1 /* Aes.c -- AES encryption / decryption
2 2013-11-12 : 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 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 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)) & 0xFF)
53
AesGenTables(void)54 void AesGenTables(void)
55 {
56 unsigned i;
57 for (i = 0; i < 256; i++)
58 InvS[Sbox[i]] = (Byte)i;
59 for (i = 0; i < 256; i++)
60 {
61 {
62 UInt32 a1 = Sbox[i];
63 UInt32 a2 = xtime(a1);
64 UInt32 a3 = a2 ^ a1;
65 T[ i] = Ui32(a2, a1, a1, a3);
66 T[0x100 + i] = Ui32(a3, a2, a1, a1);
67 T[0x200 + i] = Ui32(a1, a3, a2, a1);
68 T[0x300 + i] = Ui32(a1, a1, a3, a2);
69 }
70 {
71 UInt32 a1 = InvS[i];
72 UInt32 a2 = xtime(a1);
73 UInt32 a4 = xtime(a2);
74 UInt32 a8 = xtime(a4);
75 UInt32 a9 = a8 ^ a1;
76 UInt32 aB = a8 ^ a2 ^ a1;
77 UInt32 aD = a8 ^ a4 ^ a1;
78 UInt32 aE = a8 ^ a4 ^ a2;
79 D[ i] = Ui32(aE, a9, aD, aB);
80 D[0x100 + i] = Ui32(aB, aE, a9, aD);
81 D[0x200 + i] = Ui32(aD, aB, aE, a9);
82 D[0x300 + i] = Ui32(a9, aD, aB, aE);
83 }
84 }
85 g_AesCbc_Encode = AesCbc_Encode;
86 g_AesCbc_Decode = AesCbc_Decode;
87 g_AesCtr_Code = AesCtr_Code;
88 #ifdef MY_CPU_X86_OR_AMD64
89 if (CPU_Is_Aes_Supported())
90 {
91 g_AesCbc_Encode = AesCbc_Encode_Intel;
92 g_AesCbc_Decode = AesCbc_Decode_Intel;
93 g_AesCtr_Code = AesCtr_Code_Intel;
94 }
95 #endif
96 }
97
98 #define HT(i, x, s) (T + (x << 8))[gb ## x(s[(i + x) & 3])]
99 #define HT4(m, i, s, p) m[i] = \
100 HT(i, 0, s) ^ \
101 HT(i, 1, s) ^ \
102 HT(i, 2, s) ^ \
103 HT(i, 3, s) ^ w[p + i]
104 /* such order (2031) in HT16 is for VC6/K8 speed optimization) */
105 #define HT16(m, s, p) \
106 HT4(m, 2, s, p); \
107 HT4(m, 0, s, p); \
108 HT4(m, 3, s, p); \
109 HT4(m, 1, s, p); \
110
111 #define FT(i, x) Sbox[gb ## x(m[(i + x) & 3])]
112 #define FT4(i) dest[i] = Ui32(FT(i, 0), FT(i, 1), FT(i, 2), FT(i, 3)) ^ w[i];
113
114 #define HD(i, x, s) (D + (x << 8))[gb ## x(s[(i - x) & 3])]
115 #define HD4(m, i, s, p) m[i] = \
116 HD(i, 0, s) ^ \
117 HD(i, 1, s) ^ \
118 HD(i, 2, s) ^ \
119 HD(i, 3, s) ^ w[p + i];
120 /* such order (0231) in HD16 is for VC6/K8 speed optimization) */
121 #define HD16(m, s, p) \
122 HD4(m, 0, s, p); \
123 HD4(m, 2, s, p); \
124 HD4(m, 3, s, p); \
125 HD4(m, 1, s, p); \
126
127 #define FD(i, x) InvS[gb ## x(m[(i - x) & 3])]
128 #define FD4(i) dest[i] = Ui32(FD(i, 0), FD(i, 1), FD(i, 2), FD(i, 3)) ^ w[i];
129
Aes_SetKey_Enc(UInt32 * w,const Byte * key,unsigned keySize)130 void MY_FAST_CALL Aes_SetKey_Enc(UInt32 *w, const Byte *key, unsigned keySize)
131 {
132 unsigned i, wSize;
133 wSize = keySize + 28;
134 keySize /= 4;
135 w[0] = ((UInt32)keySize / 2) + 3;
136 w += 4;
137
138 for (i = 0; i < keySize; i++, key += 4)
139 w[i] = GetUi32(key);
140
141 for (; i < wSize; i++)
142 {
143 UInt32 t = w[i - 1];
144 unsigned rem = i % keySize;
145 if (rem == 0)
146 t = Ui32(Sbox[gb1(t)] ^ Rcon[i / keySize], Sbox[gb2(t)], Sbox[gb3(t)], Sbox[gb0(t)]);
147 else if (keySize > 6 && rem == 4)
148 t = Ui32(Sbox[gb0(t)], Sbox[gb1(t)], Sbox[gb2(t)], Sbox[gb3(t)]);
149 w[i] = w[i - keySize] ^ t;
150 }
151 }
152
Aes_SetKey_Dec(UInt32 * w,const Byte * key,unsigned keySize)153 void MY_FAST_CALL Aes_SetKey_Dec(UInt32 *w, const Byte *key, unsigned keySize)
154 {
155 unsigned i, num;
156 Aes_SetKey_Enc(w, key, keySize);
157 num = keySize + 20;
158 w += 8;
159 for (i = 0; i < num; i++)
160 {
161 UInt32 r = w[i];
162 w[i] =
163 D[ Sbox[gb0(r)]] ^
164 D[0x100 + Sbox[gb1(r)]] ^
165 D[0x200 + Sbox[gb2(r)]] ^
166 D[0x300 + Sbox[gb3(r)]];
167 }
168 }
169
170 /* Aes_Encode and Aes_Decode functions work with little-endian words.
