National-instruments AutoCode NI MATRIX Manuel d'utilisateur Page 188
- Page / 250
- Table des matières
- MARQUE LIVRES
Noté. / 5. Basé sur avis des utilisateurs
Chapter 6 Vectorized Code Generation
© National Instruments Corporation 6-25 AutoCode Reference
top[-1+i] = R_P[-1+i]*U->gain_1[-1+i];
}
/* ---------------------------- Procedure Super Block */
/* {vecproc.3} */
vecproc(&top[0], &Y->vecproc_3_1[0], &vecproc_3_i);
iinfo[0] = vecproc_3_i.iinfo[0];
if( iinfo[0] != 0 ) {
vecproc_3_i.iinfo[0] = 0; goto EXEC_ERROR;
}
A vectorized procedure interface suffers from a form of the copy-back
problem. When using the vectorized interface, each time the procedure is
called, the inputs and outputs must be an array. Therefore, if an array is not
connected as the input or output, AutoCode must generate code to copy
data into a temporary array and then pass that array. In other words, if the
inputs and/or outputs are not connected to naturally conform to a vector
interface, the copies required might add significant overhead to the
procedure call, eliminating the benefits of the vectorized interface.
Ada Array Aggregates and Slices
All of the examples presented include generated C code. AutoCode will
generate the equivalent Ada code, except that use of array aggregation and
slices are used when possible. Example 6-11 is the equivalent Ada code for
a diagram previously shown in Figure 6-2.
Note Only the relevant subsystem code is listed.
Example 6-11 Ada Generated Code (for Figure 6-2)
procedure subsys_1 is
------ Local Block Outputs. ------
Throttle : RT_FLOAT_AY(0..4);
------ Algorithmic Local Variables. ------
i_1 : RT_INTEGER;
j : RT_INTEGER;
k_1 : RT_INTEGER;
begin
-------- Initialization. --------
if SUBSYS_PREINIT(1) then
iinfo(0..3) := (0,1,1,1);
- NI MATRIXx 1
- Important Information 3
- Conventions 4
- Chapter 3 6
- Ada Language Reference 6
- Chapter 4 8
- Chapter 5 8
- Contents 10
- AutoCode Reference x ni.com 10
- Chapter 6 11
- Vectorized Code Generation 11
- AutoCode Reference xii ni.com 12
- Appendix A 13
- Introduction 14
- General Information 15
- Configuration File 15
- Language-Specific Information 15
- Using MATRIXx Help 16
- Related Publications 17
- C Language Reference 18
- Stand-Alone Library 19
- AutoCode Reference 2-4 ni.com 21
- Target-Specific Utilities 22
- AutoCode Reference 2-6 ni.com 23
- AutoCode Reference 2-8 ni.com 25
- UserCode Blocks 27
- Implementing Handwritten UCBs 30
- File Name: usr_dsp.c 32
- Fleming: usr_dsp.c 34
- Function Name: usr_dsp 34
- Variable Interface UCB 35
- Interface Ordering 35
- Function Prototype 36
- Procedure SuperBlocks 37
- Generate Reusable Procedure 40
- Function 41
- C Fixed-Point Arithmetic 43
- Fixed-Point Data Types 45
- User Types 47
- Overflow Protection 48
- Stand-Alone Files 48
- Macro Interface 49
- Function Interface 50
- Conversion Macros 52
- 8-bit number with a shift of 53
- i ALIGN so wo p (n, rp) 54
- Arithmetic Macros 55
- 32-Bit Multiplication 59
- 32-Bit Division 60
- 16-Bit by 8-Bit Division 60
- 32-Bit by 16-Bit Division 60
- Fixed-Point Relational Macros 61
- Some Relevant Issues 62
- Supplied Templates 64
- (or sa_utils.ada), the 66
- Data Types 67
- Error Procedure( ) Procedure 69
- AutoCode Reference 3-8 ni.com 70
- External_Input ( ) Procedure 72
- Calling UCBs 74
- Ada Fixed-Point Arithmetic 78
- Generic Functions 79
- Instantiated Functions 79
- Package Dependencies 80
- Compilation Example 83
- Fixed-Point Type Declarations 85
- Bit-Wise Functions 88
- Operator Instantiations 88
- Sample Package 90
- Conversion Functions 93
- Language-Defined Conversion 94
- Truncation Conversion 94
- Explicit Rounding Conversion 94
- Known Ada Compiler Problems 97
- No-Op Conversion Function 98
- Operating Systems 100
- Configuration Items 101
- Table Syntax 101
- Scheduler Priority Table 103
- Subsystem Table 103
- AutoCode Reference 4-6 ni.com 105
