1 /* Aes.c -- AES encryption / decryption
2 2017-01-24 : Igor Pavlov : Public domain */
3
4 #include "Precomp.h"
5
6 #include "Aes.h"
7 #include "CpuArch.h"
8
9 static UInt32 T[256 * 4];
10 static const Byte Sbox[256] = {
11 0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5, 0x30, 0x01, 0x67, 0x2b, 0xfe, 0xd7, 0xab, 0x76,
12 0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0, 0xad, 0xd4, 0xa2, 0xaf, 0x9c, 0xa4, 0x72, 0xc0,
13 0xb7, 0xfd, 0x93, 0x26, 0x36, 0x3f, 0xf7, 0xcc, 0x34, 0xa5, 0xe5, 0xf1, 0x71, 0xd8, 0x31, 0x15,
14 0x04, 0xc7, 0x23, 0xc3, 0x18, 0x96, 0x05, 0x9a, 0x07, 0x12, 0x80, 0xe2, 0xeb, 0x27, 0xb2, 0x75,
15 0x09, 0x83, 0x2c, 0x1a, 0x1b, 0x6e, 0x5a, 0xa0, 0x52, 0x3b, 0xd6, 0xb3, 0x29, 0xe3, 0x2f, 0x84,
16 0x53, 0xd1, 0x00, 0xed, 0x20, 0xfc, 0xb1, 0x5b, 0x6a, 0xcb, 0xbe, 0x39, 0x4a, 0x4c, 0x58, 0xcf,
17 0xd0, 0xef, 0xaa, 0xfb, 0x43, 0x4d, 0x33, 0x85, 0x45, 0xf9, 0x02, 0x7f, 0x50, 0x3c, 0x9f, 0xa8,
18 0x51, 0xa3, 0x40, 0x8f, 0x92, 0x9d, 0x38, 0xf5, 0xbc, 0xb6, 0xda, 0x21, 0x10, 0xff, 0xf3, 0xd2,
19 0xcd, 0x0c, 0x13, 0xec, 0x5f, 0x97, 0x44, 0x17, 0xc4, 0xa7, 0x7e, 0x3d, 0x64, 0x5d, 0x19, 0x73,
20 0x60, 0x81, 0x4f, 0xdc, 0x22, 0x2a, 0x90, 0x88, 0x46, 0xee, 0xb8, 0x14, 0xde, 0x5e, 0x0b, 0xdb,
21 0xe0, 0x32, 0x3a, 0x0a, 0x49, 0x06, 0x24, 0x5c, 0xc2, 0xd3, 0xac, 0x62, 0x91, 0x95, 0xe4, 0x79,
22 0xe7, 0xc8, 0x37, 0x6d, 0x8d, 0xd5, 0x4e, 0xa9, 0x6c, 0x56, 0xf4, 0xea, 0x65, 0x7a, 0xae, 0x08,
23 0xba, 0x78, 0x25, 0x2e, 0x1c, 0xa6, 0xb4, 0xc6, 0xe8, 0xdd, 0x74, 0x1f, 0x4b, 0xbd, 0x8b, 0x8a,
24 0x70, 0x3e, 0xb5, 0x66, 0x48, 0x03, 0xf6, 0x0e, 0x61, 0x35, 0x57, 0xb9, 0x86, 0xc1, 0x1d, 0x9e,
25 0xe1, 0xf8, 0x98, 0x11, 0x69, 0xd9, 0x8e, 0x94, 0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf,
26 0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68, 0x41, 0x99, 0x2d, 0x0f, 0xb0, 0x54, 0xbb, 0x16};
27
28 void MY_FAST_CALL AesCbc_Encode(UInt32 *ivAes, Byte *data, size_t numBlocks);
29 void MY_FAST_CALL AesCbc_Decode(UInt32 *ivAes, Byte *data, size_t numBlocks);
30 void MY_FAST_CALL AesCtr_Code(UInt32 *ivAes, Byte *data, size_t numBlocks);
31
