Merge pull request #156 from p12nGH/noncontiguous_harts
[riscv-isa-sim.git] / softfloat / f128_rem.c
1
2 /*============================================================================
3
4 This C source file is part of the SoftFloat IEEE Floating-Point Arithmetic
5 Package, Release 3d, by John R. Hauser.
6
7 Copyright 2011, 2012, 2013, 2014, 2015, 2016 The Regents of the University of
8 California. All rights reserved.
9
10 Redistribution and use in source and binary forms, with or without
11 modification, are permitted provided that the following conditions are met:
12
13 1. Redistributions of source code must retain the above copyright notice,
14 this list of conditions, and the following disclaimer.
15
16 2. Redistributions in binary form must reproduce the above copyright notice,
17 this list of conditions, and the following disclaimer in the documentation
18 and/or other materials provided with the distribution.
19
20 3. Neither the name of the University nor the names of its contributors may
21 be used to endorse or promote products derived from this software without
22 specific prior written permission.
23
24 THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS "AS IS", AND ANY
25 EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
26 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, ARE
27 DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE FOR ANY
28 DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
29 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
30 LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
31 ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
32 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
33 SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
34
35 =============================================================================*/
36
37 #include <stdbool.h>
38 #include <stdint.h>
39 #include "platform.h"
40 #include "internals.h"
41 #include "specialize.h"
42 #include "softfloat.h"
43
44 float128_t f128_rem( float128_t a, float128_t b )
45 {
46 union ui128_f128 uA;
47 uint_fast64_t uiA64, uiA0;
48 bool signA;
49 int_fast32_t expA;
50 struct uint128 sigA;
51 union ui128_f128 uB;
52 uint_fast64_t uiB64, uiB0;
53 int_fast32_t expB;
54 struct uint128 sigB;
55 struct exp32_sig128 normExpSig;
56 struct uint128 rem;
57 int_fast32_t expDiff;
58 uint_fast32_t q, recip32;
59 uint_fast64_t q64;
60 struct uint128 term, altRem, meanRem;
61 bool signRem;
62 struct uint128 uiZ;
63 union ui128_f128 uZ;
64
65 /*------------------------------------------------------------------------
66 *------------------------------------------------------------------------*/
67 uA.f = a;
68 uiA64 = uA.ui.v64;
69 uiA0 = uA.ui.v0;
70 signA = signF128UI64( uiA64 );
71 expA = expF128UI64( uiA64 );
72 sigA.v64 = fracF128UI64( uiA64 );
73 sigA.v0 = uiA0;
74 uB.f = b;
75 uiB64 = uB.ui.v64;
76 uiB0 = uB.ui.v0;
77 expB = expF128UI64( uiB64 );
78 sigB.v64 = fracF128UI64( uiB64 );
79 sigB.v0 = uiB0;
80 /*------------------------------------------------------------------------
81 *------------------------------------------------------------------------*/
82 if ( expA == 0x7FFF ) {
83 if (
84 (sigA.v64 | sigA.v0) || ((expB == 0x7FFF) && (sigB.v64 | sigB.v0))
85 ) {
86 goto propagateNaN;
87 }
88 goto invalid;
89 }
90 if ( expB == 0x7FFF ) {
91 if ( sigB.v64 | sigB.v0 ) goto propagateNaN;
92 return a;
93 }
94 /*------------------------------------------------------------------------
95 *------------------------------------------------------------------------*/
96 if ( ! expB ) {
