Coverage Report

Created: 2020-05-23 13:54

/src/botan/src/lib/block/sm4/sm4.cpp
Line
Count
Source (jump to first uncovered line)
1
/*
2
* SM4
3
* (C) 2017 Ribose Inc
4
* (C) 2018 Jack Lloyd
5
*
6
* Botan is released under the Simplified BSD License (see license.txt)
7
*/
8
9
#include <botan/sm4.h>
10
#include <botan/loadstor.h>
11
#include <botan/rotate.h>
12
#include <botan/cpuid.h>
13
14
namespace Botan {
15
16
namespace {
17
18
alignas(64)
19
const uint8_t SM4_SBOX[256] = {
20
0xD6, 0x90, 0xE9, 0xFE, 0xCC, 0xE1, 0x3D, 0xB7, 0x16, 0xB6, 0x14, 0xC2, 0x28, 0xFB, 0x2C, 0x05,
21
0x2B, 0x67, 0x9A, 0x76, 0x2A, 0xBE, 0x04, 0xC3, 0xAA, 0x44, 0x13, 0x26, 0x49, 0x86, 0x06, 0x99,
22
0x9C, 0x42, 0x50, 0xF4, 0x91, 0xEF, 0x98, 0x7A, 0x33, 0x54, 0x0B, 0x43, 0xED, 0xCF, 0xAC, 0x62,
23
0xE4, 0xB3, 0x1C, 0xA9, 0xC9, 0x08, 0xE8, 0x95, 0x80, 0xDF, 0x94, 0xFA, 0x75, 0x8F, 0x3F, 0xA6,
24
0x47, 0x07, 0xA7, 0xFC, 0xF3, 0x73, 0x17, 0xBA, 0x83, 0x59, 0x3C, 0x19, 0xE6, 0x85, 0x4F, 0xA8,
25
0x68, 0x6B, 0x81, 0xB2, 0x71, 0x64, 0xDA, 0x8B, 0xF8, 0xEB, 0x0F, 0x4B, 0x70, 0x56, 0x9D, 0x35,
26
0x1E, 0x24, 0x0E, 0x5E, 0x63, 0x58, 0xD1, 0xA2, 0x25, 0x22, 0x7C, 0x3B, 0x01, 0x21, 0x78, 0x87,
27
0xD4, 0x00, 0x46, 0x57, 0x9F, 0xD3, 0x27, 0x52, 0x4C, 0x36, 0x02, 0xE7, 0xA0, 0xC4, 0xC8, 0x9E,
28
0xEA, 0xBF, 0x8A, 0xD2, 0x40, 0xC7, 0x38, 0xB5, 0xA3, 0xF7, 0xF2, 0xCE, 0xF9, 0x61, 0x15, 0xA1,
29
0xE0, 0xAE, 0x5D, 0xA4, 0x9B, 0x34, 0x1A, 0x55, 0xAD, 0x93, 0x32, 0x30, 0xF5, 0x8C, 0xB1, 0xE3,
30
0x1D, 0xF6, 0xE2, 0x2E, 0x82, 0x66, 0xCA, 0x60, 0xC0, 0x29, 0x23, 0xAB, 0x0D, 0x53, 0x4E, 0x6F,
31
0xD5, 0xDB, 0x37, 0x45, 0xDE, 0xFD, 0x8E, 0x2F, 0x03, 0xFF, 0x6A, 0x72, 0x6D, 0x6C, 0x5B, 0x51,
32
0x8D, 0x1B, 0xAF, 0x92, 0xBB, 0xDD, 0xBC, 0x7F, 0x11, 0xD9, 0x5C, 0x41, 0x1F, 0x10, 0x5A, 0xD8,
33
0x0A, 0xC1, 0x31, 0x88, 0xA5, 0xCD, 0x7B, 0xBD, 0x2D, 0x74, 0xD0, 0x12, 0xB8, 0xE5, 0xB4, 0xB0,
34
0x89, 0x69, 0x97, 0x4A, 0x0C, 0x96, 0x77, 0x7E, 0x65, 0xB9, 0xF1, 0x09, 0xC5, 0x6E, 0xC6, 0x84,
35
0x18, 0xF0, 0x7D, 0xEC, 0x3A, 0xDC, 0x4D, 0x20, 0x79, 0xEE, 0x5F, 0x3E, 0xD7, 0xCB, 0x39, 0x48
36
};
37
38
/*
39
* SM4_SBOX_T[j] == L(SM4_SBOX[j]).