32 void MY_FAST_CALL AesCbc_Encode_Intel(UInt32 *ivAes, Byte *data, size_t numBlocks);
33 void MY_FAST_CALL AesCbc_Decode_Intel(UInt32 *ivAes, Byte *data, size_t numBlocks);
34 void MY_FAST_CALL AesCtr_Code_Intel(UInt32 *ivAes, Byte *data, size_t numBlocks);
35
36 AES_CODE_FUNC g_AesCbc_Encode;
37 AES_CODE_FUNC g_AesCbc_Decode;
38 AES_CODE_FUNC g_AesCtr_Code;
39
40 static UInt32 D[256 * 4];
41 static Byte InvS[256];
42
43 static const Byte Rcon[11] = { 0x00, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36 };
44
45 #define xtime(x) ((((x) << 1) ^ (((x) & 0x80) != 0 ? 0x1B : 0)) & 0xFF)
46
47 #define Ui32(a0, a1, a2, a3) ((UInt32)(a0) | ((UInt32)(a1) << 8) | ((UInt32)(a2) << 16) | ((UInt32)(a3) << 24))
48
49 #define gb0(x) ( (x) & 0xFF)
50 #define gb1(x) (((x) >> ( 8)) & 0xFF)
51 #define gb2(x) (((x) >> (16)) & 0xFF)
52 #define gb3(x) (((x) >> (24)))
53
54 #define gb(n, x) gb ## n(x)
55
56 #define TT(x) (T + (x << 8))
57 #define DD(x) (D + (x << 8))
58
59
AesGenTables(void)60 void AesGenTables(void)
61 {
62 unsigned i;
63 for (i = 0; i < 256; i++)
64 InvS[Sbox[i]] = (Byte)i;
65
66 for (i = 0; i < 256; i++)
67 {
68 {
69 UInt32 a1 = Sbox[i];
70 UInt32 a2 = xtime(a1);
71 UInt32 a3 = a2 ^ a1;
72 TT(0)[i] = Ui32(a2, a1, a1, a3);
73 TT(1)[i] = Ui32(a3, a2, a1, a1);
74 TT(2)[i] = Ui32(a1, a3, a2, a1);
75 TT(3)[i] = Ui32(a1, a1, a3, a2);
76 }
77 {
78 UInt32 a1 = InvS[i];
79 UInt32 a2 = xtime(a1);
80 UInt32 a4 = xtime(a2);
81 UInt32 a8 = xtime(a4);
82 UInt32 a9 = a8 ^ a1;
83 UInt32 aB = a8 ^ a2 ^ a1;
84 UInt32 aD = a8 ^ a4 ^ a1;
85 UInt32 aE = a8 ^ a4 ^ a2;
86 DD(0)[i] = Ui32(aE, a9, aD, aB);
87 DD(1)[i] = Ui32(aB, aE, a9, aD);
88 DD(2)[i] = Ui32(aD, aB, aE, a9);
89 DD(3)[i] = Ui32(a9, aD, aB, aE);
90 }
91 }
92
93 g_AesCbc_Encode = AesCbc_Encode;
94 g_AesCbc_Decode = AesCbc_Decode;
95 g_AesCtr_Code = AesCtr_Code;
96
97 #ifdef MY_CPU_X86_OR_AMD64
98 if (CPU_Is_Aes_Supported())
99 {
100 g_AesCbc_Encode = AesCbc_Encode_Intel;
101 g_AesCbc_Decode = AesCbc_Decode_Intel;
102 g_AesCtr_Code = AesCtr_Code_Intel;
103 }
104 #endif
105 }
106
107
108 #define HT(i, x, s) TT(x)[gb(x, s[(i + x) & 3])]
