[PM] Port MergedLoadStoreMotion to the new pass manager, take two.

This is indeed a much cleaner approach (thanks to Daniel Berlin
for pointing out), and also David/Sean for review.

Differential Revision:  http://reviews.llvm.org/D21454


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@273032 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Davide Italiano 2016-06-17 19:10:09 +00:00
parent c22fa67ab0
commit e92bed3b5e
8 changed files with 125 additions and 56 deletions

View File

@ -226,7 +226,7 @@ void initializeMemoryDependenceWrapperPassPass(PassRegistry&);
void initializeMemorySSAWrapperPassPass(PassRegistry&);
void initializeMemorySanitizerPass(PassRegistry&);
void initializeMergeFunctionsPass(PassRegistry&);
void initializeMergedLoadStoreMotionPass(PassRegistry &);
void initializeMergedLoadStoreMotionLegacyPassPass(PassRegistry &);
void initializeMetaRenamerPass(PassRegistry&);
void initializeModuleDebugInfoPrinterPass(PassRegistry&);
void initializeModuleSummaryIndexWrapperPassPass(PassRegistry &);

View File

@ -0,0 +1,39 @@
//===- MergedLoadStoreMotion.h - merge and hoist/sink load/stores ---------===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
//! \file
//! \brief This pass performs merges of loads and stores on both sides of a
// diamond (hammock). It hoists the loads and sinks the stores.
//
// The algorithm iteratively hoists two loads to the same address out of a
// diamond (hammock) and merges them into a single load in the header. Similar
// it sinks and merges two stores to the tail block (footer). The algorithm
// iterates over the instructions of one side of the diamond and attempts to
// find a matching load/store on the other side. It hoists / sinks when it
// thinks it safe to do so. This optimization helps with eg. hiding load
// latencies, triggering if-conversion, and reducing static code size.
//
//===----------------------------------------------------------------------===//
#ifndef LLVM_TRANSFORMS_SCALAR_MERGEDLOADSTOREMOTION_H
#define LLVM_TRANSFORMS_SCALAR_MERGEDLOADSTOREMOTION_H
#include "llvm/IR/Module.h"
#include "llvm/IR/PassManager.h"
namespace llvm {
class MergedLoadStoreMotionPass
: public PassInfoMixin<MergedLoadStoreMotionPass> {
public:
PreservedAnalyses run(Function &F, AnalysisManager<Function> &AM);
};
}
#endif // LLVM_TRANSFORMS_SCALAR_MERGEDLOADSTOREMOTION_H

View File

@ -123,7 +123,7 @@ void LTOCodeGenerator::initializeLTOPasses() {
initializeReversePostOrderFunctionAttrsLegacyPassPass(R);
initializeGlobalsAAWrapperPassPass(R);
initializeLICMPass(R);
initializeMergedLoadStoreMotionPass(R);
initializeMergedLoadStoreMotionLegacyPassPass(R);
initializeGVNLegacyPassPass(R);
initializeMemCpyOptLegacyPassPass(R);
initializeDCELegacyPassPass(R);

View File

@ -83,6 +83,7 @@
#include "llvm/Transforms/Scalar/LowerAtomic.h"
#include "llvm/Transforms/Scalar/LowerExpectIntrinsic.h"
#include "llvm/Transforms/Scalar/MemCpyOptimizer.h"
#include "llvm/Transforms/Scalar/MergedLoadStoreMotion.h"
#include "llvm/Transforms/Scalar/PartiallyInlineLibCalls.h"
#include "llvm/Transforms/Scalar/Reassociate.h"
#include "llvm/Transforms/Scalar/SCCP.h"

View File

@ -134,6 +134,7 @@ FUNCTION_PASS("guard-widening", GuardWideningPass())
FUNCTION_PASS("gvn", GVN())
FUNCTION_PASS("mem2reg", PromotePass())
FUNCTION_PASS("memcpyopt", MemCpyOptPass())
FUNCTION_PASS("mldst-motion", MergedLoadStoreMotionPass())
FUNCTION_PASS("jump-threading", JumpThreadingPass())
FUNCTION_PASS("partially-inline-libcalls", PartiallyInlineLibCallsPass())
FUNCTION_PASS("lcssa", LCSSAPass())

