/build/source/llvm/lib/CodeGen/CodeGenPrepare.cpp

Bug Summary

File:	build/source/llvm/lib/CodeGen/CodeGenPrepare.cpp
Warning:	line 2468, column 10 Called C++ object pointer is null
Annotated Source Code

Press '?' to see keyboard shortcuts
Show analyzer invocation
clang -cc1 -cc1 -triple x86_64-pc-linux-gnu -analyze -disable-free -clear-ast-before-backend -disable-llvm-verifier -discard-value-names -main-file-name CodeGenPrepare.cpp -analyzer-checker=core -analyzer-checker=apiModeling -analyzer-checker=unix -analyzer-checker=deadcode -analyzer-checker=cplusplus -analyzer-checker=security.insecureAPI.UncheckedReturn -analyzer-checker=security.insecureAPI.getpw -analyzer-checker=security.insecureAPI.gets -analyzer-checker=security.insecureAPI.mktemp -analyzer-checker=security.insecureAPI.mkstemp -analyzer-checker=security.insecureAPI.vfork -analyzer-checker=nullability.NullPassedToNonnull -analyzer-checker=nullability.NullReturnedFromNonnull -analyzer-output plist -w -setup-static-analyzer -analyzer-config-compatibility-mode=true -mrelocation-model pic -pic-level 2 -mframe-pointer=none -fmath-errno -ffp-contract=on -fno-rounding-math -mconstructor-aliases -funwind-tables=2 -target-cpu x86-64 -tune-cpu generic -debugger-tuning=gdb -ffunction-sections -fdata-sections -fcoverage-compilation-dir=/build/source/build-llvm -resource-dir /usr/lib/llvm-17/lib/clang/17 -D _DEBUG -D _GLIBCXX_ASSERTIONS -D _GNU_SOURCE -D _LIBCPP_ENABLE_ASSERTIONS -D __STDC_CONSTANT_MACROS -D __STDC_FORMAT_MACROS -D __STDC_LIMIT_MACROS -I lib/CodeGen -I /build/source/llvm/lib/CodeGen -I include -I /build/source/llvm/include -D _FORTIFY_SOURCE=2 -D NDEBUG -U NDEBUG -internal-isystem /usr/lib/gcc/x86_64-linux-gnu/10/../../../../include/c++/10 -internal-isystem /usr/lib/gcc/x86_64-linux-gnu/10/../../../../include/x86_64-linux-gnu/c++/10 -internal-isystem /usr/lib/gcc/x86_64-linux-gnu/10/../../../../include/c++/10/backward -internal-isystem /usr/lib/llvm-17/lib/clang/17/include -internal-isystem /usr/local/include -internal-isystem /usr/lib/gcc/x86_64-linux-gnu/10/../../../../x86_64-linux-gnu/include -internal-externc-isystem /usr/include/x86_64-linux-gnu -internal-externc-isystem /include -internal-externc-isystem /usr/include -fmacro-prefix-map=/build/source/build-llvm=build-llvm -fmacro-prefix-map=/build/source/= -fcoverage-prefix-map=/build/source/build-llvm=build-llvm -fcoverage-prefix-map=/build/source/= -source-date-epoch 1679915782 -O3 -Wno-unused-command-line-argument -Wno-unused-parameter -Wwrite-strings -Wno-missing-field-initializers -Wno-long-long -Wno-maybe-uninitialized -Wno-class-memaccess -Wno-redundant-move -Wno-pessimizing-move -Wno-noexcept-type -Wno-comment -Wno-misleading-indentation -std=c++17 -fdeprecated-macro -fdebug-compilation-dir=/build/source/build-llvm -fdebug-prefix-map=/build/source/build-llvm=build-llvm -fdebug-prefix-map=/build/source/= -fdebug-prefix-map=/build/source/build-llvm=build-llvm -fdebug-prefix-map=/build/source/= -ferror-limit 19 -fvisibility-inlines-hidden -stack-protector 2 -fgnuc-version=4.2.1 -fcolor-diagnostics -vectorize-loops -vectorize-slp -analyzer-output=html -analyzer-config stable-report-filename=true -faddrsig -D__GCC_HAVE_DWARF2_CFI_ASM=1 -o /tmp/scan-build-2023-03-27-130437-16335-1 -x c++ /build/source/llvm/lib/CodeGen/CodeGenPrepare.cpp
/build/source/llvm/lib/CodeGen/CodeGenPrepare.cpp

→
1//===- CodeGenPrepare.cpp - Prepare a function for code generation --------===//
2//
3// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4// See https://llvm.org/LICENSE.txt for license information.
5// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6//
7//===----------------------------------------------------------------------===//
8//
9// This pass munges the code in the input function to better prepare it for
10// SelectionDAG-based code generation. This works around limitations in it's
11// basic-block-at-a-time approach. It should eventually be removed.
12//
13//===----------------------------------------------------------------------===//

15#include "llvm/ADT/APInt.h"
16#include "llvm/ADT/ArrayRef.h"
17#include "llvm/ADT/DenseMap.h"
18#include "llvm/ADT/MapVector.h"
19#include "llvm/ADT/PointerIntPair.h"
20#include "llvm/ADT/STLExtras.h"
21#include "llvm/ADT/SmallPtrSet.h"
22#include "llvm/ADT/SmallVector.h"
23#include "llvm/ADT/Statistic.h"
24#include "llvm/Analysis/BlockFrequencyInfo.h"
25#include "llvm/Analysis/BranchProbabilityInfo.h"
26#include "llvm/Analysis/InstructionSimplify.h"
27#include "llvm/Analysis/LoopInfo.h"
28#include "llvm/Analysis/ProfileSummaryInfo.h"
29#include "llvm/Analysis/TargetLibraryInfo.h"
30#include "llvm/Analysis/TargetTransformInfo.h"
31#include "llvm/Analysis/ValueTracking.h"
32#include "llvm/Analysis/VectorUtils.h"
33#include "llvm/CodeGen/Analysis.h"
34#include "llvm/CodeGen/BasicBlockSectionsProfileReader.h"
35#include "llvm/CodeGen/ISDOpcodes.h"
36#include "llvm/CodeGen/SelectionDAGNodes.h"
37#include "llvm/CodeGen/TargetLowering.h"
38#include "llvm/CodeGen/TargetPassConfig.h"
39#include "llvm/CodeGen/TargetSubtargetInfo.h"
40#include "llvm/CodeGen/ValueTypes.h"
41#include "llvm/Config/llvm-config.h"
42#include "llvm/IR/Argument.h"
43#include "llvm/IR/Attributes.h"
44#include "llvm/IR/BasicBlock.h"
45#include "llvm/IR/Constant.h"
46#include "llvm/IR/Constants.h"
47#include "llvm/IR/DataLayout.h"
48#include "llvm/IR/DebugInfo.h"
49#include "llvm/IR/DerivedTypes.h"
50#include "llvm/IR/Dominators.h"
51#include "llvm/IR/Function.h"
52#include "llvm/IR/GetElementPtrTypeIterator.h"
53#include "llvm/IR/GlobalValue.h"
54#include "llvm/IR/GlobalVariable.h"
55#include "llvm/IR/IRBuilder.h"
56#include "llvm/IR/InlineAsm.h"
57#include "llvm/IR/InstrTypes.h"
58#include "llvm/IR/Instruction.h"
59#include "llvm/IR/Instructions.h"
60#include "llvm/IR/IntrinsicInst.h"
61#include "llvm/IR/Intrinsics.h"
62#include "llvm/IR/IntrinsicsAArch64.h"
63#include "llvm/IR/LLVMContext.h"
64#include "llvm/IR/MDBuilder.h"
65#include "llvm/IR/Module.h"
66#include "llvm/IR/Operator.h"
67#include "llvm/IR/PatternMatch.h"
68#include "llvm/IR/ProfDataUtils.h"
69#include "llvm/IR/Statepoint.h"
70#include "llvm/IR/Type.h"
71#include "llvm/IR/Use.h"
72#include "llvm/IR/User.h"
73#include "llvm/IR/Value.h"
74#include "llvm/IR/ValueHandle.h"
75#include "llvm/IR/ValueMap.h"
76#include "llvm/InitializePasses.h"
77#include "llvm/Pass.h"
78#include "llvm/Support/BlockFrequency.h"
79#include "llvm/Support/BranchProbability.h"
80#include "llvm/Support/Casting.h"
81#include "llvm/Support/CommandLine.h"
82#include "llvm/Support/Compiler.h"
83#include "llvm/Support/Debug.h"
84#include "llvm/Support/ErrorHandling.h"
85#include "llvm/Support/MachineValueType.h"
86#include "llvm/Support/MathExtras.h"
87#include "llvm/Support/raw_ostream.h"
88#include "llvm/Target/TargetMachine.h"
89#include "llvm/Target/TargetOptions.h"
90#include "llvm/Transforms/Utils/BasicBlockUtils.h"
91#include "llvm/Transforms/Utils/BypassSlowDivision.h"
92#include "llvm/Transforms/Utils/Local.h"
93#include "llvm/Transforms/Utils/SimplifyLibCalls.h"
94#include "llvm/Transforms/Utils/SizeOpts.h"
95#include <algorithm>
96#include <cassert>
97#include <cstdint>
98#include <iterator>
99#include <limits>
100#include <memory>
101#include <optional>
102#include <utility>
103#include <vector>

105using namespace llvm;
106using namespace llvm::PatternMatch;

108#define DEBUG_TYPE"codegenprepare" "codegenprepare"

110STATISTIC(NumBlocksElim, "Number of blocks eliminated")static llvm::Statistic NumBlocksElim = {"codegenprepare", "NumBlocksElim"
, "Number of blocks eliminated"};
111STATISTIC(NumPHIsElim, "Number of trivial PHIs eliminated")static llvm::Statistic NumPHIsElim = {"codegenprepare", "NumPHIsElim"
, "Number of trivial PHIs eliminated"};
112STATISTIC(NumGEPsElim, "Number of GEPs converted to casts")static llvm::Statistic NumGEPsElim = {"codegenprepare", "NumGEPsElim"
, "Number of GEPs converted to casts"};
113STATISTIC(NumCmpUses, "Number of uses of Cmp expressions replaced with uses of "static llvm::Statistic NumCmpUses = {"codegenprepare", "NumCmpUses"
, "Number of uses of Cmp expressions replaced with uses of " "sunken Cmps"
}
                    "sunken Cmps")static llvm::Statistic NumCmpUses = {"codegenprepare", "NumCmpUses"
, "Number of uses of Cmp expressions replaced with uses of " "sunken Cmps"
};
115STATISTIC(NumCastUses, "Number of uses of Cast expressions replaced with uses "static llvm::Statistic NumCastUses = {"codegenprepare", "NumCastUses"
, "Number of uses of Cast expressions replaced with uses " "of sunken Casts"
}
                     "of sunken Casts")static llvm::Statistic NumCastUses = {"codegenprepare", "NumCastUses"
, "Number of uses of Cast expressions replaced with uses " "of sunken Casts"
};
117STATISTIC(NumMemoryInsts, "Number of memory instructions whose address "static llvm::Statistic NumMemoryInsts = {"codegenprepare", "NumMemoryInsts"
, "Number of memory instructions whose address " "computations were sunk"
}
                        "computations were sunk")static llvm::Statistic NumMemoryInsts = {"codegenprepare", "NumMemoryInsts"
, "Number of memory instructions whose address " "computations were sunk"
};
119STATISTIC(NumMemoryInstsPhiCreated,static llvm::Statistic NumMemoryInstsPhiCreated = {"codegenprepare"
, "NumMemoryInstsPhiCreated", "Number of phis created when address "
 "computations were sunk to memory instructions"}
        "Number of phis created when address "static llvm::Statistic NumMemoryInstsPhiCreated = {"codegenprepare"
, "NumMemoryInstsPhiCreated", "Number of phis created when address "
 "computations were sunk to memory instructions"}
        "computations were sunk to memory instructions")static llvm::Statistic NumMemoryInstsPhiCreated = {"codegenprepare"
, "NumMemoryInstsPhiCreated", "Number of phis created when address "
 "computations were sunk to memory instructions"};
122STATISTIC(NumMemoryInstsSelectCreated,static llvm::Statistic NumMemoryInstsSelectCreated = {"codegenprepare"
, "NumMemoryInstsSelectCreated", "Number of select created when address "
 "computations were sunk to memory instructions"}
        "Number of select created when address "static llvm::Statistic NumMemoryInstsSelectCreated = {"codegenprepare"
, "NumMemoryInstsSelectCreated", "Number of select created when address "
 "computations were sunk to memory instructions"}
        "computations were sunk to memory instructions")static llvm::Statistic NumMemoryInstsSelectCreated = {"codegenprepare"
, "NumMemoryInstsSelectCreated", "Number of select created when address "
 "computations were sunk to memory instructions"};
125STATISTIC(NumExtsMoved, "Number of [s|z]ext instructions combined with loads")static llvm::Statistic NumExtsMoved = {"codegenprepare", "NumExtsMoved"
, "Number of [s|z]ext instructions combined with loads"};
126STATISTIC(NumExtUses, "Number of uses of [s|z]ext instructions optimized")static llvm::Statistic NumExtUses = {"codegenprepare", "NumExtUses"
, "Number of uses of [s|z]ext instructions optimized"};
127STATISTIC(NumAndsAdded,static llvm::Statistic NumAndsAdded = {"codegenprepare", "NumAndsAdded"
, "Number of and mask instructions added to form ext loads"}
        "Number of and mask instructions added to form ext loads")static llvm::Statistic NumAndsAdded = {"codegenprepare", "NumAndsAdded"
, "Number of and mask instructions added to form ext loads"};
129STATISTIC(NumAndUses, "Number of uses of and mask instructions optimized")static llvm::Statistic NumAndUses = {"codegenprepare", "NumAndUses"
, "Number of uses of and mask instructions optimized"};
130STATISTIC(NumRetsDup, "Number of return instructions duplicated")static llvm::Statistic NumRetsDup = {"codegenprepare", "NumRetsDup"
, "Number of return instructions duplicated"};
131STATISTIC(NumDbgValueMoved, "Number of debug value instructions moved")static llvm::Statistic NumDbgValueMoved = {"codegenprepare", "NumDbgValueMoved"
, "Number of debug value instructions moved"};
132STATISTIC(NumSelectsExpanded, "Number of selects turned into branches")static llvm::Statistic NumSelectsExpanded = {"codegenprepare"
, "NumSelectsExpanded", "Number of selects turned into branches"
};
133STATISTIC(NumStoreExtractExposed, "Number of store(extractelement) exposed")static llvm::Statistic NumStoreExtractExposed = {"codegenprepare"
, "NumStoreExtractExposed", "Number of store(extractelement) exposed"
};

135static cl::opt<bool> DisableBranchOpts(
  "disable-cgp-branch-opts", cl::Hidden, cl::init(false),
  cl::desc("Disable branch optimizations in CodeGenPrepare"));

139static cl::opt<bool>
  DisableGCOpts("disable-cgp-gc-opts", cl::Hidden, cl::init(false),
                cl::desc("Disable GC optimizations in CodeGenPrepare"));

143static cl::opt<bool>
  DisableSelectToBranch("disable-cgp-select2branch", cl::Hidden,
                        cl::init(false),
                        cl::desc("Disable select to branch conversion."));

148static cl::opt<bool>
  AddrSinkUsingGEPs("addr-sink-using-gep", cl::Hidden, cl::init(true),
                    cl::desc("Address sinking in CGP using GEPs."));

152static cl::opt<bool>
  EnableAndCmpSinking("enable-andcmp-sinking", cl::Hidden, cl::init(true),
                      cl::desc("Enable sinkinig and/cmp into branches."));

156static cl::opt<bool> DisableStoreExtract(
  "disable-cgp-store-extract", cl::Hidden, cl::init(false),
  cl::desc("Disable store(extract) optimizations in CodeGenPrepare"));

160static cl::opt<bool> StressStoreExtract(
  "stress-cgp-store-extract", cl::Hidden, cl::init(false),
  cl::desc("Stress test store(extract) optimizations in CodeGenPrepare"));

164static cl::opt<bool> DisableExtLdPromotion(
  "disable-cgp-ext-ld-promotion", cl::Hidden, cl::init(false),
  cl::desc("Disable ext(promotable(ld)) -> promoted(ext(ld)) optimization in "
           "CodeGenPrepare"));

169static cl::opt<bool> StressExtLdPromotion(
  "stress-cgp-ext-ld-promotion", cl::Hidden, cl::init(false),
  cl::desc("Stress test ext(promotable(ld)) -> promoted(ext(ld)) "
           "optimization in CodeGenPrepare"));

