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
use super::*;
use core::arch::x86_64::*;
#[macro_use]
#[path = "loop_macros.rs"]
mod loop_macros;
macro_rules! mm256_rorx_epi32 {
($x:expr, $shift:expr) => {{
#[allow(unused_unsafe)]
unsafe {
_mm256_xor_si256(
_mm256_srli_epi32($x, $shift),
_mm256_slli_epi32($x, 32 - $shift),
)
}
}};
}
#[inline(always)]
fn g4(
va: &mut __m256i,
vb: &mut __m256i,
vc: &mut __m256i,
vd: &mut __m256i,
x: __m256i,
y: __m256i,
) {
/*
FUNCTION G( v[0..15], a, b, c, d, x, y )
|
| v[a] := (v[a] + v[b] + x) mod 2**32
| v[d] := (v[d] ^ v[a]) >>> 16
| v[c] := (v[c] + v[d]) mod 2**32
| v[b] := (v[b] ^ v[c]) >>> 12
| v[a] := (v[a] + v[b] + y) mod 2**32
| v[d] := (v[d] ^ v[a]) >>> 8
| v[c] := (v[c] + v[d]) mod 2**32
| v[b] := (v[b] ^ v[c]) >>> 7
|
| RETURN v[0..15]
|
END FUNCTION. */
unsafe {
*va = _mm256_add_epi32(*va, _mm256_add_epi32(*vb, x));
*vd = _mm256_xor_si256(*vd, *va);
*vd = mm256_rorx_epi32!(*vd, 16);
*vc = _mm256_add_epi32(*vc, *vd);
*vb = _mm256_xor_si256(*vb, *vc);
*vb = mm256_rorx_epi32!(*vb, 12);
*va = _mm256_add_epi32(*va, _mm256_add_epi32(*vb, y));
*vd = _mm256_xor_si256(*vd, *va);
*vd = mm256_rorx_epi32!(*vd, 8);
*vc = _mm256_add_epi32(*vc, *vd);
*vb = _mm256_xor_si256(*vb, *vc);
*vb = mm256_rorx_epi32!(*vb, 7);
}
}
#[inline(always)]
pub(crate) fn compress_mb8<const CONSTANT_WORD_COUNT: usize, const PATCH_1: usize>(
v: &mut [__m256i; 16],
block_template: &[u32; 16],
patch_1: __m256i,
) {
/*
FUNCTION BLAKE3_COMPRESS( h[0..7], m[0..15], t, len, flags )
FUNCTION BLAKE3_COMPRESS( h[0..7], m[0..15], t, len, flags )
|
| // Initialize local 16-word array v[0..15]
| v[0..7] := h[0..7] // 8 words from the state.
| v[8..11] := IV[0..3] // 4 words from the IV.
|
| v[12] := t[0] // Low word of the counter.
| v[13] := t[1] // High word of the counter.
| v[14] := len // Application data length.
| v[15] := flags // Flags.
|
| // Cryptographic mixing
| FOR i = 0 TO 6 DO // 7 rounds.
| |
| | v := G( v, 0, 4, 8, 12, m[ 0], m[ 1] )
| | v := G( v, 1, 5, 9, 13, m[ 2], m[ 3] )
| | v := G( v, 2, 6, 10, 14, m[ 4], m[ 5] )
| | v := G( v, 3, 7, 11, 15, m[ 6], m[ 7] )
| |
| | v := G( v, 0, 5, 10, 15, m[ 8], m[ 9] )
| | v := G( v, 1, 6, 11, 12, m[10], m[11] )
| | v := G( v, 2, 7, 8, 13, m[12], m[13] )
| | v := G( v, 3, 4, 9, 14, m[14], m[15] )
| |
| | PERMUTE(m) // Apply the permutation.
| |
| END FOR
|
| // Compute the output state (untruncated)
| FOR i = 0 TO 7 DO
| | v[i] := v[i] ^ v[i + 8]
| | v[i + 8] := v[i + 8] ^ h[i]
| END FOR.
|
| RETURN v
|
END FUNCTION.
