Execute.bsv

Q&A

  • How does it caculate the memory access address?
    • where does it access TLB?

Global Functions:

  • function Bit#(a) arithmeticShift(Bit#(a) toShift, Bit#(b) shiftAmount)
  • function Bit#(a) arithmeticShift2(Bit#(a) toShift, Bit#(b) shiftAmount)

mkExecute module s

Input:

  • MIPSRegFileIfc rf
  • WritebackIfc writeback
  • CP0Ifc cp0
  • CoProIfc cop1
  • CapCopIfc capCop
  • FIFO#(ControlTokenT) inQ

Rules:

  • finishMultiplyOrDivide
  • deliverPendingOp

Methods:

  • enq
  • first
  • deq
  • clear

States:

  • FIFO#(ControlTokenT) outQ <- mkFIFO;
  • MulDivIfc mul <- mkMulDiv;
  • Reg#(MIPSReg) hi <- mkReg(64’b0);
  • Reg#(MIPSReg) lo <- mkReg(64’b0);
  • FIFOF#(Bool) hiLoPending <- mkFIFOF1;
  • FIFOF#(ControlTokenT) pendingOps <- mkFIFOF1;
  • Reg#(Bit#(16)) coreid <- mkConfigReg(0);

method Action enq(ControlTokenT di)

Condition:

  • !hiOrLoIsBlocking
  • !pendingOps.notEmpty

Input:

  • ControlTokenT di.{ opA, opB, pc, memSize, mem, inst, …}

Output:

  • ControlTokenT er.{opA, opB, }

Steps:

  • pre-execute instruction
    • check di.inst to see whether the instruction is for COP2, if so set cap=True;
    • get cop1 response with opA, opB
    • get PC from cap.Cop.getArchPc(er.pc, er.epoch)
    • call capCop.getArchPc(er.pc, er.epoch) assign to er.archPc
    • test the instruction is memory, or branch, or other instruction.
    • if instruction is for memory access:
    • use opA + opB, or just opB as offset to retrieve CapResponse by calling capVal <- capCop.getCapResponse(capReq, Memory) and grab the result from capVal.
    • bring result and store it at er.opA = capVal.data
    • go through coprocessor0 (MMU) by scResult <- cp0.setLlScReg(er.opA[63:0], er.test == LL, di.mem == Write)
      • cp0.setLlScReg: ???
    • if write, store conditional, get data to be written, either from cop1 (di.storeDatasrc == CoPro1) or from cop2 (di.storeDatasrc == CoPro2)
    • set write value writeVal = er.opB ?? Which one is address to be stored, opA or opB???
    • if instruction is for branch:
    • what does opA opB mean? which is the target address? It seems to be opA; and opB will be used to store the old PCC.
    • capReq.offset = unpack(er.opB)
    • capVal <- capCop.getCapResponse(capReq, Branch)
    • check capVal.valid and read capability data form Cop2 er.opB = capVal.data (the old pcc)
    • other instruction (not a memory op or a branch, classified as Arithmetic)
    • capVal <- capCop.getCapResponse(capReq, Arithmetic)
    • check capVal.valid.
    • do alu and store result to calcResult[64:0];
    • copy result to er.opA, er.carryout
    • pendingOps.enq(er)
    • outQ.enq(er)
Created Apr 5, 2020 // Last Updated May 23, 2020

If you could revise
the fundmental principles of
computer system design
to improve security...

... what would you change?