174static cl::opt<bool> DisablePreheaderProtect(
  "disable-preheader-prot", cl::Hidden, cl::init(false),
  cl::desc("Disable protection against removing loop preheaders"));

178static cl::opt<bool> ProfileGuidedSectionPrefix(
  "profile-guided-section-prefix", cl::Hidden, cl::init(true),
  cl::desc("Use profile info to add section prefix for hot/cold functions"));

182static cl::opt<bool> ProfileUnknownInSpecialSection(
  "profile-unknown-in-special-section", cl::Hidden,
  cl::desc("In profiling mode like sampleFDO, if a function doesn't have "
           "profile, we cannot tell the function is cold for sure because "
           "it may be a function newly added without ever being sampled. "
           "With the flag enabled, compiler can put such profile unknown "
           "functions into a special section, so runtime system can choose "
           "to handle it in a different way than .text section, to save "
           "RAM for example. "));

192static cl::opt<bool> BBSectionsGuidedSectionPrefix(
  "bbsections-guided-section-prefix", cl::Hidden, cl::init(true),
  cl::desc("Use the basic-block-sections profile to determine the text "
           "section prefix for hot functions. Functions with "
           "basic-block-sections profile will be placed in `.text.hot` "
           "regardless of their FDO profile info. Other functions won't be "
           "impacted, i.e., their prefixes will be decided by FDO/sampleFDO "
           "profiles."));

201static cl::opt<unsigned> FreqRatioToSkipMerge(
  "cgp-freq-ratio-to-skip-merge", cl::Hidden, cl::init(2),
  cl::desc("Skip merging empty blocks if (frequency of empty block) / "
           "(frequency of destination block) is greater than this ratio"));

206static cl::opt<bool> ForceSplitStore(
  "force-split-store", cl::Hidden, cl::init(false),
  cl::desc("Force store splitting no matter what the target query says."));

210static cl::opt<bool> EnableTypePromotionMerge(
  "cgp-type-promotion-merge", cl::Hidden,
  cl::desc("Enable merging of redundant sexts when one is dominating"
           " the other."),
  cl::init(true));

216static cl::opt<bool> DisableComplexAddrModes(
  "disable-complex-addr-modes", cl::Hidden, cl::init(false),
  cl::desc("Disables combining addressing modes with different parts "
           "in optimizeMemoryInst."));

221static cl::opt<bool>
  AddrSinkNewPhis("addr-sink-new-phis", cl::Hidden, cl::init(false),
                  cl::desc("Allow creation of Phis in Address sinking."));

225static cl::opt<bool> AddrSinkNewSelects(
  "addr-sink-new-select", cl::Hidden, cl::init(true),
  cl::desc("Allow creation of selects in Address sinking."));

229static cl::opt<bool> AddrSinkCombineBaseReg(
  "addr-sink-combine-base-reg", cl::Hidden, cl::init(true),
  cl::desc("Allow combining of BaseReg field in Address sinking."));

233static cl::opt<bool> AddrSinkCombineBaseGV(
  "addr-sink-combine-base-gv", cl::Hidden, cl::init(true),
  cl::desc("Allow combining of BaseGV field in Address sinking."));

237static cl::opt<bool> AddrSinkCombineBaseOffs(
  "addr-sink-combine-base-offs", cl::Hidden, cl::init(true),
  cl::desc("Allow combining of BaseOffs field in Address sinking."));

241static cl::opt<bool> AddrSinkCombineScaledReg(
  "addr-sink-combine-scaled-reg", cl::Hidden, cl::init(true),
  cl::desc("Allow combining of ScaledReg field in Address sinking."));

245static cl::opt<bool>
  EnableGEPOffsetSplit("cgp-split-large-offset-gep", cl::Hidden,
                       cl::init(true),
                       cl::desc("Enable splitting large offset of GEP."));

250static cl::opt<bool> EnableICMP_EQToICMP_ST(
  "cgp-icmp-eq2icmp-st", cl::Hidden, cl::init(false),
  cl::desc("Enable ICMP_EQ to ICMP_S(L|G)T conversion."));

254static cl::opt<bool>
  VerifyBFIUpdates("cgp-verify-bfi-updates", cl::Hidden, cl::init(false),
                   cl::desc("Enable BFI update verification for "
                            "CodeGenPrepare."));

259static cl::opt<bool>
  OptimizePhiTypes("cgp-optimize-phi-types", cl::Hidden, cl::init(false),
                   cl::desc("Enable converting phi types in CodeGenPrepare"));

263static cl::opt<unsigned>
  HugeFuncThresholdInCGPP("cgpp-huge-func", cl::init(10000), cl::Hidden,
                          cl::desc("Least BB number of huge function."));

267namespace {

269enum ExtType {
ZeroExtension, // Zero extension has been seen.
SignExtension, // Sign extension has been seen.
BothExtension  // This extension type is used if we saw sext after
               // ZeroExtension had been set, or if we saw zext after
               // SignExtension had been set. It makes the type
               // information of a promoted instruction invalid.
276};

278enum ModifyDT {
NotModifyDT, // Not Modify any DT.
ModifyBBDT,  // Modify the Basic Block Dominator Tree.
ModifyInstDT // Modify the Instruction Dominator in a Basic Block,
             // This usually means we move/delete/insert instruction
             // in a Basic Block. So we should re-iterate instructions
             // in such Basic Block.
285};

287using SetOfInstrs = SmallPtrSet<Instruction *, 16>;
288using TypeIsSExt = PointerIntPair<Type *, 2, ExtType>;
289using InstrToOrigTy = DenseMap<Instruction *, TypeIsSExt>;
290using SExts = SmallVector<Instruction *, 16>;
291using ValueToSExts = MapVector<Value *, SExts>;

293class TypePromotionTransaction;

295class CodeGenPrepare : public FunctionPass {
const TargetMachine *TM = nullptr;
const TargetSubtargetInfo *SubtargetInfo;
const TargetLowering *TLI = nullptr;
const TargetRegisterInfo *TRI;
const TargetTransformInfo *TTI = nullptr;
const BasicBlockSectionsProfileReader *BBSectionsProfileReader = nullptr;
const TargetLibraryInfo *TLInfo;
const LoopInfo *LI;
std::unique_ptr<BlockFrequencyInfo> BFI;
std::unique_ptr<BranchProbabilityInfo> BPI;
ProfileSummaryInfo *PSI;

/// As we scan instructions optimizing them, this is the next instruction
/// to optimize. Transforms that can invalidate this should update it.
BasicBlock::iterator CurInstIterator;

/// Keeps track of non-local addresses that have been sunk into a block.
/// This allows us to avoid inserting duplicate code for blocks with
/// multiple load/stores of the same address. The usage of WeakTrackingVH
/// enables SunkAddrs to be treated as a cache whose entries can be
/// invalidated if a sunken address computation has been erased.
ValueMap<Value *, WeakTrackingVH> SunkAddrs;

/// Keeps track of all instructions inserted for the current function.
SetOfInstrs InsertedInsts;

/// Keeps track of the type of the related instruction before their
/// promotion for the current function.
InstrToOrigTy PromotedInsts;

/// Keep track of instructions removed during promotion.
SetOfInstrs RemovedInsts;

/// Keep track of sext chains based on their initial value.
DenseMap<Value *, Instruction *> SeenChainsForSExt;

/// Keep track of GEPs accessing the same data structures such as structs or
/// arrays that are candidates to be split later because of their large
/// size.
MapVector<AssertingVH<Value>,
          SmallVector<std::pair<AssertingVH<GetElementPtrInst>, int64_t>, 32>>
    LargeOffsetGEPMap;

/// Keep track of new GEP base after splitting the GEPs having large offset.
SmallSet<AssertingVH<Value>, 2> NewGEPBases;

/// Map serial numbers to Large offset GEPs.
DenseMap<AssertingVH<GetElementPtrInst>, int> LargeOffsetGEPID;

/// Keep track of SExt promoted.
ValueToSExts ValToSExtendedUses;

/// True if the function has the OptSize attribute.
bool OptSize;

/// DataLayout for the Function being processed.
const DataLayout *DL = nullptr;

/// Building the dominator tree can be expensive, so we only build it
/// lazily and update it when required.
std::unique_ptr<DominatorTree> DT;

358public:
/// If encounter huge function, we need to limit the build time.
bool IsHugeFunc = false;

/// FreshBBs is like worklist, it collected the updated BBs which need
/// to be optimized again.
/// Note: Consider building time in this pass, when a BB updated, we need
/// to insert such BB into FreshBBs for huge function.
SmallSet<BasicBlock *, 32> FreshBBs;

static char ID; // Pass identification, replacement for typeid

CodeGenPrepare() : FunctionPass(ID) {
  initializeCodeGenPreparePass(*PassRegistry::getPassRegistry());
}

bool runOnFunction(Function &F) override;

StringRef getPassName() const override { return "CodeGen Prepare"; }

void getAnalysisUsage(AnalysisUsage &AU) const override {
  // FIXME: When we can selectively preserve passes, preserve the domtree.
  AU.addRequired<ProfileSummaryInfoWrapperPass>();
  AU.addRequired<TargetLibraryInfoWrapperPass>();
  AU.addRequired<TargetPassConfig>();
  AU.addRequired<TargetTransformInfoWrapperPass>();
  AU.addRequired<LoopInfoWrapperPass>();
  AU.addUsedIfAvailable<BasicBlockSectionsProfileReader>();
}

388private:
template <typename F>
void resetIteratorIfInvalidatedWhileCalling(BasicBlock *BB, F f) {
  // Substituting can cause recursive simplifications, which can invalidate
  // our iterator.  Use a WeakTrackingVH to hold onto it in case this
  // happens.
  Value *CurValue = &*CurInstIterator;
  WeakTrackingVH IterHandle(CurValue);

  f();

  // If the iterator instruction was recursively deleted, start over at the
  // start of the block.
  if (IterHandle != CurValue) {
    CurInstIterator = BB->begin();
    SunkAddrs.clear();
  }
}

// Get the DominatorTree, building if necessary.
DominatorTree &getDT(Function &F) {
  if (!DT)
    DT = std::make_unique<DominatorTree>(F);
  return *DT;
}

void removeAllAssertingVHReferences(Value *V);
bool eliminateAssumptions(Function &F);
bool eliminateFallThrough(Function &F);
bool eliminateMostlyEmptyBlocks(Function &F);
BasicBlock *findDestBlockOfMergeableEmptyBlock(BasicBlock *BB);
bool canMergeBlocks(const BasicBlock *BB, const BasicBlock *DestBB) const;
void eliminateMostlyEmptyBlock(BasicBlock *BB);
bool isMergingEmptyBlockProfitable(BasicBlock *BB, BasicBlock *DestBB,
                                   bool isPreheader);
bool makeBitReverse(Instruction &I);
bool optimizeBlock(BasicBlock &BB, ModifyDT &ModifiedDT);
bool optimizeInst(Instruction *I, ModifyDT &ModifiedDT);
bool optimizeMemoryInst(Instruction *MemoryInst, Value *Addr, Type *AccessTy,
                        unsigned AddrSpace);
bool optimizeGatherScatterInst(Instruction *MemoryInst, Value *Ptr);
bool optimizeInlineAsmInst(CallInst *CS);
bool optimizeCallInst(CallInst *CI, ModifyDT &ModifiedDT);
bool optimizeExt(Instruction *&I);
bool optimizeExtUses(Instruction *I);
bool optimizeLoadExt(LoadInst *Load);
bool optimizeShiftInst(BinaryOperator *BO);
bool optimizeFunnelShift(IntrinsicInst *Fsh);
bool optimizeSelectInst(SelectInst *SI);
bool optimizeShuffleVectorInst(ShuffleVectorInst *SVI);
bool optimizeSwitchType(SwitchInst *SI);
bool optimizeSwitchPhiConstants(SwitchInst *SI);
bool optimizeSwitchInst(SwitchInst *SI);
bool optimizeExtractElementInst(Instruction *Inst);
bool dupRetToEnableTailCallOpts(BasicBlock *BB, ModifyDT &ModifiedDT);
bool fixupDbgValue(Instruction *I);
bool placeDbgValues(Function &F);
bool placePseudoProbes(Function &F);
bool canFormExtLd(const SmallVectorImpl<Instruction *> &MovedExts,
                  LoadInst *&LI, Instruction *&Inst, bool HasPromoted);
bool tryToPromoteExts(TypePromotionTransaction &TPT,
                      const SmallVectorImpl<Instruction *> &Exts,
                      SmallVectorImpl<Instruction *> &ProfitablyMovedExts,
                      unsigned CreatedInstsCost = 0);
bool mergeSExts(Function &F);
bool splitLargeGEPOffsets();
bool optimizePhiType(PHINode *Inst, SmallPtrSetImpl<PHINode *> &Visited,
                     SmallPtrSetImpl<Instruction *> &DeletedInstrs);
bool optimizePhiTypes(Function &F);
bool performAddressTypePromotion(
    Instruction *&Inst, bool AllowPromotionWithoutCommonHeader,
    bool HasPromoted, TypePromotionTransaction &TPT,
    SmallVectorImpl<Instruction *> &SpeculativelyMovedExts);
bool splitBranchCondition(Function &F, ModifyDT &ModifiedDT);
bool simplifyOffsetableRelocate(GCStatepointInst &I);

bool tryToSinkFreeOperands(Instruction *I);
bool replaceMathCmpWithIntrinsic(BinaryOperator *BO, Value *Arg0, Value *Arg1,
                                 CmpInst *Cmp, Intrinsic::ID IID);
bool optimizeCmp(CmpInst *Cmp, ModifyDT &ModifiedDT);
bool combineToUSubWithOverflow(CmpInst *Cmp, ModifyDT &ModifiedDT);
bool combineToUAddWithOverflow(CmpInst *Cmp, ModifyDT &ModifiedDT);
void verifyBFIUpdates(Function &F);
471};

473} // end anonymous namespace

475char CodeGenPrepare::ID = 0;

477INITIALIZE_PASS_BEGIN(CodeGenPrepare, DEBUG_TYPE,static void *initializeCodeGenPreparePassOnce(PassRegistry &
Registry) {
                    "Optimize for code generation", false, false)static void *initializeCodeGenPreparePassOnce(PassRegistry &
Registry) {
479INITIALIZE_PASS_DEPENDENCY(BasicBlockSectionsProfileReader)initializeBasicBlockSectionsProfileReaderPass(Registry);
480INITIALIZE_PASS_DEPENDENCY(LoopInfoWrapperPass)initializeLoopInfoWrapperPassPass(Registry);
481INITIALIZE_PASS_DEPENDENCY(ProfileSummaryInfoWrapperPass)initializeProfileSummaryInfoWrapperPassPass(Registry);
482INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)initializeTargetLibraryInfoWrapperPassPass(Registry);
483INITIALIZE_PASS_DEPENDENCY(TargetPassConfig)initializeTargetPassConfigPass(Registry);
484INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)initializeTargetTransformInfoWrapperPassPass(Registry);
485INITIALIZE_PASS_END(CodeGenPrepare, DEBUG_TYPE, "Optimize for code generation",PassInfo *PI = new PassInfo( "Optimize for code generation", "codegenprepare"
, &CodeGenPrepare::ID, PassInfo::NormalCtor_t(callDefaultCtor
<CodeGenPrepare>), false, false); Registry.registerPass
(*PI, true); return PI; } static llvm::once_flag InitializeCodeGenPreparePassFlag
; void llvm::initializeCodeGenPreparePass(PassRegistry &Registry
) { llvm::call_once(InitializeCodeGenPreparePassFlag, initializeCodeGenPreparePassOnce
, std::ref(Registry)); }
                  false, false)PassInfo *PI = new PassInfo( "Optimize for code generation", "codegenprepare"
, &CodeGenPrepare::ID, PassInfo::NormalCtor_t(callDefaultCtor
<CodeGenPrepare>), false, false); Registry.registerPass
(*PI, true); return PI; } static llvm::once_flag InitializeCodeGenPreparePassFlag
; void llvm::initializeCodeGenPreparePass(PassRegistry &Registry
) { llvm::call_once(InitializeCodeGenPreparePassFlag, initializeCodeGenPreparePassOnce
, std::ref(Registry)); }

488FunctionPass *llvm::createCodeGenPreparePass() { return new CodeGenPrepare(); }

490bool CodeGenPrepare::runOnFunction(Function &F) {
if (skipFunction(F))
  return false;

DL = &F.getParent()->getDataLayout();

bool EverMadeChange = false;
// Clear per function information.
InsertedInsts.clear();
PromotedInsts.clear();
FreshBBs.clear();

