static void Transform(Sha* sha)
{
word32 W[SHA_BLOCK_SIZE / sizeof(word32)];
/* Copy context->state[] to working vars */
word32 a = sha->digest[0];
word32 b = sha->digest[1];
word32 c = sha->digest[2];
word32 d = sha->digest[3];
word32 e = sha->digest[4];
#ifdef USE_SLOW_SHA
word32 t, i;
for (i = 0; i < 16; i++) {
R0(a, b, c, d, e, i);
t = e; e = d; d = c; c = b; b = a; a = t;
}
for (; i < 20; i++) {
R1(a, b, c, d, e, i);
t = e; e = d; d = c; c = b; b = a; a = t;
}
for (; i < 40; i++) {
R2(a, b, c, d, e, i);
t = e; e = d; d = c; c = b; b = a; a = t;
}
for (; i < 60; i++) {
R3(a, b, c, d, e, i);
t = e; e = d; d = c; c = b; b = a; a = t;
}
for (; i < 80; i++) {
R4(a, b, c, d, e, i);
t = e; e = d; d = c; c = b; b = a; a = t;
}
#else
/* nearly 1 K bigger in code size but 25% faster */
/* 4 rounds of 20 operations each. Loop unrolled. */
R0(a,b,c,d,e, 0); R0(e,a,b,c,d, 1); R0(d,e,a,b,c, 2); R0(c,d,e,a,b, 3);
R0(b,c,d,e,a, 4); R0(a,b,c,d,e, 5); R0(e,a,b,c,d, 6); R0(d,e,a,b,c, 7);
R0(c,d,e,a,b, 8); R0(b,c,d,e,a, 9); R0(a,b,c,d,e,10); R0(e,a,b,c,d,11);
R0(d,e,a,b,c,12); R0(c,d,e,a,b,13); R0(b,c,d,e,a,14); R0(a,b,c,d,e,15);
R1(e,a,b,c,d,16); R1(d,e,a,b,c,17); R1(c,d,e,a,b,18); R1(b,c,d,e,a,19);
R2(a,b,c,d,e,20); R2(e,a,b,c,d,21); R2(d,e,a,b,c,22); R2(c,d,e,a,b,23);
R2(b,c,d,e,a,24); R2(a,b,c,d,e,25); R2(e,a,b,c,d,26); R2(d,e,a,b,c,27);
R2(c,d,e,a,b,28); R2(b,c,d,e,a,29); R2(a,b,c,d,e,30); R2(e,a,b,c,d,31);
R2(d,e,a,b,c,32); R2(c,d,e,a,b,33); R2(b,c,d,e,a,34); R2(a,b,c,d,e,35);
R2(e,a,b,c,d,36); R2(d,e,a,b,c,37); R2(c,d,e,a,b,38); R2(b,c,d,e,a,39);
R3(a,b,c,d,e,40); R3(e,a,b,c,d,41); R3(d,e,a,b,c,42); R3(c,d,e,a,b,43);
R3(b,c,d,e,a,44); R3(a,b,c,d,e,45); R3(e,a,b,c,d,46); R3(d,e,a,b,c,47);
R3(c,d,e,a,b,48); R3(b,c,d,e,a,49); R3(a,b,c,d,e,50); R3(e,a,b,c,d,51);
R3(d,e,a,b,c,52); R3(c,d,e,a,b,53); R3(b,c,d,e,a,54); R3(a,b,c,d,e,55);
R3(e,a,b,c,d,56); R3(d,e,a,b,c,57); R3(c,d,e,a,b,58); R3(b,c,d,e,a,59);
R4(a,b,c,d,e,60); R4(e,a,b,c,d,61); R4(d,e,a,b,c,62); R4(c,d,e,a,b,63);
R4(b,c,d,e,a,64); R4(a,b,c,d,e,65); R4(e,a,b,c,d,66); R4(d,e,a,b,c,67);
R4(c,d,e,a,b,68); R4(b,c,d,e,a,69); R4(a,b,c,d,e,70); R4(e,a,b,c,d,71);
R4(d,e,a,b,c,72); R4(c,d,e,a,b,73); R4(b,c,d,e,a,74); R4(a,b,c,d,e,75);
R4(e,a,b,c,d,76); R4(d,e,a,b,c,77); R4(c,d,e,a,b,78); R4(b,c,d,e,a,79);
#endif
/* Add the working vars back into digest state[] */
sha->digest[0] += a;
sha->digest[1] += b;
sha->digest[2] += c;
sha->digest[3] += d;
sha->digest[4] += e;
}
void
EulerAngleUpdater::initialize()
{
const auto grain_num = _grain_tracker.getTotalFeatureCount();
if (_first_time)
{
_angles.resize(grain_num);
_angles_old.resize(grain_num);
for (unsigned int i = 0; i < grain_num; ++i)
_angles[i] = _euler.getEulerAngles(i); // Read initial euler angles
}
unsigned int angle_size = _angles.size();
for (unsigned int i = angle_size; i < grain_num; ++i) // if new grains are created
_angles.push_back(_euler.getEulerAngles(i)); // Assign initial euler angles
for (unsigned int i = 0; i < grain_num; ++i)
{
if (!_first_time && !_fe_problem.converged())
_angles[i] = _angles_old[i];
RealGradient torque = _grain_torque.getTorqueValues()[i];
if (i <= angle_size) // if new grains are created
_angles_old[i] = _angles[i];
else
_angles_old.push_back(_angles[i]);
RotationTensor R0(_angles_old[i]); // RotationTensor as per old euler angles
RealVectorValue torque_rotated = R0 * torque; // Applied torque is rotated to allign with old grain axes
RealVectorValue omega = _mr / _grain_volumes[i] * torque_rotated; // Angular velocity as per old grain axes
/**
* Change in euler angles are obtained from the torque & angular velocities about the material axes.
* Change in phi1, Phi and phi2 are caused by rotation about z axis, x' axis & z'' axis, respectively.
* Components of the angular velocities across z, x' and z'' axes are obtained from the torque values.
* This yields change in euler angles due to grain rotation.
*/
RealVectorValue angle_change;
angle_change(0) = omega(2) * _dt;
angle_change(1) = (omega(0) * std::cos(angle_change(0)) + omega(1) * std::sin(angle_change(0))) * _dt;
angle_change(2) = (omega(0) * std::sin(angle_change(0)) * std::sin(angle_change(1))
- omega(1) * std::cos(angle_change(0)) * std::sin(angle_change(1))
+ omega(2) * std::cos(angle_change(1))) * _dt;
angle_change *= (180.0 / libMesh::pi);
RotationTensor R1(angle_change); // Rotation matrix due to torque
/**
* Final RotationMatrix = RotationMatrix due to applied torque X old RotationMatrix
* Updated Euler angles are obtained by back-tracking the angles from the rotation matrix
* For details about the componenets of the rotation matrix please refer to RotationTensor.C
* Phi = acos(R33); phi1 = atan2(R31,-R32); phi2 = atan2(R13,R23) for phi != 0.0 por 180.0
*/
RealTensorValue R = R1 * R0;
if (R(2,2) != 1.0 && R(2,2) != -1.0) // checks if cos(Phi) = 1 or -1
{
_angles[i].phi1 = std::atan2(R(2,0), -R(2,1)) * (180.0 / libMesh::pi) ;
_angles[i].Phi = std::acos(R(2,2)) * (180.0 / libMesh::pi);
_angles[i].phi2 = std::atan2(R(0,2), R(1,2)) * (180.0 / libMesh::pi);
}
else if (R(2,2) == 1.0) // special case for Phi = 0.0
{
if (R0(2,2) == 1.0)
// when Phi_old = 0.0; all the rotations are about z axis and angles accumulates after each rotation
_angles[i].phi1 = _angles_old[i].phi1 + _angles_old[i].phi2 + angle_change(0);
else
_angles[i].phi1 = angle_change(0);
// Comply with bunge euler angle definitions, 0.0 <= phi1 <= 360.0
if (std::abs(_angles[i].phi1) > 360.0)
{
int laps = _angles[i].phi1 / 360.0;
_angles[i].phi1 -= laps * 360.0;
}
_angles[i].Phi = 0.0;
_angles[i].phi2 = -_angles[i].phi1 + std::atan2(R(0,1), R(1,1)) * (180.0 / libMesh::pi);
}
else
{
if (R0(2,2) == 1.0)
_angles[i].phi1 = _angles_old[i].phi1 + _angles_old[i].phi2 + angle_change(0);
else
_angles[i].phi1 = angle_change(0);
// Comply with bunge euler angle definitions, 0.0 <= phi1 <= 360.0
if (std::abs(_angles[i].phi1) > 360.0)
{
int laps = _angles[i].phi1 / 360.0;
_angles[i].phi1 -= laps * 360.0;
}
_angles[i].Phi = 180.0;
_angles[i].phi2 = _angles[i].phi1 - std::atan2(-R(0,1), -R(1,1)) * (180.0 / libMesh::pi);
}
// Following checks and updates are done only to comply with bunge euler angle definitions, 0.0 <= phi1/phi2 <= 360.0
if (_angles[i].phi1 < 0.0)
_angles[i].phi1 += 360.0;
if (_angles[i].phi2 < 0.0)
_angles[i].phi2 += 360.0;
if (_angles[i].Phi < 0.0)
mooseError("Euler angle out of range.");
}
_first_time = false;
}
void md5_block_data_order (MD5_CTX * c, const void *data_, size_t num)
{
const unsigned char *data = data_;
register unsigned MD32_REG_T A, B, C, D, l;
#ifndef MD32_XARRAY
/* See comment in crypto/sha/sha_locl.h for details. */
unsigned MD32_REG_T XX0, XX1, XX2, XX3, XX4, XX5, XX6, XX7, XX8, XX9, XX10, XX11, XX12, XX13, XX14, XX15;
# define X(i) XX##i
#else
MD5_LONG XX[MD5_LBLOCK];
# define X(i) XX[i]
#endif
A = c->A;
B = c->B;
C = c->C;
D = c->D;
for (; num--;)
{
HOST_c2l (data, l);
X (0) = l;
HOST_c2l (data, l);
X (1) = l;
/* Round 0 */
R0 (A, B, C, D, X (0), 7, 0xd76aa478L);
HOST_c2l (data, l);
X (2) = l;
R0 (D, A, B, C, X (1), 12, 0xe8c7b756L);
HOST_c2l (data, l);
X (3) = l;
R0 (C, D, A, B, X (2), 17, 0x242070dbL);
HOST_c2l (data, l);
X (4) = l;
R0 (B, C, D, A, X (3), 22, 0xc1bdceeeL);
HOST_c2l (data, l);
X (5) = l;
R0 (A, B, C, D, X (4), 7, 0xf57c0fafL);
HOST_c2l (data, l);
X (6) = l;
R0 (D, A, B, C, X (5), 12, 0x4787c62aL);
HOST_c2l (data, l);
X (7) = l;
R0 (C, D, A, B, X (6), 17, 0xa8304613L);
HOST_c2l (data, l);
X (8) = l;
R0 (B, C, D, A, X (7), 22, 0xfd469501L);
HOST_c2l (data, l);
X (9) = l;
R0 (A, B, C, D, X (8), 7, 0x698098d8L);
HOST_c2l (data, l);
X (10) = l;
R0 (D, A, B, C, X (9), 12, 0x8b44f7afL);
HOST_c2l (data, l);
X (11) = l;
R0 (C, D, A, B, X (10), 17, 0xffff5bb1L);
HOST_c2l (data, l);
X (12) = l;
R0 (B, C, D, A, X (11), 22, 0x895cd7beL);
HOST_c2l (data, l);
X (13) = l;
R0 (A, B, C, D, X (12), 7, 0x6b901122L);
HOST_c2l (data, l);
X (14) = l;
R0 (D, A, B, C, X (13), 12, 0xfd987193L);
HOST_c2l (data, l);
X (15) = l;
R0 (C, D, A, B, X (14), 17, 0xa679438eL);
R0 (B, C, D, A, X (15), 22, 0x49b40821L);
/* Round 1 */
R1 (A, B, C, D, X (1), 5, 0xf61e2562L);
R1 (D, A, B, C, X (6), 9, 0xc040b340L);
R1 (C, D, A, B, X (11), 14, 0x265e5a51L);
R1 (B, C, D, A, X (0), 20, 0xe9b6c7aaL);
R1 (A, B, C, D, X (5), 5, 0xd62f105dL);
R1 (D, A, B, C, X (10), 9, 0x02441453L);
R1 (C, D, A, B, X (15), 14, 0xd8a1e681L);
R1 (B, C, D, A, X (4), 20, 0xe7d3fbc8L);
R1 (A, B, C, D, X (9), 5, 0x21e1cde6L);
R1 (D, A, B, C, X (14), 9, 0xc33707d6L);
R1 (C, D, A, B, X (3), 14, 0xf4d50d87L);
R1 (B, C, D, A, X (8), 20, 0x455a14edL);
R1 (A, B, C, D, X (13), 5, 0xa9e3e905L);
R1 (D, A, B, C, X (2), 9, 0xfcefa3f8L);
R1 (C, D, A, B, X (7), 14, 0x676f02d9L);
R1 (B, C, D, A, X (12), 20, 0x8d2a4c8aL);
/* Round 2 */
R2 (A, B, C, D, X (5), 4, 0xfffa3942L);
R2 (D, A, B, C, X (8), 11, 0x8771f681L);
R2 (C, D, A, B, X (11), 16, 0x6d9d6122L);
R2 (B, C, D, A, X (14), 23, 0xfde5380cL);
R2 (A, B, C, D, X (1), 4, 0xa4beea44L);
R2 (D, A, B, C, X (4), 11, 0x4bdecfa9L);
R2 (C, D, A, B, X (7), 16, 0xf6bb4b60L);
R2 (B, C, D, A, X (10), 23, 0xbebfbc70L);
R2 (A, B, C, D, X (13), 4, 0x289b7ec6L);
R2 (D, A, B, C, X (0), 11, 0xeaa127faL);
R2 (C, D, A, B, X (3), 16, 0xd4ef3085L);
R2 (B, C, D, A, X (6), 23, 0x04881d05L);
R2 (A, B, C, D, X (9), 4, 0xd9d4d039L);
R2 (D, A, B, C, X (12), 11, 0xe6db99e5L);
R2 (C, D, A, B, X (15), 16, 0x1fa27cf8L);
R2 (B, C, D, A, X (2), 23, 0xc4ac5665L);
/* Round 3 */
R3 (A, B, C, D, X (0), 6, 0xf4292244L);
R3 (D, A, B, C, X (7), 10, 0x432aff97L);
R3 (C, D, A, B, X (14), 15, 0xab9423a7L);
R3 (B, C, D, A, X (5), 21, 0xfc93a039L);
R3 (A, B, C, D, X (12), 6, 0x655b59c3L);
R3 (D, A, B, C, X (3), 10, 0x8f0ccc92L);
R3 (C, D, A, B, X (10), 15, 0xffeff47dL);
R3 (B, C, D, A, X (1), 21, 0x85845dd1L);
R3 (A, B, C, D, X (8), 6, 0x6fa87e4fL);
R3 (D, A, B, C, X (15), 10, 0xfe2ce6e0L);
R3 (C, D, A, B, X (6), 15, 0xa3014314L);
R3 (B, C, D, A, X (13), 21, 0x4e0811a1L);
R3 (A, B, C, D, X (4), 6, 0xf7537e82L);
R3 (D, A, B, C, X (11), 10, 0xbd3af235L);
R3 (C, D, A, B, X (2), 15, 0x2ad7d2bbL);
R3 (B, C, D, A, X (9), 21, 0xeb86d391L);
A = c->A += A;
B = c->B += B;
C = c->C += C;
D = c->D += D;
}
}
void md5_block_host_order (MD5_CTX *c, const void *data, int num)
{
const MD5_LONG *X=data;
register unsigned long A,B,C,D;
/*
* In case you wonder why A-D are declared as long and not
* as MD5_LONG. Doing so results in slight performance
* boost on LP64 architectures. The catch is we don't
* really care if 32 MSBs of a 64-bit register get polluted
* with eventual overflows as we *save* only 32 LSBs in
* *either* case. Now declaring 'em long excuses the compiler
* from keeping 32 MSBs zeroed resulting in 13% performance
* improvement under SPARC Solaris7/64 and 5% under AlphaLinux.
* Well, to be honest it should say that this *prevents*
* performance degradation.
