[OpenSPARC-T2-DV] / verif / env / common / vera / classes / sparcBenchUtils.vr

// ========== Copyright Header Begin ==========================================
// 
// OpenSPARC T2 Processor File: sparcBenchUtils.vr
// Copyright (C) 1995-2007 Sun Microsystems, Inc. All Rights Reserved
// 4150 Network Circle, Santa Clara, California 95054, U.S.A.
//
// * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. 
//
// This program is free software; you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation; version 2 of the License.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
// GNU General Public License for more details.
//
// You should have received a copy of the GNU General Public License
// along with this program; if not, write to the Free Software
// Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
// 
// For the avoidance of doubt, and except that if any non-GPL license 
// choice is available it will apply instead, Sun elects to use only 
// the General Public License version 2 (GPLv2) at this time for any 
// software where a choice of GPL license versions is made 
// available with the language indicating that GPLv2 or any later version 
// may be used, or where a choice of which version of the GPL is applied is 
// otherwise unspecified. 
//
// Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara, 
// CA 95054 USA or visit www.sun.com if you need additional information or 
// have any questions. 
// 
// ========== Copyright Header End ============================================
#include <vera_defines.vrh>
#include <std_display_defines.vri>
#include <std_display_class.vrh>
#include <plusArgMacros.vri>
#include <defines.vri>

#include <cmp.vri>
#include <globals.vri>

#ifndef FC_BENCH
#ifndef FC_SCAN_BENCH
#include <verilog_tasks_ncu.vri>
#endif
#endif

// mem slam, reg slam
#include <verilog_tasks_misc.vri>
   
#include <std_display_class.vrh>
#include <baseParamsClass.vrh>
#include <sparcParams.vrh>
#include <baseUtilsClass.vrh>
#include <memArray.vrh>

#ifndef CCM_BENCH
#ifndef FC_BENCH
#ifndef FC_SCAN_BENCH
#include <ccx_tag_class.vrh>
#endif
#endif
#endif

#include <sparcBenchUtils_if.vrh>

#define CLASSNAME SparcBenchUtils
#define CLASSNAMEQ "SparcBenchUtils"

class CLASSNAME extends BaseUtils {

  local string className = "SparcBenchUtils";
  local StandardDisplay dbg;
  local reg noNCUcheck = 0;
  local reg failNCUcheck = 0;
  
  // rands
  local randc reg [5:0] startThread;
  local randc reg [2:0] randCID;
  local rand  integer parkWait;
  // end rands

  // constraints
  constraint parkWait_c {
    parkWait <= gParam.por_delay_max;
    parkWait >= gParam.por_delay_min;
  }
  // end constraints

  
  task new(StandardDisplay dbgHndl, integer clockPeriod = 100);

  task resetDut();
  
  function reg ioSpaceAccess(reg [63:0] addr, // full addr of request
                            var reg [127:0] data, // r/w data
                             reg read = 1,    // request is read, else write
                             reg [7:0] size = 8'hff, // write byte mask
                             integer thread = 99, // please provide
                             integer myPort = 99); // optional for messaging

  function reg [127:0] copyDataByte (reg [127:0] data, 
                                     reg [7:0]   size,
                                     reg [3:0] offset);
  
  function reg [63:0] evictAddr (reg [7:0] core_enable,
                                 var reg [2:0] cidUsed,
                                 reg [3:0] cid = 4'hf,
                                 integer dCacheWeight = 60);

  function reg [127:0] evictVector (reg [7:0] core_enable,
                                    reg [63:0] evict_pa,
                                    reg [2:0] cpuId,
                                    var reg [7:0] targets);

  function reg [63:0] getThreadEnables();
  function reg [63:0] getRunStatus();
  function reg [2:0] whichBank(reg [63:0] addr);
}


//----------------------------------------------------------
// Do not call randomize() in this classes new() task.
// The final class extention of this class must (does) do that in its new()!
// Go ahead an assume that the first set of random numbers are available.
task CLASSNAME::new(StandardDisplay dbgHndl, integer clockPeriod = 100) {
  super.new(dbgHndl, clockPeriod);
  this.dbg = dbgHndl;
  if (mChkPlusarg(noNCUcheck)) noNCUcheck = 1;
  if (mChkPlusarg(failNCUcheck)) failNCUcheck = 1;
}


