From evan.cheng at apple.com Mon Mar 31 02:18:39 2008
From: evan.cheng at apple.com (Evan Cheng)
Date: Mon, 31 Mar 2008 00:18:39 -0700
Subject: [llvm-commits] Speeding up instruction selection
In-Reply-To:
References:
<97188DDC-093E-4733-B8A2-7B30B6684DE1@apple.com>
<00FB5805-1FA5-4534-BBD0-E6E34EC17BF1@apple.com>
<1C1F9E6C-5B8E-4667-A119-4590F340FDE7@apple.com>
<83CD954F-CD49-4D4B-950D-B6CB6CD1B32D@apple.com>
Message-ID: <417E9823-6AF6-44B0-83FD-89F0CC079B6A@apple.com>
Ok, as long as the compile time increase isn't noticeable I won't
sweat it. However, if it happens in a few places, perhaps we need (yet
another) a getNode() that takes iterators and hide the copying inside
that instead.
Evan
On Mar 28, 2008, at 3:14 PM, Dan Gohman wrote:
>
> On Mar 27, 2008, at 6:04 PM, Evan Cheng wrote:
>>
>>> OK. This is also my view. I don't think this temporary SmallVectors
>>> introduce any significant overhead. According to the profiler (I use
>>> Google profiling tools, which do not require to compile with -pg),
>>> they consume virtually no time and that even on big testcases,
>>> that I
>>> like so much :-)
>>
>> I am not sure about this. For all the places which pass the list of
>> SDOperand's as a ptr to an array, you want to make a copy of it into
>> an array before passing it?
>
> Hi Evan,
>
> The patch is not changing every place that passes lists of
> SDOperands to make an extra copy. The copy is only needed when
> passing an existing node's operands, which is are now stored as
> an array of SDUse, to a function that wants an SDOperand array.
> And judging by Roman's current patch, this doesn't happen in
> all that many places.
>
> Dan
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
From evan.cheng at apple.com Mon Mar 31 02:24:05 2008
From: evan.cheng at apple.com (Evan Cheng)
Date: Mon, 31 Mar 2008 07:24:05 -0000
Subject: [llvm-commits] [test-suite] r48975 -
/test-suite/trunk/MultiSource/Applications/sqlite3/Makefile
Message-ID: <200803310724.m2V7O5CD032069@zion.cs.uiuc.edu>
Author: evancheng
Date: Mon Mar 31 02:24:04 2008
New Revision: 48975
URL: http://llvm.org/viewvc/llvm-project?rev=48975&view=rev
Log:
Support building with objdir != srcdir. Patch by Julien Lerough.
Modified:
test-suite/trunk/MultiSource/Applications/sqlite3/Makefile
Modified: test-suite/trunk/MultiSource/Applications/sqlite3/Makefile
URL: http://llvm.org/viewvc/llvm-project/test-suite/trunk/MultiSource/Applications/sqlite3/Makefile?rev=48975&r1=48974&r2=48975&view=diff
==============================================================================
--- test-suite/trunk/MultiSource/Applications/sqlite3/Makefile (original)
+++ test-suite/trunk/MultiSource/Applications/sqlite3/Makefile Mon Mar 31 02:24:04 2008
@@ -17,10 +17,10 @@
# the input files are generated from a tcl script
# it needs to be generated before sqlite is run
-sqlite3.c: test15.sql
+$(SourceDir)/sqlite3.c: test15.sql
-test15.sql: speedtest.tcl
- $(TCLSH) speedtest.tcl
+test15.sql: $(SourceDir)/speedtest.tcl
+ $(TCLSH) $<
clean::
rm -f test*.sql
From evan.cheng at apple.com Mon Mar 31 02:53:30 2008
From: evan.cheng at apple.com (Evan Cheng)
Date: Mon, 31 Mar 2008 07:53:30 -0000
Subject: [llvm-commits] [llvm] r48976 -
/llvm/trunk/lib/CodeGen/LiveIntervalAnalysis.cpp
Message-ID: <200803310753.m2V7rUwA008575@zion.cs.uiuc.edu>
Author: evancheng
Date: Mon Mar 31 02:53:30 2008
New Revision: 48976
URL: http://llvm.org/viewvc/llvm-project?rev=48976&view=rev
Log:
The support for remat of instructions with a register operand is hackish, to say the least. Since the register operand guaranteed to be PIC base and that it is already live at all uses, we are making sure it will not be spilled after its uses are rematerialized for both performance and correctness reasons.
Modified:
llvm/trunk/lib/CodeGen/LiveIntervalAnalysis.cpp
Modified: llvm/trunk/lib/CodeGen/LiveIntervalAnalysis.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/CodeGen/LiveIntervalAnalysis.cpp?rev=48976&r1=48975&r2=48976&view=diff
==============================================================================
--- llvm/trunk/lib/CodeGen/LiveIntervalAnalysis.cpp (original)
+++ llvm/trunk/lib/CodeGen/LiveIntervalAnalysis.cpp Mon Mar 31 02:53:30 2008
@@ -886,15 +886,6 @@
if (MI == ReMatOrigDefMI && CanDelete) {
DOUT << "\t\t\t\tErasing re-materlizable def: ";
DOUT << MI << '\n';
- unsigned ImpUse = getReMatImplicitUse(li, MI);
- if (ImpUse) {
- // To be deleted MI has a virtual register operand, update the
- // spill weight of the register interval.
- unsigned loopDepth = loopInfo->getLoopDepth(MI->getParent());
- LiveInterval &ImpLi = getInterval(ImpUse);
- ImpLi.weight -=
- getSpillWeight(false, true, loopDepth) / ImpLi.getSize();
- }
RemoveMachineInstrFromMaps(MI);
vrm.RemoveMachineInstrFromMaps(MI);
MI->eraseFromParent();
@@ -1116,6 +1107,7 @@
MachineInstr *MI = &(*ri);
MachineOperand &O = ri.getOperand();
++ri;
+ assert(!O.isImplicit() && "Spilling register that's used as implicit use?");
unsigned index = getInstructionIndex(MI);
if (index < start || index >= end)
continue;
@@ -1147,11 +1139,10 @@
if (ImpUse && MI != ReMatDefMI) {
// Re-matting an instruction with virtual register use. Update the
- // register interval's spill weight.
- unsigned loopDepth = loopInfo->getLoopDepth(MI->getParent());
+ // register interval's spill weight to HUGE_VALF to prevent it from
+ // being spilled.
LiveInterval &ImpLi = getInterval(ImpUse);
- ImpLi.weight +=
- getSpillWeight(false, true, loopDepth) * NumUses / ImpLi.getSize();
+ ImpLi.weight = HUGE_VALF;
}
unsigned MBBId = MBB->getNumber();
@@ -1575,12 +1566,10 @@
if (ImpUse) {
// Re-matting an instruction with virtual register use. Add the
// register as an implicit use on the use MI and update the register
- // interval's spill weight.
- unsigned loopDepth = loopInfo->getLoopDepth(MI->getParent());
+ // interval's spill weight to HUGE_VALF to prevent it from being
+ // spilled.
LiveInterval &ImpLi = getInterval(ImpUse);
- ImpLi.weight +=
- getSpillWeight(false, true, loopDepth) / ImpLi.getSize();
-
+ ImpLi.weight = HUGE_VALF;
MI->addOperand(MachineOperand::CreateReg(ImpUse, false, true));
}
}
From evan.cheng at apple.com Mon Mar 31 02:54:19 2008
From: evan.cheng at apple.com (Evan Cheng)
Date: Mon, 31 Mar 2008 07:54:19 -0000
Subject: [llvm-commits] [llvm] r48977 -
/llvm/trunk/lib/Target/X86/X86InstrInfo.cpp
Message-ID: <200803310754.m2V7sJkW008612@zion.cs.uiuc.edu>
Author: evancheng
Date: Mon Mar 31 02:54:19 2008
New Revision: 48977
URL: http://llvm.org/viewvc/llvm-project?rev=48977&view=rev
Log:
Re-apply 48911.
Modified:
llvm/trunk/lib/Target/X86/X86InstrInfo.cpp
Modified: llvm/trunk/lib/Target/X86/X86InstrInfo.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86InstrInfo.cpp?rev=48977&r1=48976&r2=48977&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86InstrInfo.cpp (original)
+++ llvm/trunk/lib/Target/X86/X86InstrInfo.cpp Mon Mar 31 02:54:19 2008
@@ -748,6 +748,12 @@
}
return isPICBase;
}
+
+/// isGVStub - Return true if the GV requires an extra load to get the
+/// real address.
+static inline bool isGVStub(GlobalValue *GV, X86TargetMachine &TM) {
+ return TM.getSubtarget().GVRequiresExtraLoad(GV, TM, false);
+}
bool X86InstrInfo::isReallyTriviallyReMaterializable(MachineInstr *MI) const {
switch (MI->getOpcode()) {
@@ -769,7 +775,9 @@
if (MI->getOperand(1).isReg() &&
MI->getOperand(2).isImm() &&
MI->getOperand(3).isReg() && MI->getOperand(3).getReg() == 0 &&
- MI->getOperand(4).isCPI()) {
+ (MI->getOperand(4).isCPI() ||
+ (MI->getOperand(4).isGlobal() &&
+ isGVStub(MI->getOperand(4).getGlobal(), TM)))) {
unsigned BaseReg = MI->getOperand(1).getReg();
if (BaseReg == 0)
return true;
@@ -831,13 +839,9 @@
// Loads from constant pools are trivially invariant.
if (MO.isCPI())
return true;
-
- if (MO.isGlobal()) {
- if (TM.getSubtarget().GVRequiresExtraLoad(MO.getGlobal(),
- TM, false))
- return true;
- return false;
- }
+
+ if (MO.isGlobal())
+ return isGVStub(MO.getGlobal(), TM);
// If this is a load from an invariant stack slot, the load is a constant.
if (MO.isFI()) {
From idadesub at users.sourceforge.net Mon Mar 31 03:44:53 2008
From: idadesub at users.sourceforge.net (Erick Tryzelaar)
Date: Mon, 31 Mar 2008 08:44:53 -0000
Subject: [llvm-commits] [llvm] r48978 - in /llvm/trunk/docs/tutorial:
OCamlLangImpl3.html OCamlLangImpl4.html OCamlLangImpl5.html
OCamlLangImpl6.html OCamlLangImpl7.html
Message-ID: <200803310844.m2V8is9B010008@zion.cs.uiuc.edu>
Author: erickt
Date: Mon Mar 31 03:44:50 2008
New Revision: 48978
URL: http://llvm.org/viewvc/llvm-project?rev=48978&view=rev
Log:
Chapter 5, 6, and 7 of the ocaml/kaleidoscope tutorial
and fix some tabs in chapter 3 and 4.
Added:
llvm/trunk/docs/tutorial/OCamlLangImpl5.html
llvm/trunk/docs/tutorial/OCamlLangImpl6.html
llvm/trunk/docs/tutorial/OCamlLangImpl7.html
Modified:
llvm/trunk/docs/tutorial/OCamlLangImpl3.html
llvm/trunk/docs/tutorial/OCamlLangImpl4.html
Modified: llvm/trunk/docs/tutorial/OCamlLangImpl3.html
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/docs/tutorial/OCamlLangImpl3.html?rev=48978&r1=48977&r2=48978&view=diff
==============================================================================
--- llvm/trunk/docs/tutorial/OCamlLangImpl3.html (original)
+++ llvm/trunk/docs/tutorial/OCamlLangImpl3.html Mon Mar 31 03:44:50 2008
@@ -183,7 +183,7 @@
let i = build_fcmp Fcmp.Ult lhs_val rhs_val "cmptmp" builder in
build_uitofp i double_type "booltmp" builder
| _ -> raise (Error "invalid binary operator")
- end
+ end
@@ -280,7 +280,7 @@
(* Make the function type: double(double,double) etc. *)
let doubles = Array.make (Array.length args) double_type in
let ft = function_type double_type doubles in
- let f =
+ let f =
match lookup_function name the_module with
Modified: llvm/trunk/docs/tutorial/OCamlLangImpl4.html
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/docs/tutorial/OCamlLangImpl4.html?rev=48978&r1=48977&r2=48978&view=diff
==============================================================================
--- llvm/trunk/docs/tutorial/OCamlLangImpl4.html (original)
+++ llvm/trunk/docs/tutorial/OCamlLangImpl4.html Mon Mar 31 03:44:50 2008
@@ -237,7 +237,7 @@
let codegen_func the_fpm = function
- ...
+ ...
try
let ret_val = codegen_expr body in
@@ -316,10 +316,9 @@
...
let main () =
...
-
- (* Create the JIT. *)
+ (* Create the JIT. *)
let the_module_provider = ModuleProvider.create Codegen.the_module in
- let the_execution_engine = ExecutionEngine.create the_module_provider in
+ let the_execution_engine = ExecutionEngine.create the_module_provider in
...
@@ -508,6 +507,17 @@
LLVM JIT and optimizer. To build this example, use:
+
+
+# Compile
+ocamlbuild toy.byte
+# Run
+./toy.byte
+
Welcome to Chapter 5 of the "Implementing a language
+with LLVM" tutorial. Parts 1-4 described the implementation of the simple
+Kaleidoscope language and included support for generating LLVM IR, followed by
+optimizations and a JIT compiler. Unfortunately, as presented, Kaleidoscope is
+mostly useless: it has no control flow other than call and return. This means
+that you can't have conditional branches in the code, significantly limiting its
+power. In this episode of "build that compiler", we'll extend Kaleidoscope to
+have an if/then/else expression plus a simple 'for' loop.
+Extending Kaleidoscope to support if/then/else is quite straightforward. It
+basically requires adding lexer support for this "new" concept to the lexer,
+parser, AST, and LLVM code emitter. This example is nice, because it shows how
+easy it is to "grow" a language over time, incrementally extending it as new
+ideas are discovered.
+
+
Before we get going on "how" we add this extension, lets talk about "what" we
+want. The basic idea is that we want to be able to write this sort of thing:
+
+
+
+
+def fib(x)
+ if x < 3 then
+ 1
+ else
+ fib(x-1)+fib(x-2);
+
+
+
+
In Kaleidoscope, every construct is an expression: there are no statements.
+As such, the if/then/else expression needs to return a value like any other.
+Since we're using a mostly functional form, we'll have it evaluate its
+conditional, then return the 'then' or 'else' value based on how the condition
+was resolved. This is very similar to the C "?:" expression.
+
+
The semantics of the if/then/else expression is that it evaluates the
+condition to a boolean equality value: 0.0 is considered to be false and
+everything else is considered to be true.
+If the condition is true, the first subexpression is evaluated and returned, if
+the condition is false, the second subexpression is evaluated and returned.
+Since Kaleidoscope allows side-effects, this behavior is important to nail down.
+
+
+
Now that we know what we "want", lets break this down into its constituent
+pieces.
Now that we have the relevant tokens coming from the lexer and we have the
+AST node to build, our parsing logic is relatively straightforward. First we
+define a new parsing function:
Now that we have it parsing and building the AST, the final piece is adding
+LLVM code generation support. This is the most interesting part of the
+if/then/else example, because this is where it starts to introduce new concepts.
+All of the code above has been thoroughly described in previous chapters.
+
+
+
To motivate the code we want to produce, lets take a look at a simple
+example. Consider:
+
+
+
+extern foo();
+extern bar();
+def baz(x) if x then foo() else bar();
+
+
+
+
If you disable optimizations, the code you'll (soon) get from Kaleidoscope
+looks like this:
To visualize the control flow graph, you can use a nifty feature of the LLVM
+'opt' tool. If you put this LLVM IR
+into "t.ll" and run "llvm-as < t.ll | opt -analyze -view-cfg", a window will pop up and you'll
+see this graph:
+
+
+
+
Another way to get this is to call "Llvm_analysis.view_function_cfg
+f" or "Llvm_analysis.view_function_cfg_only f" (where f
+is a "Function") either by inserting actual calls into the code and
+recompiling or by calling these in the debugger. LLVM has many nice features
+for visualizing various graphs.
+
+
Getting back to the generated code, it is fairly simple: the entry block
+evaluates the conditional expression ("x" in our case here) and compares the
+result to 0.0 with the "fcmp one"
+instruction ('one' is "Ordered and Not Equal"). Based on the result of this
+expression, the code jumps to either the "then" or "else" blocks, which contain
+the expressions for the true/false cases.
+
+
Once the then/else blocks are finished executing, they both branch back to the
+'ifcont' block to execute the code that happens after the if/then/else. In this
+case the only thing left to do is to return to the caller of the function. The
+question then becomes: how does the code know which expression to return?
+
+
The answer to this question involves an important SSA operation: the
+Phi
+operation. If you're not familiar with SSA, the wikipedia
+article is a good introduction and there are various other introductions to
+it available on your favorite search engine. The short version is that
+"execution" of the Phi operation requires "remembering" which block control came
+from. The Phi operation takes on the value corresponding to the input control
+block. In this case, if control comes in from the "then" block, it gets the
+value of "calltmp". If control comes from the "else" block, it gets the value
+of "calltmp1".
+
+
At this point, you are probably starting to think "Oh no! This means my
+simple and elegant front-end will have to start generating SSA form in order to
+use LLVM!". Fortunately, this is not the case, and we strongly advise
+not implementing an SSA construction algorithm in your front-end
+unless there is an amazingly good reason to do so. In practice, there are two
+sorts of values that float around in code written for your average imperative
+programming language that might need Phi nodes:
+
+
+
Code that involves user variables: x = 1; x = x + 1;
+
Values that are implicit in the structure of your AST, such as the Phi node
+in this case.
+
+
+
In Chapter 7 of this tutorial ("mutable
+variables"), we'll talk about #1
+in depth. For now, just believe me that you don't need SSA construction to
+handle this case. For #2, you have the choice of using the techniques that we will
+describe for #1, or you can insert Phi nodes directly, if convenient. In this
+case, it is really really easy to generate the Phi node, so we choose to do it
+directly.
+
+
Okay, enough of the motivation and overview, lets generate code!
In order to generate code for this, we implement the Codegen method
+for IfExprAST:
+
+
+
+let rec codegen_expr = function
+ ...
+ | Ast.If (cond, then_, else_) ->
+ let cond = codegen_expr cond in
+
+ (* Convert condition to a bool by comparing equal to 0.0 *)
+ let zero = const_float double_type 0.0 in
+ let cond_val = build_fcmp Fcmp.One cond zero "ifcond" builder in
+
+
+
+
This code is straightforward and similar to what we saw before. We emit the
+expression for the condition, then compare that value to zero to get a truth
+value as a 1-bit (bool) value.
