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
|
/* -*- Mode: Asm -*- */
/* Copyright (c) 2002 Michael Stumpf <mistumpf@de.pepperl-fuchs.com>
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in
the documentation and/or other materials provided with the
distribution.
* Neither the name of the copyright holders nor the names of
contributors may be used to endorse or promote products derived
from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
POSSIBILITY OF SUCH DAMAGE.
*/
/* $Id: mulsf3x.S,v 1.7 2005/11/15 21:18:48 aesok Exp $ */
/*
mulsf3x.S is part of FPlib V 0.3.0 ported to avr-as
for details see readme.fplib
*----------------------------------------------------------------------------------------
*
*--- multiplication kernel : used by other high level functions
* multiply two extended numbers RX = AX * BX
* sign not considered, resulting stored in T by ___mulsf3
* AX rA3 : rA2:rA1:rA0:rAE
* BX rB3 : rB2:rB1:rB0:rBE
* RX rA3 : rA2:rA1:rA0:rAE
*/
#if !defined(__DOXYGEN__)
#include "gasava.inc"
#include "macros.inc"
#include "fplib.inc"
TEXT_SEG(fplib, __mulsf3x)
FUNCTION(__mulsf3x)
GLOBAL(__mulsf3x)
TST rA3
BREQ 2f
TST rB3
BREQ 2f
SUBI rA3,0x7F ; exp(A) now signed char
SUBI rB3,0x7F ; exp(B) now signed char
ADD rA3,rB3 ; add two signed registers
BRVC 1f ; no signed overflow
BRMI 100f ; signed overflow : if now negative -> positive overflow
2:
RJMP _U(__fp_zerox)
1:
SUBI rA3,0x81 ; = ADI,0x7F
CPI rA3,0xFF
BREQ 2b ; no (unsigned overflow) -> signed underflow
#ifdef __AVR_ENHANCED__
; this section of code is used by processors which have MUL instruction
; now multiply mantissa A[rA2:rA1:rA0] * B[rB2:rB1:rB0]
; result : [rr5:rr4:rr3:rr2:rr1:rr0]
; definitions used by calculation
#define rr2 rAE
#define rr3 rTI1
#define rr4 rBE
#define rr5 rB3
#define rmh r1 ; the multiply result high register
#define rml r0 ; the multiply result low register
MUL rA0, rB0 ; 0+0 = 0,1
MOV rr4, rmh ; rr1 == rr4
PUSH rml ; rr0 finished
CLR rr2
MUL rA0, rB1 ; 0+1 = 1,2
ADD rr4, rml
ADC rr2, rmh
CLR rr3
MUL rA1, rB0 ; 1+0 = 1,2,3
ADD rr4, rml
ADC rr2, rmh
ADC rr3, rr3
PUSH rr4 ; rr1 finished
CLR rr4
MUL rA0, rB2 ; 0+2 = 2,3
ADD rr2, rml
ADC rr3, rmh
MUL rA1, rB1 ; 1+1 = 2,3,4
ADD rr2, rml
ADC rr3, rmh
ADC rr4, rr4
CLR rA0
MUL rA2, rB0 ; 2+0 = 2,3,4
ADD rr2, rml
ADC rr3, rmh
ADC rr4, rA0
CLR rr5
MUL rA1, rB2 ; 1+2 = 3,4,5
ADD rr3, rml
ADC rr4, rmh
ADC rr5, rr5
MUL rA2, rB1 ; 2+1 = 3,4,5
ADD rr3, rml
ADC rr4, rmh
ADC rr5, rA0
MUL rA2, rB2 ; 2+2 = 4,5
ADD rr4, rml
ADC rr5, rmh
; converting result to original format
MOV rA2, rr5
MOV rA1, rr4
MOV rA0, rr3
; rr2 is the right one, no need to move
#undef rr5
#undef rr4
#undef rr3
; definitions used by result
#define rr5 rA2
#define rr4 rA1
#define rr3 rA0
// #define rr2 rAE
#define rr1 rT1c
#define rr0 rT0
#define rb5 rTI1
#define rb4 rBE
#define rb3 rB3
; take out the lowest bytes from stack
POP rr1
POP rr0
; +++ 44 instructions +++
#else // __AVR_ENHANCED__
; this section of code is used by processors which have not MUL instruction
; now multiply mantissa A[rA2:rA1:rA0] * B[rb5:rb4:rb3:rB2:rB1:rB0]
; result : [rr5:rr4:rr3:rr2:rr1:rr0]
#define loop rTI0
#define rr5 rA2 /* push rA2 before */
#define rr4 rA1 /* push rA1 before */
#define rr3 rA0 /* push rA0 before */
#define rr2 rAE /* not used, but right position later on */
#define rr1 rT1c /* free to use, needed after pop */
#define rr0 rT0 /* scratch register, needed after pop */
#define rb5 rTI1 /* */
#define rb4 rBE /* not used */
#define rb3 rB3 /* no longer used */
MOV rb4,rA0 ; rB4 not yet needed
CLR rr0
CLR rr1 ; not conclusive zero for high level function
CLR rr2
CLR rr3
CLR rb3
LDI loop,8 ; loop counter
1:
LSR rb4
BRCC 2f
ADD rr0,rB0
adc rr1,rB1
adc rr2,rB2
adc rr3,rb3
2:
ADD rB0,rB0
adc rB1,rB1
adc rB2,rB2
adc rb3,rb3
DEC loop
BRNE 1b
LDI loop,8 ; loop counter
MOV rb5,rA1 ; rb5 not yet needed
CLR rr4 ; rb4 is allready clear
LDI loop,8 ; loop counter
1:
LSR rb5
BRCC 2f
ADD rr1,rB1
adc rr2,rB2
adc rr3,rb3
adc rr4,rb4
2:
ADD rB1,rB1
adc rB2,rB2
adc rb3,rb3
adc rb4,rb4
DEC loop
BRNE 1b
MOV loop,rA2 ; loop no longer needed, check
CLR rr5 ; rb5 is allready clear
1:
LSR loop
BRCC 2f
ADD rr2,rB2
adc rr3,rb3
adc rr4,rb4
adc rr5,rb5
2:
ADD rB2,rB2
adc rb3,rb3
adc rb4,rb4
adc rb5,rb5
TST loop
BRNE 1b
; +++ 49 instructions +++
#endif // __AVR_ENHANCED__
; multiplication done : result in rr5:rr4:rr3:rr2:rr1:rr0
; = rA2:rA1:rA0:rAE:rr1:rr0
; normalize 1.0 * 1.0 = 1.0
; 0x800000 * 0x800000 = 0x40 00 00 00 00 00
; 1.999999 * 1.999999 = 3.99999 ~ 4.0
; 0xFFFFFF * 0xFFFFFF = 0xFF FF FE 00 00 01
TST rA2
BRPL 1f ; if MSB erg is clr
INC rA3 ;
BRNE 2f
100:
RJMP _U(__fp_nanx) ; returns to ___mulsf3 or a high level function : rT1c ok
1:
ADD rr0,rr0
adc rr1,rr1
adc rAE,rAE
adc rA0,rA0
adc rA1,rA1
adc rA2,rA2
2:
OR rr0,rr1 ; rr0 = rT0 which holds the mantissae extension beyond rAE
RET
ENDFUNC
#endif /* not __DOXYGEN__ */
|
