コード例 #1
0
    def elaborate_logic(s):

        # input register wen, wben, addr, wdata

        s.wen_reg = Reg(1)
        s.connect(s.wen_reg.in_, s.wen)

        s.wben_reg = Reg(s.num_nbytes)
        s.connect(s.wben_reg.in_, s.wben)

        s.addr_reg = Reg(s.addr_nbits)
        s.connect(s.addr_reg.in_, s.addr)

        s.wdata_reg = Reg(s.data_nbits)
        s.connect(s.wdata_reg.in_, s.wdata)

        # instantiate a s.combinational SRAM

        s.SRAM = m = SRAMBytesComb_rst_1rw(s.num_entries,
                                           s.num_nbytes,
                                           reset_value=s.reset_value)
        s.connect_dict({
            m.wen: s.wen_reg.out,
            m.wben: s.wben_reg.out,
            m.addr: s.addr_reg.out,
            m.wdata: s.wdata_reg.out,
            m.rdata: s.rdata
        })
コード例 #2
0
ファイル: queues.py プロジェクト: cornell-brg/mamba-dac2018
    def __init__(s, Type):
        s.enq = EnqIfcRTL(Type)
        s.deq = DeqIfcRTL(Type)

        s.buf = RegEn(Type)(in_=s.enq.msg)

        s.full = Reg(Bits1)

        s.byp_mux = Mux(Type, 2)(
            out=s.deq.msg,
            in_={
                0: s.enq.msg,
                1: s.buf.out,
            },
            sel=s.full.out,  # full -- buf.out, empty -- bypass
        )

        @s.update
        def up_bypq_set_enq_rdy():
            s.enq.rdy = ~s.full.out

        @s.update
        def up_bypq_set_deq_rdy():
            s.deq.rdy = s.full.out | s.enq.en  # if enq is enabled deq must be rdy

        @s.update
        def up_bypq_full():
            # enable buf <==> receiver marks deq.en=0 even if it sees deq.rdy=1
            s.buf.en = ~s.deq.en & s.enq.en
            s.full.in_ = ~s.deq.en & (s.enq.en | s.full.out)
コード例 #3
0
ファイル: queues.py プロジェクト: cornell-brg/mamba-dac2018
    def __init__(s, type_):
        s.enq = EnqIfcRTL(type_)
        s.deq = DeqIfcRTL(type_)

        s.buf = RegEn(type_)(out=s.deq.msg, in_=enq.msg)
        s.full = Reg(Bits1)

        @s.update
        def up_normq_set_both_rdy():
            s.enq.rdy = ~s.full.out
            s.deq.rdy = s.full.out

        @s.update
        def up_normq_full():
            s.buf.en = s.enq.en
            s.full.in_ = ~s.deq.en & (s.enq.en | s.full.out)
コード例 #4
0
ファイル: queues.py プロジェクト: cornell-brg/mamba-dac2018
    def __init__(s, Type):
        s.enq = EnqIfcRTL(Type)
        s.deq = DeqIfcRTL(Type)

        s.buf = RegEn(Type)(out=s.deq.msg, in_=s.enq.msg)
        s.full = Reg(Bits1)

        @s.update
        def up_pipeq_set_deq_rdy():
            s.deq.rdy = s.full.out

        @s.update
        def up_pipeq_set_enq_rdy():
            s.enq.rdy = ~s.full.out | s.deq.en

        @s.update
        def up_pipeq_full():
            s.buf.en = s.enq.en
            s.full.in_ = s.enq.en | (s.full.out & ~s.deq.en)
コード例 #5
0
    def __init__(s):

        s.req_val = InPort(1)
        s.req_rdy = OutPort(1)

        s.resp_val = OutPort(1)
        s.resp_rdy = InPort(1)

        s.req_msg_type = InPort(1)

        # Control signals (ctrl -> dpath)
        s.en_test = OutPort(1)
        s.en_train = OutPort(1)
        s.en_out = OutPort(1)
        s.sel_out = OutPort(1)
        s.sel = OutPort(6)

        # State element
        s.STATE_IDLE = 0
        s.STATE_INIT = 1
        s.STATE_CALC = 2
        s.STATE_DONE = 3

        s.state = RegRst(2, reset_value=s.STATE_IDLE)

        s.counter = Reg(DATA_NBITS)

        # State Transition Logic
        @s.combinational
        def state_transitions():

            curr_state = s.state.out
            next_state = s.state.out

            # Transition out of IDLE state
            if (curr_state == s.STATE_IDLE):
                if ((s.req_val and s.req_rdy) and (s.req_msg_type == 1)):
                    next_state = s.STATE_INIT
                elif ((s.req_val and s.req_rdy) and (s.req_msg_type == 0)):
                    next_state = s.STATE_CALC

            # Transition out of INIT state
            if (curr_state == s.STATE_INIT):
                if ((s.req_val and s.req_rdy) and (s.req_msg_type == 0)):
                    next_state = s.STATE_CALC

            # Transition out of CALC state
            if (curr_state == s.STATE_CALC):
                if ((s.counter.out + 1) == DATA_NBITS):
                    next_state = s.STATE_DONE

            # Transition out of DONE state
            if (curr_state == s.STATE_DONE):
                if (s.resp_val and s.resp_rdy):
                    next_state = s.STATE_IDLE

            s.state.in_.value = next_state

        # State Output Logic
        @s.combinational
        def state_outputs():

            current_state = s.state.out

            # IDLE state
            if current_state == s.STATE_IDLE:
                s.req_rdy.value = 1
                s.resp_val.value = 0
                s.en_out.value = 0
                s.en_test.value = 1
                s.en_train.value = 1
                s.sel.value = 0
                s.sel_out.value = 0
                s.counter.in_.value = 0

