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
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
|
/*
copyfunctions implementation in mmx
Copyright (C) 2000 Martin Vogt
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU Library General Public License as published by
the Free Software Foundation.
For more information look at the file COPYRIGHT in this package
*/
#include "copyFunctions_mmx.h"
using namespace std;
// mmx goodies
static long ADD_1[] = {0x01010101, 0x01010101};
static long ADDW_1[] = {0x00010001, 0x00010001};
static long MASK_AND[] = {0x7f7f7f7f, 0x7f7f7f7f};
void dummyCopyFunctions() {
cout << "ADD_1:"<<ADD_1<<endl;
cout << "ADDW_1:"<<ADDW_1<<endl;
cout << "MASK_AND:"<<MASK_AND<<endl;
}
CopyFunctions_MMX::CopyFunctions_MMX() {
#ifdef INTEL
lmmx=mm_support();
#else
lmmx=false;
cout << "no INTEL arch- disable MMX in copyFunctions"<<endl;
#endif
}
CopyFunctions_MMX::~CopyFunctions_MMX() {
}
int CopyFunctions_MMX::support() {
return lmmx;
}
#if defined (__GNUC__) && defined (INTEL)
void CopyFunctions_MMX::startNOFloatSection() {
}
void CopyFunctions_MMX::endNOFloatSection() {
emms();
}
void CopyFunctions_MMX::copy8_byte(unsigned char* source1,
unsigned char* dest,int inc) {
int rr=4;
asm (
"1:\n"
"movq (%0) ,%%mm0\n"
"leal (%0,%2) ,%0\n"
"movq (%0) ,%%mm1\n"
"leal (%0,%2) ,%0\n"
// Write
"movq %%mm0 ,(%1)\n"
"leal (%1,%2) ,%1\n"
"movq %%mm1 ,(%1)\n"
"leal (%1,%2) ,%1\n"
"decl %3\n"
"jnz 1b\n"
:
: "r"(source1),"r"(dest),"r"(inc),"r"(rr)
);
}
void CopyFunctions_MMX::copy8_src1linear_crop(short* source1,
unsigned char* dest,int inc) {
asm (
"movq (%1),%%mm0\n"
"packuswb 8(%1),%%mm0\n"
"movq %%mm0,(%0)\n"
"addl %2,%0\n"
"movq 16(%1),%%mm0\n"
"packuswb 24(%1),%%mm0\n"
"movq %%mm0,(%0)\n"
"addl %2,%0\n"
"movq 32(%1),%%mm0\n"
"packuswb 40(%1),%%mm0\n"
"movq %%mm0,(%0)\n"
"addl %2,%0\n"
"movq 48(%1),%%mm0\n"
"packuswb 56(%1),%%mm0\n"
"movq %%mm0,(%0)\n"
"addl %2,%0\n"
"movq 64(%1),%%mm0\n"
"packuswb 72(%1),%%mm0\n"
"movq %%mm0,(%0)\n"
"addl %2,%0\n"
"movq 80(%1),%%mm0\n"
"packuswb 88(%1),%%mm0\n"
"movq %%mm0,(%0)\n"
"addl %2,%0\n"
"movq 96(%1),%%mm0\n"
"packuswb 104(%1),%%mm0\n"
"movq %%mm0,(%0)\n"
"addl %2,%0\n"
"movq 112(%1),%%mm0\n"
"packuswb 120(%1),%%mm0\n"
"movq %%mm0,(%0)\n"
:
:"r" (dest), "r" (source1),"r" (inc)
);
}
void CopyFunctions_MMX::copy8_div2_nocrop(unsigned char* source1,
unsigned char* source2,
unsigned char* dest,int inc) {
