Trigger/TrigT1/TrigT1RPChardware/src/Matrix.cxx

Go to the documentation of this file.

/*
  Copyright (C) 2002-2026 CERN for the benefit of the ATLAS collaboration
*/
 
#include "TrigT1RPChardware/Matrix.h"
#include "TrigT1RPChardware/CMAVERSION.h"
 
#include <array>
#include <cmath>
#include <fstream>
#include <iostream>
#include <cstring> //strcpy
#include <sstream>
 
 
 
using namespace std;
 
//----------------------------------------------------------------------------//
// number of thresholds
const ubit16 Matrix::s_nthres = 3;
// number of channels in side=0 and side=1;
const ubit16 Matrix::s_nchan[2] = {32, 64};
// Bunch Crossing period, ns
const float Matrix::s_BCtime = 25.0;
// Number of DLL cycles
const ubit16 Matrix::s_NDLLCYC = 8;
// DLL period, ns
const float Matrix::s_DLLtime = s_BCtime / (float)s_NDLLCYC;
// ReadOut offset in DLL steps
const sbit16 Matrix::s_ROOffset = 1;
// number of channel in a group for the timing setting
const sbit16 Matrix::s_timeGroupA = 16;
const sbit16 Matrix::s_timeGroupB = 8;
// number of bits for a CMAword word
const sbit16 Matrix::s_wordlen = 32;
//----------------------------------------------------------------------------//
namespace {
    constexpr std::array<ubit16, 2> inds(ubit16 channel) {
        // static cast back to ubit16 because arithmetic operators implicitly cast to int (don't accept smaller types)
        return {static_cast<ubit16>(channel / Matrix::s_wordlen), static_cast<ubit16>(channel % Matrix::s_wordlen)};
    }
    constexpr ubit16 bitstatus(const CMAword *p, ubit16 channel) {
        CMAword j{1};
        std::array<ubit16, 2> id = inds(channel);
        return *(p + id[0]) & j << id[1] ? 1 : 0;
    }
}  // namespace
//----------------------------------------------------------------------------//
Matrix::Matrix(int run, int event, CMAword debug, int subsys, int proj, int sect, int padadd, int lowhig, int add[2], int locadd, int NOBXS, int BCZERO) :
    BaseObject(Hardware, "Matrix") {
    ubit16 df = 0;  // debug flag address
    m_run = run;
    m_event = event;
    //
    // debug flags first
    //
    m_matrixDebug = debug;
 
    if (m_matrixDebug & 1 << df) {
        cout << "=============================================" << endl
             << "Constructor of Matrix called with parameters:" << endl
             << subsys << " " << proj << " " << sect << " " << lowhig << " " << add[0] << add[1] << endl
             << "=============================================" << endl;
    }
    m_thisBC = 0;  // temporary initialization
 
    m_BCzero = BCZERO;
    m_Nbunch = NOBXS;
    m_nclock = s_NDLLCYC * m_Nbunch;
    // m_BCzero=(m_nclock/s_NDLLCYC)/2;      // default initialization of m_BCzero
    // user setting by setBCzero
 
    //
    //
    //                 BCzero
    //                   |
    //                   |
    //                   |
    //                   V
    //  BCID=-2  BCID=-1 .BCID= 0  BCID=+1  BCID=+2
    //                   .
    //                   .
    // ********+********+********+********+********
    // 01234567 01234567 01234567 01234567 01234567
    //
    //
    // Matrix attributes
    //
    m_subsystem = subsys;
    m_projection = proj;
    m_sector = sect;
    m_pad = padadd;
    m_lowhigh = lowhig;
    //
    m_address[0] = add[0];
    m_address[1] = add[1];
    m_localadd = locadd;
    //
    // initialize some variables
    //
    initDat();
    //
    // initialize the RPC data structure
    //
    initRPCpointers();
    //
    // initialize the CMAword registers
    //
    initRegisters();
    //
    // initialize the Configuration pointers
    //
    initPointers();
    //
    // set default CM parameters configuration
    //
    setDefaultConfiguration();
}
//-----------------------------------------------------------------------//
Matrix::~Matrix() {
    ubit16 df = 1;
    deleteRPCdata();
    if (m_matrixDebug & 1 << df) { cout << "Destructor of Matrix executed " << endl; }
}  // end-of-Matrix::~Matrix()
//-----------------------------------------------------------------------//
void Matrix::deleteRPCdata() {
    ubit16 i;
    // ubit16 df=1;
    rpcdata *rpcpnt, *rpcpntnext;
    //
    // delete the dynamic memory allocated by the Matrix object
    //
    for (i = 0; i < 2; i++) {
        rpcpnt = m_datarpc[i];
        while (rpcpnt) {
            rpcpntnext = rpcpnt->next;
            delete rpcpnt;
            rpcpnt = rpcpntnext;
        }  // end-of-while(rpcpnt)
    }      // end-of-for(i
    // if(rpcpnt) delete rpcpnt;
    //
    // set DEFAULT parameters for the Coincidence Matrix
    //
    initRPCpointers();
    //
}  // end-of-Matrix::deleteRPCdata
//-----------------------------------------------------------------------//
void Matrix::reset() {
    //
    // initialize some variables
    //
    initDat();
    //
    // delete dynamic memory used for RPC data
    //
    deleteRPCdata();
    //
    // initialize the CMAword registers
    //
    initRegisters();
    //
}  // end-of-Matrix::reset
//-----------------------------------------------------------------------//
void Matrix::initRegisters() {
 
    // helper to set arrays to zero (only use with integer arrays!)
    auto zero = [](auto& arr) { memset( arr, 0, sizeof(arr) ); };
    zero( rodat );
    zero( m_input );
    zero( m_prepr );
    zero( m_mjori );
    zero( m_trigger );
    zero( m_trigg );
    zero( m_triggerOverlap );
    zero( overlapRO );
    zero( highestthRO );
    zero( m_k_pattern );
    zero( k_readout );
    zero( m_triggerOverlapRO );
    zero( m_highestth );
    zero( m_overlap );
}  // end-of-Matrix::initRegisters
//-----------------------------------------------------------------------//
void Matrix::initRPCpointers() {
    //
    // initialize the rpcdata structure
    //
    for (ubit16 i = 0; i < 2; i++) { m_datarpc[i] = 0; }  // end-of-for(i
}  // end-of-Matrix::initRPCpointers
//-----------------------------------------------------------------------//
void Matrix::initPointers() {
    //
    // initialize the Configuration pointers
    //
    m_chdly = 0;  // pointer to channel delays
    m_width = 0;  // pointer to pulse widths
    m_locDi = 0;  // pointer to local Coincidence Direction
    m_kRead = 0;  // pointer to threshold pattern for readout
    m_roads = 0;  // pointer to the road Matrix settings
    m_major = 0;  // pointer to the majority
    m_overl = 0;  // pointer to the overlapping channel list
    m_geome = 0;  // pointer to "geometry"
}  // end-of-Matrix::initPointers
//-----------------------------------------------------------------------//
void Matrix::initDat() {
    // put the current BC at the center of the buffer
    m_BCID = -1;
}  // end-of-Matrix::initDat
//-----------------------------------------------------------------------//
void Matrix::setDefaultConfiguration() {
    ubit16 df = 0;  // debug flag address as for constructor
    ubit16 i, j, k, l;
    //
    // set DEFAULT parameters for the Coincidence Matrix
    //
    //
 
    //
    // Coincidence Windows for the the three thresholds
    //
    for (i = 0; i < s_nthres; i++) {
        for (j = 0; j < s_nchan[0]; j++) {
            m_trigRoad[i][j][0] = 0x00000000;
            m_trigRoad[i][j][1] = 0x00000000;
        }  // end-of-for(j
    }      // end-of-for(i
    //
    // Majority setting
    //
    m_majorities[0] = 2;  // majority values for threshold address 0
    m_majorities[1] = 2;  // majority values for threshold address 1
    m_majorities[2] = 2;  // majority values for threshold address 2
    //
    // threshold to be used for coincidence of the lowpt trigger with the
    // external RPC doublet
    //
    m_lowtohigh = s_nthres - 1;
    //
    // threshold to be used to provide the trigger data in the readout;
    // important for the correct functioning of the LVL2 (muFast)
    //
    m_toreadout = 0;  // threshold of the pattern to be sent to the readout
    //
    // address of the threshold to be considered in overlap (dimuon counting)
    //
    m_overlapthres = 0;
    //
    // address of the configuration for the local coincidence
    //
    m_localDirec[0] = 7;  // majority Direction
    m_localDirec[1] = 7;  // majority Direction
    //
    // default for m_matOverlap
    //
    for (i = 0; i < 2; i++) { m_matOverlap[i] = 0; }
    //
    // default for the signal delay, deadtime and pulse width arrays
    //
    for (i = 0; i < 2; i++) {                                    // side
        for (j = 0; j < 2; j++) {                                // layer
            for (k = 0; k < (s_nchan[1] / s_timeGroupB); k++) {  // group
                m_pulseWidth[i][j][k] = 8;
            }                                                    // end-of-for(k
            for (k = 0; k < (s_nchan[1] / s_timeGroupA); k++) {  // group
                m_channDelay[i][j][k] = 0;
            }                                                    // end-of-for(k
            for (k = 0; k < (s_nchan[1] / s_timeGroupB); k++) {  // group
                m_channDeadT[i][j][k] = 0;
            }  // end-of-for(k
        }      // end-of-for(j
    }          // end-of-for(i
    //
    // Masking to 0 and masking to 1
    //
    for (i = 0; i < 2; i++) {                                                // side
        for (j = 0; j < 2; j++) {                                            // layer (or "1/2"=0, "2/2"=1 in case of m_channMask1)
            for (k = 0; k < s_nchan[1]; k++) { m_channMask0[i][j][k] = 0; }  // end-of-for(k
            for (l = 0; l < s_nthres; l++) {
                m_channMask1[l][i][j][0] = 0;
                m_channMask1[l][i][j][1] = 0;
            }  // end-of-for(l=0
            m_channReadOutMask[i][j][0] = 0;
            m_channReadOutMask[i][j][1] = 0;
        }  // end-of-for(j
    }      // end-of-for(i
    //
    // default for trigger dead time
    //
    for (k = 0; k < (s_nchan[0] / s_timeGroupB); k++) {  // group
        m_trigDeadTime[k] = 8;
    }  // end-of-for(
    //
    // initialize m_BunchPhase and m_BunchOffset
    //
    m_BunchPhase = 0;  // use this setting for standard ATLAS simulation;
    // m_BunchPhase=-1;  // use this setting with comparison with the HARDWARE;
    // value fixed with VHDL comparison 1-7 august 2004.
    // use with setBCzero(0);
    m_BunchOffset = 0;  // test with hardware; use with setBCzero(0);
    //
    // default for m_diagonal
    //
    for (i = 0; i < s_nchan[0]; i++) { m_diagonal[i] = 0; }
 