            # INIT state
            elif current_state == s.STATE_INIT:
                s.req_rdy.value = 1
                s.resp_val.value = 0
                s.en_out.value = 0
                s.en_test.value = 0
                s.en_train.value = 1
                s.sel.value = 0
                s.sel_out.value = 0
                s.counter.in_.value = 0

            # CALC state
            elif current_state == s.STATE_CALC:
                s.req_rdy.value = 0
                s.resp_val.value = 0
                s.en_out.value = 1
                if (s.counter.out == 0):
                    s.sel_out.value = 0
                else:
                    s.sel_out.value = 1
                s.en_test.value = 0
                s.en_train.value = 0
                s.sel.value = s.counter.out
                s.counter.in_.value = s.counter.out + 1

            # DONE state
            elif current_state == s.STATE_DONE:
                s.req_rdy.value = 0
                s.resp_val.value = 1
                s.en_out.value = 0
                s.en_test.value = 0
                s.en_train.value = 0
                s.sel.value = 0
                s.sel_out.value = 0
                s.counter.in_.value = 0
コード例 #6
0
    def __init__(s, mapper_num=10, reducer_num=1, train_size=600):

        TRAIN_DATA = train_size
        TRAIN_LOG = int(math.ceil(math.log(TRAIN_DATA, 2)))

        # import training data and store them into array
        training_data = []
        for i in xrange(DIGIT):
            count = 0
            filename = 'data/training_set_' + str(i) + '.dat'
            with open(filename, 'r') as f:
                for L in f:
                    if (count > TRAIN_DATA - 1):
                        break
                    training_data.append(int(L.replace(',\n', ''), 16))
                    count = count + 1

        # Top Level Interface
        s.in_ = InValRdyBundle(digitrecReqMsg())
        s.out = OutValRdyBundle(digitrecRespMsg())
        s.base = InPort(32)
        s.size = InPort(32)

        # Global Memory Interface
        s.gmem_req = OutValRdyBundle(MemReqMsg(8, 32, 64))
        s.gmem_resp = InValRdyBundle(MemRespMsg(8, 64))

        # Register File Interface
        s.regf_addr = OutPort[DIGIT](TRAIN_LOG)
        s.regf_data = OutPort[DIGIT](DATA_BITS)
        s.regf_wren = OutPort[DIGIT](1)
        s.regf_rdaddr = OutPort[mapper_num / reducer_num](TRAIN_LOG)

        # Mapper Interface
        s.map_req = OutPort[mapper_num](DATA_BITS)

        # Reducer Reset
        s.red_rst = OutPort(1)

        # Merger Interface
        s.merger_resp = InPort(DIGIT_LOG)

        # States
        s.STATE_IDLE = 0  # Idle state, scheduler waiting for top level to start
        s.STATE_SOURCE = 1  # Source state, handling with Test Source, getting base, size, ref info
        s.STATE_INIT = 2  # Init state, scheduler assigns input info to each Mapper
        s.STATE_START = 3  # Start state, scheduler gets test data, starts distributing and sorting
        s.STATE_WRITE = 4  # Write state, scheduler writes merger data to memory
        s.STATE_END = 5  # End state, shceduler loads all task from global memory and it is done

        s.state = RegRst(4, reset_value=s.STATE_IDLE)

        # Counters
        s.input_count = Wire(TEST_LOG)
        s.result_count = Wire(TEST_LOG)
        s.train_count_rd = Wire(TRAIN_LOG)
        s.train_count_wr = Wire(32)
        s.train_data_wr = Wire(1)
        s.train_data_rd = Wire(1)

        # Logic to Increment Counters
        @s.tick
        def counter():

            if (s.gmem_req.val and s.gmem_req.msg.type_ == TYPE_READ):
                s.input_count.next = s.input_count + 1

            if (s.gmem_req.val and s.gmem_req.msg.type_ == TYPE_WRITE):
                s.result_count.next = s.result_count + 1

            if s.rst:
                s.train_count_rd.next = 0
            elif s.train_data_rd:
                s.train_count_rd.next = s.train_count_rd + (mapper_num / DIGIT)

            if (s.train_data_wr):
                s.train_count_wr.next = s.train_count_wr + 1

        # Signals
        s.go = Wire(1)  # go signal tells scheduler to start scheduling
        s.done = Wire(1)  # done signal indicates everything is done
        s.rst = Wire(1)  # reset train count every test data processed

        # Reference data
        s.reference = Reg(dtype=DATA_BITS)  # reference stores test data

        #---------------------------------------------------------------------
        # Initialize Register File for Training data
        #---------------------------------------------------------------------

        @s.combinational
        def traindata():

            if s.train_data_wr:
                for i in xrange(DIGIT):
                    s.regf_addr[i].value = s.train_count_wr
                    s.regf_data[i].value = training_data[i * TRAIN_DATA +
                                                         s.train_count_wr]
                    s.regf_wren[i].value = 1
            else:
                for i in xrange(DIGIT):
                    s.regf_wren[i].value = 0

        #---------------------------------------------------------------------
        # Assign Task to Mapper Combinational Logic
        #---------------------------------------------------------------------

        @s.combinational
        def mapper():

            # broadcast train data to mapper
            for i in xrange(DIGIT):
                for j in xrange(mapper_num / DIGIT):
                    if (s.train_data_rd):
                        s.map_req[j * 10 + i].value = s.reference.out
                        s.regf_rdaddr[j].value = s.train_count_rd + j

        #---------------------------------------------------------------------
        # Task State Transition Logic
        #---------------------------------------------------------------------

