Group Project 2 - Pipelined Mini-MIPS

CS 3410 Spring 2018


Schedule design doc meeting by: 11:59pm, Sunday, February 25th, 2018

Preliminary design documentation Due: 11:59pm, Thursday, March 1st, 2018

Project Due: 11:59pm, Tuesday, March 13th, 2018

Circuit Naming: Your top-level circuit must be named "MIPS32" (case-sensitive).

Late Policy: Slip days can be used for the final submission but not for the design doc submission. If a slip day is used, it will be used for both partners.

You must work with the same partner for this project and the next. You can change partners for the later projects.

Warning: Read the ENTIRE writeup before you begin.


Overview

In this and the next project we will implement a subset of the MIPS32 architecture in Logisim. We will ignore more advanced features such as the MIPS coprocessor instructions and traps/exceptions for these assignments.

In this project you will implement a functioning outline of the pipelined processor for a small set of instructions, including: decoding all the instructions you will encounter in this and the next project, implementing most of the MIPS pipeline, correct implementation of arithmetic and logic operations, and simple hazard avoidance for these instructions. In the next project you will implement a more complete set of instructions, and add more complete hazard logic.

You will also be expected to use Git throughout this project to maintain version control of your circuit and to ensure that you have an online backup available. At the end of the assignment, you will turn in a log of your git commit history in a text file. There are no specific restrictions on how often you should be committing, but it is highly recommended that you commit often and write meaningful commit messages so that you can restore old versions in case of failure. We will expect your commit log to show you made a genuine effort to accomplish this.

Academic Integrity. As one of the most widely studied architectures, MIPS has a wealth of information available on the web and in textbooks. You may consult any MIPS documentation available to you in order to learn about the instruction set, what each instruction does, etc.; however, we expect your design to be entirely your own, and your submission should cite any significant sources of information you used. If you are unsure if it is okay to borrow from some other source, just ask the TAs. If you are hesitant to ask the TAs or to cite the source, then it is probably not okay. Plagiarism in any form will not be tolerated. It is also your responsibility to make sure your sources match the material we describe here (warning: the top Google hit for "MIPS reference" contains several inaccuracies).

What to Submit

Note the different due dates.

    By Sunday, February 25th, 2018
  • Schedule a meeting with a TA to present your intended design.
    By Thursday, March 1st, 2018
  • Meet with your TA at your scheduled time to present your intended design.
  • Submit your preliminary design document.
    By Tuesday, March 13th, 2018
  • A single Logisim project file containing your processor and all needed subcomponents.
  • A text file containing your well-commented MIPS assembly test program.
  • A updated design document of your processor.
  • An assembly program containing your Fibonacci implementation.
  • A git commit history obtained using the command git log on your repository.

Fibonacci

In this part of the project you will write a function in assembly in order to test the processor that you will build.

A Fibonacci sequence is characterized by the fact that every number after the first two is the sum of the two preceding ones: 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, ... The sequence F of Fibonacci numbers is defined by the recurrence relation: F(n) = F(n-1) + F(n-2), with seed values F(0)=0 and F(1)=1. You will implement the Fibonacci function, which compute the n-th number, F(n), in the Fibonacci sequence. For instance, Fibonacci(12) returns 144. because the sequence starting at F(0)=0 and F(1)=1, and goes by 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, ...

There are several ways to compute the n-th Fibonacci number. For this assignment, you will implement a (naive, but counterintuitive) iterative version. Later, we will ask you to implement a recursive version.

  • Iterative Fibonacci:

        int i_Fibonacci(int n) {
            int f1 = 0;
            int f2 = 1;
            int fi;
    
            if(n == 0) {
              return 0;
            }
    
            if(n == 1) {
              return 1;
            }
    
            for(int i = 2; i <= n; i++) {
                fi = f1 + f2;
                f1 = f2;
                f2 = fi;
            }
            return fi;
        }

You will implement this iterative version of the Fibonacci using MIPS assembly code.
Notice that you don't have any control instructions in the Mini-MIPS processor, which will make expressing the for loop and if statements impossible. To work around this, we will simply hardcode n to a particular value and unroll the loop, simulating the arithmetic that would have happened if we had control instructions.