View File

@ -72,6 +72,7 @@
//
//===----------------------------------------------------------------------===//
#include "llvm/Transforms/Scalar/MergedLoadStoreMotion.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/Analysis/CFG.h"
@ -95,32 +96,20 @@ using namespace llvm;
//===----------------------------------------------------------------------===//
// MergedLoadStoreMotion Pass
//===----------------------------------------------------------------------===//
class MergedLoadStoreMotion {
MemoryDependenceResults *MD = nullptr;
AliasAnalysis *AA = nullptr;
namespace {
class MergedLoadStoreMotion : public FunctionPass {
AliasAnalysis *AA;
MemoryDependenceResults *MD;
// The mergeLoad/Store algorithms could have Size0 * Size1 complexity,
// where Size0 and Size1 are the #instructions on the two sides of
// the diamond. The constant chosen here is arbitrary. Compiler Time
// Control is enforced by the check Size0 * Size1 < MagicCompileTimeControl.
const int MagicCompileTimeControl = 250;
public:
static char ID; // Pass identification, replacement for typeid
MergedLoadStoreMotion()
: FunctionPass(ID), MD(nullptr), MagicCompileTimeControl(250) {
initializeMergedLoadStoreMotionPass(*PassRegistry::getPassRegistry());
}
bool runOnFunction(Function &F) override;
bool run(Function &F, MemoryDependenceResults *MD, AliasAnalysis &AA);
private:
// This transformation requires dominator postdominator info
void getAnalysisUsage(AnalysisUsage &AU) const override {
AU.setPreservesCFG();
AU.addRequired<AAResultsWrapperPass>();
AU.addPreserved<GlobalsAAWrapperPass>();
AU.addPreserved<MemoryDependenceWrapperPass>();
}
// Helper routines
///
/// \brief Remove instruction from parent and update memory dependence
/// analysis.
@ -145,30 +134,8 @@ private:
const Instruction &End, MemoryLocation Loc);
bool sinkStore(BasicBlock *BB, StoreInst *SinkCand, StoreInst *ElseInst);
bool mergeStores(BasicBlock *BB);
// The mergeLoad/Store algorithms could have Size0 * Size1 complexity,
// where Size0 and Size1 are the #instructions on the two sides of
// the diamond. The constant chosen here is arbitrary. Compiler Time
// Control is enforced by the check Size0 * Size1 < MagicCompileTimeControl.
const int MagicCompileTimeControl;
};
char MergedLoadStoreMotion::ID = 0;
} // anonymous namespace
///
/// \brief createMergedLoadStoreMotionPass - The public interface to this file.
///
FunctionPass *llvm::createMergedLoadStoreMotionPass() {
return new MergedLoadStoreMotion();
}
INITIALIZE_PASS_BEGIN(MergedLoadStoreMotion, "mldst-motion",
"MergedLoadStoreMotion", false, false)
INITIALIZE_PASS_DEPENDENCY(MemoryDependenceWrapperPass)
INITIALIZE_PASS_DEPENDENCY(AAResultsWrapperPass)
INITIALIZE_PASS_END(MergedLoadStoreMotion, "mldst-motion",
"MergedLoadStoreMotion", false, false)
///
/// \brief Remove instruction from parent and update memory dependence analysis.
///
@ -552,16 +519,10 @@ bool MergedLoadStoreMotion::mergeStores(BasicBlock *T) {
return MergedStores;
}
///
/// \brief Run the transformation for each function
///
bool MergedLoadStoreMotion::runOnFunction(Function &F) {
if (skipFunction(F))
return false;
