
+	ROLDQ(kl, kr, w, 17);
+	/* k13 */
+	sub[18] = kl;
+	/* k14 */
+	sub[19] = kr;
+	/* rotation left shift 15+30+15+17+17+17 bit */
+	ROLDQ(kl, kr, w, 17);
+	/* k17 */
+	sub[22] = kl;
+	/* k18 */
+	sub[23] = kr;
+
+	/* generate KA */
+	kl = sub[0];
+	kr = sub[1];
+	CAMELLIA_F(kl, CAMELLIA_SIGMA1, w, i);
+	kr ^= w;
+	CAMELLIA_F(kr, CAMELLIA_SIGMA2, kl, i);
+	/* current status == (kl, w) */
+	CAMELLIA_F(kl, CAMELLIA_SIGMA3, kr, i);
+	kr ^= w;
+	CAMELLIA_F(kr, CAMELLIA_SIGMA4, w, i);
+	kl ^= w;
+
+	/* generate KA dependent subkeys */
+	/* k1, k2 */
+	sub[2] = kl;
+	sub[3] = kr;
+	ROLDQ(kl, kr, w, 15);
+	/* k5,k6 */
+	sub[6] = kl;
+	sub[7] = kr;
+	ROLDQ(kl, kr, w, 15);
+	/* kl1, kl2 */
+	sub[8] = kl;
+	sub[9] = kr;
+	ROLDQ(kl, kr, w, 15);
+	/* k9 */
+	sub[12] = kl;
+	ROLDQ(kl, kr, w, 15);
+	/* k11, k12 */
+	sub[14] = kl;
+	sub[15] = kr;
+	ROLDQ(kl, kr, w, 34);
+	/* k15, k16 */
+	sub[20] = kl;
+	sub[21] = kr;
+	ROLDQ(kl, kr, w, 17);
+	/* kw3, kw4 */
+	sub[24] = kl;
+	sub[25] = kr;
+
+	/* absorb kw2 to other subkeys */
+	/* round 2 */
+	sub[3] ^= sub[1];
+	/* round 4 */
+	sub[5] ^= sub[1];
+	/* round 6 */
+	sub[7] ^= sub[1];
+	subL(1) ^= subR(1) & ~subR(9);
+	dw = subL(1) & subL(9),
+		subR(1) ^= ROL1(dw); /* modified for FLinv(kl2) */
+	/* round 8 */
+	sub[11] ^= sub[1];
+	/* round 10 */
+	sub[13] ^= sub[1];
+	/* round 12 */
+	sub[15] ^= sub[1];
+	subL(1) ^= subR(1) & ~subR(17);
+	dw = subL(1) & subL(17),
+		subR(1) ^= ROL1(dw); /* modified for FLinv(kl4) */
+	/* round 14 */
+	sub[19] ^= sub[1];
+	/* round 16 */
+	sub[21] ^= sub[1];
+	/* round 18 */
+	sub[23] ^= sub[1];
+	/* kw3 */
+	sub[24] ^= sub[1];
+
+	/* absorb kw4 to other subkeys */
+	kw4 = sub[25];
+	/* round 17 */
+	sub[22] ^= kw4;
+	/* round 15 */
+	sub[20] ^= kw4;
+	/* round 13 */
+	sub[18] ^= kw4;
+	kw4 ^= (u64)((u32)kw4 & ~subR(16)) << 32; //kw4l ^= kw4r & ~subR(16);
+	dw = (u32)(kw4 >> 32) & subL(16); // kw4l & subL[16],
+	kw4 ^= ROL1(dw); /* modified for FL(kl3) */
+	/* round 11 */
+	sub[14] ^= kw4;
+	/* round 9 */
+	sub[12] ^= kw4;
+	/* round 7 */
+	sub[10] ^= kw4;
+	kw4 ^= (u64)((u32)kw4 & ~subR(8)) << 32; //kw4l ^= kw4r & ~subR[8];
+	dw = (u32)(kw4 >> 32) & subL(8);
+	kw4 ^= ROL1(dw); /* modified for FL(kl1) */
+	/* round 5 */
+	sub[6] ^= kw4;
+	/* round 3 */
+	sub[4] ^= kw4;
+	/* round 1 */
+	sub[2] ^= kw4;
+	/* kw1 */
+	sub[0] ^= kw4;
+
+	/* key XOR is end of F-function */
+	SUBKEY(0) = sub[0] ^ sub[2];/* kw1 */
+	SUBKEY(2) = sub[3];       /* round 1 */
+	SUBKEY(3) = sub[2] ^ sub[4]; /* round 2 */
+	SUBKEY(4) = sub[3] ^ sub[5]; /* round 3 */
+	SUBKEY(5) = sub[4] ^ sub[6]; /* round 4 */
+	SUBKEY(6) = sub[5] ^ sub[7]; /* round 5 */
+	t = subL(10) ^ (subR(10) & ~subR(8)); // tl = subL[10] ^ (subR[10] & ~subR[8]);
+	dw = (u32)t & subL(8);  /* FL(kl1) */
+	t = (t << 32) | (subR(10) ^ ROL1(dw)); // tr = subR[10] ^ ROL1(dw);
+	SUBKEY(7) = sub[6] ^ t; /* round 6 */
+	SUBKEY(8) = sub[8];       /* FL(kl1) */
+	SUBKEY(9) = sub[9];       /* FLinv(kl2) */
+	t = subL(7) ^ (subR(7) & ~subR(9));
+	dw = (u32)t & subL(9);  /* FLinv(kl2) */
+	t = (t << 32) | (subR(7) ^ ROL1(dw));
+	SUBKEY(10) = t ^ sub[11]; /* round 7 */
+	SUBKEY(11) = sub[10] ^ sub[12]; /* round 8 */
+	SUBKEY(12) = sub[11] ^ sub[13]; /* round 9 */
+	SUBKEY(13) = sub[12] ^ sub[14]; /* round 10 */
+	SUBKEY(14) = sub[13] ^ sub[15]; /* round 11 */
+	t = subL(18) ^ (subR(18) & ~subR(16));
+	dw = (u32)t & subL(16); /* FL(kl3) */
+	t = (t << 32) | (subR(18) ^ ROL1(dw));
+	SUBKEY(15) = sub[14] ^ t; /* round 12 */
+	SUBKEY(16) = sub[16];     /* FL(kl3) */
+	SUBKEY(17) = sub[17];     /* FLinv(kl4) */
+	t = subL(15) ^ (subR(15) & ~subR(17));
+	dw = (u32)t & subL(17); /* FLinv(kl4) */
+	t = (t << 32) | (subR(15) ^ ROL1(dw));
+	SUBKEY(18) = t ^ sub[19]; /* round 13 */
+	SUBKEY(19) = sub[18] ^ sub[20]; /* round 14 */
+	SUBKEY(20) = sub[19] ^ sub[21]; /* round 15 */
+	SUBKEY(21) = sub[20] ^ sub[22]; /* round 16 */
+	SUBKEY(22) = sub[21] ^ sub[23]; /* round 17 */
+	SUBKEY(23) = sub[22];     /* round 18 */
+	SUBKEY(24) = sub[24] ^ sub[23]; /* kw3 */
+
+	/* apply the inverse of the last half of P-function */
+	camellia_setup_tail(subkey, 24);
+}
+
+static void camellia_setup256(const unsigned char *key, u64 *subkey)
+{
+	u64 kl, kr;        /* left half of key */
+	u64 krl, krr;      /* right half of key */
+	u64 i, t, w;       /* temporary variables */
+	u64 kw4;
+	u32 dw;
+	u64 sub[34];
+
+	/**
+	 *  key = (kl || kr || krl || krr)
+	 *  (|| is concatination)
+	 */
+	GETU64(kl,  key     );
+	GETU64(kr,  key +  8);
+	GETU64(krl, key + 16);
+	GETU64(krr, key + 24);
+
+	/* generate KL dependent subkeys */
+	/* kw1 */
+	sub[0] = kl;
+	/* kw2 */
+	sub[1] = kr;
+	ROLDQ(kl, kr, w, 45);
+	/* k9 */
+	sub[12] = kl;
+	/* k10 */
+	sub[13] = kr;
+	ROLDQ(kl, kr, w, 15);
+	/* kl3 */
+	sub[16] = kl;
+	/* kl4 */
+	sub[17] = kr;
+	ROLDQ(kl, kr, w, 17);
+	/* k17 */
+	sub[22] = kl;
+	/* k18 */
+	sub[23] = kr;
+	ROLDQ(kl, kr, w, 34);
+	/* k23 */
+	sub[30] = kl;
+	/* k24 */
+	sub[31] = kr;
+
+	/* generate KR dependent subkeys */
+	ROLDQ(krl, krr, w, 15);
+	/* k3 */
+	sub[4] = krl;
+	/* k4 */
+	sub[5] = krr;
+	ROLDQ(krl, krr, w, 15);
+	/* kl1 */
+	sub[8] = krl;
+	/* kl2 */
+	sub[9] = krr;
+	ROLDQ(krl, krr, w, 30);
+	/* k13 */
+	sub[18] = krl;
+	/* k14 */
+	sub[19] = krr;
+	ROLDQ(krl, krr, w, 34);
+	/* k19 */
+	sub[26] = krl;
+	/* k20 */
+	sub[27] = krr;
+	ROLDQ(krl, krr, w, 34);
+
+	/* generate KA */
+	kl = sub[0] ^ krl;
+	kr = sub[1] ^ krr;
+	CAMELLIA_F(kl, CAMELLIA_SIGMA1, w, i);
+	kr ^= w;
+	CAMELLIA_F(kr, CAMELLIA_SIGMA2, kl, i);
+	kl ^= krl;
+	CAMELLIA_F(kl, CAMELLIA_SIGMA3, kr, i);
+	kr ^= w ^ krr;
+	CAMELLIA_F(kr, CAMELLIA_SIGMA4, w, i);
+	kl ^= w;
+
+	/* generate KB */
+	krl ^= kl;
+	krr ^= kr;
+	CAMELLIA_F(krl, CAMELLIA_SIGMA5, w, i);
+	krr ^= w;
+	CAMELLIA_F(krr, CAMELLIA_SIGMA6, w, i);
+	krl ^= w;
+
+	/* generate KA dependent subkeys */
+	ROLDQ(kl, kr, w, 15);
+	/* k5 */
+	sub[6] = kl;
+	/* k6 */
+	sub[7] = kr;
+	ROLDQ(kl, kr, w, 30);
+	/* k11 */
+	sub[14] = kl;
+	/* k12 */
+	sub[15] = kr;
+	/* kl5 */
+	ROLDQ(kl, kr, w, 32);
+	sub[24] = kl;
+	/* kl6 */
+	sub[25] = kr;
+	/* rotation left shift 49 from k11,k12 -> k21,k22 */
+	ROLDQ(kl, kr, w, (49 - 32));
+	/* k21 */
+	sub[28] = kl;
+	/* k22 */
+	sub[29] = kr;
+
+	/* generate KB dependent subkeys */
+	/* k1 */
+	sub[2] = krl;
+	/* k2 */
+	sub[3] = krr;
+	ROLDQ(krl, krr, w, 30);
+	/* k7 */
+	sub[10] = krl;
+	/* k8 */
+	sub[11] = krr;
+	ROLDQ(krl, krr, w, 30);
+	/* k15 */
+	sub[20] = krl;
+	/* k16 */
+	sub[21] = krr;
+	ROLDQ(krl, krr, w, 51);
+	/* kw3 */
+	sub[32] = krl;
+	/* kw4 */
+	sub[33] = krr;
+
+	/* absorb kw2 to other subkeys */
+	/* round 2 */
+	sub[3] ^= sub[1];
+	/* round 4 */
+	sub[5] ^= sub[1];
+	/* round 6 */
+	sub[7] ^= sub[1];
+	subL(1) ^= subR(1) & ~subR(9);
+	dw = subL(1) & subL(9),
+		subR(1) ^= ROL1(dw); /* modified for FLinv(kl2) */
+	/* round 8 */
+	sub[11] ^= sub[1];
+	/* round 10 */
+	sub[13] ^= sub[1];
+	/* round 12 */
+	sub[15] ^= sub[1];
+	subL(1) ^= subR(1) & ~subR(17);
+	dw = subL(1) & subL(17),
+		subR(1) ^= ROL1(dw); /* modified for FLinv(kl4) */
+	/* round 14 */
+	sub[19] ^= sub[1];
+	/* round 16 */
+	sub[21] ^= sub[1];
+	/* round 18 */
+	sub[23] ^= sub[1];
+	subL(1) ^= subR(1) & ~subR(25);
+	dw = subL(1) & subL(25),
+		subR(1) ^= ROL1(dw); /* modified for FLinv(kl6) */
+	/* round 20 */
+	sub[27] ^= sub[1];
+	/* round 22 */
+	sub[29] ^= sub[1];
+	/* round 24 */
+	sub[31] ^= sub[1];
+	/* kw3 */
+	sub[32] ^= sub[1];
+
+	/* absorb kw4 to other subkeys */
+	kw4 = sub[33];
+	/* round 23 */
+	sub[30] ^= kw4;
+	/* round 21 */
+	sub[28] ^= kw4;
+	/* round 19 */
+	sub[26] ^= kw4;
+	kw4 ^= (u64)((u32)kw4 & ~subR(24)) << 32; //kw4l ^= kw4r & ~subR[24];
+	dw = (u32)(kw4 >> 32) & subL(24),
+		kw4 ^= ROL1(dw); /* modified for FL(kl5) */
+	/* round 17 */
+	sub[22] ^= kw4;
+	/* round 15 */
+	sub[20] ^= kw4;
+	/* round 13 */
+	sub[18] ^= kw4;
+	kw4 ^= (u64)((u32)kw4 & ~subR(16)) << 32;
+	dw = (u32)(kw4 >> 32) & subL(16),
+		kw4 ^= ROL1(dw); /* modified for FL(kl3) */
+	/* round 11 */
+	sub[14] ^= kw4;
+	/* round 9 */
+	sub[12] ^= kw4;
+	/* round 7 */
+	sub[10] ^= kw4;
+	kw4 ^= (u64)((u32)kw4 & ~subR(8)) << 32;
+	dw = (u32)(kw4 >> 32) & subL(8),
+		kw4 ^= ROL1(dw); /* modified for FL(kl1) */
+	/* round 5 */
+	sub[6] ^= kw4;
+	/* round 3 */
+	sub[4] ^= kw4;
+	/* round 1 */
+	sub[2] ^= kw4;
+	/* kw1 */
+	sub[0] ^= kw4;
+
+	/* key XOR is end of F-function */
+	SUBKEY(0) = sub[0] ^ sub[2];/* kw1 */
+	SUBKEY(2) = sub[3];       /* round 1 */
+	SUBKEY(3) = sub[2] ^ sub[4]; /* round 2 */
+	SUBKEY(4) = sub[3] ^ sub[5]; /* round 3 */
+	SUBKEY(5) = sub[4] ^ sub[6]; /* round 4 */
+	SUBKEY(6) = sub[5] ^ sub[7]; /* round 5 */
+	t = subL(10) ^ (subR(10) & ~subR(8)); // tl = subL[10] ^ (subR[10] & ~subR[8]);
+	dw = (u32)t & subL(8);  /* FL(kl1) */
+	t = (t << 32) | (subR(10) ^ ROL1(dw)); //tr = subR[10] ^ ROL1(dw);
+	SUBKEY(7) = sub[6] ^ t;   /* round 6 */
+	SUBKEY(8) = sub[8];       /* FL(kl1) */
+	SUBKEY(9) = sub[9];       /* FLinv(kl2) */
+	t = subL(7) ^ (subR(7) & ~subR(9));
+	dw = (u32)t & subL(9);  /* FLinv(kl2) */
+	t = (t << 32) | (subR(7) ^ ROL1(dw));
+	SUBKEY(10) = t ^ sub[11]; /* round 7 */
+	SUBKEY(11) = sub[10] ^ sub[12]; /* round 8 */
+	SUBKEY(12) = sub[11] ^ sub[13]; /* round 9 */
+	SUBKEY(13) = sub[12] ^ sub[14]; /* round 10 */
+	SUBKEY(14) = sub[13] ^ sub[15]; /* round 11 */
+	t = subL(18) ^ (subR(18) & ~subR(16));
+	dw = (u32)t & subL(16); /* FL(kl3) */
+	t = (t << 32) | (subR(18) ^ ROL1(dw));
+	SUBKEY(15) = sub[14] ^ t; /* round 12 */
+	SUBKEY(16) = sub[16];     /* FL(kl3) */
+	SUBKEY(17) = sub[17];     /* FLinv(kl4) */
+	t = subL(15) ^ (subR(15) & ~subR(17));
+	dw = (u32)t & subL(17); /* FLinv(kl4) */
+	t = (t << 32) | (subR(15) ^ ROL1(dw));
+	SUBKEY(18) = t ^ sub[19]; /* round 13 */
+	SUBKEY(19) = sub[18] ^ sub[20]; /* round 14 */
+	SUBKEY(20) = sub[19] ^ sub[21]; /* round 15 */
+	SUBKEY(21) = sub[20] ^ sub[22]; /* round 16 */
+	SUBKEY(22) = sub[21] ^ sub[23]; /* round 17 */
+	t = subL(26) ^ (subR(26) & ~subR(24));
+	dw = (u32)t & subL(24); /* FL(kl5) */
+	t = (t << 32) | (subR(26) ^ ROL1(dw));
+	SUBKEY(23) = sub[22] ^ t; /* round 18 */
+	SUBKEY(24) = sub[24];     /* FL(kl5) */
+	SUBKEY(25) = sub[25];     /* FLinv(kl6) */
+	t = subL(23) ^ (subR(23) & ~subR(25));
+	dw = (u32)t & subL(25); /* FLinv(kl6) */
+	t = (t << 32) | (subR(23) ^ ROL1(dw));
+	SUBKEY(26) = t ^ sub[27]; /* round 19 */
+	SUBKEY(27) = sub[26] ^ sub[28]; /* round 20 */
+	SUBKEY(28) = sub[27] ^ sub[29]; /* round 21 */
+	SUBKEY(29) = sub[28] ^ sub[30]; /* round 22 */
+	SUBKEY(30) = sub[29] ^ sub[31]; /* round 23 */
+	SUBKEY(31) = sub[30];     /* round 24 */
+	SUBKEY(32) = sub[32] ^ sub[31]; /* kw3 */
+
+	/* apply the inverse of the last half of P-function */
+	camellia_setup_tail(subkey, 32);
+}
+
+static void camellia_setup192(const unsigned char *key, u64 *subkey)
+{
+	unsigned char kk[32];
+	u64 krl, krr;
+
+	memcpy(kk, key, 24);
+	memcpy((unsigned char *)&krl, key+16, 8);
+	krr = ~krl;
+	memcpy(kk+24, (unsigned char *)&krr, 8);
+	camellia_setup256(kk, subkey);
+}
+
+typedef u64 key_element;
+typedef const u64 const_key_element;
+
+
+
+#else /* BITS_PER_LONG < 64 */
+
+/*
+ * Key setup implementation with 32-bit ops
+ */
 