    if (m_matrixDebug & 1 << df) {
        cout << "====================================================================" << endl
             << "Matrix::setDefaultConfiguration: "
             << "Default settings have been loaded." << std::endl
             << "===================================================================" << endl;
        dispDefaultConfiguration();
    }
}  // end-of-Matrix::setDefaultConfiguration
//-----------------------------------------------------------------------//
void Matrix::dispDefaultConfiguration() const {
    ubit16 i, j, k;
    //
    // Coincidence Windows for the the three thresholds
    //
    cout << "=================================" << std::endl
         << "Matrix::dispDefaultConfiguration:" << std::endl
         << "=================================" << std::endl;
    cout << "--------------------------------" << std::endl << "+ Coincidence Windows:          " << std::endl;
    for (i = 0; i < s_nthres; i++) {
        cout << " + Threshold address:" << i << std::endl;
        for (j = 0; j < s_nchan[0]; j++) {
            cout << "    -Channel address:" << j << " Window 63-32 " << std::hex << m_trigRoad[i][j][1] << " Window 31-00 "
                 << m_trigRoad[i][j][0] << std::dec << std::endl;
        }  // end-of-for(j
    }      // end-of-for(i
    //
    // Majority setting
    //
    cout << "--------------------------------" << std::endl
         << "+ Majority addresses:           " << std::endl
         << "--------------------------------" << std::endl;
    for (i = 0; i < s_nthres; i++) { cout << " - threshold address " << i << " value " << m_majorities[i] << std::endl; }
    //
    // threshold to be used for coincidence of the lowpt trigger with the
    // external RPC doublet
    //
    cout << "--------------------------------" << std::endl
         << "+ Threshold address low-to-high: " << m_lowtohigh << std::endl
         << "+ Threshold address low-to-readout: " << m_toreadout << std::endl
         << "+ Threshold address for overlap: " << m_overlapthres << std::endl;
    //
    // address of the configuration for the local coincidence
    //
    cout << "--------------------------------" << std::endl
         << "+ Local coincidence setting:    " << std::endl
         << " -Pivot Plane:       " << m_localDirec[0] << std::endl
         << " -Coincidence Plane: " << m_localDirec[1] << std::endl;
    //
    // Overlap default masking
    //
    cout << "--------------------------------" << std::endl
         << "+ Overlap mask setting:    " << std::endl
         << " -`right' address          " << m_matOverlap[0] << std::endl
         << " -`left ' address          " << m_matOverlap[1] << std::endl;
    //
    // default for the signal delay, deadtime and pulse width arrays
    //
    cout << "-----------------------------------------------" << std::endl
         << "+ Channel pulse-width, delay and deadtime :    " << std::endl;
    for (i = 0; i < 2; i++) {  // side
        if (!i) {
            cout << " +Pivot Plane " << std::endl;
        } else {
            cout << " +Coincidence Plane " << std::endl;
        }
        for (j = 0; j < 2; j++) {  // layer
            cout << "  +Layer " << j << std::endl;
            for (k = 0; k < (s_nchan[i] / s_timeGroupB); k++) {  // group
                cout << "   -group " << k << " pulsewidth " << m_pulseWidth[i][j][k] << std::endl;
            }                                                    // end-of-for(k
            for (k = 0; k < (s_nchan[i] / s_timeGroupA); k++) {  // group
                cout << "   -group " << k << " delay " << m_channDelay[i][j][k] << std::endl;
            }                                                    // end-of-for(k
            for (k = 0; k < (s_nchan[i] / s_timeGroupB); k++) {  // group
                cout << "   -group " << k << " deadtime " << m_channDeadT[i][j][k] << std::endl;
            }  // end-of-for(k
        }      // end-of-for(j
    }          // end-of-for(i
    //
    // Masking to 0
    //
    cout << "-----------------------------------------------" << std::endl << "+ Map of the masked-to-0 channels :    " << std::endl;
    for (i = 0; i < 2; i++) {  // side
        if (!i) {
            cout << " +Pivot Plane " << std::endl;
        } else {
            cout << " +Coincidence Plane " << std::endl;
        }
        for (j = 0; j < 2; j++) {  // layer
            cout << "  +Layer " << j << std::endl;
            for (k = 0; k < s_nchan[1]; k++) {
                if (m_channMask0[i][j][k]) { cout << "   -channel " << k << std::endl; }  // end-of-if
            }                                                                             // end-of-for(k
        }                                                                                 // end-of-for(j
    }                                                                                     // end-of-for(i
    //
    // default for trigger dead time
    //
    cout << "-----------------------------------------------" << std::endl << "+ Trigger Dead Time   " << std::endl;
    for (k = 0; k < (s_nchan[0] / s_timeGroupB); k++) {  // group
        cout << " -group " << k << " Trigger DeadTime " << m_trigDeadTime[k] << std::endl;
    }  // end-of-for(
    //
    // Simulation relevant parameters (used to align with the hardware)
    //
    cout << "-----------------------------------------------" << std::endl
         << "+ Simulation parameters to align with the hardware (not used in CM)" << std::endl;
    //
    // m_BunchPhase and m_BunchOffset
    //
    cout << "+BunchPhase  " << m_BunchPhase << std::endl
         << "  BunchPhase =  0 : to be used for standard ATLAS LVL1 simulation" << std::endl
         << "  BunchPhase = -1 : to be used to compare with the hardware; " << std::endl
         << "                    this value fixed with VHDL comparison 1-7 august 2004." << std::endl
         << "+BunchOffset  " << m_BunchOffset << std::endl
         << "  BunchOffset = 0 : to test with hardware; use this setting with setBCzero(0)." << std::endl;
    //
    // the end
    //
    cout << "======================================" << std::endl
         << "Matrix::dispDefaultConfiguration: Done" << std::endl
         << "======================================" << std::endl;
}  // end-of-Matrix::dispDefaultConfiguration
//-----------------------------------------------------------------------//
void Matrix::putData(int sidemat, int layer, int stripaddress, float time) {
    //
    sbit16 BCID;    // BC  time bin (from 0)
    ubit16 DLLID;   // DLL time bin (from 0)
    ubit16 df = 2;  // debug flag address
    rpcdata *rpcpntnew;
    if (m_matrixDebug & 1 << df) { cout << "Matrix:putData: putting data on Matrix" << endl; }
    //
    BCID = (int)(time / s_BCtime);
    if (time < 0.0) BCID--;  // to cope with negative times
                             //
                             // the next line determines the DLL clock bin associated
                             // to the given time. For historical reasons, a +1 offset has
                             // been used so far:
                             // DLLID= (int)((time-(float)BCID*s_BCtime)/s_DLLtime)+1;
                             // It should be now removed. Aleandro, 11 November 2008.
                             //
    DLLID = (int)((time - (float)BCID * s_BCtime) / s_DLLtime);
    if (DLLID == s_NDLLCYC) {
        BCID++;
        DLLID = 0;
    }  // end-of-if(DLLID
    //
    // put this digi in the dynamic store
    //
    // coverity change
    //   rpcpntnew = new rpcdata;
    //
    // check the stripaddress is consistent with the Matrix dimension
    //
    if (BCID >= m_Nbunch) return;
 
    if (stripaddress >= 0 && stripaddress < s_nchan[sidemat]) {
        // coverity change
        rpcpntnew = new rpcdata;
 
        rpcpntnew->layer = layer;
        rpcpntnew->stripadd = stripaddress;
        rpcpntnew->time = time;
        rpcpntnew->BC = BCID;
        rpcpntnew->DLL = DLLID;
        rpcpntnew->masked = 0;
        rpcpntnew->maskto1 = 0;
        rpcpntnew->deadtime = 0;
        rpcpntnew->delay = 0;
        rpcpntnew->mark = 0;
        rpcpntnew->next = m_datarpc[sidemat];
        //
        // put this element as first in the queue
        //
        m_datarpc[sidemat] = rpcpntnew;
 
    } else {
        throw std::out_of_range("Matrix::putData failure: channel addressed is " + std::to_string(stripaddress) + " for matrix side " +
                                std::to_string(sidemat));
    }  // end-of-if
}  // end-of-putData
//----------------------------------------------------------------------//
void Matrix::putPatt(const Matrix *p) {
    //
    // input:  p  pointer of the low-pt Matrix belonging to the
    //            high-pt one
    //
    CMAword k;
    ubit16 i, j;
    ubit16 df = 3;  // debug flag address
    if (m_matrixDebug & 1 << df) { cout << " method putPatt called; p is " << p << endl; }
    //
    //
    // for(i=0; i<2; i++)           {     // loop on the two majorities
    //  for(j=0; j<m_nclock; j++)     {     // loop on the clock bins
    //   for(k=0; k<2; k++)           {   // loop on buffer words
    //    for(n=0; n<s_nthres; n++) {       // loop on the "s_nthres" registers
    //     m_mjori[n][0][i][j][k] = p->m_k_pattern[j];
    //    }//end-of-for(n
    //   }//end-of-for(k
    //  }//end-of-for(j
    // }//end-of-for(i
    //
    //
    float time = 0.0;
    k = 1;
 
    //   cout<<" RITARDI-1 "<<endl;
    //   for(int iside=0;iside<=1;iside++){
    //     for(int ilayer=0;ilayer<=1;ilayer++){
    //       for(ubit16 k=0; k<(s_nchan[iside]/s_timeGroupA); k++) {
    //         cout<<" side="<<iside<<" layer="<<ilayer<<" group="<<k
    //         <<" delay="<<m_channDelay[iside][ilayer][k]<<endl;
    //       }
    //     }
    //   }//end-of-for(ubit16 k
 
