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
|
dnl AMD64 mpn_mul_2 optimised for Intel Haswell.
dnl Contributed to the GNU project by Torbjörn Granlund.
dnl Copyright 2013 Free Software Foundation, Inc.
dnl This file is part of the GNU MP Library.
dnl
dnl The GNU MP Library is free software; you can redistribute it and/or modify
dnl it under the terms of either:
dnl
dnl * the GNU Lesser General Public License as published by the Free
dnl Software Foundation; either version 3 of the License, or (at your
dnl option) any later version.
dnl
dnl or
dnl
dnl * the GNU General Public License as published by the Free Software
dnl Foundation; either version 2 of the License, or (at your option) any
dnl later version.
dnl
dnl or both in parallel, as here.
dnl
dnl The GNU MP Library is distributed in the hope that it will be useful, but
dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
dnl for more details.
dnl
dnl You should have received copies of the GNU General Public License and the
dnl GNU Lesser General Public License along with the GNU MP Library. If not,
dnl see https://www.gnu.org/licenses/.
include(`../config.m4')
C cycles/limb
C AMD K8,K9 -
C AMD K10 -
C AMD bull -
C AMD pile -
C AMD steam -
C AMD excavator -
C AMD bobcat -
C AMD jaguar -
C Intel P4 -
C Intel core -
C Intel NHM -
C Intel SBR -
C Intel IBR -
C Intel HWL 3.74
C Intel BWL 4.21
C Intel SKL 4.20
C Intel atom -
C Intel SLM -
C VIA nano -
C The loop of this code is the result of running a code generation and
C optimisation tool suite written by David Harvey and Torbjörn Granlund.
C TODO
C * Move test and jcc together, for insn fusion.
define(`rp', `%rdi')
define(`up', `%rsi')
define(`n_param',`%rdx')
define(`vp', `%rcx')
define(`v0', `%r8')
define(`v1', `%r9')
define(`w0', `%rbx')
define(`w1', `%rcx')
define(`w2', `%rbp')
define(`w3', `%r10')
define(`n', `%r11')
ABI_SUPPORT(DOS64)
ABI_SUPPORT(STD64)
ASM_START()
TEXT
ALIGN(32)
PROLOGUE(mpn_mul_2)
FUNC_ENTRY(4)
push %rbx
push %rbp
mov (vp), v0
mov 8(vp), v1
lea 3(n_param), n
shr $2, n
test $1, R8(n_param)
jnz L(bx1)
L(bx0): xor w0, w0
test $2, R8(n_param)
mov (up), %rdx
mulx( v0, w2, w1)
jz L(lo0)
L(b10): lea -16(rp), rp
lea -16(up), up
jmp L(lo2)
L(bx1): xor w2, w2
test $2, R8(n_param)
mov (up), %rdx
mulx( v0, w0, w3)
jnz L(b11)
L(b01): lea -24(rp), rp
lea 8(up), up
jmp L(lo1)
L(b11): lea -8(rp), rp
lea -8(up), up
jmp L(lo3)
ALIGN(16)
L(top): mulx( v1, %rax, w0)
add %rax, w2 C 0
mov (up), %rdx
mulx( v0, %rax, w1)
adc $0, w0 C 1
add %rax, w2 C 0
adc $0, w1 C 1
add w3, w2 C 0
L(lo0): mov w2, (rp) C 0
adc $0, w1 C 1
mulx( v1, %rax, w2)
add %rax, w0 C 1
mov 8(up), %rdx
adc $0, w2 C 2
mulx( v0, %rax, w3)
add %rax, w0 C 1
adc $0, w3 C 2
add w1, w0 C 1
L(lo3): mov w0, 8(rp) C 1
adc $0, w3 C 2
mulx( v1, %rax, w0)
add %rax, w2 C 2
mov 16(up), %rdx
mulx( v0, %rax, w1)
adc $0, w0 C 3
add %rax, w2 C 2
adc $0, w1 C 3
add w3, w2 C 2
L(lo2): mov w2, 16(rp) C 2
adc $0, w1 C 3
mulx( v1, %rax, w2)
add %rax, w0 C 3
mov 24(up), %rdx
adc $0, w2 C 4
mulx( v0, %rax, w3)
add %rax, w0 C 3
adc $0, w3 C 4
add w1, w0 C 3
lea 32(up), up
L(lo1): mov w0, 24(rp) C 3
adc $0, w3 C 4
dec n
lea 32(rp), rp
jnz L(top)
L(end): mulx( v1, %rdx, %rax)
add %rdx, w2
adc $0, %rax
add w3, w2
mov w2, (rp)
adc $0, %rax
pop %rbp
pop %rbx
FUNC_EXIT()
ret
EPILOGUE()
|