int h=8;
asm (
"movq MASK_AND, %%mm5\n"
"movq ADD_1, %%mm6\n"
"1:\t"
"movq (%1), %%mm0\n" /* 8 s */
"movq (%4), %%mm1\n" /* 8 s +lx */
"psrlw $1,%%mm0\n"
"psrlw $1,%%mm1\n"
"pand %%mm5,%%mm0\n"
"pand %%mm5,%%mm1\n"
"paddusb %%mm1,%%mm0\n"
"addl %3,%1\n"
"paddusb %%mm6,%%mm0\n"
"addl %3,%4\n"
"movq %%mm0,(%2)\n"
"decl %0\n"
"leal (%2, %3), %2\n"
"jnz 1b\n"
:
: "c"(h), "r"(source1), "r"(dest), "r"(inc), "r"(source2)
);
}
void CopyFunctions_MMX::copy8_div2_destlinear_nocrop(unsigned char* source1,
unsigned char* source2,
unsigned char* dest,
int inc) {
int h=8;
asm (
"movq MASK_AND, %%mm5\n"
"1:\t"
"movq (%1), %%mm0\n" /* 8 s */
"movq (%4), %%mm1\n" /* 8 s +lx */
"psrlw $1,%%mm0\n"
"psrlw $1,%%mm1\n"
"pand %%mm5,%%mm0\n"
"pand %%mm5,%%mm1\n"
"paddusb %%mm1,%%mm0\n"
"addl %3,%1\n"
"addl %3,%4\n"
"movq %%mm0,(%2)\n"
"decl %0\n"
"leal 8(%2), %2\n"
"jnz 1b\n"
:
: "c"(h), "r"(source1), "r"(dest), "r"(inc), "r"(source2)
);
}
void CopyFunctions_MMX::copy16_div2_destlinear_nocrop(unsigned char* source1,
unsigned char* source2,
unsigned char* dest,
int inc) {
int h=16;
inc=inc-8;
asm (
"movq MASK_AND, %%mm5\n"
"1:\t"
"movq (%1), %%mm0\n" /* 8 s */
"movq (%4), %%mm1\n" /* 8 s +lx */
"psrlw $1,%%mm0\n"
"psrlw $1,%%mm1\n"
"pand %%mm5,%%mm0\n"
"pand %%mm5,%%mm1\n"
"paddusb %%mm1,%%mm0\n"
"leal 8(%1),%1\n"
"leal 8(%4),%4\n"
"movq %%mm0,(%2)\n"
"leal 8(%2),%2\n"
"movq (%1), %%mm0\n" /* 8 s */
"movq (%4), %%mm1\n" /* 8 s +lx */
"psrlw $1,%%mm0\n"
"psrlw $1,%%mm1\n"
"pand %%mm5,%%mm0\n"
"pand %%mm5,%%mm1\n"
"paddusb %%mm1,%%mm0\n"
"leal (%3,%1),%1\n"
"leal (%3,%4),%4\n"
"movq %%mm0,(%2)\n"
"leal 8(%2),%2\n"
"decl %0\n"
"jnz 1b\n"
:
: "c"(h), "r"(source1), "r"(dest), "r"(inc), "r"(source2)
);
}
void CopyFunctions_MMX::copy8_src2linear_crop(unsigned char* source1,
short int* source2,
unsigned char* dest,int inc) {
int rr=8;
// buggy
asm (
".align 32\n"
"pxor %%mm2 ,%%mm2\n" // 0 0 0 0 0 0 0 0
"1:\n"
"movq (%0) ,%%mm0\n" // s1_7 s1_6 s1_5 s1_4 s1_3 s1_2 s1_1 s1_0
"movq (%0) ,%%mm4\n" // s1_7 s1_6 s1_5 s1_4 s1_3 s1_2 s1_1 s1_0
"punpckhbw %%mm2 ,%%mm0\n" // 0 s1_7 0 s1_6 0 s1_5 0 s1_4
"punpcklbw %%mm2 ,%%mm4\n" // 0 s1_3 0 s1_2 0 s1_1 0 s1_0
"movq (%1) ,%%mm1\n" // s23h s23l s22h s22l s21l s21h s20h s20l
"movq 8(%1) ,%%mm5\n" // s27h s27l s26h s26l s25l s25h s24h s24l
"paddw %%mm0 ,%%mm5\n" // mm4=mm4 + s3_0..3
"paddw %%mm4 ,%%mm1\n" // mm0=mm0 + s3_4..7
"packuswb %%mm5 ,%%mm1\n" // cm[...]