*
* <*****@*****.**>
*/
A=c->A;
B=c->B;
C=c->C;
D=c->D;
for (;num--;X+=HASH_LBLOCK)
{
/* Round 0 */
R0(A,B,C,D,X[ 0], 7,0xd76aa478L);
R0(D,A,B,C,X[ 1],12,0xe8c7b756L);
R0(C,D,A,B,X[ 2],17,0x242070dbL);
R0(B,C,D,A,X[ 3],22,0xc1bdceeeL);
R0(A,B,C,D,X[ 4], 7,0xf57c0fafL);
R0(D,A,B,C,X[ 5],12,0x4787c62aL);
R0(C,D,A,B,X[ 6],17,0xa8304613L);
R0(B,C,D,A,X[ 7],22,0xfd469501L);
R0(A,B,C,D,X[ 8], 7,0x698098d8L);
R0(D,A,B,C,X[ 9],12,0x8b44f7afL);
R0(C,D,A,B,X[10],17,0xffff5bb1L);
R0(B,C,D,A,X[11],22,0x895cd7beL);
R0(A,B,C,D,X[12], 7,0x6b901122L);
R0(D,A,B,C,X[13],12,0xfd987193L);
R0(C,D,A,B,X[14],17,0xa679438eL);
R0(B,C,D,A,X[15],22,0x49b40821L);
/* Round 1 */
R1(A,B,C,D,X[ 1], 5,0xf61e2562L);
R1(D,A,B,C,X[ 6], 9,0xc040b340L);
R1(C,D,A,B,X[11],14,0x265e5a51L);
R1(B,C,D,A,X[ 0],20,0xe9b6c7aaL);
R1(A,B,C,D,X[ 5], 5,0xd62f105dL);
R1(D,A,B,C,X[10], 9,0x02441453L);
R1(C,D,A,B,X[15],14,0xd8a1e681L);
R1(B,C,D,A,X[ 4],20,0xe7d3fbc8L);
R1(A,B,C,D,X[ 9], 5,0x21e1cde6L);
R1(D,A,B,C,X[14], 9,0xc33707d6L);
R1(C,D,A,B,X[ 3],14,0xf4d50d87L);
R1(B,C,D,A,X[ 8],20,0x455a14edL);
R1(A,B,C,D,X[13], 5,0xa9e3e905L);
R1(D,A,B,C,X[ 2], 9,0xfcefa3f8L);
R1(C,D,A,B,X[ 7],14,0x676f02d9L);
R1(B,C,D,A,X[12],20,0x8d2a4c8aL);
/* Round 2 */
R2(A,B,C,D,X[ 5], 4,0xfffa3942L);
R2(D,A,B,C,X[ 8],11,0x8771f681L);
R2(C,D,A,B,X[11],16,0x6d9d6122L);
R2(B,C,D,A,X[14],23,0xfde5380cL);
R2(A,B,C,D,X[ 1], 4,0xa4beea44L);
R2(D,A,B,C,X[ 4],11,0x4bdecfa9L);
R2(C,D,A,B,X[ 7],16,0xf6bb4b60L);
R2(B,C,D,A,X[10],23,0xbebfbc70L);
R2(A,B,C,D,X[13], 4,0x289b7ec6L);
R2(D,A,B,C,X[ 0],11,0xeaa127faL);
R2(C,D,A,B,X[ 3],16,0xd4ef3085L);
R2(B,C,D,A,X[ 6],23,0x04881d05L);
R2(A,B,C,D,X[ 9], 4,0xd9d4d039L);
R2(D,A,B,C,X[12],11,0xe6db99e5L);
R2(C,D,A,B,X[15],16,0x1fa27cf8L);
R2(B,C,D,A,X[ 2],23,0xc4ac5665L);
/* Round 3 */
R3(A,B,C,D,X[ 0], 6,0xf4292244L);
R3(D,A,B,C,X[ 7],10,0x432aff97L);
R3(C,D,A,B,X[14],15,0xab9423a7L);
R3(B,C,D,A,X[ 5],21,0xfc93a039L);
R3(A,B,C,D,X[12], 6,0x655b59c3L);
R3(D,A,B,C,X[ 3],10,0x8f0ccc92L);
R3(C,D,A,B,X[10],15,0xffeff47dL);
R3(B,C,D,A,X[ 1],21,0x85845dd1L);
R3(A,B,C,D,X[ 8], 6,0x6fa87e4fL);
R3(D,A,B,C,X[15],10,0xfe2ce6e0L);
R3(C,D,A,B,X[ 6],15,0xa3014314L);
R3(B,C,D,A,X[13],21,0x4e0811a1L);
R3(A,B,C,D,X[ 4], 6,0xf7537e82L);
R3(D,A,B,C,X[11],10,0xbd3af235L);
R3(C,D,A,B,X[ 2],15,0x2ad7d2bbL);
R3(B,C,D,A,X[ 9],21,0xeb86d391L);
A = c->A += A;
B = c->B += B;
C = c->C += C;
D = c->D += D;
}
}
/* Process LEN bytes of BUFFER, accumulating context into CTX.
It is assumed that LEN % 128 == 0. */
static void
sha512_process_block (const void *buffer, size_t len, struct sha512_ctx *ctx)
{
const uint64_t *words = buffer;
size_t nwords = len / sizeof (uint64_t);
uint64_t a = ctx->H[0];
uint64_t b = ctx->H[1];
uint64_t c = ctx->H[2];
uint64_t d = ctx->H[3];
uint64_t e = ctx->H[4];
uint64_t f = ctx->H[5];
uint64_t g = ctx->H[6];
uint64_t h = ctx->H[7];
/* First increment the byte count. FIPS 180-2 specifies the possible
length of the file up to 2^128 bits. Here we only compute the
number of bytes. Do a double word increment. */
#ifdef USE_TOTAL128
ctx->total128 += len;
#else
ctx->total[TOTAL128_low] += len;
if (ctx->total[TOTAL128_low] < len)
++ctx->total[TOTAL128_high];
#endif
/* Process all bytes in the buffer with 128 bytes in each round of
the loop. */
while (nwords > 0)
{
uint64_t W[80];
uint64_t a_save = a;
uint64_t b_save = b;
uint64_t c_save = c;
uint64_t d_save = d;
uint64_t e_save = e;
uint64_t f_save = f;
uint64_t g_save = g;
uint64_t h_save = h;
/* Operators defined in FIPS 180-2:4.1.2. */
#define Ch(x, y, z) ((x & y) ^ (~x & z))
#define Maj(x, y, z) ((x & y) ^ (x & z) ^ (y & z))
#define S0(x) (CYCLIC (x, 28) ^ CYCLIC (x, 34) ^ CYCLIC (x, 39))
#define S1(x) (CYCLIC (x, 14) ^ CYCLIC (x, 18) ^ CYCLIC (x, 41))
#define R0(x) (CYCLIC (x, 1) ^ CYCLIC (x, 8) ^ (x >> 7))
#define R1(x) (CYCLIC (x, 19) ^ CYCLIC (x, 61) ^ (x >> 6))
/* It is unfortunate that C does not provide an operator for
cyclic rotation. Hope the C compiler is smart enough. */
#define CYCLIC(w, s) ((w >> s) | (w << (64 - s)))
/* Compute the message schedule according to FIPS 180-2:6.3.2 step 2. */
for (unsigned int t = 0; t < 16; ++t)
{
W[t] = SWAP (*words);
++words;
}
for (unsigned int t = 16; t < 80; ++t)
W[t] = R1 (W[t - 2]) + W[t - 7] + R0 (W[t - 15]) + W[t - 16];
/* The actual computation according to FIPS 180-2:6.3.2 step 3. */
for (unsigned int t = 0; t < 80; ++t)
{
uint64_t T1 = h + S1 (e) + Ch (e, f, g) + K[t] + W[t];
uint64_t T2 = S0 (a) + Maj (a, b, c);
h = g;
g = f;
f = e;
e = d + T1;
d = c;
c = b;
b = a;
a = T1 + T2;
}
/* Add the starting values of the context according to FIPS 180-2:6.3.2
step 4. */
a += a_save;
b += b_save;
c += c_save;
d += d_save;
e += e_save;
f += f_save;
g += g_save;
h += h_save;
/* Prepare for the next round. */
nwords -= 16;
}
/* Put checksum in context given as argument. */
ctx->H[0] = a;
ctx->H[1] = b;
ctx->H[2] = c;
ctx->H[3] = d;
ctx->H[4] = e;
ctx->H[5] = f;
ctx->H[6] = g;
ctx->H[7] = h;
}
typename IntrEllipse2Ellipse2<Real>::Classification
IntrEllipse2Ellipse2<Real>::GetClassification () const
{
// Get the parameters of ellipe0.
Vector2<Real> K0 = mEllipse0->Center;
Matrix2<Real> R0(mEllipse0->Axis, true);
Matrix2<Real> D0(
((Real)1)/(mEllipse0->Extent[0]*mEllipse0->Extent[0]),
((Real)1)/(mEllipse0->Extent[1]*mEllipse0->Extent[1]));
// Get the parameters of ellipse1.
Vector2<Real> K1 = mEllipse1->Center;
Matrix2<Real> R1(mEllipse1->Axis, true);
Matrix2<Real> D1(
((Real)1)/(mEllipse1->Extent[0]*mEllipse1->Extent[0]),
((Real)1)/(mEllipse1->Extent[1]*mEllipse1->Extent[1]));
// Compute K2.
Matrix2<Real> D0NegHalf(
mEllipse0->Extent[0],
mEllipse0->Extent[1]);
Matrix2<Real> D0Half(
((Real)1)/mEllipse0->Extent[0],
((Real)1)/mEllipse0->Extent[1]);
Vector2<Real> K2 = D0Half*((K1 - K0)*R0);
// Compute M2.
Matrix2<Real> R1TR0D0NegHalf = R1.TransposeTimes(R0*D0NegHalf);
Matrix2<Real> M2 = R1TR0D0NegHalf.TransposeTimes(D1)*R1TR0D0NegHalf;
// Factor M2 = R*D*R^T.
Matrix2<Real> R, D;
M2.EigenDecomposition(R, D);
// Compute K = R^T*K2.
Vector2<Real> K = K2*R;
// Transformed ellipsoid0 is Z^T*Z = 1 and transformed ellipsoid1 is
// (Z-K)^T*D*(Z-K) = 0.
// The minimum and maximum squared distances from the origin of points on
// transformed ellipse1 are used to determine whether the ellipses
// intersect, are separated, or one contains the other.
Real minSqrDistance = Math<Real>::MAX_REAL;
Real maxSqrDistance = (Real)0;
int i;
if (K == Vector2<Real>::ZERO)
{
// The special case of common centers must be handled separately. It
// is not possible for the ellipsoids to be separated.
for (i = 0; i < 2; ++i)
{
Real invD = ((Real)1)/D[i][i];
if (invD < minSqrDistance)
{
minSqrDistance = invD;
}
if (invD > maxSqrDistance)
{
maxSqrDistance = invD;
}
}
if (maxSqrDistance < (Real)1)
{
return EC_ELLIPSE0_CONTAINS_ELLIPSE1;
}
else if (minSqrDistance > (Real)1)
{
return EC_ELLIPSE1_CONTAINS_ELLIPSE0;
}
else
{
return EC_ELLIPSES_INTERSECTING;
}
}
// The closest point P0 and farthest point P1 are solutions to
// s0*D*(P0 - K) = P0 and s1*D1*(P1 - K) = P1 for some scalars s0 and s1
// that are roots to the function
// f(s) = d0*k0^2/(d0*s-1)^2 + d1*k1^2/(d1*s-1)^2 - 1
// where D = diagonal(d0,d1) and K = (k0,k1).
Real d0 = D[0][0], d1 = D[1][1];
Real c0 = K2[0]*K2[0], c1 = K2[1]*K2[1];
// Sort the values so that d0 >= d1. This allows us to bound the roots of
// f(s), of which there are at most 4.
std::vector<std::pair<Real,Real> > param(2);
if (d0 >= d1)
{
param[0] = std::make_pair(d0, c0);
param[1] = std::make_pair(d1, c1);
}
else
{
param[0] = std::make_pair(d1, c1);
param[1] = std::make_pair(d0, c0);
}
std::vector<std::pair<Real,Real> > valid;
valid.reserve(2);
if (param[0].first > param[1].first)
{
// d0 > d1
for (i = 0; i < 2; ++i)
{
if (param[i].second > (Real)0)
{
valid.push_back(param[i]);
}
}
}
else
{
// d0 = d1
param[0].second += param[1].second;
if (param[0].second > (Real)0)
{
valid.push_back(param[0]);
}
}
size_t numValid = valid.size();
int numRoots;
Real roots[4];
if (numValid == 2)
{
GetRoots(
valid[0].first, valid[1].first,
valid[0].second, valid[1].second,
numRoots, roots);
}
else if (numValid == 1)
{
GetRoots(
valid[0].first,
valid[0].second,
numRoots, roots);
}
else
{
// numValid cannot be zero because we already handled case K = 0
assertion(false, "Unexpected condition.\n");
return EC_ELLIPSES_INTERSECTING;
}
for (int i = 0; i < numRoots; ++i)
{
Real s = roots[i];
Real p0 = d0*K[0]*s/(d0*s - (Real)1);
Real p1 = d1*K[1]*s/(d1*s - (Real)1);
Real sqrDistance = p0*p0 + p1*p1;
if (sqrDistance < minSqrDistance)
{
minSqrDistance = sqrDistance;
}
if (sqrDistance > maxSqrDistance)
{
maxSqrDistance = sqrDistance;
}
}
if (maxSqrDistance < (Real)1)
{
return EC_ELLIPSE0_CONTAINS_ELLIPSE1;
}
if (minSqrDistance > (Real)1)
{
if (d0*c0 + d1*c1 > (Real)1)
{
return EC_ELLIPSES_SEPARATED;
}
else
{
return EC_ELLIPSE1_CONTAINS_ELLIPSE0;
}
}
return EC_ELLIPSES_INTERSECTING;
}
int Rsimp(int m, int n, double **A, double *b, double *c,
double *x, int *basis, int *nonbasis,
double **R, double **Q, double *t1, double *t2){
int i,j,k,l,q,qv;
int max_steps=20;
double r,a,at;
void GQR(int,int,double**,double**);
max_steps=4*n;
for(k=0; k<=max_steps;k++){
/*
++ Step 0) load new basis matrix and factor it
*/
for(i=0;i<m;i++)for(j=0;j<m;j++)R0(i,j)=AB0(i,j);
GQR(m,m,Q,R);
/*
++ Step 1) solving system B'*w=c(basis)
++ a) forward solve R'*y=c(basis)
*/
for(i=0;i<m;i++){
Y0(i)=0.0;
for(j=0;j<i;j++)Y0(i)+=R0(j,i)*Y0(j);
if (R0(i,i)!=0.0) Y0(i)=(CB0(i)-Y0(i))/R0(i,i);
else {
printf("Warning Singular Matrix Found\n");
return LP_FAIL;
}
}
/*
++ b) find w=Q*y
++ note: B'*w=(Q*R)'*Q*y= R'*(Q'*Q)*y=R'*y=c(basis)
*/
for(i=0;i<m;i++){
W0(i)=0.0;
for(j=0;j<m;j++)W0(i)+=Q0(i,j)*Y0(j);
}
/*