// call AFTER time zero
function reg [63:0] CLASSNAME::getThreadEnables() {
  //getThreadEnables = sparcBenchUtils_if.th_check_enable; // -vcs_run_args=+thread=1
  getThreadEnables = gParam.finishMask; // this is better
}

// call AFTER time zero. gets core_running_status.
function reg [63:0] CLASSNAME::getRunStatus() {
  getRunStatus = sparcBenchUtils_if.core_running_status;
}


// return which bank this address goes to.
// for partial bank mode
function reg [2:0] CLASSNAME::whichBank(reg [63:0] addr) {

  case (gParam.bank_set_mask) {
    1: whichBank = addr[6];
    2: whichBank = {2'b01,addr[6]};
    3: whichBank = {1'b0,addr[7:6]};
    4: whichBank = {2'b10,addr[6]};
    5: whichBank = {addr[7],1'b0,addr[6]};
    6: whichBank = {addr[7],~addr[7],addr[6]};
    8: whichBank = {2'b11,addr[6]};
    9: whichBank = {addr[7],addr[7],addr[6]};
    10: whichBank = {addr[7],1'b1,addr[6]};
    12: whichBank = {1'b1,addr[7],addr[6]};
    15: whichBank = {addr[8],addr[7],addr[6]};
    default: error("bank_set_mask is illegal\n");
  }

}


//----------------------------------------------------------
// start threads
//
// rules:
// 1 - always start with the lowest core, after that randomize core selection.
// 2 - in a core, always start the lowest thread first.
// 3 - after each core has had its lowest thread started, randomize the
//     rest of the threads.
//

task CLASSNAME::resetDut () {

  
  //reg [31:0] resetThis;
  reg [63:0] threadEnables;
  integer core, tid, tmp;

  integer lowCore; // lowest core from threadEnables
  integer lowThread [8]; // lowest thread in each core
  reg [63:0] doneThreads; // have been started already
  reg [7:0] doneCores; // have been started already
  
  
#ifndef FC_BENCH
#ifndef FC_SCAN_BENCH
  
  for (tid=0; tid<8; tid++) lowThread[tid] = 32'hx;
  
  void = this.rand_mode(OFF, "startThread");

  // do not unpark too early.
  if (sparcBenchUtils_if.core_cycle_cnt < 12)
    repeat (12 - sparcBenchUtils_if.core_cycle_cnt)
      @(negedge sparcBenchUtils_if.clk);

  // range is 28-48 (38 ideal) cycles from time zero if +forcePORstate
  if (gParam.forcePORstate) {
    if (sparcBenchUtils_if.core_cycle_cnt < 38)
    repeat (38 - sparcBenchUtils_if.core_cycle_cnt)
      @(negedge sparcBenchUtils_if.clk);
  }
  
  threadEnables = sparcBenchUtils_if.th_check_enable; // -vcs_run_args=+thread=1
  if (! threadEnables) {
    PR_ALWAYS(className, MON_ALWAYS,
            psprintf("Unparking threads: th_check_enable = 0, Will not unpark ANY threads!!!"));
    return;
  }
  PR_ALWAYS(className, MON_ALWAYS,
            psprintf("Unparking threads: th_check_enable = %h",threadEnables));

  // find lowest core
  for (core=0; core<8; core++) {
    if (|threadEnables[7+(core*8):core*8]) {
      lowCore = core;
      break;
    } else doneCores[core] = 1;
  }
  
  // find lowest thread in each core
  for (core=lowCore; core<8; core++) {
    for (tid=0; tid<8; tid++) {
      if (threadEnables[tid+(8*core)] == 1) {
        lowThread[core] = tid;
        break;
      }
    }
  }

  doneThreads = ~threadEnables; // call unused tids done for starters
    
  PR_ALWAYS(className, MON_ALWAYS,
            psprintf("Unparking: low core = %0d, low threads (0-07): %1d,%1d,%1d,%1d,%1d,%1d,%1d,%1d",lowCore,lowThread[7],lowThread[6],lowThread[5],lowThread[4],lowThread[3],lowThread[2],lowThread[1],lowThread[0]));
  PR_ALWAYS(className, MON_ALWAYS,
            psprintf("Unparking: low core = %0d, low threads (0-63): %2d,%2d,%2d,%2d,%2d,%2d,%2d,%2d",lowCore,lowThread[7]+56,lowThread[6]+48,lowThread[5]+40,lowThread[4]+32,lowThread[3]+24,lowThread[2]+16,lowThread[1]+8,lowThread[0]));
    
