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
|
/* This Source Code Form is subject to the terms of the Mozilla Public
* License, v. 2.0. If a copy of the MPL was not distributed with this
* file, You can obtain one at http://mozilla.org/MPL/2.0/. */
/* This inlined version is for 32-bit ARM platform only */
#if !defined(__arm__)
#error "This is for ARM only"
#endif
/* 16-bit thumb doesn't work inlined assember version */
#if (!defined(__thumb__) || defined(__thumb2__)) && !defined(__ARM_ARCH_3__)
#include "mpi-priv.h"
#ifdef MP_ASSEMBLY_MULTIPLY
void
s_mpv_mul_d(const mp_digit *a, mp_size a_len, mp_digit b, mp_digit *c)
{
__asm__ __volatile__(
"mov r5, #0\n"
#ifdef __thumb2__
"cbz %1, 2f\n"
#else
"cmp %1, r5\n" /* r5 is 0 now */
"beq 2f\n"
#endif
"1:\n"
"mov r4, #0\n"
"ldr r6, [%0], #4\n"
"umlal r5, r4, r6, %3\n"
"str r5, [%2], #4\n"
"mov r5, r4\n"
"subs %1, #1\n"
"bne 1b\n"
"2:\n"
"str r5, [%2]\n"
: "+r"(a), "+l"(a_len), "+r"(c)
: "r"(b)
: "memory", "cc", "%r4", "%r5", "%r6");
}
void
s_mpv_mul_d_add(const mp_digit *a, mp_size a_len, mp_digit b, mp_digit *c)
{
__asm__ __volatile__(
"mov r5, #0\n"
#ifdef __thumb2__
"cbz %1, 2f\n"
#else
"cmp %1, r5\n" /* r5 is 0 now */
"beq 2f\n"
#endif
"1:\n"
"mov r4, #0\n"
"ldr r6, [%2]\n"
"adds r5, r6\n"
"adc r4, r4, #0\n"
"ldr r6, [%0], #4\n"
"umlal r5, r4, r6, %3\n"
"str r5, [%2], #4\n"
"mov r5, r4\n"
"subs %1, #1\n"
"bne 1b\n"
"2:\n"
"str r5, [%2]\n"
: "+r"(a), "+l"(a_len), "+r"(c)
: "r"(b)
: "memory", "cc", "%r4", "%r5", "%r6");
}
void
s_mpv_mul_d_add_prop(const mp_digit *a, mp_size a_len, mp_digit b, mp_digit *c)
{
if (!a_len)
return;
__asm__ __volatile__(
"mov r5, #0\n"
"1:\n"
"mov r4, #0\n"
"ldr r6, [%2]\n"
"adds r5, r6\n"
"adc r4, r4, #0\n"
"ldr r6, [%0], #4\n"
"umlal r5, r4, r6, %3\n"
"str r5, [%2], #4\n"
"mov r5, r4\n"
"subs %1, #1\n"
"bne 1b\n"
#ifdef __thumb2__
"cbz r4, 3f\n"
#else
"cmp r4, #0\n"
"beq 3f\n"
#endif
"2:\n"
"mov r4, #0\n"
"ldr r6, [%2]\n"
"adds r5, r6\n"
"adc r4, r4, #0\n"
"str r5, [%2], #4\n"
"movs r5, r4\n"
"bne 2b\n"
"3:\n"
: "+r"(a), "+l"(a_len), "+r"(c)
: "r"(b)
: "memory", "cc", "%r4", "%r5", "%r6");
}
#endif
#ifdef MP_ASSEMBLY_SQUARE
void
s_mpv_sqr_add_prop(const mp_digit *pa, mp_size a_len, mp_digit *ps)
{
if (!a_len)
return;
__asm__ __volatile__(
"mov r3, #0\n"
"1:\n"
"mov r4, #0\n"
"ldr r6, [%0], #4\n"
"ldr r5, [%2]\n"
"adds r3, r5\n"
"adc r4, r4, #0\n"
"umlal r3, r4, r6, r6\n" /* w = r3:r4 */
"str r3, [%2], #4\n"
"ldr r5, [%2]\n"
"adds r3, r4, r5\n"
"mov r4, #0\n"
"adc r4, r4, #0\n"
"str r3, [%2], #4\n"
"mov r3, r4\n"
"subs %1, #1\n"
"bne 1b\n"
#ifdef __thumb2__
"cbz r3, 3f\n"
#else
"cmp r3, #0\n"
"beq 3f\n"
#endif
"2:\n"
"mov r4, #0\n"
"ldr r5, [%2]\n"
"adds r3, r5\n"
"adc r4, r4, #0\n"
"str r3, [%2], #4\n"
"movs r3, r4\n"
"bne 2b\n"
"3:"
: "+r"(pa), "+r"(a_len), "+r"(ps)
:
: "memory", "cc", "%r3", "%r4", "%r5", "%r6");
}
#endif
#endif
|