TM = &getAnalysis<TargetPassConfig>().getTM<TargetMachine>();
SubtargetInfo = TM->getSubtargetImpl(F);
TLI = SubtargetInfo->getTargetLowering();
TRI = SubtargetInfo->getRegisterInfo();
TLInfo = &getAnalysis<TargetLibraryInfoWrapperPass>().getTLI(F);
TTI = &getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F);
LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo();
BPI.reset(new BranchProbabilityInfo(F, *LI));
BFI.reset(new BlockFrequencyInfo(F, *BPI, *LI));
PSI = &getAnalysis<ProfileSummaryInfoWrapperPass>().getPSI();
BBSectionsProfileReader =
    getAnalysisIfAvailable<BasicBlockSectionsProfileReader>();
OptSize = F.hasOptSize();
// Use the basic-block-sections profile to promote hot functions to .text.hot
// if requested.
if (BBSectionsGuidedSectionPrefix && BBSectionsProfileReader &&
    BBSectionsProfileReader->isFunctionHot(F.getName())) {
  F.setSectionPrefix("hot");
} else if (ProfileGuidedSectionPrefix) {
  // The hot attribute overwrites profile count based hotness while profile
  // counts based hotness overwrite the cold attribute.
  // This is a conservative behabvior.
  if (F.hasFnAttribute(Attribute::Hot) ||
      PSI->isFunctionHotInCallGraph(&F, *BFI))
    F.setSectionPrefix("hot");
  // If PSI shows this function is not hot, we will placed the function
  // into unlikely section if (1) PSI shows this is a cold function, or
  // (2) the function has a attribute of cold.
  else if (PSI->isFunctionColdInCallGraph(&F, *BFI) ||
           F.hasFnAttribute(Attribute::Cold))
    F.setSectionPrefix("unlikely");
  else if (ProfileUnknownInSpecialSection && PSI->hasPartialSampleProfile() &&
           PSI->isFunctionHotnessUnknown(F))
    F.setSectionPrefix("unknown");
}

/// This optimization identifies DIV instructions that can be
/// profitably bypassed and carried out with a shorter, faster divide.
if (!OptSize && !PSI->hasHugeWorkingSetSize() && TLI->isSlowDivBypassed()) {
  const DenseMap<unsigned int, unsigned int> &BypassWidths =
      TLI->getBypassSlowDivWidths();
  BasicBlock *BB = &*F.begin();
  while (BB != nullptr) {
    // bypassSlowDivision may create new BBs, but we don't want to reapply the
    // optimization to those blocks.
    BasicBlock *Next = BB->getNextNode();
    // F.hasOptSize is already checked in the outer if statement.
    if (!llvm::shouldOptimizeForSize(BB, PSI, BFI.get()))
      EverMadeChange |= bypassSlowDivision(BB, BypassWidths);
    BB = Next;
  }
}

// Get rid of @llvm.assume builtins before attempting to eliminate empty
// blocks, since there might be blocks that only contain @llvm.assume calls
// (plus arguments that we can get rid of).
EverMadeChange |= eliminateAssumptions(F);

// Eliminate blocks that contain only PHI nodes and an
// unconditional branch.
EverMadeChange |= eliminateMostlyEmptyBlocks(F);

ModifyDT ModifiedDT = ModifyDT::NotModifyDT;
if (!DisableBranchOpts)
  EverMadeChange |= splitBranchCondition(F, ModifiedDT);

// Split some critical edges where one of the sources is an indirect branch,
// to help generate sane code for PHIs involving such edges.
EverMadeChange |=
    SplitIndirectBrCriticalEdges(F, /*IgnoreBlocksWithoutPHI=*/true);

// If we are optimzing huge function, we need to consider the build time.
// Because the basic algorithm's complex is near O(N!).
IsHugeFunc = F.size() > HugeFuncThresholdInCGPP;

bool MadeChange = true;
bool FuncIterated = false;
while (MadeChange) {
  MadeChange = false;
  DT.reset();

  for (BasicBlock &BB : llvm::make_early_inc_range(F)) {
    if (FuncIterated && !FreshBBs.contains(&BB))
      continue;

    ModifyDT ModifiedDTOnIteration = ModifyDT::NotModifyDT;
    bool Changed = optimizeBlock(BB, ModifiedDTOnIteration);

    MadeChange |= Changed;
    if (IsHugeFunc) {
      // If the BB is updated, it may still has chance to be optimized.
      // This usually happen at sink optimization.
      // For example:
      //
      // bb0：
      // %and = and i32 %a, 4
      // %cmp = icmp eq i32 %and, 0
      //
      // If the %cmp sink to other BB, the %and will has chance to sink.
      if (Changed)
        FreshBBs.insert(&BB);
      else if (FuncIterated)
        FreshBBs.erase(&BB);

      if (ModifiedDTOnIteration == ModifyDT::ModifyBBDT)
        DT.reset();
    } else {
      // For small/normal functions, we restart BB iteration if the dominator
      // tree of the Function was changed.
      if (ModifiedDTOnIteration != ModifyDT::NotModifyDT)
        break;
    }
  }
  // We have iterated all the BB in the (only work for huge) function.
  FuncIterated = IsHugeFunc;

  if (EnableTypePromotionMerge && !ValToSExtendedUses.empty())
    MadeChange |= mergeSExts(F);
  if (!LargeOffsetGEPMap.empty())
    MadeChange |= splitLargeGEPOffsets();
  MadeChange |= optimizePhiTypes(F);

  if (MadeChange)
    eliminateFallThrough(F);

  // Really free removed instructions during promotion.
  for (Instruction *I : RemovedInsts)
    I->deleteValue();

  EverMadeChange |= MadeChange;
  SeenChainsForSExt.clear();
  ValToSExtendedUses.clear();
  RemovedInsts.clear();
  LargeOffsetGEPMap.clear();
  LargeOffsetGEPID.clear();
}

NewGEPBases.clear();
SunkAddrs.clear();

if (!DisableBranchOpts) {
  MadeChange = false;
  // Use a set vector to get deterministic iteration order. The order the
  // blocks are removed may affect whether or not PHI nodes in successors
  // are removed.
  SmallSetVector<BasicBlock *, 8> WorkList;
  for (BasicBlock &BB : F) {
    SmallVector<BasicBlock *, 2> Successors(successors(&BB));
    MadeChange |= ConstantFoldTerminator(&BB, true);
    if (!MadeChange)
      continue;

    for (BasicBlock *Succ : Successors)
      if (pred_empty(Succ))
        WorkList.insert(Succ);
  }

  // Delete the dead blocks and any of their dead successors.
  MadeChange |= !WorkList.empty();
  while (!WorkList.empty()) {
    BasicBlock *BB = WorkList.pop_back_val();
    SmallVector<BasicBlock *, 2> Successors(successors(BB));

    DeleteDeadBlock(BB);

    for (BasicBlock *Succ : Successors)
      if (pred_empty(Succ))
        WorkList.insert(Succ);
  }

  // Merge pairs of basic blocks with unconditional branches, connected by
  // a single edge.
  if (EverMadeChange || MadeChange)
    MadeChange |= eliminateFallThrough(F);

  EverMadeChange |= MadeChange;
}

if (!DisableGCOpts) {
  SmallVector<GCStatepointInst *, 2> Statepoints;
  for (BasicBlock &BB : F)
    for (Instruction &I : BB)
      if (auto *SP = dyn_cast<GCStatepointInst>(&I))
        Statepoints.push_back(SP);
  for (auto &I : Statepoints)
    EverMadeChange |= simplifyOffsetableRelocate(*I);
}

// Do this last to clean up use-before-def scenarios introduced by other
// preparatory transforms.
EverMadeChange |= placeDbgValues(F);
EverMadeChange |= placePseudoProbes(F);

695#ifndef NDEBUG
if (VerifyBFIUpdates)
  verifyBFIUpdates(F);
698#endif

return EverMadeChange;
701}

703bool CodeGenPrepare::eliminateAssumptions(Function &F) {
bool MadeChange = false;
for (BasicBlock &BB : F) {
  CurInstIterator = BB.begin();
  while (CurInstIterator != BB.end()) {
    Instruction *I = &*(CurInstIterator++);
    if (auto *Assume = dyn_cast<AssumeInst>(I)) {
      MadeChange = true;
      Value *Operand = Assume->getOperand(0);
      Assume->eraseFromParent();

      resetIteratorIfInvalidatedWhileCalling(&BB, [&]() {
        RecursivelyDeleteTriviallyDeadInstructions(Operand, TLInfo, nullptr);
      });
    }
  }
}
return MadeChange;
721}

723/// An instruction is about to be deleted, so remove all references to it in our
724/// GEP-tracking data strcutures.
725void CodeGenPrepare::removeAllAssertingVHReferences(Value *V) {
LargeOffsetGEPMap.erase(V);
NewGEPBases.erase(V);

auto GEP = dyn_cast<GetElementPtrInst>(V);
if (!GEP)
  return;

LargeOffsetGEPID.erase(GEP);

auto VecI = LargeOffsetGEPMap.find(GEP->getPointerOperand());
if (VecI == LargeOffsetGEPMap.end())
  return;

auto &GEPVector = VecI->second;
llvm::erase_if(GEPVector, [=](auto &Elt) { return Elt.first == GEP; });

if (GEPVector.empty())
  LargeOffsetGEPMap.erase(VecI);
744}

746// Verify BFI has been updated correctly by recomputing BFI and comparing them.
747void LLVM_ATTRIBUTE_UNUSED__attribute__((__unused__)) CodeGenPrepare::verifyBFIUpdates(Function &F) {
DominatorTree NewDT(F);
LoopInfo NewLI(NewDT);
BranchProbabilityInfo NewBPI(F, NewLI, TLInfo);
BlockFrequencyInfo NewBFI(F, NewBPI, NewLI);
NewBFI.verifyMatch(*BFI);
753}

755/// Merge basic blocks which are connected by a single edge, where one of the
756/// basic blocks has a single successor pointing to the other basic block,
757/// which has a single predecessor.
758bool CodeGenPrepare::eliminateFallThrough(Function &F) {
bool Changed = false;
// Scan all of the blocks in the function, except for the entry block.
// Use a temporary array to avoid iterator being invalidated when
// deleting blocks.
SmallVector<WeakTrackingVH, 16> Blocks;
for (auto &Block : llvm::drop_begin(F))
  Blocks.push_back(&Block);

SmallSet<WeakTrackingVH, 16> Preds;
for (auto &Block : Blocks) {
  auto *BB = cast_or_null<BasicBlock>(Block);
  if (!BB)
    continue;
  // If the destination block has a single pred, then this is a trivial
  // edge, just collapse it.
  BasicBlock *SinglePred = BB->getSinglePredecessor();

  // Don't merge if BB's address is taken.
  if (!SinglePred || SinglePred == BB || BB->hasAddressTaken())
    continue;

  BranchInst *Term = dyn_cast<BranchInst>(SinglePred->getTerminator());
  if (Term && !Term->isConditional()) {
    Changed = true;
    LLVM_DEBUG(dbgs() << "To merge:\n" << *BB << "\n\n\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "To merge:\n" << *
BB << "\n\n\n"; } } while (false);

    // Merge BB into SinglePred and delete it.
    MergeBlockIntoPredecessor(BB);
    Preds.insert(SinglePred);

    if (IsHugeFunc) {
      // Update FreshBBs to optimize the merged BB.
      FreshBBs.insert(SinglePred);
      FreshBBs.erase(BB);
    }
  }
}

// (Repeatedly) merging blocks into their predecessors can create redundant
// debug intrinsics.
for (const auto &Pred : Preds)
  if (auto *BB = cast_or_null<BasicBlock>(Pred))
    RemoveRedundantDbgInstrs(BB);

return Changed;
804}

806/// Find a destination block from BB if BB is mergeable empty block.
807BasicBlock *CodeGenPrepare::findDestBlockOfMergeableEmptyBlock(BasicBlock *BB) {
// If this block doesn't end with an uncond branch, ignore it.
BranchInst *BI = dyn_cast<BranchInst>(BB->getTerminator());
if (!BI || !BI->isUnconditional())
  return nullptr;

// If the instruction before the branch (skipping debug info) isn't a phi
// node, then other stuff is happening here.
BasicBlock::iterator BBI = BI->getIterator();
if (BBI != BB->begin()) {
  --BBI;
  while (isa<DbgInfoIntrinsic>(BBI)) {
    if (BBI == BB->begin())
      break;
    --BBI;
  }
  if (!isa<DbgInfoIntrinsic>(BBI) && !isa<PHINode>(BBI))
    return nullptr;
}

// Do not break infinite loops.
BasicBlock *DestBB = BI->getSuccessor(0);
if (DestBB == BB)
  return nullptr;

if (!canMergeBlocks(BB, DestBB))
  DestBB = nullptr;

return DestBB;
836}

838/// Eliminate blocks that contain only PHI nodes, debug info directives, and an
839/// unconditional branch. Passes before isel (e.g. LSR/loopsimplify) often split
840/// edges in ways that are non-optimal for isel. Start by eliminating these
841/// blocks so we can split them the way we want them.
842bool CodeGenPrepare::eliminateMostlyEmptyBlocks(Function &F) {
SmallPtrSet<BasicBlock *, 16> Preheaders;
SmallVector<Loop *, 16> LoopList(LI->begin(), LI->end());
while (!LoopList.empty()) {
  Loop *L = LoopList.pop_back_val();
  llvm::append_range(LoopList, *L);
  if (BasicBlock *Preheader = L->getLoopPreheader())
    Preheaders.insert(Preheader);
}

bool MadeChange = false;
// Copy blocks into a temporary array to avoid iterator invalidation issues
// as we remove them.
// Note that this intentionally skips the entry block.
SmallVector<WeakTrackingVH, 16> Blocks;
for (auto &Block : llvm::drop_begin(F))
  Blocks.push_back(&Block);

for (auto &Block : Blocks) {
  BasicBlock *BB = cast_or_null<BasicBlock>(Block);
  if (!BB)
    continue;
  BasicBlock *DestBB = findDestBlockOfMergeableEmptyBlock(BB);
  if (!DestBB ||
      !isMergingEmptyBlockProfitable(BB, DestBB, Preheaders.count(BB)))
    continue;

  eliminateMostlyEmptyBlock(BB);
  MadeChange = true;
}
return MadeChange;
873}

875bool CodeGenPrepare::isMergingEmptyBlockProfitable(BasicBlock *BB,
                                                 BasicBlock *DestBB,
                                                 bool isPreheader) {
// Do not delete loop preheaders if doing so would create a critical edge.
// Loop preheaders can be good locations to spill registers. If the
// preheader is deleted and we create a critical edge, registers may be
// spilled in the loop body instead.
if (!DisablePreheaderProtect && isPreheader &&
    !(BB->getSinglePredecessor() &&
      BB->getSinglePredecessor()->getSingleSuccessor()))
  return false;

// Skip merging if the block's successor is also a successor to any callbr
// that leads to this block.
// FIXME: Is this really needed? Is this a correctness issue?
for (BasicBlock *Pred : predecessors(BB)) {
  if (auto *CBI = dyn_cast<CallBrInst>((Pred)->getTerminator()))
    for (unsigned i = 0, e = CBI->getNumSuccessors(); i != e; ++i)
      if (DestBB == CBI->getSuccessor(i))
        return false;
}

// Try to skip merging if the unique predecessor of BB is terminated by a
// switch or indirect branch instruction, and BB is used as an incoming block
// of PHIs in DestBB. In such case, merging BB and DestBB would cause ISel to
// add COPY instructions in the predecessor of BB instead of BB (if it is not
// merged). Note that the critical edge created by merging such blocks wont be
// split in MachineSink because the jump table is not analyzable. By keeping
// such empty block (BB), ISel will place COPY instructions in BB, not in the
// predecessor of BB.
BasicBlock *Pred = BB->getUniquePredecessor();
if (!Pred || !(isa<SwitchInst>(Pred->getTerminator()) ||
               isa<IndirectBrInst>(Pred->getTerminator())))
  return true;

if (BB->getTerminator() != BB->getFirstNonPHIOrDbg())
  return true;

// We use a simple cost heuristic which determine skipping merging is
// profitable if the cost of skipping merging is less than the cost of
// merging : Cost(skipping merging) < Cost(merging BB), where the
// Cost(skipping merging) is Freq(BB) * (Cost(Copy) + Cost(Branch)), and
// the Cost(merging BB) is Freq(Pred) * Cost(Copy).
// Assuming Cost(Copy) == Cost(Branch), we could simplify it to :
//   Freq(Pred) / Freq(BB) > 2.
// Note that if there are multiple empty blocks sharing the same incoming
// value for the PHIs in the DestBB, we consider them together. In such
// case, Cost(merging BB) will be the sum of their frequencies.

if (!isa<PHINode>(DestBB->begin()))
  return true;

SmallPtrSet<BasicBlock *, 16> SameIncomingValueBBs;

