I'm trying to get the float value from a global variable and set it as an instruction's operand.
Here is what I want to do:
#a = private constant float 0x3FB99999A0000000
...
%1 = load float, float* #a ---> removed
%3 = fmul fast %1, %2 ---> %3 = fmul fast float 0x3FB99999A0000000, %2
Below is what I haved tried so far:
for (auto gv_iter = llvm_module.global_begin();gv_iter != llvm_module.global_end(); gv_iter++){
llvm::GlobalVariable* gv = &*gv_iter;
for(auto user_of_gv : gv->users()){
llvm::Instruction *instr_ld_gv = llvm::dyn_cast<llvm::Instruction>(user_of_gv);
llvm::Value *val_gv = llvm::cast<llvm::Value>(instr_ld_gv);
llvm::Constant *const_gv = gv->getInitializer();
llvm::ConstantFP *constfp_gv = llvm::dyn_cast<llvm::ConstantFP>(const_gv);
float gv_fpval = (constfp_gv->getValueAPF()).convertToFloat();
llvm::Constant *const_gv_opd = llvm::ConstantFP::get(llvm::Type::getFloatTy(llvm_context),gv_fpval);
for(auto user_of_load : val_gv->users()){
llvm::Instruction *instr_exe_gv = llvm::dyn_cast<llvm::Instruction>(user_of_load);
//P
for(int operand_num = 0;operand_num < instr_exe_gv->getNumOperands();operand_num++){
llvm::Value *val_instr_op = instr_exe_gv->getOperand(operand_num);
if(val_instr_op == val_gv){
instr_exe_gv->setOperand(operand_num,const_gv_opd);
instr_ld_gv->removeFromParent();
}
}
}
}
}
However, it'll cause segmentation fault when I tried to run my code.
I'm sure that I have accessed the global variable and instruction I wanted
by printing the value of
gv_fpval which is 0.1 because 0x3FB99999A0000000 equals 0.10000000149011612 in double
precision. It seems that the program crashes at setOperand().
Consider the following example
hello.cpp
#include <stdio.h>
// Global Constant value
float a=1.4f;
float Multiply(){
float b=2.2f;
float c=4.32f;
float d= a*c;
return d;
}
int main(int argc, char const *argv[])
{
printf("%f\n",Multiply());
return 0;
}
The module pass will loop for Floating point global variable and any use in the program will be replaced by the constant FP value. The LLVM pass are as follow ConstantReplacementPass.cpp:-
#include "llvm/Pass.h"
#include "llvm/IR/Function.h"
#include "llvm/Support/raw_ostream.h"
#include "llvm/IR/LegacyPassManager.h"
#include "llvm/IR/InstrTypes.h"
#include "llvm/Transforms/IPO/PassManagerBuilder.h"
#include "llvm/IR/IRBuilder.h"
#include "llvm/IR/Module.h"
#include "llvm/Transforms/Utils/BasicBlockUtils.h"
#include "llvm/DebugInfo/DWARF/DWARFDebugLine.h"
#include "llvm/IR/DebugLoc.h"
#include "llvm/IR/DebugInfo.h"
using namespace llvm;
/* StackOverflow : https://stackoverflow.com/questions/48212351/how-to-get-llvm-global-variable-constant-value* /
/**Bernard Nongpoh */
namespace {
class ConstantReplacementPass : public ModulePass {
public:
static char ID;
ConstantReplacementPass() : ModulePass(ID) {
srand (time(NULL));
}
virtual bool runOnModule(Module &M) {
// list to collect instruction
/*
* You cannot change an iterator while iterating over it
• To remove instructions or modify, first collect the instructions to remove/modify
•
*
* **/
// This are the list of load to delete
SmallVector<Instruction*,128> *WorkListLoad=new SmallVector<Instruction*,128>();
// This is the list of instruction to modify the source operand
SmallVector<Instruction*,128> *WorkListUserOfLoad=new SmallVector<Instruction*,128>();
for (auto gv_iter = M.global_begin();gv_iter != M.global_end(); gv_iter++) {
/* GLOBAL DATA INFO*/
GlobalVariable *gv = &*gv_iter;
Constant *const_gv = gv->getInitializer();
ConstantFP *Fvalue;
if(!const_gv->isNullValue()) {
if (ConstantFP *constfp_gv = llvm::dyn_cast<llvm::ConstantFP>(const_gv)) {
float gv_fpval = (constfp_gv->getValueAPF()).convertToFloat();
Fvalue = constfp_gv;
errs() << gv_fpval; // Value retrieved here
// Collect Instruction to modify
}
for (auto user_of_gv: gv->users()) {
// Collect in a worklist
if (llvm::Instruction *instr_ld_gv = llvm::dyn_cast<Instruction>(user_of_gv)) {
if (LoadInst *loadInst = dyn_cast<LoadInst>(instr_ld_gv)) {
WorkListLoad->push_back(loadInst);
for (auto user_of_load:loadInst->users()) {
user_of_load->dump();
