mirror of
https://github.com/RPCSX/llvm.git
synced 2024-12-03 17:31:50 +00:00
cf88e9244e
parts of the AA interface out of the base class of every single AA result object. Because this logic reformulates the query in terms of some other aspect of the API, it would easily cause O(n^2) query patterns in alias analysis. These could in turn be magnified further based on the number of call arguments, and then further based on the number of AA queries made for a particular call. This ended up causing problems for Rust that were actually noticable enough to get a bug (PR26564) and probably other places as well. When originally re-working the AA infrastructure, the desire was to regularize the pattern of refinement without losing any generality. While I think it was successful, that is clearly proving to be too costly. And the cost is needless: we gain no actual improvement for this generality of making a direct query to tbaa actually be able to re-use some other alias analysis's refinement logic for one of the other APIs, or some such. In short, this is entirely wasted work. To the extent possible, delegation to other API surfaces should be done at the aggregation layer so that we can avoid re-walking the aggregation. In fact, this significantly simplifies the logic as we no longer need to smuggle the aggregation layer into each alias analysis (or the TargetLibraryInfo into each alias analysis just so we can form argument memory locations!). However, we also have some delegation logic inside of BasicAA and some of it even makes sense. When the delegation logic is baking in specific knowledge of aliasing properties of the LLVM IR, as opposed to simply reformulating the query to utilize a different alias analysis interface entry point, it makes a lot of sense to restrict that logic to a different layer such as BasicAA. So one aspect of the delegation that was in every AA base class is that when we don't have operand bundles, we re-use function AA results as a fallback for callsite alias results. This relies on the IR properties of calls and functions w.r.t. aliasing, and so seems a better fit to BasicAA. I've lifted the logic up to that point where it seems to be a natural fit. This still does a bit of redundant work (we query function attributes twice, once via the callsite and once via the function AA query) but it is *exactly* twice here, no more. The end result is that all of the delegation logic is hoisted out of the base class and into either the aggregation layer when it is a pure retargeting to a different API surface, or into BasicAA when it relies on the IR's aliasing properties. This should fix the quadratic query pattern reported in PR26564, although I don't have a stand-alone test case to reproduce it. It also seems general goodness. Now the numerous AAs that don't need target library info don't carry it around and depend on it. I think I can even rip out the general access to the aggregation layer and only expose that in BasicAA as it is the only place where we re-query in that manner. However, this is a non-trivial change to the AA infrastructure so I want to get some additional eyes on this before it lands. Sadly, it can't wait long because we should really cherry pick this into 3.8 if we're going to go this route. Differential Revision: http://reviews.llvm.org/D17329 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@262490 91177308-0d34-0410-b5e6-96231b3b80d8
255 lines
8.3 KiB
C++
255 lines
8.3 KiB
C++
//===--- AliasAnalysisTest.cpp - Mixed TBAA unit tests --------------------===//
|
|
//
|
|
// The LLVM Compiler Infrastructure
|
|
//
|
|
// This file is distributed under the University of Illinois Open Source
|
|
// License. See LICENSE.TXT for details.
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
#include "llvm/Analysis/AliasAnalysis.h"
|
|
#include "llvm/ADT/SetVector.h"
|
|
#include "llvm/Analysis/AssumptionCache.h"
|
|
#include "llvm/Analysis/BasicAliasAnalysis.h"
|
|
#include "llvm/Analysis/TargetLibraryInfo.h"
|
|
#include "llvm/AsmParser/Parser.h"
|
|
#include "llvm/IR/Constants.h"
|
|
#include "llvm/IR/Instructions.h"
|
|
#include "llvm/IR/InstIterator.h"
|
|
#include "llvm/IR/LLVMContext.h"
|
|
#include "llvm/IR/LegacyPassManager.h"
|
|
#include "llvm/IR/Module.h"
|
|
#include "llvm/Support/CommandLine.h"
|
|
#include "llvm/Support/SourceMgr.h"
|
|
#include "gtest/gtest.h"
|
|
|
|
using namespace llvm;
|
|
|
|
// Set up some test passes.