// Find all other incoming blocks from which incoming values of all PHIs in
// DestBB are the same as the ones from BB.
for (BasicBlock *DestBBPred : predecessors(DestBB)) {
  if (DestBBPred == BB)
    continue;

  if (llvm::all_of(DestBB->phis(), [&](const PHINode &DestPN) {
        return DestPN.getIncomingValueForBlock(BB) ==
               DestPN.getIncomingValueForBlock(DestBBPred);
      }))
    SameIncomingValueBBs.insert(DestBBPred);
}

// See if all BB's incoming values are same as the value from Pred. In this
// case, no reason to skip merging because COPYs are expected to be place in
// Pred already.
if (SameIncomingValueBBs.count(Pred))
  return true;

BlockFrequency PredFreq = BFI->getBlockFreq(Pred);
BlockFrequency BBFreq = BFI->getBlockFreq(BB);

for (auto *SameValueBB : SameIncomingValueBBs)
  if (SameValueBB->getUniquePredecessor() == Pred &&
      DestBB == findDestBlockOfMergeableEmptyBlock(SameValueBB))
    BBFreq += BFI->getBlockFreq(SameValueBB);

return PredFreq.getFrequency() <=
       BBFreq.getFrequency() * FreqRatioToSkipMerge;
958}

960/// Return true if we can merge BB into DestBB if there is a single
961/// unconditional branch between them, and BB contains no other non-phi
962/// instructions.
963bool CodeGenPrepare::canMergeBlocks(const BasicBlock *BB,
                                  const BasicBlock *DestBB) const {
// We only want to eliminate blocks whose phi nodes are used by phi nodes in
// the successor.  If there are more complex condition (e.g. preheaders),
// don't mess around with them.
for (const PHINode &PN : BB->phis()) {
  for (const User *U : PN.users()) {
    const Instruction *UI = cast<Instruction>(U);
    if (UI->getParent() != DestBB || !isa<PHINode>(UI))
      return false;
    // If User is inside DestBB block and it is a PHINode then check
    // incoming value. If incoming value is not from BB then this is
    // a complex condition (e.g. preheaders) we want to avoid here.
    if (UI->getParent() == DestBB) {
      if (const PHINode *UPN = dyn_cast<PHINode>(UI))
        for (unsigned I = 0, E = UPN->getNumIncomingValues(); I != E; ++I) {
          Instruction *Insn = dyn_cast<Instruction>(UPN->getIncomingValue(I));
          if (Insn && Insn->getParent() == BB &&
              Insn->getParent() != UPN->getIncomingBlock(I))
            return false;
        }
    }
  }
}

// If BB and DestBB contain any common predecessors, then the phi nodes in BB
// and DestBB may have conflicting incoming values for the block.  If so, we
// can't merge the block.
const PHINode *DestBBPN = dyn_cast<PHINode>(DestBB->begin());
if (!DestBBPN)
  return true; // no conflict.

// Collect the preds of BB.
SmallPtrSet<const BasicBlock *, 16> BBPreds;
if (const PHINode *BBPN = dyn_cast<PHINode>(BB->begin())) {
  // It is faster to get preds from a PHI than with pred_iterator.
  for (unsigned i = 0, e = BBPN->getNumIncomingValues(); i != e; ++i)
    BBPreds.insert(BBPN->getIncomingBlock(i));
} else {
  BBPreds.insert(pred_begin(BB), pred_end(BB));
}

// Walk the preds of DestBB.
for (unsigned i = 0, e = DestBBPN->getNumIncomingValues(); i != e; ++i) {
  BasicBlock *Pred = DestBBPN->getIncomingBlock(i);
  if (BBPreds.count(Pred)) { // Common predecessor?
    for (const PHINode &PN : DestBB->phis()) {
      const Value *V1 = PN.getIncomingValueForBlock(Pred);
      const Value *V2 = PN.getIncomingValueForBlock(BB);

      // If V2 is a phi node in BB, look up what the mapped value will be.
      if (const PHINode *V2PN = dyn_cast<PHINode>(V2))
        if (V2PN->getParent() == BB)
          V2 = V2PN->getIncomingValueForBlock(Pred);

      // If there is a conflict, bail out.
      if (V1 != V2)
        return false;
    }
  }
}

return true;
1026}

1028/// Replace all old uses with new ones, and push the updated BBs into FreshBBs.
1029static void replaceAllUsesWith(Value *Old, Value *New,
                             SmallSet<BasicBlock *, 32> &FreshBBs,
                             bool IsHuge) {
auto *OldI = dyn_cast<Instruction>(Old);
if (OldI) {
  for (Value::user_iterator UI = OldI->user_begin(), E = OldI->user_end();
       UI != E; ++UI) {
    Instruction *User = cast<Instruction>(*UI);
    if (IsHuge)
      FreshBBs.insert(User->getParent());
  }
}
Old->replaceAllUsesWith(New);
1042}

1044/// Eliminate a basic block that has only phi's and an unconditional branch in
1045/// it.
1046void CodeGenPrepare::eliminateMostlyEmptyBlock(BasicBlock *BB) {
BranchInst *BI = cast<BranchInst>(BB->getTerminator());
BasicBlock *DestBB = BI->getSuccessor(0);

LLVM_DEBUG(dbgs() << "MERGING MOSTLY EMPTY BLOCKS - BEFORE:\n"do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "MERGING MOSTLY EMPTY BLOCKS - BEFORE:\n"
 << *BB << *DestBB; } } while (false)
                  << *BB << *DestBB)do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "MERGING MOSTLY EMPTY BLOCKS - BEFORE:\n"
 << *BB << *DestBB; } } while (false);

// If the destination block has a single pred, then this is a trivial edge,
// just collapse it.
if (BasicBlock *SinglePred = DestBB->getSinglePredecessor()) {
  if (SinglePred != DestBB) {
    assert(SinglePred == BB &&(static_cast <bool> (SinglePred == BB && "Single predecessor not the same as predecessor"
) ? void (0) : __assert_fail ("SinglePred == BB && \"Single predecessor not the same as predecessor\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1058, __extension__ __PRETTY_FUNCTION__
))
           "Single predecessor not the same as predecessor")(static_cast <bool> (SinglePred == BB && "Single predecessor not the same as predecessor"
) ? void (0) : __assert_fail ("SinglePred == BB && \"Single predecessor not the same as predecessor\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1058, __extension__ __PRETTY_FUNCTION__
));
    // Merge DestBB into SinglePred/BB and delete it.
    MergeBlockIntoPredecessor(DestBB);
    // Note: BB(=SinglePred) will not be deleted on this path.
    // DestBB(=its single successor) is the one that was deleted.
    LLVM_DEBUG(dbgs() << "AFTER:\n" << *SinglePred << "\n\n\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "AFTER:\n" << *SinglePred
 << "\n\n\n"; } } while (false);

    if (IsHugeFunc) {
      // Update FreshBBs to optimize the merged BB.
      FreshBBs.insert(SinglePred);
      FreshBBs.erase(DestBB);
    }
    return;
  }
}

// Otherwise, we have multiple predecessors of BB.  Update the PHIs in DestBB
// to handle the new incoming edges it is about to have.
for (PHINode &PN : DestBB->phis()) {
  // Remove the incoming value for BB, and remember it.
  Value *InVal = PN.removeIncomingValue(BB, false);

  // Two options: either the InVal is a phi node defined in BB or it is some
  // value that dominates BB.
  PHINode *InValPhi = dyn_cast<PHINode>(InVal);
  if (InValPhi && InValPhi->getParent() == BB) {
    // Add all of the input values of the input PHI as inputs of this phi.
    for (unsigned i = 0, e = InValPhi->getNumIncomingValues(); i != e; ++i)
      PN.addIncoming(InValPhi->getIncomingValue(i),
                     InValPhi->getIncomingBlock(i));
  } else {
    // Otherwise, add one instance of the dominating value for each edge that
    // we will be adding.
    if (PHINode *BBPN = dyn_cast<PHINode>(BB->begin())) {
      for (unsigned i = 0, e = BBPN->getNumIncomingValues(); i != e; ++i)
        PN.addIncoming(InVal, BBPN->getIncomingBlock(i));
    } else {
      for (BasicBlock *Pred : predecessors(BB))
        PN.addIncoming(InVal, Pred);
    }
  }
}

// The PHIs are now updated, change everything that refers to BB to use
// DestBB and remove BB.
BB->replaceAllUsesWith(DestBB);
BB->eraseFromParent();
++NumBlocksElim;

LLVM_DEBUG(dbgs() << "AFTER:\n" << *DestBB << "\n\n\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "AFTER:\n" << *DestBB
 << "\n\n\n"; } } while (false);
1108}

1110// Computes a map of base pointer relocation instructions to corresponding
1111// derived pointer relocation instructions given a vector of all relocate calls
1112static void computeBaseDerivedRelocateMap(
  const SmallVectorImpl<GCRelocateInst *> &AllRelocateCalls,
  DenseMap<GCRelocateInst *, SmallVector<GCRelocateInst *, 2>>
      &RelocateInstMap) {
// Collect information in two maps: one primarily for locating the base object
// while filling the second map; the second map is the final structure holding
// a mapping between Base and corresponding Derived relocate calls
DenseMap<std::pair<unsigned, unsigned>, GCRelocateInst *> RelocateIdxMap;
for (auto *ThisRelocate : AllRelocateCalls) {
  auto K = std::make_pair(ThisRelocate->getBasePtrIndex(),
                          ThisRelocate->getDerivedPtrIndex());
  RelocateIdxMap.insert(std::make_pair(K, ThisRelocate));
}
for (auto &Item : RelocateIdxMap) {
  std::pair<unsigned, unsigned> Key = Item.first;
  if (Key.first == Key.second)
    // Base relocation: nothing to insert
    continue;

  GCRelocateInst *I = Item.second;
  auto BaseKey = std::make_pair(Key.first, Key.first);

  // We're iterating over RelocateIdxMap so we cannot modify it.
  auto MaybeBase = RelocateIdxMap.find(BaseKey);
  if (MaybeBase == RelocateIdxMap.end())
    // TODO: We might want to insert a new base object relocate and gep off
    // that, if there are enough derived object relocates.
    continue;

  RelocateInstMap[MaybeBase->second].push_back(I);
}
1143}

1145// Accepts a GEP and extracts the operands into a vector provided they're all
1146// small integer constants
1147static bool getGEPSmallConstantIntOffsetV(GetElementPtrInst *GEP,
                                        SmallVectorImpl<Value *> &OffsetV) {
for (unsigned i = 1; i < GEP->getNumOperands(); i++) {
  // Only accept small constant integer operands
  auto *Op = dyn_cast<ConstantInt>(GEP->getOperand(i));
  if (!Op || Op->getZExtValue() > 20)
    return false;
}

for (unsigned i = 1; i < GEP->getNumOperands(); i++)
  OffsetV.push_back(GEP->getOperand(i));
return true;
1159}

1161// Takes a RelocatedBase (base pointer relocation instruction) and Targets to
1162// replace, computes a replacement, and affects it.
1163static bool
1164simplifyRelocatesOffABase(GCRelocateInst *RelocatedBase,
                        const SmallVectorImpl<GCRelocateInst *> &Targets) {
bool MadeChange = false;
// We must ensure the relocation of derived pointer is defined after
// relocation of base pointer. If we find a relocation corresponding to base
// defined earlier than relocation of base then we move relocation of base
// right before found relocation. We consider only relocation in the same
// basic block as relocation of base. Relocations from other basic block will
// be skipped by optimization and we do not care about them.
for (auto R = RelocatedBase->getParent()->getFirstInsertionPt();
     &*R != RelocatedBase; ++R)
  if (auto *RI = dyn_cast<GCRelocateInst>(R))
    if (RI->getStatepoint() == RelocatedBase->getStatepoint())
      if (RI->getBasePtrIndex() == RelocatedBase->getBasePtrIndex()) {
        RelocatedBase->moveBefore(RI);
        break;
      }

for (GCRelocateInst *ToReplace : Targets) {
  assert(ToReplace->getBasePtrIndex() == RelocatedBase->getBasePtrIndex() &&(static_cast <bool> (ToReplace->getBasePtrIndex() ==
 RelocatedBase->getBasePtrIndex() && "Not relocating a derived object of the original base object"
) ? void (0) : __assert_fail ("ToReplace->getBasePtrIndex() == RelocatedBase->getBasePtrIndex() && \"Not relocating a derived object of the original base object\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1184, __extension__ __PRETTY_FUNCTION__
))
         "Not relocating a derived object of the original base object")(static_cast <bool> (ToReplace->getBasePtrIndex() ==
 RelocatedBase->getBasePtrIndex() && "Not relocating a derived object of the original base object"
) ? void (0) : __assert_fail ("ToReplace->getBasePtrIndex() == RelocatedBase->getBasePtrIndex() && \"Not relocating a derived object of the original base object\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1184, __extension__ __PRETTY_FUNCTION__
));
  if (ToReplace->getBasePtrIndex() == ToReplace->getDerivedPtrIndex()) {
    // A duplicate relocate call. TODO: coalesce duplicates.
    continue;
  }

  if (RelocatedBase->getParent() != ToReplace->getParent()) {
    // Base and derived relocates are in different basic blocks.
    // In this case transform is only valid when base dominates derived
    // relocate. However it would be too expensive to check dominance
    // for each such relocate, so we skip the whole transformation.
    continue;
  }

  Value *Base = ToReplace->getBasePtr();
  auto *Derived = dyn_cast<GetElementPtrInst>(ToReplace->getDerivedPtr());
  if (!Derived || Derived->getPointerOperand() != Base)
    continue;

  SmallVector<Value *, 2> OffsetV;
  if (!getGEPSmallConstantIntOffsetV(Derived, OffsetV))
    continue;

  // Create a Builder and replace the target callsite with a gep
  assert(RelocatedBase->getNextNode() &&(static_cast <bool> (RelocatedBase->getNextNode() &&
 "Should always have one since it's not a terminator") ? void
 (0) : __assert_fail ("RelocatedBase->getNextNode() && \"Should always have one since it's not a terminator\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1209, __extension__ __PRETTY_FUNCTION__
))
         "Should always have one since it's not a terminator")(static_cast <bool> (RelocatedBase->getNextNode() &&
 "Should always have one since it's not a terminator") ? void
 (0) : __assert_fail ("RelocatedBase->getNextNode() && \"Should always have one since it's not a terminator\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1209, __extension__ __PRETTY_FUNCTION__
));

  // Insert after RelocatedBase
  IRBuilder<> Builder(RelocatedBase->getNextNode());
  Builder.SetCurrentDebugLocation(ToReplace->getDebugLoc());

  // If gc_relocate does not match the actual type, cast it to the right type.
  // In theory, there must be a bitcast after gc_relocate if the type does not
  // match, and we should reuse it to get the derived pointer. But it could be
  // cases like this:
  // bb1:
  //  ...
  //  %g1 = call coldcc i8 addrspace(1)*
  //  @llvm.experimental.gc.relocate.p1i8(...) br label %merge
  //
  // bb2:
  //  ...
  //  %g2 = call coldcc i8 addrspace(1)*
  //  @llvm.experimental.gc.relocate.p1i8(...) br label %merge
  //
  // merge:
  //  %p1 = phi i8 addrspace(1)* [ %g1, %bb1 ], [ %g2, %bb2 ]
  //  %cast = bitcast i8 addrspace(1)* %p1 in to i32 addrspace(1)*
  //
  // In this case, we can not find the bitcast any more. So we insert a new
  // bitcast no matter there is already one or not. In this way, we can handle
  // all cases, and the extra bitcast should be optimized away in later
  // passes.
  Value *ActualRelocatedBase = RelocatedBase;
  if (RelocatedBase->getType() != Base->getType()) {
    ActualRelocatedBase =
        Builder.CreateBitCast(RelocatedBase, Base->getType());
  }
  Value *Replacement =
      Builder.CreateGEP(Derived->getSourceElementType(), ActualRelocatedBase,
                        ArrayRef(OffsetV));
  Replacement->takeName(ToReplace);
  // If the newly generated derived pointer's type does not match the original
  // derived pointer's type, cast the new derived pointer to match it. Same
  // reasoning as above.
  Value *ActualReplacement = Replacement;
  if (Replacement->getType() != ToReplace->getType()) {
    ActualReplacement =
        Builder.CreateBitCast(Replacement, ToReplace->getType());
  }
  ToReplace->replaceAllUsesWith(ActualReplacement);
  ToReplace->eraseFromParent();