Instruction *instruction1 = dyn_cast<Instruction>(user_of_load);
instruction1->dump();
//instruction1->setOperand(0, Fvalue);
//instruction1->dump();
// if(Instruction *instruction1 = dyn_cast<Instruction>(user_of_load))
WorkListUserOfLoad->push_back(instruction1);
//instruction1->setOperand(0, Fvalue);
//instruction1->dump();
}
}
}
}
// Modify Here
while (!WorkListUserOfLoad->empty()) {
Instruction *instruction = WorkListUserOfLoad->pop_back_val();
instruction->setOperand(0, Fvalue);
instruction->dump();
}
// Removing all loads that are used by the global variable
while (!WorkListLoad->empty()) {
Instruction *instruction = WorkListLoad->pop_back_val();
instruction->eraseFromParent();
}
}
}
return true;
}
};
}
char ConstantReplacementPass::ID = 0;
static RegisterPass<ConstantReplacementPass> F0("constantREP", "Constant Replacement Pass "
, false,true);
Key points:-
Before doing any modification on the instruction. Collect first in the worklist.
perform the modification on the worklist.
you cannot do the modification while using the iterator.
I successfully tested on the above source code hello.cpp the corresponding IR after the pass is as follow:-
entry:
%b = alloca float, align 4
%c = alloca float, align 4
%d = alloca float, align 4
call void #llvm.dbg.declare(metadata float* %b, metadata !14, metadata !15),
... !dbg !16
store float 0x40019999A0000000, float* %b, align 4, !dbg !16
call void #llvm.dbg.declare(metadata float* %c, metadata !17, metadata !15),
... !dbg !18
store float 0x401147AE20000000, float* %c, align 4, !dbg !18
call void #llvm.dbg.declare(metadata float* %d, metadata !19, metadata !15),
... !dbg !20
%0 = load float, float* %c, align 4, !dbg !21
%mul = fmul float 0x3FF6666660000000, %0, !dbg !22
store float %mul, float* %d, align 4, !dbg !20
%1 = load float, float* %d, align 4, !dbg !23
ret float %1, !dbg !24
Maybe using -O3 optimization flag will wipe out everything...
Hope this helps..
Related
I am a beginner in LLVM, and I wanna get the true value of a given varibale name and line number by using LLVM pass. I have several problems.
correctly get the metadata;
get the type of String;
get dynamic value.
For example, I wanna get the value of b after line 7:
#include <iostream>
int main() {
int b;
int a;
std::cin >> a;
b = a; // line 7
return 0;
}
IR of the main function:
; Function Attrs: mustprogress noinline norecurse optnone uwtable
define dso_local i32 #main() #4 !dbg !857 {
%1 = alloca i32, align 4
%2 = alloca i32, align 4
%3 = alloca i32, align 4
store i32 0, i32* %1, align 4
call void #llvm.dbg.declare(metadata i32* %2, metadata !858, metadata !DIExpression()), !dbg !859
call void #llvm.dbg.declare(metadata i32* %3, metadata !860, metadata !DIExpression()), !dbg !861
%4 = call nonnull align 8 dereferenceable(16) %"class.std::basic_istream"* #_ZNSirsERi(%"class.std::basic_istream"* nonnull align 8 dereferenceable(16) #_ZSt3cin, i32* nonnull align 4 dereferenceable(4) %3), !dbg !862
%5 = load i32, i32* %3, align 4, !dbg !863
store i32 %5, i32* %2, align 4, !dbg !864
ret i32 0, !dbg !865
}
; METADATA
!857 = distinct !DISubprogram(name: "main", scope: !8, file: !8, line: 11, type: !539, scopeLine: 11, flags: DIFlagPrototyped, spFlags: DISPFlagDefinition, unit: !7, retainedNodes: !9)
!858 = !DILocalVariable(name: "b", scope: !857, file: !8, line: 12, type: !20)
!859 = !DILocation(line: 12, column: 9, scope: !857)
!860 = !DILocalVariable(name: "a", scope: !857, file: !8, line: 13, type: !20)
!861 = !DILocation(line: 13, column: 9, scope: !857)
!862 = !DILocation(line: 14, column: 14, scope: !857)
!863 = !DILocation(line: 15, column: 9, scope: !857)
!864 = !DILocation(line: 15, column: 7, scope: !857)
!865 = !DILocation(line: 16, column: 5, scope: !857)
I need to read dynamic value, so I instrument logvar function:
#include <iostream>
#include <string>
extern "C" void logvar(int i, std::string name) {
std::cout << "Num: " << i << "; Name: " << name << std::endl;
}
And now, my pass.cpp is as follows:
virtual bool runOnFunction(Function &F) {
// Get the function to call from our runtime library.