40
*/
41
alignas(64)
42
const uint32_t SM4_SBOX_T[256] = {
43
   0x8ED55B5B, 0xD0924242, 0x4DEAA7A7, 0x06FDFBFB, 0xFCCF3333, 0x65E28787,
44
   0xC93DF4F4, 0x6BB5DEDE, 0x4E165858, 0x6EB4DADA, 0x44145050, 0xCAC10B0B,
45
   0x8828A0A0, 0x17F8EFEF, 0x9C2CB0B0, 0x11051414, 0x872BACAC, 0xFB669D9D,
46
   0xF2986A6A, 0xAE77D9D9, 0x822AA8A8, 0x46BCFAFA, 0x14041010, 0xCFC00F0F,
47
   0x02A8AAAA, 0x54451111, 0x5F134C4C, 0xBE269898, 0x6D482525, 0x9E841A1A,
48
   0x1E061818, 0xFD9B6666, 0xEC9E7272, 0x4A430909, 0x10514141, 0x24F7D3D3,
49
   0xD5934646, 0x53ECBFBF, 0xF89A6262, 0x927BE9E9, 0xFF33CCCC, 0x04555151,
50
   0x270B2C2C, 0x4F420D0D, 0x59EEB7B7, 0xF3CC3F3F, 0x1CAEB2B2, 0xEA638989,
51
   0x74E79393, 0x7FB1CECE, 0x6C1C7070, 0x0DABA6A6, 0xEDCA2727, 0x28082020,
52
   0x48EBA3A3, 0xC1975656, 0x80820202, 0xA3DC7F7F, 0xC4965252, 0x12F9EBEB,
53
   0xA174D5D5, 0xB38D3E3E, 0xC33FFCFC, 0x3EA49A9A, 0x5B461D1D, 0x1B071C1C,
54
   0x3BA59E9E, 0x0CFFF3F3, 0x3FF0CFCF, 0xBF72CDCD, 0x4B175C5C, 0x52B8EAEA,
55
   0x8F810E0E, 0x3D586565, 0xCC3CF0F0, 0x7D196464, 0x7EE59B9B, 0x91871616,
56
   0x734E3D3D, 0x08AAA2A2, 0xC869A1A1, 0xC76AADAD, 0x85830606, 0x7AB0CACA,
57
   0xB570C5C5, 0xF4659191, 0xB2D96B6B, 0xA7892E2E, 0x18FBE3E3, 0x47E8AFAF,
58
   0x330F3C3C, 0x674A2D2D, 0xB071C1C1, 0x0E575959, 0xE99F7676, 0xE135D4D4,
59
   0x661E7878, 0xB4249090, 0x360E3838, 0x265F7979, 0xEF628D8D, 0x38596161,
60
   0x95D24747, 0x2AA08A8A, 0xB1259494, 0xAA228888, 0x8C7DF1F1, 0xD73BECEC,
61
   0x05010404, 0xA5218484, 0x9879E1E1, 0x9B851E1E, 0x84D75353, 0x00000000,
62
   0x5E471919, 0x0B565D5D, 0xE39D7E7E, 0x9FD04F4F, 0xBB279C9C, 0x1A534949,
63
   0x7C4D3131, 0xEE36D8D8, 0x0A020808, 0x7BE49F9F, 0x20A28282, 0xD4C71313,
64
   0xE8CB2323, 0xE69C7A7A, 0x42E9ABAB, 0x43BDFEFE, 0xA2882A2A, 0x9AD14B4B,
65
   0x40410101, 0xDBC41F1F, 0xD838E0E0, 0x61B7D6D6, 0x2FA18E8E, 0x2BF4DFDF,
66
   0x3AF1CBCB, 0xF6CD3B3B, 0x1DFAE7E7, 0xE5608585, 0x41155454, 0x25A38686,
67
   0x60E38383, 0x16ACBABA, 0x295C7575, 0x34A69292, 0xF7996E6E, 0xE434D0D0,
68
   0x721A6868, 0x01545555, 0x19AFB6B6, 0xDF914E4E, 0xFA32C8C8, 0xF030C0C0,
69
   0x21F6D7D7, 0xBC8E3232, 0x75B3C6C6, 0x6FE08F8F, 0x691D7474, 0x2EF5DBDB,
70
   0x6AE18B8B, 0x962EB8B8, 0x8A800A0A, 0xFE679999, 0xE2C92B2B, 0xE0618181,
71
   0xC0C30303, 0x8D29A4A4, 0xAF238C8C, 0x07A9AEAE, 0x390D3434, 0x1F524D4D,
72
