2007-10-22 07:01:42 +00:00
< !DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">
< html >
< head >
< title > Kaleidoscope: Implementing code generation to LLVM IR< / title >
< meta http-equiv = "Content-Type" content = "text/html; charset=utf-8" >
< meta name = "author" content = "Chris Lattner" >
< link rel = "stylesheet" href = "../llvm.css" type = "text/css" >
< / head >
< body >
< div class = "doc_title" > Kaleidoscope: Code generation to LLVM IR< / div >
2007-11-05 19:06:59 +00:00
< ul >
2007-11-05 20:04:56 +00:00
< li > < a href = "index.html" > Up to Tutorial Index< / a > < / li >
2007-11-05 19:06:59 +00:00
< li > Chapter 3
< ol >
< li > < a href = "#intro" > Chapter 3 Introduction< / a > < / li >
2007-11-06 07:26:32 +00:00
< li > < a href = "#basics" > Code Generation Setup< / a > < / li >
2007-11-05 19:06:59 +00:00
< li > < a href = "#exprs" > Expression Code Generation< / a > < / li >
< li > < a href = "#funcs" > Function Code Generation< / a > < / li >
< li > < a href = "#driver" > Driver Changes and Closing Thoughts< / a > < / li >
< li > < a href = "#code" > Full Code Listing< / a > < / li >
< / ol >
< / li >
2007-11-05 20:04:56 +00:00
< li > < a href = "LangImpl4.html" > Chapter 4< / a > : Adding JIT and Optimizer
Support< / li >
2007-11-05 19:06:59 +00:00
< / ul >
2007-10-22 07:01:42 +00:00
< div class = "doc_author" >
< p > Written by < a href = "mailto:sabre@nondot.org" > Chris Lattner< / a > < / p >
< / div >
<!-- *********************************************************************** -->
2007-11-05 19:06:59 +00:00
< div class = "doc_section" > < a name = "intro" > Chapter 3 Introduction< / a > < / div >
2007-10-22 07:01:42 +00:00
<!-- *********************************************************************** -->
< div class = "doc_text" >
2007-11-05 19:06:59 +00:00
< p > Welcome to Chapter 3 of the "< a href = "index.html" > Implementing a language
with LLVM< / a > " tutorial. This chapter shows you how to transform the < a
2007-11-13 07:06:30 +00:00
href="LangImpl2.html">Abstract Syntax Tree< / a > , built in Chapter 2, into LLVM IR.
2007-10-22 07:01:42 +00:00
This will teach you a little bit about how LLVM does things, as well as
demonstrate how easy it is to use. It's much more work to build a lexer and
2007-11-06 07:26:32 +00:00
parser than it is to generate LLVM IR code. :)
2007-10-22 07:01:42 +00:00
< / p >
2007-11-28 19:26:42 +00:00
< p > < b > Please note< / b > : the code in this chapter and later require LLVM 2.2 or
LLVM SVN to work. LLVM 2.1 and before will not work with it.< / p >
2007-10-22 07:01:42 +00:00
< / div >
<!-- *********************************************************************** -->
2007-11-06 07:26:32 +00:00
< div class = "doc_section" > < a name = "basics" > Code Generation Setup< / a > < / div >
2007-10-22 07:01:42 +00:00
<!-- *********************************************************************** -->
< div class = "doc_text" >
< p >
In order to generate LLVM IR, we want some simple setup to get started. First,
we define virtual codegen methods in each AST class:< / p >
< div class = "doc_code" >
< pre >
/// ExprAST - Base class for all expression nodes.
class ExprAST {
public:
virtual ~ExprAST() {}
2007-11-05 19:25:14 +00:00
< b > virtual Value *Codegen() = 0;< / b >
2007-10-22 07:01:42 +00:00
};
/// NumberExprAST - Expression class for numeric literals like "1.0".
class NumberExprAST : public ExprAST {
double Val;
public:
2007-10-23 04:27:44 +00:00
explicit NumberExprAST(double val) : Val(val) {}
2007-11-05 19:25:14 +00:00
< b > virtual Value *Codegen();< / b >
2007-10-22 07:01:42 +00:00
};
...
< / pre >
< / div >
2007-11-13 07:06:30 +00:00
< p > The Codegen() method says to emit IR for that AST node along with all the things it
2007-10-23 04:27:44 +00:00
depends on, and they all return an LLVM Value object.
"Value" is the class used to represent a "< a
href="http://en.wikipedia.org/wiki/Static_single_assignment_form">Static Single
Assignment (SSA)< / a > register" or "SSA value" in LLVM. The most distinct aspect
of SSA values is that their value is computed as the related instruction
executes, and it does not get a new value until (and if) the instruction
2007-11-13 07:06:30 +00:00
re-executes. In other words, there is no way to "change" an SSA value. For
2007-10-23 04:27:44 +00:00
more information, please read up on < a
href="http://en.wikipedia.org/wiki/Static_single_assignment_form">Static Single
Assignment< / a > - the concepts are really quite natural once you grok them.< / p >
2007-11-05 19:25:14 +00:00
< p > Note that instead of adding virtual methods to the ExprAST class hierarchy,
it could also make sense to use a visitor pattern or some other way to model
this. Again, this tutorial won't dwell on good software engineering practices:
2007-11-05 19:33:52 +00:00
for our purposes, adding a virtual method is simplest.< / p >
2007-11-05 19:25:14 +00:00
2007-10-23 04:27:44 +00:00
< p > The
2007-11-13 07:06:30 +00:00
second thing we want is an "Error" method like we used for the parser, which will
2007-10-22 07:01:42 +00:00
be used to report errors found during code generation (for example, use of an
undeclared parameter):< / p >
< div class = "doc_code" >
< pre >
Value *ErrorV(const char *Str) { Error(Str); return 0; }
static Module *TheModule;
static LLVMBuilder Builder;
static std::map< std::string, Value*> NamedValues;
< / pre >
< / div >
< p > The static variables will be used during code generation. < tt > TheModule< / tt >
is the LLVM construct that contains all of the functions and global variables in
a chunk of code. In many ways, it is the top-level structure that the LLVM IR
uses to contain code.< / p >
< p > The < tt > Builder< / tt > object is a helper object that makes it easy to generate
2007-11-05 18:02:15 +00:00
LLVM instructions. Instances of the < a
href="http://llvm.org/doxygen/LLVMBuilder_8h-source.html">< tt > LLVMBuilder< / tt >
2007-11-06 07:26:32 +00:00
class< / a > keep track of the current place to
2007-10-22 07:01:42 +00:00
insert instructions and has methods to create new instructions.< / p >
< p > The < tt > NamedValues< / tt > map keeps track of which values are defined in the
2007-11-06 07:26:32 +00:00
current scope and what their LLVM representation is (in other words, it is a
symbol table for the code). In this form of
2007-10-22 07:01:42 +00:00
Kaleidoscope, the only things that can be referenced are function parameters.
