forked from bitcoinclassic/bitcoinclassic
/
modified_scrypt_smix.cpp
380 lines (321 loc) · 15.1 KB
/
modified_scrypt_smix.cpp
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
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
/*-
* Copyright 2009 Colin Percival
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* This file was originally written by Colin Percival as part of the Tarsnap
* online backup system.
*/
#include <stdint.h>
#include <string.h>
#include "../uint256.h"
#include "../util.h"
//#include "sha256.h"
#include "sysendian.h"
#include "modified_scrypt_sha256.h"
#include "modified_scrypt_smix.h"
inline void blkcpy(void *, const void *, size_t);
inline void blkxor(void *, const void *, size_t);
inline void salsa20_8(uint32_t[16]);
inline void blockmix_salsa8(const uint32_t *, uint32_t *, uint32_t *);
inline uint64_t integerify(const void *);
inline void
blkcpy(void * dest, const void * src, size_t len)
{
size_t * D = (size_t *)dest;
const size_t * S = (const size_t *)src;
size_t L = len / sizeof(size_t);
size_t i;
for (i = 0; i < L; i++)
D[i] = S[i];
}
inline void
blkxor(void * dest, const void * src, size_t len)
{
size_t * D = (size_t *)dest;
const size_t * S = (const size_t *)src;
size_t L = len / sizeof(size_t);
size_t i;
for (i = 0; i < L; i++)
D[i] ^= S[i];
}
/**
* salsa20_8(B):
* Apply the salsa20/8 core to the provided block.
*/
inline void
salsa20_8(uint32_t B[16])
{
uint32_t x[16];
size_t i;
blkcpy(x, B, 64);
for (i = 0; i < 8; i += 2) {
#define R(a,b) (((a) << (b)) | ((a) >> (32 - (b))))
/* Operate on columns. */
x[ 4] ^= R(x[ 0]+x[12], 7); x[ 8] ^= R(x[ 4]+x[ 0], 9);
x[12] ^= R(x[ 8]+x[ 4],13); x[ 0] ^= R(x[12]+x[ 8],18);
x[ 9] ^= R(x[ 5]+x[ 1], 7); x[13] ^= R(x[ 9]+x[ 5], 9);
x[ 1] ^= R(x[13]+x[ 9],13); x[ 5] ^= R(x[ 1]+x[13],18);
x[14] ^= R(x[10]+x[ 6], 7); x[ 2] ^= R(x[14]+x[10], 9);
x[ 6] ^= R(x[ 2]+x[14],13); x[10] ^= R(x[ 6]+x[ 2],18);
x[ 3] ^= R(x[15]+x[11], 7); x[ 7] ^= R(x[ 3]+x[15], 9);
x[11] ^= R(x[ 7]+x[ 3],13); x[15] ^= R(x[11]+x[ 7],18);
/* Operate on rows. */
x[ 1] ^= R(x[ 0]+x[ 3], 7); x[ 2] ^= R(x[ 1]+x[ 0], 9);
x[ 3] ^= R(x[ 2]+x[ 1],13); x[ 0] ^= R(x[ 3]+x[ 2],18);
x[ 6] ^= R(x[ 5]+x[ 4], 7); x[ 7] ^= R(x[ 6]+x[ 5], 9);
x[ 4] ^= R(x[ 7]+x[ 6],13); x[ 5] ^= R(x[ 4]+x[ 7],18);
x[11] ^= R(x[10]+x[ 9], 7); x[ 8] ^= R(x[11]+x[10], 9);
x[ 9] ^= R(x[ 8]+x[11],13); x[10] ^= R(x[ 9]+x[ 8],18);
x[12] ^= R(x[15]+x[14], 7); x[13] ^= R(x[12]+x[15], 9);
x[14] ^= R(x[13]+x[12],13); x[15] ^= R(x[14]+x[13],18);
#undef R
}
for (i = 0; i < 16; i++)
B[i] += x[i];
}
/**
* blockmix_salsa8(Bin, Bout, X):
* Compute Bout = BlockMix_{salsa20/8}(Bin). The input Bin must be 128
* bytes in length; the output Bout must also be the same size. The
* temporary space X must be 64 bytes.