++ Step 2)find entering variable,
++ (use lexicographically first variable with negative reduced cost)
*/
q=n;
for(i=0;i<n-m;i++){
/* calculate reduced cost */
r=CN0(i);
for(j=0;j<m;j++) r-=W0(j)*AN0(j,i);
if (r<-zero_tol && (q==n || nonbasis0(i)<nonbasis0(q))) q=i;
}
/*
++ if ratios were all nonnegative current solution is optimal
*/
if (q==n){
if (verbose>0) printf("optimal solution found in %d iterations\n",k);
return LP_OPT;
}
/*
++ Step 3)Calculate translation direction for q entering
++ by solving system B*d=-A(:,nonbasis(q));
++ a) let y=-Q'*A(:,nonbasis(q));
*/
for(i=0;i<m;i++){
Y0(i)=0.0;
for(j=0;j<m;j++) Y0(i)-=Q0(j,i)*AN0(j,q);
}
/*
++ b) back solve Rd=y (d=R\y)
++ note B*d= Q*R*d=Q*y=Q*-Q'*A(:nonbasis(q))=-A(:,nonbasis(q))
*/
for(i=m-1;i>=0;i--){
D0(i)=0.0;
for(j=m-1;j>=i+1;j--)D0(i)+=R0(i,j)*D0(j);
if (R0(i,i)!=0.0) D0(i)=(Y0(i)-D0(i))/R0(i,i);
else {
printf("Warning Singular Matrix Found\n");
return LP_FAIL;
}
}
/*
++ Step 4 Choose leaving variable
++ (first variable to become negative, by moving in direction D)
++ (if none become negative, then objective function unbounded)
*/
a=0;
l=-1;
for(i=0;i<m;i++){
if (D0(i)<-zero_tol){
at=-1*XB0(i)/D0(i);
if (l==-1 || at<a){ a=at; l=i;}
}
}
if (l==-1){
if (verbose>0){
printf("Objective function Unbounded (%d iterations)\n",k);
}
return LP_UNBD;
}
/*
++ Step 5) Update solution and basis data
*/
XN0(q)=a;
for(j=0;j<m;j++) XB0(j)+=a*D0(j);
XB0(l)=0.0; /* enforce strict zeroness of nonbasis variables */
qv=nonbasis0(q);
nonbasis0(q)=basis0(l);
basis0(l)=qv;
}
if (verbose>=0){
printf("Simplex Algorithm did not Terminate in %d iterations\n",k);
}
return LP_FAIL;
}
static void sha1_transform(uint32_t h[5], const uint8_t data[64])
{
uint32_t a, b, c, d, e;
uint32_t buf[16];
/* copy state and data*/
a = h[0];
b = h[1];
c = h[2];
d = h[3];
e = h[4];
memcpy(buf, data, 64);
/* unrolled sha-1 rounds */
R0(a,b,c,d,e, 0); R0(e,a,b,c,d, 1); R0(d,e,a,b,c, 2); R0(c,d,e,a,b, 3);
R0(b,c,d,e,a, 4); R0(a,b,c,d,e, 5); R0(e,a,b,c,d, 6); R0(d,e,a,b,c, 7);
R0(c,d,e,a,b, 8); R0(b,c,d,e,a, 9); R0(a,b,c,d,e,10); R0(e,a,b,c,d,11);
R0(d,e,a,b,c,12); R0(c,d,e,a,b,13); R0(b,c,d,e,a,14); R0(a,b,c,d,e,15);
R1(e,a,b,c,d,16); R1(d,e,a,b,c,17); R1(c,d,e,a,b,18); R1(b,c,d,e,a,19);
R2(a,b,c,d,e,20); R2(e,a,b,c,d,21); R2(d,e,a,b,c,22); R2(c,d,e,a,b,23);
R2(b,c,d,e,a,24); R2(a,b,c,d,e,25); R2(e,a,b,c,d,26); R2(d,e,a,b,c,27);
R2(c,d,e,a,b,28); R2(b,c,d,e,a,29); R2(a,b,c,d,e,30); R2(e,a,b,c,d,31);
R2(d,e,a,b,c,32); R2(c,d,e,a,b,33); R2(b,c,d,e,a,34); R2(a,b,c,d,e,35);
R2(e,a,b,c,d,36); R2(d,e,a,b,c,37); R2(c,d,e,a,b,38); R2(b,c,d,e,a,39);
R3(a,b,c,d,e,40); R3(e,a,b,c,d,41); R3(d,e,a,b,c,42); R3(c,d,e,a,b,43);
R3(b,c,d,e,a,44); R3(a,b,c,d,e,45); R3(e,a,b,c,d,46); R3(d,e,a,b,c,47);
R3(c,d,e,a,b,48); R3(b,c,d,e,a,49); R3(a,b,c,d,e,50); R3(e,a,b,c,d,51);
R3(d,e,a,b,c,52); R3(c,d,e,a,b,53); R3(b,c,d,e,a,54); R3(a,b,c,d,e,55);
R3(e,a,b,c,d,56); R3(d,e,a,b,c,57); R3(c,d,e,a,b,58); R3(b,c,d,e,a,59);
R4(a,b,c,d,e,60); R4(e,a,b,c,d,61); R4(d,e,a,b,c,62); R4(c,d,e,a,b,63);
R4(b,c,d,e,a,64); R4(a,b,c,d,e,65); R4(e,a,b,c,d,66); R4(d,e,a,b,c,67);
R4(c,d,e,a,b,68); R4(b,c,d,e,a,69); R4(a,b,c,d,e,70); R4(e,a,b,c,d,71);
R4(d,e,a,b,c,72); R4(c,d,e,a,b,73); R4(b,c,d,e,a,74); R4(a,b,c,d,e,75);
R4(e,a,b,c,d,76); R4(d,e,a,b,c,77); R4(c,d,e,a,b,78); R4(b,c,d,e,a,79);
/* update state */
h[0] += a;
h[1] += b;
h[2] += c;
h[3] += d;
h[4] += e;
/* overwrite all used variables */
a = b = c = d = e = 0;
memset(buf, 0, 64);
}
int main(int argc,char **argv){
// Print GPU properties
//print_properties();
// Files to print the result after the last time step
FILE *rho_file;
FILE *E_file;
rho_file = fopen("rho_final.txt", "w");
E_file = fopen("E_final.txt", "w");
// Construct initial condition for problem
ICsinus Config(-1.0, 1.0, -1.0, 1.0);
//ICsquare Config(0.5,0.5,gasGam);
// Set initial values for Configuration 1
/*
Config.set_rho(rhoConfig19);
Config.set_pressure(pressureConfig19);
Config.set_u(uConfig19);
Config.set_v(vConfig19);
*/
// Determining global border based on left over tiles (a little hack)
int globalPadding;
globalPadding = (nx+2*border+16)/16;
globalPadding = 16*globalPadding - (nx+2*border);
//printf("Globalpad: %i\n", globalPadding);
// Change border to add padding
//border = border + globalPadding/2;
// Initiate the matrices for the unknowns in the Euler equations
cpu_ptr_2D rho(nx, ny, border,1);
cpu_ptr_2D E(nx, ny, border,1);
cpu_ptr_2D rho_u(nx, ny, border,1);
cpu_ptr_2D rho_v(nx, ny, border,1);
cpu_ptr_2D zeros(nx, ny, border,1);
// Set initial condition
Config.setIC(rho, rho_u, rho_v, E);
double timeStart = get_wall_time();
// Test
cpu_ptr_2D rho_dummy(nx, ny, border);
cpu_ptr_2D E_dummy(nx, ny, border);
/*
rho_dummy.xmin = -1.0;
rho_dummy.ymin = -1.0;
E_dummy.xmin = -1.0;
E_dummy.ymin = -1.0;
*/
// Set block and grid sizes
dim3 gridBC = dim3(1, 1, 1);
dim3 blockBC = dim3(BLOCKDIM_BC,1,1);
dim3 gridBlockFlux;
dim3 threadBlockFlux;
dim3 gridBlockRK;
dim3 threadBlockRK;
computeGridBlock(gridBlockFlux, threadBlockFlux, nx + 2*border, ny + 2*border, INNERTILEDIM_X, INNERTILEDIM_Y, BLOCKDIM_X, BLOCKDIM_Y);
computeGridBlock(gridBlockRK, threadBlockRK, nx + 2*border, ny + 2*border, BLOCKDIM_X_RK, BLOCKDIM_Y_RK, BLOCKDIM_X_RK, BLOCKDIM_Y_RK);
int nElements = gridBlockFlux.x*gridBlockFlux.y;
// Allocate memory for the GPU pointers
gpu_ptr_1D L_device(nElements);
gpu_ptr_1D dt_device(1);
gpu_ptr_2D rho_device(nx, ny, border);
gpu_ptr_2D E_device(nx, ny, border);
gpu_ptr_2D rho_u_device(nx, ny, border);
gpu_ptr_2D rho_v_device(nx, ny, border);
gpu_ptr_2D R0(nx, ny, border);
gpu_ptr_2D R1(nx, ny, border);
gpu_ptr_2D R2(nx, ny, border);
gpu_ptr_2D R3(nx, ny, border);
gpu_ptr_2D Q0(nx, ny, border);
gpu_ptr_2D Q1(nx, ny, border);
gpu_ptr_2D Q2(nx, ny, border);
gpu_ptr_2D Q3(nx, ny, border);
// Allocate pinned memory on host
init_allocate();
// Set BC arguments
set_bc_args(BCArgs[0], rho_device.getRawPtr(), rho_u_device.getRawPtr(), rho_v_device.getRawPtr(), E_device.getRawPtr(), nx+2*border, ny+2*border, border);
set_bc_args(BCArgs[1], Q0.getRawPtr(), Q1.getRawPtr(), Q2.getRawPtr(), Q3.getRawPtr(), nx+2*border, ny+2*border, border);
set_bc_args(BCArgs[2], rho_device.getRawPtr(), rho_u_device.getRawPtr(), rho_v_device.getRawPtr(), E_device.getRawPtr(), nx+2*border, ny+2*border, border);
// Set FLUX arguments
set_flux_args(fluxArgs[0], L_device.getRawPtr(), rho_device.getRawPtr(), rho_u_device.getRawPtr(), rho_v_device.getRawPtr(), E_device.getRawPtr(), R0.getRawPtr(),R1.getRawPtr(), R2.getRawPtr(), R3.getRawPtr(), nx, ny, border, rho.get_dx(), rho.get_dy(), theta, gasGam, INNERTILEDIM_X, INNERTILEDIM_Y);
set_flux_args(fluxArgs[1], L_device.getRawPtr(), Q0.getRawPtr(), Q1.getRawPtr(), Q2.getRawPtr(), Q3.getRawPtr(), R0.getRawPtr(),R1.getRawPtr(), R2.getRawPtr(), R3.getRawPtr(), nx, ny, border, rho.get_dx(), rho.get_dy(), theta, gasGam, INNERTILEDIM_X, INNERTILEDIM_Y);
// Set TIME argument
set_dt_args(dtArgs, L_device.getRawPtr(), dt_device.getRawPtr(), nElements, rho.get_dx(), rho.get_dy(), cfl_number);
// Set Rk arguments
set_rk_args(RKArgs[0], dt_device.getRawPtr(), rho_device.getRawPtr(), rho_u_device.getRawPtr(), rho_v_device.getRawPtr(), E_device.getRawPtr(), R0.getRawPtr(), R1.getRawPtr(), R2.getRawPtr(), R3.getRawPtr(), Q0.getRawPtr(), Q1.getRawPtr(), Q2.getRawPtr(), Q3.getRawPtr(), nx, ny, border);
set_rk_args(RKArgs[1], dt_device.getRawPtr(), Q0.getRawPtr(), Q1.getRawPtr(), Q2.getRawPtr(), Q3.getRawPtr(), R0.getRawPtr(), R1.getRawPtr(), R2.getRawPtr(), R3.getRawPtr(), rho_device.getRawPtr(), rho_u_device.getRawPtr(), rho_v_device.getRawPtr(), E_device.getRawPtr(), nx, ny, border);
L_device.set(FLT_MAX);
/*
R0.upload(zeros.get_ptr());
R1.upload(zeros.get_ptr());
R2.upload(zeros.get_ptr());
R3.upload(zeros.get_ptr());
Q0.upload(zeros.get_ptr());
Q1.upload(zeros.get_ptr());
Q2.upload(zeros.get_ptr());
Q3.upload(zeros.get_ptr());
*/
R0.set(0,0,0,nx,ny,border);
R1.set(0,0,0,nx,ny,border);
R2.set(0,0,0,nx,ny,border);
R3.set(0,0,0,nx,ny,border);
Q0.set(0,0,0,nx,ny,border);
Q1.set(0,0,0,nx,ny,border);
Q2.set(0,0,0,nx,ny,border);
Q3.set(0,0,0,nx,ny,border);
rho_device.upload(rho.get_ptr());
rho_u_device.upload(rho_u.get_ptr());
rho_v_device.upload(rho_v.get_ptr());
E_device.upload(E.get_ptr());
// Update boudries
callCollectiveSetBCPeriodic(gridBC, blockBC, BCArgs[0]);
//Create cuda stream
cudaStream_t stream1;
cudaStreamCreate(&stream1);
cudaEvent_t dt_complete;
cudaEventCreate(&dt_complete);
while (currentTime < timeLength && step < maxStep){
//RK1
//Compute flux
callFluxKernel(gridBlockFlux, threadBlockFlux, 0, fluxArgs[0]);
// Compute timestep (based on CFL condition)
callDtKernel(TIMETHREADS, dtArgs);
cudaMemcpyAsync(dt_host, dt_device.getRawPtr(), sizeof(float), cudaMemcpyDeviceToHost, stream1);
cudaEventRecord(dt_complete, stream1);
// Perform RK1 step
callRKKernel(gridBlockRK, threadBlockRK, 0, RKArgs[0]);
//Update boudries
callCollectiveSetBCPeriodic(gridBC, blockBC, BCArgs[1]);
//RK2
// Compute flux
callFluxKernel(gridBlockFlux, threadBlockFlux, 1, fluxArgs[1]);
//Perform RK2 step
callRKKernel(gridBlockRK, threadBlockRK, 1, RKArgs[1]);
//cudaEventRecord(srteam_sync, srteam1);
callCollectiveSetBCPeriodic(gridBC, blockBC, BCArgs[2]);
cudaEventSynchronize(dt_complete);
step++;
currentTime += *dt_host;
// printf("Step: %i, current time: %.6f dt:%.6f\n" , step,currentTime, dt_host[0]);
}
//cuProfilerStop();
//cudaProfilerStop();
printf("Elapsed time %.5f", get_wall_time() - timeStart);
E_device.download(E.get_ptr());
rho_u_device.download(rho_u.get_ptr());
rho_v_device.download(rho_v.get_ptr());
rho_device.download(rho_dummy.get_ptr());
rho_dummy.printToFile(rho_file, true, false);
