eltsai / 32bit-MIPS-CPU

MIPS CPU implemented in Verilog

32bit-MIPS-CPU

---

MIPS CPU implemented in Verilog

实现

26条指令

• add、addi & addu
     

   

• sub
     
• subu
     
• and & andi
    
• or & ori
      
• nor
      
• xor
     
• bgtz & bgtqz
     
• bne
     
• j & jal
     
• jr
     
• lw
      
• sw
      
  slv
      
  等......

31     26 25     21 20    16       10    6  4     0 | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |

结构处理

1. 流水线的“依赖”及其处理：
   数据相关：
   编译技术：插入nop
        forwarding技术
2. 分支预测： 动态预测
   当连续错误两次，认为预测失败，清洗流水线。

3. 寄存器结构

4. 控制信号

测试代码：
.data result: .word 0 : 20

  .text
  la $t0, result
  xor $t1, $t1, $t1
  xor $t2, $t2, $t2
  xor $t3, $t3, $t3
  xor $t4, $t4, $t4
  addi $t1, $t1, 1
  addi $t2, $t2, 2
  addi $t3, $t3, -1

  #test lw              0
  lw    $t4, 0($t0)
  addi  $t0, $t0, 4

  #test add addi        0
  add   $t5, $t1, $t3
  sw    $t5, 0($t0)
  addi  $t0, $t0, 4

  #test addu            3
  add   $t5, $t1, $t2
  sw    $t5, 0($t0)
  addi  $t0, $t0, 4

  #test sub             2
  sub   $t5, $t1, $t3
  sw    $t5, 0($t0)
  addi  $t0, $t0, 4

  #test subu            1
  subu  $t5, $t2, $t1
  sw    $t5, 0($t0)
  addi  $t0, $t0, 4

  #test and             1
  and   $t5, $t1, $t3
  sw    $t5, 0($t0)
  addi  $t0, $t0, 4

  #test andi            16
  andi  $t5, $t3, 16
  sw    $t5, 0($t0)
  addi  $t0, $t0, 4

  #test or              3
  or    $t5, $t1, $t2
  sw    $t5, 0($t0)
  addi  $t0, $t0, 4

  #test nor             0
  nor   $t5, $t3, $t1
  sw    $t5, 0($t0)
  addi  $t0, $t0, 4

  #test ori
  ori 	$t5, $t2, 4
  sw $t5, 0($t0)
  addi  $t0, $t0, 4
  #test xori
  xori 	$5, $2, 4
  sw $5, 0($t0)
  addi  $t0, $t0, 4



 xor $t5, $t5, $t5
 addi $t5, $t5, 1
 
 #test sll
 sll $t5, $t5, 1
 sw    $t5, 0($t0)
 addi  $t0, $t0, 4
 
 #test srl
 srl $t5, $t5, 1
 sw    $t5, 0($t0)
 addi  $t0, $t0, 4
 
 #test sllv
 sllv $t5, $t5, $t2
 sw    $t5, 0($t0)
 addi  $t0, $t0, 4
 
 #test srlv
 srlv $t5, $t5, $t2
 sw    $t5, 0($t0)
 addi  $t0, $t0, 4


 #test bltz bgez beq blez

 #test xor             -2
  xor   $t5, $t3, $t1
  sw    $t5, 0($t0)
  addi  $t0, $t0, 4
  #test bgtz j          1

LABLE: addi $t5, $t5, 1 bgtz $t5, FINISH j LABLE FINISH: sw $t5, 0($t0) addi $t0, $t0, 4

  #test bnz              0
  bne   $t5, $t1, NE     #[t5] = 1
  xor   $t5, $t5, $t5

NE: sw $t5, 0($t0) addi $t0, $t0, 4

  #test jr                8
  la    $t6, TJR
  xor   $t5, $t5, $t5
  jr    $t6
  addi  $t5, $t5, 16

TJR: addi $t5, $t5, 8 sw $t5, 0($t0) addi $t0, $t0, 4

  #test forword * 2       2
  la    $t6, result
  lw    $t1, 0($t6)       # $t1 = [$t0] = 0
  add   $t1, $t2, $t3
  add   $t3, $t1, $t1
  sw    $t3, 0($t0)