CS2.201 - Computer Systems Organization | CSO Lecture 8 notes

Arithmetic Operations:

Suffix l needs to be added to the following instructions

• Unary:
  • inc d C equiv: d++;
  • dec d C equiv: d--;
  • neg d C equiv: d = -d;
  • not d this is bitwise NOT, not to be confused with logical NOT, flips each bit individually, does not effect the FLAGS. C equiv: d = ~d;
• Binary:
  • add s, d C equiv: d = d + s;
  • sub s, d C equiv: d = d + s;
  • imul s, d C equiv: d = d * s;
  • xor s, d C equiv: d = d^s;
  • or s, d C equiv: d = d | s;
  • and s, d C equiv: d = d & s;
• Shift Operations
  • sal k d arithmetic shift left the bits of d by k
  • shl k d exactly same as sal created only because there were arithmetic and logical analogs version of shift right
  • shr k d logical right shift: The MSB in the original number might be 0 or 1, irrespective, 0s are pushed in from the MSB side.
    • sar k d Arithmetic right shift: If the MSB in the original number is 0, then 0 is pushed in, if MSB is 1, then 1 is pushed in.
  • in these instructions, k can take the value
    • of a constant (aka immediate) e.g. $0x4 or
    • from a register e.g %eax or
    • from the one-byte register %cl (the lowest significant byte of %ecx)

Multiplication with overflows

• for signed “full” multiply: imull s one of the multiplicand is s, the other will be %eax. After multiplication, the most significant 32 bits will be stored in %edx and the lower 32 significant bits in %eax. Notation: $\textbf{R[}$ %edx $\textbf{]:R[}$ %eax $\textbf{]} \leftarrow \textbf{S} \times\textbf{R[}$ %eax $\textbf{]}$
• Similarly for unsigned full multiply: mull S. Does the same thing.
• S can be a constant (aka immediate) or from a register e.g. %ebx or 8(%ecx) etc.

Division and cltd

No args. convert signed long to signed double long. (basically a quad). Will take the value stored in %eax and signextend it, and store the most significant 32 bits in %edx and least significant 32 bits in %eax. Signextending means setting all bits in %edx to the sign bit of %eax

Notation: $\textbf{R[\%edx]:R[\%eax]} \leftarrow signextend(\textbf{R[\%eax])}$

• Division instructions follow the idea, where the quotient is stored separately, in %eax and remainder in %edx The arg S that is passed, is the divisor. the dividend is the number made from $\textbf{R[\%edx]:R[\%eax]}$. A usual 32 bit number can be made to be 64 bit big moving into %eax then calling cltd
• idivl S for signed divide, functioning as above
• divl S for unsigned divide, same functioning as above.