 /* key constants */
 

@@ -329,8 +912,7 @@ static const u32 camellia_sp4404[256] = 
 /*
  *  macros
  */
-
-# define GETU32(v, pt) \
+#define GETU32(v, pt) \
     do { \
 	/* latest breed of gcc is clever enough to use move */ \
 	memcpy(&(v), (pt), 4); \

@@ -363,64 +945,25 @@ static const u32 camellia_sp4404[256] = 
 	rr = (w0 << (bits - 32)) + (w1 >> (64 - bits));	\
     } while(0)
 
-
 #define CAMELLIA_F(xl, xr, kl, kr, yl, yr, il, ir, t0, t1)	\
     do {							\
 	il = xl ^ kl;						\
 	ir = xr ^ kr;						\
 	t0 = il >> 16;						\
 	t1 = ir >> 16;						\
-	yl = camellia_sp1110[ir & 0xff]				\
-	   ^ camellia_sp0222[(t1 >> 8) & 0xff]			\
-	   ^ camellia_sp3033[t1 & 0xff]				\
-	   ^ camellia_sp4404[(ir >> 8) & 0xff];			\
-	yr = camellia_sp1110[(t0 >> 8) & 0xff]			\
-	   ^ camellia_sp0222[t0 & 0xff]				\
-	   ^ camellia_sp3033[(il >> 8) & 0xff]			\
-	   ^ camellia_sp4404[il & 0xff];			\
+	yl = camellia_sp1110[(u8)(ir     )]			\
+	   ^ camellia_sp0222[    (t1 >> 8)]			\
+	   ^ camellia_sp3033[(u8)(t1     )]			\
+	   ^ camellia_sp4404[(u8)(ir >> 8)];			\
+	yr = camellia_sp1110[    (t0 >> 8)]			\
+	   ^ camellia_sp0222[(u8)(t0     )]			\
+	   ^ camellia_sp3033[(u8)(il >> 8)]			\
+	   ^ camellia_sp4404[(u8)(il     )];			\
 	yl ^= yr;						\
 	yr = ROR8(yr);						\
 	yr ^= yl;						\
     } while(0)
 
-
-/*
- * for speed up
- *
- */
-#define CAMELLIA_FLS(ll, lr, rl, rr, kll, klr, krl, krr, t0, t1, t2, t3) \
-    do {								\
-	t0 = kll;							\
-	t2 = krr;							\
-	t0 &= ll;							\
-	t2 |= rr;							\
-	rl ^= t2;							\
-	lr ^= ROL1(t0);							\
-	t3 = krl;							\
-	t1 = klr;							\
-	t3 &= rl;							\
-	t1 |= lr;							\
-	ll ^= t1;							\
-	rr ^= ROL1(t3);							\
-    } while(0)
-
-#define CAMELLIA_ROUNDSM(xl, xr, kl, kr, yl, yr, il, ir, t0, t1)	\
-    do {								\
-	ir =  camellia_sp1110[xr & 0xff];				\
-	il =  camellia_sp1110[(xl>>24) & 0xff];				\
-	ir ^= camellia_sp0222[(xr>>24) & 0xff];				\
-	il ^= camellia_sp0222[(xl>>16) & 0xff];				\
-	ir ^= camellia_sp3033[(xr>>16) & 0xff];				\
-	il ^= camellia_sp3033[(xl>>8) & 0xff];				\
-	ir ^= camellia_sp4404[(xr>>8) & 0xff];				\
-	il ^= camellia_sp4404[xl & 0xff];				\
-	il ^= kl;							\
-	ir ^= il ^ kr;							\
-	yl ^= ir;							\
-	yr ^= ROR8(il) ^ ir;						\
-    } while(0)
-
-
 #define SUBKEY_L(INDEX) (subkey[(INDEX)*2])
 #define SUBKEY_R(INDEX) (subkey[(INDEX)*2 + 1])
 

@@ -999,8 +1542,49 @@ static void camellia_setup192(const unsi
 	camellia_setup256(kk, subkey);
 }
 
+typedef u32 key_element;
+typedef const u32 const_key_element;
+
+#endif /* 32/64-bit key setup versions */
+
+
+
+/*
+ * Encrypt/decrypt
+ */
+#define CAMELLIA_FLS(ll, lr, rl, rr, kll, klr, krl, krr, t0, t1, t2, t3) \
+    do {								\
+	t0 = kll;							\
+	t2 = krr;							\
+	t0 &= ll;							\
+	t2 |= rr;							\
+	rl ^= t2;							\
+	lr ^= ROL1(t0);							\
+	t3 = krl;							\
+	t1 = klr;							\
+	t3 &= rl;							\
+	t1 |= lr;							\
+	ll ^= t1;							\
+	rr ^= ROL1(t3);							\
+    } while(0)
+
+#define CAMELLIA_ROUNDSM(xl, xr, kl, kr, yl, yr, il, ir)		\
+    do {								\
+	ir =  camellia_sp1110[(u8)xr];					\
+	il =  camellia_sp1110[    (xl >> 24)];				\
+	ir ^= camellia_sp0222[    (xr >> 24)];				\
+	il ^= camellia_sp0222[(u8)(xl >> 16)];				\
+	ir ^= camellia_sp3033[(u8)(xr >> 16)];				\
+	il ^= camellia_sp3033[(u8)(xl >> 8)];				\
+	ir ^= camellia_sp4404[(u8)(xr >> 8)];				\
+	il ^= camellia_sp4404[(u8)xl];					\
+	il ^= kl;							\
+	ir ^= il ^ kr;							\
+	yl ^= ir;							\
+	yr ^= ROR8(il) ^ ir;						\
+    } while(0)
 
-static void camellia_encrypt128(const u32 *subkey, u32 *io_text)
+static void camellia_encrypt128(const_key_element *subkey, u32 *io_text)
 {
 	u32 il,ir,t0,t1;               /* temporary variables */
 

@@ -1015,22 +1599,22 @@ static void camellia_encrypt128(const u3
 	/* main iteration */
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(2),SUBKEY_R(2),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(3),SUBKEY_R(3),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(4),SUBKEY_R(4),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(5),SUBKEY_R(5),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(6),SUBKEY_R(6),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(7),SUBKEY_R(7),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
 		     SUBKEY_L(8),SUBKEY_R(8),

@@ -1039,22 +1623,22 @@ static void camellia_encrypt128(const u3
 
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(10),SUBKEY_R(10),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(11),SUBKEY_R(11),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(12),SUBKEY_R(12),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(13),SUBKEY_R(13),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(14),SUBKEY_R(14),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(15),SUBKEY_R(15),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
 		     SUBKEY_L(16),SUBKEY_R(16),

@@ -1063,22 +1647,22 @@ static void camellia_encrypt128(const u3
 
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(18),SUBKEY_R(18),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(19),SUBKEY_R(19),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(20),SUBKEY_R(20),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(21),SUBKEY_R(21),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(22),SUBKEY_R(22),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(23),SUBKEY_R(23),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	/* post whitening but kw4 */
 	io_text[0] = io[2] ^ SUBKEY_L(24);

@@ -1087,7 +1671,7 @@ static void camellia_encrypt128(const u3
 	io_text[3] = io[1];
 }
 
-static void camellia_decrypt128(const u32 *subkey, u32 *io_text)
+static void camellia_decrypt128(const_key_element *subkey, u32 *io_text)
 {
 	u32 il,ir,t0,t1;               /* temporary variables */
 

@@ -1102,22 +1686,22 @@ static void camellia_decrypt128(const u3
 	/* main iteration */
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(23),SUBKEY_R(23),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(22),SUBKEY_R(22),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(21),SUBKEY_R(21),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(20),SUBKEY_R(20),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(19),SUBKEY_R(19),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(18),SUBKEY_R(18),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
 		     SUBKEY_L(17),SUBKEY_R(17),

@@ -1126,22 +1710,22 @@ static void camellia_decrypt128(const u3
 
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(15),SUBKEY_R(15),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(14),SUBKEY_R(14),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(13),SUBKEY_R(13),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(12),SUBKEY_R(12),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(11),SUBKEY_R(11),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(10),SUBKEY_R(10),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
 		     SUBKEY_L(9),SUBKEY_R(9),

@@ -1150,22 +1734,22 @@ static void camellia_decrypt128(const u3
 
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(7),SUBKEY_R(7),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(6),SUBKEY_R(6),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(5),SUBKEY_R(5),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(4),SUBKEY_R(4),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(3),SUBKEY_R(3),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(2),SUBKEY_R(2),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	/* post whitening but kw4 */
 	io_text[0] = io[2] ^ SUBKEY_L(0);

@@ -1174,7 +1758,7 @@ static void camellia_decrypt128(const u3
 	io_text[3] = io[1];
 }
 
-static void camellia_encrypt256(const u32 *subkey, u32 *io_text)
+static void camellia_encrypt256(const_key_element *subkey, u32 *io_text)
 {
 	u32 il,ir,t0,t1;           /* temporary variables */
 