        @s.combinational
        def state_transitions():

            curr_state = s.state.out
            next_state = s.state.out

            if (curr_state == s.STATE_IDLE):
                if (s.in_.val):
                    next_state = s.STATE_SOURCE

            if (curr_state == s.STATE_SOURCE):
                if (s.go):
                    next_state = s.STATE_INIT
                elif (s.done):
                    next_state = s.STATE_IDLE

            if (curr_state == s.STATE_INIT):
                if (s.train_count_wr == TRAIN_DATA - 1):
                    next_state = s.STATE_START

            if (curr_state == s.STATE_START):
                if (s.train_count_rd == TRAIN_DATA - (mapper_num / DIGIT)):
                    next_state = s.STATE_WRITE

            if (curr_state == s.STATE_WRITE):
                if (s.input_count == s.size):
                    next_state = s.STATE_END
                else:
                    next_state = s.STATE_START

            if (curr_state == s.STATE_END):
                if s.gmem_resp.val:
                    next_state = s.STATE_SOURCE

            s.state.in_.value = next_state

        #---------------------------------------------------------------------
        # Task State Output Logic
        #---------------------------------------------------------------------

        @s.combinational
        def state_outputs():

            current_state = s.state.out
            s.gmem_req.val.value = 0
            s.gmem_resp.rdy.value = 0
            s.in_.rdy.value = 0
            s.out.val.value = 0

            # In IDLE state
            if (current_state == s.STATE_IDLE):
                s.input_count.value = 0
                s.train_count_rd.value = 0
                s.train_count_wr.value = 0
                s.reference.value = 0
                s.go.value = 0
                s.train_data_rd.value = 0
                s.train_data_wr.value = 0
                s.done.value = 0
                s.rst.value = 0
                s.red_rst.value = 0

            # In SOURCE state
            if (current_state == s.STATE_SOURCE):
                if (s.in_.val and s.out.rdy):
                    if (s.in_.msg.type_ == digitrecReqMsg.TYPE_WRITE):
                        if (s.in_.msg.addr == 0):  # start computing
                            s.go.value = 1
                        elif (s.in_.msg.addr == 1):  # base address
                            s.base.value = s.in_.msg.data
                        elif (s.in_.msg.addr == 2):  # size
                            s.size.value = s.in_.msg.data

                        # Send xcel response message
                        s.in_.rdy.value = 1
                        s.out.msg.type_.value = digitrecReqMsg.TYPE_WRITE
                        s.out.msg.data.value = 0
                        s.out.val.value = 1

                    elif (s.in_.msg.type_ == digitrecReqMsg.TYPE_READ):
                        # the computing is done, send response message
                        if (s.done):

                            s.out.msg.type_.value = digitrecReqMsg.TYPE_READ
                            s.out.msg.data.value = 1
                            s.in_.rdy.value = 1
                            s.out.val.value = 1

            # In INIT state
            if (current_state == s.STATE_INIT):

                s.train_data_wr.value = 1
                s.go.value = 0

                # at the end of init, send read req to global memory
                if s.train_count_wr == TRAIN_DATA - 1:
                    if s.gmem_req.rdy:
                        s.gmem_req.msg.addr.value = s.base + (8 *
                                                              s.input_count)
                        s.gmem_req.msg.type_.value = TYPE_READ
                        s.gmem_req.val.value = 1
                        s.red_rst.value = 1

            # In START state
            if (current_state == s.STATE_START):

                s.train_data_wr.value = 0
                s.train_data_rd.value = 1
                s.rst.value = 0
                s.red_rst.value = 0

                if s.gmem_resp.val:
                    # if response type is read, stores test data to reference, hold response val
                    # until everything is done, which is set in WRITE state
                    if s.gmem_resp.msg.type_ == TYPE_READ:
                        s.gmem_resp.rdy.value = 1
                        s.reference.in_.value = s.gmem_resp.msg.data
                    else:
                        # if response tyle is write, set response rdy, send another req to
                        # read test data
                        s.gmem_resp.rdy.value = 1
                        s.gmem_req.msg.addr.value = s.base + (8 *
                                                              s.input_count)
                        s.gmem_req.msg.type_.value = TYPE_READ
                        s.gmem_req.val.value = 1
                        s.red_rst.value = 1
                        s.train_data_rd.value = 0

            # In WRITE state
            if (current_state == s.STATE_WRITE):

                s.train_data_rd.value = 0
                # one test data done processed, write result from merger to memory
                if (s.gmem_req.rdy):
                    s.gmem_req.msg.addr.value = 0x2000 + (8 * s.result_count)
                    s.gmem_req.msg.data.value = s.merger_resp
                    s.gmem_req.msg.type_.value = TYPE_WRITE
                    s.gmem_req.val.value = 1
                    s.rst.value = 1

            # In END state
            if (current_state == s.STATE_END):
                if s.gmem_resp.val:
                    s.gmem_resp.rdy.value = 1
                    s.done.value = 1
コード例 #7
0
def test_reg16():
  reg_test( Reg(16) )
コード例 #8
0
def test_reg8():
  reg_test( Reg(8) )
コード例 #9
0
    def __init__(s):

        #==================================================================
        # Interfaces
        #==================================================================

        s.req_msg_a = InPort(32)
        s.req_msg_b = InPort(32)
        s.resp_msg = OutPort(32)

        # Control signals

        s.a_mux_sel = InPort(A_MUX_SEL_NBITS)
        s.b_mux_sel = InPort(B_MUX_SEL_NBITS)
        s.result_mux_sel = InPort(RES_MUX_SEL_NBITS)
        s.add_mux_sel = InPort(ADD_MUX_SEL_NBITS)
        s.result_en = InPort(1)
        s.result_sign = InPort(OUT_MUX_SEL_NBITS)

        # Status signals

        s.b_lsb = OutPort(1)
        s.a_msb = OutPort(1)
        s.b_msb = OutPort(1)

        s.to_ctrl_shamt = OutPort(6)

        # shamt input to shifters, calculated by ShamtGen

        s.shamt = Wire(6)

