1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-19 19:12:56 +02:00
llvm-mirror/lib/LTO/LTOCodeGenerator.cpp

590 lines
19 KiB
C++
Raw Normal View History

//===-LTOCodeGenerator.cpp - LLVM Link Time Optimizer ---------------------===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
2012-03-31 12:50:14 +02:00
//
//===----------------------------------------------------------------------===//
//
2012-03-31 12:50:14 +02:00
// This file implements the Link Time Optimization library. This library is
// intended to be used by linker to optimize code at link time.
//
//===----------------------------------------------------------------------===//
#include "llvm/LTO/LTOCodeGenerator.h"
#include "llvm/ADT/StringExtras.h"
#include "llvm/Analysis/Passes.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/Analysis/TargetTransformInfo.h"
#include "llvm/Bitcode/ReaderWriter.h"
#include "llvm/CodeGen/ParallelCG.h"
#include "llvm/CodeGen/RuntimeLibcalls.h"
#include "llvm/Config/config.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/DerivedTypes.h"
#include "llvm/IR/DiagnosticInfo.h"
#include "llvm/IR/DiagnosticPrinter.h"
#include "llvm/IR/LLVMContext.h"
#include "llvm/IR/LegacyPassManager.h"
#include "llvm/IR/Mangler.h"
#include "llvm/IR/Module.h"
#include "llvm/IR/Verifier.h"
#include "llvm/InitializePasses.h"
#include "llvm/LTO/LTOModule.h"
#include "llvm/Linker/Linker.h"
#include "llvm/MC/MCAsmInfo.h"
#include "llvm/MC/MCContext.h"
#include "llvm/MC/SubtargetFeature.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/FileSystem.h"
#include "llvm/Support/Host.h"
#include "llvm/Support/MemoryBuffer.h"
#include "llvm/Support/Signals.h"
#include "llvm/Support/TargetRegistry.h"
#include "llvm/Support/TargetSelect.h"
#include "llvm/Support/ToolOutputFile.h"
#include "llvm/Support/raw_ostream.h"
#include "llvm/Target/TargetLowering.h"
#include "llvm/Target/TargetOptions.h"
#include "llvm/Target/TargetRegisterInfo.h"
#include "llvm/Target/TargetSubtargetInfo.h"
#include "llvm/Transforms/IPO.h"
#include "llvm/Transforms/IPO/PassManagerBuilder.h"
#include "llvm/Transforms/ObjCARC.h"
#include <system_error>
using namespace llvm;
const char* LTOCodeGenerator::getVersionString() {
#ifdef LLVM_VERSION_INFO
return PACKAGE_NAME " version " PACKAGE_VERSION ", " LLVM_VERSION_INFO;
#else
return PACKAGE_NAME " version " PACKAGE_VERSION;
#endif
}
static void handleLTODiagnostic(const DiagnosticInfo &DI) {
DiagnosticPrinterRawOStream DP(errs());
DI.print(DP);
errs() << "\n";
}
2012-03-31 12:50:14 +02:00
LTOCodeGenerator::LTOCodeGenerator()
: Context(getGlobalContext()),
MergedModule(new Module("ld-temp.o", Context)),
IRLinker(MergedModule.get(), handleLTODiagnostic) {
initializeLTOPasses();
}
LTOCodeGenerator::LTOCodeGenerator(std::unique_ptr<LLVMContext> Context)
: OwnedContext(std::move(Context)), Context(*OwnedContext),
MergedModule(new Module("ld-temp.o", *OwnedContext)),
IRLinker(MergedModule.get(), handleLTODiagnostic) {
initializeLTOPasses();
}
LTOCodeGenerator::~LTOCodeGenerator() {}
2015-09-01 12:13:49 +02:00
// Initialize LTO passes. Please keep this function in sync with
// PassManagerBuilder::populateLTOPassManager(), and make sure all LTO
// passes are initialized.
