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
|
; mc88110 __mpn_sub_n -- Subtract two limb vectors of the same length > 0 and
; store difference in a third limb vector.
; Copyright (C) 1995, 1996 Free Software Foundation, Inc.
; This file is part of the GNU MP Library.
; The GNU MP Library is free software; you can redistribute it and/or modify
; it under the terms of the GNU Library General Public License as published by
; the Free Software Foundation; either version 2 of the License, or (at your
; option) any later version.
; The GNU MP Library is distributed in the hope that it will be useful, but
; WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
; or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Library General Public
; License for more details.
; You should have received a copy of the GNU Library General Public License
; along with the GNU MP Library; see the file COPYING.LIB. If not, write to
; the Free Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
; INPUT PARAMETERS
#define res_ptr r2
#define s1_ptr r3
#define s2_ptr r4
#define size r5
#include "sysdep.h"
text
align 16
global C_SYMBOL_NAME(__mpn_sub_n)
C_SYMBOL_NAME(__mpn_sub_n):
subu.co r0,r0,r0 ; set cy flag
xor r12,s2_ptr,res_ptr
bb1 2,r12,L1
; ** V1a **
L0: bb0 2,res_ptr,L_v1 ; branch if res_ptr is aligned
/* Add least significant limb separately to align res_ptr and s2_ptr */
ld r10,s1_ptr,0
addu s1_ptr,s1_ptr,4
ld r8,s2_ptr,0
addu s2_ptr,s2_ptr,4
subu size,size,1
subu.co r6,r10,r8
st r6,res_ptr,0
addu res_ptr,res_ptr,4
L_v1: cmp r12,size,2
bb1 lt,r12,Lend2
ld r10,s1_ptr,0
ld r12,s1_ptr,4
ld.d r8,s2_ptr,0
subu size,size,10
bcnd lt0,size,Lfin1
/* Add blocks of 8 limbs until less than 8 limbs remain */
align 8
Loop1: subu size,size,8
subu.cio r6,r10,r8
ld r10,s1_ptr,8
subu.cio r7,r12,r9
ld r12,s1_ptr,12
ld.d r8,s2_ptr,8
st.d r6,res_ptr,0
subu.cio r6,r10,r8
ld r10,s1_ptr,16
subu.cio r7,r12,r9
ld r12,s1_ptr,20
ld.d r8,s2_ptr,16
st.d r6,res_ptr,8
subu.cio r6,r10,r8
ld r10,s1_ptr,24
subu.cio r7,r12,r9
ld r12,s1_ptr,28
ld.d r8,s2_ptr,24
st.d r6,res_ptr,16
subu.cio r6,r10,r8
ld r10,s1_ptr,32
subu.cio r7,r12,r9
ld r12,s1_ptr,36
addu s1_ptr,s1_ptr,32
ld.d r8,s2_ptr,32
addu s2_ptr,s2_ptr,32
st.d r6,res_ptr,24
addu res_ptr,res_ptr,32
bcnd ge0,size,Loop1
Lfin1: addu size,size,8-2
bcnd lt0,size,Lend1
/* Add blocks of 2 limbs until less than 2 limbs remain */
Loope1: subu.cio r6,r10,r8
ld r10,s1_ptr,8
subu.cio r7,r12,r9
ld r12,s1_ptr,12
ld.d r8,s2_ptr,8
st.d r6,res_ptr,0
subu size,size,2
addu s1_ptr,s1_ptr,8
addu s2_ptr,s2_ptr,8
addu res_ptr,res_ptr,8
bcnd ge0,size,Loope1
Lend1: subu.cio r6,r10,r8
subu.cio r7,r12,r9
st.d r6,res_ptr,0
bb0 0,size,Lret1
/* Add last limb */
ld r10,s1_ptr,8
ld r8,s2_ptr,8
subu.cio r6,r10,r8
st r6,res_ptr,8
Lret1: addu.ci r2,r0,r0 ; return carry-out from most sign. limb
jmp.n r1
xor r2,r2,1
L1: xor r12,s1_ptr,res_ptr
bb1 2,r12,L2
; ** V1b **
bb0 2,res_ptr,L_v1b ; branch if res_ptr is aligned
/* Add least significant limb separately to align res_ptr and s1_ptr */
ld r10,s2_ptr,0
addu s2_ptr,s2_ptr,4
ld r8,s1_ptr,0