Config.exactSolution(E_dummy, currentTime);
E_dummy.printToFile(E_file, true, false);
float LinfError = Linf(E_dummy, rho_dummy);
float L1Error = L1(E_dummy, rho_dummy);
float L1Error2 = L1test(E_dummy, rho_dummy);
printf("nx: %i\t Linf error %.9f\t L1 error %.7f L1test erro %.7f", nx, LinfError, L1Error, L1Error2);
printf("nx: %i step: %i, current time: %.6f dt:%.6f\n" , nx, step,currentTime, dt_host[0]);
/*
cudaMemcpy(L_host, L_device, sizeof(float)*(nElements), cudaMemcpyDeviceToHost);
for (int i =0; i < nElements; i++)
printf(" %.7f ", L_host[i]);
*/
printf("%s\n", cudaGetErrorString(cudaGetLastError()));
return(0);
}
void md4_block_data_order(MD4_CTX *c, const void *data_, size_t num) {
const uint8_t *data = data_;
uint32_t A, B, C, D, l;
uint32_t X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X12, X13, X14, X15;
A = c->A;
B = c->B;
C = c->C;
D = c->D;
for (; num--;) {
HOST_c2l(data, l);
X0 = l;
HOST_c2l(data, l);
X1 = l;
/* Round 0 */
R0(A, B, C, D, X0, 3, 0);
HOST_c2l(data, l);
X2 = l;
R0(D, A, B, C, X1, 7, 0);
HOST_c2l(data, l);
X3 = l;
R0(C, D, A, B, X2, 11, 0);
HOST_c2l(data, l);
X4 = l;
R0(B, C, D, A, X3, 19, 0);
HOST_c2l(data, l);
X5 = l;
R0(A, B, C, D, X4, 3, 0);
HOST_c2l(data, l);
X6 = l;
R0(D, A, B, C, X5, 7, 0);
HOST_c2l(data, l);
X7 = l;
R0(C, D, A, B, X6, 11, 0);
HOST_c2l(data, l);
X8 = l;
R0(B, C, D, A, X7, 19, 0);
HOST_c2l(data, l);
X9 = l;
R0(A, B, C, D, X8, 3, 0);
HOST_c2l(data, l);
X10 = l;
R0(D, A, B, C, X9, 7, 0);
HOST_c2l(data, l);
X11 = l;
R0(C, D, A, B, X10, 11, 0);
HOST_c2l(data, l);
X12 = l;
R0(B, C, D, A, X11, 19, 0);
HOST_c2l(data, l);
X13 = l;
R0(A, B, C, D, X12, 3, 0);
HOST_c2l(data, l);
X14 = l;
R0(D, A, B, C, X13, 7, 0);
HOST_c2l(data, l);
X15 = l;
R0(C, D, A, B, X14, 11, 0);
R0(B, C, D, A, X15, 19, 0);
/* Round 1 */
R1(A, B, C, D, X0, 3, 0x5A827999L);
R1(D, A, B, C, X4, 5, 0x5A827999L);
R1(C, D, A, B, X8, 9, 0x5A827999L);
R1(B, C, D, A, X12, 13, 0x5A827999L);
R1(A, B, C, D, X1, 3, 0x5A827999L);
R1(D, A, B, C, X5, 5, 0x5A827999L);
R1(C, D, A, B, X9, 9, 0x5A827999L);
R1(B, C, D, A, X13, 13, 0x5A827999L);
R1(A, B, C, D, X2, 3, 0x5A827999L);
R1(D, A, B, C, X6, 5, 0x5A827999L);
R1(C, D, A, B, X10, 9, 0x5A827999L);
R1(B, C, D, A, X14, 13, 0x5A827999L);
R1(A, B, C, D, X3, 3, 0x5A827999L);
R1(D, A, B, C, X7, 5, 0x5A827999L);
R1(C, D, A, B, X11, 9, 0x5A827999L);
R1(B, C, D, A, X15, 13, 0x5A827999L);
/* Round 2 */
R2(A, B, C, D, X0, 3, 0x6ED9EBA1L);
R2(D, A, B, C, X8, 9, 0x6ED9EBA1L);
R2(C, D, A, B, X4, 11, 0x6ED9EBA1L);
R2(B, C, D, A, X12, 15, 0x6ED9EBA1L);
R2(A, B, C, D, X2, 3, 0x6ED9EBA1L);
R2(D, A, B, C, X10, 9, 0x6ED9EBA1L);
R2(C, D, A, B, X6, 11, 0x6ED9EBA1L);
R2(B, C, D, A, X14, 15, 0x6ED9EBA1L);
R2(A, B, C, D, X1, 3, 0x6ED9EBA1L);
R2(D, A, B, C, X9, 9, 0x6ED9EBA1L);
R2(C, D, A, B, X5, 11, 0x6ED9EBA1L);
R2(B, C, D, A, X13, 15, 0x6ED9EBA1L);
R2(A, B, C, D, X3, 3, 0x6ED9EBA1L);
R2(D, A, B, C, X11, 9, 0x6ED9EBA1L);
R2(C, D, A, B, X7, 11, 0x6ED9EBA1L);
R2(B, C, D, A, X15, 15, 0x6ED9EBA1L);
A = c->A += A;
B = c->B += B;
C = c->C += C;
D = c->D += D;
}
}
void MD5MAC::Transform (word32 *digest, const word32 *X, const word32 *key)
{
// #define F(x,y,z) ((x & y) | (~x & z))
#define F(x,y,z) (z ^ (x & (y^z)))
// #define G(x,y,z) ((x & z) | (y & ~z))
#define G(x,y,z) (y ^ (z & (x^y)))
#define H(x,y,z) (x ^ y ^ z)
#define I(x,y,z) (y ^ (x | ~z))
#define R0(a,b,c,d,k,s,t) { \
a+=(k+t+ F((b),(c),(d)) + key[0]); \
a = rotlFixed(word32(a), (unsigned int)(s)); \
a+=b; };\
#define R1(a,b,c,d,k,s,t) { \
a+=(k+t+ G((b),(c),(d)) + key[1]); \
a = rotlFixed(word32(a), (unsigned int)(s)); \
a+=b; };
#define R2(a,b,c,d,k,s,t) { \
a+=(k+t+ H((b),(c),(d)) + key[2]); \
a = rotlFixed(word32(a), (unsigned int)(s)); \
a+=b; };
#define R3(a,b,c,d,k,s,t) { \
a+=(k+t+ I((b),(c),(d)) + key[3]); \
a = rotlFixed(word32(a), (unsigned int)(s)); \
a+=b; };
register unsigned long A,B,C,D;
A=digest[0];
B=digest[1];
C=digest[2];
D=digest[3];
/* Round 0 */
R0(A,B,C,D,X[ 0], 7,0xd76aa478);
R0(D,A,B,C,X[ 1],12,0xe8c7b756);
R0(C,D,A,B,X[ 2],17,0x242070db);
R0(B,C,D,A,X[ 3],22,0xc1bdceee);
R0(A,B,C,D,X[ 4], 7,0xf57c0faf);
R0(D,A,B,C,X[ 5],12,0x4787c62a);
R0(C,D,A,B,X[ 6],17,0xa8304613);
R0(B,C,D,A,X[ 7],22,0xfd469501);
R0(A,B,C,D,X[ 8], 7,0x698098d8);
R0(D,A,B,C,X[ 9],12,0x8b44f7af);
R0(C,D,A,B,X[10],17,0xffff5bb1);
R0(B,C,D,A,X[11],22,0x895cd7be);
R0(A,B,C,D,X[12], 7,0x6b901122);
R0(D,A,B,C,X[13],12,0xfd987193);
R0(C,D,A,B,X[14],17,0xa679438e);
R0(B,C,D,A,X[15],22,0x49b40821);
/* Round 1 */
R1(A,B,C,D,X[ 1], 5,0xf61e2562);
R1(D,A,B,C,X[ 6], 9,0xc040b340);
R1(C,D,A,B,X[11],14,0x265e5a51);
R1(B,C,D,A,X[ 0],20,0xe9b6c7aa);
R1(A,B,C,D,X[ 5], 5,0xd62f105d);
R1(D,A,B,C,X[10], 9,0x02441453);
R1(C,D,A,B,X[15],14,0xd8a1e681);
R1(B,C,D,A,X[ 4],20,0xe7d3fbc8);
R1(A,B,C,D,X[ 9], 5,0x21e1cde6);
R1(D,A,B,C,X[14], 9,0xc33707d6);
R1(C,D,A,B,X[ 3],14,0xf4d50d87);
R1(B,C,D,A,X[ 8],20,0x455a14ed);
R1(A,B,C,D,X[13], 5,0xa9e3e905);
R1(D,A,B,C,X[ 2], 9,0xfcefa3f8);
R1(C,D,A,B,X[ 7],14,0x676f02d9);
R1(B,C,D,A,X[12],20,0x8d2a4c8a);
/* Round 2 */
R2(A,B,C,D,X[ 5], 4,0xfffa3942);
R2(D,A,B,C,X[ 8],11,0x8771f681);
R2(C,D,A,B,X[11],16,0x6d9d6122);
R2(B,C,D,A,X[14],23,0xfde5380c);
R2(A,B,C,D,X[ 1], 4,0xa4beea44);
R2(D,A,B,C,X[ 4],11,0x4bdecfa9);
R2(C,D,A,B,X[ 7],16,0xf6bb4b60);
R2(B,C,D,A,X[10],23,0xbebfbc70);
R2(A,B,C,D,X[13], 4,0x289b7ec6);
R2(D,A,B,C,X[ 0],11,0xeaa127fa);
R2(C,D,A,B,X[ 3],16,0xd4ef3085);
R2(B,C,D,A,X[ 6],23,0x04881d05);
R2(A,B,C,D,X[ 9], 4,0xd9d4d039);
R2(D,A,B,C,X[12],11,0xe6db99e5);
R2(C,D,A,B,X[15],16,0x1fa27cf8);
R2(B,C,D,A,X[ 2],23,0xc4ac5665);
/* Round 3 */
R3(A,B,C,D,X[ 0], 6,0xf4292244);
R3(D,A,B,C,X[ 7],10,0x432aff97);
R3(C,D,A,B,X[14],15,0xab9423a7);
R3(B,C,D,A,X[ 5],21,0xfc93a039);
R3(A,B,C,D,X[12], 6,0x655b59c3);
R3(D,A,B,C,X[ 3],10,0x8f0ccc92);
R3(C,D,A,B,X[10],15,0xffeff47d);
R3(B,C,D,A,X[ 1],21,0x85845dd1);
R3(A,B,C,D,X[ 8], 6,0x6fa87e4f);
R3(D,A,B,C,X[15],10,0xfe2ce6e0);
R3(C,D,A,B,X[ 6],15,0xa3014314);
R3(B,C,D,A,X[13],21,0x4e0811a1);
R3(A,B,C,D,X[ 4], 6,0xf7537e82);
R3(D,A,B,C,X[11],10,0xbd3af235);
R3(C,D,A,B,X[ 2],15,0x2ad7d2bb);
R3(B,C,D,A,X[ 9],21,0xeb86d391);
digest[0]+=A;
digest[1]+=B;
digest[2]+=C;
digest[3]+=D;
}
void GarbleBlock(unsigned char *block,uint32_t a,uint32_t b,uint32_t c,uint32_t d,uint32_t e)
{
uint32_t W[16];
for(int i=0;i<16;i++) W[i]=(block[4*i+0]<<24)|(block[4*i+1]<<16)|(block[4*i+2]<<8)|block[4*i+3];
R0(a,b,c,d,e, 0); R0(e,a,b,c,d, 1); R0(d,e,a,b,c, 2); R0(c,d,e,a,b, 3);
R0(b,c,d,e,a, 4); R0(a,b,c,d,e, 5); R0(e,a,b,c,d, 6); R0(d,e,a,b,c, 7);
R0(c,d,e,a,b, 8); R0(b,c,d,e,a, 9); R0(a,b,c,d,e,10); R0(e,a,b,c,d,11);
R0(d,e,a,b,c,12); R0(c,d,e,a,b,13); R0(b,c,d,e,a,14); R0(a,b,c,d,e,15);
R1(e,a,b,c,d,16); R1(d,e,a,b,c,17); R1(c,d,e,a,b,18); R1(b,c,d,e,a,19);
R2(a,b,c,d,e,20); R2(e,a,b,c,d,21); R2(d,e,a,b,c,22); R2(c,d,e,a,b,23);
R2(b,c,d,e,a,24); R2(a,b,c,d,e,25); R2(e,a,b,c,d,26); R2(d,e,a,b,c,27);
R2(c,d,e,a,b,28); R2(b,c,d,e,a,29); R2(a,b,c,d,e,30); R2(e,a,b,c,d,31);
R2(d,e,a,b,c,32); R2(c,d,e,a,b,33); R2(b,c,d,e,a,34); R2(a,b,c,d,e,35);
R2(e,a,b,c,d,36); R2(d,e,a,b,c,37); R2(c,d,e,a,b,38); R2(b,c,d,e,a,39);
R3(a,b,c,d,e,40); R3(e,a,b,c,d,41); R3(d,e,a,b,c,42); R3(c,d,e,a,b,43);
R3(b,c,d,e,a,44); R3(a,b,c,d,e,45); R3(e,a,b,c,d,46); R3(d,e,a,b,c,47);
R3(c,d,e,a,b,48); R3(b,c,d,e,a,49); R3(a,b,c,d,e,50); R3(e,a,b,c,d,51);
R3(d,e,a,b,c,52); R3(c,d,e,a,b,53); R3(b,c,d,e,a,54); R3(a,b,c,d,e,55);
R3(e,a,b,c,d,56); R3(d,e,a,b,c,57); R3(c,d,e,a,b,58); R3(b,c,d,e,a,59);
R4(a,b,c,d,e,60); R4(e,a,b,c,d,61); R4(d,e,a,b,c,62); R4(c,d,e,a,b,63);
R4(b,c,d,e,a,64); R4(a,b,c,d,e,65); R4(e,a,b,c,d,66); R4(d,e,a,b,c,67);
R4(c,d,e,a,b,68); R4(b,c,d,e,a,69); R4(a,b,c,d,e,70); R4(e,a,b,c,d,71);
R4(d,e,a,b,c,72); R4(c,d,e,a,b,73); R4(b,c,d,e,a,74); R4(a,b,c,d,e,75);
R4(e,a,b,c,d,76); R4(d,e,a,b,c,77); R4(c,d,e,a,b,78); R4(b,c,d,e,a,79);
for(int i=0;i<64;i++) block[i]=W[i/4]>>(i%4)*8;
}
void md5_block_host_order (MD5_CTX *c, const void *data, size_t num)
{
const MD5_LONG *X=data;
register unsigned MD32_REG_T A,B,C,D;
A=c->A;
B=c->B;
C=c->C;
D=c->D;
for (;num--;X+=HASH_LBLOCK)
{
/* Round 0 */
R0(A,B,C,D,X[ 0], 7,0xd76aa478L);
R0(D,A,B,C,X[ 1],12,0xe8c7b756L);
R0(C,D,A,B,X[ 2],17,0x242070dbL);
R0(B,C,D,A,X[ 3],22,0xc1bdceeeL);
R0(A,B,C,D,X[ 4], 7,0xf57c0fafL);
R0(D,A,B,C,X[ 5],12,0x4787c62aL);
R0(C,D,A,B,X[ 6],17,0xa8304613L);
R0(B,C,D,A,X[ 7],22,0xfd469501L);
R0(A,B,C,D,X[ 8], 7,0x698098d8L);
R0(D,A,B,C,X[ 9],12,0x8b44f7afL);
R0(C,D,A,B,X[10],17,0xffff5bb1L);
R0(B,C,D,A,X[11],22,0x895cd7beL);
R0(A,B,C,D,X[12], 7,0x6b901122L);
R0(D,A,B,C,X[13],12,0xfd987193L);
R0(C,D,A,B,X[14],17,0xa679438eL);
R0(B,C,D,A,X[15],22,0x49b40821L);
/* Round 1 */
R1(A,B,C,D,X[ 1], 5,0xf61e2562L);
R1(D,A,B,C,X[ 6], 9,0xc040b340L);
R1(C,D,A,B,X[11],14,0x265e5a51L);
R1(B,C,D,A,X[ 0],20,0xe9b6c7aaL);
R1(A,B,C,D,X[ 5], 5,0xd62f105dL);
R1(D,A,B,C,X[10], 9,0x02441453L);
R1(C,D,A,B,X[15],14,0xd8a1e681L);
R1(B,C,D,A,X[ 4],20,0xe7d3fbc8L);
R1(A,B,C,D,X[ 9], 5,0x21e1cde6L);
R1(D,A,B,C,X[14], 9,0xc33707d6L);
R1(C,D,A,B,X[ 3],14,0xf4d50d87L);
R1(B,C,D,A,X[ 8],20,0x455a14edL);
R1(A,B,C,D,X[13], 5,0xa9e3e905L);
R1(D,A,B,C,X[ 2], 9,0xfcefa3f8L);
R1(C,D,A,B,X[ 7],14,0x676f02d9L);
R1(B,C,D,A,X[12],20,0x8d2a4c8aL);
/* Round 2 */
R2(A,B,C,D,X[ 5], 4,0xfffa3942L);
R2(D,A,B,C,X[ 8],11,0x8771f681L);
R2(C,D,A,B,X[11],16,0x6d9d6122L);
R2(B,C,D,A,X[14],23,0xfde5380cL);
R2(A,B,C,D,X[ 1], 4,0xa4beea44L);
R2(D,A,B,C,X[ 4],11,0x4bdecfa9L);
R2(C,D,A,B,X[ 7],16,0xf6bb4b60L);
R2(B,C,D,A,X[10],23,0xbebfbc70L);
R2(A,B,C,D,X[13], 4,0x289b7ec6L);
R2(D,A,B,C,X[ 0],11,0xeaa127faL);
R2(C,D,A,B,X[ 3],16,0xd4ef3085L);
R2(B,C,D,A,X[ 6],23,0x04881d05L);
R2(A,B,C,D,X[ 9], 4,0xd9d4d039L);
R2(D,A,B,C,X[12],11,0xe6db99e5L);
R2(C,D,A,B,X[15],16,0x1fa27cf8L);