  // POR threads in random orders after lowest core lowest tid
  // but lowest first for each core.
  // this is lowest first...
  tmp = (lowCore*8)+lowThread[lowCore];
  verilog_write_cmp_reg({ASI_CMP_CORE,ASI_CMP_CORE_RUNNING_W1S},
                        1 << tmp, tmp);
  
  doneThreads[tmp] = 1; // mark it done
  doneCores[lowCore] = &doneThreads[7+(lowCore*8):lowCore*8]; // core done?
  
  PR_ALWAYS(className, MON_ALWAYS,
            psprintf("T%0d unparked (lowest tid in lowest core)\n",tmp));

  // quick check
  if (doneThreads !== 64'hffffffffffffffff) {

    // now the rest

    // start the lowest thread on each core (except lowest core).
    // they could be chosen at random or sequential.
    repeat (8) {
      core = randCID;
      void = this.randomize();
      if (core == lowCore || doneCores[core]) continue;
      if (lowThread[core] !== 32'hx) {
        tmp = (core*8)+lowThread[core];
        // parkWait = urandom_range(gParam.por_delay_max,gParam.por_delay_min);
        //       PR_DEBUG (className, MON_DEBUG,
        //                 psprintf("T%0d post unpark delay %d\n",get_cycle(),resetThis,del));
        repeat (parkWait) @(posedge sparcBenchUtils_if.clk);
        verilog_write_cmp_reg({ASI_CMP_CORE,ASI_CMP_CORE_RUNNING_W1S},
                              1 << tmp, tmp);
        PR_ALWAYS(className, MON_ALWAYS,
                  psprintf("T%0d unparked (lowest tid in core)\n",tmp));
        doneThreads[tmp] = 1; // mark it done
        doneCores[core] = &doneThreads[7+(core*8):core*8]; // core done?
      }
    }
    
    // start the remaining threads at random
    repeat (64) {
      if (doneThreads[startThread] == 0) {
        void = this.randomize();
        // parkWait = urandom_range(gParam.por_delay_max,gParam.por_delay_min);
        //       PR_DEBUG (className, MON_DEBUG,
        //                 psprintf("T%0d post unpark delay %d\n",get_cycle(),resetThis,del));
        repeat (parkWait) @(posedge sparcBenchUtils_if.clk);
        
        verilog_write_cmp_reg({ASI_CMP_CORE,ASI_CMP_CORE_RUNNING_W1S},
                              1 << startThread,
                              startThread);
        PR_ALWAYS(className, MON_ALWAYS,
                  psprintf("T%0d unparked (remaining non-lowest tids)\n",startThread));
        doneThreads[startThread] = 1; // mark it done

      }
      void = this.rand_mode(ON, "startThread");
      void = this.randomize();
      void = this.rand_mode(OFF, "startThread");
    }

  }
  
#endif
#endif
}

/*
task CLASSNAME::resetDut () {

  
  reg [31:0] tid, resetThis;
  reg [63:0] threadEnables;
  integer lowThread, del;

#ifndef FC_BENCH

  void = this.rand_mode(OFF, "startThread");
  
  repeat (5) @(negedge sparcBenchUtils_if.clk);

  threadEnables = sparcBenchUtils_if.th_check_enable; // -vcs_run_args=+thread=1
  
  PR_ALWAYS(className, MON_ALWAYS,
            psprintf("Unparking threads: th_check_enable = %h",threadEnables));

  // find lowest thread
  for (tid=0; tid<=63; tid++) {
    if (threadEnables[tid] == 1) {
      lowThread = tid;
      break;
    }
  }
  