As such, function parameters will be in this map when generating code for their
function body.< / p >
< p >
With these basics in place, we can start talking about how to generate code for
each expression. Note that this assumes that the < tt > Builder< / tt > has been set
up to generate code < em > into< / em > something. For now, we'll assume that this
has already been done, and we'll just use it to emit code.
< / p >
< / div >
<!-- *********************************************************************** -->
< div class = "doc_section" > < a name = "exprs" > Expression Code Generation< / a > < / div >
<!-- *********************************************************************** -->
< div class = "doc_text" >
2007-11-13 07:06:30 +00:00
< p > Generating LLVM code for expression nodes is very straightforward: less
2007-10-22 07:01:42 +00:00
than 45 lines of commented code for all four of our expression nodes. First,
we'll do numeric literals:< / p >
< div class = "doc_code" >
< pre >
Value *NumberExprAST::Codegen() {
return ConstantFP::get(Type::DoubleTy, APFloat(Val));
}
< / pre >
< / div >
2007-10-23 04:51:30 +00:00
< p > In the LLVM IR, numeric constants are represented with the
< tt > ConstantFP< / tt > class, which holds the numeric value in an < tt > APFloat< / tt >
internally (< tt > APFloat< / tt > has the capability of holding floating point
constants of < em > A< / em > rbitrary < em > P< / em > recision). This code basically just
creates and returns a < tt > ConstantFP< / tt > . Note that in the LLVM IR
2007-10-22 07:01:42 +00:00
that constants are all uniqued together and shared. For this reason, the API
2007-11-06 07:26:32 +00:00
uses "the foo::get(..)" idiom instead of "new foo(..)" or "foo::create(..)".< / p >
2007-10-22 07:01:42 +00:00
< div class = "doc_code" >
< pre >
Value *VariableExprAST::Codegen() {
// Look this variable up in the function.
Value *V = NamedValues[Name];
return V ? V : ErrorV("Unknown variable name");
}
< / pre >
< / div >
2007-11-13 07:06:30 +00:00
< p > References to variables are also quite simple using LLVM. In the simple version
2007-10-23 04:51:30 +00:00
of Kaleidoscope, we assume that the variable has already been emited somewhere
and its value is available. In practice, the only values that can be in the
< tt > NamedValues< / tt > map are function arguments. This
2007-10-22 07:01:42 +00:00
code simply checks to see that the specified name is in the map (if not, an
2007-11-06 07:26:32 +00:00
unknown variable is being referenced) and returns the value for it. In future
chapters, we'll add support for < a href = "LangImpl5.html#for" > loop induction
variables< / a > in the symbol table, and for < a
href="LangImpl7.html#localvars">local variables< / a > .< / p >
2007-10-22 07:01:42 +00:00
< div class = "doc_code" >
< pre >
Value *BinaryExprAST::Codegen() {
Value *L = LHS-> Codegen();
Value *R = RHS-> Codegen();
if (L == 0 || R == 0) return 0;
switch (Op) {
case '+': return Builder.CreateAdd(L, R, "addtmp");
case '-': return Builder.CreateSub(L, R, "subtmp");
case '*': return Builder.CreateMul(L, R, "multmp");
case '< ':
2007-11-06 01:39:12 +00:00
L = Builder.CreateFCmpULT(L, R, "cmptmp");
2007-10-22 07:01:42 +00:00
// Convert bool 0/1 to double 0.0 or 1.0
return Builder.CreateUIToFP(L, Type::DoubleTy, "booltmp");
default: return ErrorV("invalid binary operator");
}
}
< / pre >
< / div >
2007-10-23 04:51:30 +00:00
< p > Binary operators start to get more interesting. The basic idea here is that
we recursively emit code for the left-hand side of the expression, then the
right-hand side, then we compute the result of the binary expression. In this
code, we do a simple switch on the opcode to create the right LLVM instruction.
< / p >
2007-10-22 07:01:42 +00:00
2007-11-13 07:06:30 +00:00
< p > In the example above, the LLVM builder class is starting to show its value.
