]> git.proxmox.com Git - mirror_qemu.git/blob - target-tilegx/simd_helper.c
vfio: spapr: Add DMA memory preregistering (SPAPR IOMMU v2)
[mirror_qemu.git] / target-tilegx / simd_helper.c
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 "qemu-common.h"
24 #include "exec/helper-proto.h"
25
26
27 /* Broadcast a value to all elements of a vector. */
28 #define V1(X) (((X) & 0xff) * 0x0101010101010101ull)
29 #define V2(X) (((X) & 0xffff) * 0x0001000100010001ull)
30
31
32 uint64_t helper_v1multu(uint64_t a, uint64_t b)
33 {
34 uint64_t r = 0;
35 int i;
36
37 for (i = 0; i < 64; i += 8) {
38 unsigned ae = extract64(a, i, 8);
39 unsigned be = extract64(b, i, 8);
40 r = deposit64(r, i, 8, ae * be);
41 }
42 return r;
43 }
44
45 uint64_t helper_v2mults(uint64_t a, uint64_t b)
46 {
47 uint64_t r = 0;
48 int i;
49
50 /* While the instruction talks about signed inputs, with a
51 truncated result the sign of the inputs doesn't matter. */
52 for (i = 0; i < 64; i += 16) {
53 unsigned ae = extract64(a, i, 16);
54 unsigned be = extract64(b, i, 16);
55 r = deposit64(r, i, 16, ae * be);
56 }
57 return r;
58 }
59
60 uint64_t helper_v1shl(uint64_t a, uint64_t b)
61 {
62 uint64_t m;
63
64 b &= 7;
65 m = V1(0xff >> b);
66 return (a & m) << b;
67 }
68
69 uint64_t helper_v2shl(uint64_t a, uint64_t b)
70 {
71 uint64_t m;
72
73 b &= 15;
74 m = V2(0xffff >> b);
75 return (a & m) << b;
76 }
77
78 uint64_t helper_v1shru(uint64_t a, uint64_t b)
79 {
80 uint64_t m;
81
82 b &= 7;
83 m = V1(0xff << b);
84 return (a & m) >> b;
85 }
86
87 uint64_t helper_v2shru(uint64_t a, uint64_t b)
88 {
89 uint64_t m;
90
91 b &= 15;
92 m = V2(0xffff << b);
93 return (a & m) >> b;
94 }
95
96 uint64_t helper_v1shrs(uint64_t a, uint64_t b)
97 {
98 uint64_t r = 0;
99 int i;
100
101 b &= 7;
102 for (i = 0; i < 64; i += 8) {
103 r = deposit64(r, i, 8, sextract64(a, i + b, 8 - b));
104 }
105 return r;
106 }
107
108 uint64_t helper_v2shrs(uint64_t a, uint64_t b)
109 {
110 uint64_t r = 0;
111 int i;
112
113 b &= 15;
114 for (i = 0; i < 64; i += 16) {
115 r = deposit64(r, i, 16, sextract64(a, i + b, 16 - b));
116 }
117 return r;
118 }
119
120 uint64_t helper_v1int_h(uint64_t a, uint64_t b)
121 {
122 uint64_t r = 0;
123 int i;
124
125 for (i = 0; i < 32; i += 8) {
126 r = deposit64(r, 2 * i + 8, 8, extract64(a, i + 32, 8));
127 r = deposit64(r, 2 * i, 8, extract64(b, i + 32, 8));
128 }
129 return r;
130 }
131
132 uint64_t helper_v1int_l(uint64_t a, uint64_t b)
133 {
134 uint64_t r = 0;
135 int i;
136
137 for (i = 0; i < 32; i += 8) {
138 r = deposit64(r, 2 * i + 8, 8, extract64(a, i, 8));
139 r = deposit64(r, 2 * i, 8, extract64(b, i, 8));
140 }
141 return r;
142 }
143
144 uint64_t helper_v2int_h(uint64_t a, uint64_t b)
145 {
146 uint64_t r = 0;
147 int i;
148
149 for (i = 0; i < 32; i += 16) {
150 r = deposit64(r, 2 * i + 16, 16, extract64(a, i + 32, 16));
151 r = deposit64(r, 2 * i, 16, extract64(b, i + 32, 16));
152 }
153 return r;
154 }
155
156 uint64_t helper_v2int_l(uint64_t a, uint64_t b)
157 {
158 uint64_t r = 0;
159 int i;
160
161 for (i = 0; i < 32; i += 16) {
162 r = deposit64(r, 2 * i + 16, 16, extract64(a, i, 16));
163 r = deposit64(r, 2 * i, 16, extract64(b, i, 16));
164 }
165 return r;
166 }