    for (i = 0; i < s_nchan[0]; i++) {
        for (j = 0; j < m_nclock; j++) {
            if (p->m_k_pattern[j] & k) {
                time = ((float)j + 0.5) * s_DLLtime + (float)m_thisBC * s_BCtime - (float)(m_BCzero)*s_BCtime;
                putData(0, 0, i, time);
            }  // end-of-if(p->m_k_pattern
        }      // end-of-for(j
        k = k << 1;
    }  // end-of-for(i
    //
    if (m_matrixDebug & 1 << df) { cout << " copy_pivot; input matrix address " << p << endl; }  // end-of-if(m_matrixDebug&1<<df)
}  // end-of-method putPatt
//----------------------------------------------------------------------//
void Matrix::setRunEvent(int runNum, int eventNum) {
    m_run = runNum;
    m_event = eventNum;
}  // end-of-Matrix::setRunEvent
//----------------------------------------------------------------------//
void Matrix::setBCzero(ubit16 offset) {
    //
    // set the BunchCrossing=0 to the "offset" array address in memory
    //
    if (offset <= m_Nbunch - 1)
        m_BCzero = offset;
    else
        m_BCzero = m_Nbunch - 1;
}  // end-of-setBCzero
//----------------------------------------------------------------------//
void Matrix::setDeadTime(ubit16 deadt) {
    for (ubit16 iside = 0; iside < 2; iside++) {
        for (ubit16 ilayer = 0; ilayer < 2; ilayer++) {
            for (ubit16 k = 0; k < (s_nchan[iside] / s_timeGroupB); k++) { setDeadTime(iside, ilayer, k, deadt); }  // end-of-for(ubit16 k
        }  // end-of-for(ubit16 ilayer
    }      // end-of-for(ubit16 iside
}  // end-of-setDeadTime
//----------------------------------------------------------------------//
void Matrix::setDeadTime(ubit16 iside, ubit16 ilayer, ubit16 deadt) {
    for (ubit16 k = 0; k < (s_nchan[iside] / s_timeGroupB); k++) { setDeadTime(iside, ilayer, k, deadt); }  // end-of-for(ubit16 k
}  // end-of-setDeadTime
//----------------------------------------------------------------------//
void Matrix::setDeadTime(ubit16 iside, ubit16 ilayer, ubit16 igroup, ubit16 deadt) {
    if (iside < 2 && ilayer < 2 && igroup < (s_nchan[1] / s_timeGroupB)) {
        m_channDeadT[iside][ilayer][igroup] = deadt;
        if (iside == 0 && igroup > 3) {
            throw std::out_of_range(
                "Matrix::setDeadTime: problems with side and group addresses: "
                "side=" +
                std::to_string(iside) + " layer=" + std::to_string(ilayer) + " group=" + std::to_string(igroup));
        }  // end-of-if
    } else {
        throw std::out_of_range(
            "Matrix::setDeadTime: problems in adressing pulseWidth: "
            "side=" +
            std::to_string(iside) + " layer=" + std::to_string(ilayer) + " group=" + std::to_string(igroup));
    }  // end-of-if
}  // end-of-setDeadTime
//----------------------------------------------------------------------//
void Matrix::setDelay(ubit16 iside, ubit16 ilayer, ubit16 delay) {
    if (iside > 1 || ilayer > 1) {
        throw std::out_of_range(
            "Matrix::setDelay: problems with side and layer addresses: "
            "side=" +
            std::to_string(iside) + " layer=" + std::to_string(ilayer));
    } else {
        for (ubit16 k = 0; k < (s_nchan[iside] / s_timeGroupA); k++) { setDelay(iside, ilayer, k, delay); }  // end-of-for(ubit16 k
    }
}  // end-of-setDelay
//----------------------------------------------------------------------//
void Matrix::setDelay(ubit16 iside, ubit16 ilayer, ubit16 igroup, ubit16 delay) {
    if (iside < 2 && ilayer < 2 && igroup < (s_nchan[1] / s_timeGroupA)) {
        m_channDelay[iside][ilayer][igroup] = delay;
        if (iside == 0 && igroup > 3) {
            throw std::out_of_range(
                "Matrix::setDelay: problems with side and group addresses:"
                "side=" +
                std::to_string(iside) + " layer=" + std::to_string(ilayer) + " group=" + std::to_string(igroup));
        }  // end-of-if
    } else {
        throw std::out_of_range(
            "Matrix::setDelay: problems in adressing pulseWidth:"
            "side=" +
            std::to_string(iside) + " layer=" + std::to_string(ilayer) + " group=" + std::to_string(igroup));
    }  // end-of-if
}  // end-of-setDelay
//----------------------------------------------------------------------//
void Matrix::setPulseWidth(ubit16 length) {
    for (ubit16 iside = 0; iside < 2; iside++) {
        for (ubit16 ilayer = 0; ilayer < 2; ilayer++) {
            for (ubit16 k = 0; k < (s_nchan[iside] / s_timeGroupB); k++) { setPulseWidth(iside, ilayer, k, length); }  // end-of-for(ubit16
                                                                                                                       // k
        }  // end-of-for(ubit16 ilayer
    }      // end-of-for(ubit16 iside
}  // end-of-setPulseWidth
//----------------------------------------------------------------------//
void Matrix::setPulseWidth(ubit16 iside, ubit16 ilayer, ubit16 length) {
    for (ubit16 k = 0; k < (s_nchan[iside] / s_timeGroupB); k++) { setPulseWidth(iside, ilayer, k, length); }  // end-of-for(ubit16 k
}  // end-of-setPulseWidth
//----------------------------------------------------------------------//
void Matrix::setPulseWidth(ubit16 iside, ubit16 ilayer, ubit16 igroup, ubit16 length) {
    if (iside < 2 && ilayer < 2 && igroup < s_nchan[1] / s_timeGroupB) {
        m_pulseWidth[iside][ilayer][igroup] = length;
        if (iside == 0 && igroup > 3) {
            throw std::out_of_range(
                "Matrix::setDelay: problems with side and group addresses:"
                "side=" +
                std::to_string(iside) + " layer=" + std::to_string(ilayer) + " group=" + std::to_string(igroup));
        }  // end-of-if
    } else {
        throw std::out_of_range(
            "Matrix::setDelay: problems in adressing pulseWidth:"
            "side=" +
            std::to_string(iside) + " layer=" + std::to_string(ilayer) + " group=" + std::to_string(igroup));
    }  // end-of-if
}  // end-of-setPulseWidth
//----------------------------------------------------------------------//
void Matrix::setMask0(ubit16 iside, ubit16 ilayer, ubit16 ichannel) {
    if (iside > 1 || ilayer > 1 || ichannel > (s_nchan[iside] - 1)) {
        throw std::out_of_range(
            "Matrix::setMask0: problems with side/layer/channel addresses: "
            "side=" +
            std::to_string(iside) + " layer=" + std::to_string(ilayer) + " channel=" + std::to_string(ichannel));
    } else {
        m_channMask0[iside][ilayer][ichannel] = 1;
    }
}  // end-of-setMask0
//----------------------------------------------------------------------//
void Matrix::setMask1(ubit16 ithreshold, ubit16 iside, ubit16 imajority, ubit16 ichannel) {
    if (ithreshold > 2 || iside > 1 || imajority > 1 || ichannel > (s_nchan[iside] - 1)) {
        throw std::out_of_range(
            "Matrix::setMask1: problems with side/layer/channel addresses: "
            "threshold= " +
            std::to_string(ithreshold) + " side=" + std::to_string(iside) + " majority=" + std::to_string(imajority) +
            " channel=" + std::to_string(ichannel));
    } else {
        set_to_1(&m_channMask1[ithreshold][iside][imajority][0], ichannel);
    }
}  // end-of-setMask1
//----------------------------------------------------------------------//
void Matrix::setMask1(ubit16 ithreshold, ubit16 iside, ubit16 imajority) {
    for (int ichannel = 0; ichannel < s_nchan[iside]; ichannel++) { setMask1(ithreshold, iside, imajority, ichannel); }
}  // end-of-setMask1
//----------------------------------------------------------------------//
void Matrix::setMask1(ubit16 ithreshold, ubit16 iside) {
    for (int imajority = 0; imajority < 2; imajority++) { setMask1(ithreshold, iside, imajority); }
}  // end-of-setMask1
//----------------------------------------------------------------------//
void Matrix::setMaskReadOut(ubit16 iside, ubit16 ilayer, ubit16 ichannel) {
    if (iside > 1 || ilayer > 1 || ichannel > (s_nchan[iside] - 1)) {
        throw std::out_of_range(
            "Matrix::setMaskReadout: problems with side/layer/channel addresses: "
            "side=" +
            std::to_string(iside) + " layer=" + std::to_string(ilayer) + " channel=" + std::to_string(ichannel));
    } else {
        set_to_1(&m_channReadOutMask[iside][ilayer][0], ichannel);
    }
}  // end-of-setMaskReadOut
//----------------------------------------------------------------------//
void Matrix::setMaskReadOut(ubit16 iside, ubit16 ilayer, ubit16 ichannel, ubit16 nchannels) {
    for (ubit16 i = ichannel; i < (ichannel + nchannels - 1); i++) { setMaskReadOut(iside, ilayer, ichannel, i); }
}  // end-of-setMaskReadOut
//----------------------------------------------------------------------//
void Matrix::setConfig(int *l, ubit16 *p, int *k, CMAword *r, int *m, CMAword *o, sbit32 *g) {
    //
    // input:  l   pointer to local coincidence direction
    //         p   pointer to pulse width
    //         k   pointer to threshold for k-readout pattern
    //         r   pointer to the three Matrix roads;
    //         m   pointer to the three majorities;
    //         o   pointer to the overlapping channel list;
    //         g   pointer to "geometry"
    //
    ubit16 i, j, jg;
    ubit16 addGroupMax = 0;
    // ubit16 df=4;  // debug flag address
    m_locDi = l;  // local coincidence direction              //
    m_width = p;  // pulse width                              //
    m_chdly = 0;  // delay                                    //
    m_kRead = k;  // k-readout address                        //
    m_roads = r;  // programmed coincidence window address    //
    m_major = m;  // programmed majority address              //
    m_overl = o;  // programmed overlap flag address          //
    m_geome = g;  // pre-calculated geometry                  //
    for (i = 0; i < 2; i++) { setLocalDirection(i, *(m_locDi + i)); }
    // set Pulse Widths
    for (i = 0; i < 2; i++) {  // side
        addGroupMax = s_nchan[0] / s_timeGroupB;
        if (i) addGroupMax = s_nchan[1] / s_timeGroupB;
        for (j = 0; j < 2; j++) {                   // layer
            for (jg = 0; jg < addGroupMax; jg++) {  // group
                                                    // setPulseWidth(i,j,jg,*(m_width+jg+8*j+16*i));
                                                    // setDelay     (i,j,jg,*(m_chdly+jg+8*j+16*i));
            }                                       // end-of-for(jg
        }                                           // end-of-for(j
    }                                               // end-of-for(i
    // set address of threshold to be readout
    setKReadOut(*m_kRead);
    // set majority levels
    for (i = 0; i < s_nthres; i++) { setMajority(i, *(m_major + i)); }
    // set coincidence roads for the s_nthres thresholds
    for (i = 0; i < s_nthres; i++) {
        for (j = 0; j < s_nchan[0]; j++) {
            setRoad(i, j, 0, *(m_roads + s_nchan[0] * 2 * i + 2 * j + 0));
            setRoad(i, j, 1, *(m_roads + s_nchan[0] * 2 * i + 2 * j + 1));
        }  // end-of-for(j
    }      // end-of-for(i
    // set the overlap registers
    for (i = 0; i < 2; i++) { setMatOverlap(i, *(m_overl + i)); }
    for (i = 0; i < s_nchan[0]; i++) { setDiagonal(i, *(m_geome + i)); }
}  // end-of-Matrix::setConfig
//----------------------------------------------------------------------//
void Matrix::setLocalDirection(ubit16 add, int content) {
    if (add > 1) {
        throw std::out_of_range("Matrix::setLocalDirection :  add=" + std::to_string(add) + " not valid");
    } else {
        m_localDirec[add] = content;
    }  // end-of-if;
}  // end-of-Matrix::setLocalDirection
//----------------------------------------------------------------------//
void Matrix::setKReadOut(int content) {
    if (content < 0 || content >> 2) {
        throw std::out_of_range("Matrix::setKReadout :  threshold address = " + std::to_string(content) + " not valid");
    } else {
        m_toreadout = content;
    }  // end-of-if
}  // end-of-Matrix::setKReadOut
//----------------------------------------------------------------------//
void Matrix::setOverlaThres(int content) {
    if (content >= 0 && content < s_nthres) {
        m_overlapthres = content;
    } else {
        m_overlapthres = 0;
    }
}  // end-of-setOverlapThres
//----------------------------------------------------------------------//
void Matrix::setTrigDeadTime(ubit16 igroup, ubit16 deadt) {
    if (igroup < 4) {
        m_trigDeadTime[igroup] = deadt;
    } else {
        throw std::out_of_range("Matrix::setTrigDeadTime :  igroup= " + std::to_string(igroup) + " not valid");
    }
}  // end-of-setTrigDeadTime
//----------------------------------------------------------------------//
void Matrix::setTrigDeadTime(ubit16 deadt) {
    for (ubit16 k = 0; k < (s_nchan[0] / s_timeGroupB); k++) { setTrigDeadTime(k, deadt); }  // end-of-for(ubit16 k
}  // end-of-setTrigDeadTime
//----------------------------------------------------------------------//
void Matrix::setMajority(ubit16 add, int content) {
    if (add >= s_nthres) {
        throw std::out_of_range("Matrix::setMajority :  add=" + std::to_string(add) + " not valid");
    } else {
        m_majorities[add] = content;
    }  // end-of-if
}  // end-of-Matrix::setMajority
//----------------------------------------------------------------------//
void Matrix::setRoad(ubit16 addThres, ubit16 addChn, ubit16 add64, CMAword content) {
    if (addThres >= s_nthres || addChn > s_nchan[0] || add64 > 1) {
        throw std::out_of_range("Matrix::setRoad :  addThres= " + std::to_string(addThres) + " addChn= " + std::to_string(addChn) +
                                " add64= " + std::to_string(add64) + " not valid");
    } else {
        m_trigRoad[addThres][addChn][add64] = content;
    }  // end-of-if
}  // end-of-Matrix::setRoad
//----------------------------------------------------------------------//
void Matrix::setRoad(ubit16 addThres, ubit16 addChn, char road[17]) {
    if (addThres >= s_nthres || addChn > s_nchan[0]) {
        throw std::out_of_range("Matrix::setRoad :  addThres= " + std::to_string(addThres) + " addChn= " + std::to_string(addChn));
    } 
    CMAword the32[2] = {0, 0};
    ubit16 outflag = char2int(road, the32);
    if (outflag) {
        throw std::runtime_error("Matrix::setRoad; outflag from char2int is positive: " + std::to_string(outflag));
    }
    m_trigRoad[addThres][addChn][0] = the32[0];
    m_trigRoad[addThres][addChn][1] = the32[1];
}  // end-of-Matrix::setRoad
//----------------------------------------------------------------------//
void Matrix::setMatOverlap(ubit16 add, CMAword content) {
    if (add > 1) {
        throw std::out_of_range("Matrix::setMatOverlap :  add= " + std::to_string(add) + " not valid");
    } else {
        m_matOverlap[add] = content;
    }
}  // end-of-Matrix::setMatOverlap
//----------------------------------------------------------------------//
void Matrix::setDiagonal(ubit16 add, sbit32 content) {
    if (add > s_nchan[0]) {
        throw std::out_of_range("Matrix::setDiagonal :  add= " + std::to_string(add) + " not valid");
    } else {
        m_diagonal[add] = content;
    }  // end-of-if
}  // end-of-Matrix::setDiagonal
//----------------------------------------------------------------------//
CMAword Matrix::getMatOverlap(ubit16 add) const {
    CMAword output = 0;
    if (add > 1) {
        throw std::out_of_range("Matrix::getMatOverlap :  add= " + std::to_string(add) + " not valid");
    } else {
        output = m_matOverlap[add];
    }
    return output;
}  // end-of-Matrix::getMatOverlap
//----------------------------------------------------------------------//
CMAword Matrix::getRoad(ubit16 addThres, ubit16 addChn, ubit16 add64) const {
    CMAword output = 0;
    if (addThres >= s_nthres || addChn > s_nchan[0] || add64 > 1) {
        throw std::out_of_range("Matrix::getRoad :  addThres= " + std::to_string(addThres) + " addChn= " + std::to_string(addChn) +
                                " add64= " + std::to_string(add64) + " not valid");
    } else {
        output = m_trigRoad[addThres][addChn][add64];
    }  // end-of-if
    return output;
}  // end-of-Matrix::getRoad
//----------------------------------------------------------------------//
int Matrix::getMajority(ubit16 add) const {
    int output = 0;
    if (add >= s_nthres) {
        throw std::out_of_range("Matrix::getMajority :  add= " + std::to_string(add) + " not valid");
    } else {
        output = m_majorities[add];
    }  // end-of-if
    return output;
}  // end-of-Matrix::getMajority
//----------------------------------------------------------------------//
void Matrix::execute() {
    // returns coincidence flag:  0 if there is no trigger for all thresholds
    //                            1 if there is at least one threshold with trigger
    //                            2 if the lowtohigh threshold (or higher) is m_trigg
    //
    // execute this active CM
    //
    ubit16 df = 4;  // debug flag address
    if (m_matrixDebug & 1 << df) {
        cout << "====================" << endl
             << "|                  |" << endl
             << "|  Matrix::execute |" << endl
             << "|  --------------- |" << endl
             << "|                  |" << endl
             << "====================" << endl;
        show_attributes();
    }  // end-of-if(m_matrixDebug&1<<df)
    //
    // 0) programmed windows (s_nthres) and time calibration constants for this
    //    matrix
    if (m_matrixDebug & 1 << df) wind();
    // 1) store deadtime in dyn. structure; deadtime is applied by "applyDeadTime"
    storeDeadtime();
    // 2) masking;
    masking();
    // 3) delay
    delay();
    // 5) fill the CMA
    load();
    // 6) apply deadtime response to Input signals
    deadTime();
    copyDataToReadOut();
    // 7) preprocessing
    prepro();
    // 8) coincidence;
    coincide();
    //
    makeOut();
 