+
+
+
+ (* Grab the first block so that we might later add the conditional branch
+ * to it at the end of the function. *)
+ let start_bb = insertion_block builder in
+ let the_function = block_parent start_bb in
+
+ let then_bb = append_block "then" the_function in
+ position_at_end then_bb builder;
+
+
+
+
+As opposed to the C++ tutorial, we have to build
+our basic blocks bottom up since we can't have dangling BasicBlocks. We start
+off by saving a pointer to the first block (which might not be the entry
+block), which we'll need to build a conditional branch later. We do this by
+asking the builder for the current BasicBlock. The fourth line
+gets the current Function object that is being built. It gets this by the
+start_bb for its "parent" (the function it is currently embedded
+into).
+
+
Once it has that, it creates one block. It is automatically appended into
+the function's list of blocks.
+
+
+
+ (* Emit 'then' value. *)
+ position_at_end then_bb builder;
+ let then_val = codegen_expr then_ in
+
+ (* Codegen of 'then' can change the current block, update then_bb for the
+ * phi. We create a new name because one is used for the phi node, and the
+ * other is used for the conditional branch. *)
+ let new_then_bb = insertion_block builder in
+
+
+
+
We move the builder to start inserting into the "then" block. Strictly
+speaking, this call moves the insertion point to be at the end of the specified
+block. However, since the "then" block is empty, it also starts out by
+inserting at the beginning of the block. :)
+
+
Once the insertion point is set, we recursively codegen the "then" expression
+from the AST.
+
+
The final line here is quite subtle, but is very important. The basic issue
+is that when we create the Phi node in the merge block, we need to set up the
+block/value pairs that indicate how the Phi will work. Importantly, the Phi
+node expects to have an entry for each predecessor of the block in the CFG. Why
+then, are we getting the current block when we just set it to ThenBB 5 lines
+above? The problem is that the "Then" expression may actually itself change the
+block that the Builder is emitting into if, for example, it contains a nested
+"if/then/else" expression. Because calling Codegen recursively could
+arbitrarily change the notion of the current block, we are required to get an
+up-to-date value for code that will set up the Phi node.
+
+
+
+ (* Emit 'else' value. *)
+ let else_bb = append_block "else" the_function in
+ position_at_end else_bb builder;
+ let else_val = codegen_expr else_ in
+
+ (* Codegen of 'else' can change the current block, update else_bb for the
+ * phi. *)
+ let new_else_bb = insertion_block builder in
+
+
+
+
Code generation for the 'else' block is basically identical to codegen for
+the 'then' block.
+
+
+
+ (* Emit merge block. *)
+ let merge_bb = append_block "ifcont" the_function in
+ position_at_end merge_bb builder;
+ let incoming = [(then_val, new_then_bb); (else_val, new_else_bb)] in
+ let phi = build_phi incoming "iftmp" builder in
+
+
+
+
The first two lines here are now familiar: the first adds the "merge" block
+to the Function object. The second block changes the insertion point so that
+newly created code will go into the "merge" block. Once that is done, we need
+to create the PHI node and set up the block/value pairs for the PHI.
+
+
+
+ (* Return to the start block to add the conditional branch. *)
+ position_at_end start_bb builder;
+ ignore (build_cond_br cond_val then_bb else_bb builder);
+
+
+
+
Once the blocks are created, we can emit the conditional branch that chooses
+between them. Note that creating new blocks does not implicitly affect the
+LLVMBuilder, so it is still inserting into the block that the condition
+went into. This is why we needed to save the "start" block.
+
+
+
+ (* Set a unconditional branch at the end of the 'then' block and the
+ * 'else' block to the 'merge' block. *)
+ position_at_end new_then_bb builder; ignore (build_br merge_bb builder);
+ position_at_end new_else_bb builder; ignore (build_br merge_bb builder);
+
+ (* Finally, set the builder to the end of the merge block. *)
+ position_at_end merge_bb builder;
+
+ phi
+
+
+
+
To finish off the blocks, we create an unconditional branch
+to the merge block. One interesting (and very important) aspect of the LLVM IR
+is that it requires all basic blocks
+to be "terminated" with a control flow
+instruction such as return or branch. This means that all control flow,
+including fall throughs must be made explicit in the LLVM IR. If you
+violate this rule, the verifier will emit an error.
+
+
Finally, the CodeGen function returns the phi node as the value computed by
+the if/then/else expression. In our example above, this returned value will
+feed into the code for the top-level function, which will create the return
+instruction.
+
+
Overall, we now have the ability to execute conditional code in
+Kaleidoscope. With this extension, Kaleidoscope is a fairly complete language
+that can calculate a wide variety of numeric functions. Next up we'll add
+another useful expression that is familiar from non-functional languages...
Now that we know how to add basic control flow constructs to the language,
+we have the tools to add more powerful things. Lets add something more
+aggressive, a 'for' expression:
+
+
+
+ extern putchard(char);
+ def printstar(n)
+ for i = 1, i < n, 1.0 in
+ putchard(42); # ascii 42 = '*'
+
+ # print 100 '*' characters
+ printstar(100);
+
+
+
+
This expression defines a new variable ("i" in this case) which iterates from
+a starting value, while the condition ("i < n" in this case) is true,
+incrementing by an optional step value ("1.0" in this case). If the step value
+is omitted, it defaults to 1.0. While the loop is true, it executes its
+body expression. Because we don't have anything better to return, we'll just
+define the loop as always returning 0.0. In the future when we have mutable
+variables, it will get more useful.
+
+
As before, lets talk about the changes that we need to Kaleidoscope to
+support this.
The parser code is also fairly standard. The only interesting thing here is
+handling of the optional step value. The parser code handles it by checking to
+see if the second comma is present. If not, it sets the step value to null in
+the AST node:
+
+
+
+let rec parse_primary = parser
+ ...
+ (* forexpr
+ ::= 'for' identifier '=' expr ',' expr (',' expr)? 'in' expression *)
+ | [< 'Token.For;
+ 'Token.Ident id ?? "expected identifier after for";
+ 'Token.Kwd '=' ?? "expected '=' after for";
+ stream >] ->
+ begin parser
+ | [<
+ start=parse_expr;
+ 'Token.Kwd ',' ?? "expected ',' after for";
+ end_=parse_expr;
+ stream >] ->
+ let step =
+ begin parser
+ | [< 'Token.Kwd ','; step=parse_expr >] -> Some step
+ | [< >] -> None
+ end stream
+ in
+ begin parser
+ | [< 'Token.In; body=parse_expr >] ->
+ Ast.For (id, start, end_, step, body)
+ | [< >] ->
+ raise (Stream.Error "expected 'in' after for")
+ end stream
+ | [< >] ->
+ raise (Stream.Error "expected '=' after for")
+ end stream
+
Now we get to the good part: the LLVM IR we want to generate for this thing.
+With the simple example above, we get this LLVM IR (note that this dump is
+generated with optimizations disabled for clarity):
+
The first part of Codegen is very simple: we just output the start expression
+for the loop value:
+
+
+
+let rec codegen_expr = function
+ ...
+ | Ast.For (var_name, start, end_, step, body) ->
+ (* Emit the start code first, without 'variable' in scope. *)
+ let start_val = codegen_expr start in
+
+
+
+
With this out of the way, the next step is to set up the LLVM basic block
+for the start of the loop body. In the case above, the whole loop body is one
+block, but remember that the body code itself could consist of multiple blocks
+(e.g. if it contains an if/then/else or a for/in expression).
+
+
+
+ (* Make the new basic block for the loop header, inserting after current
+ * block. *)
+ let preheader_bb = insertion_block builder in
+ let the_function = block_parent preheader_bb in
+ let loop_bb = append_block "loop" the_function in
+
+ (* Insert an explicit fall through from the current block to the
+ * loop_bb. *)
+ ignore (build_br loop_bb builder);
+
+
+
+
This code is similar to what we saw for if/then/else. Because we will need
+it to create the Phi node, we remember the block that falls through into the
+loop. Once we have that, we create the actual block that starts the loop and
+create an unconditional branch for the fall-through between the two blocks.
+
+
+
+ (* Start insertion in loop_bb. *)
+ position_at_end loop_bb builder;
+
+ (* Start the PHI node with an entry for start. *)
+ let variable = build_phi [(start_val, preheader_bb)] var_name builder in
+
+
+
+
Now that the "preheader" for the loop is set up, we switch to emitting code
+for the loop body. To begin with, we move the insertion point and create the
+PHI node for the loop induction variable. Since we already know the incoming
+value for the starting value, we add it to the Phi node. Note that the Phi will
+eventually get a second value for the backedge, but we can't set it up yet
+(because it doesn't exist!).
+
+
+
+ (* Within the loop, the variable is defined equal to the PHI node. If it
+ * shadows an existing variable, we have to restore it, so save it
+ * now. *)
+ let old_val =
+ try Some (Hashtbl.find named_values var_name) with Not_found -> None
+ in
+ Hashtbl.add named_values var_name variable;
+
+ (* Emit the body of the loop. This, like any other expr, can change the
+ * current BB. Note that we ignore the value computed by the body, but
+ * don't allow an error *)
+ ignore (codegen_expr body);
+
+
+
+
Now the code starts to get more interesting. Our 'for' loop introduces a new
+variable to the symbol table. This means that our symbol table can now contain
+either function arguments or loop variables. To handle this, before we codegen
+the body of the loop, we add the loop variable as the current value for its
+name. Note that it is possible that there is a variable of the same name in the
+outer scope. It would be easy to make this an error (emit an error and return
+null if there is already an entry for VarName) but we choose to allow shadowing
+of variables. In order to handle this correctly, we remember the Value that
+we are potentially shadowing in old_val (which will be None if there is
+no shadowed variable).
+
+
Once the loop variable is set into the symbol table, the code recursively
+codegen's the body. This allows the body to use the loop variable: any
+references to it will naturally find it in the symbol table.
+
+
+
+ (* Emit the step value. *)
+ let step_val =
+ match step with
+ | Some step -> codegen_expr step
+ (* If not specified, use 1.0. *)
+ | None -> const_float double_type 1.0
+ in
+
+ let next_var = build_add variable step_val "nextvar" builder in
+
+
+
+
Now that the body is emitted, we compute the next value of the iteration
+variable by adding the step value, or 1.0 if it isn't present.
+'next_var' will be the value of the loop variable on the next iteration
+of the loop.
+
+
+
+ (* Compute the end condition. *)
+ let end_cond = codegen_expr end_ in
+
+ (* Convert condition to a bool by comparing equal to 0.0. *)
+ let zero = const_float double_type 0.0 in
+ let end_cond = build_fcmp Fcmp.One end_cond zero "loopcond" builder in
+
+
+
+
Finally, we evaluate the exit value of the loop, to determine whether the
+loop should exit. This mirrors the condition evaluation for the if/then/else
+statement.
+
+
+
+ (* Create the "after loop" block and insert it. *)
+ let loop_end_bb = insertion_block builder in
+ let after_bb = append_block "afterloop" the_function in
+
+ (* Insert the conditional branch into the end of loop_end_bb. *)
+ ignore (build_cond_br end_cond loop_bb after_bb builder);
+
+ (* Any new code will be inserted in after_bb. *)
+ position_at_end after_bb builder;
+
+
+
+
With the code for the body of the loop complete, we just need to finish up
+the control flow for it. This code remembers the end block (for the phi node), then creates the block for the loop exit ("afterloop"). Based on the value of the
+exit condition, it creates a conditional branch that chooses between executing
+the loop again and exiting the loop. Any future code is emitted in the
+"afterloop" block, so it sets the insertion position to it.
+
+
+
+ (* Add a new entry to the PHI node for the backedge. *)
+ add_incoming (next_var, loop_end_bb) variable;
+
+ (* Restore the unshadowed variable. *)
+ begin match old_val with
+ | Some old_val -> Hashtbl.add named_values var_name old_val
+ | None -> ()
+ end;
+
+ (* for expr always returns 0.0. *)
+ const_null double_type
+
+
+
+
The final code handles various cleanups: now that we have the
+"next_var" value, we can add the incoming value to the loop PHI node.
+After that, we remove the loop variable from the symbol table, so that it isn't
+in scope after the for loop. Finally, code generation of the for loop always
+returns 0.0, so that is what we return from Codegen.codegen_expr.
+
+
With this, we conclude the "adding control flow to Kaleidoscope" chapter of
+the tutorial. In this chapter we added two control flow constructs, and used
+them to motivate a couple of aspects of the LLVM IR that are important for
+front-end implementors to know. In the next chapter of our saga, we will get
+a bit crazier and add user-defined operators
+to our poor innocent language.
+(*===----------------------------------------------------------------------===
+ * Lexer Tokens
+ *===----------------------------------------------------------------------===*)
+
+(* The lexer returns these 'Kwd' if it is an unknown character, otherwise one of
+ * these others for known things. *)
+type token =
+ (* commands *)
+ | Def | Extern
+
+ (* primary *)
+ | Ident of string | Number of float
+
+ (* unknown *)
+ | Kwd of char
+
+ (* control *)
+ | If | Then | Else
+ | For | In
+
+(*===----------------------------------------------------------------------===
+ * Abstract Syntax Tree (aka Parse Tree)
+ *===----------------------------------------------------------------------===*)
+
+(* expr - Base type for all expression nodes. *)
+type expr =
+ (* variant for numeric literals like "1.0". *)
+ | Number of float
+
+ (* variant for referencing a variable, like "a". *)
+ | Variable of string
+
+ (* variant for a binary operator. *)
+ | Binary of char * expr * expr
+
+ (* variant for function calls. *)
+ | Call of string * expr array
+
+ (* variant for if/then/else. *)
+ | If of expr * expr * expr
+
+ (* variant for for/in. *)
+ | For of string * expr * expr * expr option * expr
+
+(* proto - This type represents the "prototype" for a function, which captures
+ * its name, and its argument names (thus implicitly the number of arguments the
+ * function takes). *)
+type proto = Prototype of string * string array
+
+(* func - This type represents a function definition itself. *)
+type func = Function of proto * expr
+
+
+
+
parser.ml:
+
+
+(*===---------------------------------------------------------------------===
+ * Parser
+ *===---------------------------------------------------------------------===*)
+
+(* binop_precedence - This holds the precedence for each binary operator that is
+ * defined *)
+let binop_precedence:(char, int) Hashtbl.t = Hashtbl.create 10
+
+(* precedence - Get the precedence of the pending binary operator token. *)
+let precedence c = try Hashtbl.find binop_precedence c with Not_found -> -1
+
+(* primary
+ * ::= identifier
+ * ::= numberexpr
+ * ::= parenexpr
+ * ::= ifexpr
+ * ::= forexpr *)
+let rec parse_primary = parser
+ (* numberexpr ::= number *)
+ | [< 'Token.Number n >] -> Ast.Number n
+
+ (* parenexpr ::= '(' expression ')' *)
+ | [< 'Token.Kwd '('; e=parse_expr; 'Token.Kwd ')' ?? "expected ')'" >] -> e
+
+ (* identifierexpr
+ * ::= identifier
+ * ::= identifier '(' argumentexpr ')' *)
+ | [< 'Token.Ident id; stream >] ->
+ let rec parse_args accumulator = parser
+ | [< e=parse_expr; stream >] ->
+ begin parser
+ | [< 'Token.Kwd ','; e=parse_args (e :: accumulator) >] -> e
+ | [< >] -> e :: accumulator
+ end stream
+ | [< >] -> accumulator
+ in
+ let rec parse_ident id = parser
+ (* Call. *)
+ | [< 'Token.Kwd '(';
+ args=parse_args [];
+ 'Token.Kwd ')' ?? "expected ')'">] ->
+ Ast.Call (id, Array.of_list (List.rev args))
+
+ (* Simple variable ref. *)
+ | [< >] -> Ast.Variable id
+ in
+ parse_ident id stream
+
+ (* ifexpr ::= 'if' expr 'then' expr 'else' expr *)
+ | [< 'Token.If; c=parse_expr;
+ 'Token.Then ?? "expected 'then'"; t=parse_expr;
+ 'Token.Else ?? "expected 'else'"; e=parse_expr >] ->
+ Ast.If (c, t, e)
+
+ (* forexpr
+ ::= 'for' identifier '=' expr ',' expr (',' expr)? 'in' expression *)
+ | [< 'Token.For;
+ 'Token.Ident id ?? "expected identifier after for";
+ 'Token.Kwd '=' ?? "expected '=' after for";
+ stream >] ->
+ begin parser
+ | [<
+ start=parse_expr;
+ 'Token.Kwd ',' ?? "expected ',' after for";
+ end_=parse_expr;
+ stream >] ->
+ let step =
+ begin parser
+ | [< 'Token.Kwd ','; step=parse_expr >] -> Some step
+ | [< >] -> None
+ end stream
+ in
+ begin parser
+ | [< 'Token.In; body=parse_expr >] ->
+ Ast.For (id, start, end_, step, body)
+ | [< >] ->
+ raise (Stream.Error "expected 'in' after for")
+ end stream
+ | [< >] ->
+ raise (Stream.Error "expected '=' after for")
+ end stream
+
+ | [< >] -> raise (Stream.Error "unknown token when expecting an expression.")