  • As an example, this is what the code looks like if we hardcode n to 3:

        int f1 = 0;
        int f2 = 1;
        int fi;
        fi = f1 + f2;
        f1 = f2;
        f2 = fi;
        fi = f1 + f2;
        f1 = f2;
        f2 = fi;
        return fi;

For your implementation of the unrolled version of Fibonacci(4), store f1 in register r8, also known as $t0, store f2 in register r9, also known as $t1, and store fi in register r10, also known as $t2. Compute the 4th Fibonacci number. At the end of your function, instead of returning fi, simply store the value fi in register r2, also known as $v0.

Note: if you just store the return value of Fibonacci(4) into r2, that completely defeats the purpose of the exercise. Submissions that do this will receive no credit. Do not optimize/modify the code beyond removing the control instructions. You may only use the instructions in Table A.

MIPS Pipeline

You will implement a five-stage MIPS pipeline, which is the most common organization for MIPS and is similar to what is described in the book and in class:

  1. Fetch
  2. Decode
  3. Execute
  4. Memory
  5. Writeback

The MIPS standard, volume I, section 2.6, describe several possible MIPS pipelines, none of which exactly match the five-stage pipeline that we will implement.

Memory stage. For this project, the memory stage of the pipeline will just be a passthrough doing nothing. In the next project you will implement load and store memory operations and populate the memory stage part of the pipeline.

Delay slot. In this assignment you will not implement the branch/jump delay slot or memory load delay slot. That will be added in the next project.

What to Implement

Implement a five-stage pipelined MIPS processor in Logisim. Your design should contain a program counter, a read-only program memory (ROM), a register file, our ALU, and any other components needed, along with the instruction decode and control circuits needed to connect them all together. The pipeline should: fetch instructions to execute from the program ROM and increment the program counter by 4; decode each instruction; select arguments from the register file; compute results; do nothing in the memory stage; and store results back in the register file.

Your pipeline must correctly execute all of the instructions in Table A:

Table A
Immediate arithmetic ADDIU, ANDI, ORI, XORI, SLTI, SLTIU
Register arithmetic ADDU, SUBU, AND, OR, XOR, NOR, SLT, SLTU
Move MOVN, MOVZ
Shifts (constant and variable) SLL, SRL, SRA, SLLV, SRLV, SRAV
Immediate load LUI

Note that LUI doesn't access memory and so, despite its name, it is more similar to an arithmetic or logic instruction than a memory load instruction.

Table B contains instructions which you will be implementing in the next project. These instructions should be nonfunctional in this project, so your instruction decoding and control logic must be able to recognize table B instructions, and ensure that they have no effect on the program execution. Since you will be implementing the functionality of these instructions in the near future, decoding them properly now will save you from having to completely re-implement your decoding and control logic later.

Table B
Jumps (with one delay slot) J, JR, JAL, JALR
Branches (with one delay slot) BEQ, BNE, BLEZ, BGTZ, BLTZ, BGEZ
Memory Load/Store (little endian) LW, LB, LBU, SW, SB

Our testing programs for this project will include a mixture of instructions from both Table A and Table B. When processing one of the instructions of Table B, your implementation is free to do anything as long as: (1) no value in the register file changes and (2) the execution of instructions from Table A are not interfered with in any way. Think carefully about your decode logic to make sure this second condition works out.

Deviation from the MIPS standard: You can ignore any MIPS instruction or feature not mentioned in this document, such as traps, exceptions, and system calls. You can also assume that your processor will never encounter anything but legal instructions from the lists above.

Refer to the MIPS manual, Volume 2, linked on the course web site for a full specification of what each of these operations does. Except where noted in here, the MIPS manual is the authoritative specification for this project: information you find elsewhere (e.g. Wikipedia or the book) doesn't count if it contradicts the MIPS manual.