109
110 #define HT4(m, i, s, p) m[i] = \
111 HT(i, 0, s) ^ \
112 HT(i, 1, s) ^ \
113 HT(i, 2, s) ^ \
114 HT(i, 3, s) ^ w[p + i]
115
116 #define HT16(m, s, p) \
117 HT4(m, 0, s, p); \
118 HT4(m, 1, s, p); \
119 HT4(m, 2, s, p); \
120 HT4(m, 3, s, p); \
121
122 #define FT(i, x) Sbox[gb(x, m[(i + x) & 3])]
123 #define FT4(i) dest[i] = Ui32(FT(i, 0), FT(i, 1), FT(i, 2), FT(i, 3)) ^ w[i];
124
125
126 #define HD(i, x, s) DD(x)[gb(x, s[(i - x) & 3])]
127
128 #define HD4(m, i, s, p) m[i] = \
129 HD(i, 0, s) ^ \
130 HD(i, 1, s) ^ \
131 HD(i, 2, s) ^ \
132 HD(i, 3, s) ^ w[p + i];
133
134 #define HD16(m, s, p) \
135 HD4(m, 0, s, p); \
136 HD4(m, 1, s, p); \
137 HD4(m, 2, s, p); \
138 HD4(m, 3, s, p); \
139
140 #define FD(i, x) InvS[gb(x, m[(i - x) & 3])]
141 #define FD4(i) dest[i] = Ui32(FD(i, 0), FD(i, 1), FD(i, 2), FD(i, 3)) ^ w[i];
142
Aes_SetKey_Enc(UInt32 * w,const Byte * key,unsigned keySize)143 void MY_FAST_CALL Aes_SetKey_Enc(UInt32 *w, const Byte *key, unsigned keySize)
144 {
145 unsigned i, wSize;
146 wSize = keySize + 28;
147 keySize /= 4;
148 w[0] = ((UInt32)keySize / 2) + 3;
149 w += 4;
150
151 for (i = 0; i < keySize; i++, key += 4)
152 w[i] = GetUi32(key);
153
154 for (; i < wSize; i++)
155 {
156 UInt32 t = w[(size_t)i - 1];
157 unsigned rem = i % keySize;
158 if (rem == 0)
159 t = Ui32(Sbox[gb1(t)] ^ Rcon[i / keySize], Sbox[gb2(t)], Sbox[gb3(t)], Sbox[gb0(t)]);
160 else if (keySize > 6 && rem == 4)
161 t = Ui32(Sbox[gb0(t)], Sbox[gb1(t)], Sbox[gb2(t)], Sbox[gb3(t)]);
162 w[i] = w[i - keySize] ^ t;
163 }
164 }
165
Aes_SetKey_Dec(UInt32 * w,const Byte * key,unsigned keySize)166 void MY_FAST_CALL Aes_SetKey_Dec(UInt32 *w, const Byte *key, unsigned keySize)
167 {
168 unsigned i, num;
169 Aes_SetKey_Enc(w, key, keySize);
170 num = keySize + 20;
171 w += 8;
172 for (i = 0; i < num; i++)
173 {
174 UInt32 r = w[i];
175 w[i] =
176 DD(0)[Sbox[gb0(r)]] ^
177 DD(1)[Sbox[gb1(r)]] ^
178 DD(2)[Sbox[gb2(r)]] ^
179 DD(3)[Sbox[gb3(r)]];
180 }
181 }
182
183 /* Aes_Encode and Aes_Decode functions work with little-endian words.
184 src and dest are pointers to 4 UInt32 words.
