The following code performs matrix multiplication in c$++.$

float dot$\_$product$\_$f$($float$*$ m$1,$ float$*$ m$2,$ int r$,$ int c$,$ int l$)$

$\{$

float product $=0$;

for$($int i$=0$; i $<$ l; i$++)$

$\{$

float a $=$ m$1[$i $+$ r $*$ l$]$;

float b $=$ m$2[$c $+$ i $*$ l$]$;

product $+=$ a $*$b;

$//$printf$("\%1$f$*\%1$f $+",$ a$,$ b$)$;

$\}$

$//$printf$("$

$")$;

return product;

$\}$

The following is the assembly for the above c$++$ code. The requirement is to remove redundant instructions to improve performance. Please provide new working assembly file with improved code. Thanks.

$.$text

$.$file $"$dp$\_$float.cpp$"$

$.$globl $\_$Z$13$dot$\_$product$\_$fPfS$\_$iii # $--$ Begin function $\_$Z$13$dot$\_$product$\_$fPfS$\_$iii

$.$p$2$align $4,0$x$90$

$.$type $\_$Z$13$dot$\_$product$\_$fPfS$\_$iii,@function

$\_$Z$13$dot$\_$product$\_$fPfS$\_$iii: # @$\_$Z$13$dot$\_$product$\_$fPfS$\_$iii

$.$cfi$\_$startproc

# $\%$bb$\mathrm{.}0$:

pushq $\%$rbp

$.$cfi$\_$def$\_$cfa$\_$offset $16$

$.$cfi$\_$offset $\%$rbp$,-16$

movq $\%$rsp$,\%$rbp

$.$cfi$\_$def$\_$cfa$\_$register $\%$rbp

movq $\%$rdi, $-8(\%$rbp$)$

movq $\%$rsi, $-16(\%$rbp$)$

movl $\%$edx, $-20(\%$rbp$)$

movl $\%$ecx, $-24(\%$rbp$)$

movl $\%$r$8$d$,-28(\%$rbp$)$

xorps $\%$xmm$0,\%$xmm$0$

movss $\%$xmm$0,-32(\%$rbp$)$

movl $$0,-36(\%$rbp$)$

$.$LBB$0\_1$: # $=>$This Inner Loop Header: Depth$=1$

movl $-36(\%$rbp$),\%$eax

cmpl $-28(\%$rbp$),\%$eax

jge $.$LBB$0\_4$

# $\%$bb$\mathrm{.}2$: # in Loop: Header$=$BB$0\_1$ Depth$=1$

movq $-8(\%$rbp$),\%$rax

movl $-36(\%$rbp$),\%$ecx

movl $-20(\%$rbp$),\%$edx

imull $-28(\%$rbp$),\%$edx

addl $\%$edx, $\%$ecx

movslq $\%$ecx, $\%$rcx

movss $(\%$rax,$\%$rcx$,4),\%$xmm$0$ # xmm$0=$ mem$[0],$zero,zero,zero

movss $\%$xmm$0,-40(\%$rbp$)$

movq $-16(\%$rbp$),\%$rax

movl $-24(\%$rbp$),\%$ecx

movl $-36(\%$rbp$),\%$edx

imull $-28(\%$rbp$),\%$edx

addl $\%$edx, $\%$ecx

movslq $\%$ecx, $\%$rcx

movss $(\%$rax,$\%$rcx$,4),\%$xmm$0$ # xmm$0=$ mem$[0],$zero,zero,zero

movss $\%$xmm$0,-44(\%$rbp$)$

movss $-40(\%$rbp$),\%$xmm$0$ # xmm$0=$ mem$[0],$zero,zero,zero

movss $-44(\%$rbp$),\%$xmm$2$ # xmm$2=$ mem$[0],$zero,zero,zero

movss $-32(\%$rbp$),\%$xmm$1$ # xmm$1=$ mem$[0],$zero,zero,zero

mulss $\%$xmm$2,\%$xmm$0$

addss $\%$xmm$1,\%$xmm$0$

movss $\%$xmm$0,-32(\%$rbp$)$

# $\%$bb$\mathrm{.}3$: # in Loop: Header$=$BB$0\_1$ Depth$=1$

movl $-36(\%$rbp$),\%$eax

addl $$1,\%$eax

movl $\%$eax, $-36(\%$rbp$)$

jmp $.$LBB$0\_1$

$.$LBB$0\_4$:

movss $-32(\%$rbp$),\%$xmm$0$ # xmm$0=$ mem$[0],$zero,zero,zero

popq $\%$rbp

$.$cfi$\_$def$\_$cfa $\%$rsp$,8$

retq

$.$Lfunc$\_$end$0$:

$.$size $\_$Z$13$dot$\_$product$\_$fPfS$\_$iii, $.$Lfunc$\_$end$0-\_$Z$13$dot$\_$product$\_$fPfS$\_$iii

$.$cfi$\_$endproc

# $--$ End function

$.$ident "clang version $17\mathrm{.}0\mathrm{.}6($CentOS $17\mathrm{.}0\mathrm{.}6-5.$el$9)"$

$.$section $".$note.GNU$-$stack","",@progbits

$.$addrsig