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
|
/* FreeRDP: A Remote Desktop Protocol Client
* Copy operations.
* vi:ts=4 sw=4:
*
* (c) Copyright 2012 Hewlett-Packard Development Company, L.P.
* Licensed under the Apache License, Version 2.0 (the "License"); you may
* not use this file except in compliance with the License. You may obtain
* a copy of the License at http://www.apache.org/licenses/LICENSE-2.0.
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express
* or implied. See the License for the specific language governing
* permissions and limitations under the License.
*/
#include <freerdp/config.h>
#include <string.h>
#include <freerdp/types.h>
#include <freerdp/primitives.h>
#include "prim_internal.h"
static primitives_t* generic = NULL;
/* ------------------------------------------------------------------------- */
/*static inline BOOL memory_regions_overlap_1d(*/
static BOOL memory_regions_overlap_1d(const BYTE* p1, const BYTE* p2, size_t bytes)
{
const ULONG_PTR p1m = (const ULONG_PTR)p1;
const ULONG_PTR p2m = (const ULONG_PTR)p2;
if (p1m <= p2m)
{
if (p1m + bytes > p2m)
return TRUE;
}
else
{
if (p2m + bytes > p1m)
return TRUE;
}
/* else */
return FALSE;
}
/* ------------------------------------------------------------------------- */
/*static inline BOOL memory_regions_overlap_2d( */
static BOOL memory_regions_overlap_2d(const BYTE* p1, int p1Step, int p1Size, const BYTE* p2,
int p2Step, int p2Size, int width, int height)
{
ULONG_PTR p1m = (ULONG_PTR)p1;
ULONG_PTR p2m = (ULONG_PTR)p2;
if (p1m <= p2m)
{
ULONG_PTR p1mEnd = p1m + 1ull * (height - 1) * p1Step + 1ull * width * p1Size;
if (p1mEnd > p2m)
return TRUE;
}
else
{
ULONG_PTR p2mEnd = p2m + 1ull * (height - 1) * p2Step + 1ull * width * p2Size;
if (p2mEnd > p1m)
return TRUE;
}
/* else */
return FALSE;
}
/* ------------------------------------------------------------------------- */
static pstatus_t general_copy_8u(const BYTE* pSrc, BYTE* pDst, INT32 len)
{
if (memory_regions_overlap_1d(pSrc, pDst, (size_t)len))
{
memmove((void*)pDst, (const void*)pSrc, (size_t)len);
}
else
{
memcpy((void*)pDst, (const void*)pSrc, (size_t)len);
}
return PRIMITIVES_SUCCESS;
}
/* ------------------------------------------------------------------------- */
/* Copy a block of pixels from one buffer to another.
* The addresses are assumed to have been already offset to the upper-left
* corners of the source and destination region of interest.
*/
static pstatus_t general_copy_8u_AC4r(const BYTE* pSrc, INT32 srcStep, BYTE* pDst, INT32 dstStep,
INT32 width, INT32 height)
{
const BYTE* src = (const BYTE*)pSrc;
BYTE* dst = (BYTE*)pDst;
int rowbytes = width * sizeof(UINT32);
if ((width == 0) || (height == 0))
return PRIMITIVES_SUCCESS;
if (memory_regions_overlap_2d(pSrc, srcStep, sizeof(UINT32), pDst, dstStep, sizeof(UINT32),
width, height))
{
do
{
generic->copy(src, dst, rowbytes);
src += srcStep;
dst += dstStep;
} while (--height);
}
else
{
/* TODO: do it in one operation when the rowdata is adjacent. */
do
{
/* If we find a replacement for memcpy that is consistently
* faster, this could be replaced with that.
*/
memcpy(dst, src, rowbytes);
src += srcStep;
dst += dstStep;
} while (--height);
}
return PRIMITIVES_SUCCESS;
}
/* ------------------------------------------------------------------------- */
void primitives_init_copy(primitives_t* prims)
{
/* Start with the default. */
prims->copy_8u = general_copy_8u;
prims->copy_8u_AC4r = general_copy_8u_AC4r;
/* This is just an alias with void* parameters */
prims->copy = (__copy_t)(prims->copy_8u);
}
#if defined(WITH_SSE2) || defined(WITH_NEON)
void primitives_init_copy_opt(primitives_t* prims)
{
generic = primitives_get_generic();
primitives_init_copy(prims);
/* Pick tuned versions if possible. */
/* Performance with an SSE2 version with no prefetch seemed to be
* all over the map vs. memcpy.
* Sometimes it was significantly faster, sometimes dreadfully slower,
* and it seemed to vary a lot depending on block size and processor.
* Hence, no SSE version is used here unless once can be written that
* is consistently faster than memcpy.
*/
/* This is just an alias with void* parameters */
prims->copy = (__copy_t)(prims->copy_8u);
}
#endif
|