@@ -1189,22 +1773,22 @@ static void camellia_encrypt256(const u3
 	/* main iteration */
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(2),SUBKEY_R(2),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(3),SUBKEY_R(3),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(4),SUBKEY_R(4),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(5),SUBKEY_R(5),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(6),SUBKEY_R(6),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(7),SUBKEY_R(7),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
 		     SUBKEY_L(8),SUBKEY_R(8),

@@ -1213,22 +1797,22 @@ static void camellia_encrypt256(const u3
 
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(10),SUBKEY_R(10),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(11),SUBKEY_R(11),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(12),SUBKEY_R(12),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(13),SUBKEY_R(13),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(14),SUBKEY_R(14),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(15),SUBKEY_R(15),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
 		     SUBKEY_L(16),SUBKEY_R(16),

@@ -1237,22 +1821,22 @@ static void camellia_encrypt256(const u3
 
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(18),SUBKEY_R(18),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(19),SUBKEY_R(19),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(20),SUBKEY_R(20),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(21),SUBKEY_R(21),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(22),SUBKEY_R(22),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(23),SUBKEY_R(23),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
 		     SUBKEY_L(24),SUBKEY_R(24),

@@ -1261,22 +1845,22 @@ static void camellia_encrypt256(const u3
 
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(26),SUBKEY_R(26),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(27),SUBKEY_R(27),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(28),SUBKEY_R(28),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(29),SUBKEY_R(29),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(30),SUBKEY_R(30),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(31),SUBKEY_R(31),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	/* post whitening but kw4 */
 	io_text[0] = io[2] ^ SUBKEY_L(32);

@@ -1285,7 +1869,7 @@ static void camellia_encrypt256(const u3
 	io_text[3] = io[1];
 }
 
-static void camellia_decrypt256(const u32 *subkey, u32 *io_text)
+static void camellia_decrypt256(const_key_element *subkey, u32 *io_text)
 {
 	u32 il,ir,t0,t1;           /* temporary variables */
 

@@ -1300,22 +1884,22 @@ static void camellia_decrypt256(const u3
 	/* main iteration */
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(31),SUBKEY_R(31),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(30),SUBKEY_R(30),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(29),SUBKEY_R(29),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(28),SUBKEY_R(28),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(27),SUBKEY_R(27),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(26),SUBKEY_R(26),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
 		     SUBKEY_L(25),SUBKEY_R(25),

@@ -1324,22 +1908,22 @@ static void camellia_decrypt256(const u3
 
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(23),SUBKEY_R(23),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(22),SUBKEY_R(22),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(21),SUBKEY_R(21),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(20),SUBKEY_R(20),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(19),SUBKEY_R(19),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(18),SUBKEY_R(18),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
 		     SUBKEY_L(17),SUBKEY_R(17),

@@ -1348,22 +1932,22 @@ static void camellia_decrypt256(const u3
 
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(15),SUBKEY_R(15),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(14),SUBKEY_R(14),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(13),SUBKEY_R(13),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(12),SUBKEY_R(12),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(11),SUBKEY_R(11),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(10),SUBKEY_R(10),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
 		     SUBKEY_L(9),SUBKEY_R(9),

@@ -1372,22 +1956,22 @@ static void camellia_decrypt256(const u3
 
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(7),SUBKEY_R(7),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(6),SUBKEY_R(6),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(5),SUBKEY_R(5),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(4),SUBKEY_R(4),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 	CAMELLIA_ROUNDSM(io[0],io[1],
 			 SUBKEY_L(3),SUBKEY_R(3),
-			 io[2],io[3],il,ir,t0,t1);
+			 io[2],io[3],il,ir);
 	CAMELLIA_ROUNDSM(io[2],io[3],
 			 SUBKEY_L(2),SUBKEY_R(2),
-			 io[0],io[1],il,ir,t0,t1);
+			 io[0],io[1],il,ir);
 
 	/* post whitening but kw4 */
 	io_text[0] = io[2] ^ SUBKEY_L(0);

@@ -1399,7 +1983,7 @@ static void camellia_decrypt256(const u3
 
 struct camellia_ctx {
 	int key_length;
-	u32 key_table[CAMELLIA_TABLE_BYTE_LEN / 4];
+	key_element key_table[CAMELLIA_TABLE_BYTE_LEN / sizeof(key_element)];
 };
 
 static int
maintainer
#2 lore raw

Re: [PATCH 1/4] camellia: 64-bit optimization

Quoted 8 lines · Denys Vlasenko
Quoted 6 lines · Denys Vlasenko

Patches start from 5 because pathches 1..4 are already in cryptodev-2.6.

camellia5:
adds 64-bit key setup, it is used if BITS_PER_LONG is 64.

Signed-off-by Denys Vlasenko <[email protected]>

Denys, I really appreciate your work in cleaning this algorithm up.

However, I just can't accept 600 lines of new code just to do setkey faster/smaller on 64-bit platforms. You know, some people have to study these files line-by-line in order have them pass government certifications and such.

OK I went back and looked at the figures, so we're talking about either 1K extra on 64-bit if we use the 32-bit version or 2K extra on 32-bit if we went for the 64-bit version.

Either way it doesn't seem to be a big deal. So how about just picking one and sticking with it?

Thanks,

Signature · 4 lines
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <[email protected]>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
author
#3 lore raw

Re: [PATCH 1/4] camellia: 64-bit optimization

Quoted 8 lines · Herbert Xu
Quoted 7 lines · Denys Vlasenko
Quoted 5 lines · Denys Vlasenko

Patches start from 5 because pathches 1..4 are already in cryptodev-2.6.

camellia5:
adds 64-bit key setup, it is used if BITS_PER_LONG is 64.

Quoted 8 lines
Quoted 1 line
Signed-off-by Denys Vlasenko <[email protected]>

Denys, I really appreciate your work in cleaning this algorithm up.

However, I just can't accept 600 lines of new code just to do setkey faster/smaller on 64-bit platforms. You know, some people have to study these files line-by-line in order have them pass government certifications and such.

480 lines after all patches are applied.

Quoted 6 lines

OK I went back and looked at the figures, so we're talking about either 1K extra on 64-bit if we use the 32-bit version or 2K extra on 32-bit if we went for the 64-bit version.

Either way it doesn't seem to be a big deal. So how about just picking one and sticking with it?

"No big deal" approach resulted in thish module having 30k of code. I cut it down to 20-18k.

In aic7xxx driver, similarly, I rediced size from 300k to 200k. Still not in mainline. This is depressing.

I am an -Os guy and for me 1k out of 20k module doesn't sound too small.

I will rework and repost patches so that 64-bit code is added by last patch, feel free to not apply it.
--
vda

author
#4 lore raw

PATCH 2/4 camellia: code shrink

In reply to Denys Vlasenko
Quoted 6 lines · Denys Vlasenko

camellia6:
unifies encrypt/decrypt routines for different key lengths. This reduces module size by ~25%, with tiny (less than 1%) speed impact.
Also collapses encrypt/decrypt into more readable (visually shorter) form using macros.

Signed-off-by Denys Vlasenko <[email protected]>
Signature · 1 line
vda
crypto/camellia.c +108 -405
diff -urpN linux-2.6.23.1.camellia5/crypto/camellia.c linux-2.6.23.1.camellia6/crypto/camellia.c
@@ -1584,400 +1584,115 @@ typedef const u32 const_key_element;
 	yr ^= ROR8(il) ^ ir;						\
     } while(0)
 
-static void camellia_encrypt128(const_key_element *subkey, u32 *io_text)
+/* max = 24: 128bit encrypt, max = 32: 256bit encrypt */
+static void camellia_do_encrypt(const_key_element *subkey, u32 *io, unsigned max)
 {
 	u32 il,ir,t0,t1;               /* temporary variables */
 
-	u32 io[4];
-
 	/* pre whitening but absorb kw2 */
-	io[0] = io_text[0] ^ SUBKEY_L(0);
-	io[1] = io_text[1] ^ SUBKEY_R(0);
-	io[2] = io_text[2];
-	io[3] = io_text[3];
+	io[0] ^= SUBKEY_L(0);
+	io[1] ^= SUBKEY_R(0);
 
 	/* main iteration */
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(2),SUBKEY_R(2),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(3),SUBKEY_R(3),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(4),SUBKEY_R(4),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(5),SUBKEY_R(5),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(6),SUBKEY_R(6),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(7),SUBKEY_R(7),
-			 io[0],io[1],il,ir);
-
-	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
-		     SUBKEY_L(8),SUBKEY_R(8),
-		     SUBKEY_L(9),SUBKEY_R(9),
-		     t0,t1,il,ir);
-
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(10),SUBKEY_R(10),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(11),SUBKEY_R(11),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(12),SUBKEY_R(12),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(13),SUBKEY_R(13),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(14),SUBKEY_R(14),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(15),SUBKEY_R(15),
-			 io[0],io[1],il,ir);
-
-	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
-		     SUBKEY_L(16),SUBKEY_R(16),
-		     SUBKEY_L(17),SUBKEY_R(17),
-		     t0,t1,il,ir);
-
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(18),SUBKEY_R(18),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(19),SUBKEY_R(19),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(20),SUBKEY_R(20),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(21),SUBKEY_R(21),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(22),SUBKEY_R(22),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(23),SUBKEY_R(23),
-			 io[0],io[1],il,ir);
-
-	/* post whitening but kw4 */
-	io_text[0] = io[2] ^ SUBKEY_L(24);
-	io_text[1] = io[3] ^ SUBKEY_R(24);
-	io_text[2] = io[0];
-	io_text[3] = io[1];
-}
-
-static void camellia_decrypt128(const_key_element *subkey, u32 *io_text)
-{
-	u32 il,ir,t0,t1;               /* temporary variables */
-
-	u32 io[4];
-
-	/* pre whitening but absorb kw2 */
-	io[0] = io_text[0] ^ SUBKEY_L(24);
-	io[1] = io_text[1] ^ SUBKEY_R(24);
-	io[2] = io_text[2];
-	io[3] = io_text[3];
+#define ROUNDS(i) do { \
+	CAMELLIA_ROUNDSM(io[0],io[1], \
+			 SUBKEY_L(i + 2),SUBKEY_R(i + 2), \
+			 io[2],io[3],il,ir); \
+	CAMELLIA_ROUNDSM(io[2],io[3], \
+			 SUBKEY_L(i + 3),SUBKEY_R(i + 3), \
+			 io[0],io[1],il,ir); \
+	CAMELLIA_ROUNDSM(io[0],io[1], \
+			 SUBKEY_L(i + 4),SUBKEY_R(i + 4), \
+			 io[2],io[3],il,ir); \
+	CAMELLIA_ROUNDSM(io[2],io[3], \
+			 SUBKEY_L(i + 5),SUBKEY_R(i + 5), \
+			 io[0],io[1],il,ir); \
+	CAMELLIA_ROUNDSM(io[0],io[1], \
+			 SUBKEY_L(i + 6),SUBKEY_R(i + 6), \
+			 io[2],io[3],il,ir); \
+	CAMELLIA_ROUNDSM(io[2],io[3], \
+			 SUBKEY_L(i + 7),SUBKEY_R(i + 7), \
+			 io[0],io[1],il,ir); \
+} while (0)
+#define FLS(i) do { \
+	CAMELLIA_FLS(io[0],io[1],io[2],io[3], \
+		     SUBKEY_L(i + 0),SUBKEY_R(i + 0), \
+		     SUBKEY_L(i + 1),SUBKEY_R(i + 1), \
+		     t0,t1,il,ir); \
+} while (0)
+
+	ROUNDS(0);
+	FLS(8);
+	ROUNDS(8);
+	FLS(16);
+	ROUNDS(16);
+	if (max == 32) {
+		FLS(24);
+		ROUNDS(24);
+	}
 
-	/* main iteration */
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(23),SUBKEY_R(23),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(22),SUBKEY_R(22),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(21),SUBKEY_R(21),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(20),SUBKEY_R(20),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(19),SUBKEY_R(19),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(18),SUBKEY_R(18),
-			 io[0],io[1],il,ir);
-
-	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
-		     SUBKEY_L(17),SUBKEY_R(17),
-		     SUBKEY_L(16),SUBKEY_R(16),
-		     t0,t1,il,ir);
-
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(15),SUBKEY_R(15),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(14),SUBKEY_R(14),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(13),SUBKEY_R(13),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(12),SUBKEY_R(12),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(11),SUBKEY_R(11),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(10),SUBKEY_R(10),
-			 io[0],io[1],il,ir);
-
-	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
-		     SUBKEY_L(9),SUBKEY_R(9),
-		     SUBKEY_L(8),SUBKEY_R(8),
-		     t0,t1,il,ir);
-
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(7),SUBKEY_R(7),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(6),SUBKEY_R(6),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(5),SUBKEY_R(5),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(4),SUBKEY_R(4),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(3),SUBKEY_R(3),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(2),SUBKEY_R(2),
-			 io[0],io[1],il,ir);
+#undef ROUNDS
+#undef FLS
 
 	/* post whitening but kw4 */
-	io_text[0] = io[2] ^ SUBKEY_L(0);
-	io_text[1] = io[3] ^ SUBKEY_R(0);
-	io_text[2] = io[0];
-	io_text[3] = io[1];
+	io[2] ^= SUBKEY_L(max);
+	io[3] ^= SUBKEY_R(max);
+	/* NB: io[0],[1] should be swapped with [2],[3] by caller! */
 }
 
-static void camellia_encrypt256(const_key_element *subkey, u32 *io_text)
+static void camellia_do_decrypt(const_key_element *subkey, u32 *io, unsigned i)
 {
-	u32 il,ir,t0,t1;           /* temporary variables */
-
-	u32 io[4];
+	u32 il,ir,t0,t1;               /* temporary variables */
 
 	/* pre whitening but absorb kw2 */
-	io[0] = io_text[0] ^ SUBKEY_L(0);
-	io[1] = io_text[1] ^ SUBKEY_R(0);
-	io[2] = io_text[2];
-	io[3] = io_text[3];
+	io[0] ^= SUBKEY_L(i);
+	io[1] ^= SUBKEY_R(i);
 