        # Binary representation of the multiplier

        s.bit_string = Wire(32)

        #==================================================================
        # Structure
        #==================================================================

        # A Mux

        s.in_a = Wire(32)

        # Take the absolute value of the input

        @s.combinational
        def sign_handling_a():
            s.in_a.value = s.req_msg_a if ~s.req_msg_a[31] \
                           else (~s.req_msg_a) + 1

        s.l_shift_out = Wire(32)

        s.a_mux = m = Mux(32, 2)

        s.connect_pairs(
            m.sel,
            s.a_mux_sel,
            m.in_[A_MUX_SEL_IN],
            s.in_a,
            m.in_[A_MUX_SEL_SHIFT],
            s.l_shift_out,
        )

        # A Register

        s.a_reg = m = Reg(32)

        s.connect(m.in_, s.a_mux.out)

        # Left Shifter

        s.l_shift = m = LeftLogicalShifter(32, 6)

        s.connect_pairs(
            m.in_,
            s.a_reg.out,
            m.shamt,
            s.shamt,
            m.out,
            s.l_shift_out,
        )

        # B Mux

        s.in_b = Wire(32)

        # Take the absolute value of the input

        @s.combinational
        def sign_handling_b():
            s.in_b.value = s.req_msg_b if ~s.req_msg_b[31] \
                           else (~s.req_msg_b) + 1

        s.r_shift_out = Wire(32)

        s.b_mux = m = Mux(32, 2)

        s.connect_pairs(
            m.sel,
            s.b_mux_sel,
            m.in_[B_MUX_SEL_IN],
            s.in_b,
            m.in_[B_MUX_SEL_SHIFT],
            s.r_shift_out,
        )

        # B Register

        s.b_reg = m = Reg(32)

        s.connect(m.in_, s.b_mux.out)

        # Take the higher 31 bits and add 0 in the high order
        # The ShamtGen module will generate the appropriate shamt
        # according to the number of consecutive zeros in lower s.bit_string

        @s.combinational
        def bit_string_block():
            s.bit_string.value = concat(Bits(1, 0), s.b_reg.out[1:32])

        # Right Shifter

        s.r_shift = m = RightLogicalShifter(32, 6)

        s.connect_pairs(
            m.in_,
            s.b_reg.out,
            m.shamt,
            s.shamt,
            m.out,
            s.r_shift_out,
        )

        # Result Mux

        s.add_mux_out = Wire(32)

        s.result_mux = m = Mux(32, 2)

        s.connect_pairs(
            m.sel,
            s.result_mux_sel,
            m.in_[RES_MUX_SEL_ZERO],
            0,
            m.in_[RES_MUX_SEL_ADD],
            s.add_mux_out,
        )

        # Result Register

        s.res_reg = m = RegEn(32)

        s.connect_pairs(m.in_, s.result_mux.out, m.en, s.result_en)

        # Adder

        s.adder = m = Adder(32)

        s.connect_pairs(
            m.in0,
            s.a_reg.out,
            m.in1,
            s.res_reg.out,
            m.cin,
            0,
        )

        # Add Mux

        s.add_mux = m = Mux(32, 2)

        s.connect_pairs(
            m.sel,
            s.add_mux_sel,
            m.in_[ADD_MUX_SEL_ADD],
            s.adder.out,
            m.in_[ADD_MUX_SEL_RES],
            s.res_reg.out,
            m.out,
            s.add_mux_out,
        )

        # ShamtGen

        s.shamt_gen = m = ShamtGenPRTL()

        s.connect_pairs(
            m.a,
            s.bit_string,
            m.shamt,
            s.shamt,
        )

        # Forward shamt to control unit so the counter can update
        # accordingly

        @s.combinational
        def to_ctrl_shamt_block():
            s.to_ctrl_shamt.value = s.shamt

        # Output MUX
        s.res_neg = Wire(32)

        # Generate -res in case the result is negative

        @s.combinational
        def twos_compl_block():
            s.res_neg.value = (~s.res_reg.out) + 1

        s.out_mux = m = Mux(32, 2)

        s.connect_pairs(
            m.sel,
            s.result_sign,
            m.in_[OUT_MUX_SEL_POS],
            s.res_reg.out,
            m.in_[OUT_MUX_SEL_NEG],
            s.res_neg,
            m.out,
            s.resp_msg,
        )

        # Connect status signals

        s.connect(s.b_reg.out[0], s.b_lsb)
        s.connect(s.req_msg_a[31], s.a_msb)
        s.connect(s.req_msg_b[31], s.b_msb)
コード例 #10
0
    def __init__(s):

        #==================================================================
        # Interfaces
        #==================================================================

        s.req_msg_a = InPort(32)
        s.req_msg_b = InPort(32)
        s.resp_msg = OutPort(32)

        # Control signals

        s.a_mux_sel = InPort(A_MUX_SEL_NBITS)
        s.b_mux_sel = InPort(B_MUX_SEL_NBITS)
        s.result_mux_sel = InPort(RES_MUX_SEL_NBITS)
        s.add_mux_sel = InPort(ADD_MUX_SEL_NBITS)
        s.result_en = InPort(1)
        s.result_sign = InPort(OUT_MUX_SEL_NBITS)

        # Status signals

        s.b_lsb = OutPort(1)
        s.a_msb = OutPort(1)
        s.b_msb = OutPort(1)

        #==================================================================
        # Structure
        #==================================================================

        # A Mux

        s.in_a = Wire(32)

        # Take the abs value of the input

        @s.combinational
        def sign_handling_a():
            s.in_a.value = s.req_msg_a if ~s.req_msg_a[31] \
                           else (~s.req_msg_a) + 1

        s.l_shift_out = Wire(32)

        s.a_mux = m = Mux(32, 2)

        s.connect_pairs(
            m.sel,
            s.a_mux_sel,
            m.in_[A_MUX_SEL_IN],
            s.in_a,
            m.in_[A_MUX_SEL_SHIFT],
            s.l_shift_out,
        )

