// pm_test.cpp

#include "test_case.h"

int main()
{
    // Top-level AXI4-Stream ports for the DUT
    hls::stream<coeff_axis_big_t> in_data;
    hls::stream<coeff_axis_t>     out_data;

    coeff_axis_big_t local_stream1;
    coeff_axis_t     local_stream2;

    int i;

    coeff_t actual_outputs[Nt];
    coeff_t golden_outputs[Nt];  // NEW: buffer for golden result

    // -------------------------------------------------------------------------
    // Write stimulus into input AXI4-Stream
    // -------------------------------------------------------------------------
    for (i = 0; i < Nt; i++)
    {
        coeff_t val1 = input1_vals[i];
        coeff_t val2 = input2_vals[i];

        // Packing: 2×32-bit coeffs into one 64-bit word
        ap_uint<64> word = 0;
        word |= (ap_uint<32>)val1;                         // low  32 bits
        word |= (ap_uint<64>)(ap_uint<32>)val2 << 32;   // high 32 bits

        local_stream1.data = word;
        local_stream1.keep = -1;                      // 0xFF for 64-bit TDATA
        local_stream1.strb = -1;
        local_stream1.last = (i == Nt - 1) ? 1 : 0;

        in_data.write(local_stream1);
    }

    // -------------------------------------------------------------------------
    // Call DUT
    // -------------------------------------------------------------------------
    poly_mult_dil(in_data, out_data);

    // -------------------------------------------------------------------------
    // Read result from output AXI4-Stream
    // -------------------------------------------------------------------------
    for (i = 0; i < Nt; i++)
    {
        local_stream2 = out_data.read();
        actual_outputs[i] = (coeff_t)local_stream2.data;
        // local_stream2.last could be checked here if you want
    }

    // -------------------------------------------------------------------------
    // Compute golden result (software negacyclic product)
    // -------------------------------------------------------------------------
    golden_poly_mult_dil(golden_outputs, input1_vals, input2_vals);

    // -------------------------------------------------------------------------
    // Compare against golden output
    // -------------------------------------------------------------------------
    int ret_val = 0;
    for (i = 0; i < Nt; i++)
    {
        if (golden_outputs[i] != actual_outputs[i])
        {
            ret_val++;
            std::cout << "Mismatch at i = " << i
                      << " golden = " << golden_outputs[i]
                      << " hw = "    << actual_outputs[i]
                      << std::endl;
            break;
        }
    }

    return ret_val;
}