Phase 2.2: NTT-domain polynomial pointwise multiplication. - basecase_mul.v: degree-1 base-case multiply (c0,c1) with Barrett - poly_mul_zeta_rom.v: 128-entry zeta ROM for PolyMul - poly_mul_sync.v: FSM (IDLE→LOAD 256 cycles→COMPUTE 256 cycles→DONE) Verified: 5/5 vectors bit-exact vs Python PolyMul reference
187 lines
5.2 KiB
C++
187 lines
5.2 KiB
C++
// tb_poly_mul.cpp - Verilator C++ testbench for poly_mul_sync
|
|
//
|
|
// Reads test vectors from +VECTOR_FILE= plusarg.
|
|
// Format: one line with 512 space-separated hex values:
|
|
// A[0] A[1] ... A[255] B[0] B[1] ... B[255]
|
|
// Each value is a 3-digit hex (12-bit).
|
|
//
|
|
// Drives DUT with 256 paired (A,B) coefficients, waits for valid_o,
|
|
// then reads 256 output coefficients via valid/ready handshake.
|
|
// Prints "RESULT: COEFF0 COEFF1 ... COEFF255\n" to stdout.
|
|
//
|
|
// Clock: 10ns period. Reset: 2 cycles.
|
|
|
|
#include <iostream>
|
|
#include <fstream>
|
|
#include <string>
|
|
#include <sstream>
|
|
#include <vector>
|
|
#include <cstdlib>
|
|
#include <cstring>
|
|
#include <cstdint>
|
|
#include "Vpoly_mul_sync.h"
|
|
#include "verilated.h"
|
|
|
|
#define CLK_PERIOD_NS 10.0
|
|
#define TIMEOUT_CYCLES 500000
|
|
|
|
static vluint64_t main_time = 0;
|
|
|
|
double sc_time_stamp() {
|
|
return main_time;
|
|
}
|
|
|
|
// Toggle clock: both edges + eval (one full cycle)
|
|
static void posedge(Vpoly_mul_sync* dut) {
|
|
dut->clk = 1;
|
|
main_time += (vluint64_t)(CLK_PERIOD_NS / 2.0);
|
|
dut->eval();
|
|
dut->clk = 0;
|
|
main_time += (vluint64_t)(CLK_PERIOD_NS / 2.0);
|
|
dut->eval();
|
|
}
|
|
|
|
static int hex_char_to_nibble(char c) {
|
|
if (c >= '0' && c <= '9') return c - '0';
|
|
if (c >= 'A' && c <= 'F') return c - 'A' + 10;
|
|
if (c >= 'a' && c <= 'f') return c - 'a' + 10;
|
|
return 0;
|
|
}
|
|
|
|
// Parse 3-char hex token to 12-bit value.
|
|
static uint16_t hex3_to_val(const std::string& tok) {
|
|
uint16_t val = 0;
|
|
for (size_t i = 0; i < tok.length() && i < 3; i++) {
|
|
val = (val << 4) | hex_char_to_nibble(tok[i]);
|
|
}
|
|
return val & 0xFFF;
|
|
}
|
|
|
|
// Format 12-bit value as 3-char hex.