Data Hazard: The trivial solution to data hazards is, of course, stalling. However, the performance of trivial solutions is not optimal. Implement forwarding such that your implementation will handle data hazards without sacrificing performance.

Processor Components

Build your MIPS processor as a single Logisim circuit file. Do not use Logisim's Project > Load Library > Logisim Library command to import circuits from other Logisim files.

Your top-level circuit must be named either "MIPS" or "MIPS32" (case-sensitive).

The "CS3410 Components" library folder contains several components you will need, such as a register file, an ALU, an incrementer, and a program memory. You must use these components where applicable, but you may only use one of each. Also note that you should not nest these components too deeply (inside one subcircuit is fine). Documentation and instructions for cs3410 components can be found here.

Register file. Don't attempt to create your own register file out of regular Logisim Registers. This one makes debugging and simulation much easier.

ALU. You must use the ALU from this library, rather than your ALU from project 1. Your design may only contain one ALU.

MIPS Program ROM. We will implement a Harvard Architecture design: The Program ROM component will store a read-only copy of the program instructions, and (in the next project) you will use a separate Logisim RAM to store data. The Program ROM component can load and display MIPS assembly language, which helps with debugging and saves you from having to write test programs in machine language.

Incrementer. You must use one (and only one) of these to increment the PC. Note that the incrementer only computes +1, whereas the PC increments by 4.

Logisim components. You can additionally use any of the components that come with Logisim, such as a Register for the PC, multiplexers, and so on.

Testing

Write a test program file for Logisim containing MIPS assembly that fully tests all aspects of your processor; it should be able to be run by loading it into the Program ROM. The test program should demonstrate that the processor's behavior conforms to all specifications given here and in the MIPS reference manual.
This is a critical step, and you should spend a fair amount of time developing a comprehensive test program that exercises all of the different instructions and features of your processor. The program should be well commented , indicating what it is doing and what results should be expected when running the tests, so that the course staff is convinced of the correctness of your processor. All test cases should be included in a single file, and the quality and thoroughness of the tests will be graded.

Since the MIPS processor does not use input and output pins in the same way as the ALU, you will not be able to test your cicuit as a whole using test vectors. You must use your MIPS test program to evaluate the correctness of your processor. However, test vectors may be useful for debugging some of your subcircuits.

Don't forget to include every possible instruction in your testing program. This includes table B instructions; you must ensure that your processor does nothing when executing any table B instruction, as required.

Documentation

The design document should include a block diagram showing all the major components in the processor (ALU, Register File, PC, pipeline registers, etc.), and the datapath connections between them. You need not completely draw wires for control logic signals, but should indicate which components take control inputs, and give names to all control signals. Overall, the diagram should be very similar to (and not much more complicated than) the diagrams used in lecture, but with labels for control signals and any relevant data signals. You should provide an explanation for any parts of your processor or any subcomponents that are unusual or potentially confusing.

Also include a description of your control and instruction decoding logic. For each control logic signal (or group of related control logic signals) you should provide (a) a brief description of what the signal does, e.g. what the values of the control signal mean; and (b) a truth table showing what value the signal takes for each possible opcode.

While the handling of the instructions in Table B will not be graded in the design document, it is recommended to start thinking about how you will handle the execution of these instructions. The design document meeting is your chance to ask a TA any questions you may have about this and the next project (where you will implement the rest of the instructions).

You will need to submit a preliminary design document by 11:59pm, Thursday, March 1st, 2017 via CMS. You also need to schedule a meeting with one of the TAs to discuss about your intended design. The time schedule is available on CMS. Pick one that works for you and your partner. During the meeting, you will need to bring your printed design document, and be prepared to present your design to TA.

Update your design document as you work on your circuit. Documentation provides you another avenue of communicating your intended design and implementation in the case your circuit has mistakes. For a well-designed, clearly labeled and nicely laid out, completely correct solution, a screenshot of the Logisim circuit itself can serve as a diagram, but, otherwise, use the documentation to clarify and explain any parts of your circuit we might find confusing.