*/
inline void
blockmix_salsa8(const uint32_t * Bin, uint32_t * Bout, uint32_t * X)
{
size_t i;
size_t r = 1; // Using r=1, let the compiler optimize it out rather than doing by hand
/* 1: X <-- B_{2r - 1} */
blkcpy(X, &Bin[(2 * r - 1) * 16], 64);
/* 2: for i = 0 to 2r - 1 do */
for (i = 0; i < 2 * r; i += 2) {
/* 3: X <-- H(X \xor B_i) */
blkxor(X, &Bin[i * 16], 64);
salsa20_8(X);
/* 4: Y_i <-- X */
/* 6: B' <-- (Y_0, Y_2 ... Y_{2r-2}, Y_1, Y_3 ... Y_{2r-1}) */
blkcpy(&Bout[i * 8], X, 64);
/* 3: X <-- H(X \xor B_i) */
blkxor(X, &Bin[i * 16 + 16], 64);
salsa20_8(X);
/* 4: Y_i <-- X */
/* 6: B' <-- (Y_0, Y_2 ... Y_{2r-2}, Y_1, Y_3 ... Y_{2r-1}) */
blkcpy(&Bout[i * 8 + r * 16], X, 64);
}
}
/**
* integerify(B, r):
* Return the result of parsing B_{2r-1} as a little-endian integer.
*/
inline uint64_t
integerify(const void * B)
{
size_t r = 1; // Using r=1, let the compiler optimize it out, rather than by hand
const uint32_t * X = (const uint32_t *)((const void *)((uintptr_t)(B) + (2 * r - 1) * 64));
return (((uint64_t)(X[1]) << 32) + X[0]);
}
/**
* crypto_scrypt_smix(B, N, V, XY):
* Compute B = SMix_r(B, N). The input B must be 128 bytes in length;
* the temporary storage V must be 128N bytes in length; the temporary
* storage XY must be 256 + 64 bytes in length. The value N must be a
* power of 2 greater than 1. The arrays B, V, and XY must be aligned to a
* multiple of 64 bytes.
*/
void
crypto_scrypt_smix(uint8_t * B, uint64_t N, void * _V, void * XY)
{
size_t r = 1; // Using r=1, let the compiler optimize r out rather than by hand
uint32_t * X = (uint32_t *)XY;
uint32_t * Y = (uint32_t *)((void *)((uint8_t *)(XY) + 128 * r));
uint32_t * Z = (uint32_t *)((void *)((uint8_t *)(XY) + 256 * r));
uint32_t * V = (uint32_t *)_V;
uint64_t i;
uint64_t j;
size_t k;
//uint256 Random_CachelineLow;
//uint256 Random_CachelineHigh;
uint64_t Random_Cacheline_Position;
uint64_t Random_DataMixer[8];
/* 1: X <-- B */
for (k = 0; k < 32 * r; k++)
X[k] = le32dec(&B[4 * k]);
// Explaination of scrypt modifications
//
// Script tries to be ASIC & GPU resistant by using a memory size that cannot fit onto a single CMOS chip.
// However even though memory reads are performed off-chip the reads themselves are sequential, this enables
// a hardware implementation to create a data pipeline where on each clock cycle the next memory address is
// read and data is pipelined through an ASIC/FPGA implementation.
//
// This version of scrypt adds a sequence of reads that are random, unpredicable and dynamically change with
// each hash. As a result it is not possible to create a data pipeline where data is processed on each clock
// cycle. Instead computation must follow the following steps:
//
// a) Compute next address to read from
// b) Send read request to memory address
// c) Wait 100s of clock cycles
// d) Receive single data packet
// e) Perform a very simple computation in a few (<10) clock cycles
// f) Write result back to memory and use as next address to read from
// g) Send next read request to next memory address
// h) Wait 100s of clock cycles
// --- Repeat billions of times ---
//
// The end result is a computation that cannot be parallelized (the steps must be performed sequentially)
// and which has very limited gains when implemented in an ASIC/FPGA. An ASIC/FPGA implementation will
// likely only be marginally faster than a simple CPU core and likely not worth the engineering overhead
// cost. GPU implementations are also prevented since there is no parallelization available. Additionally
// the benefits of modern CPU advances such as Out-of-Order execution are also reduced or eliminated.
// Here the out-of-order execution engine in a high-end CPU will be unable to perform additional
// optimizations since the critical path is physical data movement between a simple computation that takes
// very few clock cycles and waiting on memory for 100s of clock cycles.
//
// If the value of mining this algorithm is high enough it will obviously be optimized as much as possible.