|
|
static std::string val_to_hex3(uint16_t val) {
|
|
char buf[4];
|
|
snprintf(buf, sizeof(buf), "%03X", val & 0xFFF);
|
|
return std::string(buf);
|
|
}
|
|
|
|
int main(int argc, char** argv) {
|
|
Verilated::commandArgs(argc, argv);
|
|
|
|
// Parse +VECTOR_FILE= plusarg
|
|
const char* vector_file = NULL;
|
|
for (int i = 1; i < argc; i++) {
|
|
std::string arg(argv[i]);
|
|
if (arg.rfind("+VECTOR_FILE=", 0) == 0) {
|
|
vector_file = argv[i] + 13;
|
|
}
|
|
}
|
|
|
|
if (!vector_file) {
|
|
std::cerr << "ERROR: +VECTOR_FILE= not specified" << std::endl;
|
|
return 1;
|
|
}
|
|
|
|
std::ifstream infile(vector_file);
|
|
if (!infile.is_open()) {
|
|
std::cerr << "ERROR: Cannot open vector file: " << vector_file << std::endl;
|
|
return 1;
|
|
}
|
|
|
|
// Instantiate DUT
|
|
Vpoly_mul_sync* dut = new Vpoly_mul_sync;
|
|
|
|
// Initialize
|
|
dut->clk = 0;
|
|
dut->rst_n = 0;
|
|
dut->coeff_a_in = 0;
|
|
dut->coeff_b_in = 0;
|
|
dut->valid_i = 0;
|
|
dut->ready_i = 0;
|
|
|
|
// Reset: 2 full cycles
|
|
for (int i = 0; i < 2; i++) posedge(dut);
|
|
dut->rst_n = 1;
|
|
|
|
std::string line;
|
|
vluint64_t cycle = 0;
|
|
int vec_count = 0;
|
|
|
|
while (std::getline(infile, line)) {
|
|
if (line.empty() || line[0] == '#') continue;
|
|
|
|
// Parse: A[0] A[1] ... A[255] B[0] B[1] ... B[255]
|
|
// Total: 512 hex values, one line
|
|
std::istringstream iss(line);
|
|
std::string tok;
|
|
|
|
std::vector<uint16_t> input_coeffs(512);
|
|
int coeff_idx = 0;
|
|
while (iss >> tok && coeff_idx < 512) {
|
|
input_coeffs[coeff_idx] = hex3_to_val(tok);
|
|
coeff_idx++;
|
|
}
|
|
if (coeff_idx != 512) {
|
|
std::cerr << "ERROR: Expected 512 coefficients, got " << coeff_idx
|
|
<< " (vec " << vec_count << ")" << std::endl;
|
|
continue;
|
|
}
|
|
|
|
// ---- Load 256 A+B pairs ----
|
|
while (!dut->ready_o) {
|
|
posedge(dut); cycle++;
|
|
if (cycle > TIMEOUT_CYCLES) goto timeout_err;
|
|
}
|
|
for (int i = 0; i < 256; i++) {
|
|
dut->coeff_a_in = input_coeffs[i];
|
|
dut->coeff_b_in = input_coeffs[256 + i];
|
|
dut->valid_i = 1;
|
|
posedge(dut); cycle++;
|
|
dut->valid_i = 0;
|
|
if (cycle > TIMEOUT_CYCLES) goto timeout_err;
|
|
}
|
|
|
|
// ---- Wait for valid_o (DUT computing) ----
|
|
dut->ready_i = 1;
|
|
while (!dut->valid_o) {
|
|
posedge(dut);
|
|
cycle++;
|
|
if (cycle > TIMEOUT_CYCLES) goto timeout_err;
|
|
}
|
|
|
|
// ---- Read 256 output coefficients ----
|
|
printf("RESULT: ");
|
|
for (int i = 0; i < 256; i++) {
|
|
// Wait for valid_o to be asserted (data is valid NOW)
|
|
while (!dut->valid_o) {
|
|
posedge(dut);
|
|
cycle++;
|
|
if (cycle > TIMEOUT_CYCLES) goto timeout_err;
|
|
}
|
|
|
|
// Capture coefficient BEFORE consuming posedge
|
|
uint16_t coeff_val = (uint16_t)(dut->coeff_out & 0xFFF);
|
|
printf("%s%s", val_to_hex3(coeff_val).c_str(),
|
|
(i < 255) ? " " : "");
|
|
|
|
// Consume this coefficient: posedge with ready_i=1
|
|
posedge(dut);
|
|
cycle++;
|
|
}
|
|
printf("\n");
|
|
|
|
vec_count++;
|
|
}
|
|
|
|
std::cout << "Processed " << vec_count << " vectors" << std::endl;
|
|
infile.close();
|
|
delete dut;
|
|
return (vec_count > 0) ? 0 : 1;
|
|
|
|
timeout_err:
|
|
std::cerr << "ERROR: Timeout at cycle " << cycle
|
|
<< " (vec " << vec_count << ")" << std::endl;
|
|
infile.close();
|
|
delete dut;
|
|
return 1;
|
|
}
|