    // Code to be activated for testing the VHDL simulation.
    // makeOutPattern ();
    //
}  // end-of-method execute;
//----------------------------------------------------------------------//
void Matrix::storeDeadtime() {
    ubit16 df = 5;
    rpcdata *rpchit;
    if (m_matrixDebug & 1 << df) {
        cout << "--------------------------" << endl << "|  Matrix::storeDeadtime |" << endl << "--------------------------" << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
    for (ubit16 i = 0; i < 2; i++) {
        rpchit = m_datarpc[i];
        while (rpchit) {
            rpchit->deadtime = m_channDeadT[i][rpchit->layer][rpchit->stripadd / s_timeGroupB];
            rpchit = rpchit->next;
        }  // end-of-while(rpchit
    }      // end-of-for(ubit16 i
}  // end-of-Matrix::storeDeadtime
//----------------------------------------------------------------------//
void Matrix::masking() {
    ubit16 df = 6;
    rpcdata *rpchit;
    if (m_matrixDebug & 1 << df) {
        cout << "--------------------" << endl << "|  Matrix::masking |" << endl << "--------------------" << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
    for (ubit16 i = 0; i < 2; i++) {
        rpchit = m_datarpc[i];
        while (rpchit) {
            rpchit->masked = m_channMask0[i][rpchit->layer][rpchit->stripadd];
            rpchit = rpchit->next;
        }  // end-of-while(rpchit
    }      // end-of-for(ubit16 i
}  // end-of-Matrix::masking
//----------------------------------------------------------------------//
void Matrix::delay() {
    ubit16 df = 7;
    rpcdata *rpchit;
    if (m_matrixDebug & 1 << df) {
        cout << "--------------------" << endl << "|  Matrix::delay   |" << endl << "--------------------" << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
    for (ubit16 i = 0; i < 2; i++) {
        rpchit = m_datarpc[i];
        while (rpchit) {
            rpchit->delay = m_channDelay[i][rpchit->layer][rpchit->stripadd / s_timeGroupA];
            rpchit = rpchit->next;
        }  // end-of-while(rpchit
    }      // end-of-for(ubit16 i
}  // end-of-Matrix::delay
//----------------------------------------------------------------------//
void Matrix::load() {
    ubit16 df = 8;
    sbit32 abs_time, timeadd;
    rpcdata *rpcpnt;
    //
    if (m_matrixDebug & 1 << df) {
        cout << "--------------------" << endl << "|  Matrix::load    |" << endl << "--------------------" << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
    //
    for (ubit16 i = 0; i < 2; i++) {  // "i" is the CMA side address
        rpcpnt = m_datarpc[i];
        while (rpcpnt) {
            if (m_matrixDebug & 1 << df) {
                cout << "  Layer= " << rpcpnt->layer << " stripadd= " << rpcpnt->stripadd << " time= " << rpcpnt->time
                     << " mask= " << rpcpnt->masked << " BC= " << rpcpnt->BC << " DLL= " << rpcpnt->DLL << " delay= " << rpcpnt->delay
                     << endl;
            }  // end-of-if(m_matrixDebug&1<<df)
            abs_time = s_NDLLCYC * rpcpnt->BC + rpcpnt->DLL + rpcpnt->delay;
 
            timeadd = abs_time - s_NDLLCYC * m_thisBC  // synchronize the data to m_thisBC
                      + s_NDLLCYC * m_BCzero;          // put m_thisBC at the center of the buffer
 
            if (m_matrixDebug & 1 << df) { cout << " abs_time= " << abs_time << " timeadd= " << timeadd << endl; }
 
            //
            // store this digit if it is within time window and not masked
            //
            if (timeadd >= 0 && timeadd < m_nclock && !rpcpnt->masked) {
                if (m_matrixDebug & 1 << df) {
                    cout << " setting input with side " << i << " " << rpcpnt->layer << " " << timeadd << " 0"
                         << " for channel " << rpcpnt->stripadd << " timeadd " << timeadd << endl;
                }  // end-of-if(m_matrixDebug&1<<df)
                set_to_1(&m_input[i][rpcpnt->layer][timeadd][0], rpcpnt->stripadd);
            }  // end-of-if(timeadd
            //
            rpcpnt = rpcpnt->next;
            //
        }  // end-of-while(rpcpnt)
    }      // end-of-for(i
}  // end-of-Matrix::load()
//------------------------------------------------------------------------//
void Matrix::copyDataToReadOut() {
    //
    // copy input to rodat
    //
    for (ubit16 i = 0; i < 2; i++) {                 // side address
        for (ubit16 j = 0; j < 2; j++) {             // layer address
            for (ubit16 k = 0; k < m_nclock; k++) {  // clock bins
                for (ubit16 l = 0; l < 2; l++) {     // the two words to make 64 bits
                    rodat[i][j][k][l] = m_input[i][j][k][l] & ~m_channReadOutMask[i][j][l];
                }  // end-of-for(l
            }      // end-of-for(k
        }          // end-of-for(j
    }              // end-of-for(i
}  // end-of-copyDataToReadOut
//------------------------------------------------------------------------//
void Matrix::prepro() {
    ubit16 df = 9;
    if (m_matrixDebug & 1 << df) {
        cout << "-------------------------" << endl << "| matrix: preprocessing |" << endl << "-------------------------" << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
    //
    // 1) pulse width
    pulse_width();
    // 2) declustering
    declus();
    // 3) majority
    majori();
    // 4) mask to 1
    maskTo1();
}  // end-of-method prepro
//------------------------------------------------------------------------//
void Matrix::coincide() {
    //
    // input  none
    //
    // returns trigger flag
    //
    //
    sbit16 BCaddress = 0;
    //
    ubit16 df = 10;
    ubit16 i, j, l, m;
    ubit16 thres, chan, nconf, conf[4][2];
    for (i = 0; i < 4; i++) {
        for (j = 0; j < 2; j++) { conf[i][j] = 0; }
    }
    //
    if (m_matrixDebug & 1 << df) {
        cout << "--------------------" << endl << "| matrix: coincide |" << endl << "--------------------" << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
    //
    // nconf gives the number of possible configurations compatible with
    // the required majority;
    // conf[i][j] gives the two indices "j" for the side-x (pivot) and side-y
    // array respectively correspondent to the configurtion number "i"
    //
    if (m_matrixDebug & 1 << df) {
        cout << " Matrix::coincidence; lowhigh= " << m_lowhigh << endl
             << " Matrix::coincidence; majority array ="
             << " " << m_majorities[0] << " " << m_majorities[1] << " " << m_majorities[2] << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
    //
    for (i = 0; i < m_nclock; i++) {                  // loop on clock cycles
        for (thres = 0; thres < s_nthres; thres++) {  // loop on the three possible thresholds
                                                      // thresholds address increases with
                                                      // increasing pT
            nconf = config(thres, &conf[0][0]);       // number of config. for this threshold
                                                      //
                                                      // if(m_matrixDebug&1<<df) {
            // cout<<"nconf="<<nconf<<" conf="<<endl;
            // for(int ii=0;ii<4;ii++){ cout<<" "<<conf[ii][0]<<endl;}
            // cout<<endl;
            // for(int ii=0;ii<4;ii++){ cout<<" "<<conf[ii][1]<<endl;}
            // cout<<endl;
            //}//end-of-if(m_matrixDebug&1<<df)
 
            for (l = 0; l < nconf; l++) {           // loop on all config. for this threshold
                for (j = 0; j < s_nchan[0]; j++) {  // loop on all side-x channels
                                                    //
                                                    // coincidence evaluation:
                                                    // for the given configuration "l" and the given channel "j" in the
                                                    // pivot plane, the coincidence requires
                                                    // a) the required majority "m_mjori[thres][0][conf[l][0]][i][0]"
                                                    //    for channel "j"
                                                    // b) the required majority "m_mjori[thres][1][conf[l][0]][i][0,1]"
                                                    //    in coincidence with the programmed road in "*(m_roads+2*j)" and
                                                    //    "*(m_roads+2*j+1)
                                                    //
                                                    //    if(   bitstatus(&m_mjori[thres][0][conf[l][0]][i][0],j)
                                                    //          *( m_mjori[thres][1][conf[l][1]][i][0]&(*(m_roads+32*2*thres+2*j+0))
                                                    //    |  m_mjori[thres][1][conf[l][1]][i][1]&(*(m_roads+32*2*thres+2*j+1)))){
 
                    if (bitstatus(&m_mjori[thres][0][conf[l][0]][i][0], j) &&
                        ((m_mjori[thres][1][conf[l][1]][i][0] & m_trigRoad[thres][j][0]) |
                         (m_mjori[thres][1][conf[l][1]][i][1] & m_trigRoad[thres][j][1]))) {
                        if (m_matrixDebug & 1 << df) {
                            cout << "coincidence!!!"
                                 << " clock=" << i << " xchan=" << j << endl;
                        }  // end-of-if(m_matrixDebug
 