LLVMBuilder knows where to insert the newly created instruction, all you have to
do is specify what instruction to create (e.g. with < tt > CreateAdd< / tt > ), which
2007-10-23 04:51:30 +00:00
operands to use (< tt > L< / tt > and < tt > R< / tt > here) and optionally provide a name
for the generated instruction. One nice thing about LLVM is that the name is
just a hint: if there are multiple additions in a single function, the first
will be named "addtmp" and the second will be "autorenamed" by adding a suffix,
giving it a name like "addtmp42". Local value names for instructions are purely
optional, but it makes it much easier to read the IR dumps.< / p >
2007-11-13 07:06:30 +00:00
< p > < a href = "../LangRef.html#instref" > LLVM instructions< / a > are constrained by
2007-11-06 07:26:32 +00:00
strict rules: for example, the Left and Right operators of
2007-11-13 07:06:30 +00:00
an < a href = "../LangRef.html#i_add" > add instruction< / a > must have the same
type, and the result type of the add must match the operand types. Because
2007-11-06 07:26:32 +00:00
all values in Kaleidoscope are doubles, this makes for very simple code for add,
sub and mul.< / p >
2007-10-23 04:51:30 +00:00
< p > On the other hand, LLVM specifies that the < a
href="../LangRef.html#i_fcmp">fcmp instruction< / a > always returns an 'i1' value
2007-11-13 07:06:30 +00:00
(a one bit integer). The problem with this is that Kaleidoscope wants the value to be a 0.0 or 1.0 value. In order to get these semantics, we combine the fcmp instruction with
2007-10-23 04:51:30 +00:00
a < a href = "../LangRef.html#i_uitofp" > uitofp instruction< / a > . This instruction
converts its input integer into a floating point value by treating the input
as an unsigned value. In contrast, if we used the < a
2007-11-07 05:07:10 +00:00
href="../LangRef.html#i_sitofp">sitofp instruction< / a > , the Kaleidoscope '< '
2007-10-23 04:51:30 +00:00
operator would return 0.0 and -1.0, depending on the input value.< / p >
2007-10-22 07:01:42 +00:00
< div class = "doc_code" >
< pre >
Value *CallExprAST::Codegen() {
// Look up the name in the global module table.
Function *CalleeF = TheModule-> getFunction(Callee);
if (CalleeF == 0)
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() == 0) return 0;
}
return Builder.CreateCall(CalleeF, ArgsV.begin(), ArgsV.end(), "calltmp");
}
< / pre >
< / div >
2007-11-13 07:06:30 +00:00
< p > Code generation for function calls is quite straightforward with LLVM. The
code above initially does a function name lookup in the LLVM Module's symbol
2007-10-23 04:51:30 +00:00
table. Recall that the LLVM Module is the container that holds all of the
functions we are JIT'ing. By giving each function the same name as what the
user specifies, we can use the LLVM symbol table to resolve function names for
us.< / p >
< p > Once we have the function to call, we recursively codegen each argument that
is to be passed in, and create an LLVM < a href = "../LangRef.html#i_call" > call
instruction< / a > . Note that LLVM uses the native C calling conventions by
2007-11-13 07:06:30 +00:00
default, allowing these calls to also call into standard library functions like
"sin" and "cos", with no additional effort.< / p >
2007-10-23 04:51:30 +00:00
< p > This wraps up our handling of the four basic expressions that we have so far
in Kaleidoscope. Feel free to go in and add some more. For example, by
browsing the < a href = "../LangRef.html" > LLVM language reference< / a > you'll find
several other interesting instructions that are really easy to plug into our
basic framework.< / p >
2007-10-22 07:01:42 +00:00
< / div >
<!-- *********************************************************************** -->
2007-10-23 06:23:57 +00:00
< div class = "doc_section" > < a name = "funcs" > Function Code Generation< / a > < / div >
2007-10-22 07:01:42 +00:00
<!-- *********************************************************************** -->
< div class = "doc_text" >
2007-11-07 05:07:10 +00:00
< p > Code generation for prototypes and functions must handle a number of
details, which make their code less beautiful than expression code
generation, but allows us to illustrate some important points. First, lets
talk about code generation for prototypes: they are used both for function
bodies and external function declarations. The code starts with:< / p >
2007-10-23 06:23:57 +00:00
< div class = "doc_code" >
< pre >
Function *PrototypeAST::Codegen() {
// Make the function type: double(double,double) etc.
std::vector< const Type*> Doubles(Args.size(), Type::DoubleTy);
FunctionType *FT = FunctionType::get(Type::DoubleTy, Doubles, false);
Function *F = new Function(FT, Function::ExternalLinkage, Name, TheModule);
< / pre >
< / div >
2007-11-05 19:22:50 +00:00
< p > This code packs a lot of power into a few lines. Note first that this
2007-11-07 05:07:10 +00:00
function returns a "Function*" instead of a "Value*". Because a "prototype"
really talks about the external interface for a function (not the value computed
by an expression), it makes sense for it to return the LLVM Function it
corresponds to when codegen'd.< / p >
2007-11-05 19:22:50 +00:00
2007-11-07 05:07:10 +00:00
< p > The call to < tt > FunctionType::get< / tt > creates
2007-10-23 06:23:57 +00:00
the < tt > FunctionType< / tt > that should be used for a given Prototype. Since all
function arguments in Kaleidoscope are of type double, the first line creates
2007-11-07 05:07:10 +00:00
a vector of "N" LLVM double types. It then uses the < tt > FunctionType::get< / tt >
2007-10-23 06:23:57 +00:00
method to create a function type that takes "N" doubles as arguments, returns
one double as a result, and that is not vararg (the false parameter indicates
this). Note that Types in LLVM are uniqued just like Constants are, so you
don't "new" a type, you "get" it.< / p >
< p > The final line above actually creates the function that the prototype will
2007-11-13 07:06:30 +00:00
correspond to. This indicates the type, linkage and name to use, as well as which
2007-10-23 06:23:57 +00:00
module to insert into. "< a href = "LangRef.html#linkage" > external linkage< / a > "
means that the function may be defined outside the current module and/or that it
is callable by functions outside the module. The Name passed in is the name the
user specified: since "< tt > TheModule< / tt > " is specified, this name is registered
in "< tt > TheModule< / tt > "s symbol table, which is used by the function call code
above.< / p >
< div class = "doc_code" >
< pre >
// If F conflicted, there was already something named 'Name'. If it has a
// body, don't allow redefinition or reextern.
if (F-> getName() != Name) {
// Delete the one we just made and get the existing one.
