Arithmetic.md

Basic Arithmetic in LLVM IR

In LLVM IR we can do basic arithmetic if we use instructions like mul, add, sub and sdiv , udiv, fdiv.

Addition (+)

%variable = alloca i64, align 4
store i64 10, ptr %variable, align 4
%value = load i64, ptr %variable, align 4

%resultado = add i64 %value, i64 10 ; 10 + 10 = 20
store i64 %resultado, ptr %variable, align 4

Multiplication (*)

%variable = alloca i64, align 4
store i64 10, ptr %variable, align 4
%value = load i64, ptr %variable, align 4

%resultado = mul i64 %value, i64 10 ; 10 x 10 = 100
store i64 %resultado, ptr %variable, align 4

Subtraction (-)

%variable = alloca i64, align 4
store i64 10, ptr %variable, align 4
%value = load i64, ptr %variable, align 4

%resultado = sub i64 %value, i64 10 ; 10 - 10 = 0
store i64 %resultado, ptr %variable, align 4

Signed Integer Division (-N) (/)

%variable = alloca i64, align 4
store i64 10, ptr %variable, align 4
%value = load i64, ptr %variable, align 4

%resultado = sdiv i64 %value, i64 10 ; 10 / 10 = 0
store i64 %resultado, ptr %variable, align 4

Unsigned Integer Division (N) (/)

%variable = alloca i64, align 4
store i64 10, ptr %variable, align 4
%value = load i64, ptr %variable, align 4

%resultado = udiv i64 %value, i64 10 ; 10 / 10 = 0
store i64 %resultado, ptr %variable, align 4

Floating-Point Division (N.N) (/)

%variable = alloca float, align 4
store float 10.0, ptr %variable, align 4
%value = load float, ptr %variable, align 4

%resultado = fdiv float %value, float 10.0 ; 10.0 / 10.0 = 0
store float %resultado, ptr %variable, align 4

Each of these instructions allows us to execute simple mathematical operations in a simple and orderly manner, in cascade, similar to assembler in some variants.