  MadeChange = true;
}
return MadeChange;
1260}

1262// Turns this:
1263//
1264// %base = ...
1265// %ptr = gep %base + 15
1266// %tok = statepoint (%fun, i32 0, i32 0, i32 0, %base, %ptr)
1267// %base' = relocate(%tok, i32 4, i32 4)
1268// %ptr' = relocate(%tok, i32 4, i32 5)
1269// %val = load %ptr'
1270//
1271// into this:
1272//
1273// %base = ...
1274// %ptr = gep %base + 15
1275// %tok = statepoint (%fun, i32 0, i32 0, i32 0, %base, %ptr)
1276// %base' = gc.relocate(%tok, i32 4, i32 4)
1277// %ptr' = gep %base' + 15
1278// %val = load %ptr'
1279bool CodeGenPrepare::simplifyOffsetableRelocate(GCStatepointInst &I) {
bool MadeChange = false;
SmallVector<GCRelocateInst *, 2> AllRelocateCalls;
for (auto *U : I.users())
  if (GCRelocateInst *Relocate = dyn_cast<GCRelocateInst>(U))
    // Collect all the relocate calls associated with a statepoint
    AllRelocateCalls.push_back(Relocate);

// We need at least one base pointer relocation + one derived pointer
// relocation to mangle
if (AllRelocateCalls.size() < 2)
  return false;

// RelocateInstMap is a mapping from the base relocate instruction to the
// corresponding derived relocate instructions
DenseMap<GCRelocateInst *, SmallVector<GCRelocateInst *, 2>> RelocateInstMap;
computeBaseDerivedRelocateMap(AllRelocateCalls, RelocateInstMap);
if (RelocateInstMap.empty())
  return false;

for (auto &Item : RelocateInstMap)
  // Item.first is the RelocatedBase to offset against
  // Item.second is the vector of Targets to replace
  MadeChange = simplifyRelocatesOffABase(Item.first, Item.second);
return MadeChange;
1304}

1306/// Sink the specified cast instruction into its user blocks.
1307static bool SinkCast(CastInst *CI) {
BasicBlock *DefBB = CI->getParent();

/// InsertedCasts - Only insert a cast in each block once.
DenseMap<BasicBlock *, CastInst *> InsertedCasts;

bool MadeChange = false;
for (Value::user_iterator UI = CI->user_begin(), E = CI->user_end();
     UI != E;) {
  Use &TheUse = UI.getUse();
  Instruction *User = cast<Instruction>(*UI);

  // Figure out which BB this cast is used in.  For PHI's this is the
  // appropriate predecessor block.
  BasicBlock *UserBB = User->getParent();
  if (PHINode *PN = dyn_cast<PHINode>(User)) {
    UserBB = PN->getIncomingBlock(TheUse);
  }

  // Preincrement use iterator so we don't invalidate it.
  ++UI;

  // The first insertion point of a block containing an EH pad is after the
  // pad.  If the pad is the user, we cannot sink the cast past the pad.
  if (User->isEHPad())
    continue;

  // If the block selected to receive the cast is an EH pad that does not
  // allow non-PHI instructions before the terminator, we can't sink the
  // cast.
  if (UserBB->getTerminator()->isEHPad())
    continue;

  // If this user is in the same block as the cast, don't change the cast.
  if (UserBB == DefBB)
    continue;

  // If we have already inserted a cast into this block, use it.
  CastInst *&InsertedCast = InsertedCasts[UserBB];

  if (!InsertedCast) {
    BasicBlock::iterator InsertPt = UserBB->getFirstInsertionPt();
    assert(InsertPt != UserBB->end())(static_cast <bool> (InsertPt != UserBB->end()) ? void
 (0) : __assert_fail ("InsertPt != UserBB->end()", "llvm/lib/CodeGen/CodeGenPrepare.cpp"
, 1349, __extension__ __PRETTY_FUNCTION__));
    InsertedCast = CastInst::Create(CI->getOpcode(), CI->getOperand(0),
                                    CI->getType(), "", &*InsertPt);
    InsertedCast->setDebugLoc(CI->getDebugLoc());
  }

  // Replace a use of the cast with a use of the new cast.
  TheUse = InsertedCast;
  MadeChange = true;
  ++NumCastUses;
}

// If we removed all uses, nuke the cast.
if (CI->use_empty()) {
  salvageDebugInfo(*CI);
  CI->eraseFromParent();
  MadeChange = true;
}

return MadeChange;
1369}

1371/// If the specified cast instruction is a noop copy (e.g. it's casting from
1372/// one pointer type to another, i32->i8 on PPC), sink it into user blocks to
1373/// reduce the number of virtual registers that must be created and coalesced.
1374///
1375/// Return true if any changes are made.
1376static bool OptimizeNoopCopyExpression(CastInst *CI, const TargetLowering &TLI,
                                     const DataLayout &DL) {
// Sink only "cheap" (or nop) address-space casts.  This is a weaker condition
// than sinking only nop casts, but is helpful on some platforms.
if (auto *ASC = dyn_cast<AddrSpaceCastInst>(CI)) {
  if (!TLI.isFreeAddrSpaceCast(ASC->getSrcAddressSpace(),
                               ASC->getDestAddressSpace()))
    return false;
}

// If this is a noop copy,
EVT SrcVT = TLI.getValueType(DL, CI->getOperand(0)->getType());
EVT DstVT = TLI.getValueType(DL, CI->getType());

// This is an fp<->int conversion?
if (SrcVT.isInteger() != DstVT.isInteger())
  return false;

// If this is an extension, it will be a zero or sign extension, which
// isn't a noop.
if (SrcVT.bitsLT(DstVT))
  return false;

// If these values will be promoted, find out what they will be promoted
// to.  This helps us consider truncates on PPC as noop copies when they
// are.
if (TLI.getTypeAction(CI->getContext(), SrcVT) ==
    TargetLowering::TypePromoteInteger)
  SrcVT = TLI.getTypeToTransformTo(CI->getContext(), SrcVT);
if (TLI.getTypeAction(CI->getContext(), DstVT) ==
    TargetLowering::TypePromoteInteger)
  DstVT = TLI.getTypeToTransformTo(CI->getContext(), DstVT);

// If, after promotion, these are the same types, this is a noop copy.
if (SrcVT != DstVT)
  return false;

return SinkCast(CI);
1414}

1416// Match a simple increment by constant operation.  Note that if a sub is
1417// matched, the step is negated (as if the step had been canonicalized to
1418// an add, even though we leave the instruction alone.)
1419bool matchIncrement(const Instruction *IVInc, Instruction *&LHS,
                  Constant *&Step) {
if (match(IVInc, m_Add(m_Instruction(LHS), m_Constant(Step))) ||
    match(IVInc, m_ExtractValue<0>(m_Intrinsic<Intrinsic::uadd_with_overflow>(
                     m_Instruction(LHS), m_Constant(Step)))))
  return true;
if (match(IVInc, m_Sub(m_Instruction(LHS), m_Constant(Step))) ||
    match(IVInc, m_ExtractValue<0>(m_Intrinsic<Intrinsic::usub_with_overflow>(
                     m_Instruction(LHS), m_Constant(Step))))) {
  Step = ConstantExpr::getNeg(Step);
  return true;
}
return false;
1432}

1434/// If given \p PN is an inductive variable with value IVInc coming from the
1435/// backedge, and on each iteration it gets increased by Step, return pair
1436/// <IVInc, Step>. Otherwise, return std::nullopt.
1437static std::optional<std::pair<Instruction *, Constant *>>
1438getIVIncrement(const PHINode *PN, const LoopInfo *LI) {
const Loop *L = LI->getLoopFor(PN->getParent());
if (!L || L->getHeader() != PN->getParent() || !L->getLoopLatch())
  return std::nullopt;
auto *IVInc =
    dyn_cast<Instruction>(PN->getIncomingValueForBlock(L->getLoopLatch()));
if (!IVInc || LI->getLoopFor(IVInc->getParent()) != L)
  return std::nullopt;
Instruction *LHS = nullptr;
Constant *Step = nullptr;
if (matchIncrement(IVInc, LHS, Step) && LHS == PN)
  return std::make_pair(IVInc, Step);
return std::nullopt;
1451}

1453static bool isIVIncrement(const Value *V, const LoopInfo *LI) {
auto *I = dyn_cast<Instruction>(V);
if (!I)
  return false;
Instruction *LHS = nullptr;
Constant *Step = nullptr;
if (!matchIncrement(I, LHS, Step))
  return false;
if (auto *PN = dyn_cast<PHINode>(LHS))
  if (auto IVInc = getIVIncrement(PN, LI))
    return IVInc->first == I;
return false;
1465}

1467bool CodeGenPrepare::replaceMathCmpWithIntrinsic(BinaryOperator *BO,
                                               Value *Arg0, Value *Arg1,
                                               CmpInst *Cmp,
                                               Intrinsic::ID IID) {
auto IsReplacableIVIncrement = [this, &Cmp](BinaryOperator *BO) {
  if (!isIVIncrement(BO, LI))
    return false;
  const Loop *L = LI->getLoopFor(BO->getParent());
  assert(L && "L should not be null after isIVIncrement()")(static_cast <bool> (L && "L should not be null after isIVIncrement()"
) ? void (0) : __assert_fail ("L && \"L should not be null after isIVIncrement()\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1475, __extension__ __PRETTY_FUNCTION__
));
  // Do not risk on moving increment into a child loop.
  if (LI->getLoopFor(Cmp->getParent()) != L)
    return false;

  // Finally, we need to ensure that the insert point will dominate all
  // existing uses of the increment.

  auto &DT = getDT(*BO->getParent()->getParent());
  if (DT.dominates(Cmp->getParent(), BO->getParent()))
    // If we're moving up the dom tree, all uses are trivially dominated.
    // (This is the common case for code produced by LSR.)
    return true;

  // Otherwise, special case the single use in the phi recurrence.
  return BO->hasOneUse() && DT.dominates(Cmp->getParent(), L->getLoopLatch());
};
if (BO->getParent() != Cmp->getParent() && !IsReplacableIVIncrement(BO)) {
  // We used to use a dominator tree here to allow multi-block optimization.
  // But that was problematic because:
  // 1. It could cause a perf regression by hoisting the math op into the
  //    critical path.
  // 2. It could cause a perf regression by creating a value that was live
  //    across multiple blocks and increasing register pressure.
  // 3. Use of a dominator tree could cause large compile-time regression.
  //    This is because we recompute the DT on every change in the main CGP
  //    run-loop. The recomputing is probably unnecessary in many cases, so if
  //    that was fixed, using a DT here would be ok.
  //
  // There is one important particular case we still want to handle: if BO is
  // the IV increment. Important properties that make it profitable:
  // - We can speculate IV increment anywhere in the loop (as long as the
  //   indvar Phi is its only user);
  // - Upon computing Cmp, we effectively compute something equivalent to the
  //   IV increment (despite it loops differently in the IR). So moving it up
  //   to the cmp point does not really increase register pressure.
  return false;
}

// We allow matching the canonical IR (add X, C) back to (usubo X, -C).
if (BO->getOpcode() == Instruction::Add &&
    IID == Intrinsic::usub_with_overflow) {
  assert(isa<Constant>(Arg1) && "Unexpected input for usubo")(static_cast <bool> (isa<Constant>(Arg1) &&
 "Unexpected input for usubo") ? void (0) : __assert_fail ("isa<Constant>(Arg1) && \"Unexpected input for usubo\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1517, __extension__ __PRETTY_FUNCTION__
));
  Arg1 = ConstantExpr::getNeg(cast<Constant>(Arg1));
}

// Insert at the first instruction of the pair.
Instruction *InsertPt = nullptr;
for (Instruction &Iter : *Cmp->getParent()) {
  // If BO is an XOR, it is not guaranteed that it comes after both inputs to
  // the overflow intrinsic are defined.
  if ((BO->getOpcode() != Instruction::Xor && &Iter == BO) || &Iter == Cmp) {
    InsertPt = &Iter;
    break;
  }
}
assert(InsertPt != nullptr && "Parent block did not contain cmp or binop")(static_cast <bool> (InsertPt != nullptr && "Parent block did not contain cmp or binop"
) ? void (0) : __assert_fail ("InsertPt != nullptr && \"Parent block did not contain cmp or binop\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1531, __extension__ __PRETTY_FUNCTION__
));

IRBuilder<> Builder(InsertPt);
Value *MathOV = Builder.CreateBinaryIntrinsic(IID, Arg0, Arg1);
if (BO->getOpcode() != Instruction::Xor) {
  Value *Math = Builder.CreateExtractValue(MathOV, 0, "math");
  replaceAllUsesWith(BO, Math, FreshBBs, IsHugeFunc);
} else
  assert(BO->hasOneUse() &&(static_cast <bool> (BO->hasOneUse() && "Patterns with XOr should use the BO only in the compare"
) ? void (0) : __assert_fail ("BO->hasOneUse() && \"Patterns with XOr should use the BO only in the compare\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1540, __extension__ __PRETTY_FUNCTION__
))
         "Patterns with XOr should use the BO only in the compare")(static_cast <bool> (BO->hasOneUse() && "Patterns with XOr should use the BO only in the compare"
) ? void (0) : __assert_fail ("BO->hasOneUse() && \"Patterns with XOr should use the BO only in the compare\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1540, __extension__ __PRETTY_FUNCTION__
));
Value *OV = Builder.CreateExtractValue(MathOV, 1, "ov");
replaceAllUsesWith(Cmp, OV, FreshBBs, IsHugeFunc);
Cmp->eraseFromParent();
BO->eraseFromParent();
return true;
1546}

1548/// Match special-case patterns that check for unsigned add overflow.
1549static bool matchUAddWithOverflowConstantEdgeCases(CmpInst *Cmp,
                                                 BinaryOperator *&Add) {
// Add = add A, 1; Cmp = icmp eq A,-1 (overflow if A is max val)
// Add = add A,-1; Cmp = icmp ne A, 0 (overflow if A is non-zero)
Value *A = Cmp->getOperand(0), *B = Cmp->getOperand(1);

// We are not expecting non-canonical/degenerate code. Just bail out.
if (isa<Constant>(A))
  return false;

ICmpInst::Predicate Pred = Cmp->getPredicate();
if (Pred == ICmpInst::ICMP_EQ && match(B, m_AllOnes()))
  B = ConstantInt::get(B->getType(), 1);
else if (Pred == ICmpInst::ICMP_NE && match(B, m_ZeroInt()))
  B = ConstantInt::get(B->getType(), -1);
else
  return false;

// Check the users of the variable operand of the compare looking for an add
// with the adjusted constant.
for (User *U : A->users()) {
  if (match(U, m_Add(m_Specific(A), m_Specific(B)))) {
    Add = cast<BinaryOperator>(U);
    return true;
  }
}
return false;
1576}

1578/// Try to combine the compare into a call to the llvm.uadd.with.overflow
1579/// intrinsic. Return true if any changes were made.
1580bool CodeGenPrepare::combineToUAddWithOverflow(CmpInst *Cmp,
                                             ModifyDT &ModifiedDT) {
Value *A, *B;
BinaryOperator *Add;
if (!match(Cmp, m_UAddWithOverflow(m_Value(A), m_Value(B), m_BinOp(Add)))) {
  if (!matchUAddWithOverflowConstantEdgeCases(Cmp, Add))
    return false;
  // Set A and B in case we match matchUAddWithOverflowConstantEdgeCases.
  A = Add->getOperand(0);
  B = Add->getOperand(1);
}

if (!TLI->shouldFormOverflowOp(ISD::UADDO,
                               TLI->getValueType(*DL, Add->getType()),
                               Add->hasNUsesOrMore(2)))
  return false;

// We don't want to move around uses of condition values this late, so we
// check if it is legal to create the call to the intrinsic in the basic
// block containing the icmp.
if (Add->getParent() != Cmp->getParent() && !Add->hasOneUse())
  return false;

if (!replaceMathCmpWithIntrinsic(Add, A, B, Cmp,
                                 Intrinsic::uadd_with_overflow))
  return false;

// Reset callers - do not crash by iterating over a dead instruction.
ModifiedDT = ModifyDT::ModifyInstDT;
return true;
1610}

1612bool CodeGenPrepare::combineToUSubWithOverflow(CmpInst *Cmp,
                                             ModifyDT &ModifiedDT) {
// We are not expecting non-canonical/degenerate code. Just bail out.
Value *A = Cmp->getOperand(0), *B = Cmp->getOperand(1);
if (isa<Constant>(A) && isa<Constant>(B))
  return false;