F-> eraseFromParent();
F = TheModule-> getFunction(Name);
< / pre >
< / div >
< p > The Module symbol table works just like the Function symbol table when it
comes to name conflicts: if a new function is created with a name was previously
added to the symbol table, it will get implicitly renamed when added to the
2007-11-13 07:06:30 +00:00
Module. The code above exploits this fact to determine if there was a previous
2007-10-23 06:23:57 +00:00
definition of this function.< / p >
< p > In Kaleidoscope, I choose to allow redefinitions of functions in two cases:
2007-11-07 05:07:10 +00:00
first, we want to allow 'extern'ing a function more than once, as long as the
2007-10-23 06:23:57 +00:00
prototypes for the externs match (since all arguments have the same type, we
just have to check that the number of arguments match). Second, we want to
allow 'extern'ing a function and then definining a body for it. This is useful
when defining mutually recursive functions.< / p >
< p > In order to implement this, the code above first checks to see if there is
a collision on the name of the function. If so, it deletes the function we just
created (by calling < tt > eraseFromParent< / tt > ) and then calling
< tt > getFunction< / tt > to get the existing function with the specified name. Note
that many APIs in LLVM have "erase" forms and "remove" forms. The "remove" form
unlinks the object from its parent (e.g. a Function from a Module) and returns
it. The "erase" form unlinks the object and then deletes it.< / p >
< div class = "doc_code" >
< pre >
// If F already has a body, reject this.
if (!F-> empty()) {
ErrorF("redefinition of function");
return 0;
}
// If F took a different number of args, reject.
if (F-> arg_size() != Args.size()) {
ErrorF("redefinition of function with different # args");
return 0;
}
}
< / pre >
< / div >
2007-11-07 05:07:10 +00:00
< p > In order to verify the logic above, we first check to see if the pre-existing
2007-10-23 06:23:57 +00:00
function is "empty". In this case, empty means that it has no basic blocks in
2007-11-07 05:07:10 +00:00
it, which means it has no body. If it has no body, it is a forward
2007-10-23 06:23:57 +00:00
declaration. Since we don't allow anything after a full definition of the
function, the code rejects this case. If the previous reference to a function
was an 'extern', we simply verify that the number of arguments for that
definition and this one match up. If not, we emit an error.< / p >
< div class = "doc_code" >
< pre >
// Set names for all arguments.
unsigned Idx = 0;
for (Function::arg_iterator AI = F-> arg_begin(); Idx != Args.size();
++AI, ++Idx) {
AI-> setName(Args[Idx]);
// Add arguments to variable symbol table.
NamedValues[Args[Idx]] = AI;
}
return F;
}
< / pre >
< / div >
< p > The last bit of code for prototypes loops over all of the arguments in the
2007-11-13 07:06:30 +00:00
function, setting the name of the LLVM Argument objects to match, and registering
2007-10-23 06:23:57 +00:00
the arguments in the < tt > NamedValues< / tt > map for future use by the
< tt > VariableExprAST< / tt > AST node. Once this is set up, it returns the Function
object to the caller. Note that we don't check for conflicting
argument names here (e.g. "extern foo(a b a)"). Doing so would be very
2007-11-07 05:07:10 +00:00
straight-forward with the mechanics we have already used above.< / p >
2007-10-23 06:23:57 +00:00
< div class = "doc_code" >
< pre >
Function *FunctionAST::Codegen() {
NamedValues.clear();
Function *TheFunction = Proto-> Codegen();
if (TheFunction == 0)
return 0;
< / pre >
< / div >
2007-11-13 07:06:30 +00:00
< p > Code generation for function definitions starts out simply enough: we just
codegen the prototype (Proto) and verify that it is ok. We then clear out the
2007-10-23 06:23:57 +00:00
< tt > NamedValues< / tt > map to make sure that there isn't anything in it from the
2007-11-06 05:07:30 +00:00
last function we compiled. Code generation of the prototype ensures that there
is an LLVM Function object that is ready to go for us.< / p >
2007-10-23 06:23:57 +00:00
< div class = "doc_code" >
< pre >
// Create a new basic block to start insertion into.
BasicBlock *BB = new BasicBlock("entry", TheFunction);
Builder.SetInsertPoint(BB);
if (Value *RetVal = Body-> Codegen()) {
< / pre >
< / div >
< p > Now we get to the point where the < tt > Builder< / tt > is set up. The first
line creates a new < a href = "http://en.wikipedia.org/wiki/Basic_block" > basic
block< / a > (named "entry"), which is inserted into < tt > TheFunction< / tt > . The
second line then tells the builder that new instructions should be inserted into
the end of the new basic block. Basic blocks in LLVM are an important part
of functions that define the < a
href="http://en.wikipedia.org/wiki/Control_flow_graph">Control Flow Graph< / a > .
Since we don't have any control flow, our functions will only contain one
2007-11-13 07:06:30 +00:00
block at this point. We'll fix this in < a href = "LangImpl5.html" > Chapter 5< / a > :).< / p >
2007-10-23 06:23:57 +00:00
2007-10-25 04:30:35 +00:00
< div class = "doc_code" >
< pre >
if (Value *RetVal = Body-> Codegen()) {
// Finish off the function.
Builder.CreateRet(RetVal);
// Validate the generated code, checking for consistency.
verifyFunction(*TheFunction);
return TheFunction;
}
< / pre >
< / div >
2007-10-23 06:23:57 +00:00
< p > Once the insertion point is set up, we call the < tt > CodeGen()< / tt > method for
the root expression of the function. If no error happens, this emits code to
compute the expression into the entry block and returns the value that was
computed. Assuming no error, we then create an LLVM < a
2007-10-25 04:30:35 +00:00
href="../LangRef.html#i_ret">ret instruction< / a > , which completes the function.
2007-11-13 07:06:30 +00:00
Once the function is built, we call < tt > verifyFunction< / tt > , which
2007-10-25 04:30:35 +00:00
is provided by LLVM. This function does a variety of consistency checks on the
generated code, to determine if our compiler is doing everything right. Using
this is important: it can catch a lot of bugs. Once the function is finished
and validated, we return it.< / p >
2007-10-23 06:23:57 +00:00
< div class = "doc_code" >
< pre >
// Error reading body, remove function.