|
|
namespace llvm {
|
|
void initializeAATestPassPass(PassRegistry&);
|
|
void initializeTestCustomAAWrapperPassPass(PassRegistry&);
|
|
}
|
|
|
|
namespace {
|
|
struct AATestPass : FunctionPass {
|
|
static char ID;
|
|
AATestPass() : FunctionPass(ID) {
|
|
initializeAATestPassPass(*PassRegistry::getPassRegistry());
|
|
}
|
|
|
|
void getAnalysisUsage(AnalysisUsage &AU) const override {
|
|
AU.addRequired<AAResultsWrapperPass>();
|
|
AU.setPreservesAll();
|
|
}
|
|
|
|
bool runOnFunction(Function &F) override {
|
|
AliasAnalysis &AA = getAnalysis<AAResultsWrapperPass>().getAAResults();
|
|
|
|
SetVector<Value *> Pointers;
|
|
for (Argument &A : F.args())
|
|
if (A.getType()->isPointerTy())
|
|
Pointers.insert(&A);
|
|
for (Instruction &I : instructions(F))
|
|
if (I.getType()->isPointerTy())
|
|
Pointers.insert(&I);
|
|
|
|
for (Value *P1 : Pointers)
|
|
for (Value *P2 : Pointers)
|
|
(void)AA.alias(P1, MemoryLocation::UnknownSize, P2,
|
|
MemoryLocation::UnknownSize);
|
|
|
|
return false;
|
|
}
|
|
};
|
|
}
|
|
|
|
char AATestPass::ID = 0;
|
|
INITIALIZE_PASS_BEGIN(AATestPass, "aa-test-pas", "Alias Analysis Test Pass",
|
|
false, true)
|
|
INITIALIZE_PASS_DEPENDENCY(AAResultsWrapperPass)
|
|
INITIALIZE_PASS_END(AATestPass, "aa-test-pass", "Alias Analysis Test Pass",
|
|
false, true)
|
|
|
|
namespace {
|
|
/// A test customizable AA result. It merely accepts a callback to run whenever
|
|
/// it receives an alias query. Useful for testing that a particular AA result
|
|
/// is reached.
|
|
struct TestCustomAAResult : AAResultBase<TestCustomAAResult> {
|
|
friend AAResultBase<TestCustomAAResult>;
|
|
|
|
std::function<void()> CB;
|
|
|
|
explicit TestCustomAAResult(std::function<void()> CB)
|
|
: AAResultBase(), CB(std::move(CB)) {}
|
|
TestCustomAAResult(TestCustomAAResult &&Arg)
|
|
: AAResultBase(std::move(Arg)), CB(std::move(Arg.CB)) {}
|
|
|
|
bool invalidate(Function &, const PreservedAnalyses &) { return false; }
|
|
|
|
AliasResult alias(const MemoryLocation &LocA, const MemoryLocation &LocB) {
|
|
CB();
|
|
return MayAlias;
|
|
}
|
|
};
|
|
}
|
|
|
|
namespace {
|
|
/// A wrapper pass for the legacy pass manager to use with the above custom AA
|
|
/// result.
|
|
class TestCustomAAWrapperPass : public ImmutablePass {
|
|
std::function<void()> CB;
|
|
std::unique_ptr<TestCustomAAResult> Result;
|
|
|
|
public:
|
|
static char ID;
|
|
|
|
explicit TestCustomAAWrapperPass(
|
|
std::function<void()> CB = std::function<void()>())
|
|
: ImmutablePass(ID), CB(std::move(CB)) {
|
|
initializeTestCustomAAWrapperPassPass(*PassRegistry::getPassRegistry());
|
|
}
|
|
|
|
void getAnalysisUsage(AnalysisUsage &AU) const override {
|
|
AU.setPreservesAll();
|
|
AU.addRequired<TargetLibraryInfoWrapperPass>();
|
|
}
|
|
|
|
bool doInitialization(Module &M) override {
|
|
Result.reset(new TestCustomAAResult(std::move(CB)));
|
|
return true;
|
|
}
|
|
|
|
bool doFinalization(Module &M) override {
|
|
Result.reset();
|
|
return true;
|
|
}
|
|
|
|
TestCustomAAResult &getResult() { return *Result; }
|
|
const TestCustomAAResult &getResult() const { return *Result; }
|
|
};
|
|
}
|
|
|
|
char TestCustomAAWrapperPass::ID = 0;
|
|
INITIALIZE_PASS_BEGIN(TestCustomAAWrapperPass, "test-custom-aa",
|
|
"Test Custom AA Wrapper Pass", false, true)
|
|
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
|
|
INITIALIZE_PASS_END(TestCustomAAWrapperPass, "test-custom-aa",
|
|
"Test Custom AA Wrapper Pass", false, true)
|
|
|
|
namespace {
|
|
|
|
class AliasAnalysisTest : public testing::Test {
|
|
protected:
|
|
LLVMContext C;
|
|
Module M;
|
|
TargetLibraryInfoImpl TLII;
|
|
TargetLibraryInfo TLI;
|
|
std::unique_ptr<AssumptionCache> AC;
|
|
std::unique_ptr<BasicAAResult> BAR;
|
|
std::unique_ptr<AAResults> AAR;
|
|
|
|
AliasAnalysisTest() : M("AliasAnalysisTest", C), TLI(TLII) {}
|
|
|
|
AAResults &getAAResults(Function &F) {
|
|
// Reset the Function AA results first to clear out any references.
|
|
AAR.reset(new AAResults(TLI));
|
|
|
|
// Build the various AA results and register them.
|
|
AC.reset(new AssumptionCache(F));
|
|
BAR.reset(new BasicAAResult(M.getDataLayout(), TLI, *AC));
|
|
AAR->addAAResult(*BAR);
|
|
|
|
return *AAR;
|
|
}
|
|
};
|
|
|
|
TEST_F(AliasAnalysisTest, getModRefInfo) {
|
|
// Setup function.