185 src and dest can point to same block */
186
Aes_Encode(const UInt32 * w,UInt32 * dest,const UInt32 * src)187 static void Aes_Encode(const UInt32 *w, UInt32 *dest, const UInt32 *src)
188 {
189 UInt32 s[4];
190 UInt32 m[4];
191 UInt32 numRounds2 = w[0];
192 w += 4;
193 s[0] = src[0] ^ w[0];
194 s[1] = src[1] ^ w[1];
195 s[2] = src[2] ^ w[2];
196 s[3] = src[3] ^ w[3];
197 w += 4;
198 for (;;)
199 {
200 HT16(m, s, 0);
201 if (--numRounds2 == 0)
202 break;
203 HT16(s, m, 4);
204 w += 8;
205 }
206 w += 4;
207 FT4(0); FT4(1); FT4(2); FT4(3);
208 }
209
Aes_Decode(const UInt32 * w,UInt32 * dest,const UInt32 * src)210 static void Aes_Decode(const UInt32 *w, UInt32 *dest, const UInt32 *src)
211 {
212 UInt32 s[4];
213 UInt32 m[4];
214 UInt32 numRounds2 = w[0];
215 w += 4 + numRounds2 * 8;
216 s[0] = src[0] ^ w[0];
217 s[1] = src[1] ^ w[1];
218 s[2] = src[2] ^ w[2];
219 s[3] = src[3] ^ w[3];
220 for (;;)
221 {
222 w -= 8;
223 HD16(m, s, 4);
224 if (--numRounds2 == 0)
225 break;
226 HD16(s, m, 0);
227 }
228 FD4(0); FD4(1); FD4(2); FD4(3);
229 }
230
AesCbc_Init(UInt32 * p,const Byte * iv)231 void AesCbc_Init(UInt32 *p, const Byte *iv)
232 {
233 unsigned i;
234 for (i = 0; i < 4; i++)
235 p[i] = GetUi32(iv + i * 4);
236 }
237
AesCbc_Encode(UInt32 * p,Byte * data,size_t numBlocks)238 void MY_FAST_CALL AesCbc_Encode(UInt32 *p, Byte *data, size_t numBlocks)
239 {
240 for (; numBlocks != 0; numBlocks--, data += AES_BLOCK_SIZE)
241 {
242 p[0] ^= GetUi32(data);
243 p[1] ^= GetUi32(data + 4);
244 p[2] ^= GetUi32(data + 8);
245 p[3] ^= GetUi32(data + 12);
246
247 Aes_Encode(p + 4, p, p);
248
249 SetUi32(data, p[0]);
250 SetUi32(data + 4, p[1]);
251 SetUi32(data + 8, p[2]);
252 SetUi32(data + 12, p[3]);
253 }
254 }
255
AesCbc_Decode(UInt32 * p,Byte * data,size_t numBlocks)256 void MY_FAST_CALL AesCbc_Decode(UInt32 *p, Byte *data, size_t numBlocks)
257 {
258 UInt32 in[4], out[4];
259 for (; numBlocks != 0; numBlocks--, data += AES_BLOCK_SIZE)
260 {
261 in[0] = GetUi32(data);
262 in[1] = GetUi32(data + 4);
263 in[2] = GetUi32(data + 8);
264 in[3] = GetUi32(data + 12);
265
266 Aes_Decode(p + 4, out, in);
267
268 SetUi32(data, p[0] ^ out[0]);
269 SetUi32(data + 4, p[1] ^ out[1]);
270 SetUi32(data + 8, p[2] ^ out[2]);
271 SetUi32(data + 12, p[3] ^ out[3]);
272
273 p[0] = in[0];
274 p[1] = in[1];
275 p[2] = in[2];
276 p[3] = in[3];
277 }
278 }
279
AesCtr_Code(UInt32 * p,Byte * data,size_t numBlocks)280 void MY_FAST_CALL AesCtr_Code(UInt32 *p, Byte *data, size_t numBlocks)
281 {
282 for (; numBlocks != 0; numBlocks--)
283 {
284 UInt32 temp[4];
285 unsigned i;
286
287 if (++p[0] == 0)
288 p[1]++;
289
290 Aes_Encode(p + 4, temp, p);
291
292 for (i = 0; i < 4; i++, data += 4)
293 {
294 UInt32 t = temp[i];
295
296 #ifdef MY_CPU_LE_UNALIGN
297 *((UInt32 *)data) ^= t;
298 #else
299 data[0] ^= (t & 0xFF);
300 data[1] ^= ((t >> 8) & 0xFF);
301 data[2] ^= ((t >> 16) & 0xFF);
302 data[3] ^= ((t >> 24));
303 #endif
304 }
305 }
306 }
307