TheFunction-> eraseFromParent();
return 0;
}
< / pre >
< / div >
< p > The only piece left here is handling of the error case. For simplicity, we
2007-11-13 07:06:30 +00:00
handle this by merely deleting the function we produced with the
2007-10-23 06:23:57 +00:00
< tt > eraseFromParent< / tt > method. This allows the user to redefine a function
that they incorrectly typed in before: if we didn't delete it, it would live in
the symbol table, with a body, preventing future redefinition.< / p >
2007-11-13 07:06:30 +00:00
< p > This code does have a bug, though. Since the < tt > PrototypeAST::Codegen< / tt >
can return a previously defined forward declaration, our code can actually delete
2007-10-23 06:23:57 +00:00
a forward declaration. There are a number of ways to fix this bug, see what you
can come up with! Here is a testcase:< / p >
< div class = "doc_code" >
< pre >
extern foo(a b); # ok, defines foo.
def foo(a b) c; # error, 'c' is invalid.
def bar() foo(1, 2); # error, unknown function "foo"
< / pre >
< / div >
< / div >
<!-- *********************************************************************** -->
< div class = "doc_section" > < a name = "driver" > Driver Changes and
Closing Thoughts< / a > < / div >
<!-- *********************************************************************** -->
< div class = "doc_text" >
< p >
For now, code generation to LLVM doesn't really get us much, except that we can
look at the pretty IR calls. The sample code inserts calls to Codegen into the
"< tt > HandleDefinition< / tt > ", "< tt > HandleExtern< / tt > " etc functions, and then
dumps out the LLVM IR. This gives a nice way to look at the LLVM IR for simple
functions. For example:
< / p >
< div class = "doc_code" >
< pre >
ready> < b > 4+5< / b > ;
2007-11-07 05:07:10 +00:00
Read top-level expression:
2007-10-23 06:23:57 +00:00
define double @""() {
entry:
%addtmp = add double 4.000000e+00, 5.000000e+00
ret double %addtmp
}
< / pre >
< / div >
< p > Note how the parser turns the top-level expression into anonymous functions
2007-11-07 05:07:10 +00:00
for us. This will be handy when we add < a href = "LangImpl4.html#jit" > JIT
support< / a > in the next chapter. Also note that the code is very literally
transcribed, no optimizations are being performed. We will
< a href = "LangImpl4.html#trivialconstfold" > add optimizations< / a > explicitly in
the next chapter.< / p >
2007-10-23 06:23:57 +00:00
< div class = "doc_code" >
< pre >
ready> < b > def foo(a b) a*a + 2*a*b + b*b;< / b >
2007-11-07 05:07:10 +00:00
Read function definition:
2007-10-23 06:23:57 +00:00
define double @foo(double %a, double %b) {
entry:
%multmp = mul double %a, %a
%multmp1 = mul double 2.000000e+00, %a
%multmp2 = mul double %multmp1, %b
%addtmp = add double %multmp, %multmp2
%multmp3 = mul double %b, %b
%addtmp4 = add double %addtmp, %multmp3
ret double %addtmp4
}
< / pre >
< / div >
< p > This shows some simple arithmetic. Notice the striking similarity to the
LLVM builder calls that we use to create the instructions.< / p >
< div class = "doc_code" >
< pre >
ready> < b > def bar(a) foo(a, 4.0) + bar(31337);< / b >
2007-11-07 05:07:10 +00:00
Read function definition:
2007-10-23 06:23:57 +00:00
define double @bar(double %a) {
entry:
%calltmp = call double @foo( double %a, double 4.000000e+00 )
%calltmp1 = call double @bar( double 3.133700e+04 )
%addtmp = add double %calltmp, %calltmp1
ret double %addtmp
}
< / pre >
< / div >
2007-11-05 17:39:26 +00:00
< p > This shows some function calls. Note that this function will take a long
time to execute if you call it. In the future we'll add conditional control
2007-11-13 07:06:30 +00:00
flow to actually make recursion useful :).< / p >
2007-10-23 06:23:57 +00:00
< div class = "doc_code" >
< pre >
ready> < b > extern cos(x);< / b >
2007-11-07 05:07:10 +00:00
Read extern:
2007-10-23 06:23:57 +00:00
declare double @cos(double)
ready> < b > cos(1.234);< / b >
2007-11-07 05:07:10 +00:00
Read top-level expression:
2007-10-23 06:23:57 +00:00
define double @""() {
entry:
2007-10-23 06:30:50 +00:00
%calltmp = call double @cos( double 1.234000e+00 )
2007-10-23 06:23:57 +00:00
ret double %calltmp
}
< / pre >
< / div >
< p > This shows an extern for the libm "cos" function, and a call to it.< / p >
< div class = "doc_code" >
< pre >
ready> < b > ^D< / b >
; ModuleID = 'my cool jit'
define double @""() {
entry:
%addtmp = add double 4.000000e+00, 5.000000e+00
ret double %addtmp
}
define double @foo(double %a, double %b) {
entry:
%multmp = mul double %a, %a
%multmp1 = mul double 2.000000e+00, %a
%multmp2 = mul double %multmp1, %b
%addtmp = add double %multmp, %multmp2
%multmp3 = mul double %b, %b
%addtmp4 = add double %addtmp, %multmp3
ret double %addtmp4
}
define double @bar(double %a) {
entry:
%calltmp = call double @foo( double %a, double 4.000000e+00 )
%calltmp1 = call double @bar( double 3.133700e+04 )
%addtmp = add double %calltmp, %calltmp1
ret double %addtmp
}
declare double @cos(double)
define double @""() {
entry:
%calltmp = call double @cos( double 1.234000e+00 )
ret double %calltmp
}
< / pre >
< / div >
< p > When you quit the current demo, it dumps out the IR for the entire module
generated. Here you can see the big picture with all the functions referencing
each other.< / p >
2007-11-13 07:06:30 +00:00
< p > This wraps up the third chapter of the Kaleidoscope tutorial. Up next, we'll
2007-10-23 06:23:57 +00:00
describe how to < a href = "LangImpl4.html" > add JIT codegen and optimizer
support< / a > to this so we can actually start running code!< / p >
< / div >
<!-- *********************************************************************** -->
< div class = "doc_section" > < a name = "code" > Full Code Listing< / a > < / div >
<!-- *********************************************************************** -->
< div class = "doc_text" >
< p >
Here is the complete code listing for our running example, enhanced with the
LLVM code generator. Because this uses the LLVM libraries, we need to link
them in. To do this, we use the < a
href="http://llvm.org/cmds/llvm-config.html">llvm-config< / a > tool to inform
our makefile/command line about which options to use:< / p >
< div class = "doc_code" >
< pre >
# Compile
2007-11-07 05:07:10 +00:00
g++ -g -O3 toy.cpp `llvm-config --cppflags --ldflags --libs core` -o toy
2007-10-23 06:23:57 +00:00
# Run
./toy
< / pre >
< / div >
< p > Here is the code:< / p >
2007-10-22 07:01:42 +00:00
< div class = "doc_code" >
< pre >
// To build this:
// See example below.
