Menu

My notebook

2003-10-03
2003-10-14
  • Michael Chen

    Michael Chen - 2003-10-03

    Goal finish design doc tonight by 8 pm

    If possible, finish controller. 

     
    • Michael Chen

      Michael Chen - 2003-10-03

      I still need to look at controller to double check

      aluControl is done
      design doc is done

       
      • Michael Chen

        Michael Chen - 2003-10-10

        working with the group to put everything together

        we are behind the schedule

        groups need to read the design doc and follow the design doc

        correct the design doc to fix any errors in the input

         
    • Michael Chen

      Michael Chen - 2003-10-03

      shifter is not compilable
      Jason, take a look at >>>

      datapath, I need to go over the datapath with Brian, or we can just let the tester verify the datapath works. 

       
    • Michael Chen

      Michael Chen - 2003-10-06

      I just fixed a bug in aluControl.v about the shift signal. 

       
    • Michael Chen

      Michael Chen - 2003-10-07

      I found a bug in sltControl on sunday.  fixed it with Jason.  Need to update design doc, that sltControl now takes in cout as an input

      Brandan found a bug in controller, need to fixed xor and update design doc. 

       
    • Michael Chen

      Michael Chen - 2003-10-08

      I found another bug in the data path, we should be calculating the address of Pc+4 + immed, and pc +8.  Since we pipeline the pc, we are seq to pc+8 for regular address. 

      The ori bug is fixed in the controller. 

      we can use luts to implement a shifter, but the CAD automatically does that for us. 

      a mux 4 is faster because of dedicated logic

       
    • Michael Chen

      Michael Chen - 2003-10-10

      We can't use monitor to verify the cpu unless we know monitor has been tested and function correctly.

       
    • Michael Chen

      Michael Chen - 2003-10-11

      some people are still not reading or following the design doc.

      the design doc keeps everyone on the same sheet of music

      it's confusing if everyone name their wires and signals not following the naming convention 

      the initial plan of testing without dataforwarding and hazardDetect is slowing us down considerably

      so far we only found a lot of wiring errors and bugs, no logical bugs or controoler bugs yet. 

      we need to test stall and beq really good

       
    • Michael Chen

      Michael Chen - 2003-10-11

      we need to stall if
      lw $5, 0($6)
      sw $3, 4($5)

       
    • Michael Chen

      Michael Chen - 2003-10-13

      started working at 8:30 pm, helping the tester to do integration testing

      wrote test cases to test lw after lw and lw

      fixed monitor.v to fixed the pipe effect

      wrote test caes to test break
      fixed many errors in monitor.v

      modified testBasicCpu.s to output the to io

      wrote test cases to break

       
    • Michael Chen

      Michael Chen - 2003-10-14

      update on today's work back log

      started working at 1 pm today

      fixed the bug in monitor, fixed the jr after lw bug in hazardDetect, it now stalls the pipeline, 
      need to fix dataForwarding to get rid of the 2 extra signals.

      fixed the bug the corner.s in simulatation
      there is still a bug in corner.s on the board

       

Log in to post a comment.

MongoDB Logo MongoDB