void LTOCodeGenerator::initializeLTOPasses() {
PassRegistry &R = *PassRegistry::getPassRegistry();
initializeInternalizePassPass(R);
initializeIPSCCPPass(R);
initializeGlobalOptPass(R);
initializeConstantMergePass(R);
initializeDAHPass(R);
initializeInstructionCombiningPassPass(R);
initializeSimpleInlinerPass(R);
initializePruneEHPass(R);
initializeGlobalDCEPass(R);
initializeArgPromotionPass(R);
initializeJumpThreadingPass(R);
initializeSROAPass(R);
initializeSROA_DTPass(R);
initializeSROA_SSAUpPass(R);
initializeFunctionAttrsPass(R);
[PM/AA] Rebuild LLVM's alias analysis infrastructure in a way compatible with the new pass manager, and no longer relying on analysis groups. This builds essentially a ground-up new AA infrastructure stack for LLVM. The core ideas are the same that are used throughout the new pass manager: type erased polymorphism and direct composition. The design is as follows: - FunctionAAResults is a type-erasing alias analysis results aggregation interface to walk a single query across a range of results from different alias analyses. Currently this is function-specific as we always assume that aliasing queries are *within* a function. - AAResultBase is a CRTP utility providing stub implementations of various parts of the alias analysis result concept, notably in several cases in terms of other more general parts of the interface. This can be used to implement only a narrow part of the interface rather than the entire interface. This isn't really ideal, this logic should be hoisted into FunctionAAResults as currently it will cause a significant amount of redundant work, but it faithfully models the behavior of the prior infrastructure. - All the alias analysis passes are ported to be wrapper passes for the legacy PM and new-style analysis passes for the new PM with a shared result object. In some cases (most notably CFL), this is an extremely naive approach that we should revisit when we can specialize for the new pass manager. - BasicAA has been restructured to reflect that it is much more fundamentally a function analysis because it uses dominator trees and loop info that need to be constructed for each function. All of the references to getting alias analysis results have been updated to use the new aggregation interface. All the preservation and other pass management code has been updated accordingly. The way the FunctionAAResultsWrapperPass works is to detect the available alias analyses when run, and add them to the results object. This means that we should be able to continue to respect when various passes are added to the pipeline, for example adding CFL or adding TBAA passes should just cause their results to be available and to get folded into this. The exception to this rule is BasicAA which really needs to be a function pass due to using dominator trees and loop info. As a consequence, the FunctionAAResultsWrapperPass directly depends on BasicAA and always includes it in the aggregation. This has significant implications for preserving analyses. Generally, most passes shouldn't bother preserving FunctionAAResultsWrapperPass because rebuilding the results just updates the set of known AA passes. The exception to this rule are LoopPass instances which need to preserve all the function analyses that the loop pass manager will end up needing. This means preserving both BasicAAWrapperPass and the aggregating FunctionAAResultsWrapperPass. Now, when preserving an alias analysis, you do so by directly preserving that analysis. This is only necessary for non-immutable-pass-provided alias analyses though, and there are only three of interest: BasicAA, GlobalsAA (formerly GlobalsModRef), and SCEVAA. Usually BasicAA is preserved when needed because it (like DominatorTree and LoopInfo) is marked as a CFG-only pass. I've expanded GlobalsAA into the preserved set everywhere we previously