// SPDX-FileCopyrightText: 2024 Filip Leonarski, Paul Scherrer Institute // SPDX-License-Identifier: GPL-3.0-only #include #include "../hls_simulation/hls_cores.h" int test(size_t nframes, int16_t min, int16_t max) { std::cout << "Nframes " << nframes << " min " << min << " max " << max << std::endl; if ((nframes == 0) || (nframes > 16)) { std::cerr << "summation amount must be between 1 and 16" << std::endl; return 1; } std::vector input_frame(nframes * RAW_MODULE_SIZE); STREAM_768 input; STREAM_768 output; hls::stream compl_in; hls::stream compl_out; std::vector output_frame_ref_32(RAW_MODULE_SIZE, 0); std::vector output_frame_32(RAW_MODULE_SIZE, 0); std::mt19937 g1(1387); std::uniform_int_distribution dist(min, max); for (auto &i: input_frame) i = dist(g1); for (int n = 0; n < nframes * RAW_MODULE_SIZE; n++) { if ((input_frame[n] == INT16_MIN) || (output_frame_ref_32[n % RAW_MODULE_SIZE] == INT24_MIN)) output_frame_ref_32[n % RAW_MODULE_SIZE] = INT24_MIN; else if ((input_frame[n] == INT16_MAX) || (output_frame_ref_32[n % RAW_MODULE_SIZE] == INT24_MAX)) output_frame_ref_32[n % RAW_MODULE_SIZE] = INT24_MAX; else output_frame_ref_32[n % RAW_MODULE_SIZE] += input_frame[n]; } auto input_frame_512 = (ap_uint<512>*) input_frame.data(); ap_uint<768> action_control = 0; ACT_REG_NSUMMATION(action_control) = nframes - 1; input << packet_768_t { .data = action_control, .user = 0 }; for (int i = 0; i < nframes * RAW_MODULE_SIZE * sizeof(uint16_t) / 64; i++) { ap_int<16> tmp1[32]; ap_int<24> tmp2[32]; unpack32(input_frame_512[i], tmp1); for (int j = 0; j < 32; j++) tmp2[j] = tmp1[j]; input << packet_768_t{.data = pack32(tmp2), .user = 0}; } input << packet_768_t { .user = 1 }; ap_uint<128> packet_mask; for (int i = 0; i < 128; i++) packet_mask[i] = 1; for (int i = 0; i < nframes; i++) compl_in << axis_completion{.packet_mask = packet_mask, .frame_number = 100 + i, .packet_count = 128, .last = 0}; compl_in << axis_completion{.last = 1}; ap_uint<1> idle; frame_summation(input, output, compl_in, compl_out, idle); if (compl_in.size() != 0) { std::cout << "compl_in should be empty: " << compl_in.size() << std::endl; return 1; } if (compl_out.size() != 2) { std::cout << "compl_out should be size 2: " << compl_out.size() << std::endl; return 1; } if (input.size() != 0) return 1; if (output.size() != RAW_MODULE_SIZE * sizeof(uint16_t) / 64 + 2) return 1; output.read(); for (int i = 0; i < RAW_MODULE_SIZE * sizeof(uint16_t) / 64 ; i++) { ap_int<24> tmp24[32]; ap_uint<768> tmp = output.read().data; unpack32(tmp, tmp24); for (int j = 0; j < 32; j++) output_frame_32[i * 32 + j] = tmp24[j]; } output.read(); axis_completion cmpl; cmpl = compl_out.read(); if (cmpl.frame_number != 100/nframes) { std::cout << "Wrong output frame number" << std::endl; return 1; } if (cmpl.packet_count != 128*nframes) { std::cout << "Wrong output frame number" << std::endl; return 1; } if (cmpl.packet_mask != packet_mask) { std::cout << "Wrong packet mask" << std::endl; return 1; } compl_out.read(); if (output_frame_32 != output_frame_ref_32) { std::cout << "Input and output don't match" << std::endl; for (int i = 0; i < RAW_MODULE_SIZE; i++) { if (output_frame_32[i] != output_frame_ref_32[i]) { std::cout << output_frame_32[i] << " " << output_frame_ref_32[i] << std::endl; for (int n = 0; n < nframes; n++) std::cout << input_frame[n * RAW_MODULE_SIZE + i] << " "; std::cout << std::endl; } } return 1; } return 0; } int main() { size_t nframes = 5; if (test(nframes, 0, 5000) != 0) return 1; if (test(nframes, 0, INT16_MAX-1) != 0) return 1; //if (test(nframes, INT16_MIN, INT16_MAX) != 0) // return 1; return 0; }