#include "llvm/DerivedTypes.h"
#include "llvm/Module.h"
2007-10-25 04:30:35 +00:00
#include "llvm/Analysis/Verifier.h"
2007-10-22 07:01:42 +00:00
#include "llvm/Support/LLVMBuilder.h"
#include < cstdio>
#include < string>
#include < map>
#include < vector>
using namespace llvm;
//===----------------------------------------------------------------------===//
// Lexer
//===----------------------------------------------------------------------===//
// The lexer returns tokens [0-255] if it is an unknown character, otherwise one
// of these for known things.
enum Token {
tok_eof = -1,
// commands
tok_def = -2, tok_extern = -3,
// primary
tok_identifier = -4, tok_number = -5,
};
static std::string IdentifierStr; // Filled in if tok_identifier
static double NumVal; // Filled in if tok_number
/// gettok - Return the next token from standard input.
static int gettok() {
static int LastChar = ' ';
// Skip any whitespace.
while (isspace(LastChar))
LastChar = getchar();
if (isalpha(LastChar)) { // identifier: [a-zA-Z][a-zA-Z0-9]*
IdentifierStr = LastChar;
while (isalnum((LastChar = getchar())))
IdentifierStr += LastChar;
if (IdentifierStr == "def") return tok_def;
if (IdentifierStr == "extern") return tok_extern;
return tok_identifier;
}
if (isdigit(LastChar) || LastChar == '.') { // Number: [0-9.]+
std::string NumStr;
do {
NumStr += LastChar;
LastChar = getchar();
} while (isdigit(LastChar) || LastChar == '.');
NumVal = strtod(NumStr.c_str(), 0);
return tok_number;
}
if (LastChar == '#') {
// Comment until end of line.
do LastChar = getchar();
2007-12-02 22:46:01 +00:00
while (LastChar != EOF & & LastChar != '\n' & & LastChar != '\r');
2007-10-22 07:01:42 +00:00
if (LastChar != EOF)
return gettok();
}
// Check for end of file. Don't eat the EOF.
if (LastChar == EOF)
return tok_eof;
// Otherwise, just return the character as its ascii value.
int ThisChar = LastChar;
LastChar = getchar();
return ThisChar;
}
//===----------------------------------------------------------------------===//
// Abstract Syntax Tree (aka Parse Tree)
//===----------------------------------------------------------------------===//
/// ExprAST - Base class for all expression nodes.
class ExprAST {
public:
virtual ~ExprAST() {}
virtual Value *Codegen() = 0;
};
/// NumberExprAST - Expression class for numeric literals like "1.0".
class NumberExprAST : public ExprAST {
double Val;
public:
2007-10-23 04:27:44 +00:00
explicit NumberExprAST(double val) : Val(val) {}
2007-10-22 07:01:42 +00:00
virtual Value *Codegen();
};
/// VariableExprAST - Expression class for referencing a variable, like "a".
class VariableExprAST : public ExprAST {
std::string Name;
public:
2007-10-23 04:27:44 +00:00
explicit VariableExprAST(const std::string & name) : Name(name) {}
2007-10-22 07:01:42 +00:00
virtual Value *Codegen();
};
/// BinaryExprAST - Expression class for a binary operator.
class BinaryExprAST : public ExprAST {
char Op;
ExprAST *LHS, *RHS;
public:
BinaryExprAST(char op, ExprAST *lhs, ExprAST *rhs)
: Op(op), LHS(lhs), RHS(rhs) {}
virtual Value *Codegen();
};
/// CallExprAST - Expression class for function calls.
class CallExprAST : public ExprAST {
std::string Callee;
std::vector< ExprAST*> Args;
public:
CallExprAST(const std::string & callee, std::vector< ExprAST*> & args)
: Callee(callee), Args(args) {}
virtual Value *Codegen();
};
/// PrototypeAST - This class represents the "prototype" for a function,
/// which captures its argument names as well as if it is an operator.
class PrototypeAST {
std::string Name;
std::vector< std::string> Args;
public:
PrototypeAST(const std::string & name, const std::vector< std::string> & args)
: Name(name), Args(args) {}
Function *Codegen();
};
/// FunctionAST - This class represents a function definition itself.
class FunctionAST {
PrototypeAST *Proto;
ExprAST *Body;
public:
FunctionAST(PrototypeAST *proto, ExprAST *body)
: Proto(proto), Body(body) {}
Function *Codegen();
};
//===----------------------------------------------------------------------===//
// Parser
//===----------------------------------------------------------------------===//
/// CurTok/getNextToken - Provide a simple token buffer. CurTok is the current
/// token the parser it looking at. getNextToken reads another token from the
/// lexer and updates CurTok with its results.
static int CurTok;
static int getNextToken() {
return CurTok = gettok();
}
/// BinopPrecedence - This holds the precedence for each binary operator that is
/// defined.
static std::map< char, int> BinopPrecedence;
/// GetTokPrecedence - Get the precedence of the pending binary operator token.
static int GetTokPrecedence() {
if (!isascii(CurTok))
return -1;
// Make sure it's a declared binop.
int TokPrec = BinopPrecedence[CurTok];
if (TokPrec < = 0) return -1;
return TokPrec;
}
/// Error* - These are little helper functions for error handling.
