chore(tb): remove Verilator TBs + framework; parallelize XSIM runs

Verilator is no longer used (all verification is via Vivado XSIM). Remove:
- 10 per-module tb_*.cpp Verilator testbenches
- the entire test_framework/ Verilator harness (lib/, run_all.py, config.json,
  per-module test_plan.json/gen_vectors.py, golden vectors, reports)
- stale specs: verilator-conventions.md, test_framework/structure.md
  (index.md updated to drop the Verilator entry)

Parallelize run_tb.sh K x case execution (modules stay serial):
- new run_xsim_jobs helper: compile+elaborate once (serial, populates the
  shared xsim.dir), then run each (K,case) xsim in its own private workdir
  with a COPY of xsim.dir (~1MB) so concurrent same-snapshot runs don't clobber
  each other's runtime logs. Each workdir symlinks the repo sync_rtl tree so
  the TB's repo-relative $readmemh vector paths resolve.
- top/enc/dec runners refactored to build a (snapshot:K:case) spec list and
  hand it to run_xsim_jobs; ordered PASS/FAIL summary + per-job /tmp logs
  preserved. Bare './run_tb.sh top' now also takes the parallel path.

Speedup (20 cores): top full sweep 2:11 -> 0:51 (~2.6x), ~320% CPU.
Verified: top (11) / enc (9) / dec (9) all PASS; missing-vector runs still
fail (file-not-found guard -> exit 1).
This commit is contained in:
2026-06-29 16:05:06 +08:00
parent 030931d4e5
commit e46d2258d9
135 changed files with 347 additions and 22072 deletions

View File

@@ -1,196 +0,0 @@
// tb_sha3.cpp - Verilator C++ testbench for sha3_top
//
// Reads test vectors from +VECTOR_FILE=plusarg.
// Format: "MM DDDD..." (mode hex, then 512-bit message hex)
// MM: "00"=G, "01"=H, "10"=J
// DDDD...: 128 hex chars representing 512-bit data_i
//
// Drives DUT with mode and input, waits for ready_o/valid_o,
// prints "RESULT: OUTPUT_HEX\n" to stdout.
//
// Clock: 10ns period. Reset: 2 cycles.
#include <iostream>
#include <fstream>
#include <string>
#include <sstream>
#include <cstdlib>
#include <cstring>
#include <cstdint>
#include "Vsha3_top.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(Vsha3_top* dut) {
dut->clk = !dut->clk;
main_time += (vluint64_t)(CLK_PERIOD_NS / 2.0);
dut->eval();
dut->clk = !dut->clk;
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 a hex string into Verilator WData words (16 x 32-bit = 512 bits).
// Hex string is MSB-first (leftmost hex char = bits 511:508).
// data[0] = bits[31:0], data[15] = bits[511:480].
static void hex_to_512(const std::string& hex, uint32_t data_words[16]) {
for (int w = 0; w < 16; w++) data_words[w] = 0;
int len = (int)hex.length();
// nibble_idx increments for each hex char from RIGHT to LEFT
int nibble_idx = 0;
for (int i = len - 1; i >= 0; i--) {
char c = hex[i];
if (c == ' ' || c == '\t') continue;
int nib = hex_char_to_nibble(c);
int word_idx = nibble_idx / 8; // 8 nibbles per 32-bit word
int shift = (nibble_idx % 8) * 4; // shift within the word
if (word_idx < 16) {
data_words[word_idx] |= ((uint32_t)nib << shift);
}
nibble_idx++;
}
}
// Print hash_o as hex (MSB-first).
// data[0] = bits[31:0], data[15] = bits[511:480].
static void print_hex(uint32_t data_words[16], int bits) {
int full_words = bits / 32;
// Build hex from MSB nibble to LSB nibble
for (int w = full_words - 1; w >= 0; w--) {
uint32_t val = data_words[w];
for (int j = 28; j >= 0; j -= 4) {
int nib = (int)((val >> j) & 0xF);
printf("%01X", nib);
}
}
printf("\n");
}
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
Vsha3_top* dut = new Vsha3_top;
// Initialize
dut->clk = 0;
dut->rst_n = 0;
dut->mode = 0;
for (int w = 0; w < 16; w++) dut->data_i[w] = 0;
dut->valid_i = 0;
dut->ready_i = 0;
// multi-block absorb path disabled for single-block G/H/J tests
dut->mb_en = 0;
dut->mb_valid_i = 0;
dut->mb_last_i = 0;
// Reset: 2 full cycles
for (int i = 0; i < 2; i++) posedge(dut);
dut->rst_n = 1;
// Consumer always ready
dut->ready_i = 1;
std::string line;
vluint64_t cycle = 0;
int vec_count = 0;
while (std::getline(infile, line)) {
if (line.empty() || line[0] == '#') continue;
// Parse: MODE_HEX MESSAGE_HEX
std::istringstream iss(line);
std::string mode_str, data_hex;
if (!(iss >> mode_str >> data_hex)) continue;
if (mode_str.length() < 2) continue;
int mode_val = hex_char_to_nibble(mode_str[1]);
// Set mode
dut->mode = mode_val & 0x3;
// Set data_i from hex string
uint32_t data_words[16];
hex_to_512(data_hex, data_words);
for (int w = 0; w < 16; w++) dut->data_i[w] = data_words[w];
// Assert valid_i for one cycle
dut->valid_i = 1;
// posedge: DUT samples valid_i, starts keccak_core
posedge(dut);
cycle++;
dut->valid_i = 0;
// Wait for valid_o (keccak_core takes ~25 cycles)
do {
posedge(dut);
cycle++;
if (cycle > TIMEOUT_CYCLES) {
std::cerr << "ERROR: Timeout waiting for valid_o (vec " << vec_count << ")" << std::endl;
goto done;
}
} while (!dut->valid_o);
// Read hash_o and print
uint32_t hash_words[16];
for (int w = 0; w < 16; w++) hash_words[w] = dut->hash_o[w];
int out_bits = (mode_val == 0) ? 512 : 256;
printf("RESULT: ");
print_hex(hash_words, out_bits);
// One more cycle for valid_o handshake
posedge(dut);
cycle++;
vec_count++;
}
done:
infile.close();
delete dut;
if (vec_count == 0) {
std::cerr << "ERROR: No vectors processed" << std::endl;
return 1;
}
return 0;
}