+
+(* binoprhs
+ * ::= ('+' primary)* *)
+and parse_bin_rhs expr_prec lhs stream =
+ match Stream.peek stream with
+ (* If this is a binop, find its precedence. *)
+ | Some (Token.Kwd c) when Hashtbl.mem binop_precedence c ->
+ let token_prec = precedence c in
+
+ (* If this is a binop that binds at least as tightly as the current binop,
+ * consume it, otherwise we are done. *)
+ if token_prec < expr_prec then lhs else begin
+ (* Eat the binop. *)
+ Stream.junk stream;
+
+ (* Parse the primary expression after the binary operator. *)
+ let rhs = parse_primary stream in
+
+ (* Okay, we know this is a binop. *)
+ let rhs =
+ match Stream.peek stream with
+ | Some (Token.Kwd c2) ->
+ (* If BinOp binds less tightly with rhs than the operator after
+ * rhs, let the pending operator take rhs as its lhs. *)
+ let next_prec = precedence c2 in
+ if token_prec < next_prec
+ then parse_bin_rhs (token_prec + 1) rhs stream
+ else rhs
+ | _ -> rhs
+ in
+
+ (* Merge lhs/rhs. *)
+ let lhs = Ast.Binary (c, lhs, rhs) in
+ parse_bin_rhs expr_prec lhs stream
+ end
+ | _ -> lhs
+
+(* expression
+ * ::= primary binoprhs *)
+and parse_expr = parser
+ | [< lhs=parse_primary; stream >] -> parse_bin_rhs 0 lhs stream
+
+(* prototype
+ * ::= id '(' id* ')' *)
+let parse_prototype =
+ let rec parse_args accumulator = parser
+ | [< 'Token.Ident id; e=parse_args (id::accumulator) >] -> e
+ | [< >] -> accumulator
+ in
+
+ parser
+ | [< 'Token.Ident id;
+ 'Token.Kwd '(' ?? "expected '(' in prototype";
+ args=parse_args [];
+ 'Token.Kwd ')' ?? "expected ')' in prototype" >] ->
+ (* success. *)
+ Ast.Prototype (id, Array.of_list (List.rev args))
+
+ | [< >] ->
+ raise (Stream.Error "expected function name in prototype")
+
+(* definition ::= 'def' prototype expression *)
+let parse_definition = parser
+ | [< 'Token.Def; p=parse_prototype; e=parse_expr >] ->
+ Ast.Function (p, e)
+
+(* toplevelexpr ::= expression *)
+let parse_toplevel = parser
+ | [< e=parse_expr >] ->
+ (* Make an anonymous proto. *)
+ Ast.Function (Ast.Prototype ("", [||]), e)
+
+(* external ::= 'extern' prototype *)
+let parse_extern = parser
+ | [< 'Token.Extern; e=parse_prototype >] -> e
+
+
+
+
codegen.ml:
+
+
+(*===----------------------------------------------------------------------===
+ * Code Generation
+ *===----------------------------------------------------------------------===*)
+
+open Llvm
+
+exception Error of string
+
+let the_module = create_module "my cool jit"
+let builder = builder ()
+let named_values:(string, llvalue) Hashtbl.t = Hashtbl.create 10
+
+let rec codegen_expr = function
+ | Ast.Number n -> const_float double_type n
+ | Ast.Variable name ->
+ (try Hashtbl.find named_values name with
+ | Not_found -> raise (Error "unknown variable name"))
+ | Ast.Binary (op, lhs, rhs) ->
+ let lhs_val = codegen_expr lhs in
+ let rhs_val = codegen_expr rhs in
+ begin
+ match op with
+ | '+' -> build_add lhs_val rhs_val "addtmp" builder
+ | '-' -> build_sub lhs_val rhs_val "subtmp" builder
+ | '*' -> build_mul lhs_val rhs_val "multmp" builder
+ | '<' ->
+ (* Convert bool 0/1 to double 0.0 or 1.0 *)
+ let i = build_fcmp Fcmp.Ult lhs_val rhs_val "cmptmp" builder in
+ build_uitofp i double_type "booltmp" builder
+ | _ -> raise (Error "invalid binary operator")
+ end
+ | Ast.Call (callee, args) ->
+ (* Look up the name in the module table. *)
+ let callee =
+ match lookup_function callee the_module with
+ | Some callee -> callee
+ | None -> raise (Error "unknown function referenced")
+ in
+ let params = params callee in
+
+ (* If argument mismatch error. *)
+ if Array.length params == Array.length args then () else
+ raise (Error "incorrect # arguments passed");
+ let args = Array.map codegen_expr args in
+ build_call callee args "calltmp" builder
+ | Ast.If (cond, then_, else_) ->
+ let cond = codegen_expr cond in
+
+ (* Convert condition to a bool by comparing equal to 0.0 *)
+ let zero = const_float double_type 0.0 in
+ let cond_val = build_fcmp Fcmp.One cond zero "ifcond" builder in
+
+ (* Grab the first block so that we might later add the conditional branch
+ * to it at the end of the function. *)
+ let start_bb = insertion_block builder in
+ let the_function = block_parent start_bb in
+
+ let then_bb = append_block "then" the_function in
+
+ (* Emit 'then' value. *)
+ position_at_end then_bb builder;
+ let then_val = codegen_expr then_ in
+
+ (* Codegen of 'then' can change the current block, update then_bb for the
+ * phi. We create a new name because one is used for the phi node, and the
+ * other is used for the conditional branch. *)
+ let new_then_bb = insertion_block builder in
+
+ (* Emit 'else' value. *)
+ let else_bb = append_block "else" the_function in
+ position_at_end else_bb builder;
+ let else_val = codegen_expr else_ in
+
+ (* Codegen of 'else' can change the current block, update else_bb for the
+ * phi. *)
+ let new_else_bb = insertion_block builder in
+
+ (* Emit merge block. *)
+ let merge_bb = append_block "ifcont" the_function in
+ position_at_end merge_bb builder;
+ let incoming = [(then_val, new_then_bb); (else_val, new_else_bb)] in
+ let phi = build_phi incoming "iftmp" builder in
+
+ (* Return to the start block to add the conditional branch. *)
+ position_at_end start_bb builder;
+ ignore (build_cond_br cond_val then_bb else_bb builder);
+
+ (* Set a unconditional branch at the end of the 'then' block and the
+ * 'else' block to the 'merge' block. *)
+ position_at_end new_then_bb builder; ignore (build_br merge_bb builder);
+ position_at_end new_else_bb builder; ignore (build_br merge_bb builder);
+
+ (* Finally, set the builder to the end of the merge block. *)
+ position_at_end merge_bb builder;
+
+ phi
+ | Ast.For (var_name, start, end_, step, body) ->
+ (* Emit the start code first, without 'variable' in scope. *)
+ let start_val = codegen_expr start in
+
+ (* Make the new basic block for the loop header, inserting after current
+ * block. *)
+ let preheader_bb = insertion_block builder in
+ let the_function = block_parent preheader_bb in
+ let loop_bb = append_block "loop" the_function in
+
+ (* Insert an explicit fall through from the current block to the
+ * loop_bb. *)
+ ignore (build_br loop_bb builder);
+
+ (* Start insertion in loop_bb. *)
+ position_at_end loop_bb builder;
+
+ (* Start the PHI node with an entry for start. *)
+ let variable = build_phi [(start_val, preheader_bb)] var_name builder in
+
+ (* Within the loop, the variable is defined equal to the PHI node. If it
+ * shadows an existing variable, we have to restore it, so save it
+ * now. *)
+ let old_val =
+ try Some (Hashtbl.find named_values var_name) with Not_found -> None
+ in
+ Hashtbl.add named_values var_name variable;
+
+ (* Emit the body of the loop. This, like any other expr, can change the
+ * current BB. Note that we ignore the value computed by the body, but
+ * don't allow an error *)
+ ignore (codegen_expr body);
+
+ (* Emit the step value. *)
+ let step_val =
+ match step with
+ | Some step -> codegen_expr step
+ (* If not specified, use 1.0. *)
+ | None -> const_float double_type 1.0
+ in
+
+ let next_var = build_add variable step_val "nextvar" builder in
+
+ (* Compute the end condition. *)
+ let end_cond = codegen_expr end_ in
+
+ (* Convert condition to a bool by comparing equal to 0.0. *)
+ let zero = const_float double_type 0.0 in
+ let end_cond = build_fcmp Fcmp.One end_cond zero "loopcond" builder in
+
+ (* Create the "after loop" block and insert it. *)
+ let loop_end_bb = insertion_block builder in
+ let after_bb = append_block "afterloop" the_function in
+
+ (* Insert the conditional branch into the end of loop_end_bb. *)
+ ignore (build_cond_br end_cond loop_bb after_bb builder);
+
+ (* Any new code will be inserted in after_bb. *)
+ position_at_end after_bb builder;
+
+ (* Add a new entry to the PHI node for the backedge. *)
+ add_incoming (next_var, loop_end_bb) variable;
+
+ (* Restore the unshadowed variable. *)
+ begin match old_val with
+ | Some old_val -> Hashtbl.add named_values var_name old_val
+ | None -> ()
+ end;
+
+ (* for expr always returns 0.0. *)
+ const_null double_type
+
+let codegen_proto = function
+ | Ast.Prototype (name, args) ->
+ (* Make the function type: double(double,double) etc. *)
+ let doubles = Array.make (Array.length args) double_type in
+ let ft = function_type double_type doubles in
+ let f =
+ match lookup_function name the_module with
+ | None -> declare_function name ft the_module
+
+ (* If 'f' conflicted, there was already something named 'name'. If it
+ * has a body, don't allow redefinition or reextern. *)
+ | Some f ->
+ (* If 'f' already has a body, reject this. *)
+ if block_begin f <> At_end f then
+ raise (Error "redefinition of function");
+
+ (* If 'f' took a different number of arguments, reject. *)
+ if element_type (type_of f) <> ft then
+ raise (Error "redefinition of function with different # args");
+ f
+ in
+
+ (* Set names for all arguments. *)
+ Array.iteri (fun i a ->
+ let n = args.(i) in
+ set_value_name n a;
+ Hashtbl.add named_values n a;
+ ) (params f);
+ f
+
+let codegen_func the_fpm = function
+ | Ast.Function (proto, body) ->
+ Hashtbl.clear named_values;
+ let the_function = codegen_proto proto in
+
+ (* Create a new basic block to start insertion into. *)
+ let bb = append_block "entry" the_function in
+ position_at_end bb builder;
+
+ try
+ let ret_val = codegen_expr body in
+
+ (* Finish off the function. *)
+ let _ = build_ret ret_val builder in
+
+ (* Validate the generated code, checking for consistency. *)
+ Llvm_analysis.assert_valid_function the_function;
+
+ (* Optimize the function. *)
+ let _ = PassManager.run_function the_function the_fpm in
+
+ the_function
+ with e ->
+ delete_function the_function;
+ raise e
+
+
+
+
toplevel.ml:
+
+
+(*===----------------------------------------------------------------------===
+ * Top-Level parsing and JIT Driver
+ *===----------------------------------------------------------------------===*)
+
+open Llvm
+open Llvm_executionengine
+
+(* top ::= definition | external | expression | ';' *)
+let rec main_loop the_fpm the_execution_engine stream =
+ match Stream.peek stream with
+ | None -> ()
+
+ (* ignore top-level semicolons. *)
+ | Some (Token.Kwd ';') ->
+ Stream.junk stream;
+ main_loop the_fpm the_execution_engine stream
+
+ | Some token ->
+ begin
+ try match token with
+ | Token.Def ->
+ let e = Parser.parse_definition stream in
+ print_endline "parsed a function definition.";
+ dump_value (Codegen.codegen_func the_fpm e);
+ | Token.Extern ->
+ let e = Parser.parse_extern stream in
+ print_endline "parsed an extern.";
+ dump_value (Codegen.codegen_proto e);
+ | _ ->
+ (* Evaluate a top-level expression into an anonymous function. *)
+ let e = Parser.parse_toplevel stream in
+ print_endline "parsed a top-level expr";
+ let the_function = Codegen.codegen_func the_fpm e in
+ dump_value the_function;
+
+ (* JIT the function, returning a function pointer. *)
+ let result = ExecutionEngine.run_function the_function [||]
+ the_execution_engine in
+
+ print_string "Evaluated to ";
+ print_float (GenericValue.as_float double_type result);
+ print_newline ();
+ with Stream.Error s | Codegen.Error s ->
+ (* Skip token for error recovery. *)
+ Stream.junk stream;
+ print_endline s;
+ end;
+ print_string "ready> "; flush stdout;
+ main_loop the_fpm the_execution_engine stream
+
+
+
+
toy.ml:
+
+
+(*===----------------------------------------------------------------------===
+ * Main driver code.
+ *===----------------------------------------------------------------------===*)
+
+open Llvm
+open Llvm_executionengine
+open Llvm_target
+open Llvm_scalar_opts
+
+let main () =
+ (* Install standard binary operators.
+ * 1 is the lowest precedence. *)
+ Hashtbl.add Parser.binop_precedence '<' 10;
+ Hashtbl.add Parser.binop_precedence '+' 20;
+ Hashtbl.add Parser.binop_precedence '-' 20;
+ Hashtbl.add Parser.binop_precedence '*' 40; (* highest. *)
+
+ (* Prime the first token. *)
+ print_string "ready> "; flush stdout;
+ let stream = Lexer.lex (Stream.of_channel stdin) in
+
+ (* Create the JIT. *)
+ let the_module_provider = ModuleProvider.create Codegen.the_module in
+ let the_execution_engine = ExecutionEngine.create the_module_provider in
+ let the_fpm = PassManager.create_function the_module_provider in
+
+ (* Set up the optimizer pipeline. Start with registering info about how the
+ * target lays out data structures. *)
+ TargetData.add (ExecutionEngine.target_data the_execution_engine) the_fpm;
+
+ (* Do simple "peephole" optimizations and bit-twiddling optzn. *)
+ add_instruction_combining the_fpm;
+
+ (* reassociate expressions. *)
+ add_reassociation the_fpm;
+
+ (* Eliminate Common SubExpressions. *)
+ add_gvn the_fpm;
+
+ (* Simplify the control flow graph (deleting unreachable blocks, etc). *)
+ add_cfg_simplification the_fpm;
+
+ (* Run the main "interpreter loop" now. *)
+ Toplevel.main_loop the_fpm the_execution_engine stream;
+
+ (* Print out all the generated code. *)
+ dump_module Codegen.the_module
+;;
+
+main ()
+
+
+
+
bindings.c
+
+
+#include <stdio.h>
+
+/* putchard - putchar that takes a double and returns 0. */
+extern double putchard(double X) {
+ putchar((char)X);
+ return 0;
+}
+
Welcome to Chapter 6 of the "Implementing a language
+with LLVM" tutorial. At this point in our tutorial, we now have a fully
+functional language that is fairly minimal, but also useful. There
+is still one big problem with it, however. Our language doesn't have many
+useful operators (like division, logical negation, or even any comparisons
+besides less-than).
+
+
This chapter of the tutorial takes a wild digression into adding user-defined
+operators to the simple and beautiful Kaleidoscope language. This digression now
+gives us a simple and ugly language in some ways, but also a powerful one at the
+same time. One of the great things about creating your own language is that you
+get to decide what is good or bad. In this tutorial we'll assume that it is
+okay to use this as a way to show some interesting parsing techniques.
+
+
At the end of this tutorial, we'll run through an example Kaleidoscope
+application that renders the Mandelbrot set. This gives
+an example of what you can build with Kaleidoscope and its feature set.
+The "operator overloading" that we will add to Kaleidoscope is more general than
+languages like C++. In C++, you are only allowed to redefine existing
+operators: you can't programatically change the grammar, introduce new
+operators, change precedence levels, etc. In this chapter, we will add this
+capability to Kaleidoscope, which will let the user round out the set of
+operators that are supported.
+
+
The point of going into user-defined operators in a tutorial like this is to
+show the power and flexibility of using a hand-written parser. Thus far, the parser
+we have been implementing uses recursive descent for most parts of the grammar and
+operator precedence parsing for the expressions. See Chapter 2 for details. Without using operator
+precedence parsing, it would be very difficult to allow the programmer to
+introduce new operators into the grammar: the grammar is dynamically extensible
+as the JIT runs.
+
+
The two specific features we'll add are programmable unary operators (right
+now, Kaleidoscope has no unary operators at all) as well as binary operators.
+An example of this is:
+
+
+
+# Logical unary not.
+def unary!(v)
+ if v then
+ 0
+ else
+ 1;
+
+# Define > with the same precedence as <.
+def binary> 10 (LHS RHS)
+ RHS < LHS;
+
+# Binary "logical or", (note that it does not "short circuit")
+def binary| 5 (LHS RHS)
+ if LHS then
+ 1
+ else if RHS then
+ 1
+ else
+ 0;
+
+# Define = with slightly lower precedence than relationals.
+def binary= 9 (LHS RHS)
+ !(LHS < RHS | LHS > RHS);
+
+
+
+
Many languages aspire to being able to implement their standard runtime
+library in the language itself. In Kaleidoscope, we can implement significant
+parts of the language in the library!
+
+
We will break down implementation of these features into two parts:
+implementing support for user-defined binary operators and adding unary
+operators.
This just adds lexer support for the unary and binary keywords, like we
+did in previous chapters. One nice
+thing about our current AST, is that we represent binary operators with full
+generalisation by using their ASCII code as the opcode. For our extended
+operators, we'll use this same representation, so we don't need any new AST or
+parser support.
+
+
On the other hand, we have to be able to represent the definitions of these
+new operators, in the "def binary| 5" part of the function definition. In our
+grammar so far, the "name" for the function definition is parsed as the
+"prototype" production and into the Ast.Prototype AST node. To
+represent our new user-defined operators as prototypes, we have to extend
+the Ast.Prototype AST node like this:
+
+
+
+(* proto - This type represents the "prototype" for a function, which captures
+ * its name, and its argument names (thus implicitly the number of arguments the
+ * function takes). *)
+type proto =
+ | Prototype of string * string array
+ | BinOpPrototype of string * string array * int
+
+
+
+
Basically, in addition to knowing a name for the prototype, we now keep track
+of whether it was an operator, and if it was, what precedence level the operator
+is at. The precedence is only used for binary operators (as you'll see below,
+it just doesn't apply for unary operators). Now that we have a way to represent
+the prototype for a user-defined operator, we need to parse it:
+
+
+
+(* prototype
+ * ::= id '(' id* ')'
+ * ::= binary LETTER number? (id, id)
+ * ::= unary LETTER number? (id) *)
+let parse_prototype =
+ let rec parse_args accumulator = parser
+ | [< 'Token.Ident id; e=parse_args (id::accumulator) >] -> e
+ | [< >] -> accumulator
+ in
+ let parse_operator = parser
+ | [< 'Token.Unary >] -> "unary", 1
+ | [< 'Token.Binary >] -> "binary", 2
+ in
+ let parse_binary_precedence = parser
+ | [< 'Token.Number n >] -> int_of_float n
+ | [< >] -> 30
+ in
+ parser
+ | [< 'Token.Ident id;
+ 'Token.Kwd '(' ?? "expected '(' in prototype";
+ args=parse_args [];
+ 'Token.Kwd ')' ?? "expected ')' in prototype" >] ->
+ (* success. *)
+ Ast.Prototype (id, Array.of_list (List.rev args))
+ | [< (prefix, kind)=parse_operator;
+ 'Token.Kwd op ?? "expected an operator";
+ (* Read the precedence if present. *)
+ binary_precedence=parse_binary_precedence;
+ 'Token.Kwd '(' ?? "expected '(' in prototype";
+ args=parse_args [];
+ 'Token.Kwd ')' ?? "expected ')' in prototype" >] ->
+ let name = prefix ^ (String.make 1 op) in
+ let args = Array.of_list (List.rev args) in
+
+ (* Verify right number of arguments for operator. *)
+ if Array.length args != kind
+ then raise (Stream.Error "invalid number of operands for operator")
+ else
+ if kind == 1 then
+ Ast.Prototype (name, args)
+ else
+ Ast.BinOpPrototype (name, args, binary_precedence)
+ | [< >] ->
+ raise (Stream.Error "expected function name in prototype")
+
+
+
+
This is all fairly straightforward parsing code, and we have already seen
+a lot of similar code in the past. One interesting part about the code above is
+the couple lines that set up name for binary operators. This builds
+names like "binary@" for a newly defined "@" operator. This then takes
+advantage of the fact that symbol names in the LLVM symbol table are allowed to
+have any character in them, including embedded nul characters.
