Consider the following procedure written in C:

void scale$\_$by$\_7($int arr$[])\{$

int i;

for$($i $=0$; i $<100$; i$++)$

arr$[$i$]*=7$;

$\}$

This code has been compiled with $-$O$2$ to get the following x$86$ code:

xorl $\%$eax, $\%$eax

$.$L$3$:

movl $(\%$rdi,$\%$rax$),\%$ecx

leal $0(,\%$rcx$,8),\%$edx #$3$rd

subl $\%$ecx, $\%$edx #$4$th

movl $\%$edx, $(\%$rdi,$\%$rax$)$

addq $$4,\%$rax

cmpq $$400,\%$rax

jne $.$L$3$

ret

What optimization has the compiler performed with the third and fourth assembly instructions?

code motion

strength reduction

common subexpression elimination