        # A Register

        s.a_reg = m = Reg(32)

        s.connect(m.in_, s.a_mux.out)

        # Left Shifter

        s.l_shift = m = LeftLogicalShifter(32)

        s.connect_pairs(
            m.in_,
            s.a_reg.out,
            m.shamt,
            1,
            m.out,
            s.l_shift_out,
        )

        # B Mux

        s.in_b = Wire(32)

        # Take the abs value of the input

        @s.combinational
        def sign_handling_b():
            s.in_b.value = s.req_msg_b if ~s.req_msg_b[31] \
                           else (~s.req_msg_b) + 1

        s.r_shift_out = Wire(32)

        s.b_mux = m = Mux(32, 2)

        s.connect_pairs(
            m.sel,
            s.b_mux_sel,
            m.in_[B_MUX_SEL_IN],
            s.in_b,
            m.in_[B_MUX_SEL_SHIFT],
            s.r_shift_out,
        )

        # B Register

        s.b_reg = m = Reg(32)

        s.connect(m.in_, s.b_mux.out)

        # Right Shifter

        s.r_shift = m = RightLogicalShifter(32)

        s.connect_pairs(
            m.in_,
            s.b_reg.out,
            m.shamt,
            1,
            m.out,
            s.r_shift_out,
        )

        # Result Mux

        s.add_mux_out = Wire(32)

        s.result_mux = m = Mux(32, 2)

        s.connect_pairs(
            m.sel,
            s.result_mux_sel,
            m.in_[RES_MUX_SEL_ZERO],
            0,
            m.in_[RES_MUX_SEL_ADD],
            s.add_mux_out,
        )

        # Result Register

        s.res_reg = m = RegEn(32)

        s.connect_pairs(m.in_, s.result_mux.out, m.en, s.result_en)

        # Adder

        s.adder = m = Adder(32)

        s.connect_pairs(
            m.in0,
            s.a_reg.out,
            m.in1,
            s.res_reg.out,
            m.cin,
            0,
        )

        # Add Mux

        s.add_mux = m = Mux(32, 2)

        s.connect_pairs(
            m.sel,
            s.add_mux_sel,
            m.in_[ADD_MUX_SEL_ADD],
            s.adder.out,
            m.in_[ADD_MUX_SEL_RES],
            s.res_reg.out,
            m.out,
            s.add_mux_out,
        )

        # Output MUX
        s.res_neg = Wire(32)

        # Generate -res in case the output is negative

        @s.combinational
        def twos_compl_block():
            s.res_neg.value = (~s.res_reg.out) + 1

        s.out_mux = m = Mux(32, 2)

        s.connect_pairs(
            m.sel,
            s.result_sign,
            m.in_[OUT_MUX_SEL_POS],
            s.res_reg.out,
            m.in_[OUT_MUX_SEL_NEG],
            s.res_neg,
            m.out,
            s.resp_msg,
        )

        # Connect status signals

        s.connect(s.b_reg.out[0], s.b_lsb)
        s.connect(s.req_msg_a[31], s.a_msb)
        s.connect(s.req_msg_b[31], s.b_msb)
コード例 #11
0
ファイル: SortXcelRTL.py プロジェクト: nmtrmail/pymtl-tut-hls
    def __init__(s, mem_ifc_types=MemMsg(8, 32, 32)):

        # Interface

        s.xcelreq = InValRdyBundle(XcelReqMsg())
        s.xcelresp = OutValRdyBundle(XcelRespMsg())

        s.memreq = OutValRdyBundle(mem_ifc_types.req)
        s.memresp = InValRdyBundle(mem_ifc_types.resp)

        # Queues

        s.xcelreq_q = SingleElementPipelinedQueue(XcelReqMsg())
        s.connect(s.xcelreq, s.xcelreq_q.enq)

        s.memreq_q = SingleElementBypassQueue(MemReqMsg(8, 32, 32))
        s.connect(s.memreq, s.memreq_q.deq)

        s.memresp_q = SingleElementPipelinedQueue(MemRespMsg(8, 32))
        s.connect(s.memresp, s.memresp_q.enq)

        # Internal state

        s.base_addr = Reg(32)
        s.size = Reg(32)
        s.inner_count = Reg(32)
        s.outer_count = Reg(32)
        s.a = Reg(32)

        # Line tracing

        s.prev_state = 0
        s.xcfg_trace = "  "

        # Helpers to make memory read/write requests

        s.mk_rd = mem_ifc_types.req.mk_rd
        s.mk_wr = mem_ifc_types.req.mk_wr

        #=====================================================================
        # State Update
        #=====================================================================

        s.STATE_XCFG = 0
        s.STATE_FIRST0 = 1
        s.STATE_FIRST1 = 2
        s.STATE_BUBBLE0 = 3
        s.STATE_BUBBLE1 = 4
        s.STATE_LAST = 5

        s.state = Wire(8)

        s.go = Wire(1)