+
+
The next interesting thing to add, is codegen support for these binary
+operators. Given our current structure, this is a simple addition of a default
+case for our existing binary operator node:
+
+
+
+let codegen_expr = function
+ ...
+ | Ast.Binary (op, lhs, rhs) ->
+ let lhs_val = codegen_expr lhs in
+ let rhs_val = codegen_expr rhs in
+ begin
+ match op with
+ | '+' -> build_add lhs_val rhs_val "addtmp" builder
+ | '-' -> build_sub lhs_val rhs_val "subtmp" builder
+ | '*' -> build_mul lhs_val rhs_val "multmp" builder
+ | '<' ->
+ (* Convert bool 0/1 to double 0.0 or 1.0 *)
+ let i = build_fcmp Fcmp.Ult lhs_val rhs_val "cmptmp" builder in
+ build_uitofp i double_type "booltmp" builder
+ | _ ->
+ (* If it wasn't a builtin binary operator, it must be a user defined
+ * one. Emit a call to it. *)
+ let callee = "binary" ^ (String.make 1 op) in
+ let callee =
+ match lookup_function callee the_module with
+ | Some callee -> callee
+ | None -> raise (Error "binary operator not found!")
+ in
+ build_call callee [|lhs_val; rhs_val|] "binop" builder
+ end
+
+
+
+
As you can see above, the new code is actually really simple. It just does
+a lookup for the appropriate operator in the symbol table and generates a
+function call to it. Since user-defined operators are just built as normal
+functions (because the "prototype" boils down to a function with the right
+name) everything falls into place.
+
+
The final piece of code we are missing, is a bit of top level magic:
+
+
+
+let codegen_func the_fpm = function
+ | Ast.Function (proto, body) ->
+ Hashtbl.clear named_values;
+ let the_function = codegen_proto proto in
+
+ (* If this is an operator, install it. *)
+ begin match proto with
+ | Ast.BinOpPrototype (name, args, prec) ->
+ let op = name.[String.length name - 1] in
+ Hashtbl.add Parser.binop_precedence op prec;
+ | _ -> ()
+ end;
+
+ (* Create a new basic block to start insertion into. *)
+ let bb = append_block "entry" the_function in
+ position_at_end bb builder;
+ ...
+
+
+
+
Basically, before codegening a function, if it is a user-defined operator, we
+register it in the precedence table. This allows the binary operator parsing
+logic we already have in place to handle it. Since we are working on a
+fully-general operator precedence parser, this is all we need to do to "extend
+the grammar".
+
+
Now we have useful user-defined binary operators. This builds a lot
+on the previous framework we built for other operators. Adding unary operators
+is a bit more challenging, because we don't have any framework for it yet - lets
+see what it takes.
Since we don't currently support unary operators in the Kaleidoscope
+language, we'll need to add everything to support them. Above, we added simple
+support for the 'unary' keyword to the lexer. In addition to that, we need an
+AST node:
+
+
+
+type expr =
+ ...
+ (* variant for a unary operator. *)
+ | Unary of char * expr
+ ...
+
+
+
+
This AST node is very simple and obvious by now. It directly mirrors the
+binary operator AST node, except that it only has one child. With this, we
+need to add the parsing logic. Parsing a unary operator is pretty simple: we'll
+add a new function to do it:
+
+
+
+(* unary
+ * ::= primary
+ * ::= '!' unary *)
+and parse_unary = parser
+ (* If this is a unary operator, read it. *)
+ | [< 'Token.Kwd op when op != '(' && op != ')'; operand=parse_expr >] ->
+ Ast.Unary (op, operand)
+
+ (* If the current token is not an operator, it must be a primary expr. *)
+ | [< stream >] -> parse_primary stream
+
+
+
+
The grammar we add is pretty straightforward here. If we see a unary
+operator when parsing a primary operator, we eat the operator as a prefix and
+parse the remaining piece as another unary operator. This allows us to handle
+multiple unary operators (e.g. "!!x"). Note that unary operators can't have
+ambiguous parses like binary operators can, so there is no need for precedence
+information.
+
+
The problem with this function, is that we need to call ParseUnary from
+somewhere. To do this, we change previous callers of ParsePrimary to call
+parse_unary instead:
With these two simple changes, we are now able to parse unary operators and build the
+AST for them. Next up, we need to add parser support for prototypes, to parse
+the unary operator prototype. We extend the binary operator code above
+with:
+
+
+
+(* prototype
+ * ::= id '(' id* ')'
+ * ::= binary LETTER number? (id, id)
+ * ::= unary LETTER number? (id) *)
+let parse_prototype =
+ let rec parse_args accumulator = parser
+ | [< 'Token.Ident id; e=parse_args (id::accumulator) >] -> e
+ | [< >] -> accumulator
+ in
+ let parse_operator = parser
+ | [< 'Token.Unary >] -> "unary", 1
+ | [< 'Token.Binary >] -> "binary", 2
+ in
+ let parse_binary_precedence = parser
+ | [< 'Token.Number n >] -> int_of_float n
+ | [< >] -> 30
+ in
+ parser
+ | [< 'Token.Ident id;
+ 'Token.Kwd '(' ?? "expected '(' in prototype";
+ args=parse_args [];
+ 'Token.Kwd ')' ?? "expected ')' in prototype" >] ->
+ (* success. *)
+ Ast.Prototype (id, Array.of_list (List.rev args))
+ | [< (prefix, kind)=parse_operator;
+ 'Token.Kwd op ?? "expected an operator";
+ (* Read the precedence if present. *)
+ binary_precedence=parse_binary_precedence;
+ 'Token.Kwd '(' ?? "expected '(' in prototype";
+ args=parse_args [];
+ 'Token.Kwd ')' ?? "expected ')' in prototype" >] ->
+ let name = prefix ^ (String.make 1 op) in
+ let args = Array.of_list (List.rev args) in
+
+ (* Verify right number of arguments for operator. *)
+ if Array.length args != kind
+ then raise (Stream.Error "invalid number of operands for operator")
+ else
+ if kind == 1 then
+ Ast.Prototype (name, args)
+ else
+ Ast.BinOpPrototype (name, args, binary_precedence)
+ | [< >] ->
+ raise (Stream.Error "expected function name in prototype")
+
+
+
+
As with binary operators, we name unary operators with a name that includes
+the operator character. This assists us at code generation time. Speaking of,
+the final piece we need to add is codegen support for unary operators. It looks
+like this:
+
+
+
+let rec codegen_expr = function
+ ...
+ | Ast.Unary (op, operand) ->
+ let operand = codegen_expr operand in
+ let callee = "unary" ^ (String.make 1 op) in
+ let callee =
+ match lookup_function callee the_module with
+ | Some callee -> callee
+ | None -> raise (Error "unknown unary operator")
+ in
+ build_call callee [|operand|] "unop" builder
+
+
+
+
This code is similar to, but simpler than, the code for binary operators. It
+is simpler primarily because it doesn't need to handle any predefined operators.
+
It is somewhat hard to believe, but with a few simple extensions we've
+covered in the last chapters, we have grown a real-ish language. With this, we
+can do a lot of interesting things, including I/O, math, and a bunch of other
+things. For example, we can now add a nice sequencing operator (printd is
+defined to print out the specified value and a newline):
We can also define a bunch of other "primitive" operations, such as:
+
+
+
+# Logical unary not.
+def unary!(v)
+ if v then
+ 0
+ else
+ 1;
+
+# Unary negate.
+def unary-(v)
+ 0-v;
+
+# Define > with the same precedence as >.
+def binary> 10 (LHS RHS)
+ RHS < LHS;
+
+# Binary logical or, which does not short circuit.
+def binary| 5 (LHS RHS)
+ if LHS then
+ 1
+ else if RHS then
+ 1
+ else
+ 0;
+
+# Binary logical and, which does not short circuit.
+def binary& 6 (LHS RHS)
+ if !LHS then
+ 0
+ else
+ !!RHS;
+
+# Define = with slightly lower precedence than relationals.
+def binary = 9 (LHS RHS)
+ !(LHS < RHS | LHS > RHS);
+
+
+
+
+
+
Given the previous if/then/else support, we can also define interesting
+functions for I/O. For example, the following prints out a character whose
+"density" reflects the value passed in: the lower the value, the denser the
+character:
+
+
+
+ready>
+
+extern putchard(char)
+def printdensity(d)
+ if d > 8 then
+ putchard(32) # ' '
+ else if d > 4 then
+ putchard(46) # '.'
+ else if d > 2 then
+ putchard(43) # '+'
+ else
+ putchard(42); # '*'
+...
+ready> printdensity(1): printdensity(2): printdensity(3) :
+ printdensity(4): printdensity(5): printdensity(9): putchard(10);
+*++..
+Evaluated to 0.000000
+
+
+
+
Based on these simple primitive operations, we can start to define more
+interesting things. For example, here's a little function that solves for the
+number of iterations it takes a function in the complex plane to
+converge:
+
+
+
+# determine whether the specific location diverges.
+# Solve for z = z^2 + c in the complex plane.
+def mandleconverger(real imag iters creal cimag)
+ if iters > 255 | (real*real + imag*imag > 4) then
+ iters
+ else
+ mandleconverger(real*real - imag*imag + creal,
+ 2*real*imag + cimag,
+ iters+1, creal, cimag);
+
+# return the number of iterations required for the iteration to escape
+def mandleconverge(real imag)
+ mandleconverger(real, imag, 0, real, imag);
+
+
+
+
This "z = z2 + c" function is a beautiful little creature that is the basis
+for computation of the Mandelbrot Set. Our
+mandelconverge function returns the number of iterations that it takes
+for a complex orbit to escape, saturating to 255. This is not a very useful
+function by itself, but if you plot its value over a two-dimensional plane,
+you can see the Mandelbrot set. Given that we are limited to using putchard
+here, our amazing graphical output is limited, but we can whip together
+something using the density plotter above:
+
+
+
+# compute and plot the mandlebrot set with the specified 2 dimensional range
+# info.
+def mandelhelp(xmin xmax xstep ymin ymax ystep)
+ for y = ymin, y < ymax, ystep in (
+ (for x = xmin, x < xmax, xstep in
+ printdensity(mandleconverge(x,y)))
+ : putchard(10)
+ )
+
+# mandel - This is a convenient helper function for ploting the mandelbrot set
+# from the specified position with the specified Magnification.
+def mandel(realstart imagstart realmag imagmag)
+ mandelhelp(realstart, realstart+realmag*78, realmag,
+ imagstart, imagstart+imagmag*40, imagmag);
+
+
+
+
Given this, we can try plotting out the mandlebrot set! Lets try it out:
At this point, you may be starting to realize that Kaleidoscope is a real
+and powerful language. It may not be self-similar :), but it can be used to
+plot things that are!
+
+
With this, we conclude the "adding user-defined operators" chapter of the
+tutorial. We have successfully augmented our language, adding the ability to
+extend the language in the library, and we have shown how this can be used to
+build a simple but interesting end-user application in Kaleidoscope. At this
+point, Kaleidoscope can build a variety of applications that are functional and
+can call functions with side-effects, but it can't actually define and mutate a
+variable itself.
+
+
Strikingly, variable mutation is an important feature of some
+languages, and it is not at all obvious how to add
+support for mutable variables without having to add an "SSA construction"
+phase to your front-end. In the next chapter, we will describe how you can
+add variable mutation without building SSA in your front-end.
+(*===----------------------------------------------------------------------===
+ * Lexer Tokens
+ *===----------------------------------------------------------------------===*)
+
+(* The lexer returns these 'Kwd' if it is an unknown character, otherwise one of
+ * these others for known things. *)
+type token =
+ (* commands *)
+ | Def | Extern
+
+ (* primary *)
+ | Ident of string | Number of float
+
+ (* unknown *)
+ | Kwd of char
+
+ (* control *)
+ | If | Then | Else
+ | For | In
+
+ (* operators *)
+ | Binary | Unary
+
+(*===----------------------------------------------------------------------===
+ * Abstract Syntax Tree (aka Parse Tree)
+ *===----------------------------------------------------------------------===*)
+
+(* expr - Base type for all expression nodes. *)
+type expr =
+ (* variant for numeric literals like "1.0". *)
+ | Number of float
+
+ (* variant for referencing a variable, like "a". *)
+ | Variable of string
+
+ (* variant for a unary operator. *)
+ | Unary of char * expr
+
+ (* variant for a binary operator. *)
+ | Binary of char * expr * expr
+
+ (* variant for function calls. *)
+ | Call of string * expr array
+
+ (* variant for if/then/else. *)
+ | If of expr * expr * expr
+
+ (* variant for for/in. *)
+ | For of string * expr * expr * expr option * expr
+
+(* proto - This type represents the "prototype" for a function, which captures
+ * its name, and its argument names (thus implicitly the number of arguments the
+ * function takes). *)
+type proto =
+ | Prototype of string * string array
+ | BinOpPrototype of string * string array * int
+
+(* func - This type represents a function definition itself. *)
+type func = Function of proto * expr
+
+
+
+
parser.ml:
+
+
+(*===---------------------------------------------------------------------===
+ * Parser
+ *===---------------------------------------------------------------------===*)
+
+(* binop_precedence - This holds the precedence for each binary operator that is
+ * defined *)
+let binop_precedence:(char, int) Hashtbl.t = Hashtbl.create 10
+
+(* precedence - Get the precedence of the pending binary operator token. *)
+let precedence c = try Hashtbl.find binop_precedence c with Not_found -> -1
+
+(* primary
+ * ::= identifier
+ * ::= numberexpr
+ * ::= parenexpr
+ * ::= ifexpr
+ * ::= forexpr *)
+let rec parse_primary = parser
+ (* numberexpr ::= number *)
+ | [< 'Token.Number n >] -> Ast.Number n
+
+ (* parenexpr ::= '(' expression ')' *)
+ | [< 'Token.Kwd '('; e=parse_expr; 'Token.Kwd ')' ?? "expected ')'" >] -> e
+
+ (* identifierexpr
+ * ::= identifier
+ * ::= identifier '(' argumentexpr ')' *)
+ | [< 'Token.Ident id; stream >] ->
+ let rec parse_args accumulator = parser
+ | [< e=parse_expr; stream >] ->
+ begin parser
+ | [< 'Token.Kwd ','; e=parse_args (e :: accumulator) >] -> e
+ | [< >] -> e :: accumulator
+ end stream
+ | [< >] -> accumulator
+ in
+ let rec parse_ident id = parser
+ (* Call. *)
+ | [< 'Token.Kwd '(';
+ args=parse_args [];
+ 'Token.Kwd ')' ?? "expected ')'">] ->
+ Ast.Call (id, Array.of_list (List.rev args))
+
+ (* Simple variable ref. *)
+ | [< >] -> Ast.Variable id
+ in
+ parse_ident id stream
+
+ (* ifexpr ::= 'if' expr 'then' expr 'else' expr *)
+ | [< 'Token.If; c=parse_expr;
+ 'Token.Then ?? "expected 'then'"; t=parse_expr;
+ 'Token.Else ?? "expected 'else'"; e=parse_expr >] ->
+ Ast.If (c, t, e)
+
+ (* forexpr
+ ::= 'for' identifier '=' expr ',' expr (',' expr)? 'in' expression *)
+ | [< 'Token.For;
+ 'Token.Ident id ?? "expected identifier after for";
+ 'Token.Kwd '=' ?? "expected '=' after for";
+ stream >] ->
+ begin parser
+ | [<
+ start=parse_expr;
+ 'Token.Kwd ',' ?? "expected ',' after for";
+ end_=parse_expr;
+ stream >] ->
+ let step =
+ begin parser
+ | [< 'Token.Kwd ','; step=parse_expr >] -> Some step
+ | [< >] -> None
+ end stream
+ in
+ begin parser
+ | [< 'Token.In; body=parse_expr >] ->
+ Ast.For (id, start, end_, step, body)
+ | [< >] ->
+ raise (Stream.Error "expected 'in' after for")
+ end stream
+ | [< >] ->
+ raise (Stream.Error "expected '=' after for")
+ end stream
+
+ | [< >] -> raise (Stream.Error "unknown token when expecting an expression.")