We have provided a design doc guideline and template in your Github repositories to help you with writing the design documentation. The tex source file is also available.

Design Doc Meetings

Each team will have a 20 minute design doc meeting with a member of course staff between Febrary 25th and February 28th.

How much of my design doc should be completed before the meeting?

In general, the more of the design doc you have completed before the meeting, the more helpful the meeting is. That being said, it is more than ok to have questions for your TA about the specifics of your design.

At the very least, you should come to the design doc meeting with a preliminary plan for implementing pipelining, a diagram outlining the 5 stages of your processor as well as a basic description of each, a plan for your decode and control logic, and a general plan for testing.

Design Doc Submission

You will submit a preliminary design document to CMS by March 1st. This design doc should take into account all of the feedback you received during your design doc meeting as well as any further progress you have made on your processor. This design doc will be graded by the member of course staff you met with. While we realize this design doc may not be entirely complete, we expect it to demonstrate a strong understanding of the project requirements.

Help and Hints

Help. Ask the instructor, TAs and consultants for help. We also expect to see most students in office hours during the course of the project. Extra hours will be scheduled as needed. Also check Piazza for more help.

Bugs. If you suspect a bug in Logisim or the cs3410 library, ask the course staff for help.

Really, it shows. Do a pencil-and-paper design before implementing in Logisim. We will penalize poorly laid-out designs that are hard to understand.

Read the docs. Refer to the components guide, the design guidelines, the MIPS cheat sheet, and the MIPS manual often.

What to do first. Don't wait to get started. However, because we are still covering pipelining and pipeline hazards in lecture, you probably will not want to start right in on the final design. The following order of operations might be sensible:

  1. Find a partner ASAP. YOU ARE REQUIRED TO WORK WITH A PARTNER FOR THIS PROJECT. You must be paired up with a partner on CMS before scheduling your design doc meeting.
  2. Read the MIPS manual page for every instruction (some have unusual quirks).
  3. Start writing test programs, including code both with and without hazards.
  4. (After pipelining is covered in lecture) Design on paper, build in Logisim, and begin to test a pipelined CPU that doesn't handle pipeline hazards. Write the documentation as you go.
  5. (After pipeline hazards are covered in lecture) Modify the datapath and control logic to handle pipeline hazards.

Project Design Rules and Tips

Ensure that your circuit sticks to the following rules, as well as the Logisim design guidelines to avoid losing points!

cs3410 Components. Do not duplicate any of the provided cs3410 components in your circuit. At most one of each should be used. Note that duplicates of some of the components, such as the register file, will break the autograder, even if placed in an unused subcircuit, so ensure that any extras are removed before submission to avoid losing points.

Clock. You should only have one clock in the entirety of your circuit. Subcircuits requiring a clock signal should use input pins to connect to the processor clock. Your MIPS design should use a rising clock edge to define the boundaries of clock cycles: during the first half of each processor clock cycle the clock is 1; during the second half of each cycle the clock is 0; and the end of the cycle is when clock transitions from 0 to 1. By default, most Logisim memory components (Registers, D Flip-Flops, etc.) are triggered on the rising clock edge, so you can leave them as is. The register file is the only component that may use a falling clock edge, and can be so configured using the attributes panel. In order to avoid read-write hazards, you will either want to change this setting, or negate the clock signal going into the register file (think about why this works).

Comparators, found in the Arithmetic folder in Logisim, may be useful for implementing some of the functionality of the processor. You are allowed to use at most two comparators in your circuit. Points will be deducted if you use more than two comparators.

Demultiplexors should not be needed for this project, and you should avoid using them.

Tunnels, if used at all, should be used sparingly and only when they make your circuit significantly easier to read.

You may find that combinational analysis is useful in designing your instruction decoding logic.

Notes on implementing SLTI and SLTIU: The immediate value in these instructions should be sign-extended. The MIPS handbook describes the operation of SLTIU as sign-extending the immediate value first, and then comparing the immediate to Rs both as unsigned integers. As a result, you're limited to representing the minimum or maximum of the range of representable integers:

  • 0x0000 to 0x7fff turn into 0x00000000 to 0x00007fff: This is 0 to 32767.
  • 0x8000 to 0xffff turn into 0xffff8000 to 0xffffffff: This is maxint-32767 to maxint, where maxint = 2^32 - 1.