        @s.tick_rtl
        def block0():

            if s.reset:
                s.state.next = s.STATE_XCFG
            else:
                s.state.next = s.state

                if s.state == s.STATE_XCFG:
                    if s.go & s.xcelresp.val & s.xcelresp.rdy:
                        s.state.next = s.STATE_FIRST0

                elif s.state == s.STATE_FIRST0:
                    if s.memreq_q.enq.rdy:
                        s.state.next = s.STATE_FIRST1

                elif s.state == s.STATE_FIRST1:
                    if s.memreq_q.enq.rdy and s.memresp_q.deq.rdy:
                        s.state.next = s.STATE_BUBBLE0

                elif s.state == s.STATE_BUBBLE0:
                    if s.memreq_q.enq.rdy and s.memresp_q.deq.rdy:
                        s.state.next = s.STATE_BUBBLE1

                elif s.state == s.STATE_BUBBLE1:
                    if s.memreq_q.enq.rdy and s.memresp_q.deq.rdy:
                        if s.inner_count.out + 1 < s.size.out:
                            s.state.next = s.STATE_BUBBLE0
                        else:
                            s.state.next = s.STATE_LAST

                elif s.state == s.STATE_LAST:
                    if s.memreq_q.enq.rdy and s.memresp_q.deq.rdy:
                        if s.outer_count.out + 1 < s.size.out:
                            s.state.next = s.STATE_FIRST1
                        else:
                            s.state.next = s.STATE_XCFG

        #=====================================================================
        # State Outputs
        #=====================================================================

        @s.combinational
        def block1():

            s.xcelreq_q.deq.rdy.value = 0
            s.xcelresp.val.value = 0
            s.memreq_q.enq.val.value = 0
            s.memresp_q.deq.rdy.value = 0
            s.go.value = 0

            s.outer_count.in_.value = s.outer_count.out
            s.inner_count.in_.value = s.inner_count.out

            #-------------------------------------------------------------------
            # STATE: XCFG
            #-------------------------------------------------------------------

            if s.state == s.STATE_XCFG:
                s.xcelreq_q.deq.rdy.value = s.xcelresp.rdy
                s.xcelresp.val.value = s.xcelreq_q.deq.val

                if s.xcelreq_q.deq.val:

                    if s.xcelreq_q.deq.msg.type_ == XcelReqMsg.TYPE_WRITE:

                        if s.xcelreq_q.deq.msg.raddr == 0:
                            s.outer_count.in_.value = 0
                            s.go.value = 1

                        elif s.xcelreq_q.deq.msg.raddr == 1:
                            s.base_addr.in_.value = s.xcelreq_q.deq.msg.data

                        elif s.xcelreq_q.deq.msg.raddr == 2:
                            s.size.in_.value = s.xcelreq_q.deq.msg.data

                        # Send xcel response message

                        s.xcelresp.msg.type_.value = XcelRespMsg.TYPE_WRITE

                    else:

                        # Send xcel response message, obviously you only want to
                        # send the response message when accelerator is done

                        s.xcelresp.msg.type_.value = XcelRespMsg.TYPE_READ
                        s.xcelresp.msg.data.value = 1

            #-------------------------------------------------------------------
            # STATE: FIRST0
            #-------------------------------------------------------------------
            # Send the first memory read request for the very first
            # element in the array.

            elif s.state == s.STATE_FIRST0:
                if s.memreq_q.enq.rdy:

                    s.memreq_q.enq.val.value = 1
                    s.memreq_q.enq.msg.type_.value = MemReqMsg.TYPE_READ
                    s.memreq_q.enq.msg.opaque.value = 0
                    s.memreq_q.enq.msg.addr.value = s.base_addr.out + 4 * s.inner_count.out
                    s.memreq_q.enq.msg.len.value = 0

                    s.inner_count.in_.value = 1

            #-------------------------------------------------------------------
            # STATE: FIRST1
            #-------------------------------------------------------------------
            # Wait for the memory response for the first element in the array,
            # and once it arrives store this element in a, and send the memory
            # read request for the second element.

            elif s.state == s.STATE_FIRST1:
                if s.memreq_q.enq.rdy and s.memresp_q.deq.val:
                    s.memresp_q.deq.rdy.value = 1
                    s.a.in_.value = s.memresp_q.deq.msg.data

                    s.memreq_q.enq.val.value = 1
                    s.memreq_q.enq.msg.type_.value = MemReqMsg.TYPE_READ
                    s.memreq_q.enq.msg.opaque.value = 0
                    s.memreq_q.enq.msg.addr.value = s.base_addr.out + 4 * s.inner_count.out
                    s.memreq_q.enq.msg.len.value = 0

            #-------------------------------------------------------------------
            # STATE: BUBBLE0
            #-------------------------------------------------------------------
            # Wait for the memory read response to get the next element,
            # compare the new value to the previous max value, update b with
            # the new max value, and send a memory request to store the new min
            # value. Notice how we decrement the write address by four since we
            # want to store to the new min value _previous_ element.

            elif s.state == s.STATE_BUBBLE0:
                if s.memreq_q.enq.rdy and s.memresp_q.deq.val:
                    s.memresp_q.deq.rdy.value = 1

                    if s.a.out > s.memresp_q.deq.msg:
                        s.a.in_.value = s.a.out
                        s.memreq_q.enq.msg.data.value = s.memresp_q.deq.msg
                    else:
                        s.a.in_.value = s.memresp_q.deq.msg
                        s.memreq_q.enq.msg.data.value = s.a.out

                    s.memreq_q.enq.val.value = 1
                    s.memreq_q.enq.msg.type_.value = MemReqMsg.TYPE_WRITE
                    s.memreq_q.enq.msg.opaque.value = 0
                    s.memreq_q.enq.msg.addr.value = (s.base_addr.out + 4 *
                                                     (s.inner_count.out - 1))
                    s.memreq_q.enq.msg.len.value = 0

            #-------------------------------------------------------------------
            # STATE: BUBBLE1
            #-------------------------------------------------------------------
            # Wait for the memory write response, and then check to see if we
            # have reached the end of the array. If we have not reached the end
            # of the array, then make a new memory read request for the next
            # element; if we have reached the end of the array, then make a
            # final write request (with value from a) to update the final
            # element in the array.

            elif s.state == s.STATE_BUBBLE1:
                if s.memreq_q.enq.rdy and s.memresp_q.deq.val:
                    s.memresp_q.deq.rdy.value = 1

                    s.inner_count.in_.value = s.inner_count.out + 1
                    if s.inner_count.out + 1 < s.size.out:

                        s.memreq_q.enq.val.value = 1
                        s.memreq_q.enq.msg.type_.value = MemReqMsg.TYPE_READ
                        s.memreq_q.enq.msg.opaque.value = 0
                        s.memreq_q.enq.msg.addr.value = s.base_addr.out + 4 * (
                            s.inner_count.out + 1)
                        s.memreq_q.enq.msg.len.value = 0

                    else:

                        s.memreq_q.enq.val.value = 1
                        s.memreq_q.enq.msg.type_.value = MemReqMsg.TYPE_WRITE
                        s.memreq_q.enq.msg.opaque.value = 0
                        s.memreq_q.enq.msg.addr.value = (s.base_addr.out + 4 *
                                                         (s.inner_count.out))
                        s.memreq_q.enq.msg.len.value = 0
                        s.memreq_q.enq.msg.data.value = s.a.out

            #-------------------------------------------------------------------
            # STATE: LAST
            #-------------------------------------------------------------------
            # Wait for the last response, and then check to see if we need to
            # go through the array again. If we do need to go through array
            # again, then make a new memory read request for the very first
            # element in the array; if we do not need to go through the array
            # again, then we are all done and we can go back to accelerator
            # configuration.

            elif s.state == s.STATE_LAST:
                if s.memreq_q.enq.rdy and s.memresp_q.deq.val:
                    s.memresp_q.deq.rdy.value = 1

                    s.outer_count.in_.value = s.outer_count.out + 1
                    if s.outer_count.out + 1 < s.size.out:

                        s.memreq_q.enq.val.value = 1
                        s.memreq_q.enq.msg.type_.value = MemReqMsg.TYPE_READ
                        s.memreq_q.enq.msg.opaque.value = 0
                        s.memreq_q.enq.msg.addr.value = s.base_addr.out
                        s.memreq_q.enq.msg.len.value = 0

                        s.inner_count.in_.value = 1
コード例 #12
0
    def elaborate_logic(s):

        #---------------------------------------------------------------------
        # Stage A->B pipeline registers
        #---------------------------------------------------------------------

        s.reg_AB = [Reg(16) for x in range(4)]
        s.connect(s.reg_AB[0].in_, s.in_[0])
        s.connect(s.reg_AB[1].in_, s.in_[1])
        s.connect(s.reg_AB[2].in_, s.in_[2])
        s.connect(s.reg_AB[3].in_, s.in_[3])

        #---------------------------------------------------------------------
        # Stage B combinational logic
        #---------------------------------------------------------------------

        s.cmp_B0 = m = MinMax()
        s.connect_dict({
            m.in0: s.reg_AB[0].out,
            m.in1: s.reg_AB[1].out,
        })

        s.cmp_B1 = m = MinMax()
        s.connect_dict({
            m.in0: s.reg_AB[2].out,
            m.in1: s.reg_AB[3].out,
        })

        #---------------------------------------------------------------------
        # Stage B->C pipeline registers
        #---------------------------------------------------------------------

        s.reg_BC = [Reg(16) for x in range(4)]
        s.connect(s.reg_BC[0].in_, s.cmp_B0.min)
        s.connect(s.reg_BC[1].in_, s.cmp_B0.max)
        s.connect(s.reg_BC[2].in_, s.cmp_B1.min)
        s.connect(s.reg_BC[3].in_, s.cmp_B1.max)

        #---------------------------------------------------------------------
        # Stage C combinational logic
        #---------------------------------------------------------------------

        s.cmp_C0 = m = MinMax()
        s.connect_dict({
            m.in0: s.reg_BC[0].out,
            m.in1: s.reg_BC[2].out,
        })

        s.cmp_C1 = m = MinMax()
        s.connect_dict({
            m.in0: s.reg_BC[1].out,
            m.in1: s.reg_BC[3].out,
        })

        s.cmp_C2 = m = MinMax()
        s.connect_dict({
            m.in0: s.cmp_C0.max,
            m.in1: s.cmp_C1.min,
        })

        # Connect to output ports

        s.connect(s.cmp_C0.min, s.out[0])
        s.connect(s.cmp_C2.min, s.out[1])
        s.connect(s.cmp_C2.max, s.out[2])
        s.connect(s.cmp_C1.max, s.out[3])
コード例 #13
0
    def __init__(s, nbits):
        nbitsx2 = nbits * 2

        dtype = mk_bits(nbits)
        dtypex2 = mk_bits(nbitsx2)

        s.req_msg = InVPort(dtypex2)
        s.resp_msg = OutVPort(dtypex2)

        # Status signals

        s.sub_negative1 = OutVPort(Bits1)
        s.sub_negative2 = OutVPort(Bits1)

        # Control signals

        s.quotient_mux_sel = InVPort(Bits1)
        s.quotient_reg_en = InVPort(Bits1)

        s.remainder_mux_sel = InVPort(Bits2)
        s.remainder_reg_en = InVPort(Bits1)

        s.divisor_mux_sel = InVPort(Bits1)

        # Dpath components

        s.remainder_mux = Mux(dtypex2, 3)(sel=s.remainder_mux_sel)

        @s.update
        def up_remainder_mux_in0():
            s.remainder_mux.in_[R_MUX_SEL_IN] = dtypex2()
            s.remainder_mux.in_[R_MUX_SEL_IN][0:nbits] = s.req_msg[0:nbits]

        s.remainder_reg = RegEn(dtypex2)(
            in_=s.remainder_mux.out,
            en=s.remainder_reg_en,
        )
        # lower bits of resp_msg save the remainder
        s.connect(s.resp_msg[0:nbits], s.remainder_reg.out[0:nbits])

        s.divisor_mux = Mux(dtypex2, 2)(sel=s.divisor_mux_sel)