  //POR threads in random orders after lowest
  verilog_write_cmp_reg({ASI_CMP_CORE,ASI_CMP_CORE_RUNNING_W1S},
                        1<<lowThread,
                        lowThread);
  PR_ALWAYS(className, MON_ALWAYS,
            psprintf("T%0d unparked (first)\n",lowThread));

  repeat (64) {
    if (threadEnables[startThread] == 1 && startThread !== lowThread) {
      del = urandom_range(gParam.por_delay_max,gParam.por_delay_min);
      //       PR_DEBUG (className, MON_DEBUG,
      //                 psprintf("T%0d post unpark delay %d\n",get_cycle(),resetThis,del));
      repeat (del) @(posedge sparcBenchUtils_if.clk);
      
      verilog_write_cmp_reg({ASI_CMP_CORE,ASI_CMP_CORE_RUNNING_W1S},
                            1<<startThread,
                            startThread);
      PR_ALWAYS(className, MON_ALWAYS,
                psprintf("T%0d unparked\n",startThread));

    }
    void = this.rand_mode(ON, "startThread");
    void = this.randomize();
    void = this.rand_mode(OFF, "startThread");
  }

#endif

}
*/


//----------------------------------------------------------
// Is address allowed at NCU for a read or write?  Returns true if the
// r/w was OK.  If the return is false, the response packet MUST set the
// error indication just as a real NCU would. The caller has to do
// this!!!  Also returns the data for the address to the caller.  Zeros
// will be returned for addresses that have never been written (except
// 0x82 RNG).
//
// Certain ASI registers (CMP) may be implemented in verilog.
// We will detect that here and do the right thing.
//
// Any write to any address that isn't explicitly defined will be
// silently dropped.  That includes reserved and not supported regions.
// It also includes unimplemented addresses within a region.  (For
// example, the 0x90 region is for IO mapped ASI registers, but there
// aren't very many.  Any write to an address that isn't mapped to a
// register will get dropped by NCU RTL, *BUT* this function allows the

//
// Real world store example:
// * The LSU sends the NCU a store.
// * The NCU acks the store - always.
// * Since the store is ack'ed, software sees it as done.
// * The NCU trys to figure out where the write should go.
//   If it can't figure it out, it drops it.
// * The SOC units themselves could also drop it if the NCU
//   forwarded them a request that they don't support.
// 
// This function will use the following table to define it's behavior.
// Any range identified with ERROR *must* cause a load error response packet
// to be returned to the core. THE USER/CALLER MUST DO THIS!!! This will
// cause a trap to be taken by the core.  In order for nas to take the
// same trap, we need to add this case to the current interrupt sync
// mechanism. Reads from 0x82 will return random data. Currently, only
// 0xFF will be considered 100% read only.
//
// OLD Address [39:32]                                                behavior
// OLD ---------------                                                --------
// OLD 0x80 NCU                                                       R/W allowed
// OLD 0x81 NIU                                                       R/W allowed
// OLD * 0x82 RNG (Random Number Generator)                             IGNORE writes
// OLD * 0x83 CCU                                                       R/W allowed
// OLD 0x84 MCUs                                                      R/W allowed
// OLD 0x85 TCU (JTAG/TAP. NCU/Core will not initiate)                ERROR on read
// OLD * 0x86 TAP to ASI (not supported)                                ERROR on read
// OLD * 0x87 TAP to L2 CSR (not supported)                             ERROR on read
// OLD 0x88 DMUCSR                                                    R/W allowed
// OLD 0x89 RST                                                       R/W allowed
// OLD 0x8A-0x8F Reserved                                             ERROR on read
// OLD 0x90 ASI CPU shared registers                                  R/W allowed
// OLD 0x91-0x9F Reserved                                             ERROR on read
// OLD 0xA0-0xBF L2 CSR (never comes to NCU)                          ERROR on read
// OLD 0xC0-0xCF PCIE (64GB) / DMUPIO                                 R/W allowed
// OLD 0xD0-0xFE Reserved                                             ERROR on read
// OLD 0xFF SSI (boot ROM)                                            IGNORE writes
//
// Address [39:32]                                                behavior
// ---------------                                                --------
// 0x80 NCU                                                       R/W allowed
// 0x81 NIU                                                       R/W allowed
// 0x82 Reserved                                                  ERROR on read
// 0x83 CCU (+0x30 is Random Number Generator)                    R/W allowed
// 0x84 MCUs                                                      R/W allowed
// 0x85 TCU (JTAG/TAP)                                            R/W allowed
// 0x86 Debug                                                     R/W allowed
// 0x87 Reserved                                                  ERROR on read
// 0x88 DMU                                                       R/W allowed
// 0x89 RST                                                       R/W allowed
// 0x8A-0x8F Reserved                                             ERROR on read
// 0x90 ASI CPU shared registers                                  R/W allowed
// 0x91-0x9F Reserved                                             ERROR on read
// 0xA0-0xBF L2 CSR (never goes to NCU)                           ERROR on read
// 0xC0-0xCF PCIE (64GB) / DMUPIO                                 R/W allowed
// 0xD0-0xFE Reserved                                             ERROR on read
// 0xFF SSI (boot ROM)                                            IGNORE writes
// 
// To disable the CPX pkt error response and allow full R/W of ALL
// addresses by the NCU model (except writes to 0xFF) +noNCUcheck can be
// used.  Until Riesling/nas is updated with interrupt sync, and a
// specific knowledge of the I/O map, +noNCUcheck could possibly be used.
// The plus arg +noNCUcheck is normally not recommended because it
// presents a behavior that IS NOT at all like a real N2 (so forget I
// even mentioned it)!
// 
// To always fail the simulation on error ("ERROR on read" space was
// accessed or 0xFF was written), use +failNCUcheck. This would be a good
// way to find "bad" diags.
// 
//
function reg CLASSNAME::ioSpaceAccess(reg [63:0] addr, // full addr of request
                                      var reg [127:0] data, // r/w data
                                      reg read = 1,    // request is read, else write
                                      reg [7:0] size = 8'hff, // write mask, read size
                                      integer thread = 99,
                                      integer myPort = 99) // optional for messaging
{

