blob: a69cae3a780e392a237acb4a5941c3dfeddcf409 (
plain)
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
|
To build the applications in this directory, you should be able to run these
target scripts in this order:
./target_build_all.sh
this will build all the applications
./archive_for_target.sh
this will archive all applications and target scripts into fft.tgz for
installation on the target
At this point everything should be built. If you'd like to install and run
everything you can run this script:
./setup_target_fft_env.sh
this target script will ensure that the fft_driver is installed, ensure
that /mnt/ram has a 100MB tmpfs ram disk mounted, and then extract the
fft.tgz archive into /mnt/ram/fft and run the create_input_waveforms.sh
script and then runs the run_all.sh script
NOTE: see the README_TARGET.TXT file for additional information
Description of directory contents:
target_build_all.sh
target script to build all applications and libraries
target_build_app.sh
target script to build one application
target_build_lib.sh
target script to build one library
clean_all.sh
target script to clean all build objects
archive_for_target.sh
target script to archive all applications and scripts for target
installation
create_input_waveforms.sh
target script that executes waveform creation applications and
duplication scripts to create all input waveforms for the applications
to consume
duplicate_x128.sh
duplicate_x32.sh
duplicate_x8.sh
these target scripts duplicate a file x8, x32 or x128 times. These are
used to replicate the fundamental waveforms into longer patterns
images/
this directory contains host scripts to create PNG images of waveform
data text files using gnu plot and montage from imagemagick
run_all.sh
run_fft_256.sh
run_fft_256x32.sh
run_fft_256x32x128.sh
run_fft_4096.sh
run_stream_256x32x128.sh
run_stream_256x16x1.sh
run_stream_256x1x1.sh
these target scripts execute all of the various permutations of the demo
applications mentioned below
setup_target_fft_env.sh
this target script is used to extract the fft.tgz archive onto the ram
disk mounted at /mnt/ram
overhead.c
this is the source for the overhead library that all of the applications
are linked against
c16_256.c
c16_256x32.c
c16_256x32x128.c
c16_4096.c
these applications use the NE10 C functions to implement the FFT
algorithm for a variety of input sample sizes with 16-bit resolution
c32_256.c
c32_256x32.c
c32_256x32x128.c
c32_4096.c
these applications use the NE10 C functions to implement the FFT
algorithm for a variety of input sample sizes with 32-bit resolution
fft_256.c
fft_256x32.c
fft_256x32x128.c
fft_4096.c
these applications use the FPGA logic to implement the FFT
algorithm for a variety of input sample sizes. This variant pushes
the input samples into the FPGA and pulls the results back out using
programmed I/O from the HPS Cortex A9 processor
fftdma_256.c
fftdma_256x32.c
fftdma_256x32x128.c
fftdma_4096.c
these applications use the FPGA logic to implement the FFT
algorithm for a variety of input sample sizes. This variant moves
the input samples into the FPGA and the results back out using DMA
through the ACP port of the Cortex A9 processor
neon16_256.c
neon16_256x32.c
neon16_256x32x128.c
neon16_4096.c
these applications use the NE10 NEON functions to implement the FFT
algorithm for a variety of input sample sizes with 16-bit resolution
neon32_256.c
neon32_256x32.c
neon32_256x32x128.c
neon32_4096.c
these applications use the NE10 NEON functions to implement the FFT
algorithm for a variety of input sample sizes with 32-bit resolution
stream_fpga_256x32x128.c
stream_fpga_256x16x1.c
stream_fpga_256x1x1.c
stream_neon32_256x32x128.c
stream_neon32_256x16x1.c
stream_neon32_256x1x1.c
stream_raw_256x32x128.c
stream_raw_256x16x1.c
stream_raw_256x1x1.c
these applications stream data from the FPGA into ramdisk files, the
fpga variant DMAs the input stream thru the FPGA FFT logic before it
is DMAed through the ACP port of the Cortex A9, the neon variant DMAs
the raw input data stream through the ACP port and then computes the FFT
algorithm on it before saving to the ram disk, the raw variant DMAs the
raw input data stream through the ACP port and then directly into the
ram disk
create_real_short_sine32.c
create_real_short_square32.c
create_real_short_triangle32.c
these applications create the initial 32 samples of their respective
waveform, sine, square and triangle
real_short_to_ne10cpx_long.c
real_short_to_ne10cpx_short.c
these applications translate the real short data format into ne10cpx
long and short format which is consumed by the demo applications
ne10cpx_long_to_text.c
ne10cpx_short_to_text.c
these applications translate the ne10cpx long and short format binary
data into text format output which may be read by gnu plot for example
|