 	/* main iteration */
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(2),SUBKEY_R(2),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(3),SUBKEY_R(3),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(4),SUBKEY_R(4),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(5),SUBKEY_R(5),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(6),SUBKEY_R(6),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(7),SUBKEY_R(7),
-			 io[0],io[1],il,ir);
-
-	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
-		     SUBKEY_L(8),SUBKEY_R(8),
-		     SUBKEY_L(9),SUBKEY_R(9),
-		     t0,t1,il,ir);
-
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(10),SUBKEY_R(10),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(11),SUBKEY_R(11),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(12),SUBKEY_R(12),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(13),SUBKEY_R(13),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(14),SUBKEY_R(14),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(15),SUBKEY_R(15),
-			 io[0],io[1],il,ir);
-
-	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
-		     SUBKEY_L(16),SUBKEY_R(16),
-		     SUBKEY_L(17),SUBKEY_R(17),
-		     t0,t1,il,ir);
-
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(18),SUBKEY_R(18),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(19),SUBKEY_R(19),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(20),SUBKEY_R(20),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(21),SUBKEY_R(21),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(22),SUBKEY_R(22),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(23),SUBKEY_R(23),
-			 io[0],io[1],il,ir);
-
-	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
-		     SUBKEY_L(24),SUBKEY_R(24),
-		     SUBKEY_L(25),SUBKEY_R(25),
-		     t0,t1,il,ir);
-
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(26),SUBKEY_R(26),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(27),SUBKEY_R(27),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(28),SUBKEY_R(28),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(29),SUBKEY_R(29),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(30),SUBKEY_R(30),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(31),SUBKEY_R(31),
-			 io[0],io[1],il,ir);
-
-	/* post whitening but kw4 */
-	io_text[0] = io[2] ^ SUBKEY_L(32);
-	io_text[1] = io[3] ^ SUBKEY_R(32);
-	io_text[2] = io[0];
-	io_text[3] = io[1];
-}
-
-static void camellia_decrypt256(const_key_element *subkey, u32 *io_text)
-{
-	u32 il,ir,t0,t1;           /* temporary variables */
-
-	u32 io[4];
-
-	/* pre whitening but absorb kw2 */
-	io[0] = io_text[0] ^ SUBKEY_L(32);
-	io[1] = io_text[1] ^ SUBKEY_R(32);
-	io[2] = io_text[2];
-	io[3] = io_text[3];
+#define ROUNDS(i) do { \
+	CAMELLIA_ROUNDSM(io[0],io[1], \
+			 SUBKEY_L(i + 7),SUBKEY_R(i + 7), \
+			 io[2],io[3],il,ir); \
+	CAMELLIA_ROUNDSM(io[2],io[3], \
+			 SUBKEY_L(i + 6),SUBKEY_R(i + 6), \
+			 io[0],io[1],il,ir); \
+	CAMELLIA_ROUNDSM(io[0],io[1], \
+			 SUBKEY_L(i + 5),SUBKEY_R(i + 5), \
+			 io[2],io[3],il,ir); \
+	CAMELLIA_ROUNDSM(io[2],io[3], \
+			 SUBKEY_L(i + 4),SUBKEY_R(i + 4), \
+			 io[0],io[1],il,ir); \
+	CAMELLIA_ROUNDSM(io[0],io[1], \
+			 SUBKEY_L(i + 3),SUBKEY_R(i + 3), \
+			 io[2],io[3],il,ir); \
+	CAMELLIA_ROUNDSM(io[2],io[3], \
+			 SUBKEY_L(i + 2),SUBKEY_R(i + 2), \
+			 io[0],io[1],il,ir); \
+} while (0)
+#define FLS(i) do { \
+	CAMELLIA_FLS(io[0],io[1],io[2],io[3], \
+		     SUBKEY_L(i + 1),SUBKEY_R(i + 1), \
+		     SUBKEY_L(i + 0),SUBKEY_R(i + 0), \
+		     t0,t1,il,ir); \
+} while (0)
+
+	if (i == 32) {
+		ROUNDS(24);
+		FLS(24);
+	}
+	ROUNDS(16);
+	FLS(16);
+	ROUNDS(8);
+	FLS(8);
+	ROUNDS(0);
 
-	/* main iteration */
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(31),SUBKEY_R(31),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(30),SUBKEY_R(30),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(29),SUBKEY_R(29),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(28),SUBKEY_R(28),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(27),SUBKEY_R(27),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(26),SUBKEY_R(26),
-			 io[0],io[1],il,ir);
-
-	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
-		     SUBKEY_L(25),SUBKEY_R(25),
-		     SUBKEY_L(24),SUBKEY_R(24),
-		     t0,t1,il,ir);
-
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(23),SUBKEY_R(23),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(22),SUBKEY_R(22),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(21),SUBKEY_R(21),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(20),SUBKEY_R(20),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(19),SUBKEY_R(19),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(18),SUBKEY_R(18),
-			 io[0],io[1],il,ir);
-
-	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
-		     SUBKEY_L(17),SUBKEY_R(17),
-		     SUBKEY_L(16),SUBKEY_R(16),
-		     t0,t1,il,ir);
-
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(15),SUBKEY_R(15),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(14),SUBKEY_R(14),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(13),SUBKEY_R(13),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(12),SUBKEY_R(12),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(11),SUBKEY_R(11),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(10),SUBKEY_R(10),
-			 io[0],io[1],il,ir);
-
-	CAMELLIA_FLS(io[0],io[1],io[2],io[3],
-		     SUBKEY_L(9),SUBKEY_R(9),
-		     SUBKEY_L(8),SUBKEY_R(8),
-		     t0,t1,il,ir);
-
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(7),SUBKEY_R(7),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(6),SUBKEY_R(6),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(5),SUBKEY_R(5),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(4),SUBKEY_R(4),
-			 io[0],io[1],il,ir);
-	CAMELLIA_ROUNDSM(io[0],io[1],
-			 SUBKEY_L(3),SUBKEY_R(3),
-			 io[2],io[3],il,ir);
-	CAMELLIA_ROUNDSM(io[2],io[3],
-			 SUBKEY_L(2),SUBKEY_R(2),
-			 io[0],io[1],il,ir);
+#undef ROUNDS
+#undef FLS
 
 	/* post whitening but kw4 */
-	io_text[0] = io[2] ^ SUBKEY_L(0);
-	io_text[1] = io[3] ^ SUBKEY_R(0);
-	io_text[2] = io[0];
-	io_text[3] = io[1];
+	io[2] ^= SUBKEY_L(0);
+	io[3] ^= SUBKEY_R(0);
+	/* NB: 0,1 should be swapped with 2,3 by caller! */
 }
 
 

@@ -2029,21 +1744,15 @@ static void camellia_encrypt(struct cryp
 	tmp[2] = be32_to_cpu(src[2]);
 	tmp[3] = be32_to_cpu(src[3]);
 
-	switch (cctx->key_length) {
-	case 16:
-		camellia_encrypt128(cctx->key_table, tmp);
-		break;
-	case 24:
-		/* fall through */
-	case 32:
-		camellia_encrypt256(cctx->key_table, tmp);
-		break;
-	}
-
-	dst[0] = cpu_to_be32(tmp[0]);
-	dst[1] = cpu_to_be32(tmp[1]);
-	dst[2] = cpu_to_be32(tmp[2]);
-	dst[3] = cpu_to_be32(tmp[3]);
+	camellia_do_encrypt(cctx->key_table, tmp,
+		cctx->key_length == 16 ? 24 : 32 /* for key lengths of 24 and 32 */
+	);
+
+	/* do_encrypt returns 0,1 swapped with 2,3 */
+	dst[0] = cpu_to_be32(tmp[2]);
+	dst[1] = cpu_to_be32(tmp[3]);
+	dst[2] = cpu_to_be32(tmp[0]);
+	dst[3] = cpu_to_be32(tmp[1]);
 }
 
 static void camellia_decrypt(struct crypto_tfm *tfm, u8 *out, const u8 *in)

@@ -2059,21 +1768,15 @@ static void camellia_decrypt(struct cryp
 	tmp[2] = be32_to_cpu(src[2]);
 	tmp[3] = be32_to_cpu(src[3]);
 
-	switch (cctx->key_length) {
-	case 16:
-		camellia_decrypt128(cctx->key_table, tmp);
-		break;
-	case 24:
-		/* fall through */
-	case 32:
-		camellia_decrypt256(cctx->key_table, tmp);
-		break;
-	}
-
-	dst[0] = cpu_to_be32(tmp[0]);
-	dst[1] = cpu_to_be32(tmp[1]);
-	dst[2] = cpu_to_be32(tmp[2]);
-	dst[3] = cpu_to_be32(tmp[3]);
+	camellia_do_decrypt(cctx->key_table, tmp,
+		cctx->key_length == 16 ? 24 : 32 /* for key lengths of 24 and 32 */
+	);
+
+	/* do_decrypt returns 0,1 swapped with 2,3 */
+	dst[0] = cpu_to_be32(tmp[2]);
+	dst[1] = cpu_to_be32(tmp[3]);
+	dst[2] = cpu_to_be32(tmp[0]);
+	dst[3] = cpu_to_be32(tmp[1]);
 }
 
 static struct crypto_alg camellia_alg = {
maintainer
#5 lore raw

Re: [PATCH 2/4] camellia: code shrink

Quoted 9 lines · Denys Vlasenko
Quoted 6 lines · Denys Vlasenko

camellia6:
unifies encrypt/decrypt routines for different key lengths. This reduces module size by ~25%, with tiny (less than 1%) speed impact.
Also collapses encrypt/decrypt into more readable (visually shorter) form using macros.

Signed-off-by Denys Vlasenko <[email protected]>

Sorry, this doesn't apply without applying 1/4.

You couldn't have picked a better patch order :)

Seriously, you'll make both of us happier if you put the easy patches first and the contentious stuff at the end.

Thanks,

Signature · 4 lines
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <[email protected]>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
author
#6 lore raw

PATCH 3/4 camellia: code shrink 2

In reply to Denys Vlasenko
Quoted 4 lines · Denys Vlasenko

camellia7:
Move "key XOR is end of F-function" code part into camellia_setup_tail(), it is sufficiently similar between camellia_setup128 and camellia_setup256.

Signed-off-by Denys Vlasenko <[email protected]>
Signature · 1 line
vda
crypto/camellia.c +184 -277
diff -urpN linux-2.6.23.1.camellia6/crypto/camellia.c linux-2.6.23.1.camellia7/crypto/camellia.c
@@ -380,15 +380,80 @@ static const u32 camellia_sp4404[256] = 
 #ifdef __BIG_ENDIAN
 #define SUBKEY_L(INDEX) (((u32*)subkey)[(INDEX)*2])
 #define SUBKEY_R(INDEX) (((u32*)subkey)[(INDEX)*2 + 1])
+#define subL(INDEX) (((u32*)sub)[(INDEX)*2])
+#define subR(INDEX) (((u32*)sub)[(INDEX)*2 + 1])
 #else
 #define SUBKEY_L(INDEX) (((u32*)subkey)[(INDEX)*2 + 1])
 #define SUBKEY_R(INDEX) (((u32*)subkey)[(INDEX)*2])
+#define subL(INDEX) (((u32*)sub)[(INDEX)*2 + 1])
+#define subR(INDEX) (((u32*)sub)[(INDEX)*2])
 #endif
 
-static void camellia_setup_tail(u64 *subkey, int max)
+static void camellia_setup_tail(u64 *subkey, u64 *sub, int max)
 {
+	u64 t;
 	u32 dw;
-	int i = 2;
+	int i;
+
+	/* key XOR is end of F-function */
+	SUBKEY(0) = sub[0] ^ sub[2];/* kw1 */
+	SUBKEY(2) = sub[3];       /* round 1 */
+	SUBKEY(3) = sub[2] ^ sub[4]; /* round 2 */
+	SUBKEY(4) = sub[3] ^ sub[5]; /* round 3 */
+	SUBKEY(5) = sub[4] ^ sub[6]; /* round 4 */
+	SUBKEY(6) = sub[5] ^ sub[7]; /* round 5 */
+	t = subL(10) ^ (subR(10) & ~subR(8)); // tl = subL[10] ^ (subR[10] & ~subR[8]);
+	dw = (u32)t & subL(8);  /* FL(kl1) */
+	t = (t << 32) | (subR(10) ^ ROL1(dw)); // tr = subR[10] ^ ROL1(dw);
+	SUBKEY(7) = sub[6] ^ t;   /* round 6 */
+	SUBKEY(8) = sub[8];       /* FL(kl1) */
+	SUBKEY(9) = sub[9];       /* FLinv(kl2) */
+	t = subL(7) ^ (subR(7) & ~subR(9));
+	dw = (u32)t & subL(9);  /* FLinv(kl2) */
+	t = (t << 32) | (subR(7) ^ ROL1(dw));
+	SUBKEY(10) = t ^ sub[11]; /* round 7 */
+	SUBKEY(11) = sub[10] ^ sub[12]; /* round 8 */
+	SUBKEY(12) = sub[11] ^ sub[13]; /* round 9 */
+	SUBKEY(13) = sub[12] ^ sub[14]; /* round 10 */
+	SUBKEY(14) = sub[13] ^ sub[15]; /* round 11 */
+	t = subL(18) ^ (subR(18) & ~subR(16));
+	dw = (u32)t & subL(16); /* FL(kl3) */
+	t = (t << 32) | (subR(18) ^ ROL1(dw));
+	SUBKEY(15) = sub[14] ^ t; /* round 12 */
+	SUBKEY(16) = sub[16];     /* FL(kl3) */
+	SUBKEY(17) = sub[17];     /* FLinv(kl4) */
+	t = subL(15) ^ (subR(15) & ~subR(17));
+	dw = (u32)t & subL(17); /* FLinv(kl4) */
+	t = (t << 32) | (subR(15) ^ ROL1(dw));
+	SUBKEY(18) = t ^ sub[19]; /* round 13 */
+	SUBKEY(19) = sub[18] ^ sub[20]; /* round 14 */
+	SUBKEY(20) = sub[19] ^ sub[21]; /* round 15 */
+	SUBKEY(21) = sub[20] ^ sub[22]; /* round 16 */
+	SUBKEY(22) = sub[21] ^ sub[23]; /* round 17 */
+	if (max == 24) {
+		SUBKEY(23) = sub[22];     /* round 18 */
+		SUBKEY(24) = sub[24] ^ sub[23]; /* kw3 */
+	} else { 
+		t = subL(26) ^ (subR(26) & ~subR(24));
+		dw = (u32)t & subL(24); /* FL(kl5) */
+		t = (t << 32) | (subR(26) ^ ROL1(dw));
+		SUBKEY(23) = sub[22] ^ t; /* round 18 */
+		SUBKEY(24) = sub[24];     /* FL(kl5) */
+		SUBKEY(25) = sub[25];     /* FLinv(kl6) */
+		t = subL(23) ^ (subR(23) & ~subR(25));
+		dw = (u32)t & subL(25); /* FLinv(kl6) */
+		t = (t << 32) | (subR(23) ^ ROL1(dw));
+		SUBKEY(26) = t ^ sub[27]; /* round 19 */
+		SUBKEY(27) = sub[26] ^ sub[28]; /* round 20 */
+		SUBKEY(28) = sub[27] ^ sub[29]; /* round 21 */
+		SUBKEY(29) = sub[28] ^ sub[30]; /* round 22 */
+		SUBKEY(30) = sub[29] ^ sub[31]; /* round 23 */
+		SUBKEY(31) = sub[30];     /* round 24 */
+		SUBKEY(32) = sub[32] ^ sub[31]; /* kw3 */
+	}
+
+	/* apply the inverse of the last half of P-function */
+	i = 2;
 	do {
 		dw = SUBKEY_L(i + 0) ^ SUBKEY_R(i + 0); dw = ROL8(dw);/* round 1 */
 		SUBKEY_R(i + 0) = SUBKEY_L(i + 0) ^ dw; SUBKEY_L(i + 0) = dw;