  reg [63:0] tmp64;

#ifndef NCURTL
#ifndef FC_BENCH
#ifndef FC_SCAN_BENCH
  
  PR_INFO ("ioaccess", MON_INFO,
             psprintf("addr = %h data = %h R/W = %b size = %h thread = %0d myPort = %0d",
             addr,data,read,size,thread,myPort));

  ioSpaceAccess = 1; // no error
  addr = addr & 64'hfffffffffffffff8;
  if (read) data = 128'hdeaddeaddeaddeaddeaddeaddeaddead;
  
  // detect special registers implemented in verilog.
  if (
    (addr[IO_ASI_ADDR_NCU] == 8'h90 &&
     (addr[IO_ASI_ADDR_REG] == ASI_CMP_CORE ||
      addr[IO_ASI_ADDR_ADR] == ASI_WMR_VEC_MASK_ADR ||
      addr[IO_ASI_ADDR_ADR] == ASI_L2_IDX_HASH_EN_ADR ||
      addr[IO_ASI_ADDR_ADR] == ASI_CMP_TICK_ENABLE_ADR)) ||
     (addr == ASI_RESET_STAT)
     ) {

    if (thread == 99) thread = addr[IO_ASI_ADDR_CT];
    if (read == 0)
      verilog_write_cmp_reg(addr,data[63:0],thread);
    else
    {
      verilog_read_cmp_reg(addr,tmp64,thread);
      data = {tmp64,tmp64};
    }
    
    PR_INFO ("ioaccess", MON_INFO,
               psprintf("err = %b  Special IO Register Access (i.e. CMP reg, etc)",
               ioSpaceAccess));
    return;
  }


  case (addr[IO_ASI_ADDR_NCU]) {
    8'h80, 8'h81, 8'h83, 8'h84, 8'h85, 8'h86, 8'h88, 8'h89: {
      // 8'h83 +0x30 is Random Number Generator
      if (addr[39:0] == 40'h83_0000_0030) {
        if (read) {
          data[63:0] = {urandom(),urandom()};
          // review getting socket errors
          #ifndef GATESIM
            verilog_mem_slam(addr[39:0],
                             data[63:0],
                             8'hff,
                             "NCU BFM");
          #endif
        }
      } else {
      if (read) data = gMem.read128(addr,myPort, 1);
      else gMem.writeBM(addr, data[63:0], size, myPort);
      }
    }
    8'h82, 8'h87, 8'h8a, 8'h8b, 8'h8c, 8'h8d, 8'h8e, 8'h8f: { // reserved
      ioSpaceAccess = 0;
    }
  } // case
  
