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
// generated source. do not edit.
#![allow(non_upper_case_globals, unused_macros, unused_imports)]
use crate::low::macros::*;
// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
// SPDX-License-Identifier: Apache-2.0 OR ISC OR MIT-0
// ----------------------------------------------------------------------------
// Reduce modulo group order, z := x mod n_256
// Input x[4]; output z[4]
//
// extern void bignum_mod_n256_4(uint64_t z[static 4], const uint64_t x[static 4]);
//
// Reduction is modulo the group order of the NIST curve P-256.
//
// Standard ARM ABI: X0 = z, X1 = x
// ----------------------------------------------------------------------------
macro_rules! z {
() => {
"x0"
};
}
macro_rules! x {
() => {
"x1"
};
}
macro_rules! n0 {
() => {
"x2"
};
}
macro_rules! n1 {
() => {
"x3"
};
}
macro_rules! n2 {
() => {
"x4"
};
}
macro_rules! n3 {
() => {
"x5"
};
}
macro_rules! d0 {
() => {
"x6"
};
}
macro_rules! d1 {
() => {
"x7"
};
}
macro_rules! d2 {
() => {
"x8"
};
}
macro_rules! d3 {
() => {
"x9"
};
}
// Loading large constants
macro_rules! movbig {
($nn:expr, $n3:expr, $n2:expr, $n1:expr, $n0:expr) => { Q!(
"movz " $nn ", " $n0 ";\n"
"movk " $nn ", " $n1 ", lsl #16;\n"
"movk " $nn ", " $n2 ", lsl #32;\n"
"movk " $nn ", " $n3 ", lsl #48"
)}
}
/// Reduce modulo group order, z := x mod n_256
///
/// Input x[4]; output z[4]
///
/// Reduction is modulo the group order of the NIST curve P-256.
pub(crate) fn bignum_mod_n256(z: &mut [u64; 4], x: &[u64; 4]) {
// SAFETY: inline assembly. see [crate::low::inline_assembly_safety] for safety info.
unsafe {
core::arch::asm!(
// Load the complicated three words of n_256, the other being all 1s
movbig!(n0!(), "#0xf3b9", "#0xcac2", "#0xfc63", "#0x2551"),
movbig!(n1!(), "#0xbce6", "#0xfaad", "#0xa717", "#0x9e84"),
Q!(" mov " n3!() ", #0xffffffff00000000"),
// Load the input number
Q!(" ldp " d0!() ", " d1!() ", [" x!() "]"),
Q!(" ldp " d2!() ", " d3!() ", [" x!() ", #16]"),
// Do the subtraction. Since word 2 of n_256 is all 1s, that can be
// done by adding zero with carry, thanks to the inverted carry.
Q!(" subs " n0!() ", " d0!() ", " n0!()),
Q!(" sbcs " n1!() ", " d1!() ", " n1!()),
Q!(" adcs " n2!() ", " d2!() ", xzr"),
Q!(" sbcs " n3!() ", " d3!() ", " n3!()),
// Now if the carry is *clear* (inversion at work) the subtraction carried
// and hence we should have done nothing, so we reset each n_i = d_i
Q!(" csel " n0!() ", " d0!() ", " n0!() ", cc"),
Q!(" csel " n1!() ", " d1!() ", " n1!() ", cc"),
Q!(" csel " n2!() ", " d2!() ", " n2!() ", cc"),
Q!(" csel " n3!() ", " d3!() ", " n3!() ", cc"),
// Store the end result
Q!(" stp " n0!() ", " n1!() ", [" z!() "]"),
Q!(" stp " n2!() ", " n3!() ", [" z!() ", #16]"),
inout("x0") z.as_mut_ptr() => _,
inout("x1") x.as_ptr() => _,
// clobbers
out("x2") _,
out("x3") _,
out("x4") _,
out("x5") _,
out("x6") _,
out("x7") _,
out("x8") _,
out("x9") _,
)
};
}