LLVMContext &Ctx = F.getContext();
std::vector<Type*> paramTypes = {
Type::getInt32Ty(Ctx),
// Here I need a string type to match logvar function,
// but I don't know how to write it.
// Maybe a pointerType?
};
Type *retType = Type::getVoidTy(Ctx);
FunctionType *logFuncType = FunctionType::get(retType, paramTypes, false);
FunctionCallee logFunc =
F.getParent()->getOrInsertFunction("logvar", logFuncType);
for (auto &B : F) {
for (auto &I : B) {
if (auto *op = dyn_cast<LoadInst>(&I)) { // timo_schalachter
int number;
auto alloca = dyn_cast<AllocaInst>(op->getOperand(0));
for (auto user: alloca->users()) {
if (auto store = dyn_cast<StoreInst>(user)) {
auto constant_int = dyn_cast<ConstantInt>(store->getOperand(0));
number = constant_int->getSExtValue();
errs() << number << "\n";
}
}
}
if (auto *op = dyn_cast<StoreInst>(&I)) {
errs() << *op << ".StoreInst\n";
Value *val = op->getValueOperand();
if (auto constant_int = dyn_cast<ConstantInt>(val)) {
int number = constant_int->getSExtValue();
errs() << number << ".\n\n";
} else if (auto constant_fp = dyn_cast<ConstantFP>(val)) {
float number = constant_fp->???;
// I cannot find the function to get the value of float point here.
} // and how to deal with constant string?
// metadata:
// store i32 %4, i32* %2, align 4, !dbg !863
Value *arg1 = op->getOperand(0); // %4 = xxx
Value *arg2 = op->getOperand(1); // %2 = xxx
unsigned mk = op->getContext().getMDKindID("dbg");
MDNode *mdn = op->getMetadata(mk);
if (mdn) {
Metadata *mds = mdn->getOperand(0);
StringRef str;
if (MDString::classof(mds)) {
str = (cast<MDString>(*mds)).getString();
errs() << str;
}
} else {
errs() << "no dbg!\n";
// when I run this code, it always says: "no dbg!"
// I don't know why...
}
// instrumentation
IRBuilder<> builder(op);
builder.SetInsertPoint(&B, ++builder.GetInsertPoint());
Value* args[] = {arg1, ???};
// I cannot find the name of variable.
// I think it should be str in metadata.
// Still the problem: how to write STRING's type in LLVM?
builder.CreateCall(logFunc, args);
}
}
}
return false;
}
Problems are in code. I cannot make myself understood clearly by using English.
I think String maybe i8*. You can read this link https://llvm.org/docs/LangRef.html#function-type to get function type.
Final goal: Trying to generate CFG related information(such as topological sort) using LLVM.
Status: I'm pretty new to LLVM and kind of lost - any kind of information or blogs to help me get started towards my final goal is great!
My Question: After reading Eli's code and blog post, I get the .ll file first and run the code but I got no result.