// Convert (A u> B) to (A u< B) to simplify pattern matching.
ICmpInst::Predicate Pred = Cmp->getPredicate();
if (Pred == ICmpInst::ICMP_UGT) {
  std::swap(A, B);
  Pred = ICmpInst::ICMP_ULT;
}
// Convert special-case: (A == 0) is the same as (A u< 1).
if (Pred == ICmpInst::ICMP_EQ && match(B, m_ZeroInt())) {
  B = ConstantInt::get(B->getType(), 1);
  Pred = ICmpInst::ICMP_ULT;
}
// Convert special-case: (A != 0) is the same as (0 u< A).
if (Pred == ICmpInst::ICMP_NE && match(B, m_ZeroInt())) {
  std::swap(A, B);
  Pred = ICmpInst::ICMP_ULT;
}
if (Pred != ICmpInst::ICMP_ULT)
  return false;

// Walk the users of a variable operand of a compare looking for a subtract or
// add with that same operand. Also match the 2nd operand of the compare to
// the add/sub, but that may be a negated constant operand of an add.
Value *CmpVariableOperand = isa<Constant>(A) ? B : A;
BinaryOperator *Sub = nullptr;
for (User *U : CmpVariableOperand->users()) {
  // A - B, A u< B --> usubo(A, B)
  if (match(U, m_Sub(m_Specific(A), m_Specific(B)))) {
    Sub = cast<BinaryOperator>(U);
    break;
  }

  // A + (-C), A u< C (canonicalized form of (sub A, C))
  const APInt *CmpC, *AddC;
  if (match(U, m_Add(m_Specific(A), m_APInt(AddC))) &&
      match(B, m_APInt(CmpC)) && *AddC == -(*CmpC)) {
    Sub = cast<BinaryOperator>(U);
    break;
  }
}
if (!Sub)
  return false;

if (!TLI->shouldFormOverflowOp(ISD::USUBO,
                               TLI->getValueType(*DL, Sub->getType()),
                               Sub->hasNUsesOrMore(2)))
  return false;

if (!replaceMathCmpWithIntrinsic(Sub, Sub->getOperand(0), Sub->getOperand(1),
                                 Cmp, Intrinsic::usub_with_overflow))
  return false;

// Reset callers - do not crash by iterating over a dead instruction.
ModifiedDT = ModifyDT::ModifyInstDT;
return true;
1673}

1675/// Sink the given CmpInst into user blocks to reduce the number of virtual
1676/// registers that must be created and coalesced. This is a clear win except on
1677/// targets with multiple condition code registers (PowerPC), where it might
1678/// lose; some adjustment may be wanted there.
1679///
1680/// Return true if any changes are made.
1681static bool sinkCmpExpression(CmpInst *Cmp, const TargetLowering &TLI) {
if (TLI.hasMultipleConditionRegisters())
  return false;

// Avoid sinking soft-FP comparisons, since this can move them into a loop.
if (TLI.useSoftFloat() && isa<FCmpInst>(Cmp))
  return false;

// Only insert a cmp in each block once.
DenseMap<BasicBlock *, CmpInst *> InsertedCmps;

bool MadeChange = false;
for (Value::user_iterator UI = Cmp->user_begin(), E = Cmp->user_end();
     UI != E;) {
  Use &TheUse = UI.getUse();
  Instruction *User = cast<Instruction>(*UI);

  // Preincrement use iterator so we don't invalidate it.
  ++UI;

  // Don't bother for PHI nodes.
  if (isa<PHINode>(User))
    continue;

  // Figure out which BB this cmp is used in.
  BasicBlock *UserBB = User->getParent();
  BasicBlock *DefBB = Cmp->getParent();

  // If this user is in the same block as the cmp, don't change the cmp.
  if (UserBB == DefBB)
    continue;

  // If we have already inserted a cmp into this block, use it.
  CmpInst *&InsertedCmp = InsertedCmps[UserBB];

  if (!InsertedCmp) {
    BasicBlock::iterator InsertPt = UserBB->getFirstInsertionPt();
    assert(InsertPt != UserBB->end())(static_cast <bool> (InsertPt != UserBB->end()) ? void
 (0) : __assert_fail ("InsertPt != UserBB->end()", "llvm/lib/CodeGen/CodeGenPrepare.cpp"
, 1718, __extension__ __PRETTY_FUNCTION__));
    InsertedCmp = CmpInst::Create(Cmp->getOpcode(), Cmp->getPredicate(),
                                  Cmp->getOperand(0), Cmp->getOperand(1), "",
                                  &*InsertPt);
    // Propagate the debug info.
    InsertedCmp->setDebugLoc(Cmp->getDebugLoc());
  }

  // Replace a use of the cmp with a use of the new cmp.
  TheUse = InsertedCmp;
  MadeChange = true;
  ++NumCmpUses;
}

// If we removed all uses, nuke the cmp.
if (Cmp->use_empty()) {
  Cmp->eraseFromParent();
  MadeChange = true;
}

return MadeChange;
1739}

1741/// For pattern like:
1742///
1743///   DomCond = icmp sgt/slt CmpOp0, CmpOp1 (might not be in DomBB)
1744///   ...
1745/// DomBB:
1746///   ...
1747///   br DomCond, TrueBB, CmpBB
1748/// CmpBB: (with DomBB being the single predecessor)
1749///   ...
1750///   Cmp = icmp eq CmpOp0, CmpOp1
1751///   ...
1752///
1753/// It would use two comparison on targets that lowering of icmp sgt/slt is
1754/// different from lowering of icmp eq (PowerPC). This function try to convert
1755/// 'Cmp = icmp eq CmpOp0, CmpOp1' to ' Cmp = icmp slt/sgt CmpOp0, CmpOp1'.
1756/// After that, DomCond and Cmp can use the same comparison so reduce one
1757/// comparison.
1758///
1759/// Return true if any changes are made.
1760static bool foldICmpWithDominatingICmp(CmpInst *Cmp,
                                     const TargetLowering &TLI) {
if (!EnableICMP_EQToICMP_ST && TLI.isEqualityCmpFoldedWithSignedCmp())
  return false;

ICmpInst::Predicate Pred = Cmp->getPredicate();
if (Pred != ICmpInst::ICMP_EQ)
  return false;

// If icmp eq has users other than BranchInst and SelectInst, converting it to
// icmp slt/sgt would introduce more redundant LLVM IR.
for (User *U : Cmp->users()) {
  if (isa<BranchInst>(U))
    continue;
  if (isa<SelectInst>(U) && cast<SelectInst>(U)->getCondition() == Cmp)
    continue;
  return false;
}

// This is a cheap/incomplete check for dominance - just match a single
// predecessor with a conditional branch.
BasicBlock *CmpBB = Cmp->getParent();
BasicBlock *DomBB = CmpBB->getSinglePredecessor();
if (!DomBB)
  return false;

// We want to ensure that the only way control gets to the comparison of
// interest is that a less/greater than comparison on the same operands is
// false.
Value *DomCond;
BasicBlock *TrueBB, *FalseBB;
if (!match(DomBB->getTerminator(), m_Br(m_Value(DomCond), TrueBB, FalseBB)))
  return false;
if (CmpBB != FalseBB)
  return false;

Value *CmpOp0 = Cmp->getOperand(0), *CmpOp1 = Cmp->getOperand(1);
ICmpInst::Predicate DomPred;
if (!match(DomCond, m_ICmp(DomPred, m_Specific(CmpOp0), m_Specific(CmpOp1))))
  return false;
if (DomPred != ICmpInst::ICMP_SGT && DomPred != ICmpInst::ICMP_SLT)
  return false;

// Convert the equality comparison to the opposite of the dominating
// comparison and swap the direction for all branch/select users.
// We have conceptually converted:
// Res = (a < b) ? <LT_RES> : (a == b) ? <EQ_RES> : <GT_RES>;
// to
// Res = (a < b) ? <LT_RES> : (a > b)  ? <GT_RES> : <EQ_RES>;
// And similarly for branches.
for (User *U : Cmp->users()) {
  if (auto *BI = dyn_cast<BranchInst>(U)) {
    assert(BI->isConditional() && "Must be conditional")(static_cast <bool> (BI->isConditional() && "Must be conditional"
) ? void (0) : __assert_fail ("BI->isConditional() && \"Must be conditional\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1812, __extension__ __PRETTY_FUNCTION__
));
    BI->swapSuccessors();
    continue;
  }
  if (auto *SI = dyn_cast<SelectInst>(U)) {
    // Swap operands
    SI->swapValues();
    SI->swapProfMetadata();
    continue;
  }
  llvm_unreachable("Must be a branch or a select")::llvm::llvm_unreachable_internal("Must be a branch or a select"
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1822);
}
Cmp->setPredicate(CmpInst::getSwappedPredicate(DomPred));
return true;
1826}

1828bool CodeGenPrepare::optimizeCmp(CmpInst *Cmp, ModifyDT &ModifiedDT) {
if (sinkCmpExpression(Cmp, *TLI))
  return true;

if (combineToUAddWithOverflow(Cmp, ModifiedDT))
  return true;

if (combineToUSubWithOverflow(Cmp, ModifiedDT))
  return true;

if (foldICmpWithDominatingICmp(Cmp, *TLI))
  return true;

return false;
1842}

1844/// Duplicate and sink the given 'and' instruction into user blocks where it is
1845/// used in a compare to allow isel to generate better code for targets where
1846/// this operation can be combined.
1847///
1848/// Return true if any changes are made.
1849static bool sinkAndCmp0Expression(Instruction *AndI, const TargetLowering &TLI,
                                SetOfInstrs &InsertedInsts) {
// Double-check that we're not trying to optimize an instruction that was
// already optimized by some other part of this pass.
assert(!InsertedInsts.count(AndI) &&(static_cast <bool> (!InsertedInsts.count(AndI) &&
 "Attempting to optimize already optimized and instruction") ?
 void (0) : __assert_fail ("!InsertedInsts.count(AndI) && \"Attempting to optimize already optimized and instruction\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1854, __extension__ __PRETTY_FUNCTION__
))
       "Attempting to optimize already optimized and instruction")(static_cast <bool> (!InsertedInsts.count(AndI) &&
 "Attempting to optimize already optimized and instruction") ?
 void (0) : __assert_fail ("!InsertedInsts.count(AndI) && \"Attempting to optimize already optimized and instruction\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1854, __extension__ __PRETTY_FUNCTION__
));
(void)InsertedInsts;

// Nothing to do for single use in same basic block.
if (AndI->hasOneUse() &&
    AndI->getParent() == cast<Instruction>(*AndI->user_begin())->getParent())
  return false;

// Try to avoid cases where sinking/duplicating is likely to increase register
// pressure.
if (!isa<ConstantInt>(AndI->getOperand(0)) &&
    !isa<ConstantInt>(AndI->getOperand(1)) &&
    AndI->getOperand(0)->hasOneUse() && AndI->getOperand(1)->hasOneUse())
  return false;

for (auto *U : AndI->users()) {
  Instruction *User = cast<Instruction>(U);

  // Only sink 'and' feeding icmp with 0.
  if (!isa<ICmpInst>(User))
    return false;

  auto *CmpC = dyn_cast<ConstantInt>(User->getOperand(1));
  if (!CmpC || !CmpC->isZero())
    return false;
}

if (!TLI.isMaskAndCmp0FoldingBeneficial(*AndI))
  return false;

LLVM_DEBUG(dbgs() << "found 'and' feeding only icmp 0;\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "found 'and' feeding only icmp 0;\n"
; } } while (false);
LLVM_DEBUG(AndI->getParent()->dump())do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { AndI->getParent()->dump(); } } while
 (false);

// Push the 'and' into the same block as the icmp 0.  There should only be
// one (icmp (and, 0)) in each block, since CSE/GVN should have removed any
// others, so we don't need to keep track of which BBs we insert into.
for (Value::user_iterator UI = AndI->user_begin(), E = AndI->user_end();
     UI != E;) {
  Use &TheUse = UI.getUse();
  Instruction *User = cast<Instruction>(*UI);

  // Preincrement use iterator so we don't invalidate it.
  ++UI;

  LLVM_DEBUG(dbgs() << "sinking 'and' use: " << *User << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "sinking 'and' use: " <<
 *User << "\n"; } } while (false);

  // Keep the 'and' in the same place if the use is already in the same block.
  Instruction *InsertPt =
      User->getParent() == AndI->getParent() ? AndI : User;
  Instruction *InsertedAnd =
      BinaryOperator::Create(Instruction::And, AndI->getOperand(0),
                             AndI->getOperand(1), "", InsertPt);
  // Propagate the debug info.
  InsertedAnd->setDebugLoc(AndI->getDebugLoc());

  // Replace a use of the 'and' with a use of the new 'and'.
  TheUse = InsertedAnd;
  ++NumAndUses;
  LLVM_DEBUG(User->getParent()->dump())do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { User->getParent()->dump(); } } while
 (false);
}

// We removed all uses, nuke the and.
AndI->eraseFromParent();
return true;
1918}

1920/// Check if the candidates could be combined with a shift instruction, which
1921/// includes:
1922/// 1. Truncate instruction
1923/// 2. And instruction and the imm is a mask of the low bits:
1924/// imm & (imm+1) == 0
1925static bool isExtractBitsCandidateUse(Instruction *User) {
if (!isa<TruncInst>(User)) {
  if (User->getOpcode() != Instruction::And ||
      !isa<ConstantInt>(User->getOperand(1)))
    return false;

  const APInt &Cimm = cast<ConstantInt>(User->getOperand(1))->getValue();

  if ((Cimm & (Cimm + 1)).getBoolValue())
    return false;
}
return true;
1937}

1939/// Sink both shift and truncate instruction to the use of truncate's BB.
1940static bool
1941SinkShiftAndTruncate(BinaryOperator *ShiftI, Instruction *User, ConstantInt *CI,
                   DenseMap<BasicBlock *, BinaryOperator *> &InsertedShifts,
                   const TargetLowering &TLI, const DataLayout &DL) {
BasicBlock *UserBB = User->getParent();
DenseMap<BasicBlock *, CastInst *> InsertedTruncs;
auto *TruncI = cast<TruncInst>(User);
bool MadeChange = false;

for (Value::user_iterator TruncUI = TruncI->user_begin(),
                          TruncE = TruncI->user_end();
     TruncUI != TruncE;) {

  Use &TruncTheUse = TruncUI.getUse();
  Instruction *TruncUser = cast<Instruction>(*TruncUI);
  // Preincrement use iterator so we don't invalidate it.

  ++TruncUI;

  int ISDOpcode = TLI.InstructionOpcodeToISD(TruncUser->getOpcode());
  if (!ISDOpcode)
    continue;

  // If the use is actually a legal node, there will not be an
  // implicit truncate.
  // FIXME: always querying the result type is just an
  // approximation; some nodes' legality is determined by the
  // operand or other means. There's no good way to find out though.
  if (TLI.isOperationLegalOrCustom(
          ISDOpcode, TLI.getValueType(DL, TruncUser->getType(), true)))
    continue;

  // Don't bother for PHI nodes.
  if (isa<PHINode>(TruncUser))
    continue;

  BasicBlock *TruncUserBB = TruncUser->getParent();

  if (UserBB == TruncUserBB)
    continue;

  BinaryOperator *&InsertedShift = InsertedShifts[TruncUserBB];
  CastInst *&InsertedTrunc = InsertedTruncs[TruncUserBB];

  if (!InsertedShift && !InsertedTrunc) {
    BasicBlock::iterator InsertPt = TruncUserBB->getFirstInsertionPt();
    assert(InsertPt != TruncUserBB->end())(static_cast <bool> (InsertPt != TruncUserBB->end())
 ? void (0) : __assert_fail ("InsertPt != TruncUserBB->end()"
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1986, __extension__ __PRETTY_FUNCTION__
));
    // Sink the shift
    if (ShiftI->getOpcode() == Instruction::AShr)
      InsertedShift = BinaryOperator::CreateAShr(ShiftI->getOperand(0), CI,
                                                 "", &*InsertPt);
    else
      InsertedShift = BinaryOperator::CreateLShr(ShiftI->getOperand(0), CI,
                                                 "", &*InsertPt);
    InsertedShift->setDebugLoc(ShiftI->getDebugLoc());

    // Sink the trunc
    BasicBlock::iterator TruncInsertPt = TruncUserBB->getFirstInsertionPt();
    TruncInsertPt++;
    assert(TruncInsertPt != TruncUserBB->end())(static_cast <bool> (TruncInsertPt != TruncUserBB->end
()) ? void (0) : __assert_fail ("TruncInsertPt != TruncUserBB->end()"
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 1999, __extension__ __PRETTY_FUNCTION__
));

    InsertedTrunc = CastInst::Create(TruncI->getOpcode(), InsertedShift,
                                     TruncI->getType(), "", &*TruncInsertPt);
    InsertedTrunc->setDebugLoc(TruncI->getDebugLoc());

    MadeChange = true;