171 src and dest are pointers to 4 UInt32 words.
172 arc and dest can point to same block */
173
Aes_Encode(const UInt32 * w,UInt32 * dest,const UInt32 * src)174 static void Aes_Encode(const UInt32 *w, UInt32 *dest, const UInt32 *src)
175 {
176 UInt32 s[4];
177 UInt32 m[4];
178 UInt32 numRounds2 = w[0];
179 w += 4;
180 s[0] = src[0] ^ w[0];
181 s[1] = src[1] ^ w[1];
182 s[2] = src[2] ^ w[2];
183 s[3] = src[3] ^ w[3];
184 w += 4;
185 for (;;)
186 {
187 HT16(m, s, 0);
188 if (--numRounds2 == 0)
189 break;
190 HT16(s, m, 4);
191 w += 8;
192 }
193 w += 4;
194 FT4(0); FT4(1); FT4(2); FT4(3);
195 }
196
Aes_Decode(const UInt32 * w,UInt32 * dest,const UInt32 * src)197 static void Aes_Decode(const UInt32 *w, UInt32 *dest, const UInt32 *src)
198 {
199 UInt32 s[4];
200 UInt32 m[4];
201 UInt32 numRounds2 = w[0];
202 w += 4 + numRounds2 * 8;
203 s[0] = src[0] ^ w[0];
204 s[1] = src[1] ^ w[1];
205 s[2] = src[2] ^ w[2];
206 s[3] = src[3] ^ w[3];
207 for (;;)
208 {
209 w -= 8;
210 HD16(m, s, 4);
211 if (--numRounds2 == 0)
212 break;
213 HD16(s, m, 0);
214 }
215 FD4(0); FD4(1); FD4(2); FD4(3);
216 }
217
AesCbc_Init(UInt32 * p,const Byte * iv)218 void AesCbc_Init(UInt32 *p, const Byte *iv)
219 {
220 unsigned i;
221 for (i = 0; i < 4; i++)
222 p[i] = GetUi32(iv + i * 4);
223 }
224
AesCbc_Encode(UInt32 * p,Byte * data,size_t numBlocks)225 void MY_FAST_CALL AesCbc_Encode(UInt32 *p, Byte *data, size_t numBlocks)
226 {
227 for (; numBlocks != 0; numBlocks--, data += AES_BLOCK_SIZE)
228 {
229 p[0] ^= GetUi32(data);
230 p[1] ^= GetUi32(data + 4);
231 p[2] ^= GetUi32(data + 8);
232 p[3] ^= GetUi32(data + 12);
233
234 Aes_Encode(p + 4, p, p);
235
236 SetUi32(data, p[0]);
237 SetUi32(data + 4, p[1]);
238 SetUi32(data + 8, p[2]);
239 SetUi32(data + 12, p[3]);
240 }
241 }
242
AesCbc_Decode(UInt32 * p,Byte * data,size_t numBlocks)243 void MY_FAST_CALL AesCbc_Decode(UInt32 *p, Byte *data, size_t numBlocks)
244 {
245 UInt32 in[4], out[4];
246 for (; numBlocks != 0; numBlocks--, data += AES_BLOCK_SIZE)
247 {
248 in[0] = GetUi32(data);
249 in[1] = GetUi32(data + 4);
250 in[2] = GetUi32(data + 8);
251 in[3] = GetUi32(data + 12);
252
253 Aes_Decode(p + 4, out, in);
254
255 SetUi32(data, p[0] ^ out[0]);
256 SetUi32(data + 4, p[1] ^ out[1]);
257 SetUi32(data + 8, p[2] ^ out[2]);
258 SetUi32(data + 12, p[3] ^ out[3]);
259
260 p[0] = in[0];
261 p[1] = in[1];
262 p[2] = in[2];
263 p[3] = in[3];
264 }
265 }
266
AesCtr_Code(UInt32 * p,Byte * data,size_t numBlocks)267 void MY_FAST_CALL AesCtr_Code(UInt32 *p, Byte *data, size_t numBlocks)
268 {
269 for (; numBlocks != 0; numBlocks--)
270 {
271 UInt32 temp[4];
272 Byte buf[16];
273 int i;
274 if (++p[0] == 0)
275 p[1]++;
276 Aes_Encode(p + 4, temp, p);
277 SetUi32(buf, temp[0]);
278 SetUi32(buf + 4, temp[1]);
279 SetUi32(buf + 8, temp[2]);
280 SetUi32(buf + 12, temp[3]);
281 for (i = 0; i < 16; i++)
282 *data++ ^= buf[i];
283 }
284 }
285