|
END FUNCTION. */
unsafe {
repeat7!(i, {
macro_rules! g4 {
($f:ident; $a:literal, $b:literal, $c:literal, $d:literal, $x:literal, $y:literal) => {{
let [va, vb, vc, vd] = v.get_disjoint_unchecked_mut([$a, $b, $c, $d]);
let ix = MESSAGE_SCHEDULE[i][$x];
let iy = MESSAGE_SCHEDULE[i][$y];
$f(
va,
vb,
vc,
vd,
if ix == PATCH_1 {
patch_1
} else {
_mm256_set1_epi32(block_template[ix] as _)
},
if iy == PATCH_1 {
patch_1
} else {
_mm256_set1_epi32(block_template[iy] as _)
},
);
}};
($a:literal, $b:literal, $c:literal, $d:literal, $x:literal, $y:literal) => {{
g4!(g4; $a, $b, $c, $d, $x, $y);
}};
}
if i > 0 || CONSTANT_WORD_COUNT < 2 {
g4!(0, 4, 8, 12, 0, 1);
}
if i > 0 || CONSTANT_WORD_COUNT < 4 {
g4!(1, 5, 9, 13, 2, 3);
}
if i > 0 || CONSTANT_WORD_COUNT < 6 {
g4!(2, 6, 10, 14, 4, 5);
}
if i > 0 || CONSTANT_WORD_COUNT < 8 {
g4!(3, 7, 11, 15, 6, 7);
}
if i > 0 || CONSTANT_WORD_COUNT < 10 {
g4!(0, 5, 10, 15, 8, 9);
}
if i > 0 || CONSTANT_WORD_COUNT < 12 {
g4!(1, 6, 11, 12, 10, 11);
}
if i > 0 || CONSTANT_WORD_COUNT < 14 {
g4!(2, 7, 8, 13, 12, 13);
}
if i > 0 || CONSTANT_WORD_COUNT < 16 {
g4!(3, 4, 9, 14, 14, 15);
}
});
repeat8!(i, {
v[i] = _mm256_xor_si256(v[i], v[i + 8]);
});
}
}
#[cfg(target_feature = "avx2")]
#[cfg(test)]
mod tests {
use super::*;
use blake3::Hasher;
#[test]
fn test_g_function() {
let mut state = core::array::from_fn(|i| crate::sha256::IV[i % 8].wrapping_add(i as u32));
let mut state_v: [_; 16] =
core::array::from_fn(|i| unsafe { _mm256_set1_epi32(state[i] as _) });
g(
&mut state,
0,
4,
8,
12,
crate::sha256::IV[0],
crate::sha256::IV[1],
);
let [va, vb, vc, vd] = state_v.get_disjoint_mut([0, 4, 8, 12]).unwrap();
g4(
va,
vb,
vc,
vd,
unsafe { _mm256_set1_epi32(crate::sha256::IV[0] as _) },
unsafe { _mm256_set1_epi32(crate::sha256::IV[1] as _) },
);
for i in 0..16 {
assert_eq!(
unsafe { _mm_extract_epi32(_mm256_castsi256_si128(state_v[i]), 0) as u32 },
state[i],
"word {}: expected: {:08x}, results: {:08x}",
i,
state[i],
unsafe { _mm_extract_epi32(_mm256_castsi256_si128(state_v[i]), 0) as u32 }
);
}
}
#[test]
fn test_compress_mb8() {
let mut v = [0u32; 16];
v[..8].copy_from_slice(&crate::blake3::IV);
v[8..12].copy_from_slice(&crate::blake3::IV[..4]);
v[12] = 0;
v[13] = 0;
v[14] = 4;
v[15] = 0x0b;
let mut v = core::array::from_fn(|i| unsafe { _mm256_set1_epi32(v[i] as _) });
let mut block = [0u32; 16];
block[0] = u32::from_le_bytes(*b"IETF");
compress_mb8::<0, 8>(&mut v, &block, unsafe { _mm256_setzero_si256() });
let expected = [
0x1edea283, 0xabe6f4e6, 0x24896868, 0xcfc04e8f, 0x9470c54c, 0xff82a646, 0xd6b4cbd1,
0xe2815116,
];
let mut results = [0u32; 8];
for i in 0..8 {
unsafe {
results[i] = _mm_extract_epi32(_mm256_castsi256_si128(v[i]), 0) as u32;
}
}
assert_eq!(
results, expected,
"expected: {:08x?}, results: {:08x?}",
expected, results
);
let mut hasher = Hasher::new();
hasher.update(b"IETF");
let hash = hasher.finalize();
let hash = hash.as_bytes();
let mut expected = [0u32; 8];
for i in 0..8 {
expected[i] = u32::from_le_bytes(hash[i * 4..i * 4 + 4].try_into().unwrap());
}
assert_eq!(results, expected);
}
}