addu s1_ptr,s1_ptr,4
subu size,size,1
subu.co r6,r8,r10
st r6,res_ptr,0
addu res_ptr,res_ptr,4
L_v1b: cmp r12,size,2
bb1 lt,r12,Lend2
ld r10,s2_ptr,0
ld r12,s2_ptr,4
ld.d r8,s1_ptr,0
subu size,size,10
bcnd lt0,size,Lfin1b
/* Add blocks of 8 limbs until less than 8 limbs remain */
align 8
Loop1b: subu size,size,8
subu.cio r6,r8,r10
ld r10,s2_ptr,8
subu.cio r7,r9,r12
ld r12,s2_ptr,12
ld.d r8,s1_ptr,8
st.d r6,res_ptr,0
subu.cio r6,r8,r10
ld r10,s2_ptr,16
subu.cio r7,r9,r12
ld r12,s2_ptr,20
ld.d r8,s1_ptr,16
st.d r6,res_ptr,8
subu.cio r6,r8,r10
ld r10,s2_ptr,24
subu.cio r7,r9,r12
ld r12,s2_ptr,28
ld.d r8,s1_ptr,24
st.d r6,res_ptr,16
subu.cio r6,r8,r10
ld r10,s2_ptr,32
subu.cio r7,r9,r12
ld r12,s2_ptr,36
addu s2_ptr,s2_ptr,32
ld.d r8,s1_ptr,32
addu s1_ptr,s1_ptr,32
st.d r6,res_ptr,24
addu res_ptr,res_ptr,32
bcnd ge0,size,Loop1b
Lfin1b: addu size,size,8-2
bcnd lt0,size,Lend1b
/* Add blocks of 2 limbs until less than 2 limbs remain */
Loope1b:subu.cio r6,r8,r10
ld r10,s2_ptr,8
subu.cio r7,r9,r12
ld r12,s2_ptr,12
ld.d r8,s1_ptr,8
st.d r6,res_ptr,0
subu size,size,2
addu s1_ptr,s1_ptr,8
addu s2_ptr,s2_ptr,8
addu res_ptr,res_ptr,8
bcnd ge0,size,Loope1b
Lend1b: subu.cio r6,r8,r10
subu.cio r7,r9,r12
st.d r6,res_ptr,0
bb0 0,size,Lret1b
/* Add last limb */
ld r10,s2_ptr,8
ld r8,s1_ptr,8
subu.cio r6,r8,r10
st r6,res_ptr,8
Lret1b: addu.ci r2,r0,r0 ; return carry-out from most sign. limb
jmp.n r1
xor r2,r2,1
; ** V2 **
/* If we come here, the alignment of s1_ptr and res_ptr as well as the
alignment of s2_ptr and res_ptr differ. Since there are only two ways
things can be aligned (that we care about) we now know that the alignment
of s1_ptr and s2_ptr are the same. */
L2: cmp r12,size,1
bb1 eq,r12,Ljone
bb0 2,s1_ptr,L_v2 ; branch if s1_ptr is aligned
/* Add least significant limb separately to align res_ptr and s2_ptr */
ld r10,s1_ptr,0
addu s1_ptr,s1_ptr,4
ld r8,s2_ptr,0
addu s2_ptr,s2_ptr,4
subu size,size,1
subu.co r6,r10,r8
st r6,res_ptr,0
addu res_ptr,res_ptr,4
L_v2: subu size,size,8
bcnd lt0,size,Lfin2
/* Add blocks of 8 limbs until less than 8 limbs remain */
align 8
Loop2: subu size,size,8
ld.d r8,s1_ptr,0
ld.d r6,s2_ptr,0
subu.cio r8,r8,r6
st r8,res_ptr,0
subu.cio r9,r9,r7
st r9,res_ptr,4
ld.d r8,s1_ptr,8
ld.d r6,s2_ptr,8
subu.cio r8,r8,r6
st r8,res_ptr,8
subu.cio r9,r9,r7
st r9,res_ptr,12
ld.d r8,s1_ptr,16
ld.d r6,s2_ptr,16
subu.cio r8,r8,r6
st r8,res_ptr,16
subu.cio r9,r9,r7
st r9,res_ptr,20
ld.d r8,s1_ptr,24
ld.d r6,s2_ptr,24
subu.cio r8,r8,r6
st r8,res_ptr,24
subu.cio r9,r9,r7
st r9,res_ptr,28
addu s1_ptr,s1_ptr,32
addu s2_ptr,s2_ptr,32
addu res_ptr,res_ptr,32
bcnd ge0,size,Loop2
Lfin2: addu size,size,8-2
bcnd lt0,size,Lend2
Loope2: ld.d r8,s1_ptr,0
ld.d r6,s2_ptr,0
subu.cio r8,r8,r6
st r8,res_ptr,0
subu.cio r9,r9,r7
st r9,res_ptr,4
subu size,size,2
addu s1_ptr,s1_ptr,8
addu s2_ptr,s2_ptr,8
addu res_ptr,res_ptr,8
bcnd ge0,size,Loope2
Lend2: bb0 0,size,Lret2
/* Add last limb */
Ljone: ld r10,s1_ptr,0
ld r8,s2_ptr,0
subu.cio r6,r10,r8
st r6,res_ptr,0
Lret2: addu.ci r2,r0,r0 ; return carry-out from most sign. limb
jmp.n r1
xor r2,r2,1
|