   0x764F3939, 0xD36EBDBD, 0x81D65757, 0xB7D86F6F, 0xEB37DCDC, 0x51441515,
73
   0xA6DD7B7B, 0x09FEF7F7, 0xB68C3A3A, 0x932FBCBC, 0x0F030C0C, 0x03FCFFFF,
74
   0xC26BA9A9, 0xBA73C9C9, 0xD96CB5B5, 0xDC6DB1B1, 0x375A6D6D, 0x15504545,
75
   0xB98F3636, 0x771B6C6C, 0x13ADBEBE, 0xDA904A4A, 0x57B9EEEE, 0xA9DE7777,
76
   0x4CBEF2F2, 0x837EFDFD, 0x55114444, 0xBDDA6767, 0x2C5D7171, 0x45400505,
77
   0x631F7C7C, 0x50104040, 0x325B6969, 0xB8DB6363, 0x220A2828, 0xC5C20707,
78
   0xF531C4C4, 0xA88A2222, 0x31A79696, 0xF9CE3737, 0x977AEDED, 0x49BFF6F6,
79
   0x992DB4B4, 0xA475D1D1, 0x90D34343, 0x5A124848, 0x58BAE2E2, 0x71E69797,
80
   0x64B6D2D2, 0x70B2C2C2, 0xAD8B2626, 0xCD68A5A5, 0xCB955E5E, 0x624B2929,
81
   0x3C0C3030, 0xCE945A5A, 0xAB76DDDD, 0x867FF9F9, 0xF1649595, 0x5DBBE6E6,
82
   0x35F2C7C7, 0x2D092424, 0xD1C61717, 0xD66FB9B9, 0xDEC51B1B, 0x94861212,
83
   0x78186060, 0x30F3C3C3, 0x897CF5F5, 0x5CEFB3B3, 0xD23AE8E8, 0xACDF7373,
84
   0x794C3535, 0xA0208080, 0x9D78E5E5, 0x56EDBBBB, 0x235E7D7D, 0xC63EF8F8,
85
   0x8BD45F5F, 0xE7C82F2F, 0xDD39E4E4, 0x68492121 };
86
87
inline uint32_t SM4_T_slow(uint32_t b)
88
0
   {
89
0
   const uint32_t t = make_uint32(SM4_SBOX[get_byte(0,b)],
90
0
                                  SM4_SBOX[get_byte(1,b)],
91
0
                                  SM4_SBOX[get_byte(2,b)],
92
0
                                  SM4_SBOX[get_byte(3,b)]);
93
0
94
0
   // L linear transform
95
0
   return t ^ rotl<2>(t) ^ rotl<10>(t) ^ rotl<18>(t) ^ rotl<24>(t);
96
0
   }
97
98
inline uint32_t SM4_T(uint32_t b)
99
0
   {
100
0
   return          SM4_SBOX_T[get_byte(0,b)]  ^
101
0
          rotr< 8>(SM4_SBOX_T[get_byte(1,b)]) ^
102
0
          rotr<16>(SM4_SBOX_T[get_byte(2,b)]) ^
103
0
          rotr<24>(SM4_SBOX_T[get_byte(3,b)]);
104
0
   }
105
106
// Variant of T for key schedule
107
inline uint32_t SM4_Tp(uint32_t b)
108
0
   {
109
0
   const uint32_t t = make_uint32(SM4_SBOX[get_byte(0,b)],
110
0
                                  SM4_SBOX[get_byte(1,b)],
111
0
                                  SM4_SBOX[get_byte(2,b)],
112
0
                                  SM4_SBOX[get_byte(3,b)]);
113
0
114
0
   // L' linear transform
115
0
   return t ^ rotl<13>(t) ^ rotl<23>(t);
116
0
   }
117
118
0
#define SM4_E_RNDS(B, R, F) do {                           \
119
0
   B##0 ^= F(B##1 ^ B##2 ^ B##3 ^ m_RK[4*R+0]);            \
