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
/*
* Copyright 1993-2010 NVIDIA Corporation. All rights reserved.
*
* NVIDIA Corporation and its licensors retain all intellectual
* property and proprietary rights in and to this software and
* related documentation and any modifications thereto.
* Any use, reproduction, disclosure, or distribution of this
* software and related documentation without an express license
* agreement from NVIDIA Corporation is strictly prohibited.
*/
#if !defined CUDA_DISABLER
#include "TestCompact.h"
TestCompact::TestCompact(std::string testName_, NCVTestSourceProvider<Ncv32u> &src_,
Ncv32u length_, Ncv32u badElem_, Ncv32u badElemPercentage_)
:
NCVTestProvider(testName_),
src(src_),
length(length_),
badElem(badElem_),
badElemPercentage(badElemPercentage_ > 100 ? 100 : badElemPercentage_)
{
}
bool TestCompact::toString(std::ofstream &strOut)
{
strOut << "length=" << length << std::endl;
strOut << "badElem=" << badElem << std::endl;
strOut << "badElemPercentage=" << badElemPercentage << std::endl;
return true;
}
bool TestCompact::init()
{
return true;
}
bool TestCompact::process()
{
NCVStatus ncvStat;
bool rcode = false;
NCVVectorAlloc<Ncv32u> h_vecSrc(*this->allocatorCPU.get(), this->length);
ncvAssertReturn(h_vecSrc.isMemAllocated(), false);
NCVVectorAlloc<Ncv32u> d_vecSrc(*this->allocatorGPU.get(), this->length);
ncvAssertReturn(d_vecSrc.isMemAllocated(), false);
NCVVectorAlloc<Ncv32u> h_vecDst(*this->allocatorCPU.get(), this->length);
ncvAssertReturn(h_vecDst.isMemAllocated(), false);
NCVVectorAlloc<Ncv32u> d_vecDst(*this->allocatorGPU.get(), this->length);
ncvAssertReturn(d_vecDst.isMemAllocated(), false);
NCVVectorAlloc<Ncv32u> h_vecDst_d(*this->allocatorCPU.get(), this->length);
ncvAssertReturn(h_vecDst_d.isMemAllocated(), false);
NCV_SET_SKIP_COND(this->allocatorGPU.get()->isCounting());
NCV_SKIP_COND_BEGIN
ncvAssertReturn(this->src.fill(h_vecSrc), false);
for (Ncv32u i=0; i<this->length; i++)
{
Ncv32u tmp = (h_vecSrc.ptr()[i]) & 0xFF;
tmp = tmp * 99 / 255;
if (tmp < this->badElemPercentage)
{
h_vecSrc.ptr()[i] = this->badElem;
}
}
NCV_SKIP_COND_END
NCVVectorAlloc<Ncv32u> h_dstLen(*this->allocatorCPU.get(), 1);
ncvAssertReturn(h_dstLen.isMemAllocated(), false);
Ncv32u bufSize;
ncvStat = nppsStCompactGetSize_32u(this->length, &bufSize, this->devProp);
ncvAssertReturn(NPPST_SUCCESS == ncvStat, false);
NCVVectorAlloc<Ncv8u> d_tmpBuf(*this->allocatorGPU.get(), bufSize);
ncvAssertReturn(d_tmpBuf.isMemAllocated(), false);
Ncv32u h_outElemNum_h = 0;
NCV_SKIP_COND_BEGIN
ncvStat = h_vecSrc.copySolid(d_vecSrc, 0);
ncvAssertReturn(ncvStat == NPPST_SUCCESS, false);
ncvStat = nppsStCompact_32u(d_vecSrc.ptr(), this->length,
d_vecDst.ptr(), h_dstLen.ptr(), this->badElem,
d_tmpBuf.ptr(), bufSize, this->devProp);
ncvAssertReturn(ncvStat == NPPST_SUCCESS, false);
ncvStat = d_vecDst.copySolid(h_vecDst_d, 0);
ncvAssertReturn(ncvStat == NPPST_SUCCESS, false);
ncvStat = nppsStCompact_32u_host(h_vecSrc.ptr(), this->length, h_vecDst.ptr(), &h_outElemNum_h, this->badElem);
ncvAssertReturn(ncvStat == NPPST_SUCCESS, false);
NCV_SKIP_COND_END
//bit-to-bit check
bool bLoopVirgin = true;
NCV_SKIP_COND_BEGIN
if (h_dstLen.ptr()[0] != h_outElemNum_h)
{
bLoopVirgin = false;
}
else
{
for (Ncv32u i=0; bLoopVirgin && i < h_outElemNum_h; i++)
{
if (h_vecDst.ptr()[i] != h_vecDst_d.ptr()[i])
{
bLoopVirgin = false;
}
}
}
NCV_SKIP_COND_END
if (bLoopVirgin)
{
rcode = true;
}
return rcode;
}
bool TestCompact::deinit()
{
return true;
}
#endif /* CUDA_DISABLER */