    TruncTheUse = InsertedTrunc;
  }
}
return MadeChange;
2011}

2013/// Sink the shift *right* instruction into user blocks if the uses could
2014/// potentially be combined with this shift instruction and generate BitExtract
2015/// instruction. It will only be applied if the architecture supports BitExtract
2016/// instruction. Here is an example:
2017/// BB1:
2018///   %x.extract.shift = lshr i64 %arg1, 32
2019/// BB2:
2020///   %x.extract.trunc = trunc i64 %x.extract.shift to i16
2021/// ==>
2022///
2023/// BB2:
2024///   %x.extract.shift.1 = lshr i64 %arg1, 32
2025///   %x.extract.trunc = trunc i64 %x.extract.shift.1 to i16
2026///
2027/// CodeGen will recognize the pattern in BB2 and generate BitExtract
2028/// instruction.
2029/// Return true if any changes are made.
2030static bool OptimizeExtractBits(BinaryOperator *ShiftI, ConstantInt *CI,
                              const TargetLowering &TLI,
                              const DataLayout &DL) {
BasicBlock *DefBB = ShiftI->getParent();

/// Only insert instructions in each block once.
DenseMap<BasicBlock *, BinaryOperator *> InsertedShifts;

bool shiftIsLegal = TLI.isTypeLegal(TLI.getValueType(DL, ShiftI->getType()));

bool MadeChange = false;
for (Value::user_iterator UI = ShiftI->user_begin(), E = ShiftI->user_end();
     UI != E;) {
  Use &TheUse = UI.getUse();
  Instruction *User = cast<Instruction>(*UI);
  // Preincrement use iterator so we don't invalidate it.
  ++UI;

  // Don't bother for PHI nodes.
  if (isa<PHINode>(User))
    continue;

  if (!isExtractBitsCandidateUse(User))
    continue;

  BasicBlock *UserBB = User->getParent();

  if (UserBB == DefBB) {
    // If the shift and truncate instruction are in the same BB. The use of
    // the truncate(TruncUse) may still introduce another truncate if not
    // legal. In this case, we would like to sink both shift and truncate
    // instruction to the BB of TruncUse.
    // for example:
    // BB1:
    // i64 shift.result = lshr i64 opnd, imm
    // trunc.result = trunc shift.result to i16
    //
    // BB2:
    //   ----> We will have an implicit truncate here if the architecture does
    //   not have i16 compare.
    // cmp i16 trunc.result, opnd2
    //
    if (isa<TruncInst>(User) &&
        shiftIsLegal
        // If the type of the truncate is legal, no truncate will be
        // introduced in other basic blocks.
        && (!TLI.isTypeLegal(TLI.getValueType(DL, User->getType()))))
      MadeChange =
          SinkShiftAndTruncate(ShiftI, User, CI, InsertedShifts, TLI, DL);

    continue;
  }
  // If we have already inserted a shift into this block, use it.
  BinaryOperator *&InsertedShift = InsertedShifts[UserBB];

  if (!InsertedShift) {
    BasicBlock::iterator InsertPt = UserBB->getFirstInsertionPt();
    assert(InsertPt != UserBB->end())(static_cast <bool> (InsertPt != UserBB->end()) ? void
 (0) : __assert_fail ("InsertPt != UserBB->end()", "llvm/lib/CodeGen/CodeGenPrepare.cpp"
, 2087, __extension__ __PRETTY_FUNCTION__));

    if (ShiftI->getOpcode() == Instruction::AShr)
      InsertedShift = BinaryOperator::CreateAShr(ShiftI->getOperand(0), CI,
                                                 "", &*InsertPt);
    else
      InsertedShift = BinaryOperator::CreateLShr(ShiftI->getOperand(0), CI,
                                                 "", &*InsertPt);
    InsertedShift->setDebugLoc(ShiftI->getDebugLoc());

    MadeChange = true;
  }

  // Replace a use of the shift with a use of the new shift.
  TheUse = InsertedShift;
}

// If we removed all uses, or there are none, nuke the shift.
if (ShiftI->use_empty()) {
  salvageDebugInfo(*ShiftI);
  ShiftI->eraseFromParent();
  MadeChange = true;
}

return MadeChange;
2112}

2114/// If counting leading or trailing zeros is an expensive operation and a zero
2115/// input is defined, add a check for zero to avoid calling the intrinsic.
2116///
2117/// We want to transform:
2118///     %z = call i64 @llvm.cttz.i64(i64 %A, i1 false)
2119///
2120/// into:
2121///   entry:
2122///     %cmpz = icmp eq i64 %A, 0
2123///     br i1 %cmpz, label %cond.end, label %cond.false
2124///   cond.false:
2125///     %z = call i64 @llvm.cttz.i64(i64 %A, i1 true)
2126///     br label %cond.end
2127///   cond.end:
2128///     %ctz = phi i64 [ 64, %entry ], [ %z, %cond.false ]
2129///
2130/// If the transform is performed, return true and set ModifiedDT to true.
2131static bool despeculateCountZeros(IntrinsicInst *CountZeros,
                                const TargetLowering *TLI,
                                const DataLayout *DL, ModifyDT &ModifiedDT,
                                SmallSet<BasicBlock *, 32> &FreshBBs,
                                bool IsHugeFunc) {
// If a zero input is undefined, it doesn't make sense to despeculate that.
if (match(CountZeros->getOperand(1), m_One()))
  return false;

// If it's cheap to speculate, there's nothing to do.
Type *Ty = CountZeros->getType();
auto IntrinsicID = CountZeros->getIntrinsicID();
if ((IntrinsicID == Intrinsic::cttz && TLI->isCheapToSpeculateCttz(Ty)) ||
    (IntrinsicID == Intrinsic::ctlz && TLI->isCheapToSpeculateCtlz(Ty)))
  return false;

// Only handle legal scalar cases. Anything else requires too much work.
unsigned SizeInBits = Ty->getScalarSizeInBits();
if (Ty->isVectorTy() || SizeInBits > DL->getLargestLegalIntTypeSizeInBits())
  return false;

// Bail if the value is never zero.
Use &Op = CountZeros->getOperandUse(0);
if (isKnownNonZero(Op, *DL))
  return false;

// The intrinsic will be sunk behind a compare against zero and branch.
BasicBlock *StartBlock = CountZeros->getParent();
BasicBlock *CallBlock = StartBlock->splitBasicBlock(CountZeros, "cond.false");
if (IsHugeFunc)
  FreshBBs.insert(CallBlock);

// Create another block after the count zero intrinsic. A PHI will be added
// in this block to select the result of the intrinsic or the bit-width
// constant if the input to the intrinsic is zero.
BasicBlock::iterator SplitPt = ++(BasicBlock::iterator(CountZeros));
BasicBlock *EndBlock = CallBlock->splitBasicBlock(SplitPt, "cond.end");
if (IsHugeFunc)
  FreshBBs.insert(EndBlock);

// Set up a builder to create a compare, conditional branch, and PHI.
IRBuilder<> Builder(CountZeros->getContext());
Builder.SetInsertPoint(StartBlock->getTerminator());
Builder.SetCurrentDebugLocation(CountZeros->getDebugLoc());

// Replace the unconditional branch that was created by the first split with
// a compare against zero and a conditional branch.
Value *Zero = Constant::getNullValue(Ty);
// Avoid introducing branch on poison. This also replaces the ctz operand.
if (!isGuaranteedNotToBeUndefOrPoison(Op))
  Op = Builder.CreateFreeze(Op, Op->getName() + ".fr");
Value *Cmp = Builder.CreateICmpEQ(Op, Zero, "cmpz");
Builder.CreateCondBr(Cmp, EndBlock, CallBlock);
StartBlock->getTerminator()->eraseFromParent();

// Create a PHI in the end block to select either the output of the intrinsic
// or the bit width of the operand.
Builder.SetInsertPoint(&EndBlock->front());
PHINode *PN = Builder.CreatePHI(Ty, 2, "ctz");
replaceAllUsesWith(CountZeros, PN, FreshBBs, IsHugeFunc);
Value *BitWidth = Builder.getInt(APInt(SizeInBits, SizeInBits));
PN->addIncoming(BitWidth, StartBlock);
PN->addIncoming(CountZeros, CallBlock);

// We are explicitly handling the zero case, so we can set the intrinsic's
// undefined zero argument to 'true'. This will also prevent reprocessing the
// intrinsic; we only despeculate when a zero input is defined.
CountZeros->setArgOperand(1, Builder.getTrue());
ModifiedDT = ModifyDT::ModifyBBDT;
return true;
2201}

2203bool CodeGenPrepare::optimizeCallInst(CallInst *CI, ModifyDT &ModifiedDT) {
BasicBlock *BB = CI->getParent();

// Lower inline assembly if we can.
// If we found an inline asm expession, and if the target knows how to
// lower it to normal LLVM code, do so now.
if (CI->isInlineAsm()) {
  if (TLI->ExpandInlineAsm(CI)) {
    // Avoid invalidating the iterator.
    CurInstIterator = BB->begin();
    // Avoid processing instructions out of order, which could cause
    // reuse before a value is defined.
    SunkAddrs.clear();
    return true;
  }
  // Sink address computing for memory operands into the block.
  if (optimizeInlineAsmInst(CI))
    return true;
}

// Align the pointer arguments to this call if the target thinks it's a good
// idea
unsigned MinSize;
Align PrefAlign;
if (TLI->shouldAlignPointerArgs(CI, MinSize, PrefAlign)) {
  for (auto &Arg : CI->args()) {
    // We want to align both objects whose address is used directly and
    // objects whose address is used in casts and GEPs, though it only makes
    // sense for GEPs if the offset is a multiple of the desired alignment and
    // if size - offset meets the size threshold.
    if (!Arg->getType()->isPointerTy())
      continue;
    APInt Offset(DL->getIndexSizeInBits(
                     cast<PointerType>(Arg->getType())->getAddressSpace()),
                 0);
    Value *Val = Arg->stripAndAccumulateInBoundsConstantOffsets(*DL, Offset);
    uint64_t Offset2 = Offset.getLimitedValue();
    if (!isAligned(PrefAlign, Offset2))
      continue;
    AllocaInst *AI;
    if ((AI = dyn_cast<AllocaInst>(Val)) && AI->getAlign() < PrefAlign &&
        DL->getTypeAllocSize(AI->getAllocatedType()) >= MinSize + Offset2)
      AI->setAlignment(PrefAlign);
    // Global variables can only be aligned if they are defined in this
    // object (i.e. they are uniquely initialized in this object), and
    // over-aligning global variables that have an explicit section is
    // forbidden.
    GlobalVariable *GV;
    if ((GV = dyn_cast<GlobalVariable>(Val)) && GV->canIncreaseAlignment() &&
        GV->getPointerAlignment(*DL) < PrefAlign &&
        DL->getTypeAllocSize(GV->getValueType()) >= MinSize + Offset2)
      GV->setAlignment(PrefAlign);
  }
}
// If this is a memcpy (or similar) then we may be able to improve the
// alignment.
if (MemIntrinsic *MI = dyn_cast<MemIntrinsic>(CI)) {
  Align DestAlign = getKnownAlignment(MI->getDest(), *DL);
  MaybeAlign MIDestAlign = MI->getDestAlign();
  if (!MIDestAlign || DestAlign > *MIDestAlign)
    MI->setDestAlignment(DestAlign);
  if (MemTransferInst *MTI = dyn_cast<MemTransferInst>(MI)) {
    MaybeAlign MTISrcAlign = MTI->getSourceAlign();
    Align SrcAlign = getKnownAlignment(MTI->getSource(), *DL);
    if (!MTISrcAlign || SrcAlign > *MTISrcAlign)
      MTI->setSourceAlignment(SrcAlign);
  }
}

// If we have a cold call site, try to sink addressing computation into the
// cold block.  This interacts with our handling for loads and stores to
// ensure that we can fold all uses of a potential addressing computation
// into their uses.  TODO: generalize this to work over profiling data
if (CI->hasFnAttr(Attribute::Cold) && !OptSize &&
    !llvm::shouldOptimizeForSize(BB, PSI, BFI.get()))
  for (auto &Arg : CI->args()) {
    if (!Arg->getType()->isPointerTy())
      continue;
    unsigned AS = Arg->getType()->getPointerAddressSpace();
    if (optimizeMemoryInst(CI, Arg, Arg->getType(), AS))
      return true;
  }

IntrinsicInst *II = dyn_cast<IntrinsicInst>(CI);
if (II) {
  switch (II->getIntrinsicID()) {
  default:
    break;
  case Intrinsic::assume:
    llvm_unreachable("llvm.assume should have been removed already")::llvm::llvm_unreachable_internal("llvm.assume should have been removed already"
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 2292);
  case Intrinsic::experimental_widenable_condition: {
    // Give up on future widening oppurtunties so that we can fold away dead
    // paths and merge blocks before going into block-local instruction
    // selection.
    if (II->use_empty()) {
      II->eraseFromParent();
      return true;
    }
    Constant *RetVal = ConstantInt::getTrue(II->getContext());
    resetIteratorIfInvalidatedWhileCalling(BB, [&]() {
      replaceAndRecursivelySimplify(CI, RetVal, TLInfo, nullptr);
    });
    return true;
  }
  case Intrinsic::objectsize:
    llvm_unreachable("llvm.objectsize.* should have been lowered already")::llvm::llvm_unreachable_internal("llvm.objectsize.* should have been lowered already"
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 2308);
  case Intrinsic::is_constant:
    llvm_unreachable("llvm.is.constant.* should have been lowered already")::llvm::llvm_unreachable_internal("llvm.is.constant.* should have been lowered already"
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 2310);
  case Intrinsic::aarch64_stlxr:
  case Intrinsic::aarch64_stxr: {
    ZExtInst *ExtVal = dyn_cast<ZExtInst>(CI->getArgOperand(0));
    if (!ExtVal || !ExtVal->hasOneUse() ||
        ExtVal->getParent() == CI->getParent())
      return false;
    // Sink a zext feeding stlxr/stxr before it, so it can be folded into it.
    ExtVal->moveBefore(CI);
    // Mark this instruction as "inserted by CGP", so that other
    // optimizations don't touch it.
    InsertedInsts.insert(ExtVal);
    return true;
  }

  case Intrinsic::launder_invariant_group:
  case Intrinsic::strip_invariant_group: {
    Value *ArgVal = II->getArgOperand(0);
    auto it = LargeOffsetGEPMap.find(II);
    if (it != LargeOffsetGEPMap.end()) {
      // Merge entries in LargeOffsetGEPMap to reflect the RAUW.
      // Make sure not to have to deal with iterator invalidation
      // after possibly adding ArgVal to LargeOffsetGEPMap.
      auto GEPs = std::move(it->second);
      LargeOffsetGEPMap[ArgVal].append(GEPs.begin(), GEPs.end());
      LargeOffsetGEPMap.erase(II);
    }

    replaceAllUsesWith(II, ArgVal, FreshBBs, IsHugeFunc);
    II->eraseFromParent();
    return true;
  }
  case Intrinsic::cttz:
  case Intrinsic::ctlz:
    // If counting zeros is expensive, try to avoid it.
    return despeculateCountZeros(II, TLI, DL, ModifiedDT, FreshBBs,
                                 IsHugeFunc);
  case Intrinsic::fshl:
  case Intrinsic::fshr:
    return optimizeFunnelShift(II);
  case Intrinsic::dbg_assign:
  case Intrinsic::dbg_value:
    return fixupDbgValue(II);
  case Intrinsic::masked_gather:
    return optimizeGatherScatterInst(II, II->getArgOperand(0));
  case Intrinsic::masked_scatter:
    return optimizeGatherScatterInst(II, II->getArgOperand(1));
  }

  SmallVector<Value *, 2> PtrOps;
  Type *AccessTy;
  if (TLI->getAddrModeArguments(II, PtrOps, AccessTy))
    while (!PtrOps.empty()) {
      Value *PtrVal = PtrOps.pop_back_val();
      unsigned AS = PtrVal->getType()->getPointerAddressSpace();
      if (optimizeMemoryInst(II, PtrVal, AccessTy, AS))
        return true;
    }
}

// From here on out we're working with named functions.
if (!CI->getCalledFunction())
  return false;