@@ -406,31 +471,21 @@ static void camellia_setup_tail(u64 *sub
 	} while (i < max);
 }
 
-#ifdef __BIG_ENDIAN
-#define subL(INDEX) (((u32*)sub)[(INDEX)*2])
-#define subR(INDEX) (((u32*)sub)[(INDEX)*2 + 1])
-#else
-#define subL(INDEX) (((u32*)sub)[(INDEX)*2 + 1])
-#define subR(INDEX) (((u32*)sub)[(INDEX)*2])
-#endif
-
 static void camellia_setup128(const unsigned char *key, u64 *subkey)
 {
 	u64 kl, kr;
-	u64 i, t, w;
+	u64 i, w;
 	u64 kw4;
 	u32 dw;
 	u64 sub[26];
 
 	/**
-	 *  k == kl || kr (|| is concatination)
+	 *  k == kl || kr (|| is concatenation)
 	 */
 	GETU64(kl, key     );
 	GETU64(kr, key +  8);
 
-	/**
-	 * generate KL dependent subkeys
-	 */
+	/* generate KL dependent subkeys */
 	/* kw1 */
 	sub[0] = kl;
 	/* kw2 */

@@ -567,60 +622,21 @@ static void camellia_setup128(const unsi
 	/* kw1 */
 	sub[0] ^= kw4;
 
-	/* key XOR is end of F-function */
-	SUBKEY(0) = sub[0] ^ sub[2];/* kw1 */
-	SUBKEY(2) = sub[3];       /* round 1 */
-	SUBKEY(3) = sub[2] ^ sub[4]; /* round 2 */
-	SUBKEY(4) = sub[3] ^ sub[5]; /* round 3 */
-	SUBKEY(5) = sub[4] ^ sub[6]; /* round 4 */
-	SUBKEY(6) = sub[5] ^ sub[7]; /* round 5 */
-	t = subL(10) ^ (subR(10) & ~subR(8)); // tl = subL[10] ^ (subR[10] & ~subR[8]);
-	dw = (u32)t & subL(8);  /* FL(kl1) */
-	t = (t << 32) | (subR(10) ^ ROL1(dw)); // tr = subR[10] ^ ROL1(dw);
-	SUBKEY(7) = sub[6] ^ t; /* round 6 */
-	SUBKEY(8) = sub[8];       /* FL(kl1) */
-	SUBKEY(9) = sub[9];       /* FLinv(kl2) */
-	t = subL(7) ^ (subR(7) & ~subR(9));
-	dw = (u32)t & subL(9);  /* FLinv(kl2) */
-	t = (t << 32) | (subR(7) ^ ROL1(dw));
-	SUBKEY(10) = t ^ sub[11]; /* round 7 */
-	SUBKEY(11) = sub[10] ^ sub[12]; /* round 8 */
-	SUBKEY(12) = sub[11] ^ sub[13]; /* round 9 */
-	SUBKEY(13) = sub[12] ^ sub[14]; /* round 10 */
-	SUBKEY(14) = sub[13] ^ sub[15]; /* round 11 */
-	t = subL(18) ^ (subR(18) & ~subR(16));
-	dw = (u32)t & subL(16); /* FL(kl3) */
-	t = (t << 32) | (subR(18) ^ ROL1(dw));
-	SUBKEY(15) = sub[14] ^ t; /* round 12 */
-	SUBKEY(16) = sub[16];     /* FL(kl3) */
-	SUBKEY(17) = sub[17];     /* FLinv(kl4) */
-	t = subL(15) ^ (subR(15) & ~subR(17));
-	dw = (u32)t & subL(17); /* FLinv(kl4) */
-	t = (t << 32) | (subR(15) ^ ROL1(dw));
-	SUBKEY(18) = t ^ sub[19]; /* round 13 */
-	SUBKEY(19) = sub[18] ^ sub[20]; /* round 14 */
-	SUBKEY(20) = sub[19] ^ sub[21]; /* round 15 */
-	SUBKEY(21) = sub[20] ^ sub[22]; /* round 16 */
-	SUBKEY(22) = sub[21] ^ sub[23]; /* round 17 */
-	SUBKEY(23) = sub[22];     /* round 18 */
-	SUBKEY(24) = sub[24] ^ sub[23]; /* kw3 */
-
-	/* apply the inverse of the last half of P-function */
-	camellia_setup_tail(subkey, 24);
+	camellia_setup_tail(subkey, sub, 24);
 }
 
 static void camellia_setup256(const unsigned char *key, u64 *subkey)
 {
 	u64 kl, kr;        /* left half of key */
 	u64 krl, krr;      /* right half of key */
-	u64 i, t, w;       /* temporary variables */
+	u64 i, w;          /* temporary variables */
 	u64 kw4;
 	u32 dw;
 	u64 sub[34];
 
 	/**
 	 *  key = (kl || kr || krl || krr)
-	 *  (|| is concatination)
+	 *  (|| is concatenation)
 	 */
 	GETU64(kl,  key     );
 	GETU64(kr,  key +  8);

@@ -786,8 +802,8 @@ static void camellia_setup256(const unsi
 	/* round 19 */
 	sub[26] ^= kw4;
 	kw4 ^= (u64)((u32)kw4 & ~subR(24)) << 32; //kw4l ^= kw4r & ~subR[24];
-	dw = (u32)(kw4 >> 32) & subL(24),
-		kw4 ^= ROL1(dw); /* modified for FL(kl5) */
+	dw = (u32)(kw4 >> 32) & subL(24);
+	kw4 ^= ROL1(dw); /* modified for FL(kl5) */
 	/* round 17 */
 	sub[22] ^= kw4;
 	/* round 15 */

@@ -795,8 +811,8 @@ static void camellia_setup256(const unsi
 	/* round 13 */
 	sub[18] ^= kw4;
 	kw4 ^= (u64)((u32)kw4 & ~subR(16)) << 32;
-	dw = (u32)(kw4 >> 32) & subL(16),
-		kw4 ^= ROL1(dw); /* modified for FL(kl3) */
+	dw = (u32)(kw4 >> 32) & subL(16);
+	kw4 ^= ROL1(dw); /* modified for FL(kl3) */
 	/* round 11 */
 	sub[14] ^= kw4;
 	/* round 9 */

@@ -804,8 +820,8 @@ static void camellia_setup256(const unsi
 	/* round 7 */
 	sub[10] ^= kw4;
 	kw4 ^= (u64)((u32)kw4 & ~subR(8)) << 32;
-	dw = (u32)(kw4 >> 32) & subL(8),
-		kw4 ^= ROL1(dw); /* modified for FL(kl1) */
+	dw = (u32)(kw4 >> 32) & subL(8);
+	kw4 ^= ROL1(dw); /* modified for FL(kl1) */
 	/* round 5 */
 	sub[6] ^= kw4;
 	/* round 3 */

@@ -815,60 +831,7 @@ static void camellia_setup256(const unsi
 	/* kw1 */
 	sub[0] ^= kw4;
 
-	/* key XOR is end of F-function */
-	SUBKEY(0) = sub[0] ^ sub[2];/* kw1 */
-	SUBKEY(2) = sub[3];       /* round 1 */
-	SUBKEY(3) = sub[2] ^ sub[4]; /* round 2 */
-	SUBKEY(4) = sub[3] ^ sub[5]; /* round 3 */
-	SUBKEY(5) = sub[4] ^ sub[6]; /* round 4 */
-	SUBKEY(6) = sub[5] ^ sub[7]; /* round 5 */
-	t = subL(10) ^ (subR(10) & ~subR(8)); // tl = subL[10] ^ (subR[10] & ~subR[8]);
-	dw = (u32)t & subL(8);  /* FL(kl1) */
-	t = (t << 32) | (subR(10) ^ ROL1(dw)); //tr = subR[10] ^ ROL1(dw);
-	SUBKEY(7) = sub[6] ^ t;   /* round 6 */
-	SUBKEY(8) = sub[8];       /* FL(kl1) */
-	SUBKEY(9) = sub[9];       /* FLinv(kl2) */
-	t = subL(7) ^ (subR(7) & ~subR(9));
-	dw = (u32)t & subL(9);  /* FLinv(kl2) */
-	t = (t << 32) | (subR(7) ^ ROL1(dw));
-	SUBKEY(10) = t ^ sub[11]; /* round 7 */
-	SUBKEY(11) = sub[10] ^ sub[12]; /* round 8 */
-	SUBKEY(12) = sub[11] ^ sub[13]; /* round 9 */
-	SUBKEY(13) = sub[12] ^ sub[14]; /* round 10 */
-	SUBKEY(14) = sub[13] ^ sub[15]; /* round 11 */
-	t = subL(18) ^ (subR(18) & ~subR(16));
-	dw = (u32)t & subL(16); /* FL(kl3) */
-	t = (t << 32) | (subR(18) ^ ROL1(dw));
-	SUBKEY(15) = sub[14] ^ t; /* round 12 */
-	SUBKEY(16) = sub[16];     /* FL(kl3) */
-	SUBKEY(17) = sub[17];     /* FLinv(kl4) */
-	t = subL(15) ^ (subR(15) & ~subR(17));
-	dw = (u32)t & subL(17); /* FLinv(kl4) */
-	t = (t << 32) | (subR(15) ^ ROL1(dw));
-	SUBKEY(18) = t ^ sub[19]; /* round 13 */
-	SUBKEY(19) = sub[18] ^ sub[20]; /* round 14 */
-	SUBKEY(20) = sub[19] ^ sub[21]; /* round 15 */
-	SUBKEY(21) = sub[20] ^ sub[22]; /* round 16 */
-	SUBKEY(22) = sub[21] ^ sub[23]; /* round 17 */
-	t = subL(26) ^ (subR(26) & ~subR(24));
-	dw = (u32)t & subL(24); /* FL(kl5) */
-	t = (t << 32) | (subR(26) ^ ROL1(dw));
-	SUBKEY(23) = sub[22] ^ t; /* round 18 */
-	SUBKEY(24) = sub[24];     /* FL(kl5) */
-	SUBKEY(25) = sub[25];     /* FLinv(kl6) */
-	t = subL(23) ^ (subR(23) & ~subR(25));
-	dw = (u32)t & subL(25); /* FLinv(kl6) */
-	t = (t << 32) | (subR(23) ^ ROL1(dw));
-	SUBKEY(26) = t ^ sub[27]; /* round 19 */
-	SUBKEY(27) = sub[26] ^ sub[28]; /* round 20 */
-	SUBKEY(28) = sub[27] ^ sub[29]; /* round 21 */
-	SUBKEY(29) = sub[28] ^ sub[30]; /* round 22 */
-	SUBKEY(30) = sub[29] ^ sub[31]; /* round 23 */
-	SUBKEY(31) = sub[30];     /* round 24 */
-	SUBKEY(32) = sub[32] ^ sub[31]; /* kw3 */
-
-	/* apply the inverse of the last half of P-function */
-	camellia_setup_tail(subkey, 32);
+	camellia_setup_tail(subkey, sub, 32);
 }
 
 static void camellia_setup192(const unsigned char *key, u64 *subkey)

@@ -967,10 +930,104 @@ typedef const u64 const_key_element;
 #define SUBKEY_L(INDEX) (subkey[(INDEX)*2])
 #define SUBKEY_R(INDEX) (subkey[(INDEX)*2 + 1])
 