120
0
   B##1 ^= F(B##0 ^ B##2 ^ B##3 ^ m_RK[4*R+1]);            \
121
0
   B##2 ^= F(B##0 ^ B##1 ^ B##3 ^ m_RK[4*R+2]);            \
122
0
   B##3 ^= F(B##0 ^ B##1 ^ B##2 ^ m_RK[4*R+3]);            \
123
0
   } while(0)
124
125
0
#define SM4_D_RNDS(B, R, F) do {                           \
126
0
   B##0 ^= F(B##1 ^ B##2 ^ B##3 ^ m_RK[4*R+3]);            \
127
0
   B##1 ^= F(B##0 ^ B##2 ^ B##3 ^ m_RK[4*R+2]);            \
128
0
   B##2 ^= F(B##0 ^ B##1 ^ B##3 ^ m_RK[4*R+1]);            \
129
0
   B##3 ^= F(B##0 ^ B##1 ^ B##2 ^ m_RK[4*R+0]);            \
130
0
   } while(0)
131
132
}
133
134
/*
135
* SM4 Encryption
136
*/
137
void SM4::encrypt_n(const uint8_t in[], uint8_t out[], size_t blocks) const
138
0
   {
139
0
   verify_key_set(m_RK.empty() == false);
140
0
141
#if defined(BOTAN_HAS_SM4_ARMV8)
142
   if(CPUID::has_arm_sm4())
143
      return sm4_armv8_encrypt(in, out, blocks);
144
#endif
145
146
0
   while(blocks >= 2)
147
0
      {
148
0
      uint32_t B0 = load_be<uint32_t>(in, 0);
149
0
      uint32_t B1 = load_be<uint32_t>(in, 1);
150
0
      uint32_t B2 = load_be<uint32_t>(in, 2);
151
0
      uint32_t B3 = load_be<uint32_t>(in, 3);
152
0
153
0
      uint32_t C0 = load_be<uint32_t>(in, 4);
154
0
      uint32_t C1 = load_be<uint32_t>(in, 5);
155
0
      uint32_t C2 = load_be<uint32_t>(in, 6);
156
0
      uint32_t C3 = load_be<uint32_t>(in, 7);
157
0
158
0
      SM4_E_RNDS(B, 0, SM4_T_slow);
159
0
      SM4_E_RNDS(C, 0, SM4_T_slow);
160
0
      SM4_E_RNDS(B, 1, SM4_T);
161
0
      SM4_E_RNDS(C, 1, SM4_T);
162
0
      SM4_E_RNDS(B, 2, SM4_T);
163
0
      SM4_E_RNDS(C, 2, SM4_T);
164
0
      SM4_E_RNDS(B, 3, SM4_T);
165
0
      SM4_E_RNDS(C, 3, SM4_T);
166
0
      SM4_E_RNDS(B, 4, SM4_T);
167
0
      SM4_E_RNDS(C, 4, SM4_T);
168
0
      SM4_E_RNDS(B, 5, SM4_T);
169
0
      SM4_E_RNDS(C, 5, SM4_T);
170
0
      SM4_E_RNDS(B, 6, SM4_T);
171
0
      SM4_E_RNDS(C, 6, SM4_T);
172
0
      SM4_E_RNDS(B, 7, SM4_T_slow);
173
0
      SM4_E_RNDS(C, 7, SM4_T_slow);
174
0
175
0
      store_be(out, B3, B2, B1, B0, C3, C2, C1, C0);
176
0
177
0
      in += 2*BLOCK_SIZE;
178
0
      out += 2*BLOCK_SIZE;
179
0
      blocks -= 2;
180
0
      }
181
0
182
0
   for(size_t i = 0; i != blocks; ++i)
183
0
      {
184
0
      uint32_t B0 = load_be<uint32_t>(in, 0);
185
0
      uint32_t B1 = load_be<uint32_t>(in, 1);