were preserving all of AliasAnalysis, and I've added SCEVAA in the intersection of that with where we preserve SCEV itself. One significant challenge to all of this is that the CGSCC passes were actually using the alias analysis implementations by taking advantage of a pretty amazing set of loop holes in the old pass manager's analysis management code which allowed analysis groups to slide through in many cases. Moving away from analysis groups makes this problem much more obvious. To fix it, I've leveraged the flexibility the design of the new PM components provides to just directly construct the relevant alias analyses for the relevant functions in the IPO passes that need them. This is a bit hacky, but should go away with the new pass manager, and is already in many ways cleaner than the prior state. Another significant challenge is that various facilities of the old alias analysis infrastructure just don't fit any more. The most significant of these is the alias analysis 'counter' pass. That pass relied on the ability to snoop on AA queries at different points in the analysis group chain. Instead, I'm planning to build printing functionality directly into the aggregation layer. I've not included that in this patch merely to keep it smaller. Note that all of this needs a nearly complete rewrite of the AA documentation. I'm planning to do that, but I'd like to make sure the new design settles, and to flesh out a bit more of what it looks like in the new pass manager first. Differential Revision: http://reviews.llvm.org/D12080 llvm-svn: 247167
2015-09-09 19:55:00 +02:00
initializeGlobalsAAWrapperPassPass(R);
initializeLICMPass(R);
initializeMergedLoadStoreMotionPass(R);
initializeGVNPass(R);
initializeMemCpyOptPass(R);
initializeDCEPass(R);
initializeCFGSimplifyPassPass(R);
}
bool LTOCodeGenerator::addModule(LTOModule *Mod) {
assert(&Mod->getModule().getContext() == &Context &&
"Expected module in same context");
bool ret = IRLinker.linkInModule(&Mod->getModule());
const std::vector<const char *> &undefs = Mod->getAsmUndefinedRefs();
for (int i = 0, e = undefs.size(); i != e; ++i)
AsmUndefinedRefs[undefs[i]] = 1;
return !ret;
}
2012-03-31 12:50:14 +02:00
void LTOCodeGenerator::setModule(std::unique_ptr<LTOModule> Mod) {
assert(&Mod->getModule().getContext() == &Context &&
"Expected module in same context");
AsmUndefinedRefs.clear();
MergedModule = Mod->takeModule();
IRLinker.setModule(MergedModule.get());
const std::vector<const char*> &Undefs = Mod->getAsmUndefinedRefs();
for (int I = 0, E = Undefs.size(); I != E; ++I)
AsmUndefinedRefs[Undefs[I]] = 1;
}
void LTOCodeGenerator::setTargetOptions(TargetOptions Options) {
this->Options = Options;
}
void LTOCodeGenerator::setDebugInfo(lto_debug_model Debug) {
switch (Debug) {
case LTO_DEBUG_MODEL_NONE:
EmitDwarfDebugInfo = false;
return;
case LTO_DEBUG_MODEL_DWARF:
EmitDwarfDebugInfo = true;
return;
}
llvm_unreachable("Unknown debug format!");
}
void LTOCodeGenerator::setOptLevel(unsigned Level) {
OptLevel = Level;
switch (OptLevel) {
case 0:
CGOptLevel = CodeGenOpt::None;
break;
case 1:
CGOptLevel = CodeGenOpt::Less;
break;
case 2:
CGOptLevel = CodeGenOpt::Default;
break;
case 3:
CGOptLevel = CodeGenOpt::Aggressive;
break;
}
}
bool LTOCodeGenerator::writeMergedModules(const char *Path,
std::string &ErrMsg) {
if (!determineTarget(ErrMsg))
return false;
// mark which symbols can not be internalized
applyScopeRestrictions();
// create output file
std::error_code EC;
tool_output_file Out(Path, EC, sys::fs::F_None);
if (EC) {
ErrMsg = "could not open bitcode file for writing: ";
ErrMsg += Path;
return false;
}
2012-03-31 12:50:14 +02:00
// write bitcode to it
WriteBitcodeToFile(MergedModule.get(), Out.os(), ShouldEmbedUselists);
Out.os().close();
if (Out.os().has_error()) {
ErrMsg = "could not write bitcode file: ";
ErrMsg += Path;
Out.os().clear_error();