R2(B,C,D,A,X[ 2],23,0xc4ac5665L);
/* Round 3 */
R3(A,B,C,D,X[ 0], 6,0xf4292244L);
R3(D,A,B,C,X[ 7],10,0x432aff97L);
R3(C,D,A,B,X[14],15,0xab9423a7L);
R3(B,C,D,A,X[ 5],21,0xfc93a039L);
R3(A,B,C,D,X[12], 6,0x655b59c3L);
R3(D,A,B,C,X[ 3],10,0x8f0ccc92L);
R3(C,D,A,B,X[10],15,0xffeff47dL);
R3(B,C,D,A,X[ 1],21,0x85845dd1L);
R3(A,B,C,D,X[ 8], 6,0x6fa87e4fL);
R3(D,A,B,C,X[15],10,0xfe2ce6e0L);
R3(C,D,A,B,X[ 6],15,0xa3014314L);
R3(B,C,D,A,X[13],21,0x4e0811a1L);
R3(A,B,C,D,X[ 4], 6,0xf7537e82L);
R3(D,A,B,C,X[11],10,0xbd3af235L);
R3(C,D,A,B,X[ 2],15,0x2ad7d2bbL);
R3(B,C,D,A,X[ 9],21,0xeb86d391L);
A = c->A += A;
B = c->B += B;
C = c->C += C;
D = c->D += D;
}
}
/* transform one 512bit block. this is the core of the algorithm. */
static void
sha1_transform(uint32_t state[5], char buf[SHA1_BLOCKSIZE])
{
uint32_t a, b, c, d, e;
uint32_t *block = (uint32_t *)buf;
#ifdef SHA1_SHORTCODE
uint8_t i;
#endif
/* copy state to working vars */
a = state[0];
b = state[1];
c = state[2];
d = state[3];
e = state[4];
#ifdef SHA1_SHORTCODE
for (i = 0; i < 80; i++) {
uint32_t t;
if (i < 20)
t = ((b & (c ^ d)) ^ d) + 0x5A827999;
else if (i < 40)
t = (b ^ c ^ d) + 0x6ED9EBA1;
else if (i < 60)
t = (((b | c) & d) | (b & c)) + 0x8F1BBCDC;
else
t = (b ^ c ^ d) + 0xCA62C1D6;
t += (i < 16) ? blk0(i) : blk(i);
t += e + rol(a, 5);
e = d;
d = c;
c = rol(b, 30);
b = a;
a = t;
}
#else
/* R0 and R1, R2, R3, R4 are the different operations used in SHA1 */
#define R0(v,w,x,y,z,i) \
z += ((w&(x^y))^y) + blk0(i) + 0x5A827999 + rol(v, 5); w = rol(w, 30)
#define R1(v,w,x,y,z,i) \
z += ((w&(x^y))^y) + blk(i) + 0x5A827999 + rol(v, 5); w = rol(w, 30)
#define R2(v,w,x,y,z,i) \
z += (w^x^y) + blk(i) + 0x6ED9EBA1 + rol(v, 5); w = rol(w, 30)
#define R3(v,w,x,y,z,i) \
z += (((w|x)&y)|(w&x)) + blk(i) + 0x8F1BBCDC + rol(v, 5); w = rol(w, 30)
#define R4(v,w,x,y,z,i) \
z += (w^x^y) + blk(i) + 0xCA62C1D6 + rol(v, 5); w = rol(w, 30)
/* 4 rounds of 20 operations each. loop unrolled. */
R0(a,b,c,d,e, 0); R0(e,a,b,c,d, 1); R0(d,e,a,b,c, 2); R0(c,d,e,a,b, 3);
R0(b,c,d,e,a, 4); R0(a,b,c,d,e, 5); R0(e,a,b,c,d, 6); R0(d,e,a,b,c, 7);
R0(c,d,e,a,b, 8); R0(b,c,d,e,a, 9); R0(a,b,c,d,e,10); R0(e,a,b,c,d,11);
R0(d,e,a,b,c,12); R0(c,d,e,a,b,13); R0(b,c,d,e,a,14); R0(a,b,c,d,e,15);
R1(e,a,b,c,d,16); R1(d,e,a,b,c,17); R1(c,d,e,a,b,18); R1(b,c,d,e,a,19);
R2(a,b,c,d,e,20); R2(e,a,b,c,d,21); R2(d,e,a,b,c,22); R2(c,d,e,a,b,23);
R2(b,c,d,e,a,24); R2(a,b,c,d,e,25); R2(e,a,b,c,d,26); R2(d,e,a,b,c,27);
R2(c,d,e,a,b,28); R2(b,c,d,e,a,29); R2(a,b,c,d,e,30); R2(e,a,b,c,d,31);
R2(d,e,a,b,c,32); R2(c,d,e,a,b,33); R2(b,c,d,e,a,34); R2(a,b,c,d,e,35);
R2(e,a,b,c,d,36); R2(d,e,a,b,c,37); R2(c,d,e,a,b,38); R2(b,c,d,e,a,39);
R3(a,b,c,d,e,40); R3(e,a,b,c,d,41); R3(d,e,a,b,c,42); R3(c,d,e,a,b,43);
R3(b,c,d,e,a,44); R3(a,b,c,d,e,45); R3(e,a,b,c,d,46); R3(d,e,a,b,c,47);
R3(c,d,e,a,b,48); R3(b,c,d,e,a,49); R3(a,b,c,d,e,50); R3(e,a,b,c,d,51);
R3(d,e,a,b,c,52); R3(c,d,e,a,b,53); R3(b,c,d,e,a,54); R3(a,b,c,d,e,55);
R3(e,a,b,c,d,56); R3(d,e,a,b,c,57); R3(c,d,e,a,b,58); R3(b,c,d,e,a,59);
R4(a,b,c,d,e,60); R4(e,a,b,c,d,61); R4(d,e,a,b,c,62); R4(c,d,e,a,b,63);
R4(b,c,d,e,a,64); R4(a,b,c,d,e,65); R4(e,a,b,c,d,66); R4(d,e,a,b,c,67);
R4(c,d,e,a,b,68); R4(b,c,d,e,a,69); R4(a,b,c,d,e,70); R4(e,a,b,c,d,71);
R4(d,e,a,b,c,72); R4(c,d,e,a,b,73); R4(b,c,d,e,a,74); R4(a,b,c,d,e,75);
R4(e,a,b,c,d,76); R4(d,e,a,b,c,77); R4(c,d,e,a,b,78); R4(b,c,d,e,a,79);
#endif
/* add the working vars back into state */
state[0] += a;
state[1] += b;
state[2] += c;
state[3] += d;
state[4] += e;
}
void GQR(int r, int c, double **Q, double **R){
int i,j,k;
double s,s1,s2;
double t1,t2;
for(i=0;i<r;i++){
for(k=0;k<r;k++)Q0(i,k)=0.0;
Q0(i,i)=1.0;
}
for (i=0;i<c;i++)
for (k=i+1;k<r;k++)
/* performing givens rotations to zero A[k][i] */
if (R0(k,i)!=0){
s=sqrt(R0(i,i)*R0(i,i)+R0(k,i)*R0(k,i));
s1=R0(i,i)/s;
s2=R0(k,i)/s;
for(j=0;j<c;j++) {
t1=R0(i,j);
t2=R0(k,j);
R0(i,j)=s1*t1+s2*t2;
R0(k,j)=-s2*t1+s1*t2;
}
/* actually doing givens row rotations on Q */
for(j=0;j<r;j++){
t1=Q0(j,i);
t2=Q0(j,k);
Q0(j,i)=s1*t1+s2*t2;
Q0(j,k)=-s2*t1+s1*t2;
}
}
}
void SHA1::Transform(word32 *state, const word32 *data)
{
word32 W[16];
/* Copy context->state[] to working vars */
word32 a = state[0];
word32 b = state[1];
word32 c = state[2];
word32 d = state[3];
word32 e = state[4];
/* 4 rounds of 20 operations each. Loop unrolled. */
R0(a,b,c,d,e, 0); R0(e,a,b,c,d, 1); R0(d,e,a,b,c, 2); R0(c,d,e,a,b, 3);
R0(b,c,d,e,a, 4); R0(a,b,c,d,e, 5); R0(e,a,b,c,d, 6); R0(d,e,a,b,c, 7);
R0(c,d,e,a,b, 8); R0(b,c,d,e,a, 9); R0(a,b,c,d,e,10); R0(e,a,b,c,d,11);
R0(d,e,a,b,c,12); R0(c,d,e,a,b,13); R0(b,c,d,e,a,14); R0(a,b,c,d,e,15);
R1(e,a,b,c,d,16); R1(d,e,a,b,c,17); R1(c,d,e,a,b,18); R1(b,c,d,e,a,19);
R2(a,b,c,d,e,20); R2(e,a,b,c,d,21); R2(d,e,a,b,c,22); R2(c,d,e,a,b,23);
R2(b,c,d,e,a,24); R2(a,b,c,d,e,25); R2(e,a,b,c,d,26); R2(d,e,a,b,c,27);
R2(c,d,e,a,b,28); R2(b,c,d,e,a,29); R2(a,b,c,d,e,30); R2(e,a,b,c,d,31);
R2(d,e,a,b,c,32); R2(c,d,e,a,b,33); R2(b,c,d,e,a,34); R2(a,b,c,d,e,35);
R2(e,a,b,c,d,36); R2(d,e,a,b,c,37); R2(c,d,e,a,b,38); R2(b,c,d,e,a,39);
R3(a,b,c,d,e,40); R3(e,a,b,c,d,41); R3(d,e,a,b,c,42); R3(c,d,e,a,b,43);
R3(b,c,d,e,a,44); R3(a,b,c,d,e,45); R3(e,a,b,c,d,46); R3(d,e,a,b,c,47);
R3(c,d,e,a,b,48); R3(b,c,d,e,a,49); R3(a,b,c,d,e,50); R3(e,a,b,c,d,51);
R3(d,e,a,b,c,52); R3(c,d,e,a,b,53); R3(b,c,d,e,a,54); R3(a,b,c,d,e,55);
R3(e,a,b,c,d,56); R3(d,e,a,b,c,57); R3(c,d,e,a,b,58); R3(b,c,d,e,a,59);
R4(a,b,c,d,e,60); R4(e,a,b,c,d,61); R4(d,e,a,b,c,62); R4(c,d,e,a,b,63);
R4(b,c,d,e,a,64); R4(a,b,c,d,e,65); R4(e,a,b,c,d,66); R4(d,e,a,b,c,67);
R4(c,d,e,a,b,68); R4(b,c,d,e,a,69); R4(a,b,c,d,e,70); R4(e,a,b,c,d,71);
R4(d,e,a,b,c,72); R4(c,d,e,a,b,73); R4(b,c,d,e,a,74); R4(a,b,c,d,e,75);
R4(e,a,b,c,d,76); R4(d,e,a,b,c,77); R4(c,d,e,a,b,78); R4(b,c,d,e,a,79);
/* Add the working vars back into context.state[] */
state[0] += a;
state[1] += b;
state[2] += c;
state[3] += d;
state[4] += e;
}
void
transform(
std::uint32_t digest[], std::uint32_t block[BLOCK_INTS])
{
std::uint32_t a = digest[0];
std::uint32_t b = digest[1];
std::uint32_t c = digest[2];
std::uint32_t d = digest[3];
std::uint32_t e = digest[4];
R0(block, a, b, c, d, e, 0);
R0(block, e, a, b, c, d, 1);
R0(block, d, e, a, b, c, 2);
R0(block, c, d, e, a, b, 3);
R0(block, b, c, d, e, a, 4);
R0(block, a, b, c, d, e, 5);
R0(block, e, a, b, c, d, 6);
R0(block, d, e, a, b, c, 7);
R0(block, c, d, e, a, b, 8);
R0(block, b, c, d, e, a, 9);
R0(block, a, b, c, d, e, 10);
R0(block, e, a, b, c, d, 11);
R0(block, d, e, a, b, c, 12);
R0(block, c, d, e, a, b, 13);
R0(block, b, c, d, e, a, 14);
R0(block, a, b, c, d, e, 15);
R1(block, e, a, b, c, d, 0);
R1(block, d, e, a, b, c, 1);
R1(block, c, d, e, a, b, 2);
R1(block, b, c, d, e, a, 3);
R2(block, a, b, c, d, e, 4);
R2(block, e, a, b, c, d, 5);
R2(block, d, e, a, b, c, 6);
R2(block, c, d, e, a, b, 7);
R2(block, b, c, d, e, a, 8);
R2(block, a, b, c, d, e, 9);
R2(block, e, a, b, c, d, 10);
R2(block, d, e, a, b, c, 11);
R2(block, c, d, e, a, b, 12);
R2(block, b, c, d, e, a, 13);
R2(block, a, b, c, d, e, 14);
R2(block, e, a, b, c, d, 15);
R2(block, d, e, a, b, c, 0);
R2(block, c, d, e, a, b, 1);
R2(block, b, c, d, e, a, 2);
R2(block, a, b, c, d, e, 3);
R2(block, e, a, b, c, d, 4);
R2(block, d, e, a, b, c, 5);
R2(block, c, d, e, a, b, 6);
R2(block, b, c, d, e, a, 7);
R3(block, a, b, c, d, e, 8);
R3(block, e, a, b, c, d, 9);
R3(block, d, e, a, b, c, 10);
R3(block, c, d, e, a, b, 11);
R3(block, b, c, d, e, a, 12);
R3(block, a, b, c, d, e, 13);
R3(block, e, a, b, c, d, 14);
R3(block, d, e, a, b, c, 15);
R3(block, c, d, e, a, b, 0);
R3(block, b, c, d, e, a, 1);
R3(block, a, b, c, d, e, 2);
R3(block, e, a, b, c, d, 3);
R3(block, d, e, a, b, c, 4);
R3(block, c, d, e, a, b, 5);
R3(block, b, c, d, e, a, 6);
R3(block, a, b, c, d, e, 7);
R3(block, e, a, b, c, d, 8);
R3(block, d, e, a, b, c, 9);
R3(block, c, d, e, a, b, 10);
R3(block, b, c, d, e, a, 11);
R4(block, a, b, c, d, e, 12);
R4(block, e, a, b, c, d, 13);
R4(block, d, e, a, b, c, 14);
R4(block, c, d, e, a, b, 15);
R4(block, b, c, d, e, a, 0);
R4(block, a, b, c, d, e, 1);
R4(block, e, a, b, c, d, 2);
R4(block, d, e, a, b, c, 3);
R4(block, c, d, e, a, b, 4);
R4(block, b, c, d, e, a, 5);
R4(block, a, b, c, d, e, 6);
R4(block, e, a, b, c, d, 7);
R4(block, d, e, a, b, c, 8);
R4(block, c, d, e, a, b, 9);
R4(block, b, c, d, e, a, 10);
R4(block, a, b, c, d, e, 11);
R4(block, e, a, b, c, d, 12);
R4(block, d, e, a, b, c, 13);
R4(block, c, d, e, a, b, 14);
R4(block, b, c, d, e, a, 15);
digest[0] += a;
digest[1] += b;
digest[2] += c;
digest[3] += d;
digest[4] += e;
}
static void sha1_transform(uint32_t state[5], const uint8_t buffer[64])
{
uint32_t block[80];
unsigned int i, a, b, c, d, e;
a = state[0];
b = state[1];
c = state[2];
d = state[3];
e = state[4];
#if CONFIG_SMALL
for (i = 0; i < 80; i++) {
int t;
if (i < 16)
t = AV_RB32(buffer + 4 * i);
else
t = rol(block[i-3] ^ block[i-8] ^ block[i-14] ^ block[i-16], 1);
block[i] = t;
t += e + rol(a, 5);
if (i < 40) {
if (i < 20)
t += ((b&(c^d))^d) + 0x5A827999;
else
t += ( b^c ^d) + 0x6ED9EBA1;
} else {
if (i < 60)
t += (((b|c)&d)|(b&c)) + 0x8F1BBCDC;
else
t += ( b^c ^d) + 0xCA62C1D6;
}
e = d;
d = c;
c = rol(b, 30);
b = a;
a = t;
}
#else
#define R1_0 \
R0(a, b, c, d, e, 0 + i); \
R0(e, a, b, c, d, 1 + i); \
R0(d, e, a, b, c, 2 + i); \
R0(c, d, e, a, b, 3 + i); \
R0(b, c, d, e, a, 4 + i); \
i += 5
i = 0;
R1_0; R1_0; R1_0;
R0(a, b, c, d, e, 15);
R1(e, a, b, c, d, 16);
R1(d, e, a, b, c, 17);
R1(c, d, e, a, b, 18);
R1(b, c, d, e, a, 19);
#define R1_20 \
R2(a, b, c, d, e, 0 + i); \
R2(e, a, b, c, d, 1 + i); \
R2(d, e, a, b, c, 2 + i); \
R2(c, d, e, a, b, 3 + i); \
R2(b, c, d, e, a, 4 + i); \
i += 5
i = 20;