  case (addr[39:36]) {
    4'h9: {// 0x90 ASI CPU shared registers
      if (addr[IO_ASI_ADDR_NCU] == 8'h90) { // 0x90 ASI CPU shared registers
        // Certain ASI registers (CMP) may be implemented in verilog.
        // The code that calls this function MUST filter those out
        // and do the right thing. We will not attempt that here!!!
        if (read) data = gMem.read128(addr,myPort, 1);
        else gMem.writeBM(addr, data[63:0], size, myPort);            
      } else {// 0x91-0x9F Reserved ERROR
        ioSpaceAccess = 0;
      }  
    }
    4'hA: {// 0xA0-0xBF L2 CSR (handled by CCX directly and does not come to NCU) ERROR
      PR_ERROR (CLASSNAMEQ, MON_ERR, psprintf ("T%d Bench ERROR FAIL: - L2 CSR address seen at NCU, this is bad! (addr=%0h (%0h),data=%0h,mask=%0h,read=%0h",thread,addr,addr[39:32],data,size,read));
    }
    4'hB: {// 0xA0-0xBF L2 CSR (handled by CCX directly and does not come to NCU) ERROR
      PR_ERROR (CLASSNAMEQ, MON_ERR, psprintf ("T%d Bench ERROR FAIL: - L2 CSR address seen at NCU, this is bad! (addr=%0h (%0h),data=%0h,mask=%0h,read=%0h",thread,addr,addr[39:32],data,size,read));

    }
    4'hC: {// 0xC0-0xCF PCIE (64GB) / DMUPIO
      if (read) data = gMem.read128(addr,myPort, 1);
      else gMem.writeBM(addr, data[63:0], size, myPort);  
    }
    4'hD: {// 0xD0-0xFE Reserved ERROR
      ioSpaceAccess = 0;
    }
    4'hE: {// 0xD0-0xFE Reserved ERROR
      ioSpaceAccess = 0;
    }
    4'hF: {// 0xD0-0xFE Reserved ERROR
      if (addr[39:32] == 8'hFF) {// 0xFF SSI (boot ROM) ERROR on write
        if (read) {
          if (read) data = gMem.read128(addr,myPort, 1);
        } else ioSpaceAccess = 0;
      } else {// 0xD0-0xFE Reserved ERROR
        ioSpaceAccess = 0;
      }
    }
  } // case


  // if +failNCUcheck fail the sim right now!
  if (ioSpaceAccess == 0 && failNCUcheck == 1)
    PR_ERROR (CLASSNAMEQ, MON_ERR, psprintf ("T%d Bench ERROR FAIL: - Attempt to access I/O address/ASI in a way that is NOT allowed. (addr=%0h (%0h),data=%0h,mask=%0h,read=%0h",thread,addr,addr[39:32],data,size,read));

  // if +noNCUcheck, always return a happy value
  // and (almost) always do the access
  if (noNCUcheck) {
    ioSpaceAccess = 1;
    if (read == 1 && addr[39:32] !== 8'h82)
      data = gMem.read128(addr,myPort, 1);
    if (read == 0 && addr[39:32] !== 8'hFF)
        gMem.writeBM(addr, data[63:0], size, myPort);
  }

  PR_INFO ("ioaccess", MON_INFO,
               psprintf("err = %b Normal IO Register Access",
               ioSpaceAccess));


#endif
#endif
#endif
}

//----------------------------------------------------------
// replicate data bytes across all data bytes depending on size.
// big endian, so the high bit bytes replicate from high bit
// end to low bit end.
function reg[127:0]  CLASSNAME::copyDataByte (reg [127:0] data, 
                                              reg [7:0] size,
                                              reg [3:0] offset)
{

  reg [127:0] tmpData = 0;

