Hello all,
I have a code from the Intel HLS examples. There is a C++ file for the master test bench operators. That tutorial demonstrate use of the getInterfaceAtIndex(int) operator on a memory-mapped master. Howver, in that code there is C++ template usage with a function. I am trying to understand the purpose of template usage. I will be grateful if any one can tell about it... Below is the whole code
#include <iostream>
#include <stdio.h>
#include <stdlib.h>
#include "HLS/hls.h"
using namespace std;
using namespace ihc;
typedef mm_master<unsigned int, dwidth<32>, awidth<64>, aspace<1>, latency<0>, maxburst<1> > mm_src_t;
typedef mm_master<unsigned int, dwidth<32>, awidth<64>, aspace<3>, latency<0>, maxburst<1> > mm_dst_t;
#define SEED 3
#define NUM_SAMPLES 8
//#define DUMP_BUFFERS
int verify_data(int *src, int *dst, int count, unsigned int x=0) {
int errors = 0;
for(int i = 0; i < count; i++) {
if(dst[i] != (src[i] + x)) {
errors++;
printf("mismatch at index %d: 0x%llx != 0x%llx\n", i, (unsigned long long)dst[i], (unsigned long long)src[i] + x);
}
}
return errors;
}
template<typename T> void dump_buffers(const char* title, T *ref, T *dst, unsigned size) {
#ifdef DUMP_BUFFERS
// Title
printf("%s:\n", title);
if(size == 0) return;
// Reference
printf(" REF:");
for(unsigned i=0; i<size-1; ++i) {
printf(" 0x%04x,", ref[i]);
if(i % 8 == 7) printf("\n ");
}
printf(" 0x%04x\n", ref[size-1]);
// Destination
printf(" DST:");
for(unsigned i=0; i<size-1; ++i) {
printf(" 0x%04x,", dst[i]);
if(i % 8 == 7) printf("\n ");
}
printf(" 0x%04x\n", dst[size-1]);
#endif
}
component void add_x(mm_src_t &src, mm_dst_t &dst, unsigned int x) {
*dst = *src + x;
}
int main () {
int x = 0;
int errors = 0;
int src_mem[NUM_SAMPLES];
int dst_mem[NUM_SAMPLES];
// initialize data
srand(SEED);
for (int i=0; i<NUM_SAMPLES; ++i) {
src_mem[i] = rand() % 2048;
dst_mem[i] = -1;
}
dump_buffers("Initial", src_mem, dst_mem, NUM_SAMPLES);
/*
1)
This code creates a new mm_master interface for each iteration of the loop.
*/
x = 0x10;
for (int i=0; i<NUM_SAMPLES; ++i) {
mm_src_t src_mm_i(&src_mem[i], 1*sizeof(int));
mm_dst_t dst_mm_i(&dst_mem[i], 1*sizeof(int));
ihc_hls_enqueue_noret(&add_x, src_mm_i, dst_mm_i, x);
}
ihc_hls_component_run_all(add_x);
dump_buffers("New mm_master for each index", src_mem, dst_mem, NUM_SAMPLES);
errors += verify_data(src_mem, dst_mem, NUM_SAMPLES, x);
/*
2)
This code demonstartes use of the getInterfaceAtIndex() function which can
be used to index into a mm_master object. This can be useful when iterating
over an array and invoking a component on different indices of the array.
*/
x = 0x20;
mm_src_t src_mm(src_mem, NUM_SAMPLES*sizeof(int));
mm_dst_t dst_mm(dst_mem, NUM_SAMPLES*sizeof(int));
for (int i=0; i<NUM_SAMPLES; ++i) {
ihc_hls_enqueue_noret(&add_x, src_mm.getInterfaceAtIndex(i), dst_mm.getInterfaceAtIndex(i), x);
}
ihc_hls_component_run_all(add_x);
dump_buffers("Use getInterfaceAtIndex()", src_mem, dst_mem, NUM_SAMPLES);
errors += verify_data(src_mem, dst_mem, NUM_SAMPLES, x);
if (errors) {
printf("FAILED with %d errors\n", errors);
} else {
printf("PASSED\n");
}
}