+
+(* unary
+ * ::= primary
+ * ::= '!' unary *)
+and parse_unary = parser
+ (* If this is a unary operator, read it. *)
+ | [< 'Token.Kwd op when op != '(' && op != ')'; operand=parse_expr >] ->
+ Ast.Unary (op, operand)
+
+ (* If the current token is not an operator, it must be a primary expr. *)
+ | [< stream >] -> parse_primary stream
+
+(* binoprhs
+ * ::= ('+' primary)* *)
+and parse_bin_rhs expr_prec lhs stream =
+ match Stream.peek stream with
+ (* If this is a binop, find its precedence. *)
+ | Some (Token.Kwd c) when Hashtbl.mem binop_precedence c ->
+ let token_prec = precedence c in
+
+ (* If this is a binop that binds at least as tightly as the current binop,
+ * consume it, otherwise we are done. *)
+ if token_prec < expr_prec then lhs else begin
+ (* Eat the binop. *)
+ Stream.junk stream;
+
+ (* Parse the unary expression after the binary operator. *)
+ let rhs = parse_unary stream in
+
+ (* Okay, we know this is a binop. *)
+ let rhs =
+ match Stream.peek stream with
+ | Some (Token.Kwd c2) ->
+ (* If BinOp binds less tightly with rhs than the operator after
+ * rhs, let the pending operator take rhs as its lhs. *)
+ let next_prec = precedence c2 in
+ if token_prec < next_prec
+ then parse_bin_rhs (token_prec + 1) rhs stream
+ else rhs
+ | _ -> rhs
+ in
+
+ (* Merge lhs/rhs. *)
+ let lhs = Ast.Binary (c, lhs, rhs) in
+ parse_bin_rhs expr_prec lhs stream
+ end
+ | _ -> lhs
+
+(* expression
+ * ::= primary binoprhs *)
+and parse_expr = parser
+ | [< lhs=parse_unary; stream >] -> parse_bin_rhs 0 lhs stream
+
+(* prototype
+ * ::= id '(' id* ')'
+ * ::= binary LETTER number? (id, id)
+ * ::= unary LETTER number? (id) *)
+let parse_prototype =
+ let rec parse_args accumulator = parser
+ | [< 'Token.Ident id; e=parse_args (id::accumulator) >] -> e
+ | [< >] -> accumulator
+ in
+ let parse_operator = parser
+ | [< 'Token.Unary >] -> "unary", 1
+ | [< 'Token.Binary >] -> "binary", 2
+ in
+ let parse_binary_precedence = parser
+ | [< 'Token.Number n >] -> int_of_float n
+ | [< >] -> 30
+ in
+ parser
+ | [< 'Token.Ident id;
+ 'Token.Kwd '(' ?? "expected '(' in prototype";
+ args=parse_args [];
+ 'Token.Kwd ')' ?? "expected ')' in prototype" >] ->
+ (* success. *)
+ Ast.Prototype (id, Array.of_list (List.rev args))
+ | [< (prefix, kind)=parse_operator;
+ 'Token.Kwd op ?? "expected an operator";
+ (* Read the precedence if present. *)
+ binary_precedence=parse_binary_precedence;
+ 'Token.Kwd '(' ?? "expected '(' in prototype";
+ args=parse_args [];
+ 'Token.Kwd ')' ?? "expected ')' in prototype" >] ->
+ let name = prefix ^ (String.make 1 op) in
+ let args = Array.of_list (List.rev args) in
+
+ (* Verify right number of arguments for operator. *)
+ if Array.length args != kind
+ then raise (Stream.Error "invalid number of operands for operator")
+ else
+ if kind == 1 then
+ Ast.Prototype (name, args)
+ else
+ Ast.BinOpPrototype (name, args, binary_precedence)
+ | [< >] ->
+ raise (Stream.Error "expected function name in prototype")
+
+(* definition ::= 'def' prototype expression *)
+let parse_definition = parser
+ | [< 'Token.Def; p=parse_prototype; e=parse_expr >] ->
+ Ast.Function (p, e)
+
+(* toplevelexpr ::= expression *)
+let parse_toplevel = parser
+ | [< e=parse_expr >] ->
+ (* Make an anonymous proto. *)
+ Ast.Function (Ast.Prototype ("", [||]), e)
+
+(* external ::= 'extern' prototype *)
+let parse_extern = parser
+ | [< 'Token.Extern; e=parse_prototype >] -> e
+
+
+
+
codegen.ml:
+
+
+(*===----------------------------------------------------------------------===
+ * Code Generation
+ *===----------------------------------------------------------------------===*)
+
+open Llvm
+
+exception Error of string
+
+let the_module = create_module "my cool jit"
+let builder = builder ()
+let named_values:(string, llvalue) Hashtbl.t = Hashtbl.create 10
+
+let rec codegen_expr = function
+ | Ast.Number n -> const_float double_type n
+ | Ast.Variable name ->
+ (try Hashtbl.find named_values name with
+ | Not_found -> raise (Error "unknown variable name"))
+ | Ast.Unary (op, operand) ->
+ let operand = codegen_expr operand in
+ let callee = "unary" ^ (String.make 1 op) in
+ let callee =
+ match lookup_function callee the_module with
+ | Some callee -> callee
+ | None -> raise (Error "unknown unary operator")
+ in
+ build_call callee [|operand|] "unop" builder
+ | Ast.Binary (op, lhs, rhs) ->
+ let lhs_val = codegen_expr lhs in
+ let rhs_val = codegen_expr rhs in
+ begin
+ match op with
+ | '+' -> build_add lhs_val rhs_val "addtmp" builder
+ | '-' -> build_sub lhs_val rhs_val "subtmp" builder
+ | '*' -> build_mul lhs_val rhs_val "multmp" builder
+ | '<' ->
+ (* Convert bool 0/1 to double 0.0 or 1.0 *)
+ let i = build_fcmp Fcmp.Ult lhs_val rhs_val "cmptmp" builder in
+ build_uitofp i double_type "booltmp" builder
+ | _ ->
+ (* If it wasn't a builtin binary operator, it must be a user defined
+ * one. Emit a call to it. *)
+ let callee = "binary" ^ (String.make 1 op) in
+ let callee =
+ match lookup_function callee the_module with
+ | Some callee -> callee
+ | None -> raise (Error "binary operator not found!")
+ in
+ build_call callee [|lhs_val; rhs_val|] "binop" builder
+ end
+ | Ast.Call (callee, args) ->
+ (* Look up the name in the module table. *)
+ let callee =
+ match lookup_function callee the_module with
+ | Some callee -> callee
+ | None -> raise (Error "unknown function referenced")
+ in
+ let params = params callee in
+
+ (* If argument mismatch error. *)
+ if Array.length params == Array.length args then () else
+ raise (Error "incorrect # arguments passed");
+ let args = Array.map codegen_expr args in
+ build_call callee args "calltmp" builder
+ | Ast.If (cond, then_, else_) ->
+ let cond = codegen_expr cond in
+
+ (* Convert condition to a bool by comparing equal to 0.0 *)
+ let zero = const_float double_type 0.0 in
+ let cond_val = build_fcmp Fcmp.One cond zero "ifcond" builder in
+
+ (* Grab the first block so that we might later add the conditional branch
+ * to it at the end of the function. *)
+ let start_bb = insertion_block builder in
+ let the_function = block_parent start_bb in
+
+ let then_bb = append_block "then" the_function in
+
+ (* Emit 'then' value. *)
+ position_at_end then_bb builder;
+ let then_val = codegen_expr then_ in
+
+ (* Codegen of 'then' can change the current block, update then_bb for the
+ * phi. We create a new name because one is used for the phi node, and the
+ * other is used for the conditional branch. *)
+ let new_then_bb = insertion_block builder in
+
+ (* Emit 'else' value. *)
+ let else_bb = append_block "else" the_function in
+ position_at_end else_bb builder;
+ let else_val = codegen_expr else_ in
+
+ (* Codegen of 'else' can change the current block, update else_bb for the
+ * phi. *)
+ let new_else_bb = insertion_block builder in
+
+ (* Emit merge block. *)
+ let merge_bb = append_block "ifcont" the_function in
+ position_at_end merge_bb builder;
+ let incoming = [(then_val, new_then_bb); (else_val, new_else_bb)] in
+ let phi = build_phi incoming "iftmp" builder in
+
+ (* Return to the start block to add the conditional branch. *)
+ position_at_end start_bb builder;
+ ignore (build_cond_br cond_val then_bb else_bb builder);
+
+ (* Set a unconditional branch at the end of the 'then' block and the
+ * 'else' block to the 'merge' block. *)
+ position_at_end new_then_bb builder; ignore (build_br merge_bb builder);
+ position_at_end new_else_bb builder; ignore (build_br merge_bb builder);
+
+ (* Finally, set the builder to the end of the merge block. *)
+ position_at_end merge_bb builder;
+
+ phi
+ | Ast.For (var_name, start, end_, step, body) ->
+ (* Emit the start code first, without 'variable' in scope. *)
+ let start_val = codegen_expr start in
+
+ (* Make the new basic block for the loop header, inserting after current
+ * block. *)
+ let preheader_bb = insertion_block builder in
+ let the_function = block_parent preheader_bb in
+ let loop_bb = append_block "loop" the_function in
+
+ (* Insert an explicit fall through from the current block to the
+ * loop_bb. *)
+ ignore (build_br loop_bb builder);
+
+ (* Start insertion in loop_bb. *)
+ position_at_end loop_bb builder;
+
+ (* Start the PHI node with an entry for start. *)
+ let variable = build_phi [(start_val, preheader_bb)] var_name builder in
+
+ (* Within the loop, the variable is defined equal to the PHI node. If it
+ * shadows an existing variable, we have to restore it, so save it
+ * now. *)
+ let old_val =
+ try Some (Hashtbl.find named_values var_name) with Not_found -> None
+ in
+ Hashtbl.add named_values var_name variable;
+
+ (* Emit the body of the loop. This, like any other expr, can change the
+ * current BB. Note that we ignore the value computed by the body, but
+ * don't allow an error *)
+ ignore (codegen_expr body);
+
+ (* Emit the step value. *)
+ let step_val =
+ match step with
+ | Some step -> codegen_expr step
+ (* If not specified, use 1.0. *)
+ | None -> const_float double_type 1.0
+ in
+
+ let next_var = build_add variable step_val "nextvar" builder in
+
+ (* Compute the end condition. *)
+ let end_cond = codegen_expr end_ in
+
+ (* Convert condition to a bool by comparing equal to 0.0. *)
+ let zero = const_float double_type 0.0 in
+ let end_cond = build_fcmp Fcmp.One end_cond zero "loopcond" builder in
+
+ (* Create the "after loop" block and insert it. *)
+ let loop_end_bb = insertion_block builder in
+ let after_bb = append_block "afterloop" the_function in
+
+ (* Insert the conditional branch into the end of loop_end_bb. *)
+ ignore (build_cond_br end_cond loop_bb after_bb builder);
+
+ (* Any new code will be inserted in after_bb. *)
+ position_at_end after_bb builder;
+
+ (* Add a new entry to the PHI node for the backedge. *)
+ add_incoming (next_var, loop_end_bb) variable;
+
+ (* Restore the unshadowed variable. *)
+ begin match old_val with
+ | Some old_val -> Hashtbl.add named_values var_name old_val
+ | None -> ()
+ end;
+
+ (* for expr always returns 0.0. *)
+ const_null double_type
+
+let codegen_proto = function
+ | Ast.Prototype (name, args) | Ast.BinOpPrototype (name, args, _) ->
+ (* Make the function type: double(double,double) etc. *)
+ let doubles = Array.make (Array.length args) double_type in
+ let ft = function_type double_type doubles in
+ let f =
+ match lookup_function name the_module with
+ | None -> declare_function name ft the_module
+
+ (* If 'f' conflicted, there was already something named 'name'. If it
+ * has a body, don't allow redefinition or reextern. *)
+ | Some f ->
+ (* If 'f' already has a body, reject this. *)
+ if block_begin f <> At_end f then
+ raise (Error "redefinition of function");
+
+ (* If 'f' took a different number of arguments, reject. *)
+ if element_type (type_of f) <> ft then
+ raise (Error "redefinition of function with different # args");
+ f
+ in
+
+ (* Set names for all arguments. *)
+ Array.iteri (fun i a ->
+ let n = args.(i) in
+ set_value_name n a;
+ Hashtbl.add named_values n a;
+ ) (params f);
+ f
+
+let codegen_func the_fpm = function
+ | Ast.Function (proto, body) ->
+ Hashtbl.clear named_values;
+ let the_function = codegen_proto proto in
+
+ (* If this is an operator, install it. *)
+ begin match proto with
+ | Ast.BinOpPrototype (name, args, prec) ->
+ let op = name.[String.length name - 1] in
+ Hashtbl.add Parser.binop_precedence op prec;
+ | _ -> ()
+ end;
+
+ (* Create a new basic block to start insertion into. *)
+ let bb = append_block "entry" the_function in
+ position_at_end bb builder;
+
+ try
+ let ret_val = codegen_expr body in
+
+ (* Finish off the function. *)
+ let _ = build_ret ret_val builder in
+
+ (* Validate the generated code, checking for consistency. *)
+ Llvm_analysis.assert_valid_function the_function;
+
+ (* Optimize the function. *)
+ let _ = PassManager.run_function the_function the_fpm in
+
+ the_function
+ with e ->
+ delete_function the_function;
+ raise e
+
+
+
+
toplevel.ml:
+
+
+(*===----------------------------------------------------------------------===
+ * Top-Level parsing and JIT Driver
+ *===----------------------------------------------------------------------===*)
+
+open Llvm
+open Llvm_executionengine
+
+(* top ::= definition | external | expression | ';' *)
+let rec main_loop the_fpm the_execution_engine stream =
+ match Stream.peek stream with
+ | None -> ()
+
+ (* ignore top-level semicolons. *)
+ | Some (Token.Kwd ';') ->
+ Stream.junk stream;
+ main_loop the_fpm the_execution_engine stream
+
+ | Some token ->
+ begin
+ try match token with
+ | Token.Def ->
+ let e = Parser.parse_definition stream in
+ print_endline "parsed a function definition.";
+ dump_value (Codegen.codegen_func the_fpm e);
+ | Token.Extern ->
+ let e = Parser.parse_extern stream in
+ print_endline "parsed an extern.";
+ dump_value (Codegen.codegen_proto e);
+ | _ ->
+ (* Evaluate a top-level expression into an anonymous function. *)
+ let e = Parser.parse_toplevel stream in
+ print_endline "parsed a top-level expr";
+ let the_function = Codegen.codegen_func the_fpm e in
+ dump_value the_function;
+
+ (* JIT the function, returning a function pointer. *)
+ let result = ExecutionEngine.run_function the_function [||]
+ the_execution_engine in
+
+ print_string "Evaluated to ";
+ print_float (GenericValue.as_float double_type result);
+ print_newline ();
+ with Stream.Error s | Codegen.Error s ->
+ (* Skip token for error recovery. *)
+ Stream.junk stream;
+ print_endline s;
+ end;
+ print_string "ready> "; flush stdout;
+ main_loop the_fpm the_execution_engine stream
+
+
+
+
toy.ml:
+
+
+(*===----------------------------------------------------------------------===
+ * Main driver code.
+ *===----------------------------------------------------------------------===*)
+
+open Llvm
+open Llvm_executionengine
+open Llvm_target
+open Llvm_scalar_opts
+
+let main () =
+ (* Install standard binary operators.
+ * 1 is the lowest precedence. *)
+ Hashtbl.add Parser.binop_precedence '<' 10;
+ Hashtbl.add Parser.binop_precedence '+' 20;
+ Hashtbl.add Parser.binop_precedence '-' 20;
+ Hashtbl.add Parser.binop_precedence '*' 40; (* highest. *)
+
+ (* Prime the first token. *)
+ print_string "ready> "; flush stdout;
+ let stream = Lexer.lex (Stream.of_channel stdin) in
+
+ (* Create the JIT. *)
+ let the_module_provider = ModuleProvider.create Codegen.the_module in
+ let the_execution_engine = ExecutionEngine.create the_module_provider in
+ let the_fpm = PassManager.create_function the_module_provider in
+
+ (* Set up the optimizer pipeline. Start with registering info about how the
+ * target lays out data structures. *)
+ TargetData.add (ExecutionEngine.target_data the_execution_engine) the_fpm;
+
+ (* Do simple "peephole" optimizations and bit-twiddling optzn. *)
+ add_instruction_combining the_fpm;
+
+ (* reassociate expressions. *)
+ add_reassociation the_fpm;
+
+ (* Eliminate Common SubExpressions. *)
+ add_gvn the_fpm;
+
+ (* Simplify the control flow graph (deleting unreachable blocks, etc). *)
+ add_cfg_simplification the_fpm;
+
+ (* Run the main "interpreter loop" now. *)
+ Toplevel.main_loop the_fpm the_execution_engine stream;
+
+ (* Print out all the generated code. *)
+ dump_module Codegen.the_module
+;;
+
+main ()
+
+
+
+
bindings.c
+
+
+#include <stdio.h>
+
+/* putchard - putchar that takes a double and returns 0. */
+extern double putchard(double X) {
+ putchar((char)X);
+ return 0;
+}
+
+/* printd - printf that takes a double prints it as "%f\n", returning 0. */
+extern double printd(double X) {
+ printf("%f\n", X);
+ return 0;
+}
+
Welcome to Chapter 7 of the "Implementing a language
+with LLVM" tutorial. In chapters 1 through 6, we've built a very
+respectable, albeit simple, functional
+programming language. In our journey, we learned some parsing techniques,
+how to build and represent an AST, how to build LLVM IR, and how to optimize
+the resultant code as well as JIT compile it.
+
+
While Kaleidoscope is interesting as a functional language, the fact that it
+is functional makes it "too easy" to generate LLVM IR for it. In particular, a
+functional language makes it very easy to build LLVM IR directly in SSA form.
+Since LLVM requires that the input code be in SSA form, this is a very nice
+property and it is often unclear to newcomers how to generate code for an
+imperative language with mutable variables.
+
+
The short (and happy) summary of this chapter is that there is no need for
+your front-end to build SSA form: LLVM provides highly tuned and well tested
+support for this, though the way it works is a bit unexpected for some.
+To understand why mutable variables cause complexities in SSA construction,
+consider this extremely simple C example:
+
+
+
+
+int G, H;
+int test(_Bool Condition) {
+ int X;
+ if (Condition)
+ X = G;
+ else
+ X = H;
+ return X;
+}
+
+
+
+
In this case, we have the variable "X", whose value depends on the path
+executed in the program. Because there are two different possible values for X
+before the return instruction, a PHI node is inserted to merge the two values.
+The LLVM IR that we want for this example looks like this:
+
+
+
+ at G = weak global i32 0 ; type of @G is i32*
+ at H = weak global i32 0 ; type of @H is i32*
+
+define i32 @test(i1 %Condition) {
+entry:
+ br i1 %Condition, label %cond_true, label %cond_false
+
+cond_true:
+ %X.0 = load i32* @G
+ br label %cond_next
+
+cond_false:
+ %X.1 = load i32* @H
+ br label %cond_next
+
+cond_next:
+ %X.2 = phi i32 [ %X.1, %cond_false ], [ %X.0, %cond_true ]
+ ret i32 %X.2
+}
+
+
+
+
In this example, the loads from the G and H global variables are explicit in
+the LLVM IR, and they live in the then/else branches of the if statement
+(cond_true/cond_false). In order to merge the incoming values, the X.2 phi node
+in the cond_next block selects the right value to use based on where control
+flow is coming from: if control flow comes from the cond_false block, X.2 gets
+the value of X.1. Alternatively, if control flow comes from cond_true, it gets
+the value of X.0. The intent of this chapter is not to explain the details of
+SSA form. For more information, see one of the many online
+references.
+
+
The question for this article is "who places the phi nodes when lowering
+assignments to mutable variables?". The issue here is that LLVM
+requires that its IR be in SSA form: there is no "non-ssa" mode for it.
+However, SSA construction requires non-trivial algorithms and data structures,
+so it is inconvenient and wasteful for every front-end to have to reproduce this
+logic.
The 'trick' here is that while LLVM does require all register values to be
+in SSA form, it does not require (or permit) memory objects to be in SSA form.
+In the example above, note that the loads from G and H are direct accesses to
+G and H: they are not renamed or versioned. This differs from some other
+compiler systems, which do try to version memory objects. In LLVM, instead of
+encoding dataflow analysis of memory into the LLVM IR, it is handled with Analysis Passes which are computed on
+demand.