return false;
}
2012-03-31 12:50:14 +02:00
Out.keep();
return true;
}
bool LTOCodeGenerator::compileOptimizedToFile(const char **Name,
std::string &ErrMsg) {
// make unique temp .o file to put generated object file
SmallString<128> Filename;
int FD;
std::error_code EC =
sys::fs::createTemporaryFile("lto-llvm", "o", FD, Filename);
if (EC) {
ErrMsg = EC.message();
return false;
}
// generate object file
tool_output_file objFile(Filename.c_str(), FD);
bool genResult = compileOptimized(&objFile.os(), ErrMsg);
objFile.os().close();
if (objFile.os().has_error()) {
objFile.os().clear_error();
sys::fs::remove(Twine(Filename));
return false;
}
objFile.keep();
if (!genResult) {
sys::fs::remove(Twine(Filename));
return false;
}
NativeObjectPath = Filename.c_str();
*Name = NativeObjectPath.c_str();
return true;
}
std::unique_ptr<MemoryBuffer>
LTOCodeGenerator::compileOptimized(std::string &ErrMsg) {
const char *name;
if (!compileOptimizedToFile(&name, ErrMsg))
return nullptr;
// read .o file into memory buffer
ErrorOr<std::unique_ptr<MemoryBuffer>> BufferOrErr =
MemoryBuffer::getFile(name, -1, false);
if (std::error_code EC = BufferOrErr.getError()) {
ErrMsg = EC.message();
sys::fs::remove(NativeObjectPath);
return nullptr;
}
// remove temp files
sys::fs::remove(NativeObjectPath);
return std::move(*BufferOrErr);
}
bool LTOCodeGenerator::compile_to_file(const char **Name, bool DisableInline,
bool DisableGVNLoadPRE,
bool DisableVectorization,
std::string &ErrMsg) {
if (!optimize(DisableInline, DisableGVNLoadPRE, DisableVectorization, ErrMsg))
return false;
return compileOptimizedToFile(Name, ErrMsg);
}
std::unique_ptr<MemoryBuffer>
LTOCodeGenerator::compile(bool DisableInline, bool DisableGVNLoadPRE,
bool DisableVectorization, std::string &ErrMsg) {
if (!optimize(DisableInline, DisableGVNLoadPRE, DisableVectorization, ErrMsg))
return nullptr;
return compileOptimized(ErrMsg);
}
bool LTOCodeGenerator::determineTarget(std::string &ErrMsg) {
if (TargetMach)
return true;
std::string TripleStr = MergedModule->getTargetTriple();
if (TripleStr.empty()) {
TripleStr = sys::getDefaultTargetTriple();
MergedModule->setTargetTriple(TripleStr);
}
llvm::Triple Triple(TripleStr);
// create target machine from info for merged modules
const Target *march = TargetRegistry::lookupTarget(TripleStr, ErrMsg);
if (!march)
return false;
// Construct LTOModule, hand over ownership of module and target. Use MAttr as
// the default set of features.
SubtargetFeatures Features(MAttr);
Features.getDefaultSubtargetFeatures(Triple);
FeatureStr = Features.getString();
// Set a default CPU for Darwin triples.
if (MCpu.empty() && Triple.isOSDarwin()) {
if (Triple.getArch() == llvm::Triple::x86_64)
MCpu = "core2";
else if (Triple.getArch() == llvm::Triple::x86)
MCpu = "yonah";
else if (Triple.getArch() == llvm::Triple::aarch64)
MCpu = "cyclone";
}
TargetMach.reset(march->createTargetMachine(TripleStr, MCpu, FeatureStr,
Options, RelocModel,
CodeModel::Default, CGOptLevel));
return true;
}
void LTOCodeGenerator::
applyRestriction(GlobalValue &GV,
ArrayRef<StringRef> Libcalls,
std::vector<const char*> &MustPreserveList,
SmallPtrSetImpl<GlobalValue*> &AsmUsed,
Mangler &Mangler) {
// There are no restrictions to apply to declarations.
if (GV.isDeclaration())
return;
// There is nothing more restrictive than private linkage.
if (GV.hasPrivateLinkage())
return;
SmallString<64> Buffer;
TargetMach->getNameWithPrefix(Buffer, &GV, Mangler);
if (MustPreserveSymbols.count(Buffer))
MustPreserveList.push_back(GV.getName().data());
if (AsmUndefinedRefs.count(Buffer))
AsmUsed.insert(&GV);
// Conservatively append user-supplied runtime library functions to
// llvm.compiler.used. These could be internalized and deleted by
// optimizations like -globalopt, causing problems when later optimizations
// add new library calls (e.g., llvm.memset => memset and printf => puts).