ExprAST *Error(const char *Str) { fprintf(stderr, "Error: %s\n", Str);return 0;}
PrototypeAST *ErrorP(const char *Str) { Error(Str); return 0; }
FunctionAST *ErrorF(const char *Str) { Error(Str); return 0; }
static ExprAST *ParseExpression();
/// identifierexpr
2007-11-05 17:54:34 +00:00
/// ::= identifier
/// ::= identifier '(' expression* ')'
2007-10-22 07:01:42 +00:00
static ExprAST *ParseIdentifierExpr() {
std::string IdName = IdentifierStr;
2007-11-05 17:54:34 +00:00
getNextToken(); // eat identifier.
2007-10-22 07:01:42 +00:00
if (CurTok != '(') // Simple variable ref.
return new VariableExprAST(IdName);
// Call.
getNextToken(); // eat (
std::vector< ExprAST*> Args;
2007-11-06 01:39:12 +00:00
if (CurTok != ')') {
while (1) {
ExprAST *Arg = ParseExpression();
if (!Arg) return 0;
Args.push_back(Arg);
2007-10-22 07:01:42 +00:00
2007-11-06 01:39:12 +00:00
if (CurTok == ')') break;
2007-10-22 07:01:42 +00:00
2007-11-06 01:39:12 +00:00
if (CurTok != ',')
return Error("Expected ')'");
getNextToken();
}
2007-10-22 07:01:42 +00:00
}
// Eat the ')'.
getNextToken();
return new CallExprAST(IdName, Args);
}
/// numberexpr ::= number
static ExprAST *ParseNumberExpr() {
ExprAST *Result = new NumberExprAST(NumVal);
getNextToken(); // consume the number
return Result;
}
/// parenexpr ::= '(' expression ')'
static ExprAST *ParseParenExpr() {
getNextToken(); // eat (.
ExprAST *V = ParseExpression();
if (!V) return 0;
if (CurTok != ')')
return Error("expected ')'");
getNextToken(); // eat ).
return V;
}
/// primary
/// ::= identifierexpr
/// ::= numberexpr
/// ::= parenexpr
static ExprAST *ParsePrimary() {
switch (CurTok) {
default: return Error("unknown token when expecting an expression");
case tok_identifier: return ParseIdentifierExpr();
case tok_number: return ParseNumberExpr();
case '(': return ParseParenExpr();
}
}
/// binoprhs
/// ::= ('+' primary)*
static ExprAST *ParseBinOpRHS(int ExprPrec, ExprAST *LHS) {
// If this is a binop, find its precedence.
while (1) {
int TokPrec = GetTokPrecedence();
// If this is a binop that binds at least as tightly as the current binop,
// consume it, otherwise we are done.
if (TokPrec < ExprPrec)
return LHS;
// Okay, we know this is a binop.
int BinOp = CurTok;
getNextToken(); // eat binop
// Parse the primary expression after the binary operator.
ExprAST *RHS = ParsePrimary();
if (!RHS) return 0;
// If BinOp binds less tightly with RHS than the operator after RHS, let
// the pending operator take RHS as its LHS.
int NextPrec = GetTokPrecedence();
if (TokPrec < NextPrec) {
RHS = ParseBinOpRHS(TokPrec+1, RHS);
if (RHS == 0) return 0;
}
// Merge LHS/RHS.
LHS = new BinaryExprAST(BinOp, LHS, RHS);
}
}
/// expression
/// ::= primary binoprhs
///
static ExprAST *ParseExpression() {
ExprAST *LHS = ParsePrimary();
if (!LHS) return 0;
return ParseBinOpRHS(0, LHS);
}
/// prototype
/// ::= id '(' id* ')'
static PrototypeAST *ParsePrototype() {
if (CurTok != tok_identifier)
return ErrorP("Expected function name in prototype");
std::string FnName = IdentifierStr;
getNextToken();
if (CurTok != '(')
return ErrorP("Expected '(' in prototype");
std::vector< std::string> ArgNames;
while (getNextToken() == tok_identifier)
ArgNames.push_back(IdentifierStr);
if (CurTok != ')')
return ErrorP("Expected ')' in prototype");
// success.
getNextToken(); // eat ')'.
return new PrototypeAST(FnName, ArgNames);
}
/// definition ::= 'def' prototype expression
static FunctionAST *ParseDefinition() {
getNextToken(); // eat def.
PrototypeAST *Proto = ParsePrototype();
if (Proto == 0) return 0;
if (ExprAST *E = ParseExpression())
return new FunctionAST(Proto, E);
return 0;
}
/// toplevelexpr ::= expression
static FunctionAST *ParseTopLevelExpr() {
if (ExprAST *E = ParseExpression()) {
// Make an anonymous proto.
PrototypeAST *Proto = new PrototypeAST("", std::vector< std::string> ());
return new FunctionAST(Proto, E);
}
return 0;
}
/// external ::= 'extern' prototype
static PrototypeAST *ParseExtern() {
getNextToken(); // eat extern.
return ParsePrototype();
}
//===----------------------------------------------------------------------===//
// Code Generation
//===----------------------------------------------------------------------===//
static Module *TheModule;
static LLVMBuilder Builder;
static std::map< std::string, Value*> NamedValues;
Value *ErrorV(const char *Str) { Error(Str); return 0; }
Value *NumberExprAST::Codegen() {
return ConstantFP::get(Type::DoubleTy, APFloat(Val));
}
Value *VariableExprAST::Codegen() {
// Look this variable up in the function.
