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
|
/* $Id: cpl_mask_binary.h,v 1.7 2012-01-20 13:52:19 llundin Exp $
*
* This file is part of the ESO Common Pipeline Library
* Copyright (C) 2001-2008 European Southern Observatory
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
*/
/*----------------------------------------------------------------------------*/
/**
@internal
@brief Perform the bitwise operation
@param self Pre-allocated buffer to hold the result
@param first First operand, or NULL for an in-place operation
@param second Second operand
@param nxy The number of elements
@return void
@note No error checking in this internal function!
*/
/*----------------------------------------------------------------------------*/
void APPENDOPER(CPL_MASK_BINARY_WOPER)(cpl_binary * self,
const cpl_binary * first,
const cpl_binary * second,
size_t nxy)
{
CPL_MASK_REGISTER_TYPE * word0 = (CPL_MASK_REGISTER_TYPE*)self;
const CPL_MASK_REGISTER_TYPE* word1 = first
? (const CPL_MASK_REGISTER_TYPE*)first
: (const CPL_MASK_REGISTER_TYPE*)self;
const CPL_MASK_REGISTER_TYPE* word2 = (const CPL_MASK_REGISTER_TYPE*)second;
const size_t nword = nxy / CPL_MASK_REGISTER_SIZE;
size_t i;
for (i = 0; i < nword; i++) {
#if CPL_MASK_REGISTER_SIZE == 16
#ifndef __SSE2__
#error "For 16-byte processing __SSE2__ must be defined"
#endif
/* Assume unaligned data. With load of aligned data gcc can perform the
operation directly on an address operand and thus avoid one load */
const __m128i v1 = _mm_loadu_si128(word1 + i);
const __m128i v2 = _mm_loadu_si128(word2 + i);
const __m128i v0 = OPER2MM(CPL_MASK_BINARY_WOPER)(v1, v2);
_mm_storeu_si128(word0 + i, v0);
#else
word0[i] = word1[i] CPL_MASK_BINARY_OPER word2[i];
#endif
}
i *= CPL_MASK_REGISTER_SIZE;
if (first == NULL) {
for (; i < nxy; i++) {
self[i] = self[i] CPL_MASK_BINARY_OPER second[i];
}
} else {
for (; i < nxy; i++) {
self[i] = first[i] CPL_MASK_BINARY_OPER second[i];
}
}
}
/*----------------------------------------------------------------------------*/
/**
@internal
@brief Perform the bitwise operation
@param self Pre-allocated buffer to hold the result
@param first First operand, or NULL for an in-place operation
@param second Second operand (scalar)
@param nxy The number of elements
@return void
@note No error checking in this internal function!
*/
/*----------------------------------------------------------------------------*/
void APPENDOPERS(CPL_MASK_BINARY_WOPER)(cpl_binary * self,
const cpl_binary * first,
cpl_bitmask second,
size_t nxy)
{
CPL_MASK_REGISTER_TYPE * word0 = (CPL_MASK_REGISTER_TYPE*)self;
const CPL_MASK_REGISTER_TYPE* word1 = first
? (const CPL_MASK_REGISTER_TYPE*)first
: (const CPL_MASK_REGISTER_TYPE*)self;
const size_t nword = nxy / CPL_MASK_REGISTER_SIZE;
size_t i;
for (i = 0; i < nword; i++) {
#if CPL_MASK_REGISTER_SIZE == 16
#ifndef __SSE2__
#error "For 16-byte processing __SSE2__ must be defined"
#endif
/* Process two cpl_bitmasks in each iteration */
/* Assume unaligned data. With load of aligned data gcc can perform the
operation directly on an address operand and thus avoid one load */
const __m128i v1 = _mm_loadu_si128(word1 + i);
const __m128i v2 = _mm_set1_epi64(cpl_m_from_int64(second));
const __m128i v0 = OPER2MM(CPL_MASK_BINARY_WOPER)(v1, v2);
_mm_storeu_si128(word0 + i, v0);
#else
word0[i] = word1[i] CPL_MASK_BINARY_OPER (CPL_MASK_REGISTER_TYPE)second;
#endif
}
i *= CPL_MASK_REGISTER_SIZE;
#if CPL_MASK_REGISTER_SIZE > 4
#if CPL_MASK_REGISTER_SIZE > 8
if (i + 8 <= nxy) {
/* Handle the remaining uint64_t */
uint64_t * self8 = (uint64_t *)self;
const uint64_t * first8 = (const uint64_t *)(first ? first : self);
self8[i/8] = first8[i/8] CPL_MASK_BINARY_OPER (uint64_t)second;
i += 8;
}
#endif
if (i + 4 <= nxy) {
/* Handle the remaining uint32_t */
uint32_t * self4 = (uint32_t *)self;
const uint32_t * first4 = (const uint32_t *)(first ? first : self);
self4[i/4] = first4[i/4] CPL_MASK_BINARY_OPER (uint32_t)second;
i += 4;
}
#endif
if (first == NULL) {
for (; i < nxy; i++) {
self[i] = self[i] CPL_MASK_BINARY_OPER (cpl_binary)second;
}
} else {
for (; i < nxy; i++) {
self[i] = first[i] CPL_MASK_BINARY_OPER (cpl_binary)second;
}
}
}
|