// Leave it to the linker to remove any dead code (e.g. with -dead_strip).
if (isa<Function>(GV) &&
std::binary_search(Libcalls.begin(), Libcalls.end(), GV.getName()))
AsmUsed.insert(&GV);
}
static void findUsedValues(GlobalVariable *LLVMUsed,
SmallPtrSetImpl<GlobalValue*> &UsedValues) {
if (!LLVMUsed) return;
ConstantArray *Inits = cast<ConstantArray>(LLVMUsed->getInitializer());
for (unsigned i = 0, e = Inits->getNumOperands(); i != e; ++i)
2012-03-31 12:50:14 +02:00
if (GlobalValue *GV =
dyn_cast<GlobalValue>(Inits->getOperand(i)->stripPointerCasts()))
UsedValues.insert(GV);
}
// Collect names of runtime library functions. User-defined functions with the
// same names are added to llvm.compiler.used to prevent them from being
// deleted by optimizations.
static void accumulateAndSortLibcalls(std::vector<StringRef> &Libcalls,
const TargetLibraryInfo& TLI,
const Module &Mod,
const TargetMachine &TM) {
// TargetLibraryInfo has info on C runtime library calls on the current
// target.
for (unsigned I = 0, E = static_cast<unsigned>(LibFunc::NumLibFuncs);
I != E; ++I) {
LibFunc::Func F = static_cast<LibFunc::Func>(I);
if (TLI.has(F))
Libcalls.push_back(TLI.getName(F));
}
SmallPtrSet<const TargetLowering *, 1> TLSet;
for (const Function &F : Mod) {
const TargetLowering *Lowering =
TM.getSubtargetImpl(F)->getTargetLowering();
if (Lowering && TLSet.insert(Lowering).second)
// TargetLowering has info on library calls that CodeGen expects to be
// available, both from the C runtime and compiler-rt.
for (unsigned I = 0, E = static_cast<unsigned>(RTLIB::UNKNOWN_LIBCALL);
I != E; ++I)
if (const char *Name =
Lowering->getLibcallName(static_cast<RTLIB::Libcall>(I)))
Libcalls.push_back(Name);
}
array_pod_sort(Libcalls.begin(), Libcalls.end());
Libcalls.erase(std::unique(Libcalls.begin(), Libcalls.end()),
Libcalls.end());
}
void LTOCodeGenerator::applyScopeRestrictions() {
if (ScopeRestrictionsDone || !ShouldInternalize)
return;
// Start off with a verification pass.
legacy::PassManager passes;
passes.add(createVerifierPass());
2012-03-31 12:50:14 +02:00
// mark which symbols can not be internalized
Mangler Mangler;
std::vector<const char*> MustPreserveList;
SmallPtrSet<GlobalValue*, 8> AsmUsed;
std::vector<StringRef> Libcalls;
[PM] Rework how the TargetLibraryInfo pass integrates with the new pass manager to support the actual uses of it. =] When I ported instcombine to the new pass manager I discover that it didn't work because TLI wasn't available in the right places. This is a somewhat surprising and/or subtle aspect of the new pass manager design that came up before but I think is useful to be reminded of: While the new pass manager *allows* a function pass to query a module analysis, it requires that the module analysis is already run and cached prior to the function pass manager starting up, possibly with a 'require<foo>' style utility in the pass pipeline. This is an intentional hurdle because using a module analysis from a function pass *requires* that the module analysis is run prior to entering the function pass manager. Otherwise the other functions in the module could be in who-knows-what state, etc. A somewhat surprising consequence of this design decision (at least to me) is that you have to design a function pass that leverages a module analysis to do so as an optional feature. Even if that means your function pass does no work in the absence of the module analysis, you have to handle that possibility and remain conservatively correct. This is a natural consequence of things being able to invalidate the module analysis and us being unable to re-run it. And it's a generally good thing because it lets us reorder passes arbitrarily without breaking correctness, etc. This ends up causing problems in one case. What if we have a module analysis that is *definitionally* impossible to invalidate. In the places this might come up, the analysis is usually also definitionally trivial to run even while other transformation passes run on the module, regardless of the state of anything. And so, it follows that it is natural to have a hard requirement on such analyses from