        @s.update
        def up_divisor_mux_in0():
            s.divisor_mux.in_[D_MUX_SEL_IN] = dtypex2()
            s.divisor_mux.in_[D_MUX_SEL_IN][nbits - 1:nbitsx2 -
                                            1] = s.req_msg[nbits:nbitsx2]

        s.divisor_reg = Reg(dtypex2)(in_=s.divisor_mux.out)

        s.quotient_mux = Mux(dtype, 2)(sel=s.quotient_mux_sel)
        s.connect(s.quotient_mux.in_[Q_MUX_SEL_0], 0)

        s.quotient_reg = RegEn(dtype)(
            in_=s.quotient_mux.out,
            en=s.quotient_reg_en,
            # higher bits of resp_msg save the quotient
            out=s.resp_msg[nbits:nbitsx2],
        )

        # shamt should be 2 bits!
        s.quotient_lsh = LShifter(dtype, 2)(in_=s.quotient_reg.out)
        s.connect(s.quotient_lsh.shamt, 2)

        s.inc = Wire(Bits2)
        s.connect(s.sub_negative1, s.inc[1])
        s.connect(s.sub_negative2, s.inc[0])

        @s.update
        def up_quotient_inc():
            s.quotient_mux.in_[Q_MUX_SEL_LSH] = s.quotient_lsh.out + ~s.inc

        # stage 1/2

        s.sub1 = Subtractor(dtypex2)(
            in0=s.remainder_reg.out,
            in1=s.divisor_reg.out,
            out=s.remainder_mux.in_[R_MUX_SEL_SUB1],
        )
        s.connect(s.sub_negative1, s.sub1.out[nbitsx2 - 1])

        s.remainder_mid_mux = Mux(dtypex2, 2)(
            in_={
                0: s.sub1.out,
                1: s.remainder_reg.out,
            },
            sel=s.sub_negative1,
        )

        s.divisor_rsh1 = RShifter(dtypex2, 1)(in_=s.divisor_reg.out, )
        s.connect(s.divisor_rsh1.shamt, 1)

        # stage 2/2

        s.sub2 = Subtractor(dtypex2)(
            in0=s.remainder_mid_mux.out,
            in1=s.divisor_rsh1.out,
            out=s.remainder_mux.in_[R_MUX_SEL_SUB2],
        )

        s.connect(s.sub_negative2, s.sub2.out[nbitsx2 - 1])

        s.divisor_rsh2 = RShifter(dtypex2, 1)(
            in_=s.divisor_rsh1.out,
            out=s.divisor_mux.in_[D_MUX_SEL_RSH],
        )
        s.connect(s.divisor_rsh2.shamt, 1)
コード例 #14
0
    def __init__(s, nbits):
        s.req_val = InVPort(Bits1)
        s.req_rdy = OutVPort(Bits1)
        s.resp_val = OutVPort(Bits1)
        s.resp_rdy = InVPort(Bits1)

        # Status signals

        s.sub_negative1 = InVPort(Bits1)
        s.sub_negative2 = InVPort(Bits1)

        # Control signals

        s.quotient_mux_sel = OutVPort(Bits1)
        s.quotient_reg_en = OutVPort(Bits1)

        s.remainder_mux_sel = OutVPort(Bits2)
        s.remainder_reg_en = OutVPort(Bits1)

        s.divisor_mux_sel = OutVPort(Bits1)

        state_dtype = mk_bits(1 + clog2(nbits))
        s.state = Reg(state_dtype)

        s.STATE_IDLE = state_dtype(0)
        s.STATE_DONE = state_dtype(1)
        s.STATE_CALC = state_dtype(1 + nbits / 2)

        @s.update
        def state_transitions():

            curr_state = s.state.out

            if curr_state == s.STATE_IDLE:
                if s.req_val and s.req_rdy:
                    s.state.in_ = s.STATE_CALC

            elif curr_state == s.STATE_DONE:
                if s.resp_val and s.resp_rdy:
                    s.state.in_ = s.STATE_IDLE

            else:
                s.state.in_ = curr_state - 1

        @s.update
        def state_outputs():

            curr_state = s.state.out

            if curr_state == s.STATE_IDLE:
                s.req_rdy = Bits1(1)
                s.resp_val = Bits1(0)

                s.remainder_mux_sel = Bits2(R_MUX_SEL_IN)
                s.remainder_reg_en = Bits1(1)

                s.quotient_mux_sel = Bits2(Q_MUX_SEL_0)
                s.quotient_reg_en = Bits1(1)

                s.divisor_mux_sel = Bits1(D_MUX_SEL_IN)

            elif curr_state == s.STATE_DONE:
                s.req_rdy = Bits1(0)
                s.resp_val = Bits1(1)

                s.quotient_mux_sel = Bits2(Q_MUX_SEL_0)
                s.quotient_reg_en = Bits1(0)

                s.remainder_mux_sel = Bits2(R_MUX_SEL_IN)
                s.remainder_reg_en = Bits1(0)

                s.divisor_mux_sel = Bits1(D_MUX_SEL_IN)

            else:  # calculating
                s.req_rdy = Bits1(0)
                s.resp_val = Bits1(0)

                s.remainder_reg_en = ~(s.sub_negative1 & s.sub_negative2)
                if s.sub_negative2:
                    s.remainder_mux_sel = Bits2(R_MUX_SEL_SUB1)
                else:
                    s.remainder_mux_sel = Bits2(R_MUX_SEL_SUB2)

                s.quotient_reg_en = Bits1(1)
                s.quotient_mux_sel = Bits1(Q_MUX_SEL_LSH)

                s.divisor_mux_sel = Bits1(D_MUX_SEL_RSH)