                        set_to_1(&m_trigg[thres][i], j);
                        BCaddress = ((i + m_BunchPhase) / s_NDLLCYC) + m_BunchOffset;
                        if (BCaddress >= 0 && BCaddress < m_Nbunch)
                            m_trigger[thres][BCaddress] = 1;  // set "m_trigger" to 1 for this threshold...
                                                              // ...and this Bunch Crossing
 
                    }  // end-of-if(bitstatus...
                }      // end-of-for(j
            }          // end-of-for(l
            //
            //  for(m=0;m<2;m++) {   // loop on left-right sides to evaluate overlap
            //   if(!m_triggerOverlapRO[i][m])
            //       m_triggerOverlapRO[i][m]       = (m_trigg[thres][i] & m_matOverlap[m]);
            //   if(!m_triggerOverlap[i/s_NDLLCYC][m])
            //       m_triggerOverlap[i/s_NDLLCYC][m] = (m_trigg[thres][i] & m_matOverlap[m]);
            //  }//end-of-for(m
        }  // end-of-for(thres
    }      // end-of-for(i
 
    if (CMAVERSION == 2004) {
        //
        // build the rise pulse for "m_trigg" (320 MHz trigger output) signals
        //
        CMAword previousTime = 0;
        for (chan = 0; chan < s_nchan[0]; chan++) {
            for (thres = 0; thres < s_nthres; thres++) {  // loop on the three possible thresholds
                for (i = m_nclock - 1; i > 0; i--) {      // loop on clock cycles
                    previousTime = bitstatus(&m_trigg[thres][i - 1], chan);
                    if (bitstatus(&m_trigg[thres][i], chan) && previousTime) { set_to_0(&m_trigg[thres][i], chan); }  // end-of-if(bitstatus
                }                                                                                                     // end-of-for(i
            }                                                                                                         // end-of-for(thres
        }                                                                                                             // end-of-for(chan
    }  // end-of-if(CMAVERSION
    // - - - - - - - - - - -
 
    for (chan = 0; chan < s_nchan[0]; chan++) {
        for (thres = 0; thres < s_nthres; thres++) {  // loop on the three possible thresholds
            for (i = m_nclock - 1; i > 0; i--) {      // loop on clock cycles
                if (bitstatus(&m_trigg[thres][i], chan)) {
                    for (ubit16 idead = 1; idead < m_trigDeadTime[chan / s_timeGroupB]; idead++) {
                        set_to_0(&m_trigg[thres][i + idead], chan);
                    }  // end-of-for(
                }      // end-of-bitstatus
            }          // end-of-for(i
        }              // end-of-for(thres
    }                  // end-of-for(chan
 
    if (CMAVERSION == 2004) {
        //
        // ...now determine the correspondent "m_trigger" (40 MHz trigger output)
        //
        for (thres = 0; thres < s_nthres; thres++) {  // loop on the three possible thresholds
            // reset trigger
            for (j = 0; j < m_Nbunch; j++) { m_trigger[thres][j] = 0; }  // end-of-for(j
            // compute new "m_trigger"
            for (i = 0; i < m_nclock; i++) {  // loop on clock cycles
                BCaddress = ((i + m_BunchPhase) / s_NDLLCYC) + m_BunchOffset;
                if (BCaddress >= 0 && BCaddress < m_Nbunch) {
                    if (m_trigg[thres][i]) m_trigger[thres][BCaddress] = 1;
                }  // emd-of-if(BCadd
            }      // end-of-for(i
        }          // end-of-for(thres
    }              // end-of-if(CMAVERSION
    // - - - - - - - - - - -
    //
    // find triggers in overlap regions
    //
    // for(thres=0; thres<s_nthres; thres++){ // loop on the three possible thresholds
    thres = m_overlapthres;
    for (i = 0; i < m_nclock; i++) {  // loop on clock cycles
        if (m_trigg[thres][i]) {
            for (m = 0; m < 2; m++) {  // loop on left-right sides to evaluate overlap
                if (!m_triggerOverlapRO[i][m]) { m_triggerOverlapRO[i][m] = (m_trigg[thres][i] & m_matOverlap[m]); }
                BCaddress = ((i + m_BunchPhase) / s_NDLLCYC) + m_BunchOffset;
                if (BCaddress >= 0 && BCaddress < m_Nbunch) {
                    if (!m_triggerOverlap[BCaddress][m]) m_triggerOverlap[BCaddress][m] = (m_trigg[thres][i] & m_matOverlap[m]);
                }  // end-of-if(BCaddress
            }      // end-of-for(m
 
        }  // end-of-for(m_trigg
    }      // end-of-for(i
    //}//end-of-for(thres
    //
    // normalize m_triggerOverlapRO
    //
    for (i = 0; i < m_nclock; i++) {  // loop on clock cycles
        for (m = 0; m < 2; m++) {     // normalize to 1 overlap flags
            m_triggerOverlapRO[i][m] = m_triggerOverlapRO[i][m] ? 1 : 0;
        }  // end-of-for(m
    }      // end-of-for(i
    //
    // normalize m_triggerOverlap
    //
    for (i = 0; i < m_Nbunch; i++) {  // loop on bunches
        for (m = 0; m < 2; m++) {     // normalize to 1 overlap flags
            m_triggerOverlap[i][m] = m_triggerOverlap[i][m] ? 1 : 0;
        }  // end-of-for(m
    }      // end-of-for(i
}  // end-of-Matrix::coincide
//------------------------------------------------------------------------//
void Matrix::maskTo1() {
    ubit16 df = 11;
    if (m_matrixDebug & 1 << df) {
        cout << "---------------------" << endl << "| Matrix::mask_to_1 |" << endl << "---------------------" << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
 
    ubit16 i, j, k, l, m;
    for (m = 0; m < s_nthres; m++) {                // thresholds
        for (i = 0; i < 2; i++) {                   // side address
            for (j = 0; j < 2; j++) {               // majority address
                for (l = 0; l < s_nchan[i]; l++) {  // channel
                    if (bitstatus(&m_channMask1[m][i][j][0], l)) {
                        for (k = 0; k < m_nclock; k++) {  // clock bins
                            set_to_1(&m_mjori[m][i][j][k][0], l);
                        }  // end-of-for(k
                    }      // end-of-if(m_channMask1
                }          // end-of-for(l
            }              // end-of-for(j
        }                  // end-of-for(i
    }                      // end-of-for(m
 
