#include <linux/init.h>
#include <linux/kernel.h>
#include <linux/module.h>
+#include <linux/bitops.h>
+#include <asm/unaligned.h>
static const u32 camellia_sp1110[256] = {
0x70707000,0x82828200,0x2c2c2c00,0xececec00,
/*
* macros
*/
-#define GETU32(v, pt) \
- do { \
- /* latest breed of gcc is clever enough to use move */ \
- memcpy(&(v), (pt), 4); \
- (v) = be32_to_cpu(v); \
- } while(0)
-
-/* rotation right shift 1byte */
-#define ROR8(x) (((x) >> 8) + ((x) << 24))
-/* rotation left shift 1bit */
-#define ROL1(x) (((x) << 1) + ((x) >> 31))
-/* rotation left shift 1byte */
-#define ROL8(x) (((x) << 8) + ((x) >> 24))
-
#define ROLDQ(ll, lr, rl, rr, w0, w1, bits) \
do { \
w0 = ll; \
^ camellia_sp3033[(u8)(il >> 8)] \
^ camellia_sp4404[(u8)(il )]; \
yl ^= yr; \
- yr = ROR8(yr); \
+ yr = ror32(yr, 8); \
yr ^= yl; \
} while(0)
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] ^= rol32(dw, 1); /* 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] ^= rol32(dw, 1); /* 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] ^= rol32(dw, 1); /* 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 ^= rol32(dw, 1); /* 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 ^= rol32(dw, 1); /* 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 ^= rol32(dw, 1); /* 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];
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);
+ tr = subR[10] ^ rol32(dw, 1);
SUBKEY_L(7) = subL[6] ^ tl; /* round 6 */
SUBKEY_R(7) = subR[6] ^ tr;
SUBKEY_L(8) = subL[8]; /* FL(kl1) */
SUBKEY_R(9) = subR[9];
tl = subL[7] ^ (subR[7] & ~subR[9]);
dw = tl & subL[9], /* FLinv(kl2) */
- tr = subR[7] ^ ROL1(dw);
+ tr = subR[7] ^ rol32(dw, 1);
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(14) = subR[13] ^ subR[15];
tl = subL[18] ^ (subR[18] & ~subR[16]);
dw = tl & subL[16], /* FL(kl3) */
- tr = subR[18] ^ ROL1(dw);
+ tr = subR[18] ^ rol32(dw, 1);
SUBKEY_L(15) = subL[14] ^ tl; /* round 12 */
SUBKEY_R(15) = subR[14] ^ tr;
SUBKEY_L(16) = subL[16]; /* FL(kl3) */
SUBKEY_R(17) = subR[17];
tl = subL[15] ^ (subR[15] & ~subR[17]);
dw = tl & subL[17], /* FLinv(kl4) */
- tr = subR[15] ^ ROL1(dw);
+ tr = subR[15] ^ rol32(dw, 1);
SUBKEY_L(18) = tl ^ subL[19]; /* round 13 */
SUBKEY_R(18) = tr ^ subR[19];
SUBKEY_L(19) = subL[18] ^ subL[20]; /* round 14 */
} else {
tl = subL[26] ^ (subR[26] & ~subR[24]);
dw = tl & subL[24], /* FL(kl5) */
- tr = subR[26] ^ ROL1(dw);
+ tr = subR[26] ^ rol32(dw, 1);
SUBKEY_L(23) = subL[22] ^ tl; /* round 18 */
SUBKEY_R(23) = subR[22] ^ tr;
SUBKEY_L(24) = subL[24]; /* FL(kl5) */
SUBKEY_R(25) = subR[25];
tl = subL[23] ^ (subR[23] & ~subR[25]);
dw = tl & subL[25], /* FLinv(kl6) */
- tr = subR[23] ^ ROL1(dw);
+ tr = subR[23] ^ rol32(dw, 1);
SUBKEY_L(26) = tl ^ subL[27]; /* round 19 */
SUBKEY_R(26) = tr ^ subR[27];
SUBKEY_L(27) = subL[26] ^ subL[28]; /* round 20 */
/* 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 */
+ dw = SUBKEY_L(i + 0) ^ SUBKEY_R(i + 0); dw = rol32(dw, 8);/* 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 */
+ dw = SUBKEY_L(i + 1) ^ SUBKEY_R(i + 1); dw = rol32(dw, 8);/* 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 */
+ dw = SUBKEY_L(i + 2) ^ SUBKEY_R(i + 2); dw = rol32(dw, 8);/* 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 */
+ dw = SUBKEY_L(i + 3) ^ SUBKEY_R(i + 3); dw = rol32(dw, 8);/* 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 */
+ dw = SUBKEY_L(i + 4) ^ SUBKEY_R(i + 4); dw = rol32(dw, 9);/* 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 */
+ dw = SUBKEY_L(i + 5) ^ SUBKEY_R(i + 5); dw = rol32(dw, 8);/* round 6 */
SUBKEY_R(i + 5) = SUBKEY_L(i + 5) ^ dw; SUBKEY_L(i + 5) = dw;
i += 8;
} while (i < max);
{
u32 kll, klr, krl, krr;
u32 il, ir, t0, t1, w0, w1;
- u32 kw4l, kw4r, dw;
u32 subL[26];
u32 subR[26];
/**
* k == kll || klr || krl || krr (|| is concatenation)
*/
- GETU32(kll, key );
- GETU32(klr, key + 4);
- GETU32(krl, key + 8);
- GETU32(krr, key + 12);
+ kll = get_unaligned_be32(key);
+ klr = get_unaligned_be32(key + 4);
+ krl = get_unaligned_be32(key + 8);
+ krr = get_unaligned_be32(key + 12);
/* generate KL dependent subkeys */
/* kw1 */
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);
}
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];
* key = (kll || klr || krl || krr || krll || krlr || krrl || krrr)
* (|| is concatenation)
*/
- GETU32(kll, key );
- GETU32(klr, key + 4);
- GETU32(krl, key + 8);
- GETU32(krr, key + 12);
- GETU32(krll, key + 16);
- GETU32(krlr, key + 20);
- GETU32(krrl, key + 24);
- GETU32(krrr, key + 28);
+ kll = get_unaligned_be32(key);
+ klr = get_unaligned_be32(key + 4);
+ krl = get_unaligned_be32(key + 8);
+ krr = get_unaligned_be32(key + 12);
+ krll = get_unaligned_be32(key + 16);
+ krlr = get_unaligned_be32(key + 20);
+ krrl = get_unaligned_be32(key + 24);
+ krrr = get_unaligned_be32(key + 28);
/* generate KL dependent subkeys */
/* kw1 */
/* 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);
}
t0 &= ll; \
t2 |= rr; \
rl ^= t2; \
- lr ^= ROL1(t0); \
+ lr ^= rol32(t0, 1); \
t3 = krl; \
t1 = klr; \
t3 &= rl; \
t1 |= lr; \
ll ^= t1; \
- rr ^= ROL1(t3); \
+ rr ^= rol32(t3, 1); \
} while(0)
#define CAMELLIA_ROUNDSM(xl, xr, kl, kr, yl, yr, il, ir) \
il ^= kl; \
ir ^= il ^ kr; \
yl ^= ir; \
- yr ^= ROR8(il) ^ ir; \
+ yr ^= ror32(il, 8) ^ ir; \
} while(0)
/* max = 24: 128bit encrypt, max = 32: 256bit encrypt */