186
0
      uint32_t B2 = load_be<uint32_t>(in, 2);
187
0
      uint32_t B3 = load_be<uint32_t>(in, 3);
188
0
189
0
      SM4_E_RNDS(B, 0, SM4_T_slow);
190
0
      SM4_E_RNDS(B, 1, SM4_T);
191
0
      SM4_E_RNDS(B, 2, SM4_T);
192
0
      SM4_E_RNDS(B, 3, SM4_T);
193
0
      SM4_E_RNDS(B, 4, SM4_T);
194
0
      SM4_E_RNDS(B, 5, SM4_T);
195
0
      SM4_E_RNDS(B, 6, SM4_T);
196
0
      SM4_E_RNDS(B, 7, SM4_T_slow);
197
0
198
0
      store_be(out, B3, B2, B1, B0);
199
0
200
0
      in += BLOCK_SIZE;
201
0
      out += BLOCK_SIZE;
202
0
      }
203
0
   }
204
205
/*
206
* SM4 Decryption
207
*/
208
void SM4::decrypt_n(const uint8_t in[], uint8_t out[], size_t blocks) const
209
0
   {
210
0
   verify_key_set(m_RK.empty() == false);
211
0
212
#if defined(BOTAN_HAS_SM4_ARMV8)
213
   if(CPUID::has_arm_sm4())
214
      return sm4_armv8_decrypt(in, out, blocks);
215
#endif
216
217
0
   while(blocks >= 2)
218
0
      {
219
0
      uint32_t B0 = load_be<uint32_t>(in, 0);
220
0
      uint32_t B1 = load_be<uint32_t>(in, 1);
221
0
      uint32_t B2 = load_be<uint32_t>(in, 2);
222
0
      uint32_t B3 = load_be<uint32_t>(in, 3);
223
0
224
0
      uint32_t C0 = load_be<uint32_t>(in, 4);
225
0
      uint32_t C1 = load_be<uint32_t>(in, 5);
226
0
      uint32_t C2 = load_be<uint32_t>(in, 6);
227
0
      uint32_t C3 = load_be<uint32_t>(in, 7);
228
0
229
0
      SM4_D_RNDS(B, 7, SM4_T_slow);
230
0
      SM4_D_RNDS(C, 7, SM4_T_slow);
231
0
      SM4_D_RNDS(B, 6, SM4_T);
232
0
      SM4_D_RNDS(C, 6, SM4_T);
233
0
      SM4_D_RNDS(B, 5, SM4_T);
234
0
      SM4_D_RNDS(C, 5, SM4_T);
235
0
      SM4_D_RNDS(B, 4, SM4_T);
236
0
      SM4_D_RNDS(C, 4, SM4_T);
237
0
      SM4_D_RNDS(B, 3, SM4_T);
238
0
      SM4_D_RNDS(C, 3, SM4_T);
239
0
      SM4_D_RNDS(B, 2, SM4_T);
240
0
      SM4_D_RNDS(C, 2, SM4_T);
241
0
      SM4_D_RNDS(B, 1, SM4_T);
242
0
      SM4_D_RNDS(C, 1, SM4_T);
243
0
      SM4_D_RNDS(B, 0, SM4_T_slow);
244
0
      SM4_D_RNDS(C, 0, SM4_T_slow);
245
0
246
0
      store_be(out, B3, B2, B1, B0, C3, C2, C1, C0);
247
0
248
0
      in += 2*BLOCK_SIZE;
249
0
      out += 2*BLOCK_SIZE;
250
0
      blocks -= 2;
251
0
      }
252
0
253
0
   for(size_t i = 0; i != blocks; ++i)
254
0
      {
255
0
      uint32_t B0 = load_be<uint32_t>(in, 0);
256
0
      uint32_t B1 = load_be<uint32_t>(in, 1);
257
0