+
+
+With this in mind, the high-level idea is that we want to make a stack variable
+(which lives in memory, because it is on the stack) for each mutable object in
+a function. To take advantage of this trick, we need to talk about how LLVM
+represents stack variables.
+
+
+
In LLVM, all memory accesses are explicit with load/store instructions, and
+it is carefully designed not to have (or need) an "address-of" operator. Notice
+how the type of the @G/@H global variables is actually "i32*" even though the
+variable is defined as "i32". What this means is that @G defines space
+for an i32 in the global data area, but its name actually refers to the
+address for that space. Stack variables work the same way, except that instead of
+being declared with global variable definitions, they are declared with the
+LLVM alloca instruction:
+
+
+
+define i32 @example() {
+entry:
+ %X = alloca i32 ; type of %X is i32*.
+ ...
+ %tmp = load i32* %X ; load the stack value %X from the stack.
+ %tmp2 = add i32 %tmp, 1 ; increment it
+ store i32 %tmp2, i32* %X ; store it back
+ ...
+
+
+
+
This code shows an example of how you can declare and manipulate a stack
+variable in the LLVM IR. Stack memory allocated with the alloca instruction is
+fully general: you can pass the address of the stack slot to functions, you can
+store it in other variables, etc. In our example above, we could rewrite the
+example to use the alloca technique to avoid using a PHI node:
+
+
+
+ at G = weak global i32 0 ; type of @G is i32*
+ at H = weak global i32 0 ; type of @H is i32*
+
+define i32 @test(i1 %Condition) {
+entry:
+ %X = alloca i32 ; type of %X is i32*.
+ br i1 %Condition, label %cond_true, label %cond_false
+
+cond_true:
+ %X.0 = load i32* @G
+ store i32 %X.0, i32* %X ; Update X
+ br label %cond_next
+
+cond_false:
+ %X.1 = load i32* @H
+ store i32 %X.1, i32* %X ; Update X
+ br label %cond_next
+
+cond_next:
+ %X.2 = load i32* %X ; Read X
+ ret i32 %X.2
+}
+
+
+
+
With this, we have discovered a way to handle arbitrary mutable variables
+without the need to create Phi nodes at all:
+
+
+
Each mutable variable becomes a stack allocation.
+
Each read of the variable becomes a load from the stack.
+
Each update of the variable becomes a store to the stack.
+
Taking the address of a variable just uses the stack address directly.
+
+
+
While this solution has solved our immediate problem, it introduced another
+one: we have now apparently introduced a lot of stack traffic for very simple
+and common operations, a major performance problem. Fortunately for us, the
+LLVM optimizer has a highly-tuned optimization pass named "mem2reg" that handles
+this case, promoting allocas like this into SSA registers, inserting Phi nodes
+as appropriate. If you run this example through the pass, for example, you'll
+get:
The mem2reg pass implements the standard "iterated dominance frontier"
+algorithm for constructing SSA form and has a number of optimizations that speed
+up (very common) degenerate cases. The mem2reg optimization pass is the answer
+to dealing with mutable variables, and we highly recommend that you depend on
+it. Note that mem2reg only works on variables in certain circumstances:
+
+
+
mem2reg is alloca-driven: it looks for allocas and if it can handle them, it
+promotes them. It does not apply to global variables or heap allocations.
+
+
mem2reg only looks for alloca instructions in the entry block of the
+function. Being in the entry block guarantees that the alloca is only executed
+once, which makes analysis simpler.
+
+
mem2reg only promotes allocas whose uses are direct loads and stores. If
+the address of the stack object is passed to a function, or if any funny pointer
+arithmetic is involved, the alloca will not be promoted.
+
+
mem2reg only works on allocas of first class
+values (such as pointers, scalars and vectors), and only if the array size
+of the allocation is 1 (or missing in the .ll file). mem2reg is not capable of
+promoting structs or arrays to registers. Note that the "scalarrepl" pass is
+more powerful and can promote structs, "unions", and arrays in many cases.
+
+
+
+
+All of these properties are easy to satisfy for most imperative languages, and
+we'll illustrate it below with Kaleidoscope. The final question you may be
+asking is: should I bother with this nonsense for my front-end? Wouldn't it be
+better if I just did SSA construction directly, avoiding use of the mem2reg
+optimization pass? In short, we strongly recommend that you use this technique
+for building SSA form, unless there is an extremely good reason not to. Using
+this technique is:
+
+
+
Proven and well tested: llvm-gcc and clang both use this technique for local
+mutable variables. As such, the most common clients of LLVM are using this to
+handle a bulk of their variables. You can be sure that bugs are found fast and
+fixed early.
+
+
Extremely Fast: mem2reg has a number of special cases that make it fast in
+common cases as well as fully general. For example, it has fast-paths for
+variables that are only used in a single block, variables that only have one
+assignment point, good heuristics to avoid insertion of unneeded phi nodes, etc.
+
+
+
Needed for debug info generation:
+Debug information in LLVM relies on having the address of the variable
+exposed so that debug info can be attached to it. This technique dovetails
+very naturally with this style of debug info.
+
+
+
If nothing else, this makes it much easier to get your front-end up and
+running, and is very simple to implement. Lets extend Kaleidoscope with mutable
+variables now!
+
Now that we know the sort of problem we want to tackle, lets see what this
+looks like in the context of our little Kaleidoscope language. We're going to
+add two features:
+
+
+
The ability to mutate variables with the '=' operator.
+
The ability to define new variables.
+
+
+
While the first item is really what this is about, we only have variables
+for incoming arguments as well as for induction variables, and redefining those only
+goes so far :). Also, the ability to define new variables is a
+useful thing regardless of whether you will be mutating them. Here's a
+motivating example that shows how we could use these:
+
+
+
+# Define ':' for sequencing: as a low-precedence operator that ignores operands
+# and just returns the RHS.
+def binary : 1 (x y) y;
+
+# Recursive fib, we could do this before.
+def fib(x)
+ if (x < 3) then
+ 1
+ else
+ fib(x-1)+fib(x-2);
+
+# Iterative fib.
+def fibi(x)
+ var a = 1, b = 1, c in
+ (for i = 3, i < x in
+ c = a + b :
+ a = b :
+ b = c) :
+ b;
+
+# Call it.
+fibi(10);
+
+
+
+
+In order to mutate variables, we have to change our existing variables to use
+the "alloca trick". Once we have that, we'll add our new operator, then extend
+Kaleidoscope to support new variable definitions.
+
+The symbol table in Kaleidoscope is managed at code generation time by the
+'named_values' map. This map currently keeps track of the LLVM
+"Value*" that holds the double value for the named variable. In order to
+support mutation, we need to change this slightly, so that it
+named_values holds the memory location of the variable in
+question. Note that this change is a refactoring: it changes the structure of
+the code, but does not (by itself) change the behavior of the compiler. All of
+these changes are isolated in the Kaleidoscope code generator.
+
+
+At this point in Kaleidoscope's development, it only supports variables for two
+things: incoming arguments to functions and the induction variable of 'for'
+loops. For consistency, we'll allow mutation of these variables in addition to
+other user-defined variables. This means that these will both need memory
+locations.
+
+
+
To start our transformation of Kaleidoscope, we'll change the
+named_values map so that it maps to AllocaInst* instead of Value*.
+Once we do this, the C++ compiler will tell us what parts of the code we need to
+update:
+
+
Note: the ocaml bindings currently model both Value*s and
+AllocInst*s as Llvm.llvalues, but this may change in the
+future to be more type safe.
Also, since we will need to create these alloca's, we'll use a helper
+function that ensures that the allocas are created in the entry block of the
+function:
+
+
+
+(* Create an alloca instruction in the entry block of the function. This
+ * is used for mutable variables etc. *)
+let create_entry_block_alloca the_function var_name =
+ let builder = builder_at (instr_begin (entry_block the_function)) in
+ build_alloca double_type var_name builder
+
+
+
+
This funny looking code creates an Llvm.llbuilder object that is
+pointing at the first instruction of the entry block. It then creates an alloca
+with the expected name and returns it. Because all values in Kaleidoscope are
+doubles, there is no need to pass in a type to use.
+
+
With this in place, the first functionality change we want to make is to
+variable references. In our new scheme, variables live on the stack, so code
+generating a reference to them actually needs to produce a load from the stack
+slot:
+
+
+
+let rec codegen_expr = function
+ ...
+ | Ast.Variable name ->
+ let v = try Hashtbl.find named_values name with
+ | Not_found -> raise (Error "unknown variable name")
+ in
+ (* Load the value. *)
+ build_load v name builder
+
+
+
+
As you can see, this is pretty straightforward. Now we need to update the
+things that define the variables to set up the alloca. We'll start with
+codegen_expr Ast.For ... (see the full code listing
+for the unabridged code):
+
+
+
+ | Ast.For (var_name, start, end_, step, body) ->
+ let the_function = block_parent (insertion_block builder) in
+
+ (* Create an alloca for the variable in the entry block. *)
+ let alloca = create_entry_block_alloca the_function var_name in
+
+ (* Emit the start code first, without 'variable' in scope. *)
+ let start_val = codegen_expr start in
+
+ (* Store the value into the alloca. *)
+ ignore(build_store start_val alloca builder);
+
+ ...
+
+ (* Within the loop, the variable is defined equal to the PHI node. If it
+ * shadows an existing variable, we have to restore it, so save it
+ * now. *)
+ let old_val =
+ try Some (Hashtbl.find named_values var_name) with Not_found -> None
+ in
+ Hashtbl.add named_values var_name alloca;
+
+ ...
+
+ (* Compute the end condition. *)
+ let end_cond = codegen_expr end_ in
+
+ (* Reload, increment, and restore the alloca. This handles the case where
+ * the body of the loop mutates the variable. *)
+ let cur_var = build_load alloca var_name builder in
+ let next_var = build_add cur_var step_val "nextvar" builder in
+ ignore(build_store next_var alloca builder);
+ ...
+
+
+
+
This code is virtually identical to the code before we allowed mutable variables.
+The big difference is that we no longer have to construct a PHI node, and we use
+load/store to access the variable as needed.
+
+
To support mutable argument variables, we need to also make allocas for them.
+The code for this is also pretty simple:
+
+
+
+(* Create an alloca for each argument and register the argument in the symbol
+ * table so that references to it will succeed. *)
+let create_argument_allocas the_function proto =
+ let args = match proto with
+ | Ast.Prototype (_, args) | Ast.BinOpPrototype (_, args, _) -> args
+ in
+ Array.iteri (fun i ai ->
+ let var_name = args.(i) in
+ (* Create an alloca for this variable. *)
+ let alloca = create_entry_block_alloca the_function var_name in
+
+ (* Store the initial value into the alloca. *)
+ ignore(build_store ai alloca builder);
+
+ (* Add arguments to variable symbol table. *)
+ Hashtbl.add named_values var_name alloca;
+ ) (params the_function)
+
+
+
+
For each argument, we make an alloca, store the input value to the function
+into the alloca, and register the alloca as the memory location for the
+argument. This method gets invoked by Codegen.codegen_func right after
+it sets up the entry block for the function.
+
+
The final missing piece is adding the mem2reg pass, which allows us to get
+good codegen once again:
+
+
+
+let main () =
+ ...
+ let the_fpm = PassManager.create_function the_module_provider in
+
+ (* Set up the optimizer pipeline. Start with registering info about how the
+ * target lays out data structures. *)
+ TargetData.add (ExecutionEngine.target_data the_execution_engine) the_fpm;
+
+ (* Promote allocas to registers. *)
+ add_memory_to_register_promotion the_fpm;
+
+ (* Do simple "peephole" optimizations and bit-twiddling optzn. *)
+ add_instruction_combining the_fpm;
+
+ (* reassociate expressions. *)
+ add_reassociation the_fpm;
+
+
+
+
It is interesting to see what the code looks like before and after the
+mem2reg optimization runs. For example, this is the before/after code for our
+recursive fib function. Before the optimization:
Here there is only one variable (x, the input argument) but you can still
+see the extremely simple-minded code generation strategy we are using. In the
+entry block, an alloca is created, and the initial input value is stored into
+it. Each reference to the variable does a reload from the stack. Also, note
+that we didn't modify the if/then/else expression, so it still inserts a PHI
+node. While we could make an alloca for it, it is actually easier to create a
+PHI node for it, so we still just make the PHI.
This is a trivial case for mem2reg, since there are no redefinitions of the
+variable. The point of showing this is to calm your tension about inserting
+such blatent inefficiencies :).
Here we see that the simplifycfg pass decided to clone the return instruction
+into the end of the 'else' block. This allowed it to eliminate some branches
+and the PHI node.
+
+
Now that all symbol table references are updated to use stack variables,
+we'll add the assignment operator.
With our current framework, adding a new assignment operator is really
+simple. We will parse it just like any other binary operator, but handle it
+internally (instead of allowing the user to define it). The first step is to
+set a precedence:
Now that the parser knows the precedence of the binary operator, it takes
+care of all the parsing and AST generation. We just need to implement codegen
+for the assignment operator. This looks like:
+
+
+
+let rec codegen_expr = function
+ begin match op with
+ | '=' ->
+ (* Special case '=' because we don't want to emit the LHS as an
+ * expression. *)
+ let name =
+ match lhs with
+ | Ast.Variable name -> name
+ | _ -> raise (Error "destination of '=' must be a variable")
+ in
+
+
+
+
Unlike the rest of the binary operators, our assignment operator doesn't
+follow the "emit LHS, emit RHS, do computation" model. As such, it is handled
+as a special case before the other binary operators are handled. The other
+strange thing is that it requires the LHS to be a variable. It is invalid to
+have "(x+1) = expr" - only things like "x = expr" are allowed.
+
+
+
+
+
+ (* Codegen the rhs. *)
+ let val_ = codegen_expr rhs in
+
+ (* Lookup the name. *)
+ let variable = try Hashtbl.find named_values name with
+ | Not_found -> raise (Error "unknown variable name")
+ in
+ ignore(build_store val_ variable builder);
+ val_
+ | _ ->
+ ...
+
+
+
+
Once we have the variable, codegen'ing the assignment is straightforward:
+we emit the RHS of the assignment, create a store, and return the computed
+value. Returning a value allows for chained assignments like "X = (Y = Z)".
+
+
Now that we have an assignment operator, we can mutate loop variables and
+arguments. For example, we can now run code like this:
+
+
+
+# Function to print a double.
+extern printd(x);
+
+# Define ':' for sequencing: as a low-precedence operator that ignores operands
+# and just returns the RHS.
+def binary : 1 (x y) y;
+
+def test(x)
+ printd(x) :
+ x = 4 :
+ printd(x);
+
+test(123);
+
+
+
+
When run, this example prints "123" and then "4", showing that we did
+actually mutate the value! Okay, we have now officially implemented our goal:
+getting this to work requires SSA construction in the general case. However,
+to be really useful, we want the ability to define our own local variables, lets
+add this next!
+
Adding var/in is just like any other other extensions we made to
+Kaleidoscope: we extend the lexer, the parser, the AST and the code generator.
+The first step for adding our new 'var/in' construct is to extend the lexer.
+As before, this is pretty trivial, the code looks like this:
The next step is to define the AST node that we will construct. For var/in,
+it looks like this:
+
+
+
+type expr =
+ ...
+ (* variant for var/in. *)
+ | Var of (string * expr option) array * expr
+ ...
+
+
+
+
var/in allows a list of names to be defined all at once, and each name can
+optionally have an initializer value. As such, we capture this information in
+the VarNames vector. Also, var/in has a body, this body is allowed to access
+the variables defined by the var/in.
+
+
With this in place, we can define the parser pieces. The first thing we do
+is add it as a primary expression:
+
+
+
+(* primary
+ * ::= identifier
+ * ::= numberexpr
+ * ::= parenexpr
+ * ::= ifexpr
+ * ::= forexpr
+ * ::= varexpr *)
+let rec parse_primary = parser
+ ...
+ (* varexpr
+ * ::= 'var' identifier ('=' expression?
+ * (',' identifier ('=' expression)?)* 'in' expression *)
+ | [< 'Token.Var;
+ (* At least one variable name is required. *)
+ 'Token.Ident id ?? "expected identifier after var";
+ init=parse_var_init;
+ var_names=parse_var_names [(id, init)];
+ (* At this point, we have to have 'in'. *)
+ 'Token.In ?? "expected 'in' keyword after 'var'";
+ body=parse_expr >] ->
+ Ast.Var (Array.of_list (List.rev var_names), body)
+
+...
+
+and parse_var_init = parser
+ (* read in the optional initializer. *)
+ | [< 'Token.Kwd '='; e=parse_expr >] -> Some e
+ | [< >] -> None
+
+and parse_var_names accumulator = parser
+ | [< 'Token.Kwd ',';
+ 'Token.Ident id ?? "expected identifier list after var";
+ init=parse_var_init;
+ e=parse_var_names ((id, init) :: accumulator) >] -> e
+ | [< >] -> accumulator
+
+
+
+
Now that we can parse and represent the code, we need to support emission of
+LLVM IR for it. This code starts out with:
+
+
+
+let rec codegen_expr = function
+ ...
+ | Ast.Var (var_names, body)
+ let old_bindings = ref [] in
+
+ let the_function = block_parent (insertion_block builder) in
+
+ (* Register all variables and emit their initializer. *)
+ Array.iter (fun (var_name, init) ->
+
+
+
+
Basically it loops over all the variables, installing them one at a time.
+For each variable we put into the symbol table, we remember the previous value
+that we replace in OldBindings.
+
+
+
+ (* Emit the initializer before adding the variable to scope, this
+ * prevents the initializer from referencing the variable itself, and
+ * permits stuff like this:
+ * var a = 1 in
+ * var a = a in ... # refers to outer 'a'. *)
+ let init_val =
+ match init with
+ | Some init -> codegen_expr init
+ (* If not specified, use 0.0. *)
+ | None -> const_float double_type 0.0
+ in
+
+ let alloca = create_entry_block_alloca the_function var_name in
+ ignore(build_store init_val alloca builder);
+
+ (* Remember the old variable binding so that we can restore the binding
+ * when we unrecurse. *)
+
+ begin
+ try
+ let old_value = Hashtbl.find named_values var_name in
+ old_bindings := (var_name, old_value) :: !old_bindings;
+ with Not_found > ()
+ end;
+
+ (* Remember this binding. *)
+ Hashtbl.add named_values var_name alloca;
+ ) var_names;
+
+
+
+
There are more comments here than code. The basic idea is that we emit the
+initializer, create the alloca, then update the symbol table to point to it.