Here is the .ll file example:
; ModuleID = 'CWE15_External_Control_of_System_or_Configuration_Setting__w32_83a.cpp'
source_filename = "CWE15_External_Control_of_System_or_Configuration_Setting__w32_83a.cpp"
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-linux-gnu"
%"class.CWE15_External_Control_of_System_or_Configuration_Setting__w32_83::CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_bad" = type { i8* }
%"class.CWE15_External_Control_of_System_or_Configuration_Setting__w32_83::CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_goodG2B" = type { i8* }
; Function Attrs: noinline optnone uwtable
define void #_ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_833badEv() #0 {
%1 = alloca i8*, align 8
%2 = alloca [100 x i8], align 16
%3 = alloca %"class.CWE15_External_Control_of_System_or_Configuration_Setting__w32_83::CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_bad", align 8
%4 = bitcast [100 x i8]* %2 to i8*
call void #llvm.memset.p0i8.i64(i8* %4, i8 0, i64 100, i32 16, i1 false)
%5 = getelementptr inbounds [100 x i8], [100 x i8]* %2, i32 0, i32 0
store i8* %5, i8** %1, align 8
%6 = load i8*, i8** %1, align 8
call void #_ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_8369CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_badC1EPc(%"class.CWE15_External_Control_of_System_or_Configuration_Setting__w32_83::CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_bad"* %3, i8* %6)
call void #_ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_8369CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_badD1Ev(%"class.CWE15_External_Control_of_System_or_Configuration_Setting__w32_83::CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_bad"* %3) #4
ret void
}
; Function Attrs: argmemonly nounwind
declare void #llvm.memset.p0i8.i64(i8* nocapture writeonly, i8, i64, i32, i1) #1
declare void #_ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_8369CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_badC1EPc(%"class.CWE15_External_Control_of_System_or_Configuration_Setting__w32_83::CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_bad"*, i8*) unnamed_addr #2
; Function Attrs: nounwind
declare void #_ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_8369CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_badD1Ev(%"class.CWE15_External_Control_of_System_or_Configuration_Setting__w32_83::CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_bad"*) unnamed_addr #3
; Function Attrs: noinline optnone uwtable
define void #_ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_834goodEv() #0 {
call void #_ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_83L7goodG2BEv()
ret void
}
; Function Attrs: noinline optnone uwtable
define internal void #_ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_83L7goodG2BEv() #0 {
%1 = alloca i8*, align 8
%2 = alloca [100 x i8], align 16
%3 = alloca %"class.CWE15_External_Control_of_System_or_Configuration_Setting__w32_83::CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_goodG2B", align 8
%4 = bitcast [100 x i8]* %2 to i8*
call void #llvm.memset.p0i8.i64(i8* %4, i8 0, i64 100, i32 16, i1 false)
%5 = getelementptr inbounds [100 x i8], [100 x i8]* %2, i32 0, i32 0
store i8* %5, i8** %1, align 8
%6 = load i8*, i8** %1, align 8
call void #_ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_8373CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_goodG2BC1EPc(%"class.CWE15_External_Control_of_System_or_Configuration_Setting__w32_83::CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_goodG2B"* %3, i8* %6)
call void #_ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_8373CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_goodG2BD1Ev(%"class.CWE15_External_Control_of_System_or_Configuration_Setting__w32_83::CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_goodG2B"* %3) #4
ret void
}
declare void #_ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_8373CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_goodG2BC1EPc(%"class.CWE15_External_Control_of_System_or_Configuration_Setting__w32_83::CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_goodG2B"*, i8*) unnamed_addr #2
; Function Attrs: nounwind
declare void #_ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_8373CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_goodG2BD1Ev(%"class.CWE15_External_Control_of_System_or_Configuration_Setting__w32_83::CWE15_External_Control_of_System_or_Configuration_Setting__w32_83_goodG2B"*) unnamed_addr #3
attributes #0 = { noinline optnone uwtable "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+fxsr,+mmx,+sse,+sse2,+x87" "unsafe-fp-math"="false" "use-soft-float"="false" }
attributes #1 = { argmemonly nounwind }
attributes #2 = { "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+fxsr,+mmx,+sse,+sse2,+x87" "unsafe-fp-math"="false" "use-soft-float"="false" }
attributes #3 = { nounwind "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+fxsr,+mmx,+sse,+sse2,+x87" "unsafe-fp-math"="false" "use-soft-float"="false" }
attributes #4 = { nounwind }
!llvm.module.flags = !{!0}
!llvm.ident = !{!1}
!0 = !{i32 1, !"wchar_size", i32 4}
!1 = !{!"clang version 5.0.0 (tags/RELEASE_500/final 375507)"}
and here is the .cpp file which generate sort information:
//------------------------------------------------------------------------------
// bb_toposort_sccs LLVM sample. Demonstrates:
//
// * How to implement DFS & topological sort over the control-flow graph (CFG)
// of a function.
// * How to use po_iterator for post-order iteration over basic blocks.
// * How to use scc_iterator for post-order iteration over strongly-connected
// components in the graph of basic blocks.
