1 /*
2 * QEMU TILE-Gx helpers
3 *
4 * Copyright (c) 2015 Chen Gang
5 *
6 * This library is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
10 *
11 * This library is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * Lesser General Public License for more details.
15 *
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with this library; if not, see
18 * <http://www.gnu.org/licenses/lgpl-2.1.html>
19 */
20
21 #include "qemu/osdep.h"
22 #include "cpu.h"
23 #include "exec/helper-proto.h"
24
25
26 /* Broadcast a value to all elements of a vector. */
27 #define V1(X) (((X) & 0xff) * 0x0101010101010101ull)
28 #define V2(X) (((X) & 0xffff) * 0x0001000100010001ull)
29
30
helper_v1multu(uint64_t a,uint64_t b)31 uint64_t helper_v1multu(uint64_t a, uint64_t b)
32 {
33 uint64_t r = 0;
34 int i;
35
36 for (i = 0; i < 64; i += 8) {
37 unsigned ae = extract64(a, i, 8);
38 unsigned be = extract64(b, i, 8);
39 r = deposit64(r, i, 8, ae * be);
40 }
41 return r;
42 }
43
helper_v2mults(uint64_t a,uint64_t b)44 uint64_t helper_v2mults(uint64_t a, uint64_t b)
45 {
46 uint64_t r = 0;
47 int i;
48
49 /* While the instruction talks about signed inputs, with a
50 truncated result the sign of the inputs doesn't matter. */
51 for (i = 0; i < 64; i += 16) {
52 unsigned ae = extract64(a, i, 16);
53 unsigned be = extract64(b, i, 16);
54 r = deposit64(r, i, 16, ae * be);
55 }
56 return r;
57 }
58
helper_v1shl(uint64_t a,uint64_t b)59 uint64_t helper_v1shl(uint64_t a, uint64_t b)
60 {
61 uint64_t m;
62
63 b &= 7;
64 m = V1(0xff >> b);
65 return (a & m) << b;
66 }
67
helper_v2shl(uint64_t a,uint64_t b)68 uint64_t helper_v2shl(uint64_t a, uint64_t b)
69 {
70 uint64_t m;
71
72 b &= 15;
73 m = V2(0xffff >> b);
74 return (a & m) << b;
75 }
76
helper_v1shru(uint64_t a,uint64_t b)77 uint64_t helper_v1shru(uint64_t a, uint64_t b)
78 {
79 uint64_t m;
80
81 b &= 7;
82 m = V1(0xff << b);
83 return (a & m) >> b;
84 }
85
helper_v2shru(uint64_t a,uint64_t b)86 uint64_t helper_v2shru(uint64_t a, uint64_t b)
87 {
88 uint64_t m;
89
90 b &= 15;
91 m = V2(0xffff << b);
92 return (a & m) >> b;
93 }
94
helper_v1shrs(uint64_t a,uint64_t b)95 uint64_t helper_v1shrs(uint64_t a, uint64_t b)
96 {
97 uint64_t r = 0;
98 int i;
99
100 b &= 7;
101 for (i = 0; i < 64; i += 8) {
102 r = deposit64(r, i, 8, sextract64(a, i + b, 8 - b));
103 }
104 return r;
105 }
106
helper_v2shrs(uint64_t a,uint64_t b)107 uint64_t helper_v2shrs(uint64_t a, uint64_t b)
108 {
109 uint64_t r = 0;
110 int i;
111
112 b &= 15;
113 for (i = 0; i < 64; i += 16) {
114 r = deposit64(r, i, 16, sextract64(a, i + b, 16 - b));
115 }
116 return r;
117 }
118
helper_v1int_h(uint64_t a,uint64_t b)119 uint64_t helper_v1int_h(uint64_t a, uint64_t b)
120 {
121 uint64_t r = 0;
122 int i;
123
124 for (i = 0; i < 32; i += 8) {
125 r = deposit64(r, 2 * i + 8, 8, extract64(a, i + 32, 8));
126 r = deposit64(r, 2 * i, 8, extract64(b, i + 32, 8));
127 }
128 return r;
129 }
130
helper_v1int_l(uint64_t a,uint64_t b)131 uint64_t helper_v1int_l(uint64_t a, uint64_t b)
132 {
133 uint64_t r = 0;
134 int i;
135
136 for (i = 0; i < 32; i += 8) {
137 r = deposit64(r, 2 * i + 8, 8, extract64(a, i, 8));
138 r = deposit64(r, 2 * i, 8, extract64(b, i, 8));
139 }
140 return r;
141 }
142
helper_v2int_h(uint64_t a,uint64_t b)143 uint64_t helper_v2int_h(uint64_t a, uint64_t b)
144 {
145 uint64_t r = 0;
146 int i;
147
148 for (i = 0; i < 32; i += 16) {
149 r = deposit64(r, 2 * i + 16, 16, extract64(a, i + 32, 16));
150 r = deposit64(r, 2 * i, 16, extract64(b, i + 32, 16));
151 }
152 return r;
153 }
154
helper_v2int_l(uint64_t a,uint64_t b)155 uint64_t helper_v2int_l(uint64_t a, uint64_t b)
156 {
157 uint64_t r = 0;
158 int i;
159
160 for (i = 0; i < 32; i += 16) {
161 r = deposit64(r, 2 * i + 16, 16, extract64(a, i, 16));
162 r = deposit64(r, 2 * i, 16, extract64(b, i, 16));
163 }
164 return r;
165 }
166