- Processor IP Name Table 106
- Version Table 107
- Using the Configuration File 107
- Generated Code Architecture 109
- Signal Naming 110
- Duplicate Names 110
- Selection of a Signal Name 110
- Global Storage 111
- -vbco option is not used 112
- Subsystems 113
- Top-Level SuperBlock 114
- Block Ordering 114
- Interface Layers 114
- Sys_ExtIn and Sys_ExtOut, to 115
- Single-Rate System 116
- Multi-Rate System 116
- Sample and Hold 116
- Static Data Within Subsystems 117
- R_P and I_P 117
- State Data 117
- Procedure Data 118
- Pre-init Phase 118
- Copy Back and Duplicates 118
- Standard Procedures 119
- Unrolled Interface 120
- Phases and Error Handling 120
- Referenced Percent Variables 120
- Procedure Arguments 123
- U, Y, S, and I 123
- Caller Identification 126
- Compatibility Issues 127
- Macro Procedure 128
- Asynchronous Procedures 129
- Condition Block 130
- BlockScript Block 130
- Inputs and Outputs 131
- Environment Variables 132
- Local Variables 132
- Default Phase 134
- Local Variables and Phases 135
- Discrete Semantics 135
- Continuous Semantics 137
- Looping Concepts 137
- Terminology 137
- Loops and Scalar Code 137
- Vectorized Code 139
- Types of Loops 139
- Parameters 141
- Optimizations 143
- Dead Code Elimination 143
- Implicit Type Conversion 144
- Special Directives 144
- UserCode Block 145
- Parameterized UCB Callout 146
- Software Constructs 147
- BREAK Block 148
- CONTINUE Block 148
- Local Variable Block 148
- Continuous Subsystem 149
- Explicit Phases 150
- Integrator 150
- Limitations 150
- Shared Memory Architecture 151
- Shared Memory Callouts 152
- Callout Naming Convention 152
- Mapping Command Options 153
- Definitions and Conventions 153
- From Local Memory 154
- From Shared Memory 154
- option is not used 155
- @offset = offset plus 1@@ 156
- ) data types 158
- Callout Pairs 159
- syntax to call 160
- options 161
- Scalar Gain Block Example 165
- Vectorized Gain Block Example 166
- Array Subscripts 167
- Signal Connectivity 168
- Block Outputs 168
- Block Inputs 168
- AutoCode Reference 6-6 ni.com 169
- Vectorization Modes 170
- Vector Labels and Names 171
- Vectorization Features 177
- Split-Merge Inefficiency 180
- Split Vector 180
- External Outputs 184
- Copy-Back 184
- Eliminating Copy-Back 186
- Other Copy-Back Scenarios 186
- Matrix Outputs 191
- Code Optimization 192
- Chapter 7 Code Optimization 193
- AutoCode Reference 7-2 ni.com 193
- AutoCode Reference 7-4 ni.com 195
- Restart Capability 196
- AutoCode Reference 7-6 ni.com 197
- Merging INIT Sections 199
- -Oreuse 1 202
- Constant Propagation 204
- Optimizing with Matrix Blocks 207
- AutoCode Sim Cdelay Scheduler 211
- Task Posting Policies 212
- Standard AutoCode Scheduler 213
- AutoCode Reference 8-4 ni.com 214
- Scheduler Pipeline 215
- AutoCode Reference 8-6 ni.com 216
- AutoCode Reference 8-8 ni.com 218
- Sim Cdelay Scheduler 219
- Implementation Details 222
- DataStore Priority Problem 223
- Global Scope Signals and 228
- Parameterless Procedures 228
- Data Monitoring/Injection 229
- AutoCode Reference 9-4 ni.com 231
- Parameterless Procedure 232
- Output Specification 233
- Global Output Connection 233
- Issues and Limitations 235
- SystemBuild Simulator 235
- Connection to External Output 236
- Recommendations 236
- Naming Convention 236
- Model Documentation 236
- Explicit Sequencing 236
- Command Options 237
- Technical Support and 238
- Professional Services 238
- AutoCode Reference I-2 ni.com 240
- AutoCode Reference I-4 ni.com 242
- AutoCode Reference I-6 ni.com 244
- AutoCode Reference I-8 ni.com 246
Produits connexes et manuels pour Matériel National-instruments AutoCode NI MATRIX
(17 pages)© 2020, manymanuals.fr. Tous droits réservés | 0.039 s |
Manymanuals.com
Manymanuals.de
Manymanuals.fr
Manymanuals.it
Manymanuals.pl
Manymanuals.cz
Manymanuals.es
Manymanuals-pt.com
Commentaires sur ces manuels