      uint32_t B2 = load_be<uint32_t>(in, 2);
258
0
      uint32_t B3 = load_be<uint32_t>(in, 3);
259
0
260
0
      SM4_D_RNDS(B, 7, SM4_T_slow);
261
0
      SM4_D_RNDS(B, 6, SM4_T);
262
0
      SM4_D_RNDS(B, 5, SM4_T);
263
0
      SM4_D_RNDS(B, 4, SM4_T);
264
0
      SM4_D_RNDS(B, 3, SM4_T);
265
0
      SM4_D_RNDS(B, 2, SM4_T);
266
0
      SM4_D_RNDS(B, 1, SM4_T);
267
0
      SM4_D_RNDS(B, 0, SM4_T_slow);
268
0
269
0
      store_be(out, B3, B2, B1, B0);
270
0
271
0
      in += BLOCK_SIZE;
272
0
      out += BLOCK_SIZE;
273
0
      }
274
0
   }
275
276
#undef SM4_E_RNDS
277
#undef SM4_D_RNDS
278
279
/*
280
* SM4 Key Schedule
281
*/
282
void SM4::key_schedule(const uint8_t key[], size_t)
283
0
   {
284
0
   // System parameter or family key
285
0
   const uint32_t FK[4] = { 0xa3b1bac6, 0x56aa3350, 0x677d9197, 0xb27022dc };
286
0
287
0
   const uint32_t CK[32] = {
288
0
      0x00070E15, 0x1C232A31, 0x383F464D, 0x545B6269,
289
0
      0x70777E85, 0x8C939AA1, 0xA8AFB6BD, 0xC4CBD2D9,
290
0
      0xE0E7EEF5, 0xFC030A11, 0x181F262D, 0x343B4249,
291
0
      0x50575E65, 0x6C737A81, 0x888F969D, 0xA4ABB2B9,
292
0
      0xC0C7CED5, 0xDCE3EAF1, 0xF8FF060D, 0x141B2229,
293
0
      0x30373E45, 0x4C535A61, 0x686F767D, 0x848B9299,
294
0
      0xA0A7AEB5, 0xBCC3CAD1, 0xD8DFE6ED, 0xF4FB0209,
295
0
      0x10171E25, 0x2C333A41, 0x484F565D, 0x646B7279
296
0
   };
297
0
298
0
   secure_vector<uint32_t> K(4);
299
0
   K[0] = load_be<uint32_t>(key, 0) ^ FK[0];
300
0
   K[1] = load_be<uint32_t>(key, 1) ^ FK[1];
301
0
   K[2] = load_be<uint32_t>(key, 2) ^ FK[2];
302
0
   K[3] = load_be<uint32_t>(key, 3) ^ FK[3];
303
0
304
0
   m_RK.resize(32);
305
0
   for(size_t i = 0; i != 32; ++i)
306
0
      {
307
0
      K[i % 4] ^= SM4_Tp(K[(i+1)%4] ^ K[(i+2)%4] ^ K[(i+3)%4] ^ CK[i]);
308
0
      m_RK[i] = K[i % 4];
309
0
      }
310
0
   }
311
312
void SM4::clear()
313
0
   {
314
0
   zap(m_RK);
315
0
   }
316
317
size_t SM4::parallelism() const
318
0
   {
319
#if defined(BOTAN_HAS_SM4_ARMV8)
320
   if(CPUID::has_arm_sm4())
321
      {
322
      return 4;
323
      }
324
#endif
325
326
0
   return 1;
327
0
   }
328
329
std::string SM4::provider() const
330
0
   {
331
#if defined(BOTAN_HAS_SM4_ARMV8)
332
   if(CPUID::has_arm_sm4())
333
      {
334
      return "armv8";
335
      }
336
#endif
337
338
0
   return "base";
339
0
   }
340
341
}