//
// Eli Bendersky (eliben#gmail.com)
// This code is in the public domain
//------------------------------------------------------------------------------
#include "llvm/ADT/DenseMap.h"
#include "llvm/ADT/PostOrderIterator.h"
#include "llvm/ADT/SCCIterator.h"
#include "llvm/IR/CFG.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/Instruction.h"
#include "llvm/IR/Instructions.h"
#include "llvm/IR/LegacyPassManager.h"
#include "llvm/IR/Module.h"
#include "llvm/IRReader/IRReader.h"
#include "llvm/Pass.h"
#include "llvm/Support/SourceMgr.h"
#include "llvm/Support/raw_ostream.h"
#include <string>
#include <vector>
using namespace llvm;
// Runs a topological sort on the basic blocks of the given function. Uses
// the simple recursive DFS from "Introduction to algorithms", with 3-coloring
// of vertices. The coloring enables detecting cycles in the graph with a simple
// test.
class TopoSorter {
public:
void runToposort(const Function &F) {
outs() << "Topological sort of " << F.getName() << ":\n";
// Initialize the color map by marking all the vertices white.
for (Function::const_iterator I = F.begin(), IE = F.end(); I != IE; ++I) {
ColorMap[&*I] = TopoSorter::WHITE;
}
// The BB graph has a single entry vertex from which the other BBs should
// be discoverable - the function entry block.
bool success = recursiveDFSToposort(&F.getEntryBlock());
if (success) {
// Now we have all the BBs inside SortedBBs in reverse topological order.
for (BBVector::const_reverse_iterator RI = SortedBBs.rbegin(),
RE = SortedBBs.rend();
RI != RE; ++RI) {
outs() << " " << (*RI)->getName() << "\n";
}
} else {
outs() << " Sorting failed\n";
}
}
private:
enum Color { WHITE, GREY, BLACK };
// Color marks per vertex (BB).
typedef DenseMap<const BasicBlock *, Color> BBColorMap;
// Collects vertices (BBs) in "finish" order. The first finished vertex is
// first, and so on.
typedef SmallVector<const BasicBlock *, 32> BBVector;
BBColorMap ColorMap;
BBVector SortedBBs;
// Helper function to recursively run topological sort from a given BB.
// Returns true if the sort succeeded and false otherwise; topological sort
// may fail if, for example, the graph is not a DAG (detected a cycle).
bool recursiveDFSToposort(const BasicBlock *BB) {
ColorMap[BB] = TopoSorter::GREY;
// For demonstration, using the lowest-level APIs here. A BB's successors
// are determined by looking at its terminator instruction.
const TerminatorInst *TInst = BB->getTerminator();
for (unsigned I = 0, NSucc = TInst->getNumSuccessors(); I < NSucc; ++I) {
BasicBlock *Succ = TInst->getSuccessor(I);
Color SuccColor = ColorMap[Succ];
if (SuccColor == TopoSorter::WHITE) {
if (!recursiveDFSToposort(Succ))
return false;
} else if (SuccColor == TopoSorter::GREY) {
// This detects a cycle because grey vertices are all ancestors of the
// currently explored vertex (in other words, they're "on the stack").
outs() << " Detected cycle: edge from " << BB->getName() << " to "
<< Succ->getName() << "\n";
return false;
}
}
// This BB is finished (fully explored), so we can add it to the vector.
ColorMap[BB] = TopoSorter::BLACK;
SortedBBs.push_back(BB);
return true;
}
};
class AnalyzeBBGraph : public FunctionPass {
public:
AnalyzeBBGraph(const std::string &AnalysisKind)
: FunctionPass(ID), AnalysisKind(AnalysisKind) {}
virtual bool runOnFunction(Function &F) {
if (AnalysisKind == "-topo") {
TopoSorter TS;
TS.runToposort(F);
} else if (AnalysisKind == "-po") {
// Use LLVM's post-order iterator to produce a reverse topological sort.
// Note that this doesn't detect cycles so if the graph is not a DAG, the
// result is not a true topological sort.
outs() << "Basic blocks of " << F.getName() << " in post-order:\n";
for (po_iterator<BasicBlock *> I = po_begin(&F.getEntryBlock()),
IE = po_end(&F.getEntryBlock());
I != IE; ++I) {
outs() << " " << (*I)->getName() << "\n";
}
} else if (AnalysisKind == "-scc") {
// Use LLVM's Strongly Connected Components (SCCs) iterator to produce
// a reverse topological sort of SCCs.
outs() << "SCCs for " << F.getName() << " in post-order:\n";
for (scc_iterator<Function *> I = scc_begin(&F), IE = scc_end(&F);
I != IE; ++I) {
// Obtain the vector of BBs in this SCC and print it out.
const std::vector<BasicBlock *> &SCCBBs = *I;
outs() << " SCC: ";
for (std::vector<BasicBlock *>::const_iterator BBI = SCCBBs.begin(),
BBIE = SCCBBs.end();
BBI != BBIE; ++BBI) {
outs() << (*BBI)->getName() << " ";
}
outs() << "\n";
}
} else {
outs() << "Unknown analysis kind: " << AnalysisKind << "\n";
}
return false;
}
// The address of this member is used to uniquely identify the class. This is
// used by LLVM's own RTTI mechanism.
static char ID;
private:
std::string AnalysisKind;
};
char AnalyzeBBGraph::ID = 0;
int main(int argc, char **argv) {
if (argc < 3) {
// Using very basic command-line argument parsing here...
errs() << "Usage: " << argv[0] << " -[topo|po|scc] <IR file>\n";
return 1;
}
// Parse the input LLVM IR file into a module.