R1_20; R1_20; R1_20; R1_20;
#define R1_40 \
R3(a, b, c, d, e, 0 + i); \
R3(e, a, b, c, d, 1 + i); \
R3(d, e, a, b, c, 2 + i); \
R3(c, d, e, a, b, 3 + i); \
R3(b, c, d, e, a, 4 + i); \
i += 5
R1_40; R1_40; R1_40; R1_40;
#define R1_60 \
R4(a, b, c, d, e, 0 + i); \
R4(e, a, b, c, d, 1 + i); \
R4(d, e, a, b, c, 2 + i); \
R4(c, d, e, a, b, 3 + i); \
R4(b, c, d, e, a, 4 + i); \
i += 5
R1_60; R1_60; R1_60; R1_60;
#endif
state[0] += a;
state[1] += b;
state[2] += c;
state[3] += d;
state[4] += e;
}
static void ripemd_256_transform(akmos_ripemd_t *ctx, const uint8_t *block, size_t nb)
{
uint32_t a, b, c, d, aa, bb, cc, dd, t, *x, *state;
size_t i;
state = ctx->state;
x = ctx->x;
for(i = 0; i < nb; i++, block += AKMOS_RIPEMD_BLKLEN) {
memcpy(x, block, AKMOS_RIPEMD_BLKLEN);
a = state[0];
b = state[1];
c = state[2];
d = state[3];
aa = state[4];
bb = state[5];
cc = state[6];
dd = state[7];
/* Round 1 */
R0(a, b, c, d, F0, K0, 11, 0);
R0(d, a, b, c, F0, K0, 14, 1);
R0(c, d, a, b, F0, K0, 15, 2);
R0(b, c, d, a, F0, K0, 12, 3);
R0(a, b, c, d, F0, K0, 5, 4);
R0(d, a, b, c, F0, K0, 8, 5);
R0(c, d, a, b, F0, K0, 7, 6);
R0(b, c, d, a, F0, K0, 9, 7);
R0(a, b, c, d, F0, K0, 11, 8);
R0(d, a, b, c, F0, K0, 13, 9);
R0(c, d, a, b, F0, K0, 14, 10);
R0(b, c, d, a, F0, K0, 15, 11);
R0(a, b, c, d, F0, K0, 6, 12);
R0(d, a, b, c, F0, K0, 7, 13);
R0(c, d, a, b, F0, K0, 9, 14);
R0(b, c, d, a, F0, K0, 8, 15);
R0(aa, bb, cc, dd, F3, KK0, 8, 5);
R0(dd, aa, bb, cc, F3, KK0, 9, 14);
R0(cc, dd, aa, bb, F3, KK0, 9, 7);
R0(bb, cc, dd, aa, F3, KK0, 11, 0);
R0(aa, bb, cc, dd, F3, KK0, 13, 9);
R0(dd, aa, bb, cc, F3, KK0, 15, 2);
R0(cc, dd, aa, bb, F3, KK0, 15, 11);
R0(bb, cc, dd, aa, F3, KK0, 5, 4);
R0(aa, bb, cc, dd, F3, KK0, 7, 13);
R0(dd, aa, bb, cc, F3, KK0, 7, 6);
R0(cc, dd, aa, bb, F3, KK0, 8, 15);
R0(bb, cc, dd, aa, F3, KK0, 11, 8);
R0(aa, bb, cc, dd, F3, KK0, 14, 1);
R0(dd, aa, bb, cc, F3, KK0, 14, 10);
R0(cc, dd, aa, bb, F3, KK0, 12, 3);
R0(bb, cc, dd, aa, F3, KK0, 6, 12); /* #15 */
t = a; a = aa; aa = t;
/* Round 2 */
R0(a, b, c, d, F1, K1, 7, 7);
R0(d, a, b, c, F1, K1, 6, 4);
R0(c, d, a, b, F1, K1, 8, 13);
R0(b, c, d, a, F1, K1, 13, 1);
R0(a, b, c, d, F1, K1, 11, 10);
R0(d, a, b, c, F1, K1, 9, 6);
R0(c, d, a, b, F1, K1, 7, 15);
R0(b, c, d, a, F1, K1, 15, 3);
R0(a, b, c, d, F1, K1, 7, 12);
R0(d, a, b, c, F1, K1, 12, 0);
R0(c, d, a, b, F1, K1, 15, 9);
R0(b, c, d, a, F1, K1, 9, 5);
R0(a, b, c, d, F1, K1, 11, 2);
R0(d, a, b, c, F1, K1, 7, 14);
R0(c, d, a, b, F1, K1, 13, 11);
R0(b, c, d, a, F1, K1, 12, 8);
R0(aa, bb, cc, dd, F2, KK1, 9, 6);
R0(dd, aa, bb, cc, F2, KK1, 13, 11);
R0(cc, dd, aa, bb, F2, KK1, 15, 3);
R0(bb, cc, dd, aa, F2, KK1, 7, 7);
R0(aa, bb, cc, dd, F2, KK1, 12, 0);
R0(dd, aa, bb, cc, F2, KK1, 8, 13);
R0(cc, dd, aa, bb, F2, KK1, 9, 5);
R0(bb, cc, dd, aa, F2, KK1, 11, 10);
R0(aa, bb, cc, dd, F2, KK1, 7, 14);
R0(dd, aa, bb, cc, F2, KK1, 7, 15);
R0(cc, dd, aa, bb, F2, KK1, 12, 8);
R0(bb, cc, dd, aa, F2, KK1, 7, 12);
R0(aa, bb, cc, dd, F2, KK1, 6, 4);
R0(dd, aa, bb, cc, F2, KK1, 15, 9);
R0(cc, dd, aa, bb, F2, KK1, 13, 1);
R0(bb, cc, dd, aa, F2, KK1, 11, 2); /* #31 */
t = b; b = bb; bb = t;
/* Round 3 */
R0(a, b, c, d, F2, K2, 11, 3);
R0(d, a, b, c, F2, K2, 13, 10);
R0(c, d, a, b, F2, K2, 6, 14);
R0(b, c, d, a, F2, K2, 7, 4);
R0(a, b, c, d, F2, K2, 14, 9);
R0(d, a, b, c, F2, K2, 9, 15);
R0(c, d, a, b, F2, K2, 13, 8);
R0(b, c, d, a, F2, K2, 15, 1);
R0(a, b, c, d, F2, K2, 14, 2);
R0(d, a, b, c, F2, K2, 8, 7);
R0(c, d, a, b, F2, K2, 13, 0);
R0(b, c, d, a, F2, K2, 6, 6);
R0(a, b, c, d, F2, K2, 5, 13);
R0(d, a, b, c, F2, K2, 12, 11);
R0(c, d, a, b, F2, K2, 7, 5);
R0(b, c, d, a, F2, K2, 5, 12);
R0(aa, bb, cc, dd, F1, KK2, 9, 15);
R0(dd, aa, bb, cc, F1, KK2, 7, 5);
R0(cc, dd, aa, bb, F1, KK2, 15, 1);
R0(bb, cc, dd, aa, F1, KK2, 11, 3);
R0(aa, bb, cc, dd, F1, KK2, 8, 7);
R0(dd, aa, bb, cc, F1, KK2, 6, 14);
R0(cc, dd, aa, bb, F1, KK2, 6, 6);
R0(bb, cc, dd, aa, F1, KK2, 14, 9);
R0(aa, bb, cc, dd, F1, KK2, 12, 11);
R0(dd, aa, bb, cc, F1, KK2, 13, 8);
R0(cc, dd, aa, bb, F1, KK2, 5, 12);
R0(bb, cc, dd, aa, F1, KK2, 14, 2);
R0(aa, bb, cc, dd, F1, KK2, 13, 10);
R0(dd, aa, bb, cc, F1, KK2, 13, 0);
R0(cc, dd, aa, bb, F1, KK2, 7, 4);
R0(bb, cc, dd, aa, F1, KK2, 5, 13); /* #47 */
t = c; c = cc; cc = t;
/* Round 4 */
R0(a, b, c, d, F3, K3, 11, 1);
R0(d, a, b, c, F3, K3, 12, 9);
R0(c, d, a, b, F3, K3, 14, 11);
R0(b, c, d, a, F3, K3, 15, 10);
R0(a, b, c, d, F3, K3, 14, 0);
R0(d, a, b, c, F3, K3, 15, 8);
R0(c, d, a, b, F3, K3, 9, 12);
R0(b, c, d, a, F3, K3, 8, 4);
R0(a, b, c, d, F3, K3, 9, 13);
R0(d, a, b, c, F3, K3, 14, 3);
R0(c, d, a, b, F3, K3, 5, 7);
R0(b, c, d, a, F3, K3, 6, 15);
R0(a, b, c, d, F3, K3, 8, 14);
R0(d, a, b, c, F3, K3, 6, 5);
R0(c, d, a, b, F3, K3, 5, 6);
R0(b, c, d, a, F3, K3, 12, 2);
R0(aa, bb, cc, dd, F0, KK4, 15, 8);
R0(dd, aa, bb, cc, F0, KK4, 5, 6);
R0(cc, dd, aa, bb, F0, KK4, 8, 4);
R0(bb, cc, dd, aa, F0, KK4, 11, 1);
R0(aa, bb, cc, dd, F0, KK4, 14, 3);
R0(dd, aa, bb, cc, F0, KK4, 14, 11);
R0(cc, dd, aa, bb, F0, KK4, 6, 15);
R0(bb, cc, dd, aa, F0, KK4, 14, 0);
R0(aa, bb, cc, dd, F0, KK4, 6, 5);
R0(dd, aa, bb, cc, F0, KK4, 9, 12);
R0(cc, dd, aa, bb, F0, KK4, 12, 2);
R0(bb, cc, dd, aa, F0, KK4, 9, 13);
R0(aa, bb, cc, dd, F0, KK4, 12, 9);
R0(dd, aa, bb, cc, F0, KK4, 5, 7);
R0(cc, dd, aa, bb, F0, KK4, 15, 10);
R0(bb, cc, dd, aa, F0, KK4, 8, 14); /* #63 */
t = d; d = dd; dd = t;
state[0] += a;
state[1] += b;
state[2] += c;
state[3] += d;
state[4] += aa;
state[5] += bb;
state[6] += cc;
state[7] += dd;
}
}
void md4_block_host_order (MD4_CTX *c, const void *data, size_t num)
{
const MD4_LONG *X=data;
register unsigned MD32_REG_T A,B,C,D;
A=c->A;
B=c->B;
C=c->C;
D=c->D;
for (;num--;X+=HASH_LBLOCK)
{
/* Round 0 */
R0(A,B,C,D,X[ 0], 3,0);
R0(D,A,B,C,X[ 1], 7,0);
R0(C,D,A,B,X[ 2],11,0);
R0(B,C,D,A,X[ 3],19,0);
R0(A,B,C,D,X[ 4], 3,0);
R0(D,A,B,C,X[ 5], 7,0);
R0(C,D,A,B,X[ 6],11,0);
R0(B,C,D,A,X[ 7],19,0);
R0(A,B,C,D,X[ 8], 3,0);
R0(D,A,B,C,X[ 9], 7,0);
R0(C,D,A,B,X[10],11,0);
R0(B,C,D,A,X[11],19,0);
R0(A,B,C,D,X[12], 3,0);
R0(D,A,B,C,X[13], 7,0);
R0(C,D,A,B,X[14],11,0);
R0(B,C,D,A,X[15],19,0);
/* Round 1 */
R1(A,B,C,D,X[ 0], 3,0x5A827999L);
R1(D,A,B,C,X[ 4], 5,0x5A827999L);
R1(C,D,A,B,X[ 8], 9,0x5A827999L);
R1(B,C,D,A,X[12],13,0x5A827999L);
R1(A,B,C,D,X[ 1], 3,0x5A827999L);
R1(D,A,B,C,X[ 5], 5,0x5A827999L);
R1(C,D,A,B,X[ 9], 9,0x5A827999L);
R1(B,C,D,A,X[13],13,0x5A827999L);
R1(A,B,C,D,X[ 2], 3,0x5A827999L);
R1(D,A,B,C,X[ 6], 5,0x5A827999L);
R1(C,D,A,B,X[10], 9,0x5A827999L);
R1(B,C,D,A,X[14],13,0x5A827999L);
R1(A,B,C,D,X[ 3], 3,0x5A827999L);
R1(D,A,B,C,X[ 7], 5,0x5A827999L);
R1(C,D,A,B,X[11], 9,0x5A827999L);
R1(B,C,D,A,X[15],13,0x5A827999L);
/* Round 2 */
R2(A,B,C,D,X[ 0], 3,0x6ED9EBA1);
R2(D,A,B,C,X[ 8], 9,0x6ED9EBA1);
R2(C,D,A,B,X[ 4],11,0x6ED9EBA1);
R2(B,C,D,A,X[12],15,0x6ED9EBA1);
R2(A,B,C,D,X[ 2], 3,0x6ED9EBA1);
R2(D,A,B,C,X[10], 9,0x6ED9EBA1);
R2(C,D,A,B,X[ 6],11,0x6ED9EBA1);
R2(B,C,D,A,X[14],15,0x6ED9EBA1);
R2(A,B,C,D,X[ 1], 3,0x6ED9EBA1);
R2(D,A,B,C,X[ 9], 9,0x6ED9EBA1);
R2(C,D,A,B,X[ 5],11,0x6ED9EBA1);
R2(B,C,D,A,X[13],15,0x6ED9EBA1);
R2(A,B,C,D,X[ 3], 3,0x6ED9EBA1);
R2(D,A,B,C,X[11], 9,0x6ED9EBA1);
R2(C,D,A,B,X[ 7],11,0x6ED9EBA1);
R2(B,C,D,A,X[15],15,0x6ED9EBA1);
A = c->A += A;
B = c->B += B;
C = c->C += C;
D = c->D += D;
}
}
void sha512_hash_block(sha512_ctx *ctx, const unsigned char data[128], int perform_endian_swap)
{
ARCH_WORD_64 A, B, C, D, E, F, G, H, tmp, W[80];
int i;
#if ARCH_LITTLE_ENDIAN
if (perform_endian_swap) {
for(i = 0; i < 16; i++) {
W[i] = JOHNSWAP64(*((ARCH_WORD_64 *)&(data[i<<3])));
}
} else
#endif
{
i = 16;
memcpy(W, data, 128);
}
for(; i < 80; i++)
W[i] = R1(W[i - 2]) + W[i - 7] + R0(W[i - 15]) + W[i - 16];
A = ctx->h[0];
B = ctx->h[1];
C = ctx->h[2];
D = ctx->h[3];
E = ctx->h[4];
F = ctx->h[5];
G = ctx->h[6];
H = ctx->h[7];
R( 0, A, B, C, D, E, F, G, H);
R( 1, H, A, B, C, D, E, F, G);
R( 2, G, H, A, B, C, D, E, F);
R( 3, F, G, H, A, B, C, D, E);
R( 4, E, F, G, H, A, B, C, D);
R( 5, D, E, F, G, H, A, B, C);
R( 6, C, D, E, F, G, H, A, B);
R( 7, B, C, D, E, F, G, H, A);
R( 8, A, B, C, D, E, F, G, H);
R( 9, H, A, B, C, D, E, F, G);
R(10, G, H, A, B, C, D, E, F);
R(11, F, G, H, A, B, C, D, E);
R(12, E, F, G, H, A, B, C, D);
R(13, D, E, F, G, H, A, B, C);
R(14, C, D, E, F, G, H, A, B);
R(15, B, C, D, E, F, G, H, A);
R(16, A, B, C, D, E, F, G, H);
R(17, H, A, B, C, D, E, F, G);
R(18, G, H, A, B, C, D, E, F);
R(19, F, G, H, A, B, C, D, E);
R(20, E, F, G, H, A, B, C, D);
R(21, D, E, F, G, H, A, B, C);
R(22, C, D, E, F, G, H, A, B);
R(23, B, C, D, E, F, G, H, A);
R(24, A, B, C, D, E, F, G, H);
R(25, H, A, B, C, D, E, F, G);
R(26, G, H, A, B, C, D, E, F);
R(27, F, G, H, A, B, C, D, E);
R(28, E, F, G, H, A, B, C, D);
R(29, D, E, F, G, H, A, B, C);
R(30, C, D, E, F, G, H, A, B);
R(31, B, C, D, E, F, G, H, A);
R(32, A, B, C, D, E, F, G, H);
R(33, H, A, B, C, D, E, F, G);
R(34, G, H, A, B, C, D, E, F);
R(35, F, G, H, A, B, C, D, E);
R(36, E, F, G, H, A, B, C, D);
R(37, D, E, F, G, H, A, B, C);
R(38, C, D, E, F, G, H, A, B);
R(39, B, C, D, E, F, G, H, A);
R(40, A, B, C, D, E, F, G, H);
R(41, H, A, B, C, D, E, F, G);
R(42, G, H, A, B, C, D, E, F);
R(43, F, G, H, A, B, C, D, E);
R(44, E, F, G, H, A, B, C, D);
R(45, D, E, F, G, H, A, B, C);
R(46, C, D, E, F, G, H, A, B);
R(47, B, C, D, E, F, G, H, A);
R(48, A, B, C, D, E, F, G, H);
R(49, H, A, B, C, D, E, F, G);
R(50, G, H, A, B, C, D, E, F);
R(51, F, G, H, A, B, C, D, E);
R(52, E, F, G, H, A, B, C, D);
R(53, D, E, F, G, H, A, B, C);
R(54, C, D, E, F, G, H, A, B);
R(55, B, C, D, E, F, G, H, A);
R(56, A, B, C, D, E, F, G, H);
R(57, H, A, B, C, D, E, F, G);
R(58, G, H, A, B, C, D, E, F);
R(59, F, G, H, A, B, C, D, E);
R(60, E, F, G, H, A, B, C, D);
R(61, D, E, F, G, H, A, B, C);