+Once all the variables are installed in the symbol table, we evaluate the body
+of the var/in expression:
+
+
+
+ (* Codegen the body, now that all vars are in scope. *)
+ let body_val = codegen_expr body in
+
+
+
+
Finally, before returning, we restore the previous variable bindings:
+
+
+
+ (* Pop all our variables from scope. *)
+ List.iter (fun (var_name, old_value) ->
+ Hashtbl.add named_values var_name old_value
+ ) !old_bindings;
+
+ (* Return the body computation. *)
+ body_val
+
+
+
+
The end result of all of this is that we get properly scoped variable
+definitions, and we even (trivially) allow mutation of them :).
+
+
With this, we completed what we set out to do. Our nice iterative fib
+example from the intro compiles and runs just fine. The mem2reg pass optimizes
+all of our stack variables into SSA registers, inserting PHI nodes where needed,
+and our front-end remains simple: no "iterated dominance frontier" computation
+anywhere in sight.
+(*===----------------------------------------------------------------------===
+ * Lexer Tokens
+ *===----------------------------------------------------------------------===*)
+
+(* The lexer returns these 'Kwd' if it is an unknown character, otherwise one of
+ * these others for known things. *)
+type token =
+ (* commands *)
+ | Def | Extern
+
+ (* primary *)
+ | Ident of string | Number of float
+
+ (* unknown *)
+ | Kwd of char
+
+ (* control *)
+ | If | Then | Else
+ | For | In
+
+ (* operators *)
+ | Binary | Unary
+
+ (* var definition *)
+ | Var
+
+(*===----------------------------------------------------------------------===
+ * Abstract Syntax Tree (aka Parse Tree)
+ *===----------------------------------------------------------------------===*)
+
+(* expr - Base type for all expression nodes. *)
+type expr =
+ (* variant for numeric literals like "1.0". *)
+ | Number of float
+
+ (* variant for referencing a variable, like "a". *)
+ | Variable of string
+
+ (* variant for a unary operator. *)
+ | Unary of char * expr
+
+ (* variant for a binary operator. *)
+ | Binary of char * expr * expr
+
+ (* variant for function calls. *)
+ | Call of string * expr array
+
+ (* variant for if/then/else. *)
+ | If of expr * expr * expr
+
+ (* variant for for/in. *)
+ | For of string * expr * expr * expr option * expr
+
+ (* variant for var/in. *)
+ | Var of (string * expr option) array * expr
+
+(* proto - This type represents the "prototype" for a function, which captures
+ * its name, and its argument names (thus implicitly the number of arguments the
+ * function takes). *)
+type proto =
+ | Prototype of string * string array
+ | BinOpPrototype of string * string array * int
+
+(* func - This type represents a function definition itself. *)
+type func = Function of proto * expr
+
+
+
+
parser.ml:
+
+
+(*===---------------------------------------------------------------------===
+ * Parser
+ *===---------------------------------------------------------------------===*)
+
+(* binop_precedence - This holds the precedence for each binary operator that is
+ * defined *)
+let binop_precedence:(char, int) Hashtbl.t = Hashtbl.create 10
+
+(* precedence - Get the precedence of the pending binary operator token. *)
+let precedence c = try Hashtbl.find binop_precedence c with Not_found -> -1
+
+(* primary
+ * ::= identifier
+ * ::= numberexpr
+ * ::= parenexpr
+ * ::= ifexpr
+ * ::= forexpr
+ * ::= varexpr *)
+let rec parse_primary = parser
+ (* numberexpr ::= number *)
+ | [< 'Token.Number n >] -> Ast.Number n
+
+ (* parenexpr ::= '(' expression ')' *)
+ | [< 'Token.Kwd '('; e=parse_expr; 'Token.Kwd ')' ?? "expected ')'" >] -> e
+
+ (* identifierexpr
+ * ::= identifier
+ * ::= identifier '(' argumentexpr ')' *)
+ | [< 'Token.Ident id; stream >] ->
+ let rec parse_args accumulator = parser
+ | [< e=parse_expr; stream >] ->
+ begin parser
+ | [< 'Token.Kwd ','; e=parse_args (e :: accumulator) >] -> e
+ | [< >] -> e :: accumulator
+ end stream
+ | [< >] -> accumulator
+ in
+ let rec parse_ident id = parser
+ (* Call. *)
+ | [< 'Token.Kwd '(';
+ args=parse_args [];
+ 'Token.Kwd ')' ?? "expected ')'">] ->
+ Ast.Call (id, Array.of_list (List.rev args))
+
+ (* Simple variable ref. *)
+ | [< >] -> Ast.Variable id
+ in
+ parse_ident id stream
+
+ (* ifexpr ::= 'if' expr 'then' expr 'else' expr *)
+ | [< 'Token.If; c=parse_expr;
+ 'Token.Then ?? "expected 'then'"; t=parse_expr;
+ 'Token.Else ?? "expected 'else'"; e=parse_expr >] ->
+ Ast.If (c, t, e)
+
+ (* forexpr
+ ::= 'for' identifier '=' expr ',' expr (',' expr)? 'in' expression *)
+ | [< 'Token.For;
+ 'Token.Ident id ?? "expected identifier after for";
+ 'Token.Kwd '=' ?? "expected '=' after for";
+ stream >] ->
+ begin parser
+ | [<
+ start=parse_expr;
+ 'Token.Kwd ',' ?? "expected ',' after for";
+ end_=parse_expr;
+ stream >] ->
+ let step =
+ begin parser
+ | [< 'Token.Kwd ','; step=parse_expr >] -> Some step
+ | [< >] -> None
+ end stream
+ in
+ begin parser
+ | [< 'Token.In; body=parse_expr >] ->
+ Ast.For (id, start, end_, step, body)
+ | [< >] ->
+ raise (Stream.Error "expected 'in' after for")
+ end stream
+ | [< >] ->
+ raise (Stream.Error "expected '=' after for")
+ end stream
+
+ (* varexpr
+ * ::= 'var' identifier ('=' expression?
+ * (',' identifier ('=' expression)?)* 'in' expression *)
+ | [< 'Token.Var;
+ (* At least one variable name is required. *)
+ 'Token.Ident id ?? "expected identifier after var";
+ init=parse_var_init;
+ var_names=parse_var_names [(id, init)];
+ (* At this point, we have to have 'in'. *)
+ 'Token.In ?? "expected 'in' keyword after 'var'";
+ body=parse_expr >] ->
+ Ast.Var (Array.of_list (List.rev var_names), body)
+
+ | [< >] -> raise (Stream.Error "unknown token when expecting an expression.")
+
+(* unary
+ * ::= primary
+ * ::= '!' unary *)
+and parse_unary = parser
+ (* If this is a unary operator, read it. *)
+ | [< 'Token.Kwd op when op != '(' && op != ')'; operand=parse_expr >] ->
+ Ast.Unary (op, operand)
+
+ (* If the current token is not an operator, it must be a primary expr. *)
+ | [< stream >] -> parse_primary stream
+
+(* binoprhs
+ * ::= ('+' primary)* *)
+and parse_bin_rhs expr_prec lhs stream =
+ match Stream.peek stream with
+ (* If this is a binop, find its precedence. *)
+ | Some (Token.Kwd c) when Hashtbl.mem binop_precedence c ->
+ let token_prec = precedence c in
+
+ (* If this is a binop that binds at least as tightly as the current binop,
+ * consume it, otherwise we are done. *)
+ if token_prec < expr_prec then lhs else begin
+ (* Eat the binop. *)
+ Stream.junk stream;
+
+ (* Parse the primary expression after the binary operator. *)
+ let rhs = parse_unary stream in
+
+ (* Okay, we know this is a binop. *)
+ let rhs =
+ match Stream.peek stream with
+ | Some (Token.Kwd c2) ->
+ (* If BinOp binds less tightly with rhs than the operator after
+ * rhs, let the pending operator take rhs as its lhs. *)
+ let next_prec = precedence c2 in
+ if token_prec < next_prec
+ then parse_bin_rhs (token_prec + 1) rhs stream
+ else rhs
+ | _ -> rhs
+ in
+
+ (* Merge lhs/rhs. *)
+ let lhs = Ast.Binary (c, lhs, rhs) in
+ parse_bin_rhs expr_prec lhs stream
+ end
+ | _ -> lhs
+
+and parse_var_init = parser
+ (* read in the optional initializer. *)
+ | [< 'Token.Kwd '='; e=parse_expr >] -> Some e
+ | [< >] -> None
+
+and parse_var_names accumulator = parser
+ | [< 'Token.Kwd ',';
+ 'Token.Ident id ?? "expected identifier list after var";
+ init=parse_var_init;
+ e=parse_var_names ((id, init) :: accumulator) >] -> e
+ | [< >] -> accumulator
+
+(* expression
+ * ::= primary binoprhs *)
+and parse_expr = parser
+ | [< lhs=parse_unary; stream >] -> parse_bin_rhs 0 lhs stream
+
+(* prototype
+ * ::= id '(' id* ')'
+ * ::= binary LETTER number? (id, id)
+ * ::= unary LETTER number? (id) *)
+let parse_prototype =
+ let rec parse_args accumulator = parser
+ | [< 'Token.Ident id; e=parse_args (id::accumulator) >] -> e
+ | [< >] -> accumulator
+ in
+ let parse_operator = parser
+ | [< 'Token.Unary >] -> "unary", 1
+ | [< 'Token.Binary >] -> "binary", 2
+ in
+ let parse_binary_precedence = parser
+ | [< 'Token.Number n >] -> int_of_float n
+ | [< >] -> 30
+ in
+ parser
+ | [< 'Token.Ident id;
+ 'Token.Kwd '(' ?? "expected '(' in prototype";
+ args=parse_args [];
+ 'Token.Kwd ')' ?? "expected ')' in prototype" >] ->
+ (* success. *)
+ Ast.Prototype (id, Array.of_list (List.rev args))
+ | [< (prefix, kind)=parse_operator;
+ 'Token.Kwd op ?? "expected an operator";
+ (* Read the precedence if present. *)
+ binary_precedence=parse_binary_precedence;
+ 'Token.Kwd '(' ?? "expected '(' in prototype";
+ args=parse_args [];
+ 'Token.Kwd ')' ?? "expected ')' in prototype" >] ->
+ let name = prefix ^ (String.make 1 op) in
+ let args = Array.of_list (List.rev args) in
+
+ (* Verify right number of arguments for operator. *)
+ if Array.length args != kind
+ then raise (Stream.Error "invalid number of operands for operator")
+ else
+ if kind == 1 then
+ Ast.Prototype (name, args)
+ else
+ Ast.BinOpPrototype (name, args, binary_precedence)
+ | [< >] ->
+ raise (Stream.Error "expected function name in prototype")
+
+(* definition ::= 'def' prototype expression *)
+let parse_definition = parser
+ | [< 'Token.Def; p=parse_prototype; e=parse_expr >] ->
+ Ast.Function (p, e)
+
+(* toplevelexpr ::= expression *)
+let parse_toplevel = parser
+ | [< e=parse_expr >] ->
+ (* Make an anonymous proto. *)
+ Ast.Function (Ast.Prototype ("", [||]), e)
+
+(* external ::= 'extern' prototype *)
+let parse_extern = parser
+ | [< 'Token.Extern; e=parse_prototype >] -> e
+
+
+
+
codegen.ml:
+
+
+(*===----------------------------------------------------------------------===
+ * Code Generation
+ *===----------------------------------------------------------------------===*)
+
+open Llvm
+
+exception Error of string
+
+let the_module = create_module "my cool jit"
+let builder = builder ()
+let named_values:(string, llvalue) Hashtbl.t = Hashtbl.create 10
+
+(* Create an alloca instruction in the entry block of the function. This
+ * is used for mutable variables etc. *)
+let create_entry_block_alloca the_function var_name =
+ let builder = builder_at (instr_begin (entry_block the_function)) in
+ build_alloca double_type var_name builder
+
+let rec codegen_expr = function
+ | Ast.Number n -> const_float double_type n
+ | Ast.Variable name ->
+ let v = try Hashtbl.find named_values name with
+ | Not_found -> raise (Error "unknown variable name")
+ in
+ (* Load the value. *)
+ build_load v name builder
+ | Ast.Unary (op, operand) ->
+ let operand = codegen_expr operand in
+ let callee = "unary" ^ (String.make 1 op) in
+ let callee =
+ match lookup_function callee the_module with
+ | Some callee -> callee
+ | None -> raise (Error "unknown unary operator")
+ in
+ build_call callee [|operand|] "unop" builder
+ | Ast.Binary (op, lhs, rhs) ->
+ begin match op with
+ | '=' ->
+ (* Special case '=' because we don't want to emit the LHS as an
+ * expression. *)
+ let name =
+ match lhs with
+ | Ast.Variable name -> name
+ | _ -> raise (Error "destination of '=' must be a variable")
+ in
+
+ (* Codegen the rhs. *)
+ let val_ = codegen_expr rhs in
+
+ (* Lookup the name. *)
+ let variable = try Hashtbl.find named_values name with
+ | Not_found -> raise (Error "unknown variable name")
+ in
+ ignore(build_store val_ variable builder);
+ val_
+ | _ ->
+ let lhs_val = codegen_expr lhs in
+ let rhs_val = codegen_expr rhs in
+ begin
+ match op with
+ | '+' -> build_add lhs_val rhs_val "addtmp" builder
+ | '-' -> build_sub lhs_val rhs_val "subtmp" builder
+ | '*' -> build_mul lhs_val rhs_val "multmp" builder
+ | '<' ->
+ (* Convert bool 0/1 to double 0.0 or 1.0 *)
+ let i = build_fcmp Fcmp.Ult lhs_val rhs_val "cmptmp" builder in
+ build_uitofp i double_type "booltmp" builder
+ | _ ->
+ (* If it wasn't a builtin binary operator, it must be a user defined
+ * one. Emit a call to it. *)
+ let callee = "binary" ^ (String.make 1 op) in
+ let callee =
+ match lookup_function callee the_module with
+ | Some callee -> callee
+ | None -> raise (Error "binary operator not found!")
+ in
+ build_call callee [|lhs_val; rhs_val|] "binop" builder
+ end
+ end
+ | Ast.Call (callee, args) ->
+ (* Look up the name in the module table. *)
+ let callee =
+ match lookup_function callee the_module with
+ | Some callee -> callee
+ | None -> raise (Error "unknown function referenced")
+ in
+ let params = params callee in
+
+ (* If argument mismatch error. *)
+ if Array.length params == Array.length args then () else
+ raise (Error "incorrect # arguments passed");
+ let args = Array.map codegen_expr args in
+ build_call callee args "calltmp" builder
+ | Ast.If (cond, then_, else_) ->
+ let cond = codegen_expr cond in
+
+ (* Convert condition to a bool by comparing equal to 0.0 *)
+ let zero = const_float double_type 0.0 in
+ let cond_val = build_fcmp Fcmp.One cond zero "ifcond" builder in
+
+ (* Grab the first block so that we might later add the conditional branch
+ * to it at the end of the function. *)
+ let start_bb = insertion_block builder in
+ let the_function = block_parent start_bb in
+
+ let then_bb = append_block "then" the_function in
+
+ (* Emit 'then' value. *)
+ position_at_end then_bb builder;
+ let then_val = codegen_expr then_ in
+
+ (* Codegen of 'then' can change the current block, update then_bb for the
+ * phi. We create a new name because one is used for the phi node, and the
+ * other is used for the conditional branch. *)
+ let new_then_bb = insertion_block builder in
+
+ (* Emit 'else' value. *)
+ let else_bb = append_block "else" the_function in
+ position_at_end else_bb builder;
+ let else_val = codegen_expr else_ in
+
+ (* Codegen of 'else' can change the current block, update else_bb for the
+ * phi. *)
+ let new_else_bb = insertion_block builder in
+
+ (* Emit merge block. *)
+ let merge_bb = append_block "ifcont" the_function in
+ position_at_end merge_bb builder;
+ let incoming = [(then_val, new_then_bb); (else_val, new_else_bb)] in
+ let phi = build_phi incoming "iftmp" builder in
+
+ (* Return to the start block to add the conditional branch. *)
+ position_at_end start_bb builder;
+ ignore (build_cond_br cond_val then_bb else_bb builder);
+
+ (* Set a unconditional branch at the end of the 'then' block and the
+ * 'else' block to the 'merge' block. *)
+ position_at_end new_then_bb builder; ignore (build_br merge_bb builder);
+ position_at_end new_else_bb builder; ignore (build_br merge_bb builder);
+
+ (* Finally, set the builder to the end of the merge block. *)
+ position_at_end merge_bb builder;
+
+ phi
+ | Ast.For (var_name, start, end_, step, body) ->
+ (* Output this as:
+ * var = alloca double
+ * ...
+ * start = startexpr
+ * store start -> var
+ * goto loop
+ * loop:
+ * ...
+ * bodyexpr
+ * ...