|
|
FunctionType *FTy =
|
|
FunctionType::get(Type::getVoidTy(C), std::vector<Type *>(), false);
|
|
auto *F = cast<Function>(M.getOrInsertFunction("f", FTy));
|
|
auto *BB = BasicBlock::Create(C, "entry", F);
|
|
auto IntType = Type::getInt32Ty(C);
|
|
auto PtrType = Type::getInt32PtrTy(C);
|
|
auto *Value = ConstantInt::get(IntType, 42);
|
|
auto *Addr = ConstantPointerNull::get(PtrType);
|
|
|
|
auto *Store1 = new StoreInst(Value, Addr, BB);
|
|
auto *Load1 = new LoadInst(Addr, "load", BB);
|
|
auto *Add1 = BinaryOperator::CreateAdd(Value, Value, "add", BB);
|
|
auto *VAArg1 = new VAArgInst(Addr, PtrType, "vaarg", BB);
|
|
auto *CmpXChg1 = new AtomicCmpXchgInst(Addr, ConstantInt::get(IntType, 0),
|
|
ConstantInt::get(IntType, 1),
|
|
Monotonic, Monotonic, CrossThread, BB);
|
|
auto *AtomicRMW =
|
|
new AtomicRMWInst(AtomicRMWInst::Xchg, Addr, ConstantInt::get(IntType, 1),
|
|
Monotonic, CrossThread, BB);
|
|
|
|
ReturnInst::Create(C, nullptr, BB);
|
|
|
|
auto &AA = getAAResults(*F);
|
|
|
|
// Check basic results
|
|
EXPECT_EQ(AA.getModRefInfo(Store1, MemoryLocation()), MRI_Mod);
|
|
EXPECT_EQ(AA.getModRefInfo(Store1), MRI_Mod);
|
|
EXPECT_EQ(AA.getModRefInfo(Load1, MemoryLocation()), MRI_Ref);
|
|
EXPECT_EQ(AA.getModRefInfo(Load1), MRI_Ref);
|
|
EXPECT_EQ(AA.getModRefInfo(Add1, MemoryLocation()), MRI_NoModRef);
|
|
EXPECT_EQ(AA.getModRefInfo(Add1), MRI_NoModRef);
|
|
EXPECT_EQ(AA.getModRefInfo(VAArg1, MemoryLocation()), MRI_ModRef);
|
|
EXPECT_EQ(AA.getModRefInfo(VAArg1), MRI_ModRef);
|
|
EXPECT_EQ(AA.getModRefInfo(CmpXChg1, MemoryLocation()), MRI_ModRef);
|
|
EXPECT_EQ(AA.getModRefInfo(CmpXChg1), MRI_ModRef);
|
|
EXPECT_EQ(AA.getModRefInfo(AtomicRMW, MemoryLocation()), MRI_ModRef);
|
|
EXPECT_EQ(AA.getModRefInfo(AtomicRMW), MRI_ModRef);
|
|
}
|
|
|
|
class AAPassInfraTest : public testing::Test {
|
|
protected:
|
|
LLVMContext &C;
|
|
SMDiagnostic Err;
|
|
std::unique_ptr<Module> M;
|
|
|
|
public:
|
|
AAPassInfraTest()
|
|
: C(getGlobalContext()),
|
|
M(parseAssemblyString("define i32 @f(i32* %x, i32* %y) {\n"
|
|
"entry:\n"
|
|
" %lx = load i32, i32* %x\n"
|
|
" %ly = load i32, i32* %y\n"
|
|
" %sum = add i32 %lx, %ly\n"
|
|
" ret i32 %sum\n"
|
|
"}\n",
|
|
Err, C)) {
|
|
assert(M && "Failed to build the module!");
|
|
}
|
|
};
|
|
|
|
TEST_F(AAPassInfraTest, injectExternalAA) {
|
|
legacy::PassManager PM;
|
|
|
|
// Register our custom AA's wrapper pass manually.
|
|
bool IsCustomAAQueried = false;
|
|
PM.add(new TestCustomAAWrapperPass([&] { IsCustomAAQueried = true; }));
|
|
|
|
// Now add the external AA wrapper with a lambda which queries for the
|
|
// wrapper around our custom AA and adds it to the results.
|
|
PM.add(createExternalAAWrapperPass([](Pass &P, Function &, AAResults &AAR) {
|
|
if (auto *WrapperPass = P.getAnalysisIfAvailable<TestCustomAAWrapperPass>())
|
|
AAR.addAAResult(WrapperPass->getResult());
|
|
}));
|
|
|
|
// And run a pass that will make some alias queries. This will automatically
|
|
// trigger the rest of the alias analysis stack to be run. It is analagous to
|
|
// building a full pass pipeline with any of the existing pass manager
|
|
// builders.
|
|
PM.add(new AATestPass());
|
|
PM.run(*M);
|
|
|
|
// Finally, ensure that our custom AA was indeed queried.
|
|
EXPECT_TRUE(IsCustomAAQueried);
|
|
}
|
|
|
|
} // end anonymous namspace
|