For details please refer to MIPS handbook Vol 2, page 204 ~ 205.

Additionally, note that there are some typos in MIPS Reference Vol. 2 pages 196 and 197. For instructions SLTI and SLTIU, each GPR[rd] should be replaced with a GPR[rt] in the explanation in the Operation section.


On Data Hazard Detection

All forwarding logic should be handled in the Execute stage. Data hazards are possible when

1a. EX/MEM.RegisterRd == ID/EX.RegisterRs
1b. EX/MEM.RegisterRD == ID/EX.RegisterRt
2a. MEM/WB.RegisterRd == ID/EX.RegisterRs
2b. MEM/WB.RegisterRd == ID/EX.RegisterRt

To resolve the first two data hazards, we can forward value from the EX/MEM pipeline registers. To resolve the latter two hazards, we can forward value from the MEM/WB pipeline registers. In order to figure out the correct forwarding control logic, we note that forward will happen:

1. only if the forwarding instruction will write to a register
  EX/MEM.Regwrite, MEM/WB.Regwrite
2. AND only if Rd for that instruction is not $zero
  EX/MEM.RegisterRd != 0, MEM/WB.RegisterRd != 0
3. AND only if forwarding instruction is not a load in MEM stage
  EX/MEM.MemRead == 0

From the above, we can derive the following forwarding conditions:

EX Hazard
if (EX/MEM.RegWrite and (EX/MEM.RegisterRd != 0) and (EX/MEM.RegisterRd == ID/EX.RegisterRs)) ForwardA = 10
if (EX/MEM.RegWrite and (EX/MEM.RegisterRd != 0) and (EX/MEM.RegisterRd == ID/EX.RegisterRt)) ForwardB = 10
MEM Hazard
if (MEM/WB.RegWrite and (MEM/WB.RegisterRd != 0) and (MEM/WB.RegisterRd == ID/EX.RegisterRs)) ForwardA = 01
if (MEM/WB.RegWrite and (MEM/WB.RegisterRd != 0) and (MEM/WB.RegisterRd == ID/EX.RegisterRt)) ForwardB = 01

However, if we have a sequence of code that causes both data hazards to happen:

add $1, $1, $2
sub $1, $1, $3
or  $1, $1, $4

then we want to use the most recent result--the result from the sub instruction.
Let's revise the forwarding conditions for MEM hazard to take care of 'double' data hazards:

if (MEM/WB.RegWrite and (MEM/WB.RegisterRd != 0)
  and not (EX/MEM.RegWrite and (EX/MEM.RegisterRd != 0)
    and (EX/MEM.RegisterRd == ID/EX.RegisterRs))
  and (MEM/WB.RegisterRd == ID/EX.RegisterRs))
    ForwardA = 01

if (MEM/WB.RegWrite and (MEM/WB.RegisterRd != 0)
  and not (EX/MEM.RegWrite and (EX/MEM.RegisterRd != 0)
    and (EX/MEM.RegisterRd == ID/EX.RegisterRt))
  and (MEM/WB.RegisterRd == ID/EX.RegisterRt))
    ForwardB = 01

MIPS (subset) Assembly Syntax

The MIPS Program ROM component has a built-in assembler that understands all of the instructions you will implement. The syntax is standard MIPS syntax. Labels are case sensitive, everything else ignores case. Anything following a pound ('#') is a comment. In this project, you will only use a few of the instructions listed here.