// However the optimial design will likely be a standard light-weight CPU core connected to DRAM. This
// is a common non-speciallized configuration and should be easily available. Additionally unlike SHA256
// mining which is a concentrated logic core and dominated by electricity costs, this modified scrypt
// algorithm will use much less electricty as a component of the total cost of operation since most
// hardware will usually be idle and waiting for data without consuming electricty.
//
// The standard scrypt algorithm consists of 2 sequential passes over the memory data set:
// 1) The first pass initializes memory sequentially
// 2) The second pass performs a moderately complex operation on each data chunk randomly
//
// This modified version consists of 2 passes as follows:
// 1) The first pass initializes memory sequentially (same as above, needed for initialization)
// 2-3) The next 2 passes perform simple computation on random address regions, as described above
// 4) The tenth pass performs the same operation as in pass 2 in standard scrypt
//
// The end result is a computation that scans an off-chip memory data set 2 times, with 25% of the scans
// happening sequentially (and thus are pipeline-able) and 75% of the scans are fully random and
// unpredicable. As a result creating a hardware pipeline for the first initialization pass can only
// improve performance by 25% at the limit, and likely much less. A GPU could in theory work on pass
// 1 in parallel, but again the speed-up is bounded.
//
// The configuration selected is 128MB and 2 added random passes for 3-out-of-4 passes being random, which
// requies 1-2 seconds to compute per hash. It turns out the bitcoin client internally computes the hash
// several times for each block found and received throughout the code as part of it's checking. This means
// new blocks require 10-20 seconds to process today, which is too much. Rather than weaken the Hash algorithm
// the code should be improved to reduce the number of times the block hash is re-computed to check validity.
//
// If over time the configuration is optimized to a GPU or ASIC, this branch will change the PoW by doing one
// or both of the following: a) Increase the memory size, b) Increase the number of light-weight random
// passes (i.e. increase from 2 passes to 8). With a large enough memory footprint the middle set of passes
// should keep a light-weight CPU core as the most cost effective implementation.
/*** Pass 1 - Same as standard scrypt Pass 1 ***/
/* 2: for i = 0 to N - 1 do */
for (i = 0; i < N; i += 2) {
/* 3: V_i <-- X */
blkcpy(&V[i * (32 * r)], X, 128 * r);
/* 4: X <-- H(X) */
blockmix_salsa8(X, Y, Z);
/* 3: V_i <-- X */
blkcpy(&V[(i + 1) * (32 * r)], Y, 128 * r);
/* 4: X <-- H(X) */
blockmix_salsa8(Y, X, Z);
}
/*** Passes 2-3 - New random and simple operations over the dataset 2 times ***/
//Random_CachelineLow = ((uint256 *)Z)[0]; // uint256 has no ^ operator.... using uint64_t instead
//Random_CachelineHigh = ((uint256 *)Z)[1];
Random_Cacheline_Position = *((uint64_t *)X);
Random_DataMixer[0] = *((uint64_t *)(&Y[0]));
Random_DataMixer[1] = *((uint64_t *)(&Y[1*2]));
Random_DataMixer[2] = *((uint64_t *)(&Y[2*2]));
Random_DataMixer[3] = *((uint64_t *)(&Y[3*2]));
Random_DataMixer[4] = *((uint64_t *)(&Y[4*2]));
Random_DataMixer[5] = *((uint64_t *)(&Y[5*2]));
Random_DataMixer[6] = *((uint64_t *)(&Y[6*2]));
Random_DataMixer[7] = *((uint64_t *)(&Y[7*2]));