}  // end-of-Matrix::mask_to_1
//------------------------------------------------------------------------//
void Matrix::deadTime() {
    ubit16 i, j, k, l;
    ubit16 temp[s_NDLLCYC*8];
    //
    for (i = 0; i < 2; i++) {                           // loop on both matrix sides
        for (j = 0; j < 2; j++) {                       // loop on both trigger layers
            for (k = 0; k < s_nchan[i]; k++) {          // loop on all channels
                                                        //
                                                        // set to 1 the bins in "temp" where signals are "on" for the first time
                                                        //
                for (l = 0; l < (m_nclock - 1); l++) {  // loop on clock bins
                    ((!bitstatus(&m_input[i][j][l][0], k)) && bitstatus(&m_input[i][j][l + 1][0], k)) ? temp[l + 1] = 1 : temp[l + 1] = 0;
                }  // end-of-for(l
                temp[0] = bitstatus(&m_input[i][j][0][0], k);
                //
                // transfer to "input" the signals in "temp" far enough each other in time
                //
                sbit16 lastUp = -1;
                for (l = 0; l < m_nclock; l++) {  // loop on clock bins
                                                  //
                    if (!temp[l]) {
                        set_to_0(&m_input[i][j][l][0], k);
                    } else {
                        //
                        if ((lastUp < 0) || (l - lastUp) >= m_channDeadT[i][j][k / s_timeGroupB]) {
                            lastUp = l;
                            set_to_1(&m_input[i][j][l][0], k);
                        } else {
                            set_to_0(&m_input[i][j][l][0], k);
                        }  // end-of-if
                    }      // end-of-if
                    //
                }  // end-of-for(l
            }      // end-of-for(k
        }          // end-of-for(j
    }              // end-of-for(i
}  // end-of-Matrix::deadTime
//------------------------------------------------------------------------//
void Matrix::pulse_width() {
    ubit16 df = 12;
    ubit16 i, j, l, m;
    sbit16 k;
    if (m_matrixDebug & 1 << df) {
        cout << "-----------------------" << endl << "| Matrix::pulse_width |" << endl << "-----------------------" << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
    //
    for (i = 0; i < 2; i++) {                          // loop on the two Matrix sides
        for (j = 0; j < 2; j++) {                      // loop on both layers
            for (l = 0; l < s_nchan[i]; l++) {         // loop on all channels
                for (k = m_nclock - 1; k >= 0; k--) {  // loop on the m_nclock cycles backwards
                    if (bitstatus(&m_input[i][j][k][0], l)) {
                        // loop on all time bins to be set to 1
                        for (m = k + 1; m < k + m_pulseWidth[i][j][l / s_timeGroupB]; m++) {
                            if (m < m_nclock) {
                                set_to_1(&m_input[i][j][m][0], l);
                            } else {
                                break;
                            }  // end-of-if(m
                        }      // end-of-for(m
                    }          // end-of-if(bitstatus
                }              // end-of-for(k
            }                  // end-of-for(l
        }                      // end-of-for(j
    }                          // end-of-for(i
    //
}  // end-of-Matrix::pulse_width
//------------------------------------------------------------------------//
void Matrix::majori() {
    //
    // input   none
    // returns  nothing
    //
    // Fills the majority registers the 1/2 and 2/2 majority logic
    //
    // Dynamic scanning of "on" channels has not to be done
    // in this first part of the method
    //
    ubit16 i, j, k, l, n;
    ubit16 df = 13;  // debug flag address
    CMAword buffi[2], buffo[2];
    if (m_matrixDebug & 1 << df) {
        cout << "-------------------" << endl << "| Matrix::majori  |" << endl << "-------------------" << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
    //
    // the loop on the CMA sides has to be made as follows:
    // 1) side=0 and side=1 as long as the lowpt Matrix is concerned;
    // 2) side=1 only as long as the highpt Matrix is concerned.
    // To do this, loop from lowhigh address.
    //
    for (n = 0; n < s_nthres; n++) {          // the s_nthres thresholds
        for (i = 0; i < 2; i++) {             // the two Matrix sides
            for (j = 0; j < m_nclock; j++) {  // the clock cycles
                for (k = 0; k < 2; k++) {     // the two words to make 64 bits
                                              // copy layer with address 0,1 to buffi; initialize buffoutput
                    buffi[k] = m_prepr[n][i][1][j][k];
                    buffo[k] = 0;
                }
                shift(&buffi[0], &buffo[0], i);
                for (k = 0; k < 2; k++) {  // the two words
                                           // 1/2 majority
                    m_mjori[n][i][0][j][k] = m_prepr[n][i][0][j][k] | m_prepr[n][i][1][j][k];
                    // 2/2 majority
                    m_mjori[n][i][1][j][k] = m_prepr[n][i][0][j][k] & buffo[k];
                }  // end-of-for(k
                //
                // complete preprocessing...
                //
                for (l = 0; l < s_nchan[i]; l++) {  // loop in the number of channel for this side
                    if (bitstatus(&m_mjori[n][i][1][j][0], l)) {
                        if ((l > 0) && (!bitstatus(&m_mjori[n][i][1][j][0], l - 1))) set_to_0(&m_mjori[n][i][0][j][0], l - 1);
                        if ((l < s_nchan[i] - 1) && (!bitstatus(&m_mjori[n][i][1][j][0], l + 1))) set_to_0(&m_mjori[n][i][0][j][0], l + 1);
                    }  // end-of-if(
                }      // end-of-for(l
                // preprocessing completed
            }  // end-of-for(j
        }      // end-of-for(i
    }          // end-of-for(n
}  // end-of-method majori
//------------------------------------------------------------------------//
void Matrix::shift(CMAword *buffi, CMAword *buffo, ubit16 i) const {
    //
    ubit16 k;
    switch (m_localDirec[i]) {
        case 1:  // n(layer0)-->n(layer1)
            for (k = 0; k < 2; k++) { *(buffo + k) = *(buffi + k); }
            break;
        case 2:  // n(0)-->n-1(1)
            for (k = 0; k < 2; k++) { *(buffo + k) = *(buffi + k) << 1; }
            *(buffo + 1) = (*(buffi + 0) & 0x80000000) ? (*(buffo + 1) | 0x1) : (*(buffo + 1) | 0);
            break;
        case 3:  // case 2 plus case 1
            for (k = 0; k < 2; k++) { *(buffo + k) = *(buffi + k) | (*(buffi + k) << 1); }
            *(buffo + 1) = (*(buffi + 0) & 0x80000000) ? (*(buffo + 1) | 0x1) : (*(buffo + 1) | 0);
            break;
        case 4:  // n(0)-->n+1(1)
            for (k = 0; k < 2; k++) { *(buffo + k) = *(buffi + k) >> 1; }
            *(buffo + 0) = (*(buffi + 1) & 0x1) ? (*(buffo + 0) | 0x80000000) : (*(buffo + 0) | 0);
            break;
        case 5:  // case 4 plus case 1
            for (k = 0; k < 2; k++) { *(buffo + k) = *(buffi + k) | (*(buffi + k) >> 1); }
            *(buffo + 0) = (*(buffi + 1) & 0x1) ? (*(buffo + 0) | 0x80000000) : (*(buffo + 0) | 0);
            break;
        case 6:  // case 4 plus case 2
            for (k = 0; k < 2; k++) { *(buffo + k) = (*(buffi + k) >> 1) | (*(buffi + k) << 1); }
            *(buffo + 0) = (*(buffi + 1) & 0x1) ? (*(buffo + 0) | 0x80000000) : (*(buffo + 0) | 0);
            *(buffo + 1) = (*(buffi + 0) & 0x80000000) ? (*(buffo + 1) | 0x1) : (*(buffo + 1) | 0);
            break;
        case 7:  // case 4 plus case 2 pluse case 1
            for (k = 0; k < 2; k++) { *(buffo + k) = *(buffi + k) | (*(buffi + k) >> 1) | (*(buffi + k) << 1); }
            *(buffo + 0) = (*(buffi + 1) & 0x1) ? (*(buffo + 0) | 0x80000000) : (*(buffo + 0) | 0);
            *(buffo + 1) = (*(buffi + 0) & 0x80000000) ? (*(buffo + 1) | 0x1) : (*(buffo + 1) | 0);
            break;
        default:
            cout << " Matrix::shift -- m_localDirec[" << i << "]=" << m_localDirec[i] << " not expected; m_localDirec forced to be 1 "
                 << endl;
            for (k = 0; k < 2; k++) { *(buffo + k) = *(buffi + k); }
    }
    //
    // correct patter in *(buffo+1) in case side of Matrix (="i" in this method)
    // is = 1
    //
    if (!i) *(buffo + 1) = 0;
    //
}  // end-of-Matrix::shift(buff0, buffi, buffo)
//------------------------------------------------------------------------//
void Matrix::declus() {
    //
    // Dynamic scanning of "on" channels has not to be done
    // in this first part of the method
    //
    ubit16 df = 14;
    ubit16 nup, first, i, j, k, l;
    first = 0;
    if (m_matrixDebug & 1 << df) {
        cout << "------------------------" << endl << "| matrix: declustering |" << endl << "------------------------" << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
    //
    // loop on m_input data
    //
    for (i = 0; i < 2; i++) {                       // loop on the two sides
        for (j = 0; j < 2; j++) {                   // loop on the two layers
            for (k = 0; k < m_nclock; k++) {        // loop on the time bins
                nup = 0;                            // counter of consecutive "on" channels
                for (l = 0; l < s_nchan[i]; l++) {  // loop on the Matrix channels
                    if (bitstatus(&m_input[i][j][k][0], l)) {
                        nup++;
                        if (nup == 1) first = l;
                    }  // end-of-if(bitstatus
                    if (!bitstatus(&m_input[i][j][k][0], l) || (l == (s_nchan[i] - 1))) {
                        if (nup) {
                            reduce(i, j, k, 0, nup, first);
                            nup = 0;
                        }  // end-of-if(nup
                    }      // end-of-if(bitstatus
                }          // end-of-for(l
            }              // end-of-for(k
        }                  // end-of-for(j
    }                      // end-of-for(i
}  // end-of-method declustering
//------------------------------------------------------------------------//
void Matrix::reduce(ubit16 ia, ubit16 ja, ubit16 ka, ubit16 la, ubit16 nup, ubit16 first) {
    //
    // input  ia, ja, ka, la  indices of the Matrix data
    //        nup    cluster size = number of consecutive channels on
    //        first  address of the first channel on in the cluster (from 0)
    // returns  nothing
    //
    // It copies into m_prepr the channels from input selected by the
    // declustering logic.
    //
    //
    ubit16 ncop, i, j, na;
    ubit16 df = 15;
    ncop = 0;
    j = 0;
    if (m_matrixDebug & 1 << df) {
        cout << " --------------------" << endl
             << " |  Matrix::reduce  |" << endl
             << " --------------------" << endl
             << " nup= " << nup << " first " << first << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
    //
    // analyse nup value and apply the cluster reduction according to it.
    // j     is the first channel of the cluster retained after declustering;
    // ncop  is the cluster size after declustering
    //
    if (nup <= 2) {
        j = first;
        ncop = nup;
    } else if (nup == 3) {
        j = first + 1;
        ncop = 1;
    } else if (nup == 4) {
        j = first + 1;
        ncop = 2;
    } else if (nup > 4) {
        j = first + 2;
        ncop = nup - 4;
    }                                                                                   // end-of-if
    if (m_matrixDebug & 1 << df) { cout << " j= " << j << " ncop= " << ncop << endl; }  // end-of-if(m_matrixDebug&1<<df)
    //
    // copy the reduced cluster into the "s_nthres" m_prepr registers
    //
    for (na = 0; na < s_nthres; na++) {
        for (i = j; i < j + ncop; i++) {  // loop on each element of the reduced cluster
            set_to_1(&m_prepr[na][ia][ja][ka][la], i);
        }  // end-of-for(i
    }      // end-of-for(na
}  // end-of-reduce
//------------------------------------------------------------------------//
void Matrix::makeOut() {
    ubit16 i, j, k;
    //
    // fill k-pattern
    //
    for (j = 0; j < m_nclock; j++) {  // loop on the clock bins
        m_k_pattern[j] = m_trigg[m_lowtohigh][j];
        k_readout[j] = m_trigg[m_toreadout][j];
    }  // end-of-for(j
    //
    // find the highest satisfied threshold;
    // identify Bunch Crossing and put it in BCID.
    //
    for (j = 0; j < m_Nbunch; j++) {
        for (i = 0; i < s_nthres; i++) {
            if (m_trigger[i][j]) {
                m_highestth[j] = i + 1;  // put threshold address+1 (correct threshold value)
                if (m_BCID < 0)
                    m_BCID = j            // local Bunch Crossing IDentifier
                             + m_thisBC;  // re-syncronize to absolute BCID
                                          //             - m_BCzero; // remove offset used
            }                             // end-of-if(m_trigger
        }                                 // end-of-for(i
    }                                     // end-of-for(j
    //
    // Trigger in Overlapping channels is reported in m_triggerOverlap
    //
    for (j = 0; j < m_Nbunch; j++) { m_overlap[j] = m_triggerOverlap[j][0] + 2 * m_triggerOverlap[j][1]; }  // end-of-for(j
    //
    // find the highest satisfied threshold for ReadOut pourposes;
    //
    for (j = 0; j < m_nclock; j++) {            // loop on the clock bins
        for (i = 0; i < s_nthres; i++) {        // loop on thresholds
            for (k = 0; k < s_nchan[0]; k++) {  // loop on channels (pivot side)
                if (m_trigg[i][j] & (1 << k)) highestthRO[j] = i + 1;
            }  // end-of-for(k
        }      // end-of-for(i
    }          // end-of-for(j
    //
    //
    // Trigger in Overlapping channels is reported in m_triggerOverlapRO
    //
    for (j = 0; j < m_nclock; j++) { overlapRO[j] = m_triggerOverlapRO[j][0] + 2 * m_triggerOverlapRO[j][1]; }  // end-of-for(j
    //
}  // end-of-Matrix::makeOut
//------------------------------------------------------------------------//
void Matrix::makeTestPattern(ubit16 mode, ubit16 ktimes) {
    //
    int df = 16;
    const ubit16 maxchan = 100;
    const ubit16 maxtimes = 1000;
    ubit16 i, j, l;
    ubit16 IJ[maxtimes][4]{};
    //uses >800k stack space in this function
    //coverity[STACK_USE]
    ubit16 channels[maxtimes][4][maxchan]{};
    float times[maxtimes]{};
    char plane[4][3]{};
    const float timeOffsetHit = 114.675;
    //
    rpcdata *rpcpnt;
    float timemin;
    //
    strcpy(plane[0], "I0");
    strcpy(plane[1], "I1");
    strcpy(plane[2], "J0");
    strcpy(plane[3], "J1");
    if (m_matrixDebug & 1 << df) {
        cout << "-------------------------------" << endl
             << "|  Matrix::makeTestPattern    |" << endl
             << "-------------------------------" << endl;
    }  // end-of-if(m_matrixDebug&1<<df)
    //
    int ntimes = 0;
    ubit16 completed = 0;
    //
    // first reset the marker flags
    //
    for (i = 0; i < 2; i++) {
        rpcpnt = m_datarpc[i];
        while (rpcpnt) {
            rpcpnt->mark = 0;
            rpcpnt = rpcpnt->next;
        }  // end-of-while(rpcpnt
    }      // end-of-for(i
    //
    while (!completed) {
        completed = 1;
        timemin = 999999999.;
        for (i = 0; i < 2; i++) {  // "i" is the CMA side address
            rpcpnt = m_datarpc[i];
            while (rpcpnt) {
                if (rpcpnt->time < timemin && !rpcpnt->mark) {
                    timemin = rpcpnt->time;
                    completed = 0;
                }  // end-of-if(rpcnt
                rpcpnt = rpcpnt->next;
            }  // end-of-while(rpcpnt)
        }      // end-of-for(i
        if (!completed) {
            if (ntimes < maxtimes) ntimes += 1;
            times[ntimes - 1] = timemin;
            for (i = 0; i < 2; i++) {  // "i" is the CMA side address
                rpcpnt = m_datarpc[i];
                while (rpcpnt) {
                    if (rpcpnt->time == timemin) {
                        rpcpnt->mark = 1;
                        if (IJ[ntimes - 1][rpcpnt->layer + 2 * i] < maxchan) { IJ[ntimes - 1][rpcpnt->layer + 2 * i] += 1; }
                        channels[ntimes - 1][rpcpnt->layer + 2 * i][IJ[ntimes - 1][rpcpnt->layer + 2 * i] - 1] = rpcpnt->stripadd;
                    }  // end-of-if(rpcnt
                    rpcpnt = rpcpnt->next;
                }  // end-of-while(rpcpnt)
            }      // end-of-for(i
        }          // end-of-if(!completed
    }              // end-of-while(!completed)
    //
    //
    // open output file
    //
    ofstream vhdlinput;
    vhdlinput.open("k-trigger.output", ios::app);
    if (!vhdlinput) {
        cout << " File for vhdl analysis not opened. " << endl << " ==================================" << endl << endl;
    } else {
        if (m_matrixDebug & 1 << df) {
            cout << " File for vhdl analysis correctly opened" << endl << endl;
        }  // end-of-if(m_matrixDebug&1<<df)
    }      // end-of-if(!vhdlinput
    if (mode) {
        vhdlinput << " RUN " << m_run << " EVENT " << m_event << " WINDOW " << m_Nbunch;
        vhdlinput << " LINES " << (ntimes + ktimes) << std::endl;
    }  // end-of-if(mode
    for (l = 0; l < ntimes; l++) {
        vhdlinput << " TIME " << times[l] + timeOffsetHit << " ";
        for (i = 0; i < 4; i++) {
            vhdlinput << plane[i][0] << plane[i][1] << " " << IJ[l][i] << " ";
            for (j = 0; j < IJ[l][i]; j++) { vhdlinput << channels[l][i][IJ[l][i] - 1 - j] << " "; }  // end-of-for(j
        }                                                                                             // end-of-for(i
        vhdlinput << std::endl;
    }  // end-of-for(l
    //
    vhdlinput.close();
}  // end-of-makeTestPattern
//------------------------------------------------------------------------//
void Matrix::makeOutPattern() {
    // int df=17;
    const float timeOffsetKPa = 168.125;
    const float timeOffsetThr = 210.500;
    ubit16 i, l;
    CMAword bit;
    ubit16 chanHistory[32]{};
    const ubit16 maxchan = 100;
    const ubit16 maxtimes = m_nclock;
    ubit16 ntimes, newtime;
    //uses 50k stack space in this function
    //coverity[STACK_USE]
    ubit16 nchannels[s_NDLLCYC*8][2][2], channels[s_NDLLCYC*8][2][2][maxchan];
    float time, times[s_NDLLCYC*8]{0};
    //
    // trigger registers: k-trigger (historically was k-pattern)
    //
    ntimes = 0;
    for (i = 0; i < m_nclock; i++) { nchannels[i][0][0] = 0; }
    time = (float)m_thisBC * s_BCtime - ((float)(m_BCzero * s_NDLLCYC)) * s_DLLtime;
    for (i = 0; i < m_nclock; time += s_DLLtime, i++) {
        bit = 1;
        newtime = 1;
        for (l = 0; l < s_nchan[0]; l++) {
            if (m_k_pattern[i] & bit) {
                if (!chanHistory[l]) {
                    if (newtime) {
                        if (ntimes < maxtimes) ntimes += 1;
                        times[ntimes - 1] = time;
                    }  // end-of-if(newtime
                    if (nchannels[ntimes - 1][0][0] < maxchan) nchannels[ntimes - 1][0][0] += 1;
                    channels[ntimes - 1][0][0][nchannels[ntimes - 1][0][0] - 1] = l;
                    newtime = 0;
                }  // end-of-if(!chanHistory
                chanHistory[l] = 1;
            } else {  // if(m_k_pattern
                chanHistory[l] = 0;
            }  // end-of-if(m_k_pattern
            bit = bit << 1;
        }  // end-of-for(l
    }      // end-of-for(i
 
