-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathINSTALL
167 lines (117 loc) · 5.75 KB
/
INSTALL
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
Installation Instructions
*************************
Copyright (C) 2011, 2012 Free Software Foundation, Inc.
This file is free documentation; the Free Software Foundation gives
unlimited permission to copy, distribute and modify it.
System Requirements
===================
- NVIDIA Graphics card with CUDA 1.3 support
An up-to-date list of CUDA-enabled GPUs can be found on the NVIDIA CUDA Web
site at http://developer.nvidia.com/cuda-gpus or
http://en.wikipedia.org/wiki/CUDA )
To verify which video adapter your system uses, find the model number by going
to your distribution's equivalent of System Properties, or, from the command
line, enter:
`lspci | grep -i nvidia'
If you do not see any settings, update the PCI hardware database that Linux
maintains by entering update-pciids (generally found in /sbin) at the command
line and rerun the previous lspci command.
- Verify you have a supported version of Linux. The CUDA Development Tools are
only supported on some specific distributions of Linux. These are listed in
the CUDA Toolkit release notes. To determine which distribution and release
number you're running, type the following at the command line:
`uname -m && cat /etc/*release'
- For build and install the library you need cmake 2.6.2 or higher.
Verify your version at the command line:
`cmake --version'
- Verify That gcc Is Installed: `gcc --version'
Basic Installation
==================
Go to the /build directory to create an OUT-OF-SOURCE-Build.
Briefly, the shell commands `cmake ../; make; make install' should configure,
build, and install this package.
All intern cmake files will generated inside the build- directory, also the
makefiles. If you should have cmake update problems, only remove all files and
directory in /build by using "rm -rf *" and install again.
The simplest way to compile this package is:
1. `cd' to the directory containing the package's source code and the build
directory. `cd' to the build directory and type `cmake ../' to configure
the package for your system.
Running `cmake' might take a while. While running, it prints
some messages telling which features it is checking for.
2. Type `make' to compile the package.
3. Type `make install' to install the programs and any data files and
documentation.
4. You can remove all generated file, the program binaries and object files
from the build directory by typing `make distclean'.
6. Often, you can also type `make uninstall' to remove the installed
files again.
After successful installation go to ../src/ to find the executable "simpleTest".
For project file structure see also README.
Defining Variables
==================
Variables can be set in the environment passed to `cmake'. You should set
them in the `cmake' command line, using `-DVAR=value'. For example:
cmake ../ -DCUDA_SDK_ROOT_DIR=/usr/local/NVIDIA_GPU_Computing_SDK/
Optional Features for /inc/config.h
===================================
- set otional features in the `cmake' command line, using `-DVar=value'.
For example
cmake ../ -DMEASURED_TIMES=1
or
cmake ../ -DMEASURED_TIMES=ON
LARGE_INPUT
1[ON] : uint_t,sint_t -> ptrdiff_t
0[OFF]: uint_t, sint_t -> int
MEASURED_TIMES:
1[ON] : measured times for different steps and write it to struct members
cunfft_plan->CUNFFTTimes
0[OFF]: measured times only for the whole transformation. Write it to struct
cunfft_plan->CUNFFTTimes->runTime
MILLI_SEC:
1[ON] : write time to stdout in ms
0[OFF]: write time to stdout in sec
DEBUG:
1[ON] : Verify cuda function calls, shows errors
0[OFF]: off
PRINT_CONFIG
1[ON] : Print launch configuration of kernel calls to stdout
0[OFF]: off
COM_FG_PSI
1[ON] : use fast gaussian gridding
0[OFF]: fast gaussian gridding off
CUNFFT_DOUBLE_PRECISION
1[ON] : computation in double precision
0[OFF]: computation in single precision
CUT_OFF (default 6)
- cut-off parameter m for filtersize
THREAD_DIM_X (default 16)
- manuell restriction for maximum numbers of threads per block in x direction for 2D launch grid
- see technical data of your grafic card for restrictions
THREAD_DIM_Y (default 16)
- manuell restriction for maximum numbers of threads per block in y direction for 2D launch grid
- see technical data of your grafic card for restrictions
THREAD_DIM (default 512)
- manuell restriction for maximum numbers of threads per block in x direction for 1D launch grid (y direction=1)
- see also technical data of your grafic card for restrictions
MAX_NUM_THREADS: (default 1024)
- card restriction for maximum number of threads inside a block
- see also technical data of your grafic card for restrictions
MAX_BLOCK_DIM_X (default 1024)
- card restriction for maximum number of threads per block in x direction
- see also technical data of your grafic card for restrictions
MAX_BLOCK_DIM_Y (default is 1024)
- card restriction for maximum numbers of threads per block in y direction
- see also technical data of your grafic card for restrictions
MAX_BLOCK_DIM_Z (default is 64)
- card restriction for maximum numbers of threads per block in z direction
- see also technical data of your grafic card for restrictions
MAX_GRID_DIM_X (default 65535)
- card restriction for maximum number of blocks per grid in x direction
- see also technical data of your grafic card for restrictions
MAX_GRID_DIM_Y (default is 65535)
- card restriction for maximum numbers of blocks per grid in y direction
- see also technical data of your grafic card for restrictions
MAX_GRID_DIM_Z (default is 1)
- card restriction for maximum numbers of blocks per grid in z direction
- see also technical data of your grafic card for restrictions