for( i = 0; i < 2; i++ ) { // 0 to 7 (there are 8 full passes)
for( j = 0; j < N*2; j++ ) { // 0 to 2*N-1 (there are 2*N cachelines)
//Random_Cacheline_Position = (*((uint64_t *)(&Random_CachelineLow))) & 0x000fffff;
//Random_CachelineLow ^= *((uint256 *)(&V[Random_Cacheline_Position*16]));
//Random_CachelineHigh ^= *((uint256 *)(&V[Random_Cacheline_Position*16 + 8]));
//*((uint256 *)(&V[Random_Cacheline_Position*16])) = Random_CachelineLow;
//*((uint256 *)(&V[Random_Cacheline_Position*16 + 8])) = Random_CachelineHigh;
Random_Cacheline_Position ^= Random_DataMixer[0];
Random_Cacheline_Position = Random_Cacheline_Position & 0x00000000001fffffULL; // 0 to 2*N-1 (there are 2*N cachelines)
Random_DataMixer[0] ^= *((uint64_t *)(&V[Random_Cacheline_Position*16]));
Random_DataMixer[1] ^= *((uint64_t *)(&V[Random_Cacheline_Position*16 + 1*2]));
Random_DataMixer[2] ^= *((uint64_t *)(&V[Random_Cacheline_Position*16 + 2*2]));
Random_DataMixer[3] ^= *((uint64_t *)(&V[Random_Cacheline_Position*16 + 3*2]));
Random_DataMixer[4] ^= *((uint64_t *)(&V[Random_Cacheline_Position*16 + 4*2]));
Random_DataMixer[5] ^= *((uint64_t *)(&V[Random_Cacheline_Position*16 + 5*2]));
Random_DataMixer[6] ^= *((uint64_t *)(&V[Random_Cacheline_Position*16 + 6*2]));
Random_DataMixer[7] ^= *((uint64_t *)(&V[Random_Cacheline_Position*16 + 7*2]));
*((uint64_t *)(&V[Random_Cacheline_Position*16])) = Random_DataMixer[7];
*((uint64_t *)(&V[Random_Cacheline_Position*16 + 1*2])) = Random_DataMixer[6];
*((uint64_t *)(&V[Random_Cacheline_Position*16 + 2*2])) = Random_DataMixer[5];
*((uint64_t *)(&V[Random_Cacheline_Position*16 + 3*2])) = Random_DataMixer[4];
*((uint64_t *)(&V[Random_Cacheline_Position*16 + 4*2])) = Random_DataMixer[3];
*((uint64_t *)(&V[Random_Cacheline_Position*16 + 5*2])) = Random_DataMixer[2];
*((uint64_t *)(&V[Random_Cacheline_Position*16 + 6*2])) = Random_DataMixer[1];
*((uint64_t *)(&V[Random_Cacheline_Position*16 + 7*2])) = Random_DataMixer[0];
}
}
/*** Pass 4 - Same as standard scrypt Pass 2 ***/
/* 6: for i = 0 to N - 1 do */
for (i = 0; i < N; i += 2) {
/* 7: j <-- Integerify(X) mod N */
j = integerify(X) & (N - 1);
/* 8: X <-- H(X \xor V_j) */
blkxor(X, &V[j * (32 * r)], 128 * r);
blockmix_salsa8(X, Y, Z);
/* 7: j <-- Integerify(X) mod N */
j = integerify(Y) & (N - 1);
/* 8: X <-- H(X \xor V_j) */
blkxor(Y, &V[j * (32 * r)], 128 * r);
blockmix_salsa8(Y, X, Z);
}
/* 10: B' <-- X */
for (k = 0; k < 32 * r; k++)
le32enc(&B[4 * k], X[k]);
}
int
crypto_1M_1_1_256_scrypt(const uint8_t * passwd, size_t passwdlen,
void * V0, uint8_t * buf, size_t buflen )
{
uint8_t * B;
uint32_t * V;
uint32_t * XY;
size_t r = 1, p = 1; // Not using r and p in this version, let compiler optimize out
uint64_t N = 1024*1024; // 1024*1024=1M count * 128 = 128MB memory space
uint32_t i;
/* Align to 64-byte cachelines */
char B0[128 + 63];
B = (uint8_t *)(((uintptr_t)(B0) + 63) & ~ (uintptr_t)(63));
char XY0[256 + 64 + 63];
XY = (uint32_t *)(((uintptr_t)(XY0) + 63) & ~ (uintptr_t)(63));
V = (uint32_t *)(((uintptr_t)(V0) + 63) & ~ (uintptr_t)(63));
/* 1: (B_0 ... B_{p-1}) <-- PBKDF2(P, S, 1, p * MFLen) */
PBKDF2_SHA256(passwd, passwdlen, passwd, passwdlen, 1, B, p * 128 * r);
/* 2: for i = 0 to p - 1 do */
for (i = 0; i < p; i++) {
/* 3: B_i <-- MF(B_i, N) */
crypto_scrypt_smix(&B[i * 128 * r], N, V, XY);
}
/* 5: DK <-- PBKDF2(P, B, 1, dkLen) */
PBKDF2_SHA256(passwd, passwdlen, B, p * 128 * r, 1, buf, buflen);
/* Success! */
return (0);
}