The instruction syntax is the same as given in the MIPS standard (and different from the output of gcc and many other tools). Registers are written as $0, $1, ..., $31, and the destination register goes on the left, followed by source registers and immediate values on the right. Most integer arguments (immediates, shift amounts, jump targets) can be specified in hex (i.e. 0x12ab), in decimal (i.e. 1234 or -1234), a label, or the special constant PC. The assembler will replace PC with the address of the instruction itself. Most constants have some restrictions: jump destinations must have the same upper 4 bits as the PC+4, and must be a multiple of 4; branch destinations must be a multiple of 4 and fit in a signed 18 bit immediate; etc. As a special case, when a branch target is specified symbolically as a label, the assembler will automatically subtract the current PC value to obtain a signed offset.

By default, the first instruction will be placed at address 0, and subsequent instructions are placed at at addresses 4, 8, 12, etc.

Assembler directives. The Program ROM assembler understands two standard MIPS assembler directives, .text and .word, both of which take integer (hex or decimal) arguments. For example, .text 0x50000000 will direct the assembler to place subsequent instructions starting at address 0x50000000. And .word 0x24030005 directs the assembler to use the value 0x24030005 as the next machine instruction, which happens to be the machine code for ADDIU $3, $0, 5.

Symbolic register names. The assembler built into the MIPS Program ROM accepts standard MIPS register names: $zero, $at, $v0, $v1, $a0 - $a4, $s0 - $s7, $t0 - $t9, $k0, $k1, $sp, $gp, $fp, and $ra.

Some examples of instructions are:

Immediate Arithmetic ADDIU $12, $0, PC
Register Arithmetic ADDU $13, $0, $20
Immediate Load LUI $14, 0x123
Shifts SLL $13, $13, 2
SLLV $15, $14, $3
Jumps J 0x24
J my_label
JR $5
JALR $31, $5
JALR $5
Branches BEQ $5, $6, -12
BEQ $5, $6, my_loop_top
BLEZ $9, 16
BLEZ $9, my_loop_done
Memory Load/Store LW $12, -4($30)
SW $12, 0($30)

MIPS (subset) Opcode Summary (from the MIPS Handbook)

Items in white are required for this and the next project. Make sure to decode all of them. The grayed out items are not implemented in either project, and you do not need to worry about them.

Table 1: MIPS32 Encoding of the Opcode Field
opcode bits 28..26
bits 31..29 0 1 2 3 4 5 6 7
000 001 010 011 100 101 110 111
0 000 SPECIAL δ REGIMM δ J JAL BEQ BNE BLEZ BGTZ
1 001 ADDI ADDIU SLTI SLTIU ANDI ORI XORI LUI
2 010 COP0 δ COP1 δ COP2 θδ COP3 θδ BEQL φ BNEL φ BLEZL φ BGTZL Φ
3 011 β β β β SPECIAL2 δ JALX ε ε *
4 100 LB LH LWL LW LBU LHU LWR β
5 101 SB SH SWL SW β β SWR CACHE
6 110 LL LWC1 LWC2 θ PREF β LDC1 LDC2 θ β
7 111 SC SWC1 SWC2 θ * β SDC1 SDC2 θ β
Table 2: MIPS32 SPECIAL Opcode Encoding of the Function Field
function bits 2..0
bits 5..3 0 1 2 3 4 5 6 7
000 001 010 011 100 101 110 111
0 000 SLL MOVCI δ SRL SRA SLLV * SRLV SRAV
1 001 JR JALR MOVZ MOVN SYSCALL BREAK * SYNC
2 010 MFHI MTHI MFLO MTLO β * β β
3 011 MULT MULTU DIV DIVU β β β β
4 100 ADD ADDU SUB SUBU AND OR XOR NOR
5 101 * * SLT SLTU β β β β
6 110 TGE TGEU TLT TLTU TEQ * TNE *
7 111 β * β β β * β β
Table 3: MIPS32 SPECIAL REGIMM Encoding of the rt Field
rt bits 18..16
bits 20..19 0 1 2 3 4 5 6 7
000 001 010 011 100 101 110 111
0 00 BLTZ BGEZ BLTZL BGEZL * * * *
1 01 TGEI TGEIU TLTI TLTIU TEQI * TNEI *
2 10 BLTZAL BGETAL BLTZALL BGETALL * * * *
3 11 * * * * * * * *