  // start with the addressed byte at byte 0
  data = data << (8*offset);
    
  case (size) {
    8'h0: {// 1 byte
      repeat (16) {
        tmpData = tmpData >> 8; // seems backwards but works
        tmpData[127:120] = data[127:120];
      }
    }
    8'h1: {// 2 bytes
      repeat (8) {
        tmpData = tmpData >> 16;        
        tmpData[127:112] = data[127:112];
      }
    }
    8'h2: {// 4 bytes
      repeat (4) {
        tmpData = tmpData >> 32;
        tmpData[127:96] = data[127:96];
      }
    }
    8'h3: {// 8 bytes
      repeat (2) {
        tmpData = tmpData >> 64;
        tmpData[127:64] = data[127:64];
      }
    }
    8'h4: {// 16 bytes
      tmpData = data;
    }
    
    default: PR_ERROR("copyDataByte", MON_ERR, psprintf ("ERROR bad size=%b passed in.",size));

  }

  copyDataByte = tmpData;
}


// Evictions use 2 functions, evictAddr & evictVector.
// "Eviction loop" code will call evictAddr to get address, then BFM will call
// evictVector to get the vector and modify the shadow tag.
// "Eviction loop" code will call task enqueueEvict in the correct bfm to
// request the invalidate after evictAddr has been called.
// The evict User Event that takes PA, will call task enqueueEvict in the
// correct bfm to request the invalidate.


//--------------------
// Pick a random core and search for a line with a valid entry.
// mem_index - itag 0..63, dtag 0..127
// evict_index is normalized index between 0..31
// used to search for tag in all lines in the group (aka L2 cache line)
// returns a valid address.

// Inputs:
//    core_enable = 8 bit vector, 1 bit per core
//    cid = N, to limit search to that core.
//    dCacheWeight = N to weight I/D caches. 100 % based
//
// Outputs:
//    found address if any, all F's otherwise.
//    cidUsed = a core that had the address
function reg [63:0] CLASSNAME::evictAddr(reg [7:0] core_enable,
                                         var reg [2:0] cidUsed,
                                         reg [3:0] cid = 4'hf,
                                         integer dCacheWeight = 60) // 100 % based
{
  reg         evict=0;       // =1, if valid entry found in the table
  reg [6:0]   mem_index=0;   // raw index from tag table search
  reg [28:0]  evict_tag=0;   // tag that is being evicted
  reg [4:0]   evict_index=0; // Normalized between 0..31  
  reg         junk;
  reg dCacheAddr = 0;
  
#ifndef CCM_BENCH
#ifndef FC_BENCH
#ifndef FC_SCAN_BENCH
  
  evictAddr = 64'hFFFFFFFFFFFFFFFF;
  cidUsed = 0;
  
  if (cid ==  4'hf) {
    // Pick cid out of the enabled cores.
    // This runs in zero time so randCID is safe from other calls to randomize().
    while (!core_enable[randCID]) {
      void = this.randomize(); 
    }
    cidUsed = randCID;
  } else {
    // check the cid
    cidUsed = cid;
    if (!core_enable[cidUsed])
      PR_ERROR("evict", MON_ERR,
               psprintf (" BENCH ERROR - selected core not available!"));
  }

  randcase {
    100-dCacheWeight: {
      // Get raw index that has a valid entry
      itag[cidUsed].search_tagmem(evict,mem_index);  // set evict=1, if valid entry found
      // found nothing
      if (! evict) return;
      // Normalize between 0..31
      evict_index = mem_index / 2;
      // Get the tag that will be evicted
      itag[cidUsed].get_tag(mem_index,junk,evict_tag);
    }
    dCacheWeight: {
      // Get raw index that has a valid entry
      dtag[cidUsed].search_tagmem(evict,mem_index);  // set evict=1, if valid entry found
      // found nothing
      if (! evict) return;     
      // Normalize between 0..31
      evict_index = mem_index / 4;
      // Get the tag that will be evicted
      dtag[cidUsed].get_tag(mem_index,junk,evict_tag);
      dCacheAddr = 1;
    }
  }

  evictAddr = {evict_tag,evict_index,6'b0};

  PR_INFO ("evict", MON_INFO,
           psprintf("EVICTION evictAddr PA{[39:6],6'b0}=%h evict_tag(pa[39:11])=%h evict_index(pa[10:6])=%h core=%0d dCacheAddr=%0d",
                    evictAddr,evict_tag,evict_index,cidUsed,dCacheAddr));