a function pass. It turns out, that TargetLibraryInfo is just such an analysis, and InstCombine has a hard requirement on it. The approach I've taken here is to produce an analysis that models this flexibility by making it both a module and a function analysis. This exposes the fact that it is in fact safe to compute at any point. We can even make it a valid CGSCC analysis at some point if that is useful. However, we don't want to have a copy of the actual target library info state for each function! This state is specific to the triple. The somewhat direct and blunt approach here is to turn TLI into a pimpl, with the state and mutators in the implementation class and the query routines primarily in the wrapper. Then the analysis can lazily construct and cache the implementations, keyed on the triple, and on-demand produce wrappers of them for each function. One minor annoyance is that we will end up with a wrapper for each function in the module. While this is a bit wasteful (one pointer per function) it seems tolerable. And it has the advantage of ensuring that we pay the absolute minimum synchronization cost to access this information should we end up with a nice parallel function pass manager in the future. We could look into trying to mark when analysis results are especially cheap to recompute and more eagerly GC-ing the cached results, or we could look at supporting a variant of analyses whose results are specifically *not* cached and expected to just be used and discarded by the consumer. Either way, these seem like incremental enhancements that should happen when we start profiling the memory and CPU usage of the new pass manager and not before. The other minor annoyance is that if we end up using the TLI in both a module pass and a function pass, those will be produced by two separate analyses, and thus will point to separate copies of the implementation state. While a minor issue, I dislike this and would like to find a way to cleanly allow a single analysis instance to be used across multiple IR unit managers. But I don't have a good solution to this today, and I don't want to hold up all of the work waiting to come up with one. This too seems like a reasonable thing to incrementally improve later. llvm-svn: 226981
2015-01-24 03:06:09 +01:00
TargetLibraryInfoImpl TLII(Triple(TargetMach->getTargetTriple()));
TargetLibraryInfo TLI(TLII);
accumulateAndSortLibcalls(Libcalls, TLI, *MergedModule, *TargetMach);
for (Function &f : *MergedModule)
applyRestriction(f, Libcalls, MustPreserveList, AsmUsed, Mangler);
for (GlobalVariable &v : MergedModule->globals())
applyRestriction(v, Libcalls, MustPreserveList, AsmUsed, Mangler);
for (GlobalAlias &a : MergedModule->aliases())
applyRestriction(a, Libcalls, MustPreserveList, AsmUsed, Mangler);
GlobalVariable *LLVMCompilerUsed =
MergedModule->getGlobalVariable("llvm.compiler.used");
findUsedValues(LLVMCompilerUsed, AsmUsed);
if (LLVMCompilerUsed)
LLVMCompilerUsed->eraseFromParent();
if (!AsmUsed.empty()) {
llvm::Type *i8PTy = llvm::Type::getInt8PtrTy(Context);
std::vector<Constant*> asmUsed2;
2014-05-05 22:06:41 +02:00
for (auto *GV : AsmUsed) {
Constant *c = ConstantExpr::getBitCast(GV, i8PTy);
asmUsed2.push_back(c);
}
llvm::ArrayType *ATy = llvm::ArrayType::get(i8PTy, asmUsed2.size());
LLVMCompilerUsed =
new llvm::GlobalVariable(*MergedModule, ATy, false,
llvm::GlobalValue::AppendingLinkage,
llvm::ConstantArray::get(ATy, asmUsed2),
"llvm.compiler.used");
LLVMCompilerUsed->setSection("llvm.metadata");
}
passes.add(createInternalizePass(MustPreserveList));
// apply scope restrictions
passes.run(*MergedModule);
2012-03-31 12:50:14 +02:00
ScopeRestrictionsDone = true;
}
/// Optimize merged modules using various IPO passes
bool LTOCodeGenerator::optimize(bool DisableInline, bool DisableGVNLoadPRE,
bool DisableVectorization,
std::string &ErrMsg) {
if (!this->determineTarget(ErrMsg))
return false;
// Mark which symbols can not be internalized
this->applyScopeRestrictions();
// Instantiate the pass manager to organize the passes.
legacy::PassManager passes;
// Add an appropriate DataLayout instance for this module...