97 if ( ! (sigB.v64 | sigB.v0) ) goto invalid;
98 normExpSig = softfloat_normSubnormalF128Sig( sigB.v64, sigB.v0 );
99 expB = normExpSig.exp;
100 sigB = normExpSig.sig;
101 }
102 if ( ! expA ) {
103 if ( ! (sigA.v64 | sigA.v0) ) return a;
104 normExpSig = softfloat_normSubnormalF128Sig( sigA.v64, sigA.v0 );
105 expA = normExpSig.exp;
106 sigA = normExpSig.sig;
107 }
108 /*------------------------------------------------------------------------
109 *------------------------------------------------------------------------*/
110 sigA.v64 |= UINT64_C( 0x0001000000000000 );
111 sigB.v64 |= UINT64_C( 0x0001000000000000 );
112 rem = sigA;
113 expDiff = expA - expB;
114 if ( expDiff < 1 ) {
115 if ( expDiff < -1 ) return a;
116 if ( expDiff ) {
117 --expB;
118 sigB = softfloat_add128( sigB.v64, sigB.v0, sigB.v64, sigB.v0 );
119 q = 0;
120 } else {
121 q = softfloat_le128( sigB.v64, sigB.v0, rem.v64, rem.v0 );
122 if ( q ) {
123 rem = softfloat_sub128( rem.v64, rem.v0, sigB.v64, sigB.v0 );
124 }
125 }
126 } else {
127 recip32 = softfloat_approxRecip32_1( sigB.v64>>17 );
128 expDiff -= 30;
129 for (;;) {
130 q64 = (uint_fast64_t) (uint32_t) (rem.v64>>19) * recip32;
131 if ( expDiff < 0 ) break;
132 q = (q64 + 0x80000000)>>32;
133 rem = softfloat_shortShiftLeft128( rem.v64, rem.v0, 29 );
134 term = softfloat_mul128By32( sigB.v64, sigB.v0, q );
135 rem = softfloat_sub128( rem.v64, rem.v0, term.v64, term.v0 );
136 if ( rem.v64 & UINT64_C( 0x8000000000000000 ) ) {
137 rem = softfloat_add128( rem.v64, rem.v0, sigB.v64, sigB.v0 );
138 }
139 expDiff -= 29;
140 }
141 /*--------------------------------------------------------------------
142 | (`expDiff' cannot be less than -29 here.)
143 *--------------------------------------------------------------------*/
144 q = (uint32_t) (q64>>32)>>(~expDiff & 31);
145 rem = softfloat_shortShiftLeft128( rem.v64, rem.v0, expDiff + 30 );
146 term = softfloat_mul128By32( sigB.v64, sigB.v0, q );
147 rem = softfloat_sub128( rem.v64, rem.v0, term.v64, term.v0 );
148 if ( rem.v64 & UINT64_C( 0x8000000000000000 ) ) {
149 altRem = softfloat_add128( rem.v64, rem.v0, sigB.v64, sigB.v0 );
150 goto selectRem;
151 }
152 }
153 /*------------------------------------------------------------------------
154 *------------------------------------------------------------------------*/
155 do {
156 altRem = rem;
157 ++q;
158 rem = softfloat_sub128( rem.v64, rem.v0, sigB.v64, sigB.v0 );
159 } while ( ! (rem.v64 & UINT64_C( 0x8000000000000000 )) );
160 selectRem:
161 meanRem = softfloat_add128( rem.v64, rem.v0, altRem.v64, altRem.v0 );
162 if (
163 (meanRem.v64 & UINT64_C( 0x8000000000000000 ))
164 || (! (meanRem.v64 | meanRem.v0) && (q & 1))
165 ) {
166 rem = altRem;
167 }
168 signRem = signA;
169 if ( rem.v64 & UINT64_C( 0x8000000000000000 ) ) {
170 signRem = ! signRem;
171 rem = softfloat_sub128( 0, 0, rem.v64, rem.v0 );
172 }
173 return softfloat_normRoundPackToF128( signRem, expB - 1, rem.v64, rem.v0 );
174 /*------------------------------------------------------------------------
175 *------------------------------------------------------------------------*/
176 propagateNaN:
177 uiZ = softfloat_propagateNaNF128UI( uiA64, uiA0, uiB64, uiB0 );
178 goto uiZ;
179 /*------------------------------------------------------------------------
180 *------------------------------------------------------------------------*/
181 invalid:
182 softfloat_raiseFlags( softfloat_flag_invalid );
183 uiZ.v64 = defaultNaNF128UI64;
184 uiZ.v0 = defaultNaNF128UI0;
185 uiZ:
186 uZ.ui = uiZ;
187 return uZ.f;
188
189 }
190