SMDiagnostic Err;
LLVMContext Context;
std::unique_ptr<Module> Mod(parseIRFile(argv[2], Err, Context));
if (!Mod) {
Err.print(argv[0], errs());
return 1;
}
// Create a pass manager and fill it with the passes we want to run.
legacy::PassManager PM;
PM.add(new AnalyzeBBGraph(std::string(argv[1])));
PM.run(*Mod);
return 0;
}
here is my result looks like when I try to get topo sort:
Topological sort of _ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_833badEv:
Topological sort of _ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_834goodEv:
Topological sort of _ZN65CWE15_External_Control_of_System_or_Configuration_Setting__w32_83L7goodG2BEv:
I tried something else and I found out all the "getName()" print is nothing,
Is there something wrong with my code or somrthing wrong with my llvm IR ?
for(BasicBlock &BB : F){
outs() << "BB:" <<BB.getName() <<" has" <<BB.size() <<"instructions.\n";
for(Instruction &I : BB){
outs() << "details: "<<I <<"name: "<<I.getName() <<"\n";
for(Use &U : I.operands()){
Value *v = U.get();
outs() <<"value:"<< v<<"name:" << v->getName() <<"\n";
}
}
}
Any thoughts are appreciated!
You might want to run the instruction namer pass after you obtain your IR (the .ll file) in order to assign names to the basic blocks and the registers, because currently, they do not have any names apart from their numeric assignment (e.g. %1, etc). You can run this with:
opt -instnamer foo.bc -o foo-named.bc
The 3 functions definitions have only one basic block in them, so you are not going to get anything interesting, but it is a good start.
Another point that you might want to address now before moving on, is the optimization level. All your functions are annotated with optnone that will disable any kind of optimization from other LLVM passes. A more flexible way to get unoptimized IR, but one that will allow further optimizations to take effect on it is to extracted using:
clang -emit-llvm -O1 -Xclang -disable-llvm-passes foo.c
I'm using LLVM C++ api to generate LLVM IR and using JIT to run the function.
The function I want to simulate is
double foofor(double a)
{
for (int i = 0; i < 10; i++) {
a += 2;
}
return a;
}
The corresponding llvm IR code I generate is (using TheFunction->print)
define double #foofor(double %a) {
entry:
br label %loop
loop: ; preds = %loop, %entry
%a1 = phi double [ 0.000000e+00, %entry ], [ %nextvar, %loop ]
%addtmp = fadd double %a, 2.000000e+00
%nextvar = fadd double 1.000000e+00, %a1
%cmptmp = fcmp ult double %a1, 1.000000e+01
br i1 %cmptmp, label %loop, label %afterloop
afterloop: ; preds = %loop
ret double %addtmp
}
When I call foofor(40), I get 42 not 60(40 + 20).