R(62, C, D, E, F, G, H, A, B);
R(63, B, C, D, E, F, G, H, A);
R(64, A, B, C, D, E, F, G, H);
R(65, H, A, B, C, D, E, F, G);
R(66, G, H, A, B, C, D, E, F);
R(67, F, G, H, A, B, C, D, E);
R(68, E, F, G, H, A, B, C, D);
R(69, D, E, F, G, H, A, B, C);
R(70, C, D, E, F, G, H, A, B);
R(71, B, C, D, E, F, G, H, A);
R(72, A, B, C, D, E, F, G, H);
R(73, H, A, B, C, D, E, F, G);
R(74, G, H, A, B, C, D, E, F);
R(75, F, G, H, A, B, C, D, E);
R(76, E, F, G, H, A, B, C, D);
R(77, D, E, F, G, H, A, B, C);
R(78, C, D, E, F, G, H, A, B);
R(79, B, C, D, E, F, G, H, A);
ctx->h[0] += A;
ctx->h[1] += B;
ctx->h[2] += C;
ctx->h[3] += D;
ctx->h[4] += E;
ctx->h[5] += F;
ctx->h[6] += G;
ctx->h[7] += H;
}
void md4_block_data_order(uint32_t *state, const uint8_t *data, size_t num) {
uint32_t A, B, C, D, l;
uint32_t X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X12, X13, X14, X15;
A = state[0];
B = state[1];
C = state[2];
D = state[3];
for (; num--;) {
HOST_c2l(data, l);
X0 = l;
HOST_c2l(data, l);
X1 = l;
/* Round 0 */
R0(A, B, C, D, X0, 3, 0);
HOST_c2l(data, l);
X2 = l;
R0(D, A, B, C, X1, 7, 0);
HOST_c2l(data, l);
X3 = l;
R0(C, D, A, B, X2, 11, 0);
HOST_c2l(data, l);
X4 = l;
R0(B, C, D, A, X3, 19, 0);
HOST_c2l(data, l);
X5 = l;
R0(A, B, C, D, X4, 3, 0);
HOST_c2l(data, l);
X6 = l;
R0(D, A, B, C, X5, 7, 0);
HOST_c2l(data, l);
X7 = l;
R0(C, D, A, B, X6, 11, 0);
HOST_c2l(data, l);
X8 = l;
R0(B, C, D, A, X7, 19, 0);
HOST_c2l(data, l);
X9 = l;
R0(A, B, C, D, X8, 3, 0);
HOST_c2l(data, l);
X10 = l;
R0(D, A, B, C, X9, 7, 0);
HOST_c2l(data, l);
X11 = l;
R0(C, D, A, B, X10, 11, 0);
HOST_c2l(data, l);
X12 = l;
R0(B, C, D, A, X11, 19, 0);
HOST_c2l(data, l);
X13 = l;
R0(A, B, C, D, X12, 3, 0);
HOST_c2l(data, l);
X14 = l;
R0(D, A, B, C, X13, 7, 0);
HOST_c2l(data, l);
X15 = l;
R0(C, D, A, B, X14, 11, 0);
R0(B, C, D, A, X15, 19, 0);
/* Round 1 */
R1(A, B, C, D, X0, 3, 0x5A827999L);
R1(D, A, B, C, X4, 5, 0x5A827999L);
R1(C, D, A, B, X8, 9, 0x5A827999L);
R1(B, C, D, A, X12, 13, 0x5A827999L);
R1(A, B, C, D, X1, 3, 0x5A827999L);
R1(D, A, B, C, X5, 5, 0x5A827999L);
R1(C, D, A, B, X9, 9, 0x5A827999L);
R1(B, C, D, A, X13, 13, 0x5A827999L);
R1(A, B, C, D, X2, 3, 0x5A827999L);
R1(D, A, B, C, X6, 5, 0x5A827999L);
R1(C, D, A, B, X10, 9, 0x5A827999L);
R1(B, C, D, A, X14, 13, 0x5A827999L);
R1(A, B, C, D, X3, 3, 0x5A827999L);
R1(D, A, B, C, X7, 5, 0x5A827999L);
R1(C, D, A, B, X11, 9, 0x5A827999L);
R1(B, C, D, A, X15, 13, 0x5A827999L);
/* Round 2 */
R2(A, B, C, D, X0, 3, 0x6ED9EBA1L);
R2(D, A, B, C, X8, 9, 0x6ED9EBA1L);
R2(C, D, A, B, X4, 11, 0x6ED9EBA1L);
R2(B, C, D, A, X12, 15, 0x6ED9EBA1L);
R2(A, B, C, D, X2, 3, 0x6ED9EBA1L);
R2(D, A, B, C, X10, 9, 0x6ED9EBA1L);
R2(C, D, A, B, X6, 11, 0x6ED9EBA1L);
R2(B, C, D, A, X14, 15, 0x6ED9EBA1L);
R2(A, B, C, D, X1, 3, 0x6ED9EBA1L);
R2(D, A, B, C, X9, 9, 0x6ED9EBA1L);
R2(C, D, A, B, X5, 11, 0x6ED9EBA1L);
R2(B, C, D, A, X13, 15, 0x6ED9EBA1L);
R2(A, B, C, D, X3, 3, 0x6ED9EBA1L);
R2(D, A, B, C, X11, 9, 0x6ED9EBA1L);
R2(C, D, A, B, X7, 11, 0x6ED9EBA1L);
R2(B, C, D, A, X15, 15, 0x6ED9EBA1L);
A = state[0] += A;
B = state[1] += B;
C = state[2] += C;
D = state[3] += D;
}
}
void SHA1Transform(uint32 state[5], const unsigned char buffer[SHA1_BLOCK_LENGTH])
{
uint32 a, b, c, d, e;
typedef union {
unsigned char c[64];
uint32 l[16];
} CHAR64LONG16;
CHAR64LONG16 block[1]; /* use array to appear as a pointer */
memcpy(block, buffer, SHA1_BLOCK_LENGTH);
/* Copy context->state[] to working vars */
a = state[0];
b = state[1];
c = state[2];
d = state[3];
e = state[4];
/* 4 rounds of 20 operations each. Loop unrolled. */
R0(a,b,c,d,e, 0); R0(e,a,b,c,d, 1); R0(d,e,a,b,c, 2); R0(c,d,e,a,b, 3);
R0(b,c,d,e,a, 4); R0(a,b,c,d,e, 5); R0(e,a,b,c,d, 6); R0(d,e,a,b,c, 7);
R0(c,d,e,a,b, 8); R0(b,c,d,e,a, 9); R0(a,b,c,d,e,10); R0(e,a,b,c,d,11);
R0(d,e,a,b,c,12); R0(c,d,e,a,b,13); R0(b,c,d,e,a,14); R0(a,b,c,d,e,15);
R1(e,a,b,c,d,16); R1(d,e,a,b,c,17); R1(c,d,e,a,b,18); R1(b,c,d,e,a,19);
R2(a,b,c,d,e,20); R2(e,a,b,c,d,21); R2(d,e,a,b,c,22); R2(c,d,e,a,b,23);
R2(b,c,d,e,a,24); R2(a,b,c,d,e,25); R2(e,a,b,c,d,26); R2(d,e,a,b,c,27);
R2(c,d,e,a,b,28); R2(b,c,d,e,a,29); R2(a,b,c,d,e,30); R2(e,a,b,c,d,31);
R2(d,e,a,b,c,32); R2(c,d,e,a,b,33); R2(b,c,d,e,a,34); R2(a,b,c,d,e,35);
R2(e,a,b,c,d,36); R2(d,e,a,b,c,37); R2(c,d,e,a,b,38); R2(b,c,d,e,a,39);
R3(a,b,c,d,e,40); R3(e,a,b,c,d,41); R3(d,e,a,b,c,42); R3(c,d,e,a,b,43);
R3(b,c,d,e,a,44); R3(a,b,c,d,e,45); R3(e,a,b,c,d,46); R3(d,e,a,b,c,47);
R3(c,d,e,a,b,48); R3(b,c,d,e,a,49); R3(a,b,c,d,e,50); R3(e,a,b,c,d,51);
R3(d,e,a,b,c,52); R3(c,d,e,a,b,53); R3(b,c,d,e,a,54); R3(a,b,c,d,e,55);
R3(e,a,b,c,d,56); R3(d,e,a,b,c,57); R3(c,d,e,a,b,58); R3(b,c,d,e,a,59);
R4(a,b,c,d,e,60); R4(e,a,b,c,d,61); R4(d,e,a,b,c,62); R4(c,d,e,a,b,63);
R4(b,c,d,e,a,64); R4(a,b,c,d,e,65); R4(e,a,b,c,d,66); R4(d,e,a,b,c,67);
R4(c,d,e,a,b,68); R4(b,c,d,e,a,69); R4(a,b,c,d,e,70); R4(e,a,b,c,d,71);
R4(d,e,a,b,c,72); R4(c,d,e,a,b,73); R4(b,c,d,e,a,74); R4(a,b,c,d,e,75);
R4(e,a,b,c,d,76); R4(d,e,a,b,c,77); R4(c,d,e,a,b,78); R4(b,c,d,e,a,79);
/* Add the working vars back into context.state[] */
state[0] += a;
state[1] += b;
state[2] += c;
state[3] += d;
state[4] += e;
/* Wipe variables */
a = b = c = d = e = 0;
#ifdef SHA1HANDSOFF
memset(block, '\0', sizeof(block));
#endif
}
void md5_block_data_order (MD5_CTX *c, const void *data_, int num)
{
const unsigned char *data=data_;
register unsigned long A,B,C,D,l;
/*
* In case you wonder why A-D are declared as long and not
* as MD5_LONG. Doing so results in slight performance
* boost on LP64 architectures. The catch is we don't
* really care if 32 MSBs of a 64-bit register get polluted
* with eventual overflows as we *save* only 32 LSBs in
* *either* case. Now declaring 'em long excuses the compiler
* from keeping 32 MSBs zeroed resulting in 13% performance
* improvement under SPARC Solaris7/64 and 5% under AlphaLinux.
* Well, to be honest it should say that this *prevents*
* performance degradation.
*
* <*****@*****.**>
*/
#ifndef MD32_XARRAY
/* See comment in crypto/sha/sha_locl.h for details. */
unsigned long XX0, XX1, XX2, XX3, XX4, XX5, XX6, XX7,
XX8, XX9,XX10,XX11,XX12,XX13,XX14,XX15;
# define X(i) XX##i
#else
MD5_LONG XX[MD5_LBLOCK];
# define X(i) XX[i]
#endif
A=c->A;
B=c->B;
C=c->C;
D=c->D;
for (;num--;)
{
HOST_c2l(data,l); X( 0)=l; HOST_c2l(data,l); X( 1)=l;
/* Round 0 */
R0(A,B,C,D,X( 0), 7,0xd76aa478L); HOST_c2l(data,l); X( 2)=l;
R0(D,A,B,C,X( 1),12,0xe8c7b756L); HOST_c2l(data,l); X( 3)=l;
R0(C,D,A,B,X( 2),17,0x242070dbL); HOST_c2l(data,l); X( 4)=l;
R0(B,C,D,A,X( 3),22,0xc1bdceeeL); HOST_c2l(data,l); X( 5)=l;
R0(A,B,C,D,X( 4), 7,0xf57c0fafL); HOST_c2l(data,l); X( 6)=l;
R0(D,A,B,C,X( 5),12,0x4787c62aL); HOST_c2l(data,l); X( 7)=l;
R0(C,D,A,B,X( 6),17,0xa8304613L); HOST_c2l(data,l); X( 8)=l;
R0(B,C,D,A,X( 7),22,0xfd469501L); HOST_c2l(data,l); X( 9)=l;
R0(A,B,C,D,X( 8), 7,0x698098d8L); HOST_c2l(data,l); X(10)=l;
R0(D,A,B,C,X( 9),12,0x8b44f7afL); HOST_c2l(data,l); X(11)=l;
R0(C,D,A,B,X(10),17,0xffff5bb1L); HOST_c2l(data,l); X(12)=l;
R0(B,C,D,A,X(11),22,0x895cd7beL); HOST_c2l(data,l); X(13)=l;
R0(A,B,C,D,X(12), 7,0x6b901122L); HOST_c2l(data,l); X(14)=l;
R0(D,A,B,C,X(13),12,0xfd987193L); HOST_c2l(data,l); X(15)=l;
R0(C,D,A,B,X(14),17,0xa679438eL);
R0(B,C,D,A,X(15),22,0x49b40821L);
/* Round 1 */
R1(A,B,C,D,X( 1), 5,0xf61e2562L);
R1(D,A,B,C,X( 6), 9,0xc040b340L);
R1(C,D,A,B,X(11),14,0x265e5a51L);
R1(B,C,D,A,X( 0),20,0xe9b6c7aaL);
R1(A,B,C,D,X( 5), 5,0xd62f105dL);
R1(D,A,B,C,X(10), 9,0x02441453L);
R1(C,D,A,B,X(15),14,0xd8a1e681L);
R1(B,C,D,A,X( 4),20,0xe7d3fbc8L);
R1(A,B,C,D,X( 9), 5,0x21e1cde6L);
R1(D,A,B,C,X(14), 9,0xc33707d6L);
R1(C,D,A,B,X( 3),14,0xf4d50d87L);
R1(B,C,D,A,X( 8),20,0x455a14edL);
R1(A,B,C,D,X(13), 5,0xa9e3e905L);
R1(D,A,B,C,X( 2), 9,0xfcefa3f8L);
R1(C,D,A,B,X( 7),14,0x676f02d9L);
R1(B,C,D,A,X(12),20,0x8d2a4c8aL);
/* Round 2 */
R2(A,B,C,D,X( 5), 4,0xfffa3942L);
R2(D,A,B,C,X( 8),11,0x8771f681L);
R2(C,D,A,B,X(11),16,0x6d9d6122L);
R2(B,C,D,A,X(14),23,0xfde5380cL);
R2(A,B,C,D,X( 1), 4,0xa4beea44L);
R2(D,A,B,C,X( 4),11,0x4bdecfa9L);
R2(C,D,A,B,X( 7),16,0xf6bb4b60L);
R2(B,C,D,A,X(10),23,0xbebfbc70L);
R2(A,B,C,D,X(13), 4,0x289b7ec6L);
R2(D,A,B,C,X( 0),11,0xeaa127faL);
R2(C,D,A,B,X( 3),16,0xd4ef3085L);
R2(B,C,D,A,X( 6),23,0x04881d05L);
R2(A,B,C,D,X( 9), 4,0xd9d4d039L);
R2(D,A,B,C,X(12),11,0xe6db99e5L);
R2(C,D,A,B,X(15),16,0x1fa27cf8L);
R2(B,C,D,A,X( 2),23,0xc4ac5665L);
/* Round 3 */
R3(A,B,C,D,X( 0), 6,0xf4292244L);
R3(D,A,B,C,X( 7),10,0x432aff97L);
R3(C,D,A,B,X(14),15,0xab9423a7L);
R3(B,C,D,A,X( 5),21,0xfc93a039L);
R3(A,B,C,D,X(12), 6,0x655b59c3L);
R3(D,A,B,C,X( 3),10,0x8f0ccc92L);
R3(C,D,A,B,X(10),15,0xffeff47dL);
R3(B,C,D,A,X( 1),21,0x85845dd1L);
R3(A,B,C,D,X( 8), 6,0x6fa87e4fL);
R3(D,A,B,C,X(15),10,0xfe2ce6e0L);
R3(C,D,A,B,X( 6),15,0xa3014314L);
R3(B,C,D,A,X(13),21,0x4e0811a1L);
R3(A,B,C,D,X( 4), 6,0xf7537e82L);
R3(D,A,B,C,X(11),10,0xbd3af235L);
R3(C,D,A,B,X( 2),15,0x2ad7d2bbL);
R3(B,C,D,A,X( 9),21,0xeb86d391L);
A = c->A += A;
B = c->B += B;
C = c->C += C;
D = c->D += D;
}
}
void rtsmb_md4_block_data_order (RTSMB_MD4_CTX *c, const void *data_, int num)
{
const unsigned char *data;
register unsigned long A,B,C,D,l;
/*
* In case you wonder why A-D are declared as long and not
* as RTSMB_MD4_LONG. Doing so results in slight performance
* boost on LP64 architectures. The catch is we don't
* really care if 32 MSBs of a 64-bit register get polluted
* with eventual overflows as we *save* only 32 LSBs in
* *either* case. Now declaring 'em long excuses the compiler
* from keeping 32 MSBs zeroed resulting in 13% performance
* improvement under SPARC Solaris7/64 and 5% under AlphaLinux.
* Well, to be honest it should say that this *prevents*
* performance degradation.