-static void camellia_setup_tail(u32 *subkey, int max)
+static void camellia_setup_tail(u32 *subkey, u32 *subL, u32 *subR, int max)
 {
-	u32 dw;
-	int i = 2;
+	u32 dw, tl, tr;
+	int i;
+
+	/* key XOR is end of F-function */
+	SUBKEY_L(0) = subL[0] ^ subL[2];/* kw1 */
+	SUBKEY_R(0) = subR[0] ^ subR[2];
+	SUBKEY_L(2) = subL[3];       /* round 1 */
+	SUBKEY_R(2) = subR[3];
+	SUBKEY_L(3) = subL[2] ^ subL[4]; /* round 2 */
+	SUBKEY_R(3) = subR[2] ^ subR[4];
+	SUBKEY_L(4) = subL[3] ^ subL[5]; /* round 3 */
+	SUBKEY_R(4) = subR[3] ^ subR[5];
+	SUBKEY_L(5) = subL[4] ^ subL[6]; /* round 4 */
+	SUBKEY_R(5) = subR[4] ^ subR[6];
+	SUBKEY_L(6) = subL[5] ^ subL[7]; /* round 5 */
+	SUBKEY_R(6) = subR[5] ^ subR[7];
+	tl = subL[10] ^ (subR[10] & ~subR[8]);
+	dw = tl & subL[8],  /* FL(kl1) */
+		tr = subR[10] ^ ROL1(dw);
+	SUBKEY_L(7) = subL[6] ^ tl; /* round 6 */
+	SUBKEY_R(7) = subR[6] ^ tr;
+	SUBKEY_L(8) = subL[8];       /* FL(kl1) */
+	SUBKEY_R(8) = subR[8];
+	SUBKEY_L(9) = subL[9];       /* FLinv(kl2) */
+	SUBKEY_R(9) = subR[9];
+	tl = subL[7] ^ (subR[7] & ~subR[9]);
+	dw = tl & subL[9],  /* FLinv(kl2) */
+		tr = subR[7] ^ ROL1(dw);
+	SUBKEY_L(10) = tl ^ subL[11]; /* round 7 */
+	SUBKEY_R(10) = tr ^ subR[11];
+	SUBKEY_L(11) = subL[10] ^ subL[12]; /* round 8 */
+	SUBKEY_R(11) = subR[10] ^ subR[12];
+	SUBKEY_L(12) = subL[11] ^ subL[13]; /* round 9 */
+	SUBKEY_R(12) = subR[11] ^ subR[13];
+	SUBKEY_L(13) = subL[12] ^ subL[14]; /* round 10 */
+	SUBKEY_R(13) = subR[12] ^ subR[14];
+	SUBKEY_L(14) = subL[13] ^ subL[15]; /* round 11 */
+	SUBKEY_R(14) = subR[13] ^ subR[15];
+	tl = subL[18] ^ (subR[18] & ~subR[16]);
+	dw = tl & subL[16], /* FL(kl3) */
+		tr = subR[18] ^ ROL1(dw);
+	SUBKEY_L(15) = subL[14] ^ tl; /* round 12 */
+	SUBKEY_R(15) = subR[14] ^ tr;
+	SUBKEY_L(16) = subL[16];     /* FL(kl3) */
+	SUBKEY_R(16) = subR[16];
+	SUBKEY_L(17) = subL[17];     /* FLinv(kl4) */
+	SUBKEY_R(17) = subR[17];
+	tl = subL[15] ^ (subR[15] & ~subR[17]);
+	dw = tl & subL[17], /* FLinv(kl4) */
+		tr = subR[15] ^ ROL1(dw);
+	SUBKEY_L(18) = tl ^ subL[19]; /* round 13 */
+	SUBKEY_R(18) = tr ^ subR[19];
+	SUBKEY_L(19) = subL[18] ^ subL[20]; /* round 14 */
+	SUBKEY_R(19) = subR[18] ^ subR[20];
+	SUBKEY_L(20) = subL[19] ^ subL[21]; /* round 15 */
+	SUBKEY_R(20) = subR[19] ^ subR[21];
+	SUBKEY_L(21) = subL[20] ^ subL[22]; /* round 16 */
+	SUBKEY_R(21) = subR[20] ^ subR[22];
+	SUBKEY_L(22) = subL[21] ^ subL[23]; /* round 17 */
+	SUBKEY_R(22) = subR[21] ^ subR[23];
+	if (max == 24) {
+		SUBKEY_L(23) = subL[22];     /* round 18 */
+		SUBKEY_R(23) = subR[22];
+		SUBKEY_L(24) = subL[24] ^ subL[23]; /* kw3 */
+		SUBKEY_R(24) = subR[24] ^ subR[23];
+	} else {
+		tl = subL[26] ^ (subR[26] & ~subR[24]);
+		dw = tl & subL[24], /* FL(kl5) */
+			tr = subR[26] ^ ROL1(dw);
+		SUBKEY_L(23) = subL[22] ^ tl; /* round 18 */
+		SUBKEY_R(23) = subR[22] ^ tr;
+		SUBKEY_L(24) = subL[24];     /* FL(kl5) */
+		SUBKEY_R(24) = subR[24];
+		SUBKEY_L(25) = subL[25];     /* FLinv(kl6) */
+		SUBKEY_R(25) = subR[25];
+		tl = subL[23] ^ (subR[23] & ~subR[25]);
+		dw = tl & subL[25], /* FLinv(kl6) */
+			tr = subR[23] ^ ROL1(dw);
+		SUBKEY_L(26) = tl ^ subL[27]; /* round 19 */
+		SUBKEY_R(26) = tr ^ subR[27];
+		SUBKEY_L(27) = subL[26] ^ subL[28]; /* round 20 */
+		SUBKEY_R(27) = subR[26] ^ subR[28];
+		SUBKEY_L(28) = subL[27] ^ subL[29]; /* round 21 */
+		SUBKEY_R(28) = subR[27] ^ subR[29];
+		SUBKEY_L(29) = subL[28] ^ subL[30]; /* round 22 */
+		SUBKEY_R(29) = subR[28] ^ subR[30];
+		SUBKEY_L(30) = subL[29] ^ subL[31]; /* round 23 */
+		SUBKEY_R(30) = subR[29] ^ subR[31];
+		SUBKEY_L(31) = subL[30];     /* round 24 */
+		SUBKEY_R(31) = subR[30];
+		SUBKEY_L(32) = subL[32] ^ subL[31]; /* kw3 */
+		SUBKEY_R(32) = subR[32] ^ subR[31];
+	}
+
+	/* apply the inverse of the last half of P-function */
+	i = 2;
 	do {
 		dw = SUBKEY_L(i + 0) ^ SUBKEY_R(i + 0); dw = ROL8(dw);/* round 1 */
 		SUBKEY_R(i + 0) = SUBKEY_L(i + 0) ^ dw; SUBKEY_L(i + 0) = dw;

@@ -992,21 +1049,19 @@ static void camellia_setup128(const unsi
 {
 	u32 kll, klr, krl, krr;
 	u32 il, ir, t0, t1, w0, w1;
-	u32 kw4l, kw4r, dw, tl, tr;
+	u32 kw4l, kw4r, dw;
 	u32 subL[26];
 	u32 subR[26];
 
 	/**
-	 *  k == kll || klr || krl || krr (|| is concatination)
+	 *  k == kll || klr || krl || krr (|| is concatenation)
 	 */
 	GETU32(kll, key     );
 	GETU32(klr, key +  4);
 	GETU32(krl, key +  8);
 	GETU32(krr, key + 12);
 
-	/**
-	 * generate KL dependent subkeys
-	 */
+	/* generate KL dependent subkeys */
 	/* kw1 */
 	subL[0] = kll; subR[0] = klr;
 	/* kw2 */

@@ -1151,70 +1206,7 @@ static void camellia_setup128(const unsi
 	/* kw1 */
 	subL[0] ^= kw4l; subR[0] ^= kw4r;
 
-	/* key XOR is end of F-function */
-	SUBKEY_L(0) = subL[0] ^ subL[2];/* kw1 */
-	SUBKEY_R(0) = subR[0] ^ subR[2];
-	SUBKEY_L(2) = subL[3];       /* round 1 */
-	SUBKEY_R(2) = subR[3];
-	SUBKEY_L(3) = subL[2] ^ subL[4]; /* round 2 */
-	SUBKEY_R(3) = subR[2] ^ subR[4];
-	SUBKEY_L(4) = subL[3] ^ subL[5]; /* round 3 */
-	SUBKEY_R(4) = subR[3] ^ subR[5];
-	SUBKEY_L(5) = subL[4] ^ subL[6]; /* round 4 */
-	SUBKEY_R(5) = subR[4] ^ subR[6];
-	SUBKEY_L(6) = subL[5] ^ subL[7]; /* round 5 */
-	SUBKEY_R(6) = subR[5] ^ subR[7];
-	tl = subL[10] ^ (subR[10] & ~subR[8]);
-	dw = tl & subL[8],  /* FL(kl1) */
-		tr = subR[10] ^ ROL1(dw);
-	SUBKEY_L(7) = subL[6] ^ tl; /* round 6 */
-	SUBKEY_R(7) = subR[6] ^ tr;
-	SUBKEY_L(8) = subL[8];       /* FL(kl1) */
-	SUBKEY_R(8) = subR[8];
-	SUBKEY_L(9) = subL[9];       /* FLinv(kl2) */
-	SUBKEY_R(9) = subR[9];
-	tl = subL[7] ^ (subR[7] & ~subR[9]);
-	dw = tl & subL[9],  /* FLinv(kl2) */
-		tr = subR[7] ^ ROL1(dw);
-	SUBKEY_L(10) = tl ^ subL[11]; /* round 7 */
-	SUBKEY_R(10) = tr ^ subR[11];
-	SUBKEY_L(11) = subL[10] ^ subL[12]; /* round 8 */
-	SUBKEY_R(11) = subR[10] ^ subR[12];
-	SUBKEY_L(12) = subL[11] ^ subL[13]; /* round 9 */
-	SUBKEY_R(12) = subR[11] ^ subR[13];
-	SUBKEY_L(13) = subL[12] ^ subL[14]; /* round 10 */
-	SUBKEY_R(13) = subR[12] ^ subR[14];
-	SUBKEY_L(14) = subL[13] ^ subL[15]; /* round 11 */
-	SUBKEY_R(14) = subR[13] ^ subR[15];
-	tl = subL[18] ^ (subR[18] & ~subR[16]);
-	dw = tl & subL[16], /* FL(kl3) */
-		tr = subR[18] ^ ROL1(dw);
-	SUBKEY_L(15) = subL[14] ^ tl; /* round 12 */
-	SUBKEY_R(15) = subR[14] ^ tr;
-	SUBKEY_L(16) = subL[16];     /* FL(kl3) */
-	SUBKEY_R(16) = subR[16];
-	SUBKEY_L(17) = subL[17];     /* FLinv(kl4) */
-	SUBKEY_R(17) = subR[17];
-	tl = subL[15] ^ (subR[15] & ~subR[17]);
-	dw = tl & subL[17], /* FLinv(kl4) */
-		tr = subR[15] ^ ROL1(dw);
-	SUBKEY_L(18) = tl ^ subL[19]; /* round 13 */
-	SUBKEY_R(18) = tr ^ subR[19];
-	SUBKEY_L(19) = subL[18] ^ subL[20]; /* round 14 */
-	SUBKEY_R(19) = subR[18] ^ subR[20];
-	SUBKEY_L(20) = subL[19] ^ subL[21]; /* round 15 */
-	SUBKEY_R(20) = subR[19] ^ subR[21];
-	SUBKEY_L(21) = subL[20] ^ subL[22]; /* round 16 */
-	SUBKEY_R(21) = subR[20] ^ subR[22];
-	SUBKEY_L(22) = subL[21] ^ subL[23]; /* round 17 */
-	SUBKEY_R(22) = subR[21] ^ subR[23];
-	SUBKEY_L(23) = subL[22];     /* round 18 */
-	SUBKEY_R(23) = subR[22];
-	SUBKEY_L(24) = subL[24] ^ subL[23]; /* kw3 */
-	SUBKEY_R(24) = subR[24] ^ subR[23];
-
-	/* apply the inverse of the last half of P-function */
-	camellia_setup_tail(subkey, 24);
+	camellia_setup_tail(subkey, subL, subR, 24);
 }
 
 static void camellia_setup256(const unsigned char *key, u32 *subkey)

@@ -1222,13 +1214,13 @@ static void camellia_setup256(const unsi
 	u32 kll, klr, krl, krr;        /* left half of key */
 	u32 krll, krlr, krrl, krrr;    /* right half of key */
 	u32 il, ir, t0, t1, w0, w1;    /* temporary variables */
-	u32 kw4l, kw4r, dw, tl, tr;
+	u32 kw4l, kw4r, dw;
 	u32 subL[34];
 	u32 subR[34];
 
 	/**
 	 *  key = (kll || klr || krl || krr || krll || krlr || krrl || krrr)
-	 *  (|| is concatination)
+	 *  (|| is concatenation)
 	 */
 	GETU32(kll,  key     );
 	GETU32(klr,  key +  4);

@@ -1439,92 +1431,7 @@ static void camellia_setup256(const unsi
 	/* kw1 */
 	subL[0] ^= kw4l; subR[0] ^= kw4r;
 