The code I have write(without AST)
#include "./llvm-8.0.1.src/examples/Kaleidoscope/include/KaleidoscopeJIT.h"
#include "llvm/ADT/APFloat.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/IR/BasicBlock.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DerivedTypes.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/IRBuilder.h"
#include "llvm/IR/LLVMContext.h"
#include "llvm/IR/LegacyPassManager.h"
#include "llvm/IR/Module.h"
#include "llvm/IR/Type.h"
#include "llvm/IR/Verifier.h"
#include "llvm/Support/TargetSelect.h"
#include "llvm/Target/TargetMachine.h"
#include "llvm/Transforms/InstCombine/InstCombine.h"
#include "llvm/Transforms/Scalar.h"
#include "llvm/Transforms/Scalar/GVN.h"
#include <string>
#include <vector>
#include <iostream>
using namespace std;
using namespace llvm;
using namespace llvm::orc;
//LLVM items
static LLVMContext TheContext;
static IRBuilder<> Builder(TheContext);
static std::unique_ptr<Module> TheModule;
//JIT
static std::unique_ptr<KaleidoscopeJIT> TheJIT;
/*
*double foofor(double a)
*{
* for(i = 0; i < a; i++) {
* a = a + 1
* }
* return a
*}
*
*/
int main()
{
InitializeNativeTarget();
InitializeNativeTargetAsmPrinter();
//init module
TheModule = llvm::make_unique<Module>("myjit", TheContext);
//used to be runned by jit later
TheJIT = llvm::make_unique<KaleidoscopeJIT>();
TheModule->setDataLayout(TheJIT->getTargetMachine().createDataLayout());
//define the args
vector<std::string> ArgNames;
//foocfor has 1 args
ArgNames.push_back(string("a"));
//make the 1 args attach to LLVM Type::double
std::vector<Type *> Doubles(ArgNames.size(), Type::getDoubleTy(TheContext));
//generate llvm function type
FunctionType *FT = FunctionType::get(Type::getDoubleTy(TheContext), Doubles, false);
//Create function whose FunctionType is FT
Function *TheFunction = Function::Create(FT, Function::ExternalLinkage, "foofor", TheModule.get());
//give the name for Function args and save the args in innerargs
unsigned Idx = 0;
std::vector<Value *>innerargs;
for (auto &Arg : TheFunction->args()) {
Arg.setName(ArgNames[Idx++]);
innerargs.push_back(&Arg);
}
//this function's basic block
BasicBlock *BB = BasicBlock::Create(TheContext, "entry", TheFunction);
//create the loop BasicBlock
BasicBlock *LoopBB = BasicBlock::Create(TheContext, "loop", TheFunction);
//exit the block
BasicBlock *AfterBB = BasicBlock::Create(TheContext, "afterloop", TheFunction);
Builder.SetInsertPoint(BB);
//add goto LoopBB in BB
Builder.CreateBr(LoopBB);
//start with LoopBB
Builder.SetInsertPoint(LoopBB);
//start with 0
Value *StartVal = ConstantFP::get(TheContext, APFloat(0.0));
//step is 1
Value *StepVal = ConstantFP::get(TheContext, APFloat(1.0));
//local Variable which name is a
PHINode *Variable = Builder.CreatePHI(Type::getDoubleTy(TheContext), 1, "a");
//if it's from start block, set Variable 0
Variable->addIncoming(StartVal, BB);
//do the body then do cond
//emit loop body in LoopBB
//body: arg_a += 1
Value *ret = Builder.CreateFAdd(innerargs[0], ConstantFP::get(TheContext, APFloat(2.0)), "addtmp");
//do the cond,if Variable >= 10 then break(goto AfterBB)
Value *NextVar = Builder.CreateFAdd(StepVal, Variable, "nextvar");
//if Variable < 10 then goto LoopBB or gotot AfterBB
Value *cond = Builder.CreateFCmpULT(Variable, ConstantFP::get(TheContext, APFloat(10.0)), "cmptmp");
Builder.CreateCondBr(cond, LoopBB, AfterBB);
Builder.SetInsertPoint(AfterBB);
Variable->addIncoming(NextVar, LoopBB);
Builder.CreateRet(ret);
TheFunction->print(errs());
//using jit to run this code
auto H = TheJIT->addModule(std::move(TheModule));
auto ExprSymbol = TheJIT->findSymbol("foofor");
double (*foofor)(double) = (double (*)(double))(intptr_t)cantFail(ExprSymbol.getAddress());
cout <<foofor(40)<<endl;
}
Here's the LLVM IR form for a variable that increases:
define double #foofor(double %a) {
entry:
br label %loop
loop: ; preds = %loop, %entry
%a1 = phi double [ 0.000000e+00, %entry ], [ %nextvar, %loop ]
%a2 = phi double [ 0.000000e+00, %a ], [ %addtmp, %loop ]
%addtmp = fadd double %a2, 2.000000e+00
%nextvar = fadd double 1.000000e+00, %a1
%cmptmp = fcmp ult double %a1, 1.000000e+01
br i1 %cmptmp, label %loop, label %afterloop
afterloop: ; preds = %loop
ret double %addtmp
}
%a1/%nextvar ought to be ints as well, not doubles, but I left that since that's not the subject of your question.
I'm writing a LLVM IR generator for a pseudo code language. This language should allow redefinition of function.