*
* <*****@*****.**>
*/
#ifndef MD32_XARRAY
/* See comment in crypto/sha/sha_locl.h for details. */
unsigned long XX0, XX1, XX2, XX3, XX4, XX5, XX6, XX7,
XX8, XX9,XX10,XX11,XX12,XX13,XX14,XX15;
# define X(i) XX##i
#else
RTSMB_MD4_LONG XX[RTSMB_MD4_LBLOCK];
# define X(i) XX[i]
#endif
data =data_;
A=c->A;
B=c->B;
C=c->C;
D=c->D;
for (;num--;)
{
HOST_c2l(data,l); X( 0)=l; HOST_c2l(data,l); X( 1)=l;
/* Round 0 */
R0(A,B,C,D,X( 0), 3,0); HOST_c2l(data,l); X( 2)=l;
R0(D,A,B,C,X( 1), 7,0); HOST_c2l(data,l); X( 3)=l;
R0(C,D,A,B,X( 2),11,0); HOST_c2l(data,l); X( 4)=l;
R0(B,C,D,A,X( 3),19,0); HOST_c2l(data,l); X( 5)=l;
R0(A,B,C,D,X( 4), 3,0); HOST_c2l(data,l); X( 6)=l;
R0(D,A,B,C,X( 5), 7,0); HOST_c2l(data,l); X( 7)=l;
R0(C,D,A,B,X( 6),11,0); HOST_c2l(data,l); X( 8)=l;
R0(B,C,D,A,X( 7),19,0); HOST_c2l(data,l); X( 9)=l;
R0(A,B,C,D,X( 8), 3,0); HOST_c2l(data,l); X(10)=l;
R0(D,A,B,C,X( 9), 7,0); HOST_c2l(data,l); X(11)=l;
R0(C,D,A,B,X(10),11,0); HOST_c2l(data,l); X(12)=l;
R0(B,C,D,A,X(11),19,0); HOST_c2l(data,l); X(13)=l;
R0(A,B,C,D,X(12), 3,0); HOST_c2l(data,l); X(14)=l;
R0(D,A,B,C,X(13), 7,0); HOST_c2l(data,l); X(15)=l;
R0(C,D,A,B,X(14),11,0);
R0(B,C,D,A,X(15),19,0);
/* Round 1 */
R1(A,B,C,D,X( 0), 3,0x5A827999L);
R1(D,A,B,C,X( 4), 5,0x5A827999L);
R1(C,D,A,B,X( 8), 9,0x5A827999L);
R1(B,C,D,A,X(12),13,0x5A827999L);
R1(A,B,C,D,X( 1), 3,0x5A827999L);
R1(D,A,B,C,X( 5), 5,0x5A827999L);
R1(C,D,A,B,X( 9), 9,0x5A827999L);
R1(B,C,D,A,X(13),13,0x5A827999L);
R1(A,B,C,D,X( 2), 3,0x5A827999L);
R1(D,A,B,C,X( 6), 5,0x5A827999L);
R1(C,D,A,B,X(10), 9,0x5A827999L);
R1(B,C,D,A,X(14),13,0x5A827999L);
R1(A,B,C,D,X( 3), 3,0x5A827999L);
R1(D,A,B,C,X( 7), 5,0x5A827999L);
R1(C,D,A,B,X(11), 9,0x5A827999L);
R1(B,C,D,A,X(15),13,0x5A827999L);
/* Round 2 */
R2(A,B,C,D,X( 0), 3,0x6ED9EBA1L);
R2(D,A,B,C,X( 8), 9,0x6ED9EBA1L);
R2(C,D,A,B,X( 4),11,0x6ED9EBA1L);
R2(B,C,D,A,X(12),15,0x6ED9EBA1L);
R2(A,B,C,D,X( 2), 3,0x6ED9EBA1L);
R2(D,A,B,C,X(10), 9,0x6ED9EBA1L);
R2(C,D,A,B,X( 6),11,0x6ED9EBA1L);
R2(B,C,D,A,X(14),15,0x6ED9EBA1L);
R2(A,B,C,D,X( 1), 3,0x6ED9EBA1L);
R2(D,A,B,C,X( 9), 9,0x6ED9EBA1L);
R2(C,D,A,B,X( 5),11,0x6ED9EBA1L);
R2(B,C,D,A,X(13),15,0x6ED9EBA1L);
R2(A,B,C,D,X( 3), 3,0x6ED9EBA1L);
R2(D,A,B,C,X(11), 9,0x6ED9EBA1L);
R2(C,D,A,B,X( 7),11,0x6ED9EBA1L);
R2(B,C,D,A,X(15),15,0x6ED9EBA1L);
A = c->A += A;
B = c->B += B;
C = c->C += C;
D = c->D += D;
}
}
void md4_block_data_order (MD4_CTX *c, const void *data_, size_t num)
{
const unsigned char *data=data_;
register unsigned MD32_REG_T A,B,C,D,l;
#ifndef MD32_XARRAY
/* See comment in crypto/sha/sha_locl.h for details. */
unsigned MD32_REG_T XX0, XX1, XX2, XX3, XX4, XX5, XX6, XX7,
XX8, XX9,XX10,XX11,XX12,XX13,XX14,XX15;
# define X(i) XX##i
#else
MD4_LONG XX[MD4_LBLOCK];
# define X(i) XX[i]
#endif
A=c->A;
B=c->B;
C=c->C;
D=c->D;
for (;num--;)
{
HOST_c2l(data,l); X( 0)=l; HOST_c2l(data,l); X( 1)=l;
/* Round 0 */
R0(A,B,C,D,X( 0), 3,0); HOST_c2l(data,l); X( 2)=l;
R0(D,A,B,C,X( 1), 7,0); HOST_c2l(data,l); X( 3)=l;
R0(C,D,A,B,X( 2),11,0); HOST_c2l(data,l); X( 4)=l;
R0(B,C,D,A,X( 3),19,0); HOST_c2l(data,l); X( 5)=l;
R0(A,B,C,D,X( 4), 3,0); HOST_c2l(data,l); X( 6)=l;
R0(D,A,B,C,X( 5), 7,0); HOST_c2l(data,l); X( 7)=l;
R0(C,D,A,B,X( 6),11,0); HOST_c2l(data,l); X( 8)=l;
R0(B,C,D,A,X( 7),19,0); HOST_c2l(data,l); X( 9)=l;
R0(A,B,C,D,X( 8), 3,0); HOST_c2l(data,l); X(10)=l;
R0(D,A,B,C,X( 9), 7,0); HOST_c2l(data,l); X(11)=l;
R0(C,D,A,B,X(10),11,0); HOST_c2l(data,l); X(12)=l;
R0(B,C,D,A,X(11),19,0); HOST_c2l(data,l); X(13)=l;
R0(A,B,C,D,X(12), 3,0); HOST_c2l(data,l); X(14)=l;
R0(D,A,B,C,X(13), 7,0); HOST_c2l(data,l); X(15)=l;
R0(C,D,A,B,X(14),11,0);
R0(B,C,D,A,X(15),19,0);
/* Round 1 */
R1(A,B,C,D,X( 0), 3,0x5A827999L);
R1(D,A,B,C,X( 4), 5,0x5A827999L);
R1(C,D,A,B,X( 8), 9,0x5A827999L);
R1(B,C,D,A,X(12),13,0x5A827999L);
R1(A,B,C,D,X( 1), 3,0x5A827999L);
R1(D,A,B,C,X( 5), 5,0x5A827999L);
R1(C,D,A,B,X( 9), 9,0x5A827999L);
R1(B,C,D,A,X(13),13,0x5A827999L);
R1(A,B,C,D,X( 2), 3,0x5A827999L);
R1(D,A,B,C,X( 6), 5,0x5A827999L);
R1(C,D,A,B,X(10), 9,0x5A827999L);
R1(B,C,D,A,X(14),13,0x5A827999L);
R1(A,B,C,D,X( 3), 3,0x5A827999L);
R1(D,A,B,C,X( 7), 5,0x5A827999L);
R1(C,D,A,B,X(11), 9,0x5A827999L);
R1(B,C,D,A,X(15),13,0x5A827999L);
/* Round 2 */
R2(A,B,C,D,X( 0), 3,0x6ED9EBA1L);
R2(D,A,B,C,X( 8), 9,0x6ED9EBA1L);
R2(C,D,A,B,X( 4),11,0x6ED9EBA1L);
R2(B,C,D,A,X(12),15,0x6ED9EBA1L);
R2(A,B,C,D,X( 2), 3,0x6ED9EBA1L);
R2(D,A,B,C,X(10), 9,0x6ED9EBA1L);
R2(C,D,A,B,X( 6),11,0x6ED9EBA1L);
R2(B,C,D,A,X(14),15,0x6ED9EBA1L);
R2(A,B,C,D,X( 1), 3,0x6ED9EBA1L);
R2(D,A,B,C,X( 9), 9,0x6ED9EBA1L);
R2(C,D,A,B,X( 5),11,0x6ED9EBA1L);
R2(B,C,D,A,X(13),15,0x6ED9EBA1L);
R2(A,B,C,D,X( 3), 3,0x6ED9EBA1L);
R2(D,A,B,C,X(11), 9,0x6ED9EBA1L);
R2(C,D,A,B,X( 7),11,0x6ED9EBA1L);
R2(B,C,D,A,X(15),15,0x6ED9EBA1L);
A = c->A += A;
B = c->B += B;
C = c->C += C;
D = c->D += D;
}
}
void rtsmb_md4_block_host_order (RTSMB_MD4_CTX *c, const void *data, int num)
{
const RTSMB_MD4_LONG *X;
register unsigned long A,B,C,D;
/*
* In case you wonder why A-D are declared as long and not
* as RTSMB_MD4_LONG. Doing so results in slight performance
* boost on LP64 architectures. The catch is we don't
* really care if 32 MSBs of a 64-bit register get polluted
* with eventual overflows as we *save* only 32 LSBs in
* *either* case. Now declaring 'em long excuses the compiler
* from keeping 32 MSBs zeroed resulting in 13% performance
* improvement under SPARC Solaris7/64 and 5% under AlphaLinux.
* Well, to be honest it should say that this *prevents*
* performance degradation.
*
* <*****@*****.**>
*/
X=data;
A=c->A;
B=c->B;
C=c->C;
D=c->D;
for (;num--;X+=HASH_LBLOCK)
{
/* Round 0 */
R0(A,B,C,D,X[ 0], 3,0);
R0(D,A,B,C,X[ 1], 7,0);
R0(C,D,A,B,X[ 2],11,0);
R0(B,C,D,A,X[ 3],19,0);
R0(A,B,C,D,X[ 4], 3,0);
R0(D,A,B,C,X[ 5], 7,0);
R0(C,D,A,B,X[ 6],11,0);
R0(B,C,D,A,X[ 7],19,0);
R0(A,B,C,D,X[ 8], 3,0);
R0(D,A,B,C,X[ 9], 7,0);
R0(C,D,A,B,X[10],11,0);
R0(B,C,D,A,X[11],19,0);
R0(A,B,C,D,X[12], 3,0);
R0(D,A,B,C,X[13], 7,0);
R0(C,D,A,B,X[14],11,0);
R0(B,C,D,A,X[15],19,0);
/* Round 1 */
R1(A,B,C,D,X[ 0], 3,0x5A827999L);
R1(D,A,B,C,X[ 4], 5,0x5A827999L);
R1(C,D,A,B,X[ 8], 9,0x5A827999L);
R1(B,C,D,A,X[12],13,0x5A827999L);
R1(A,B,C,D,X[ 1], 3,0x5A827999L);
R1(D,A,B,C,X[ 5], 5,0x5A827999L);
R1(C,D,A,B,X[ 9], 9,0x5A827999L);
R1(B,C,D,A,X[13],13,0x5A827999L);
R1(A,B,C,D,X[ 2], 3,0x5A827999L);
R1(D,A,B,C,X[ 6], 5,0x5A827999L);
R1(C,D,A,B,X[10], 9,0x5A827999L);
R1(B,C,D,A,X[14],13,0x5A827999L);
R1(A,B,C,D,X[ 3], 3,0x5A827999L);
R1(D,A,B,C,X[ 7], 5,0x5A827999L);
R1(C,D,A,B,X[11], 9,0x5A827999L);
R1(B,C,D,A,X[15],13,0x5A827999L);
/* Round 2 */
R2(A,B,C,D,X[ 0], 3,0x6ED9EBA1);
R2(D,A,B,C,X[ 8], 9,0x6ED9EBA1);
R2(C,D,A,B,X[ 4],11,0x6ED9EBA1);
R2(B,C,D,A,X[12],15,0x6ED9EBA1);
R2(A,B,C,D,X[ 2], 3,0x6ED9EBA1);
R2(D,A,B,C,X[10], 9,0x6ED9EBA1);
R2(C,D,A,B,X[ 6],11,0x6ED9EBA1);
R2(B,C,D,A,X[14],15,0x6ED9EBA1);
R2(A,B,C,D,X[ 1], 3,0x6ED9EBA1);
R2(D,A,B,C,X[ 9], 9,0x6ED9EBA1);
R2(C,D,A,B,X[ 5],11,0x6ED9EBA1);
R2(B,C,D,A,X[13],15,0x6ED9EBA1);
R2(A,B,C,D,X[ 3], 3,0x6ED9EBA1);
R2(D,A,B,C,X[11], 9,0x6ED9EBA1);
R2(C,D,A,B,X[ 7],11,0x6ED9EBA1);
R2(B,C,D,A,X[15],15,0x6ED9EBA1);
A = c->A += A;
B = c->B += B;
C = c->C += C;
D = c->D += D;
}
}
//MD4 Block data order setting
void __fastcall MD4_BlockDataOrder(
MD4_CTX *c,
const void *data_,
size_t num)
{
const unsigned char *data = (const unsigned char *)data_;
register uint32_t A = 0, B = 0, C = 0, D = 0, l = 0;
uint32_t XX0 = 0, XX1 = 0, XX2 = 0, XX3 = 0, XX4 = 0, XX5 = 0, XX6 = 0, XX7 = 0, XX8 = 0, XX9 = 0, XX10 = 0, XX11 = 0, XX12 = 0, XX13 = 0, XX14 = 0, XX15 = 0;
#define X(i) XX ## i
A = c->A;
B = c->B;
C = c->C;
D = c->D;
for (;num--;)
{
(void)HOST_c2l(data, l); X(0) = l;
(void)HOST_c2l(data, l); X(1) = l;
//Round 0
R0(A, B, C, D, X(0), 3, 0); (void)HOST_c2l(data, l); X(2) = l;
R0(D, A, B, C, X(1), 7, 0); (void)HOST_c2l(data, l); X(3) = l;
R0(C, D, A, B, X(2), 11, 0); (void)HOST_c2l(data, l); X(4) = l;
R0(B, C, D, A, X(3), 19, 0); (void)HOST_c2l(data, l); X(5) = l;
R0(A, B, C, D, X(4), 3, 0); (void)HOST_c2l(data, l); X(6) = l;
R0(D, A, B, C, X(5), 7, 0); (void)HOST_c2l(data, l); X(7) = l;
R0(C, D, A, B, X(6), 11, 0); (void)HOST_c2l(data, l); X(8) = l;
R0(B, C, D, A, X(7), 19, 0); (void)HOST_c2l(data, l); X(9) = l;
R0(A, B, C, D, X(8), 3, 0); (void)HOST_c2l(data, l); X(10) = l;
R0(D, A, B, C, X(9), 7, 0); (void)HOST_c2l(data, l); X(11) = l;
R0(C, D, A, B, X(10), 11, 0); (void)HOST_c2l(data, l); X(12) = l;
R0(B, C, D, A, X(11), 19, 0); (void)HOST_c2l(data, l); X(13) = l;
R0(A, B, C, D, X(12), 3, 0); (void)HOST_c2l(data, l); X(14) = l;
R0(D, A, B, C, X(13), 7, 0); (void)HOST_c2l(data, l); X(15) = l;
R0(C, D, A, B, X(14), 11, 0);
R0(B, C, D, A, X(15), 19, 0);
//Round 1
R1(A, B, C, D, X(0), 3, 0x5A827999L);
R1(D, A, B, C, X(4), 5, 0x5A827999L);
R1(C, D, A, B, X(8), 9, 0x5A827999L);
R1(B, C, D, A, X(12), 13, 0x5A827999L);
R1(A, B, C, D, X(1), 3, 0x5A827999L);
R1(D, A, B, C, X(5), 5, 0x5A827999L);
R1(C, D, A, B, X(9), 9, 0x5A827999L);
R1(B, C, D, A, X(13), 13, 0x5A827999L);
R1(A, B, C, D, X(2), 3, 0x5A827999L);
R1(D, A, B, C, X(6), 5, 0x5A827999L);
R1(C, D, A, B, X(10), 9, 0x5A827999L);
R1(B, C, D, A, X(14), 13, 0x5A827999L);
R1(A, B, C, D, X(3), 3, 0x5A827999L);
R1(D, A, B, C, X(7), 5, 0x5A827999L);
R1(C, D, A, B, X(11), 9, 0x5A827999L);
R1(B, C, D, A, X(15), 13, 0x5A827999L);
//Round 2
R2(A, B, C, D, X(0), 3, 0x6ED9EBA1L);
R2(D, A, B, C, X(8), 9, 0x6ED9EBA1L);
R2(C, D, A, B, X(4), 11, 0x6ED9EBA1L);
R2(B, C, D, A, X(12), 15, 0x6ED9EBA1L);
R2(A, B, C, D, X(2), 3, 0x6ED9EBA1L);
R2(D, A, B, C, X(10), 9, 0x6ED9EBA1L);
R2(C, D, A, B, X(6), 11, 0x6ED9EBA1L);
R2(B, C, D, A, X(14), 15, 0x6ED9EBA1L);
R2(A, B, C, D, X(1), 3, 0x6ED9EBA1L);
R2(D, A, B, C, X(9), 9, 0x6ED9EBA1L);
R2(C, D, A, B, X(5), 11, 0x6ED9EBA1L);
R2(B, C, D, A, X(13), 15, 0x6ED9EBA1L);
R2(A, B, C, D, X(3), 3, 0x6ED9EBA1L);
R2(D, A, B, C, X(11), 9, 0x6ED9EBA1L);
R2(C, D, A, B, X(7), 11, 0x6ED9EBA1L);
R2(B, C, D, A, X(15), 15, 0x6ED9EBA1L);
A = c->A += A;
B = c->B += B;
C = c->C += C;
D = c->D += D;
}
return;
}
/* Hash a single 512-bit block. This is the core of the algorithm. */
static void SHATransform(ULONG State[5], UCHAR Buffer[64])
{
ULONG a, b, c, d, e;
ULONG *Block;
Block = (ULONG*)Buffer;
/* Copy Context->State[] to working variables */
a = State[0];
b = State[1];
c = State[2];
d = State[3];
e = State[4];
/* 4 rounds of 20 operations each. Loop unrolled. */
R0(a,b,c,d,e, 0); R0(e,a,b,c,d, 1); R0(d,e,a,b,c, 2); R0(c,d,e,a,b, 3);
R0(b,c,d,e,a, 4); R0(a,b,c,d,e, 5); R0(e,a,b,c,d, 6); R0(d,e,a,b,c, 7);
R0(c,d,e,a,b, 8); R0(b,c,d,e,a, 9); R0(a,b,c,d,e,10); R0(e,a,b,c,d,11);
R0(d,e,a,b,c,12); R0(c,d,e,a,b,13); R0(b,c,d,e,a,14); R0(a,b,c,d,e,15);
R1(e,a,b,c,d,16); R1(d,e,a,b,c,17); R1(c,d,e,a,b,18); R1(b,c,d,e,a,19);
R2(a,b,c,d,e,20); R2(e,a,b,c,d,21); R2(d,e,a,b,c,22); R2(c,d,e,a,b,23);
R2(b,c,d,e,a,24); R2(a,b,c,d,e,25); R2(e,a,b,c,d,26); R2(d,e,a,b,c,27);
R2(c,d,e,a,b,28); R2(b,c,d,e,a,29); R2(a,b,c,d,e,30); R2(e,a,b,c,d,31);
R2(d,e,a,b,c,32); R2(c,d,e,a,b,33); R2(b,c,d,e,a,34); R2(a,b,c,d,e,35);
R2(e,a,b,c,d,36); R2(d,e,a,b,c,37); R2(c,d,e,a,b,38); R2(b,c,d,e,a,39);
R3(a,b,c,d,e,40); R3(e,a,b,c,d,41); R3(d,e,a,b,c,42); R3(c,d,e,a,b,43);
R3(b,c,d,e,a,44); R3(a,b,c,d,e,45); R3(e,a,b,c,d,46); R3(d,e,a,b,c,47);
R3(c,d,e,a,b,48); R3(b,c,d,e,a,49); R3(a,b,c,d,e,50); R3(e,a,b,c,d,51);
R3(d,e,a,b,c,52); R3(c,d,e,a,b,53); R3(b,c,d,e,a,54); R3(a,b,c,d,e,55);
R3(e,a,b,c,d,56); R3(d,e,a,b,c,57); R3(c,d,e,a,b,58); R3(b,c,d,e,a,59);
R4(a,b,c,d,e,60); R4(e,a,b,c,d,61); R4(d,e,a,b,c,62); R4(c,d,e,a,b,63);
R4(b,c,d,e,a,64); R4(a,b,c,d,e,65); R4(e,a,b,c,d,66); R4(d,e,a,b,c,67);
R4(c,d,e,a,b,68); R4(b,c,d,e,a,69); R4(a,b,c,d,e,70); R4(e,a,b,c,d,71);
R4(d,e,a,b,c,72); R4(c,d,e,a,b,73); R4(b,c,d,e,a,74); R4(a,b,c,d,e,75);
R4(e,a,b,c,d,76); R4(d,e,a,b,c,77); R4(c,d,e,a,b,78); R4(b,c,d,e,a,79);
/* Add the working variables back into Context->State[] */
State[0] += a;
State[1] += b;
State[2] += c;
State[3] += d;
State[4] += e;
/* Wipe variables */
a = b = c = d = e = 0;
}