    ubit16 nthresPass, thresPass[8], overlPass[8], BCidentifier[8];
    nthresPass = 0;
    for (i = 0; i < m_Nbunch; i++) {
        if (m_highestth[i]) {
            thresPass[nthresPass] = m_highestth[i];
            overlPass[nthresPass] = m_overlap[i];
            BCidentifier[nthresPass] = i;
            nthresPass += 1;
        }
    }
    makeTestPattern(1, ntimes + 2 * nthresPass);
    //
    // open output file
    //
    ofstream out_k_trigger;
    out_k_trigger.open("k-trigger.output", ios::app);
    //
    // code to print out m_k_pattern
    //
    for (i = 0; i < ntimes; i++) {
        out_k_trigger << " TIME " << times[i] + timeOffsetKPa << " K ";
        out_k_trigger << nchannels[i][0][0] << " ";
        for (l = 0; l < nchannels[i][0][0]; l++) { out_k_trigger << channels[i][0][0][l] << " "; }  // end-of-for(l
        out_k_trigger << endl;
    }  // end-of-for(i
    //
    if (nthresPass) {
        for (i = 0; i < nthresPass; i++) {
            out_k_trigger << " TIME " << BCidentifier[i] * 25. + timeOffsetThr;
            out_k_trigger << " THR " << thresPass[i] << endl;
            if (overlPass[i]) {
                out_k_trigger << " TIME " << BCidentifier[i] * 25. + timeOffsetThr;
                out_k_trigger << " OVL " << overlPass[i] << std::endl;
            }
        }
    }
    //
    out_k_trigger.close();
    //
}  // end-of-makeOutPattern
//------------------------------------------------------------------------//
ubit16 Matrix::config(ubit16 i, ubit16 *arr) const {
    //
    // input:  i     threshold address
    //         *arr  pointer to bidimen. array
    //
    // returns the number of possible configurations compatible with the
    // required majority pointed by majorities[i]
    //
    // Data *arr(k) and *arr(k+1) give the majority addresses for the side-x
    // and side-y respectevely.
    //
    ubit16 nconf = 0;
    //
    // cout<<"lowhig="<<m_lowhigh<< "majorities= "<<m_majorities[0]
    //     <<" "<< m_majorities[1]<<" "<<m_majorities[2]<<endl;
    //
    switch (m_lowhigh) {
        case 0:                         // low-pt trigger matrix
            switch (m_majorities[i]) {  // select the required majority for each thresh.;
                case 0:                 // 1-out-of-4 majority: no implementation yet
                                        // gives 1/2 in I OR 1/2 in J
                    nconf = 0;
                    break;
                case 1:  // 2-out-of-4 majority
                    nconf = 1;
                    *(arr + 0) = 0;
                    *(arr + 1) = 0;
                    break;
                case 2:  // 3-out-of-4 majority
                    nconf = 2;
                    *(arr + 0) = 0;
                    *(arr + 1) = 1;
                    *(arr + 2) = 1;
                    *(arr + 3) = 0;
                    break;
                case 3:  // 4-out-of-4 majority
                    nconf = 1;
                    *(arr + 0) = 1;
                    *(arr + 1) = 1;
                    break;
                default: throw std::runtime_error("Matrix::config: the majority " + std::to_string(m_majorities[i]) + " is unforeseen");
            }
            break;
        case 1:  // high-pt trigger matrix
                 // high-pt trigger
            switch (m_majorities[i]) {
                case 1:  // 1-out-of-2 majority
                    nconf = 1;
                    *(arr + 0) = 0;
                    *(arr + 1) = 0;
                    break;
                case 2:  // 2-out-of-2 majority
                    nconf = 1;
                    *(arr + 0) = 0;
                    *(arr + 1) = 1;
                    break;
                default: throw std::runtime_error("Matrix::config: the majority " + std::to_string(m_majorities[i]) + " is unforeseen");
            }
            break;
        default: throw std::runtime_error("Matrix::config: lowhighpt " + std::to_string(m_lowhigh) + " is unforeseen");
    }
    return nconf;
}  // end-of-method-config
//------------------------------------------------------------------------//
int Matrix::getSubsystem() const { return m_subsystem; }
//------------------------------------------------------------------------//
int Matrix::getProjection() const { return m_projection; }
//------------------------------------------------------------------------//
int Matrix::getSector() const { return m_sector; }
//------------------------------------------------------------------------//
int Matrix::getPad() const { return m_pad; }
//------------------------------------------------------------------------//
int Matrix::getLowHigh() const { return m_lowhigh; }
//------------------------------------------------------------------------//
int Matrix::getAddress0() const { return m_address[0]; }
//------------------------------------------------------------------------//
int Matrix::getAddress1() const { return m_address[1]; }
//------------------------------------------------------------------------//
int Matrix::getLocalAdd() const { return m_localadd; }
//------------------------------------------------------------------------//
ubit16 Matrix::getOutputThres(ubit16 bunch) const { return m_highestth[bunch]; }  // end-of-ubit16-getOutput
//------------------------------------------------------------------------//
ubit16 Matrix::getOutputOverl(ubit16 bunch) const { return m_overlap[bunch]; }  // end-of-ubit16-getOutput
//------------------------------------------------------------------------//
sbit16 Matrix::getBunchPhase() const { return m_BunchPhase; }  // end-of-ubit16-getBunchPhase
//------------------------------------------------------------------------//
sbit16 Matrix::getBunchOffset() const { return m_BunchOffset; }  // end-of-ubit16-getBunchOffset
//------------------------------------------------------------------------//
void Matrix::set_to_1(CMAword *p, sbit16 channel) const {
    CMAword j{1};
    std::array<ubit16, 2> i = inds(channel);
    if (!(channel < 0)) {
        *(p + i[0]) = *(p + i[0]) | j << i[1];
    } else {
        throw std::out_of_range("Matrix::set_to_1: channel is negative; channel=" + std::to_string(channel));
    }  // end-of-if(!(channel<0
}  // end-of-Matrix::set_to_1
//----------------------------------------------------------------------//
void Matrix::set_to_0(CMAword *p, sbit16 channel) const {
    CMAword j{1};
    std::array<ubit16, 2> i = inds(channel);
    if (!(channel < 0)) {
        *(p + i[0]) = *(p + i[0]) & ~(j << i[1]);
    } else {
        throw std::out_of_range("Matrix::set_to_1: channel is negative; channel=" + std::to_string(channel));
    }  // end-of-if(!(channel<0
}  // end-of-Matrix::set_to_1
 