Here is one case that I have two functions both named "f" but they have different parameters.
function f(int i, float r) returns int { return i; }
function f(float r, float r2) returns int {return i; }
I thought LLVM could distinct that, but I get
error: invalid redefinition of function
And the code I generated is:
define i32 #f(i32 %i, float %r) {
%var.i.0 = alloca i32
store i32 %i, i32* %var.i.0
%var.r.1 = alloca float
store float %r, float* %var.r.1
%int.2 = load i32* %var.i.0
ret i32 %int.2
; -- 0 :: i32
%int.3 = add i32 0, 0
ret i32 %int.3
}
define i32 #f(float %r, float %r2) {
%var.r.2 = alloca float
store float %r, float* %var.r.2
%var.r2.3 = alloca float
store float %r2, float* %var.r2.3
%var.i.4 = alloca i32
%float.3 = load float* %var.r.2
%int.7 = fptosi float %float.3 to i32
store i32 %int.7, i32* %var.i.4
%int.8 = load i32* %var.i.4
ret i32 %int.8
; -- 0 :: i32
%int.9 = add i32 0, 0
ret i32 %int.9
}
So, I think LLVM do not allow function overloading? Then is it a good idea that I generate a sequential counter, and distinct all these functions by adding this sequential counter as a suffix i.e define i32 #f.1() and define i32 #f.2()?
You're correct that LLVM IR doesn't have function overloading.
Using a sequential counter is probably not a good idea depending on how code in your language is organized. If you're just assigning incrementing integers, those may not be deterministic across the compilation of different files. For example, in C++, you might imagine something like
// library.cpp
int f(int i, float r) { ... }
int f(float r, float r2) { ... }
// user.cpp
extern int f(float r, float r2);
int foo() { return f(1.0, 2.0); }
When compiling user.cpp, there would be no way for the compiler to know that the f being referenced will actually be named f.2.
The typical way to implement function overloading is to use name mangling, somehow encoding the type signature of the function into its name so that it'll be unique in the presence of overloads.
My generator was written in java, so every time I parse a function definition, I will increase the counter for the same function name if the function name has already existed in the scope table.
My table is defined by Map with function name as key, and a list of function def as value:
Map<String,ArrayList<functionSymbol>> = new HashMap<>();
and then the constructor will look like:
static int counter = 0;
public FunctionSymbol(String functionName, Type retType, List<Variable> paramList){
this.functionName = functionName+counter;
this.paramList = paramList;
this.retType = retType;
counter++;
}
I am pretty new with llvm and having trouble digging deep into the following IR line:
%call2 = call float bitcast (float (float, i32*)* #function to float (float, i32 addrspace(1)*)*)(float %11, i32 addrspace(1)* %arrayidx)
What I need to extract from this is line the type of the arguments of the function (i.e., (float %11, i32 addrspace(1)* %arrayidx))
I have tried the following, and played arround with ConstExpr a little as well, but cannot get to extract that addrspace(1)
for (Function::iterator block = F.begin(), blockEnd = F.end(); block != blockEnd; ++block) {
for (BasicBlock::iterator inst = block->begin(), instEnd = block->end(); inst != instEnd; ++inst) {
if (CallInst *call = dyn_cast<CallInst>(inst)) {
Function *calledFunction = call->getCalledFunction();
if (calledFunction == NULL) { // Called function is wrapped in a bitcast
Value* v = call->getCalledValue();
calledFunction = dyn_cast<Function>(v->stripPointerCasts());
FunctionType *ft = calledFunction->getFunctionType(); // This gives me the type "from" (the args without addrspace(1)
for( Function::arg_iterator arg = calledFunction->arg_begin(), marg_end = calledFunction->arg_end(); arg != marg_end ; arg++){
Type *argTy = arg->getType();
if (PointerType *ptrTy = dyn_cast<PointerType>(argTy)) {
if( ptrTy->getAddressSpace() !=0)
...
}
}
}
}
}
}
The above code gives me the types (float, i32*) and not (float, i32 addrspace(1)*)
Any help please?
The llvm ir
%call2 = call float bitcast (float (float, i32*)* #function to float (float, i32 addrspace(1)*)*)(float %11, i32 addrspace(1)* %arrayidx)
is casting function type float (float, i32*) to float (float, i32 addrspace(1)*) and calling it with argument (%11, %arrayidx).
If you want the types of argument you can check it using callInst::getArgOperand to get arguments in call instruction itself.
for (Function::iterator block = F.begin(), blockEnd = F.end(); block != blockEnd; ++block) {
for (BasicBlock::iterator inst = block->begin(), instEnd = block->end(); inst != instEnd; ++inst) {
if (CallInst *call = dyn_cast<CallInst>(inst)) {
Value *val11 = call->getArgOperand(0);
Value *valarrayIdx = call->getArgOperand(1);
Type *val11ty = val11->getType(); // this should be of float
Type *valarrayIdx = valarrayIdx->getType(); // this should be of i32 address(1)*
}
}
}
CallInst::getCalledFunction will give you the function.
For more info you can go through http://llvm.org/docs/doxygen/html/classllvm_1_1CallInst.html