auto *MDWP = getAnalysisIfAvailable<MemoryDependenceWrapperPass>();
MD = MDWP ? &MDWP->getMemDep() : nullptr;
AA = &getAnalysis<AAResultsWrapperPass>().getAAResults();
bool MergedLoadStoreMotion::run(Function &F, MemoryDependenceResults *MD,
AliasAnalysis &AA) {
this->MD = MD;
this->AA = &AA;
bool Changed = false;
DEBUG(dbgs() << "Instruction Merger\n");
@ -580,3 +541,67 @@ bool MergedLoadStoreMotion::runOnFunction(Function &F) {
}
return Changed;
}
namespace {
class MergedLoadStoreMotionLegacyPass : public FunctionPass {
public:
static char ID; // Pass identification, replacement for typeid
MergedLoadStoreMotionLegacyPass() : FunctionPass(ID) {
initializeMergedLoadStoreMotionLegacyPassPass(
*PassRegistry::getPassRegistry());
}
///
/// \brief Run the transformation for each function
///
bool runOnFunction(Function &F) override {
if (skipFunction(F))
return false;
MergedLoadStoreMotion Impl;
auto *MDWP = getAnalysisIfAvailable<MemoryDependenceWrapperPass>();
return Impl.run(F, MDWP ? &MDWP->getMemDep() : nullptr,
getAnalysis<AAResultsWrapperPass>().getAAResults());
}
private:
// This transformation requires dominator postdominator info
void getAnalysisUsage(AnalysisUsage &AU) const override {
AU.setPreservesCFG();
AU.addRequired<AAResultsWrapperPass>();
AU.addPreserved<GlobalsAAWrapperPass>();
AU.addPreserved<MemoryDependenceWrapperPass>();
}
};
char MergedLoadStoreMotionLegacyPass::ID = 0;
} // anonymous namespace
///
/// \brief createMergedLoadStoreMotionPass - The public interface to this file.
///
FunctionPass *llvm::createMergedLoadStoreMotionPass() {
return new MergedLoadStoreMotionLegacyPass();
}
INITIALIZE_PASS_BEGIN(MergedLoadStoreMotionLegacyPass, "mldst-motion",
"MergedLoadStoreMotion", false, false)
INITIALIZE_PASS_DEPENDENCY(MemoryDependenceWrapperPass)
INITIALIZE_PASS_DEPENDENCY(AAResultsWrapperPass)
INITIALIZE_PASS_END(MergedLoadStoreMotionLegacyPass, "mldst-motion",
"MergedLoadStoreMotion", false, false)
PreservedAnalyses
MergedLoadStoreMotionPass::run(Function &F, AnalysisManager<Function> &AM) {
MergedLoadStoreMotion Impl;
auto *MD = AM.getCachedResult<MemoryDependenceAnalysis>(F);
auto &AA = AM.getResult<AAManager>(F);
if (!Impl.run(F, MD, AA))
return PreservedAnalyses::all();
// FIXME: This pass should also 'preserve the CFG'.
// The new pass manager has currently no way to do it.
PreservedAnalyses PA;
PA.preserve<GlobalsAA>();
PA.preserve<MemoryDependenceAnalysis>();
return PA;
}

View File

@ -65,7 +65,7 @@ void llvm::initializeScalarOpts(PassRegistry &Registry) {
initializeLowerExpectIntrinsicPass(Registry);
initializeLowerGuardIntrinsicPass(Registry);
initializeMemCpyOptLegacyPassPass(Registry);
initializeMergedLoadStoreMotionPass(Registry);
initializeMergedLoadStoreMotionLegacyPassPass(Registry);
initializeNaryReassociatePass(Registry);
initializePartiallyInlineLibCallsLegacyPassPass(Registry);
initializeReassociateLegacyPassPass(Registry);

View File

@ -1,4 +1,7 @@
; RUN: opt -basicaa -memdep -mldst-motion -S < %s | FileCheck %s
; RUN: opt -aa-pipeline=basic-aa -passes='require<memdep>',mldst-motion \
; RUN: -S < %s | FileCheck %s
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-linux-gnu"