#endif
#endif
#endif
  
}


//----------------------------------------------------------
// Evict a PA from the L1 cache in all cores that are enabled.
// The value that is returned as evictVector has the format as defined
// in the CCX packet spec as Vinv.
//
// Pass in address. Will check all enabled cores for hit at that address and return
// an eviction vector for all matching/hitting cores.
//
// Inputs:
//    core_enable = 8 bit vector, 1 bit per core
//    evict_pa = address to evict
//    cpuId = CPU associated with evict_pa being cached
// Output:
//    evictVector = 0, no entries to evict.
//    targets = cores that will be the target of the evict packet.
function reg [127:0] CLASSNAME::evictVector(reg [7:0] core_enable,
                                            reg [63:0] evict_pa,
                                            reg [2:0] cpuId,
                                            var reg [7:0] targets)
{
  reg         evict=0;       // 1, if valid entry found in the table
  reg [6:0]   mem_index=0;   // raw index from tag table search
  reg [28:0]  evict_tag=0;   // tag that is being evicted
  reg [4:0]   evict_index=0; // Normalized between 0..31
  reg [3:0]   dmatch=0;
  reg [3:0]   imatch=0,i;
  reg [111:0] i_vect=0,d_vect=0,e_vect=0;

#ifndef CCM_BENCH
#ifndef FC_BENCH
#ifndef FC_SCAN_BENCH
  
  targets = 0;
  evictVector = 0;

  // Get index & tag from User argument
  evict_tag = evict_pa[39:11];
  evict_index = evict_pa[10:6];
  mem_index = evict_pa[10:4]; // tag class array index

  
  //-------------------- 
  // If valid tag found in tag table, evict it in both itag & dtag
 
  PR_INFO ("evict", MON_INFO,
           psprintf("EVICTION evictVector PA{[39:6],6'b0}=%h evict_tag(pa[39:11])=%h evict_index(pa[10:6])=%h core=%0d mem_index(pa[10:4])=%h",
                    evict_pa,evict_tag,evict_index,cpuId,mem_index));
  
  //-------------------- 
  // If valid tag found in tag table, evict it in both itag & dtag
  i_vect = 0;
  d_vect = 0;
  e_vect = 0;

  // evict the pa in all cores L1 cache
  for (i=0; i<=7; i=i+1) {
    if (core_enable[i]) {
      itag[i].evict_group(evict_tag, evict_index, i_vect);
      dtag[i].evict_group(evict_tag, evict_index, d_vect);

      if ((i_vect!=0)|(d_vect!=0)) {
        targets[i] = 1'b1;
      }

      // Create invalidation vector that is returned in CPX pkt
      e_vect = e_vect | i_vect | d_vect;
    }
  }


  if (e_vect) {
    //--------------------
    // Debug messages

    PR_INFO ("evict", MON_INFO,
             psprintf("EVICTION evictVector PA[39:0]=%h tag=%h evict_index(pa[10:6])=%h mem_index(pa[10:4])=%h+",evict_pa,evict_tag,evict_index,mem_index));
    PR_INFO ("evict", MON_INFO,
             psprintf("EVICTION \taddr:        0x%12h     0x%12h   0x%12h    0x%12h", evict_pa+48,evict_pa+32, evict_pa+16, evict_pa));
    PR_INFO ("evict", MON_INFO,
             psprintf("EVICTION \titag: Vinv[111:88] = %h [87:56] = %h [55:32] = %h [31:0] = %h ",
                      i_vect[111:88],i_vect[87:56],i_vect[55:32],i_vect[31:0]));
    PR_INFO ("evict", MON_INFO,
             psprintf("EVICTION \tdtag: Vinv[111:88] = %h [87:56] = %h [55:32] = %h [31:0] = %h ",
                      d_vect[111:88],d_vect[87:56],d_vect[55:32],d_vect[31:0]));
    PR_INFO ("evict", MON_INFO,
             psprintf("EVICTION \t      Vinv[111:88] = %h [87:56] = %h [55:32] = %h [31:0] = %h ",
                      e_vect[111:88],e_vect[87:56],e_vect[55:32],e_vect[31:0]));


    // Create Vinv as defined in CCX packet spec
    evictVector[127:117] = 11'b0;
    evictVector[116:112] = evict_index;
    evictVector[111:0]   = e_vect;
    
  } else {
    // not an error because time elapsed and the L1 tags have changed!
  }

#endif
#endif
#endif
  
}