1 / 10

CS170 Computer Organization and Architecture I

CS170 Computer Organization and Architecture I. Ayman Abdel-Hamid Department of Computer Science Old Dominion University Lecture 15: 10/24/2002. Outline. Loops Should cover section 3.5. A Loop with a variable array index. Loop: g = g + A[i]; i = i + j; If (i != h) go to Loop

katelin-vaddock
Télécharger la présentation

CS170 Computer Organization and Architecture I

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. CS170 Computer Organization and Architecture I Ayman Abdel-Hamid Department of Computer Science Old Dominion University Lecture 15: 10/24/2002 CS170 Fall 2002

  2. Outline • Loops • Should cover section 3.5 CS170 Fall 2002

  3. A Loop with a variable array index • Loop: g = g + A[i]; • i = i + j; • If (i != h) go to Loop • Variables g, h, i, and j associated with $s1  $s4 • Array base is in $s5. What is the MIPS assembly code? • Loop: add $t1, $s3, $s3 # $t1  2 *i • add $t1, $t1, $t1 # $t1  4 *i • add $t1, $t1, $s5 # $t1  address of A[i] • lw $t0, 0($t1) # $t0  A[i] • add $s1, $s1, $t0 # g  g + A[i] • add $s3, $s3, $s4 # i = i+ j • bne $s3, $s2, Loop # go to Loop if i != h CS170 Fall 2002

  4. Save[i] == k? No Yes i = i +j Exit: A While Loop1/2 while (save[i] == k) i = i + j; i, j, and k associated with $s3  $s5. Base of array save is in $s6. What is MIPS Assembly code? • Loop: add $t1, $s3, $s3 # $t1  2 *i • add $t1, $t1, $t1 # $t1  4 *i • add $t1, $t1, $s6 # $t1  address of save[i] • lw $t0, 0($t1) # $t0  save[i] • bne $t0, $s5, Exit # go to Exit if save [i] != k • add $s3, $s3, $s4 #i = i + j • j Loop # go to Loop • Exit: CS170 Fall 2002

  5. A While Loop2/2 • Solve problem 3.9 (see handout) • Previous assembly code in while example uses a conditional branch and an unconditional jump each time through the loop, which is inefficient • Optimize the code so that it uses at most one branch or jump each time through the loop CS170 Fall 2002

  6. A less than test if ( a < b) go to Less a, b associated with $s0, and $s1, respectively. What is the MIPS assembly code? • To test for less than we introduce a new instruction • slt $t0, $s0, $s1 # slt is set on less than • Register $t0 is set to 1 if the value in register $s0 is less than the value in register $s1, otherwise, register $t0 is set to 0 • Register $zero always contains zero • Comparing $t0 to $zero gives us the effect of branching if a less than b • Combining slt with bne implements branch on less than slt $t0, $s0, $s1 # $t0  1 if a < b, otherwise $t0  0 bne $t0, $zero, Less # go to Less if $t0 != 0 …… Less: Why not provide one instruction that performs branch on less than? CS170 Fall 2002

  7. Case/Switch Statement1/3 • Switch statement allows the programmer to select one of many alternatives depending on a single value • Can be implemented as a series of if –then-else statements (not efficient) • Alternatively (or more efficiently) • First instruction within each case sequence is assigned a label (the label holds the address of the first instruction in the sequence) • These addresses are stored in a jump table in memory • Use new instruction • jr register (jump register) • Unconditional jump to the address specified in register • Program loads appropriate entry from jump table into a register, then jump to proper address using a jump register CS170 Fall 2002

  8. Jump table stored in memory L3 1036 1032 L2 1028 $t4 L1 1024 1024 L0 Case/Switch Statement2/3 • switch (k) { • case 0: f = i + j; break; • case 1: f = g + h; break; • case 2: f = g - h; break; • case 3: f = i – j; break; • } • Variables f through k associated with $s0  $s5. Register $t2 contains 4. • Register $t4 contains address of jump table in memory. MIPS Assembly code? Address of instruction f = i + j  label L0 Address of instruction f = g+ h  label L1 Address of instruction f = g - h  label L2 Address of instruction f = i - j  label L3 CS170 Fall 2002

  9. Case/Switch Statement3/3 slt $t3, $s5, $zero # test if k < 0 bne $t3, $zero, Exit # if k < 0 go to Exit slt $t3,$s5,$t2 # test if k < 4 beq $t3, $zero, Exit # if k >= 4 go to Exit add $t1,$s5,$s5 # $t1  2 *k (Why?) add $t1,$t1,$t1 # $t1  4 *k add $t1,$t1,$t4 # $t1  address of JumpTable[k] lw $t0,0($t1) # $t0  JumpTable[k] jr $t0 # jump to address found in register $t0 L0: add $s0, $s3, $s4 # (k = 0) then f = i +j j Exit # break (go to Exit) L1: add $s0,$s1,$s2 #(k = 1) then f = g+h j Exit # break (go to Exit) L2: sub $s0,$s1,$s2 #(k=2) then f = g-h j Exit L3: sub $s0,$s3,$s4 #(k=3) then f = i- j Exit: # End of switch statement CS170 Fall 2002

  10. What we know so far • Fig. 3.9 page 131 • instruction format for unconditional jump j (J-format) • <opcode (6 bits), target address (26 bits)> (more on target address in section 3.8) • The jr instruction  (R-format) • opcode = 0, funct = 8, rs is jump register • The slt instruction  (R-format) • Opcode = 0, funct = 42, rs, rd, rt • The beq, bne instructions  (I-format) • $zero always contains the value 0 • The instruction beq $s1, $s2, 100 is an I-format instruction. The resulting machine code • Op| rs | rt | address • 4 | 17 | 18 | 25 • Why the value in address field is 25 and not 100? Will discuss this when we cover section 3.8 CS170 Fall 2002

More Related