-	/* key XOR is end of F-function */
-	SUBKEY_L(0) = subL[0] ^ subL[2];/* kw1 */
-	SUBKEY_R(0) = subR[0] ^ subR[2];
-	SUBKEY_L(2) = subL[3];       /* round 1 */
-	SUBKEY_R(2) = subR[3];
-	SUBKEY_L(3) = subL[2] ^ subL[4]; /* round 2 */
-	SUBKEY_R(3) = subR[2] ^ subR[4];
-	SUBKEY_L(4) = subL[3] ^ subL[5]; /* round 3 */
-	SUBKEY_R(4) = subR[3] ^ subR[5];
-	SUBKEY_L(5) = subL[4] ^ subL[6]; /* round 4 */
-	SUBKEY_R(5) = subR[4] ^ subR[6];
-	SUBKEY_L(6) = subL[5] ^ subL[7]; /* round 5 */
-	SUBKEY_R(6) = subR[5] ^ subR[7];
-	tl = subL[10] ^ (subR[10] & ~subR[8]);
-	dw = tl & subL[8],  /* FL(kl1) */
-		tr = subR[10] ^ ROL1(dw);
-	SUBKEY_L(7) = subL[6] ^ tl; /* round 6 */
-	SUBKEY_R(7) = subR[6] ^ tr;
-	SUBKEY_L(8) = subL[8];       /* FL(kl1) */
-	SUBKEY_R(8) = subR[8];
-	SUBKEY_L(9) = subL[9];       /* FLinv(kl2) */
-	SUBKEY_R(9) = subR[9];
-	tl = subL[7] ^ (subR[7] & ~subR[9]);
-	dw = tl & subL[9],  /* FLinv(kl2) */
-		tr = subR[7] ^ ROL1(dw);
-	SUBKEY_L(10) = tl ^ subL[11]; /* round 7 */
-	SUBKEY_R(10) = tr ^ subR[11];
-	SUBKEY_L(11) = subL[10] ^ subL[12]; /* round 8 */
-	SUBKEY_R(11) = subR[10] ^ subR[12];
-	SUBKEY_L(12) = subL[11] ^ subL[13]; /* round 9 */
-	SUBKEY_R(12) = subR[11] ^ subR[13];
-	SUBKEY_L(13) = subL[12] ^ subL[14]; /* round 10 */
-	SUBKEY_R(13) = subR[12] ^ subR[14];
-	SUBKEY_L(14) = subL[13] ^ subL[15]; /* round 11 */
-	SUBKEY_R(14) = subR[13] ^ subR[15];
-	tl = subL[18] ^ (subR[18] & ~subR[16]);
-	dw = tl & subL[16], /* FL(kl3) */
-		tr = subR[18] ^ ROL1(dw);
-	SUBKEY_L(15) = subL[14] ^ tl; /* round 12 */
-	SUBKEY_R(15) = subR[14] ^ tr;
-	SUBKEY_L(16) = subL[16];     /* FL(kl3) */
-	SUBKEY_R(16) = subR[16];
-	SUBKEY_L(17) = subL[17];     /* FLinv(kl4) */
-	SUBKEY_R(17) = subR[17];
-	tl = subL[15] ^ (subR[15] & ~subR[17]);
-	dw = tl & subL[17], /* FLinv(kl4) */
-		tr = subR[15] ^ ROL1(dw);
-	SUBKEY_L(18) = tl ^ subL[19]; /* round 13 */
-	SUBKEY_R(18) = tr ^ subR[19];
-	SUBKEY_L(19) = subL[18] ^ subL[20]; /* round 14 */
-	SUBKEY_R(19) = subR[18] ^ subR[20];
-	SUBKEY_L(20) = subL[19] ^ subL[21]; /* round 15 */
-	SUBKEY_R(20) = subR[19] ^ subR[21];
-	SUBKEY_L(21) = subL[20] ^ subL[22]; /* round 16 */
-	SUBKEY_R(21) = subR[20] ^ subR[22];
-	SUBKEY_L(22) = subL[21] ^ subL[23]; /* round 17 */
-	SUBKEY_R(22) = subR[21] ^ subR[23];
-	tl = subL[26] ^ (subR[26] & ~subR[24]);
-	dw = tl & subL[24], /* FL(kl5) */
-		tr = subR[26] ^ ROL1(dw);
-	SUBKEY_L(23) = subL[22] ^ tl; /* round 18 */
-	SUBKEY_R(23) = subR[22] ^ tr;
-	SUBKEY_L(24) = subL[24];     /* FL(kl5) */
-	SUBKEY_R(24) = subR[24];
-	SUBKEY_L(25) = subL[25];     /* FLinv(kl6) */
-	SUBKEY_R(25) = subR[25];
-	tl = subL[23] ^ (subR[23] & ~subR[25]);
-	dw = tl & subL[25], /* FLinv(kl6) */
-		tr = subR[23] ^ ROL1(dw);
-	SUBKEY_L(26) = tl ^ subL[27]; /* round 19 */
-	SUBKEY_R(26) = tr ^ subR[27];
-	SUBKEY_L(27) = subL[26] ^ subL[28]; /* round 20 */
-	SUBKEY_R(27) = subR[26] ^ subR[28];
-	SUBKEY_L(28) = subL[27] ^ subL[29]; /* round 21 */
-	SUBKEY_R(28) = subR[27] ^ subR[29];
-	SUBKEY_L(29) = subL[28] ^ subL[30]; /* round 22 */
-	SUBKEY_R(29) = subR[28] ^ subR[30];
-	SUBKEY_L(30) = subL[29] ^ subL[31]; /* round 23 */
-	SUBKEY_R(30) = subR[29] ^ subR[31];
-	SUBKEY_L(31) = subL[30];     /* round 24 */
-	SUBKEY_R(31) = subR[30];
-	SUBKEY_L(32) = subL[32] ^ subL[31]; /* kw3 */
-	SUBKEY_R(32) = subR[32] ^ subR[31];
-
-	/* apply the inverse of the last half of P-function */
-	camellia_setup_tail(subkey, 32);
+	camellia_setup_tail(subkey, subL, subR, 32);
 }
 
 static void camellia_setup192(const unsigned char *key, u32 *subkey)
author
#7 lore raw

PATCH 4/4 camellia: code shrink 3

In reply to Denys Vlasenko
Quoted 6 lines · Denys Vlasenko

camellia8:
Analogously to camellia7 patch, move
"absorb kw2 to other subkeys" and "absorb kw4 to other subkeys" code parts into camellia_setup_tail(). This further reduces source and object code size at the cost of two brances in key setup code.

I also tried to more more code into tail, but it becomes more complicated. Up to this point is was straightforward.

Signed-off-by Denys Vlasenko <[email protected]>

I noticed that in [PATCH 2/4] and [PATCH 3/4] mails I attached linux-2.6.23.1xxxxxxxxxx.diff files.

This is not a problem, the patches are actually identical to cryptodev-2.6xxxxxxxxxxxx.diff ones, except for leading directory prefix.
I verified that they do apply without rejects, offsets or fuzz to cryptodev-2.6.

Signature · 1 line
vda
crypto/camellia.c +168 -270
diff -urpN cryptodev-2.6.camellia7/crypto/camellia.c cryptodev-2.6.camellia8/crypto/camellia.c
@@ -391,10 +391,94 @@ static const u32 camellia_sp4404[256] = 
 
 static void camellia_setup_tail(u64 *subkey, u64 *sub, int max)
 {
+	u64 kw4;
 	u64 t;
 	u32 dw;
 	int i;
 
+	/* absorb kw2 to other subkeys */
+	/* round 2 */
+	sub[3] ^= sub[1];
+	/* round 4 */
+	sub[5] ^= sub[1];
+	/* round 6 */
+	sub[7] ^= sub[1];
+	subL(1) ^= subR(1) & ~subR(9);
+	dw = subL(1) & subL(9),
+		subR(1) ^= ROL1(dw); /* modified for FLinv(kl2) */
+	/* round 8 */
+	sub[11] ^= sub[1];
+	/* round 10 */
+	sub[13] ^= sub[1];
+	/* round 12 */
+	sub[15] ^= sub[1];
+	subL(1) ^= subR(1) & ~subR(17);
+	dw = subL(1) & subL(17),
+		subR(1) ^= ROL1(dw); /* modified for FLinv(kl4) */
+	/* round 14 */
+	sub[19] ^= sub[1];
+	/* round 16 */
+	sub[21] ^= sub[1];
+	/* round 18 */
+	sub[23] ^= sub[1];
+	if (max == 24) {
+		/* kw3 */
+		sub[24] ^= sub[1];
+
+	/* absorb kw4 to other subkeys */
+		kw4 = sub[25];
+	} else {
+		subL(1) ^= subR(1) & ~subR(25);
+		dw = subL(1) & subL(25),
+			subR(1) ^= ROL1(dw); /* modified for FLinv(kl6) */
+		/* round 20 */
+		sub[27] ^= sub[1];
+		/* round 22 */
+		sub[29] ^= sub[1];
+		/* round 24 */
+		sub[31] ^= sub[1];
+		/* kw3 */
+		sub[32] ^= sub[1];
+
+	/* absorb kw4 to other subkeys */
+		kw4 = sub[33];
+		/* round 23 */
+		sub[30] ^= kw4;
+		/* round 21 */
+		sub[28] ^= kw4;
+		/* round 19 */
+		sub[26] ^= kw4;
+		kw4 ^= (u64)((u32)kw4 & ~subR(24)) << 32; //kw4l ^= kw4r & ~subR[24];
+		dw = (u32)(kw4 >> 32) & subL(24);
+		kw4 ^= ROL1(dw); /* modified for FL(kl5) */
+	}
+	/* round 17 */
+	sub[22] ^= kw4;
+	/* round 15 */
+	sub[20] ^= kw4;
+	/* round 13 */
+	sub[18] ^= kw4;
+	kw4 ^= (u64)((u32)kw4 & ~subR(16)) << 32; //kw4l ^= kw4r & ~subR(16);
+	dw = (u32)(kw4 >> 32) & subL(16); // kw4l & subL[16],
+	kw4 ^= ROL1(dw); /* modified for FL(kl3) */
+	/* round 11 */
+	sub[14] ^= kw4;
+	/* round 9 */
+	sub[12] ^= kw4;
+	/* round 7 */
+	sub[10] ^= kw4;
+	kw4 ^= (u64)((u32)kw4 & ~subR(8)) << 32; //kw4l ^= kw4r & ~subR[8];
+	dw = (u32)(kw4 >> 32) & subL(8);
+	kw4 ^= ROL1(dw); /* modified for FL(kl1) */
+	/* round 5 */
+	sub[6] ^= kw4;
+	/* round 3 */
+	sub[4] ^= kw4;
+	/* round 1 */
+	sub[2] ^= kw4;
+	/* kw1 */
+	sub[0] ^= kw4;
+
 	/* key XOR is end of F-function */
 	SUBKEY(0) = sub[0] ^ sub[2];/* kw1 */
 	SUBKEY(2) = sub[3];       /* round 1 */

@@ -475,8 +559,6 @@ static void camellia_setup128(const unsi
 {
 	u64 kl, kr;
 	u64 i, w;
-	u64 kw4;
-	u32 dw;
 	u64 sub[26];
 
 	/**

@@ -565,63 +647,6 @@ static void camellia_setup128(const unsi
 	sub[24] = kl;
 	sub[25] = kr;
 
-	/* absorb kw2 to other subkeys */
-	/* round 2 */
-	sub[3] ^= sub[1];
-	/* round 4 */
-	sub[5] ^= sub[1];
-	/* round 6 */
-	sub[7] ^= sub[1];
-	subL(1) ^= subR(1) & ~subR(9);
-	dw = subL(1) & subL(9),
-		subR(1) ^= ROL1(dw); /* modified for FLinv(kl2) */
-	/* round 8 */
-	sub[11] ^= sub[1];
-	/* round 10 */
-	sub[13] ^= sub[1];
-	/* round 12 */
-	sub[15] ^= sub[1];
-	subL(1) ^= subR(1) & ~subR(17);
-	dw = subL(1) & subL(17),
-		subR(1) ^= ROL1(dw); /* modified for FLinv(kl4) */
-	/* round 14 */
-	sub[19] ^= sub[1];
-	/* round 16 */
-	sub[21] ^= sub[1];
-	/* round 18 */
-	sub[23] ^= sub[1];
-	/* kw3 */
-	sub[24] ^= sub[1];
-
-	/* absorb kw4 to other subkeys */
-	kw4 = sub[25];
-	/* round 17 */
-	sub[22] ^= kw4;
-	/* round 15 */
-	sub[20] ^= kw4;
-	/* round 13 */
-	sub[18] ^= kw4;
-	kw4 ^= (u64)((u32)kw4 & ~subR(16)) << 32; //kw4l ^= kw4r & ~subR(16);
-	dw = (u32)(kw4 >> 32) & subL(16); // kw4l & subL[16],
-	kw4 ^= ROL1(dw); /* modified for FL(kl3) */
-	/* round 11 */
-	sub[14] ^= kw4;
-	/* round 9 */
-	sub[12] ^= kw4;
-	/* round 7 */
-	sub[10] ^= kw4;
-	kw4 ^= (u64)((u32)kw4 & ~subR(8)) << 32; //kw4l ^= kw4r & ~subR[8];
-	dw = (u32)(kw4 >> 32) & subL(8);
-	kw4 ^= ROL1(dw); /* modified for FL(kl1) */
-	/* round 5 */
-	sub[6] ^= kw4;
-	/* round 3 */
-	sub[4] ^= kw4;
-	/* round 1 */
-	sub[2] ^= kw4;
-	/* kw1 */
-	sub[0] ^= kw4;
-
 	camellia_setup_tail(subkey, sub, 24);
 }
 

@@ -630,8 +655,6 @@ static void camellia_setup256(const unsi
 	u64 kl, kr;        /* left half of key */
 	u64 krl, krr;      /* right half of key */
 	u64 i, w;          /* temporary variables */
-	u64 kw4;
-	u32 dw;
 	u64 sub[34];
 
 	/**

@@ -756,81 +779,6 @@ static void camellia_setup256(const unsi
 	/* kw4 */
 	sub[33] = krr;
 
