1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
|
#include "rar.hpp"
#include "sha256.hpp"
static const uint32 K[64] =
{
0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5,
0x3956c25b, 0x59f111f1, 0x923f82a4, 0xab1c5ed5,
0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3,
0x72be5d74, 0x80deb1fe, 0x9bdc06a7, 0xc19bf174,
0xe49b69c1, 0xefbe4786, 0x0fc19dc6, 0x240ca1cc,
0x2de92c6f, 0x4a7484aa, 0x5cb0a9dc, 0x76f988da,
0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7,
0xc6e00bf3, 0xd5a79147, 0x06ca6351, 0x14292967,
0x27b70a85, 0x2e1b2138, 0x4d2c6dfc, 0x53380d13,
0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85,
0xa2bfe8a1, 0xa81a664b, 0xc24b8b70, 0xc76c51a3,
0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070,
0x19a4c116, 0x1e376c08, 0x2748774c, 0x34b0bcb5,
0x391c0cb3, 0x4ed8aa4a, 0x5b9cca4f, 0x682e6ff3,
0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208,
0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2
};
#define ROTL(x, n) (((x) << (n)) | ((x) >> (32 - (n))))
#define ROTR(x, n) (((x) >> (n)) | ((x) << (32 - (n))))
// SHA-256 functions. We could optimize Ch and Maj a little,
// but with no visible speed benefit.
#define Ch(x, y, z) ((x & y) ^ (~x & z))
#define Maj(x, y, z) ((x & y) ^ (x & z) ^ (y & z))
// Sigma functions.
#define Sg0(x) (ROTR(x, 2) ^ ROTR(x,13) ^ ROTR(x, 22))
#define Sg1(x) (ROTR(x, 6) ^ ROTR(x,11) ^ ROTR(x, 25))
#define sg0(x) (ROTR(x, 7) ^ ROTR(x,18) ^ (x >> 3))
#define sg1(x) (ROTR(x,17) ^ ROTR(x,19) ^ (x >> 10))
void sha256_init(sha256_context *ctx)
{
ctx->H[0] = 0x6a09e667; // Set the initial hash value.
ctx->H[1] = 0xbb67ae85;
ctx->H[2] = 0x3c6ef372;
ctx->H[3] = 0xa54ff53a;
ctx->H[4] = 0x510e527f;
ctx->H[5] = 0x9b05688c;
ctx->H[6] = 0x1f83d9ab;
ctx->H[7] = 0x5be0cd19;
ctx->Count = 0; // Processed data counter.
}
inline uint32 b2i(const byte *b) // Big endian bytes to integer.
{
#if defined(_MSC_VER)/* && defined(LITTLE_ENDIAN)*/
return _byteswap_ulong(*(uint32 *)b);
#elif (__GNUC__ > 3) && (__GNUC_MINOR__ > 2)
return __builtin_bswap32(*(uint32 *)b);
#else
return uint32(b[0]<<24) | uint32(b[1]<<16) | uint32(b[2]<<8) | b[3];
#endif
}
static void sha256_transform(sha256_context *ctx)
{
uint32 W[64]; // Words of message schedule.
uint32 v[8]; // FIPS a, b, c, d, e, f, g, h working variables.
if (ctx == NULL) // Clean variables and return.
{
cleandata(v,sizeof(v));
cleandata(W,sizeof(W));
return;
}
// Prepare message schedule. Loop unrolling provides some small gain here.
W[0] = b2i(ctx->Data + 0 * 4 ); W[1] = b2i(ctx->Data + 1 * 4 );
W[2] = b2i(ctx->Data + 2 * 4 ); W[3] = b2i(ctx->Data + 3 * 4 );
W[4] = b2i(ctx->Data + 4 * 4 ); W[5] = b2i(ctx->Data + 5 * 4 );
W[6] = b2i(ctx->Data + 6 * 4 ); W[7] = b2i(ctx->Data + 7 * 4 );
W[8] = b2i(ctx->Data + 8 * 4 ); W[9] = b2i(ctx->Data + 9 * 4 );
W[10] = b2i(ctx->Data + 10 * 4 ); W[11] = b2i(ctx->Data + 11 * 4 );
W[12] = b2i(ctx->Data + 12 * 4 ); W[13] = b2i(ctx->Data + 13 * 4 );
W[14] = b2i(ctx->Data + 14 * 4 ); W[15] = b2i(ctx->Data + 15 * 4 );
for (uint I = 16; I < 64; I++)
W[I] = sg1(W[I-2]) + W[I-7] + sg0(W[I-15]) + W[I-16];
uint32 *H=ctx->H;
v[0]=H[0]; v[1]=H[1]; v[2]=H[2]; v[3]=H[3];
v[4]=H[4]; v[5]=H[5]; v[6]=H[6]; v[7]=H[7];
// MSVC -O2 partially unrolls this loop automatically.
for (uint I = 0; I < 64; I++)
{
uint T1 = v[7] + Sg1(v[4]) + Ch(v[4], v[5], v[6]) + K[I] + W[I];
// It is possible to eliminate variable copying if we unroll loop
// and rename variables every time. But my test did not show any speed
// gain on i7 for such full or partial unrolling.
v[7] = v[6];
v[6] = v[5];
v[5] = v[4];
v[4] = v[3] + T1;
// It works a little faster when moved here from beginning of loop.
uint T2 = Sg0(v[0]) + Maj(v[0], v[1], v[2]);
v[3] = v[2];
v[2] = v[1];
v[1] = v[0];
v[0] = T1 + T2;
}
H[0]+=v[0]; H[1]+=v[1]; H[2]+=v[2]; H[3]+=v[3];
H[4]+=v[4]; H[5]+=v[5]; H[6]+=v[6]; H[7]+=v[7];
}
void sha256_process(sha256_context *ctx, const void *Data, size_t Size)
{
const byte *Src=(const byte *)Data;
size_t BufPos = (uint)ctx->Count & 0x3f;
ctx->Count+=Size;
while (Size > 0)
{
size_t BufSpace=sizeof(ctx->Buffer)-BufPos;
size_t CopySize=Size>BufSpace ? BufSpace:Size;
if (CopySize == 64)
ctx->Data=Src; // Point to source data instead of copying it to buffer.
else
{
ctx->Data=ctx->Buffer;
memcpy(ctx->Buffer+BufPos,Src,CopySize);
}
Src+=CopySize;
BufPos+=CopySize;
Size-=CopySize;
if (BufPos == 64)
{
BufPos = 0;
sha256_transform(ctx);
}
}
sha256_transform(NULL);
}
void sha256_done(sha256_context *ctx, byte *Digest)
{
ctx->Data = ctx->Buffer;
uint64 BitLength = ctx->Count * 8;
uint BufPos = (uint)ctx->Count & 0x3f;
ctx->Buffer[BufPos++] = 0x80; // Padding the message with "1" bit.
while (BufPos != 56) // We need 56 bytes block followed by 8 byte length.
{
BufPos &= 0x3f;
if (BufPos == 0)
sha256_transform(ctx);
ctx->Buffer[BufPos++] = 0;
}
for (uint i = 0; i < 8; i++) // Store bit length of entire message.
{
ctx->Buffer[BufPos++] = (byte)(BitLength >> 56);
BitLength <<= 8;
}
sha256_transform(ctx);
for (uint i = 0; i < 32; i++)
Digest[i] = byte(ctx->H[i / 4] >> ((3 - i % 4) * 8));
sha256_init(ctx);
sha256_transform(NULL);
cleandata(ctx->Buffer, sizeof(ctx->Buffer));
}
|