+ * loopend:
+ * step = stepexpr
+ * endcond = endexpr
+ *
+ * curvar = load var
+ * nextvar = curvar + step
+ * store nextvar -> var
+ * br endcond, loop, endloop
+ * outloop: *)
+
+ let the_function = block_parent (insertion_block builder) in
+
+ (* Create an alloca for the variable in the entry block. *)
+ let alloca = create_entry_block_alloca the_function var_name in
+
+ (* Emit the start code first, without 'variable' in scope. *)
+ let start_val = codegen_expr start in
+
+ (* Store the value into the alloca. *)
+ ignore(build_store start_val alloca builder);
+
+ (* Make the new basic block for the loop header, inserting after current
+ * block. *)
+ let loop_bb = append_block "loop" the_function in
+
+ (* Insert an explicit fall through from the current block to the
+ * loop_bb. *)
+ ignore (build_br loop_bb builder);
+
+ (* Start insertion in loop_bb. *)
+ position_at_end loop_bb builder;
+
+ (* Within the loop, the variable is defined equal to the PHI node. If it
+ * shadows an existing variable, we have to restore it, so save it
+ * now. *)
+ let old_val =
+ try Some (Hashtbl.find named_values var_name) with Not_found -> None
+ in
+ Hashtbl.add named_values var_name alloca;
+
+ (* Emit the body of the loop. This, like any other expr, can change the
+ * current BB. Note that we ignore the value computed by the body, but
+ * don't allow an error *)
+ ignore (codegen_expr body);
+
+ (* Emit the step value. *)
+ let step_val =
+ match step with
+ | Some step -> codegen_expr step
+ (* If not specified, use 1.0. *)
+ | None -> const_float double_type 1.0
+ in
+
+ (* Compute the end condition. *)
+ let end_cond = codegen_expr end_ in
+
+ (* Reload, increment, and restore the alloca. This handles the case where
+ * the body of the loop mutates the variable. *)
+ let cur_var = build_load alloca var_name builder in
+ let next_var = build_add cur_var step_val "nextvar" builder in
+ ignore(build_store next_var alloca builder);
+
+ (* Convert condition to a bool by comparing equal to 0.0. *)
+ let zero = const_float double_type 0.0 in
+ let end_cond = build_fcmp Fcmp.One end_cond zero "loopcond" builder in
+
+ (* Create the "after loop" block and insert it. *)
+ let after_bb = append_block "afterloop" the_function in
+
+ (* Insert the conditional branch into the end of loop_end_bb. *)
+ ignore (build_cond_br end_cond loop_bb after_bb builder);
+
+ (* Any new code will be inserted in after_bb. *)
+ position_at_end after_bb builder;
+
+ (* Restore the unshadowed variable. *)
+ begin match old_val with
+ | Some old_val -> Hashtbl.add named_values var_name old_val
+ | None -> ()
+ end;
+
+ (* for expr always returns 0.0. *)
+ const_null double_type
+ | Ast.Var (var_names, body) ->
+ let old_bindings = ref [] in
+
+ let the_function = block_parent (insertion_block builder) in
+
+ (* Register all variables and emit their initializer. *)
+ Array.iter (fun (var_name, init) ->
+ (* Emit the initializer before adding the variable to scope, this
+ * prevents the initializer from referencing the variable itself, and
+ * permits stuff like this:
+ * var a = 1 in
+ * var a = a in ... # refers to outer 'a'. *)
+ let init_val =
+ match init with
+ | Some init -> codegen_expr init
+ (* If not specified, use 0.0. *)
+ | None -> const_float double_type 0.0
+ in
+
+ let alloca = create_entry_block_alloca the_function var_name in
+ ignore(build_store init_val alloca builder);
+
+ (* Remember the old variable binding so that we can restore the binding
+ * when we unrecurse. *)
+ begin
+ try
+ let old_value = Hashtbl.find named_values var_name in
+ old_bindings := (var_name, old_value) :: !old_bindings;
+ with Not_found -> ()
+ end;
+
+ (* Remember this binding. *)
+ Hashtbl.add named_values var_name alloca;
+ ) var_names;
+
+ (* Codegen the body, now that all vars are in scope. *)
+ let body_val = codegen_expr body in
+
+ (* Pop all our variables from scope. *)
+ List.iter (fun (var_name, old_value) ->
+ Hashtbl.add named_values var_name old_value
+ ) !old_bindings;
+
+ (* Return the body computation. *)
+ body_val
+
+let codegen_proto = function
+ | Ast.Prototype (name, args) | Ast.BinOpPrototype (name, args, _) ->
+ (* Make the function type: double(double,double) etc. *)
+ let doubles = Array.make (Array.length args) double_type in
+ let ft = function_type double_type doubles in
+ let f =
+ match lookup_function name the_module with
+ | None -> declare_function name ft the_module
+
+ (* If 'f' conflicted, there was already something named 'name'. If it
+ * has a body, don't allow redefinition or reextern. *)
+ | Some f ->
+ (* If 'f' already has a body, reject this. *)
+ if block_begin f <> At_end f then
+ raise (Error "redefinition of function");
+
+ (* If 'f' took a different number of arguments, reject. *)
+ if element_type (type_of f) <> ft then
+ raise (Error "redefinition of function with different # args");
+ f
+ in
+
+ (* Set names for all arguments. *)
+ Array.iteri (fun i a ->
+ let n = args.(i) in
+ set_value_name n a;
+ Hashtbl.add named_values n a;
+ ) (params f);
+ f
+
+(* Create an alloca for each argument and register the argument in the symbol
+ * table so that references to it will succeed. *)
+let create_argument_allocas the_function proto =
+ let args = match proto with
+ | Ast.Prototype (_, args) | Ast.BinOpPrototype (_, args, _) -> args
+ in
+ Array.iteri (fun i ai ->
+ let var_name = args.(i) in
+ (* Create an alloca for this variable. *)
+ let alloca = create_entry_block_alloca the_function var_name in
+
+ (* Store the initial value into the alloca. *)
+ ignore(build_store ai alloca builder);
+
+ (* Add arguments to variable symbol table. *)
+ Hashtbl.add named_values var_name alloca;
+ ) (params the_function)
+
+let codegen_func the_fpm = function
+ | Ast.Function (proto, body) ->
+ Hashtbl.clear named_values;
+ let the_function = codegen_proto proto in
+
+ (* If this is an operator, install it. *)
+ begin match proto with
+ | Ast.BinOpPrototype (name, args, prec) ->
+ let op = name.[String.length name - 1] in
+ Hashtbl.add Parser.binop_precedence op prec;
+ | _ -> ()
+ end;
+
+ (* Create a new basic block to start insertion into. *)
+ let bb = append_block "entry" the_function in
+ position_at_end bb builder;
+
+ try
+ (* Add all arguments to the symbol table and create their allocas. *)
+ create_argument_allocas the_function proto;
+
+ let ret_val = codegen_expr body in
+
+ (* Finish off the function. *)
+ let _ = build_ret ret_val builder in
+
+ (* Validate the generated code, checking for consistency. *)
+ Llvm_analysis.assert_valid_function the_function;
+
+ (* Optimize the function. *)
+ let _ = PassManager.run_function the_function the_fpm in
+
+ the_function
+ with e ->
+ delete_function the_function;
+ raise e
+
+
+
+
toplevel.ml:
+
+
+(*===----------------------------------------------------------------------===
+ * Top-Level parsing and JIT Driver
+ *===----------------------------------------------------------------------===*)
+
+open Llvm
+open Llvm_executionengine
+
+(* top ::= definition | external | expression | ';' *)
+let rec main_loop the_fpm the_execution_engine stream =
+ match Stream.peek stream with
+ | None -> ()
+
+ (* ignore top-level semicolons. *)
+ | Some (Token.Kwd ';') ->
+ Stream.junk stream;
+ main_loop the_fpm the_execution_engine stream
+
+ | Some token ->
+ begin
+ try match token with
+ | Token.Def ->
+ let e = Parser.parse_definition stream in
+ print_endline "parsed a function definition.";
+ dump_value (Codegen.codegen_func the_fpm e);
+ | Token.Extern ->
+ let e = Parser.parse_extern stream in
+ print_endline "parsed an extern.";
+ dump_value (Codegen.codegen_proto e);
+ | _ ->
+ (* Evaluate a top-level expression into an anonymous function. *)
+ let e = Parser.parse_toplevel stream in
+ print_endline "parsed a top-level expr";
+ let the_function = Codegen.codegen_func the_fpm e in
+ dump_value the_function;
+
+ (* JIT the function, returning a function pointer. *)
+ let result = ExecutionEngine.run_function the_function [||]
+ the_execution_engine in
+
+ print_string "Evaluated to ";
+ print_float (GenericValue.as_float double_type result);
+ print_newline ();
+ with Stream.Error s | Codegen.Error s ->
+ (* Skip token for error recovery. *)
+ Stream.junk stream;
+ print_endline s;
+ end;
+ print_string "ready> "; flush stdout;
+ main_loop the_fpm the_execution_engine stream
+
+
+
+
toy.ml:
+
+
+(*===----------------------------------------------------------------------===
+ * Main driver code.
+ *===----------------------------------------------------------------------===*)
+
+open Llvm
+open Llvm_executionengine
+open Llvm_target
+open Llvm_scalar_opts
+
+let main () =
+ (* Install standard binary operators.
+ * 1 is the lowest precedence. *)
+ Hashtbl.add Parser.binop_precedence '=' 2;
+ Hashtbl.add Parser.binop_precedence '<' 10;
+ Hashtbl.add Parser.binop_precedence '+' 20;
+ Hashtbl.add Parser.binop_precedence '-' 20;
+ Hashtbl.add Parser.binop_precedence '*' 40; (* highest. *)
+
+ (* Prime the first token. *)
+ print_string "ready> "; flush stdout;
+ let stream = Lexer.lex (Stream.of_channel stdin) in
+
+ (* Create the JIT. *)
+ let the_module_provider = ModuleProvider.create Codegen.the_module in
+ let the_execution_engine = ExecutionEngine.create the_module_provider in
+ let the_fpm = PassManager.create_function the_module_provider in
+
+ (* Set up the optimizer pipeline. Start with registering info about how the
+ * target lays out data structures. *)
+ TargetData.add (ExecutionEngine.target_data the_execution_engine) the_fpm;
+
+ (* Promote allocas to registers. *)
+ add_memory_to_register_promotion the_fpm;
+
+ (* Do simple "peephole" optimizations and bit-twiddling optzn. *)
+ add_instruction_combining the_fpm;
+
+ (* reassociate expressions. *)
+ add_reassociation the_fpm;
+
+ (* Eliminate Common SubExpressions. *)
+ add_gvn the_fpm;
+
+ (* Simplify the control flow graph (deleting unreachable blocks, etc). *)
+ add_cfg_simplification the_fpm;
+
+ (* Run the main "interpreter loop" now. *)
+ Toplevel.main_loop the_fpm the_execution_engine stream;
+
+ (* Print out all the generated code. *)
+ dump_module Codegen.the_module
+;;
+
+main ()
+
+
+
+
bindings.c
+
+
+#include <stdio.h>
+
+/* putchard - putchar that takes a double and returns 0. */
+extern double putchard(double X) {
+ putchar((char)X);
+ return 0;
+}
+
+/* printd - printf that takes a double prints it as "%f\n", returning 0. */
+extern double printd(double X) {
+ printf("%f\n", X);
+ return 0;
+}
+
Advanced Topics
From baldrick at free.fr Mon Mar 31 06:44:10 2008
From: baldrick at free.fr (Duncan Sands)
Date: Mon, 31 Mar 2008 13:44:10 +0200
Subject: [llvm-commits]
=?iso-8859-1?q?=5Bllvm-gcc-4=2E2=5D_r48608_-_in=09?=
=?iso-8859-1?q?/llvm-gcc-4=2E2/trunk/gcc=3A_config/i386/llvm-i386-target?=
=?iso-8859-1?q?=2Eh=09config/i386/llvm-i386=2Ecpp_config/rs6000/llvm-rs60?=
=?iso-8859-1?q?00=2Ecpp=09config/rs6000/rs6000=2Eh_llvm-abi=2Eh_llvm-conv?=
=?iso-8859-1?q?ert=2Ecpp?=
In-Reply-To:
References: <200803201842.m2KIgwT9002122@zion.cs.uiuc.edu>
<200803251803.53233.baldrick@free.fr>
Message-ID: <200803311344.11770.baldrick@free.fr>
Hi Dale,
> > Hi Dale, I've cleaned up the return value part
> > in commit 48778. Can you please check that it
> > works properly for you. I haven't yet decided
> > the best way of handling the argument case.
>
> There are some regressions but they all appear to be a different
> problem. We'll need to get that cleaned up to be sure.
while cleaning up the call argument part, I discovered an existing bug.
Compiling SingleSource/UnitTests/Vector/SSE/sse.isamax.c on x86-32 linux
gives the assertion
cc1: Instructions.cpp:258: void llvm::CallInst::init(llvm::Value*, llvm::Value* const*, unsigned int): Assertion `(NumParams == FTy->getNumParams() || (FTy->isVarArg() && NumParams > FTy->getNumParams())) && "Calling a function with bad signature!"' failed.
The reason is that @llvm.x86.sse.movmsk.ps is declared as
declare i32 @llvm.x86.sse.movmsk.ps(<4 x float>) nounwind readnone
but the vector is being passed in integer registers in the call, namely
as two i64's.
Ciao,
Duncan.
From gordonhenriksen at mac.com Mon Mar 31 07:04:46 2008
From: gordonhenriksen at mac.com (Gordon Henriksen)
Date: Mon, 31 Mar 2008 08:04:46 -0400
Subject: [llvm-commits] [PATCH] Expose Function::viewCFG and
Function::viewCFGOnly to bindings.
In-Reply-To: <1206916389-84564-1-git-send-email-idadesub@users.sourceforge.net>
References: <1206916389-84564-1-git-send-email-idadesub@users.sourceforge.net>
Message-ID:
On 2008-03-30, at 18:33, Erick Tryzelaar wrote:
> These are some other functions referenced in the tutorial, so we
> should
> expose them in ocaml. I put them in analysis because having them in
> VMCore created a dependency cycle.
Looks good. Please commit.
? Gordon
From romix.llvm at googlemail.com Mon Mar 31 07:09:04 2008
From: romix.llvm at googlemail.com (Roman Levenstein)
Date: Mon, 31 Mar 2008 13:09:04 +0100
Subject: [llvm-commits] Speeding up instruction selection
In-Reply-To: <417E9823-6AF6-44B0-83FD-89F0CC079B6A@apple.com>
References:
<00FB5805-1FA5-4534-BBD0-E6E34EC17BF1@apple.com>
<1C1F9E6C-5B8E-4667-A119-4590F340FDE7@apple.com>
<83CD954F-CD49-4D4B-950D-B6CB6CD1B32D@apple.com>
<417E9823-6AF6-44B0-83FD-89F0CC079B6A@apple.com>
Message-ID:
Hi Dan, Hi Evan,
Here is the updated patch.
1) I introduced the SDOperandPtr class, as I suggested in my last
email. This class is a sort of an "intelligent" pointer. With this
class, there is no need to copy the SDOperand arrays into the
temporary SmallVectors. Therefore there is no additional overhead.
2) Additionally, I also finally made SDOperand a member of the SDUse,
instead of using it as a base class. Now we can be sure that no
unintentional conversion of SDUse into SDOperand happens.
3) I also changed the return types of some methods from SDUse to
SDOperand, as it was suggested during the review.
This patch passes all DejaGnu tests.
-Roman
2008/3/31, Evan Cheng :
> Ok, as long as the compile time increase isn't noticeable I won't
> sweat it. However, if it happens in a few places, perhaps we need (yet
> another) a getNode() that takes iterators and hide the copying inside
> that instead.
>
>
> Evan
>
>
> On Mar 28, 2008, at 3:14 PM, Dan Gohman wrote:
>
> >
> > On Mar 27, 2008, at 6:04 PM, Evan Cheng wrote:
> >>
> >>> OK. This is also my view. I don't think this temporary SmallVectors
> >>> introduce any significant overhead. According to the profiler (I use
> >>> Google profiling tools, which do not require to compile with -pg),
> >>> they consume virtually no time and that even on big testcases,
> >>> that I
> >>> like so much :-)
> >>
> >> I am not sure about this. For all the places which pass the list of
> >> SDOperand's as a ptr to an array, you want to make a copy of it into
> >> an array before passing it?
> >
> > Hi Evan,
> >
> > The patch is not changing every place that passes lists of
> > SDOperands to make an extra copy. The copy is only needed when
> > passing an existing node's operands, which is are now stored as
> > an array of SDUse, to a function that wants an SDOperand array.
> > And judging by Roman's current patch, this doesn't happen in
> > all that many places.
> >
> > Dan
-------------- next part --------------
A non-text attachment was scrubbed...
Name: SDNodeUses.patch
Type: text/x-patch
Size: 55907 bytes
Desc: not available
Url : http://lists.cs.uiuc.edu/pipermail/llvm-commits/attachments/20080331/c78088e6/attachment.bin
From baldrick at free.fr Mon Mar 31 07:39:29 2008
From: baldrick at free.fr (Duncan Sands)
Date: Mon, 31 Mar 2008 14:39:29 +0200
Subject: [llvm-commits]
=?iso-8859-1?q?=5Bllvm-gcc-4=2E2=5D_r48608_-_in=09?=
=?iso-8859-1?q?/llvm-gcc-4=2E2/trunk/gcc=3A_config/i386/llvm-i386-target?=
=?iso-8859-1?q?=2Eh=09config/i386/llvm-i386=2Ecpp_config/rs6000/llvm-rs60?=
=?iso-8859-1?q?00=2Ecpp=09config/rs6000/rs6000=2Eh_llvm-abi=2Eh_llvm-conv?=
=?iso-8859-1?q?ert=2Ecpp?=
In-Reply-To: <200803311344.11770.baldrick@free.fr>
References: <200803201842.m2KIgwT9002122@zion.cs.uiuc.edu>
<200803311344.11770.baldrick@free.fr>
Message-ID: <200803311439.30161.baldrick@free.fr>
> while cleaning up the call argument part, I discovered an existing bug.
> Compiling SingleSource/UnitTests/Vector/SSE/sse.isamax.c on x86-32 linux
> gives the assertion
>
> cc1: Instructions.cpp:258: void llvm::CallInst::init(llvm::Value*, llvm::Value* const*, unsigned int): Assertion `(NumParams == FTy->getNumParams() || (FTy->isVarArg() && NumParams > FTy->getNumParams())) && "Calling a function with bad signature!"' failed.
>
> The reason is that @llvm.x86.sse.movmsk.ps is declared as
>
> declare i32 @llvm.x86.sse.movmsk.ps(<4 x float>) nounwind readnone
>
> but the vector is being passed in integer registers in the call, namely
> as two i64's.
How about giving DefaultABI a isBuiltin parameter, not just
HandleReturnType, and use that for parameters? That said, since
builtins really have their own ABI, maybe the right thing to do
is to introduce a new ABI template for use with builtins rather
than complicating DefaultABI. Not sure it's worth it though.
Ciao,
Duncan.
From jon at alacatialabs.com Mon Mar 31 10:43:02 2008
From: jon at alacatialabs.com (Jonathan Johnson)
Date: Mon, 31 Mar 2008 10:43:02 -0500
Subject: [llvm-commits] CVS: llvm-www/devmtg/current/index.html
Message-ID: <200803311543.m2VFh2xp022505@zion.cs.uiuc.edu>
Changes in directory llvm-www/devmtg/current:
index.html updated: 1.7 -> 1.8
---
Log message:
Added date and location for conference, removed poll. Added section for agenda, with link to email if interested in presenting.
---
Diffs of the changes: (+9 -14)
index.html | 23 +++++++++--------------
1 files changed, 9 insertions(+), 14 deletions(-)
Index: llvm-www/devmtg/current/index.html
diff -u llvm-www/devmtg/current/index.html:1.7 llvm-www/devmtg/current/index.html:1.8
--- llvm-www/devmtg/current/index.html:1.7 Sun Mar 23 15:50:28 2008
+++ llvm-www/devmtg/current/index.html Mon Mar 31 10:42:21 2008
@@ -3,8 +3,8 @@
The meeting is tentatively scheduled between 8:00 AM and 5:00 PM Pacific Time. The schedule of talks has not yet been decided. If you are interested in speaking, please let us know what topic you'd like to present.