MergedModule->setDataLayout(TargetMach->createDataLayout());
2012-03-31 12:50:14 +02:00
passes.add(
createTargetTransformInfoWrapperPass(TargetMach->getTargetIRAnalysis()));
Triple TargetTriple(TargetMach->getTargetTriple());
PassManagerBuilder PMB;
PMB.DisableGVNLoadPRE = DisableGVNLoadPRE;
PMB.LoopVectorize = !DisableVectorization;
PMB.SLPVectorize = !DisableVectorization;
if (!DisableInline)
PMB.Inliner = createFunctionInliningPass();
[PM] Rework how the TargetLibraryInfo pass integrates with the new pass manager to support the actual uses of it. =] When I ported instcombine to the new pass manager I discover that it didn't work because TLI wasn't available in the right places. This is a somewhat surprising and/or subtle aspect of the new pass manager design that came up before but I think is useful to be reminded of: While the new pass manager *allows* a function pass to query a module analysis, it requires that the module analysis is already run and cached prior to the function pass manager starting up, possibly with a 'require<foo>' style utility in the pass pipeline. This is an intentional hurdle because using a module analysis from a function pass *requires* that the module analysis is run prior to entering the function pass manager. Otherwise the other functions in the module could be in who-knows-what state, etc. A somewhat surprising consequence of this design decision (at least to me) is that you have to design a function pass that leverages a module analysis to do so as an optional feature. Even if that means your function pass does no work in the absence of the module analysis, you have to handle that possibility and remain conservatively correct. This is a natural consequence of things being able to invalidate the module analysis and us being unable to re-run it. And it's a generally good thing because it lets us reorder passes arbitrarily without breaking correctness, etc. This ends up causing problems in one case. What if we have a module analysis that is *definitionally* impossible to invalidate. In the places this might come up, the analysis is usually also definitionally trivial to run even while other transformation passes run on the module, regardless of the state of anything. And so, it follows that it is natural to have a hard requirement on such analyses from a function pass. It turns out, that TargetLibraryInfo is just such an analysis, and InstCombine has a hard requirement on it. The approach I've taken here is to produce an analysis that models this flexibility by making it both a module and a function analysis. This exposes the fact that it is in fact safe to compute at any point. We can even make it a valid CGSCC analysis at some point if that is useful. However, we don't want to have a copy of the actual target library info state for each function! This state is specific to the triple. The somewhat direct and blunt approach here is to turn TLI into a pimpl, with the state and mutators in the implementation class and the query routines primarily in the wrapper. Then the analysis can lazily construct and cache the implementations, keyed on the triple, and on-demand produce wrappers of them for each function. One minor annoyance is that we will end up with a wrapper for each function in the module. While this is a bit wasteful (one pointer per function) it seems tolerable. And it has the advantage of ensuring that we pay the absolute minimum synchronization cost to access this information should we end up with a nice parallel function pass manager in the future. We could look into trying to mark when analysis results are especially cheap to recompute and more eagerly GC-ing the cached results, or we could look at supporting a variant of analyses whose results are specifically *not* cached and expected to just be used and discarded by the consumer. Either way, these seem like incremental enhancements that should happen when we start profiling the memory and CPU usage of the new pass manager and not before. The other minor annoyance is that if we end up using the TLI in both a module pass and a function pass, those will be produced by two separate analyses, and thus will point to separate copies of the implementation state. While a minor issue, I dislike this and would like to find a way to cleanly allow a single analysis instance to be used across multiple IR unit managers. But I don't have a good solution to this today, and I don't want to hold up all of the work waiting to come up with one. This too seems like a reasonable thing to incrementally improve later. llvm-svn: 226981
2015-01-24 03:06:09 +01:00
PMB.LibraryInfo = new TargetLibraryInfoImpl(TargetTriple);
PMB.OptLevel = OptLevel;
PMB.VerifyInput = true;
PMB.VerifyOutput = true;
PMB.populateLTOPassManager(passes);
// Run our queue of passes all at once now, efficiently.
