I'm following llvm's tutorial for their own simple programming language "Kaleidoscope" and there's an obvious functionality in my language which this tutorial doesn't seem to cover. I simply want to print any double to standard output pretty much as C++ would do:
std::cout << 5.0;
my language would do something like
print(5.0);
Third chapter of llvm's tutorial covers function calls. The code they use is:
Value *CallExprAST::codegen() {
// Look up the name in the global module table.
Function *CalleeF = TheModule->getFunction(Callee);
if (!CalleeF)
return ErrorV("Unknown function referenced");
// If argument mismatch error.
if (CalleeF->arg_size() != Args.size())
return ErrorV("Incorrect # arguments passed");
std::vector<Value *> ArgsV;
for (unsigned i = 0, e = Args.size(); i != e; ++i) {
ArgsV.push_back(Args[i]->codegen());
if (!ArgsV.back())
return nullptr;
}
return Builder.CreateCall(CalleeF, ArgsV, "calltmp");
}
How could I implement codegen()
method for specific function call print(any fp number)
?
LLVM currently supports compiling of Ada, C, C++, D, Delphi, Fortran, Haskell, Julia, Objective-C, Rust, and Swift using various front ends. Widespread interest in LLVM has led to several efforts to develop new front ends for a variety of languages.
• nsw – no signed wrap for add, sub, mul, shl. • nuw – no unsigned wrap for add, sub, mul, shl. • exact – no remainder for sdiv, udiv, lshr, ashr.
The LLVM Project is a collection of modular and reusable compiler and toolchain technologies, with a primary focus on these two compilers: The LLVM back-end compiler and core libraries provide a modern source- and target-independent optimizer, along with code generation support for the RHEL CPU architectures.
below is the llvm ir code generated for printf("%f", a); using clang. printf signature is int printf(const char*, ...);
@.str = private unnamed_addr constant [3 x i8] c"%f\00", align 1
; Function Attrs: nounwind uwtable
define i32 @main() #0 {
%a = alloca double, align 8
%1 = load double* %a, align 8
%2 = call i32 (i8*, ...)* @printf(i8* getelementptr inbounds ([3 x i8]* @.str, i32 0, i32 0), double %1)
ret i32 0
}
declare i32 @printf(i8*, ...) #1
to implement in codegen you first need to check if the function is already present in module or not. if not then you need to add the declaration, you can do both in one call.
Function *CalleeF = TheModule->getOrInsertFunction("printf",
FunctionType::get(IntegerType::getInt32Ty(Context), PointerType::get(Type::getInt8Ty(Context), 0), true /* this is var arg func type*/)
);
above will get or add you the handle to function declaration
declare i32 @printf(i8*, ...) #1
then you can call function via matching params.
std::vector<Value *> ArgsV;
for (unsigned i = 0, e = Args.size(); i != e; ++i)
ArgsV.push_back(Args[i]->codegen());
return Builder.CreateCall(CalleeF, ArgsV, "printfCall");
You'd first check if Callee == "print"
and then insert any instructions you want.
LLVM IR has no concept of "printing" since that's not really a language consideration -- it's a facility provided by the OS. Probably the simplest option for you would be to translate the call into a call to printf
, so that e.g. print(5.0)
becomes printf("%f\n", 5.0)
.
The tutorial you linked does show how external function calls work -- you'd have to insert a declaration for printf
with the correct signature, then build a call to that.
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With