divsi3.S
5.4 KB
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
/*
* Copyright 2004-2009 Analog Devices Inc.
*
* Licensed under the Clear BSD license or the GPL-2 (or later)
*
* 16 / 32 bit signed division.
* Special cases :
* 1) If(numerator == 0)
* return 0
* 2) If(denominator ==0)
* return positive max = 0x7fffffff
* 3) If(numerator == denominator)
* return 1
* 4) If(denominator ==1)
* return numerator
* 5) If(denominator == -1)
* return -numerator
*
* Operand : R0 - Numerator (i)
* R1 - Denominator (i)
* R0 - Quotient (o)
* Registers Used : R2-R7,P0-P2
*
*/
.global ___divsi3;
.type ___divsi3, STT_FUNC;
#ifdef CONFIG_ARITHMETIC_OPS_L1
.section .l1.text
#else
.text
#endif
.align 2;
___divsi3 :
R3 = R0 ^ R1;
R0 = ABS R0;
CC = V;
r3 = rot r3 by -1;
r1 = abs r1; /* now both positive, r3.30 means "negate result",
** r3.31 means overflow, add one to result
*/
cc = r0 < r1;
if cc jump .Lret_zero;
r2 = r1 >> 15;
cc = r2;
if cc jump .Lidents;
r2 = r1 << 16;
cc = r2 <= r0;
if cc jump .Lidents;
DIVS(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
DIVQ(R0, R1);
R0 = R0.L (Z);
r1 = r3 >> 31; /* add overflow issue back in */
r0 = r0 + r1;
r1 = -r0;
cc = bittst(r3, 30);
if cc r0 = r1;
RTS;
/* Can't use the primitives. Test common identities.
** If the identity is true, return the value in R2.
*/
.Lidents:
CC = R1 == 0; /* check for divide by zero */
IF CC JUMP .Lident_return;
CC = R0 == 0; /* check for division of zero */
IF CC JUMP .Lzero_return;
CC = R0 == R1; /* check for identical operands */
IF CC JUMP .Lident_return;
CC = R1 == 1; /* check for divide by 1 */
IF CC JUMP .Lident_return;
R2.L = ONES R1;
R2 = R2.L (Z);
CC = R2 == 1;
IF CC JUMP .Lpower_of_two;
/* Identities haven't helped either.
** Perform the full division process.
*/
P1 = 31; /* Set loop counter */
[--SP] = (R7:5); /* Push registers R5-R7 */
R2 = -R1;
[--SP] = R2;
R2 = R0 << 1; /* R2 lsw of dividend */
R6 = R0 ^ R1; /* Get sign */
R5 = R6 >> 31; /* Shift sign to LSB */
R0 = 0 ; /* Clear msw partial remainder */
R2 = R2 | R5; /* Shift quotient bit */
R6 = R0 ^ R1; /* Get new quotient bit */
LSETUP(.Llst,.Llend) LC0 = P1; /* Setup loop */
.Llst: R7 = R2 >> 31; /* record copy of carry from R2 */
R2 = R2 << 1; /* Shift 64 bit dividend up by 1 bit */
R0 = R0 << 1 || R5 = [SP];
R0 = R0 | R7; /* and add carry */
CC = R6 < 0; /* Check quotient(AQ) */
/* we might be subtracting divisor (AQ==0) */
IF CC R5 = R1; /* or we might be adding divisor (AQ==1)*/
R0 = R0 + R5; /* do add or subtract, as indicated by AQ */
R6 = R0 ^ R1; /* Generate next quotient bit */
R5 = R6 >> 31;
/* Assume AQ==1, shift in zero */
BITTGL(R5,0); /* tweak AQ to be what we want to shift in */
.Llend: R2 = R2 + R5; /* and then set shifted-in value to
** tweaked AQ.
*/
r1 = r3 >> 31;
r2 = r2 + r1;
cc = bittst(r3,30);
r0 = -r2;
if !cc r0 = r2;
SP += 4;
(R7:5)= [SP++]; /* Pop registers R6-R7 */
RTS;
.Lident_return:
CC = R1 == 0; /* check for divide by zero => 0x7fffffff */
R2 = -1 (X);
R2 >>= 1;
IF CC JUMP .Ltrue_ident_return;
CC = R0 == R1; /* check for identical operands => 1 */
R2 = 1 (Z);
IF CC JUMP .Ltrue_ident_return;
R2 = R0; /* assume divide by 1 => numerator */
/*FALLTHRU*/
.Ltrue_ident_return:
R0 = R2; /* Return an identity value */
R2 = -R2;
CC = bittst(R3,30);
IF CC R0 = R2;
.Lzero_return:
RTS; /* ...including zero */
.Lpower_of_two:
/* Y has a single bit set, which means it's a power of two.
** That means we can perform the division just by shifting
** X to the right the appropriate number of bits
*/
/* signbits returns the number of sign bits, minus one.
** 1=>30, 2=>29, ..., 0x40000000=>0. Which means we need
** to shift right n-signbits spaces. It also means 0x80000000
** is a special case, because that *also* gives a signbits of 0
*/
R2 = R0 >> 31;
CC = R1 < 0;
IF CC JUMP .Ltrue_ident_return;
R1.l = SIGNBITS R1;
R1 = R1.L (Z);
R1 += -30;
R0 = LSHIFT R0 by R1.L;
r1 = r3 >> 31;
r0 = r0 + r1;
R2 = -R0; // negate result if necessary
CC = bittst(R3,30);
IF CC R0 = R2;
RTS;
.Lret_zero:
R0 = 0;
RTS;
.size ___divsi3, .-___divsi3