passes.run(*MergedModule);
return true;
}
bool LTOCodeGenerator::compileOptimized(ArrayRef<raw_pwrite_stream *> Out,
std::string &ErrMsg) {
if (!this->determineTarget(ErrMsg))
return false;
legacy::PassManager preCodeGenPasses;
// If the bitcode files contain ARC code and were compiled with optimization,
// the ObjCARCContractPass must be run, so do it unconditionally here.
preCodeGenPasses.add(createObjCARCContractPass());
preCodeGenPasses.run(*MergedModule);
// Do code generation. We need to preserve the module in case the client calls
// writeMergedModules() after compilation, but we only need to allow this at
// parallelism level 1. This is achieved by having splitCodeGen return the
// original module at parallelism level 1 which we then assign back to
// MergedModule.
MergedModule =
splitCodeGen(std::move(MergedModule), Out, MCpu, FeatureStr, Options,
RelocModel, CodeModel::Default, CGOptLevel);
return true;
}
/// setCodeGenDebugOptions - Set codegen debugging options to aid in debugging
/// LTO problems.
void LTOCodeGenerator::setCodeGenDebugOptions(const char *Options) {
for (std::pair<StringRef, StringRef> o = getToken(Options); !o.first.empty();
o = getToken(o.second))
CodegenOptions.push_back(o.first);
}
void LTOCodeGenerator::parseCodeGenDebugOptions() {
// if options were requested, set them
if (!CodegenOptions.empty()) {
// ParseCommandLineOptions() expects argv[0] to be program name.
std::vector<const char *> CodegenArgv(1, "libLLVMLTO");
for (std::string &Arg : CodegenOptions)
CodegenArgv.push_back(Arg.c_str());
cl::ParseCommandLineOptions(CodegenArgv.size(), CodegenArgv.data());
}
}
void LTOCodeGenerator::DiagnosticHandler(const DiagnosticInfo &DI,
void *Context) {
((LTOCodeGenerator *)Context)->DiagnosticHandler2(DI);
}
void LTOCodeGenerator::DiagnosticHandler2(const DiagnosticInfo &DI) {
// Map the LLVM internal diagnostic severity to the LTO diagnostic severity.
lto_codegen_diagnostic_severity_t Severity;
switch (DI.getSeverity()) {
case DS_Error:
Severity = LTO_DS_ERROR;
break;
case DS_Warning:
Severity = LTO_DS_WARNING;
break;
case DS_Remark:
Severity = LTO_DS_REMARK;
break;
case DS_Note:
Severity = LTO_DS_NOTE;
break;
}
// Create the string that will be reported to the external diagnostic handler.
std::string MsgStorage;
raw_string_ostream Stream(MsgStorage);
DiagnosticPrinterRawOStream DP(Stream);
DI.print(DP);
Stream.flush();
// If this method has been called it means someone has set up an external
// diagnostic handler. Assert on that.
assert(DiagHandler && "Invalid diagnostic handler");
(*DiagHandler)(Severity, MsgStorage.c_str(), DiagContext);
}
void
LTOCodeGenerator::setDiagnosticHandler(lto_diagnostic_handler_t DiagHandler,
void *Ctxt) {
this->DiagHandler = DiagHandler;
this->DiagContext = Ctxt;
if (!DiagHandler)
return Context.setDiagnosticHandler(nullptr, nullptr);
// Register the LTOCodeGenerator stub in the LLVMContext to forward the
// diagnostic to the external DiagHandler.
Context.setDiagnosticHandler(LTOCodeGenerator::DiagnosticHandler, this,
/* RespectFilters */ true);
}