-	/* absorb kw2 to other subkeys */
-	/* round 2 */
-	sub[3] ^= sub[1];
-	/* round 4 */
-	sub[5] ^= sub[1];
-	/* round 6 */
-	sub[7] ^= sub[1];
-	subL(1) ^= subR(1) & ~subR(9);
-	dw = subL(1) & subL(9),
-		subR(1) ^= ROL1(dw); /* modified for FLinv(kl2) */
-	/* round 8 */
-	sub[11] ^= sub[1];
-	/* round 10 */
-	sub[13] ^= sub[1];
-	/* round 12 */
-	sub[15] ^= sub[1];
-	subL(1) ^= subR(1) & ~subR(17);
-	dw = subL(1) & subL(17),
-		subR(1) ^= ROL1(dw); /* modified for FLinv(kl4) */
-	/* round 14 */
-	sub[19] ^= sub[1];
-	/* round 16 */
-	sub[21] ^= sub[1];
-	/* round 18 */
-	sub[23] ^= sub[1];
-	subL(1) ^= subR(1) & ~subR(25);
-	dw = subL(1) & subL(25),
-		subR(1) ^= ROL1(dw); /* modified for FLinv(kl6) */
-	/* round 20 */
-	sub[27] ^= sub[1];
-	/* round 22 */
-	sub[29] ^= sub[1];
-	/* round 24 */
-	sub[31] ^= sub[1];
-	/* kw3 */
-	sub[32] ^= sub[1];
-
-	/* absorb kw4 to other subkeys */
-	kw4 = sub[33];
-	/* round 23 */
-	sub[30] ^= kw4;
-	/* round 21 */
-	sub[28] ^= kw4;
-	/* round 19 */
-	sub[26] ^= kw4;
-	kw4 ^= (u64)((u32)kw4 & ~subR(24)) << 32; //kw4l ^= kw4r & ~subR[24];
-	dw = (u32)(kw4 >> 32) & subL(24);
-	kw4 ^= ROL1(dw); /* modified for FL(kl5) */
-	/* round 17 */
-	sub[22] ^= kw4;
-	/* round 15 */
-	sub[20] ^= kw4;
-	/* round 13 */
-	sub[18] ^= kw4;
-	kw4 ^= (u64)((u32)kw4 & ~subR(16)) << 32;
-	dw = (u32)(kw4 >> 32) & subL(16);
-	kw4 ^= ROL1(dw); /* modified for FL(kl3) */
-	/* round 11 */
-	sub[14] ^= kw4;
-	/* round 9 */
-	sub[12] ^= kw4;
-	/* round 7 */
-	sub[10] ^= kw4;
-	kw4 ^= (u64)((u32)kw4 & ~subR(8)) << 32;
-	dw = (u32)(kw4 >> 32) & subL(8);
-	kw4 ^= ROL1(dw); /* modified for FL(kl1) */
-	/* round 5 */
-	sub[6] ^= kw4;
-	/* round 3 */
-	sub[4] ^= kw4;
-	/* round 1 */
-	sub[2] ^= kw4;
-	/* kw1 */
-	sub[0] ^= kw4;
-
 	camellia_setup_tail(subkey, sub, 32);
 }
 

@@ -933,8 +881,92 @@ typedef const u64 const_key_element;
 static void camellia_setup_tail(u32 *subkey, u32 *subL, u32 *subR, int max)
 {
 	u32 dw, tl, tr;
+	u32 kw4l, kw4r;
 	int i;
 
+	/* absorb kw2 to other subkeys */
+	/* round 2 */
+	subL[3] ^= subL[1]; subR[3] ^= subR[1];
+	/* round 4 */
+	subL[5] ^= subL[1]; subR[5] ^= subR[1];
+	/* round 6 */
+	subL[7] ^= subL[1]; subR[7] ^= subR[1];
+	subL[1] ^= subR[1] & ~subR[9];
+	dw = subL[1] & subL[9],
+		subR[1] ^= ROL1(dw); /* modified for FLinv(kl2) */
+	/* round 8 */
+	subL[11] ^= subL[1]; subR[11] ^= subR[1];
+	/* round 10 */
+	subL[13] ^= subL[1]; subR[13] ^= subR[1];
+	/* round 12 */
+	subL[15] ^= subL[1]; subR[15] ^= subR[1];
+	subL[1] ^= subR[1] & ~subR[17];
+	dw = subL[1] & subL[17],
+		subR[1] ^= ROL1(dw); /* modified for FLinv(kl4) */
+	/* round 14 */
+	subL[19] ^= subL[1]; subR[19] ^= subR[1];
+	/* round 16 */
+	subL[21] ^= subL[1]; subR[21] ^= subR[1];
+	/* round 18 */
+	subL[23] ^= subL[1]; subR[23] ^= subR[1];
+	if (max == 24) {
+		/* kw3 */
+		subL[24] ^= subL[1]; subR[24] ^= subR[1];
+
+	/* absorb kw4 to other subkeys */
+		kw4l = subL[25]; kw4r = subR[25];
+	} else {
+		subL[1] ^= subR[1] & ~subR[25];
+		dw = subL[1] & subL[25],
+			subR[1] ^= ROL1(dw); /* modified for FLinv(kl6) */
+		/* round 20 */
+		subL[27] ^= subL[1]; subR[27] ^= subR[1];
+		/* round 22 */
+		subL[29] ^= subL[1]; subR[29] ^= subR[1];
+		/* round 24 */
+		subL[31] ^= subL[1]; subR[31] ^= subR[1];
+		/* kw3 */
+		subL[32] ^= subL[1]; subR[32] ^= subR[1];
+
+	/* absorb kw4 to other subkeys */
+		kw4l = subL[33]; kw4r = subR[33];
+		/* round 23 */
+		subL[30] ^= kw4l; subR[30] ^= kw4r;
+		/* round 21 */
+		subL[28] ^= kw4l; subR[28] ^= kw4r;
+		/* round 19 */
+		subL[26] ^= kw4l; subR[26] ^= kw4r;
+		kw4l ^= kw4r & ~subR[24];
+		dw = kw4l & subL[24],
+			kw4r ^= ROL1(dw); /* modified for FL(kl5) */
+	}
+	/* round 17 */
+	subL[22] ^= kw4l; subR[22] ^= kw4r;
+	/* round 15 */
+	subL[20] ^= kw4l; subR[20] ^= kw4r;
+	/* round 13 */
+	subL[18] ^= kw4l; subR[18] ^= kw4r;
+	kw4l ^= kw4r & ~subR[16];
+	dw = kw4l & subL[16],
+		kw4r ^= ROL1(dw); /* modified for FL(kl3) */
+	/* round 11 */
+	subL[14] ^= kw4l; subR[14] ^= kw4r;
+	/* round 9 */
+	subL[12] ^= kw4l; subR[12] ^= kw4r;
+	/* round 7 */
+	subL[10] ^= kw4l; subR[10] ^= kw4r;
+	kw4l ^= kw4r & ~subR[8];
+	dw = kw4l & subL[8],
+		kw4r ^= ROL1(dw); /* modified for FL(kl1) */
+	/* round 5 */
+	subL[6] ^= kw4l; subR[6] ^= kw4r;
+	/* round 3 */
+	subL[4] ^= kw4l; subR[4] ^= kw4r;
+	/* round 1 */
+	subL[2] ^= kw4l; subR[2] ^= kw4r;
+	/* kw1 */
+	subL[0] ^= kw4l; subR[0] ^= kw4r;
+
 	/* key XOR is end of F-function */
 	SUBKEY_L(0) = subL[0] ^ subL[2];/* kw1 */
 	SUBKEY_R(0) = subR[0] ^ subR[2];

@@ -1049,7 +1081,6 @@ static void camellia_setup128(const unsi
 {
 	u32 kll, klr, krl, krr;
 	u32 il, ir, t0, t1, w0, w1;
-	u32 kw4l, kw4r, dw;
 	u32 subL[26];
 	u32 subR[26];
 

@@ -1149,63 +1180,6 @@ static void camellia_setup128(const unsi
 	subL[24] = kll; subR[24] = klr;
 	subL[25] = krl; subR[25] = krr;
 
-	/* absorb kw2 to other subkeys */
-	/* round 2 */
-	subL[3] ^= subL[1]; subR[3] ^= subR[1];
-	/* round 4 */
-	subL[5] ^= subL[1]; subR[5] ^= subR[1];
-	/* round 6 */
-	subL[7] ^= subL[1]; subR[7] ^= subR[1];
-	subL[1] ^= subR[1] & ~subR[9];
-	dw = subL[1] & subL[9],
-		subR[1] ^= ROL1(dw); /* modified for FLinv(kl2) */
-	/* round 8 */
-	subL[11] ^= subL[1]; subR[11] ^= subR[1];
-	/* round 10 */
-	subL[13] ^= subL[1]; subR[13] ^= subR[1];
-	/* round 12 */
-	subL[15] ^= subL[1]; subR[15] ^= subR[1];
-	subL[1] ^= subR[1] & ~subR[17];
-	dw = subL[1] & subL[17],
-		subR[1] ^= ROL1(dw); /* modified for FLinv(kl4) */
-	/* round 14 */
-	subL[19] ^= subL[1]; subR[19] ^= subR[1];
-	/* round 16 */
-	subL[21] ^= subL[1]; subR[21] ^= subR[1];
-	/* round 18 */
-	subL[23] ^= subL[1]; subR[23] ^= subR[1];
-	/* kw3 */
-	subL[24] ^= subL[1]; subR[24] ^= subR[1];
-
-	/* absorb kw4 to other subkeys */
-	kw4l = subL[25]; kw4r = subR[25];
-	/* round 17 */
-	subL[22] ^= kw4l; subR[22] ^= kw4r;
-	/* round 15 */
-	subL[20] ^= kw4l; subR[20] ^= kw4r;
-	/* round 13 */
-	subL[18] ^= kw4l; subR[18] ^= kw4r;
-	kw4l ^= kw4r & ~subR[16];
-	dw = kw4l & subL[16],
-		kw4r ^= ROL1(dw); /* modified for FL(kl3) */
-	/* round 11 */
-	subL[14] ^= kw4l; subR[14] ^= kw4r;
-	/* round 9 */
-	subL[12] ^= kw4l; subR[12] ^= kw4r;
-	/* round 7 */
-	subL[10] ^= kw4l; subR[10] ^= kw4r;
-	kw4l ^= kw4r & ~subR[8];
-	dw = kw4l & subL[8],
-		kw4r ^= ROL1(dw); /* modified for FL(kl1) */
-	/* round 5 */
-	subL[6] ^= kw4l; subR[6] ^= kw4r;
-	/* round 3 */
-	subL[4] ^= kw4l; subR[4] ^= kw4r;
-	/* round 1 */
-	subL[2] ^= kw4l; subR[2] ^= kw4r;
-	/* kw1 */
-	subL[0] ^= kw4l; subR[0] ^= kw4r;
-
 	camellia_setup_tail(subkey, subL, subR, 24);
 }
 

@@ -1214,7 +1188,6 @@ static void camellia_setup256(const unsi
 	u32 kll, klr, krl, krr;        /* left half of key */
 	u32 krll, krlr, krrl, krrr;    /* right half of key */
 	u32 il, ir, t0, t1, w0, w1;    /* temporary variables */
-	u32 kw4l, kw4r, dw;
 	u32 subL[34];
 	u32 subR[34];
 

@@ -1356,81 +1329,6 @@ static void camellia_setup256(const unsi
 	/* kw4 */
 	subL[33] = krrl; subR[33] = krrr;
 
-	/* absorb kw2 to other subkeys */
-	/* round 2 */
-	subL[3] ^= subL[1]; subR[3] ^= subR[1];
-	/* round 4 */
-	subL[5] ^= subL[1]; subR[5] ^= subR[1];
-	/* round 6 */
-	subL[7] ^= subL[1]; subR[7] ^= subR[1];
-	subL[1] ^= subR[1] & ~subR[9];
-	dw = subL[1] & subL[9],
-		subR[1] ^= ROL1(dw); /* modified for FLinv(kl2) */
-	/* round 8 */
-	subL[11] ^= subL[1]; subR[11] ^= subR[1];
-	/* round 10 */
-	subL[13] ^= subL[1]; subR[13] ^= subR[1];
-	/* round 12 */
-	subL[15] ^= subL[1]; subR[15] ^= subR[1];
-	subL[1] ^= subR[1] & ~subR[17];
-	dw = subL[1] & subL[17],
-		subR[1] ^= ROL1(dw); /* modified for FLinv(kl4) */
-	/* round 14 */
-	subL[19] ^= subL[1]; subR[19] ^= subR[1];
-	/* round 16 */
-	subL[21] ^= subL[1]; subR[21] ^= subR[1];
-	/* round 18 */
-	subL[23] ^= subL[1]; subR[23] ^= subR[1];
-	subL[1] ^= subR[1] & ~subR[25];
-	dw = subL[1] & subL[25],
-		subR[1] ^= ROL1(dw); /* modified for FLinv(kl6) */
-	/* round 20 */
-	subL[27] ^= subL[1]; subR[27] ^= subR[1];
-	/* round 22 */
-	subL[29] ^= subL[1]; subR[29] ^= subR[1];
-	/* round 24 */
-	subL[31] ^= subL[1]; subR[31] ^= subR[1];
-	/* kw3 */
-	subL[32] ^= subL[1]; subR[32] ^= subR[1];
-
-	/* absorb kw4 to other subkeys */
-	kw4l = subL[33]; kw4r = subR[33];
-	/* round 23 */
-	subL[30] ^= kw4l; subR[30] ^= kw4r;
-	/* round 21 */
-	subL[28] ^= kw4l; subR[28] ^= kw4r;
-	/* round 19 */
-	subL[26] ^= kw4l; subR[26] ^= kw4r;
-	kw4l ^= kw4r & ~subR[24];
-	dw = kw4l & subL[24],
-		kw4r ^= ROL1(dw); /* modified for FL(kl5) */
-	/* round 17 */
-	subL[22] ^= kw4l; subR[22] ^= kw4r;
-	/* round 15 */
-	subL[20] ^= kw4l; subR[20] ^= kw4r;
-	/* round 13 */
-	subL[18] ^= kw4l; subR[18] ^= kw4r;
-	kw4l ^= kw4r & ~subR[16];
-	dw = kw4l & subL[16],
-		kw4r ^= ROL1(dw); /* modified for FL(kl3) */
-	/* round 11 */
-	subL[14] ^= kw4l; subR[14] ^= kw4r;
-	/* round 9 */
-	subL[12] ^= kw4l; subR[12] ^= kw4r;
-	/* round 7 */
-	subL[10] ^= kw4l; subR[10] ^= kw4r;
-	kw4l ^= kw4r & ~subR[8];
-	dw = kw4l & subL[8],
-		kw4r ^= ROL1(dw); /* modified for FL(kl1) */
-	/* round 5 */
-	subL[6] ^= kw4l; subR[6] ^= kw4r;
-	/* round 3 */
-	subL[4] ^= kw4l; subR[4] ^= kw4r;
-	/* round 1 */
-	subL[2] ^= kw4l; subR[2] ^= kw4r;
-	/* kw1 */
-	subL[0] ^= kw4l; subR[0] ^= kw4r;
-
 	camellia_setup_tail(subkey, subL, subR, 32);
 }