"movq %%mm1 ,(%2)\n" // wrote out
"leal (%0,%3), %0\n" // source1+=inc
"leal 16(%1) , %1\n" // source2+=inc
"leal (%2,%3), %2\n" // dest+=inc
"decl %4\n"
"jnz 1b\n"
//"emms\n"
:
: "r"(source1), "r"(source2), "r"(dest),"r"(inc),"r"(rr)
);
}
void CopyFunctions_MMX::copy8_div2_src3linear_crop(unsigned char* source1,
unsigned char* source2,
short int* source3,
unsigned char* dest,
int inc){
// buggy
int rr=8;
asm (
"pxor %%mm2 ,%%mm2\n" // 0 0 0 0 0 0 0 0
"pxor %%mm3 ,%%mm3\n" // 0 0 0 0 0 0 0 0
"movq ADDW_1,%%mm6\n" // 0 1 0 1 0 1 0 1
"1:\n"
"movq (%0) ,%%mm0\n" // s1_7 s1_6 s1_5 s1_4 s1_3 s1_2 s1_1 s1_0
"movq (%1) ,%%mm1\n" // s2_7 s2_6 s2_5 s2_4 s2_3 s2_2 s2_1 s2_0
"movq %%mm0 ,%%mm4\n" // s1_7 s1_6 s1_5 s1_4 s1_3 s1_2 s1_1 s1_0
"movq %%mm1 ,%%mm5\n" // s2_7 s2_6 s2_5 s2_4 s2_3 s2_2 s2_1 s2_0
"punpckhbw %%mm2 ,%%mm0\n" // 0 s1_7 0 s1_6 0 s1_5 0 s1_4
"punpckhbw %%mm3 ,%%mm1\n" // 0 s2_7 0 s2_6 0 s2_5 0 s2_4
"punpcklbw %%mm2 ,%%mm4\n" // 0 s1_3 0 s1_2 0 s1_1 0 s1_0
"punpcklbw %%mm3 ,%%mm5\n" // 0 s2_3 0 s2_2 0 s2_1 0 s2_0
"paddusw %%mm4 ,%%mm5\n" // mm5=s1_0..3 + s2_0..3
"paddusw %%mm0 ,%%mm1\n" // mm1=s1_4..7 + s2_4..7
"paddusw %%mm6 ,%%mm5\n" // mm5=mm5 + 1
"paddusw %%mm6 ,%%mm1\n" // mm1=mm1 + 1
"psraw $1 ,%%mm1\n" // mm1=mm1/2
"psraw $1 ,%%mm5\n" // mm5=mm5/2
"movq (%2) ,%%mm0\n" // s33h s33l s32h s32l s31l s31h s30h s30l
"movq 8(%2) ,%%mm4\n" // s37h s37l s36h s36l s35l s35h s34h s34l
"paddw %%mm0 ,%%mm5\n" // mm5=mm5 + s3_0..3
"paddw %%mm4 ,%%mm1\n" // mm1=mm1 + s3_4..7
"packuswb %%mm1 ,%%mm5\n" // cm[...]
"movq %%mm5 ,(%3)\n" // wrote out
"leal (%0,%4), %0\n" // source1+=inc
"leal (%1,%4), %1\n" // source2+=inc
"addl $16 ,%2 \n" // source3+8
"leal (%3,%4), %3\n" // dest+=inc
"decl %5\n"
"jnz 1b\n"
:
: "r"(source1), "r"(source2), "r"(source3),
"r"(dest),"r"(inc),"m"(rr)
);
}
#endif
|