Value *V = NamedValues[Name];
return V ? V : ErrorV("Unknown variable name");
}
Value *BinaryExprAST::Codegen() {
Value *L = LHS-> Codegen();
Value *R = RHS-> Codegen();
if (L == 0 || R == 0) return 0;
switch (Op) {
case '+': return Builder.CreateAdd(L, R, "addtmp");
case '-': return Builder.CreateSub(L, R, "subtmp");
case '*': return Builder.CreateMul(L, R, "multmp");
case '< ':
2007-11-06 01:39:12 +00:00
L = Builder.CreateFCmpULT(L, R, "cmptmp");
2007-10-22 07:01:42 +00:00
// Convert bool 0/1 to double 0.0 or 1.0
return Builder.CreateUIToFP(L, Type::DoubleTy, "booltmp");
default: return ErrorV("invalid binary operator");
}
}
Value *CallExprAST::Codegen() {
// Look up the name in the global module table.
Function *CalleeF = TheModule-> getFunction(Callee);
if (CalleeF == 0)
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() == 0) return 0;
}
return Builder.CreateCall(CalleeF, ArgsV.begin(), ArgsV.end(), "calltmp");
}
Function *PrototypeAST::Codegen() {
// Make the function type: double(double,double) etc.
2007-10-23 06:23:57 +00:00
std::vector< const Type*> Doubles(Args.size(), Type::DoubleTy);
FunctionType *FT = FunctionType::get(Type::DoubleTy, Doubles, false);
2007-10-22 07:01:42 +00:00
Function *F = new Function(FT, Function::ExternalLinkage, Name, TheModule);
// If F conflicted, there was already something named 'Name'. If it has a
// body, don't allow redefinition or reextern.
if (F-> getName() != Name) {
// Delete the one we just made and get the existing one.
F-> eraseFromParent();
F = TheModule-> getFunction(Name);
// If F already has a body, reject this.
if (!F-> empty()) {
ErrorF("redefinition of function");
return 0;
}
// If F took a different number of args, reject.
if (F-> arg_size() != Args.size()) {
ErrorF("redefinition of function with different # args");
return 0;
}
}
// Set names for all arguments.
unsigned Idx = 0;
for (Function::arg_iterator AI = F-> arg_begin(); Idx != Args.size();
++AI, ++Idx) {
AI-> setName(Args[Idx]);
// Add arguments to variable symbol table.
NamedValues[Args[Idx]] = AI;
}
return F;
}
Function *FunctionAST::Codegen() {
NamedValues.clear();
Function *TheFunction = Proto-> Codegen();
if (TheFunction == 0)
return 0;
// Create a new basic block to start insertion into.
2007-10-23 06:23:57 +00:00
BasicBlock *BB = new BasicBlock("entry", TheFunction);
Builder.SetInsertPoint(BB);
2007-10-22 07:01:42 +00:00
if (Value *RetVal = Body-> Codegen()) {
// Finish off the function.
Builder.CreateRet(RetVal);
2007-10-25 04:30:35 +00:00
// Validate the generated code, checking for consistency.
verifyFunction(*TheFunction);
2007-10-22 07:01:42 +00:00
return TheFunction;
}
// Error reading body, remove function.
TheFunction-> eraseFromParent();
return 0;
}
//===----------------------------------------------------------------------===//
// Top-Level parsing and JIT Driver
//===----------------------------------------------------------------------===//
static void HandleDefinition() {
if (FunctionAST *F = ParseDefinition()) {
if (Function *LF = F-> Codegen()) {
fprintf(stderr, "Read function definition:");
LF-> dump();
}
} else {
// Skip token for error recovery.
getNextToken();
}
}
static void HandleExtern() {
if (PrototypeAST *P = ParseExtern()) {
if (Function *F = P-> Codegen()) {
fprintf(stderr, "Read extern: ");
F-> dump();
}
} else {
// Skip token for error recovery.
getNextToken();
}
}
static void HandleTopLevelExpression() {
// Evaluate a top level expression into an anonymous function.
if (FunctionAST *F = ParseTopLevelExpr()) {
if (Function *LF = F-> Codegen()) {
fprintf(stderr, "Read top-level expression:");
LF-> dump();
}
} else {
// Skip token for error recovery.
getNextToken();
}
}
/// top ::= definition | external | expression | ';'
static void MainLoop() {
while (1) {
fprintf(stderr, "ready> ");
switch (CurTok) {
case tok_eof: return;
case ';': getNextToken(); break; // ignore top level semicolons.
case tok_def: HandleDefinition(); break;
case tok_extern: HandleExtern(); break;
default: HandleTopLevelExpression(); break;
}
}
}
//===----------------------------------------------------------------------===//
// "Library" functions that can be "extern'd" from user code.
//===----------------------------------------------------------------------===//
/// putchard - putchar that takes a double and returns 0.
extern "C"
double putchard(double X) {
putchar((char)X);
return 0;
}
//===----------------------------------------------------------------------===//
// Main driver code.
//===----------------------------------------------------------------------===//
int main() {
TheModule = new Module("my cool jit");
// Install standard binary operators.
// 1 is lowest precedence.
BinopPrecedence['< '] = 10;
BinopPrecedence['+'] = 20;
BinopPrecedence['-'] = 20;
BinopPrecedence['*'] = 40; // highest.
// Prime the first token.
fprintf(stderr, "ready> ");
getNextToken();
MainLoop();
TheModule-> dump();
return 0;
}
< / pre >
< / div >
< / div >
<!-- *********************************************************************** -->
< hr >
< address >
< a href = "http://jigsaw.w3.org/css-validator/check/referer" > < img
src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!">< / a >
< a href = "http://validator.w3.org/check/referer" > < img
2007-10-23 06:30:50 +00:00
src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!">< / a >
2007-10-22 07:01:42 +00:00
< a href = "mailto:sabre@nondot.org" > Chris Lattner< / a > < br >
< a href = "http://llvm.org" > The LLVM Compiler Infrastructure< / a > < br >
Last modified: $Date: 2007-10-17 11:05:13 -0700 (Wed, 17 Oct 2007) $
< / address >
< / body >
< / html >