//----------------------------------------------------------------------//
void Matrix::wind() const {
    sbit16 i, j;
    cout << "-----------------------" << endl
         << "|     Matrix::wind    |" << endl
         << "-----------------------" << endl
         << " Matrix Roads " << endl;
    for (i = 0; i < s_nthres; i++) {
        for (j = 0; j < s_nchan[0]; j++) {
            cout << " thres. " << i << " channel "
                 << j
                 //       <<" Road0 "<<hex<<(*(m_roads+32*2*i+2*j+0))<<dec
                 //       <<" Road1 "<<hex<<(*(m_roads+32*2*i+2*j+1))<<dec<<endl;
                 << " Road0 " << hex << (m_trigRoad[i][j][0]) << dec << " Road1 " << hex << (m_trigRoad[i][j][1]) << dec << endl;
        }
    }
    cout << " majorities: " << endl;
    for (i = 0; i < 3; i++) { cout << m_majorities[i] << " " << endl; }
    cout << endl
         << " number of overlapping ' low' channels: " << m_matOverlap[0] << endl
         << " number of overlapping 'high' channels: " << m_matOverlap[1] << endl;
    for (i = 0; i < s_nchan[0]; i++) { cout << " channel " << i << " in coincidence with " << m_diagonal[i] << endl; }  // end-of-for(i
}  // end-of-method-wind
//----------------------------------------------------------------------//
void Matrix::display() const {
    ubit16 i;
    ubit16 df = 19;
    rpcdata *rpcpnt;
    //
    // if(this) {
    cout << "=======================" << endl << "||   Matrix Display  ||" << endl << "=======================" << endl << endl;
    show_attributes();
 
    cout << endl << " All raw data " << endl;
    for (i = 0; i < 2; i++) {
        cout << " Matrix Side is " << i << endl;
        rpcpnt = m_datarpc[i];
        while (rpcpnt) {
            cout << "  Layer= " << rpcpnt->layer << " stripadd= " << rpcpnt->stripadd << " time= " << rpcpnt->time
                 << " mask= " << rpcpnt->masked << " BC= " << rpcpnt->BC << " DLL= " << rpcpnt->DLL << " delay= " << rpcpnt->delay << endl;
            rpcpnt = rpcpnt->next;
        }  // end-of-while(rpcpnt)
    }      // end-of-for(i
    //
    if (m_matrixDebug & 1 << (df + 0)) {
        cout << " Display Matrix Input " << endl;
        disp_CMAreg(0);  // display the input registers
    }
    //
    //
    if (m_matrixDebug & 1 << (df + 1)) {
        cout << " Display Matrix Preprocessing " << endl;
        disp_CMAreg(1);  // display the prepro registers
    }
    //
    if (m_matrixDebug & 1 << (df + 2)) {
        cout << " Display Matrix Majority " << endl;
        disp_CMAreg(2);  // display the majority registers
    }
    //
    if (m_matrixDebug & 1 << (df + 3)) {
        cout << " Display Trigger  " << endl;
        disp_CMAreg(3);  // display the trigger registers
    }
    //
    //} else {
    // cout<<"======================="<<endl
    //     <<"||    Matrix EMPTY   ||"<<endl
    //     <<"======================="<<endl;
    //}//end-of-Matrix::display
}  // end-of-method display
//------------------------------------------------------------------------//
void Matrix::show_attributes() const {
    cout << " Matrix Attributes: " << endl
         << " Subsystem " << m_subsystem << "; Projection " << m_projection << "; Sector " << m_sector << "; Pad " << m_pad << "; LowHig "
         << m_lowhigh << "; addresses: " << m_address[0] << "  " << m_address[1] << endl;
}  // end-of-Matrix::attributes
//------------------------------------------------------------------------//
void Matrix::disp_CMAreg(ubit16 id) const {
    //
    // display the CMA registers
    //
    ubit16 i, j, k;
    if (id < 2) {
        for (i = 0; i < 2; i++) {  // loop on the two Matrix sides
            cout << " CMA Side (0=side-x; 1=side-y) " << i << endl;
            for (j = 0; j < 2; j++) {  // loop on the two Matrix layers
                switch (id) {
                    case 0:
                        cout << " Layer " << j << endl;
                        dispRegister(&m_input[i][j][0][0], i);
                        break;
                    case 1:
                        cout << " Layer " << j << endl;
                        dispRegister(&m_prepr[0][i][j][0][0], i);
                        break;
                }  // end-of-switch
            }      // end-of-for(j
        }          // end-of-for(i
 
    } else {
        switch (id) {
            case 2:
                for (i = 0; i < s_nthres; i++) {  // loop on threshold
                    cout << " Threshold address " << i << endl;
                    for (j = 0; j < 2; j++) {  // loop on matrix sides
                        cout << " CMA Side (0=side-x; 1=side-y) " << j << endl;
                        for (k = 0; k < 2; k++) {  // loop on majority types
                            cout << " Majority type (0=1/2; 1=2/2) " << k << endl;
                            dispRegister(&m_mjori[i][j][k][0][0], j);
                        }  // end-of-for(k
                    }      // end-of-for(j
                }          // end-of-for(i
                break;
            case 3:
                for (i = 0; i < s_nthres; i++) {  // loop on the three thresholds
                    cout << " Trigger Threshold address " << i << endl;
                    dispTrigger(&m_trigg[i][0]);
                }  // end-of-for(i
                cout << " ReadOut Buffer " << endl;
                dispRegister(&rodat[0][0][0][0], 0);
                break;
            default: cout << " Matrix::disp_CMAreg id value " << id << " not foreseen " << endl;
        }  // end-of-switch (id)
 
    }  // end-of-if(id
    cout << " " << endl;
}  // end-of-Matrix::disp_CMAreg
//------------------------------------------------------------------------//
void Matrix::dispRegister(const CMAword *p, ubit16 side) const {
    ubit16 n, j, k;
    //
    // allocation for oststream strdisp
    //
    std::ostringstream strdisp;
    n = (s_nchan[side] - 1) / s_wordlen + 1;
    strdisp << "  ";
 
    for (j = 0; j < s_nchan[side]; j += 2) { strdisp << " " << j % 10; }  // end-of-for
    strdisp << " " << endl;
    for (j = 0; j < m_nclock; j++) {  // loop on the m_nclock cycles
        strdisp << " " << j % 10 << " ";
        for (k = 0; k < n; k++) {  // loop on the buffer words
            dispBinary(p + k + 2 * j, strdisp);
        }  // end-of-for(k
        strdisp << " " << endl;
 
    }  // end-of-for(j
 
    cout << strdisp.str() << endl;
}  // end-of-Matrix::dispRegister
//------------------------------------------------------------------------//
void Matrix::dispTrigger(const CMAword *p) const {
    ubit16 j;
    //
    // allocation for oststream strdisp
    //
 
    std::ostringstream strdisp;
    strdisp << "  ";
 
    for (j = 0; j < s_nchan[0]; j += 2) { strdisp << " " << j % 10; }  // end-of-for
    strdisp << " " << endl;
    for (j = 0; j < m_nclock; j++) {  // loop on the m_nclock cycles
        strdisp << " " << j % 10 << " ";
        dispBinary(p + j, strdisp);
        strdisp << " " << endl;
    }  // end-of-for(j
 
    cout << strdisp.str() << endl;
}  // end-of-Matrix::dispTrigger
//------------------------------------------------------------------------//
void Matrix::dispBinary(const CMAword *p, std::ostream &strdisp) const {
    ubit16 i;
    CMAword j;
    j = 1;
    for (i = 0; i < s_wordlen; i++) {
        if ((*p) & j) {
            strdisp << "|";
        } else {
            strdisp << ".";
        }  // end-of-if
        j = j << 1;
    }  // end-of-for(
}  // end-of-Matrix::dispBinary
//------------------------------------------------------------------------//
void Matrix::dispWind() const {
    for (ubit16 i = 0; i < s_nthres; i++) { dispWind(i); }
}  // end-of-dispWind
//------------------------------------------------------------------------//
void Matrix::dispWind(ubit16 thres) const {
    std::ostringstream strdisp;
 
    strdisp << endl
            << " =========================" << endl
            << " =                       =" << endl
            << " =  Matrix::dispWind     =" << endl
            << " =  Threshold address " << thres << "  =" << endl
            << " =                       =" << endl
            << " =========================" << endl
            << endl;
    //
    for (sbit16 j = s_nchan[1] - 1; j >= 0; j--) {
        ubit16 ad1 = j / 32;
        ubit16 ad2 = j % 32;
        if (j < 10) {
            strdisp << " " << j << " ";
        } else {
            strdisp << j << " ";
        }
        for (ubit16 k = 0; k < s_nchan[0]; k++) {
            if ((m_trigRoad[thres][k][ad1]) & (1 << ad2)) {
                strdisp << "*";
            } else {
                strdisp << ".";
            }  // end-of-if
        }      // end-of-for(k
        strdisp << " " << endl;
    }  // end-of-for(j
    strdisp << " " << endl;
    strdisp << "   00000000001111111111222222222233" << endl << "   01234567890123456789012345678901" << endl;
    //
    cout << strdisp.str() << endl;
}  // end-of-dispWind
 
//----------------------------------------------------------------------------//
ubit16 Matrix::char2int(const char *str, CMAword the32[2]) {
    ubit16 outflag = 0;
    the32[0] = 0;
    the32[1] = 0;
    ubit16 stringLength = strlen(str);
    if (stringLength > 16) {
        outflag = 1;
    } else {
        for (ubit16 i = 0; i < stringLength; i++) {
            //        cout<<" Reading character "<<*(str+stringLength-1-i)<<endl;
            if (*(str + stringLength - 1 - i) == '0')
                the32[i / 8] = the32[i / 8] + 0;
            else if (*(str + stringLength - 1 - i) == '1')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 1;
            else if (*(str + stringLength - 1 - i) == '2')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 2;
            else if (*(str + stringLength - 1 - i) == '3')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 3;
            else if (*(str + stringLength - 1 - i) == '4')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 4;
            else if (*(str + stringLength - 1 - i) == '5')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 5;
            else if (*(str + stringLength - 1 - i) == '6')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 6;
            else if (*(str + stringLength - 1 - i) == '7')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 7;
            else if (*(str + stringLength - 1 - i) == '8')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 8;
            else if (*(str + stringLength - 1 - i) == '9')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 9;
            else if (*(str + stringLength - 1 - i) == 'a' || *(str + stringLength - 1 - i) == 'A')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 10;
            else if (*(str + stringLength - 1 - i) == 'b' || *(str + stringLength - 1 - i) == 'B')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 11;
            else if (*(str + stringLength - 1 - i) == 'c' || *(str + stringLength - 1 - i) == 'C')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 12;
            else if (*(str + stringLength - 1 - i) == 'd' || *(str + stringLength - 1 - i) == 'D')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 13;
            else if (*(str + stringLength - 1 - i) == 'e' || *(str + stringLength - 1 - i) == 'E')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 14;
            else if (*(str + stringLength - 1 - i) == 'f' || *(str + stringLength - 1 - i) == 'F')
                the32[i / 8] = the32[i / 8] + (intPow(16, i % 8)) * 15;
            else
                outflag = 2;
        }  // end-of-for
    }
 
    // if(outflag) {
    //  the32[0]=0;
    //  the32[1]=1;
    // } else {
    //  CMAword temp = the32[1];
    //  the32[1]=the32[0];
    //  the32[0]=temp;
    // }
    return outflag;
}  // end-of-char2int
//----------------------------------------------------------------------------//
CMAword Matrix::intPow(const ubit16 base, const ubit16 expo) const {
    CMAword output = 1;
    if (expo) {
        for (ubit16 i = 1; i <= expo; i++) { output = output * base; }  // end-of-for
    }
    return output;
}  // end-of-CMAword Matrix::intPow