// Lower all default uses of _chk calls.  This is very similar
// to what InstCombineCalls does, but here we are only lowering calls
// to fortified library functions (e.g. __memcpy_chk) that have the default
// "don't know" as the objectsize.  Anything else should be left alone.
FortifiedLibCallSimplifier Simplifier(TLInfo, true);
IRBuilder<> Builder(CI);
if (Value *V = Simplifier.optimizeCall(CI, Builder)) {
  replaceAllUsesWith(CI, V, FreshBBs, IsHugeFunc);
  CI->eraseFromParent();
  return true;
}

return false;
2387}

2389/// Look for opportunities to duplicate return instructions to the predecessor
2390/// to enable tail call optimizations. The case it is currently looking for is:
2391/// @code
2392/// bb0:
2393///   %tmp0 = tail call i32 @f0()
2394///   br label %return
2395/// bb1:
2396///   %tmp1 = tail call i32 @f1()
2397///   br label %return
2398/// bb2:
2399///   %tmp2 = tail call i32 @f2()
2400///   br label %return
2401/// return:
2402///   %retval = phi i32 [ %tmp0, %bb0 ], [ %tmp1, %bb1 ], [ %tmp2, %bb2 ]
2403///   ret i32 %retval
2404/// @endcode
2405///
2406/// =>
2407///
2408/// @code
2409/// bb0:
2410///   %tmp0 = tail call i32 @f0()
2411///   ret i32 %tmp0
2412/// bb1:
2413///   %tmp1 = tail call i32 @f1()
2414///   ret i32 %tmp1
2415/// bb2:
2416///   %tmp2 = tail call i32 @f2()
2417///   ret i32 %tmp2
2418/// @endcode
2419bool CodeGenPrepare::dupRetToEnableTailCallOpts(BasicBlock *BB,
                                              ModifyDT &ModifiedDT) {
if (!BB->getTerminator())
1
Assuming the condition is false→
2
←
Taking false branch→
  return false;

ReturnInst *RetI = dyn_cast<ReturnInst>(BB->getTerminator());
3
←
Assuming the object is a 'CastReturnType'→
if (!RetI3.1
'RetI' is non-null
1
'RetI' is non-null
)
4
←
Taking false branch→
  return false;

PHINode *PN = nullptr;
ExtractValueInst *EVI = nullptr;
BitCastInst *BCI = nullptr;
Value *V = RetI->getReturnValue();
5
←
Calling 'ReturnInst::getReturnValue'→
9
←
Returning from 'ReturnInst::getReturnValue'→
if (V9.1
'V' is null
1
'V' is null
) {
  BCI = dyn_cast<BitCastInst>(V);
  if (BCI)
    V = BCI->getOperand(0);

  EVI = dyn_cast<ExtractValueInst>(V);
  if (EVI) {
    V = EVI->getOperand(0);
    if (!llvm::all_of(EVI->indices(), [](unsigned idx) { return idx == 0; }))
      return false;
  }

  PN = dyn_cast<PHINode>(V);
  if (!PN)
    return false;
}

if (PN9.2
'PN' is null
2
'PN' is null
 && PN->getParent() != BB)
  return false;

auto isLifetimeEndOrBitCastFor = [](const Instruction *Inst) {
  const BitCastInst *BC = dyn_cast<BitCastInst>(Inst);
  if (BC && BC->hasOneUse())
    Inst = BC->user_back();

  if (const IntrinsicInst *II = dyn_cast<IntrinsicInst>(Inst))
    return II->getIntrinsicID() == Intrinsic::lifetime_end;
  return false;
};

// Make sure there are no instructions between the first instruction
// and return.
const Instruction *BI = BB->getFirstNonPHI();
10
←
'BI' initialized here→
// Skip over debug and the bitcast.
while (isa<DbgInfoIntrinsic>(BI) || BI == BCI || BI == EVI ||
11
←
Assuming 'BI' is not a 'DbgInfoIntrinsic'→
12
←
Assuming 'BI' is equal to 'BCI'→
       isa<PseudoProbeInst>(BI) || isLifetimeEndOrBitCastFor(BI))
  BI = BI->getNextNode();
13
←
Called C++ object pointer is null
if (BI != RetI)
  return false;

/// Only dup the ReturnInst if the CallInst is likely to be emitted as a tail
/// call.
const Function *F = BB->getParent();
SmallVector<BasicBlock *, 4> TailCallBBs;
if (PN) {
  for (unsigned I = 0, E = PN->getNumIncomingValues(); I != E; ++I) {
    // Look through bitcasts.
    Value *IncomingVal = PN->getIncomingValue(I)->stripPointerCasts();
    CallInst *CI = dyn_cast<CallInst>(IncomingVal);
    BasicBlock *PredBB = PN->getIncomingBlock(I);
    // Make sure the phi value is indeed produced by the tail call.
    if (CI && CI->hasOneUse() && CI->getParent() == PredBB &&
        TLI->mayBeEmittedAsTailCall(CI) &&
        attributesPermitTailCall(F, CI, RetI, *TLI))
      TailCallBBs.push_back(PredBB);
  }
} else {
  SmallPtrSet<BasicBlock *, 4> VisitedBBs;
  for (BasicBlock *Pred : predecessors(BB)) {
    if (!VisitedBBs.insert(Pred).second)
      continue;
    if (Instruction *I = Pred->rbegin()->getPrevNonDebugInstruction(true)) {
      CallInst *CI = dyn_cast<CallInst>(I);
      if (CI && CI->use_empty() && TLI->mayBeEmittedAsTailCall(CI) &&
          attributesPermitTailCall(F, CI, RetI, *TLI))
        TailCallBBs.push_back(Pred);
    }
  }
}

bool Changed = false;
for (auto const &TailCallBB : TailCallBBs) {
  // Make sure the call instruction is followed by an unconditional branch to
  // the return block.
  BranchInst *BI = dyn_cast<BranchInst>(TailCallBB->getTerminator());
  if (!BI || !BI->isUnconditional() || BI->getSuccessor(0) != BB)
    continue;

  // Duplicate the return into TailCallBB.
  (void)FoldReturnIntoUncondBranch(RetI, BB, TailCallBB);
  assert(!VerifyBFIUpdates ||(static_cast <bool> (!VerifyBFIUpdates || BFI->getBlockFreq
(BB) >= BFI->getBlockFreq(TailCallBB)) ? void (0) : __assert_fail
 ("!VerifyBFIUpdates || BFI->getBlockFreq(BB) >= BFI->getBlockFreq(TailCallBB)"
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 2513, __extension__ __PRETTY_FUNCTION__
))
         BFI->getBlockFreq(BB) >= BFI->getBlockFreq(TailCallBB))(static_cast <bool> (!VerifyBFIUpdates || BFI->getBlockFreq
(BB) >= BFI->getBlockFreq(TailCallBB)) ? void (0) : __assert_fail
 ("!VerifyBFIUpdates || BFI->getBlockFreq(BB) >= BFI->getBlockFreq(TailCallBB)"
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 2513, __extension__ __PRETTY_FUNCTION__
));
  BFI->setBlockFreq(
      BB,
      (BFI->getBlockFreq(BB) - BFI->getBlockFreq(TailCallBB)).getFrequency());
  ModifiedDT = ModifyDT::ModifyBBDT;
  Changed = true;
  ++NumRetsDup;
}

// If we eliminated all predecessors of the block, delete the block now.
if (Changed && !BB->hasAddressTaken() && pred_empty(BB))
  BB->eraseFromParent();

return Changed;
2527}

2529//===----------------------------------------------------------------------===//
2530// Memory Optimization
2531//===----------------------------------------------------------------------===//

2533namespace {

2535/// This is an extended version of TargetLowering::AddrMode
2536/// which holds actual Value*'s for register values.
2537struct ExtAddrMode : public TargetLowering::AddrMode {
Value *BaseReg = nullptr;
Value *ScaledReg = nullptr;
Value *OriginalValue = nullptr;
bool InBounds = true;

enum FieldName {
  NoField = 0x00,
  BaseRegField = 0x01,
  BaseGVField = 0x02,
  BaseOffsField = 0x04,
  ScaledRegField = 0x08,
  ScaleField = 0x10,
  MultipleFields = 0xff
};

ExtAddrMode() = default;

void print(raw_ostream &OS) const;
void dump() const;

FieldName compare(const ExtAddrMode &other) {
  // First check that the types are the same on each field, as differing types
  // is something we can't cope with later on.
  if (BaseReg && other.BaseReg &&
      BaseReg->getType() != other.BaseReg->getType())
    return MultipleFields;
  if (BaseGV && other.BaseGV && BaseGV->getType() != other.BaseGV->getType())
    return MultipleFields;
  if (ScaledReg && other.ScaledReg &&
      ScaledReg->getType() != other.ScaledReg->getType())
    return MultipleFields;

  // Conservatively reject 'inbounds' mismatches.
  if (InBounds != other.InBounds)
    return MultipleFields;

  // Check each field to see if it differs.
  unsigned Result = NoField;
  if (BaseReg != other.BaseReg)
    Result |= BaseRegField;
  if (BaseGV != other.BaseGV)
    Result |= BaseGVField;
  if (BaseOffs != other.BaseOffs)
    Result |= BaseOffsField;
  if (ScaledReg != other.ScaledReg)
    Result |= ScaledRegField;
  // Don't count 0 as being a different scale, because that actually means
  // unscaled (which will already be counted by having no ScaledReg).
  if (Scale && other.Scale && Scale != other.Scale)
    Result |= ScaleField;

  if (llvm::popcount(Result) > 1)
    return MultipleFields;
  else
    return static_cast<FieldName>(Result);
}

// An AddrMode is trivial if it involves no calculation i.e. it is just a base
// with no offset.
bool isTrivial() {
  // An AddrMode is (BaseGV + BaseReg + BaseOffs + ScaleReg * Scale) so it is
  // trivial if at most one of these terms is nonzero, except that BaseGV and
  // BaseReg both being zero actually means a null pointer value, which we
  // consider to be 'non-zero' here.
  return !BaseOffs && !Scale && !(BaseGV && BaseReg);
}

Value *GetFieldAsValue(FieldName Field, Type *IntPtrTy) {
  switch (Field) {
  default:
    return nullptr;
  case BaseRegField:
    return BaseReg;
  case BaseGVField:
    return BaseGV;
  case ScaledRegField:
    return ScaledReg;
  case BaseOffsField:
    return ConstantInt::get(IntPtrTy, BaseOffs);
  }
}

void SetCombinedField(FieldName Field, Value *V,
                      const SmallVectorImpl<ExtAddrMode> &AddrModes) {
  switch (Field) {
  default:
    llvm_unreachable("Unhandled fields are expected to be rejected earlier")::llvm::llvm_unreachable_internal("Unhandled fields are expected to be rejected earlier"
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 2624);
    break;
  case ExtAddrMode::BaseRegField:
    BaseReg = V;
    break;
  case ExtAddrMode::BaseGVField:
    // A combined BaseGV is an Instruction, not a GlobalValue, so it goes
    // in the BaseReg field.
    assert(BaseReg == nullptr)(static_cast <bool> (BaseReg == nullptr) ? void (0) : __assert_fail
 ("BaseReg == nullptr", "llvm/lib/CodeGen/CodeGenPrepare.cpp"
, 2632, __extension__ __PRETTY_FUNCTION__));
    BaseReg = V;
    BaseGV = nullptr;
    break;
  case ExtAddrMode::ScaledRegField:
    ScaledReg = V;
    // If we have a mix of scaled and unscaled addrmodes then we want scale
    // to be the scale and not zero.
    if (!Scale)
      for (const ExtAddrMode &AM : AddrModes)
        if (AM.Scale) {
          Scale = AM.Scale;
          break;
        }
    break;
  case ExtAddrMode::BaseOffsField:
    // The offset is no longer a constant, so it goes in ScaledReg with a
    // scale of 1.
    assert(ScaledReg == nullptr)(static_cast <bool> (ScaledReg == nullptr) ? void (0) :
 __assert_fail ("ScaledReg == nullptr", "llvm/lib/CodeGen/CodeGenPrepare.cpp"
, 2650, __extension__ __PRETTY_FUNCTION__));
    ScaledReg = V;
    Scale = 1;
    BaseOffs = 0;
    break;
  }
}
2657};

2659#ifndef NDEBUG
2660static inline raw_ostream &operator<<(raw_ostream &OS, const ExtAddrMode &AM) {
AM.print(OS);
return OS;
2663}
2664#endif

2666#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
2667void ExtAddrMode::print(raw_ostream &OS) const {
bool NeedPlus = false;
OS << "[";
if (InBounds)
  OS << "inbounds ";
if (BaseGV) {
  OS << "GV:";
  BaseGV->printAsOperand(OS, /*PrintType=*/false);
  NeedPlus = true;
}

if (BaseOffs) {
  OS << (NeedPlus ? " + " : "") << BaseOffs;
  NeedPlus = true;
}

if (BaseReg) {
  OS << (NeedPlus ? " + " : "") << "Base:";
  BaseReg->printAsOperand(OS, /*PrintType=*/false);
  NeedPlus = true;
}
if (Scale) {
  OS << (NeedPlus ? " + " : "") << Scale << "*";
  ScaledReg->printAsOperand(OS, /*PrintType=*/false);
}

OS << ']';
2694}

2696LLVM_DUMP_METHOD__attribute__((noinline)) __attribute__((__used__)) void ExtAddrMode::dump() const {
print(dbgs());
dbgs() << '\n';
2699}
2700#endif

2702} // end anonymous namespace

2704namespace {

2706/// This class provides transaction based operation on the IR.
2707/// Every change made through this class is recorded in the internal state and
2708/// can be undone (rollback) until commit is called.
2709/// CGP does not check if instructions could be speculatively executed when
2710/// moved. Preserving the original location would pessimize the debugging
2711/// experience, as well as negatively impact the quality of sample PGO.
2712class TypePromotionTransaction {
/// This represents the common interface of the individual transaction.
/// Each class implements the logic for doing one specific modification on
/// the IR via the TypePromotionTransaction.
class TypePromotionAction {
protected:
  /// The Instruction modified.
  Instruction *Inst;

public:
  /// Constructor of the action.
  /// The constructor performs the related action on the IR.
  TypePromotionAction(Instruction *Inst) : Inst(Inst) {}

  virtual ~TypePromotionAction() = default;

  /// Undo the modification done by this action.
  /// When this method is called, the IR must be in the same state as it was
  /// before this action was applied.
  /// \pre Undoing the action works if and only if the IR is in the exact same
  /// state as it was directly after this action was applied.
  virtual void undo() = 0;

  /// Advocate every change made by this action.
  /// When the results on the IR of the action are to be kept, it is important
  /// to call this function, otherwise hidden information may be kept forever.
  virtual void commit() {
    // Nothing to be done, this action is not doing anything.
  }
};

/// Utility to remember the position of an instruction.
class InsertionHandler {
  /// Position of an instruction.
  /// Either an instruction:
  /// - Is the first in a basic block: BB is used.
  /// - Has a previous instruction: PrevInst is used.
  union {
    Instruction *PrevInst;
    BasicBlock *BB;
  } Point;

  /// Remember whether or not the instruction had a previous instruction.
  bool HasPrevInstruction;

public:
  /// Record the position of \p Inst.
  InsertionHandler(Instruction *Inst) {
    BasicBlock::iterator It = Inst->getIterator();
    HasPrevInstruction = (It != (Inst->getParent()->begin()));
    if (HasPrevInstruction)
      Point.PrevInst = &*--It;
    else
      Point.BB = Inst->getParent();
  }

  /// Insert \p Inst at the recorded position.
  void insert(Instruction *Inst) {
    if (HasPrevInstruction) {
      if (Inst->getParent())
        Inst->removeFromParent();
      Inst->insertAfter(Point.PrevInst);
    } else {
      Instruction *Position = &*Point.BB->getFirstInsertionPt();
      if (Inst->getParent())
        Inst->moveBefore(Position);
      else
        Inst->insertBefore(Position);
    }
  }
};

/// Move an instruction before another.
class InstructionMoveBefore : public TypePromotionAction {
  /// Original position of the instruction.
  InsertionHandler Position;

public:
  /// Move \p Inst before \p Before.
  InstructionMoveBefore(Instruction *Inst, Instruction *Before)
      : TypePromotionAction(Inst), Position(Inst) {
    LLVM_DEBUG(dbgs() << "Do: move: " << *Inst << "\nbefore: " << *Beforedo { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: move: " << *
Inst << "\nbefore: " << *Before << "\n"; } }
 while (false)
                      << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: move: " << *
Inst << "\nbefore: " << *Before << "\n"; } }
 while (false);
    Inst->moveBefore(Before);
  }

  /// Move the instruction back to its original position.
  void undo() override {
    LLVM_DEBUG(dbgs() << "Undo: moveBefore: " << *Inst << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Undo: moveBefore: " <<
 *Inst << "\n"; } } while (false);
    Position.insert(Inst);
  }
};

/// Set the operand of an instruction with a new value.
class OperandSetter : public TypePromotionAction {
  /// Original operand of the instruction.
  Value *Origin;

  /// Index of the modified instruction.
  unsigned Idx;

public:
  /// Set \p Idx operand of \p Inst with \p NewVal.
  OperandSetter(Instruction *Inst, unsigned Idx, Value *NewVal)
      : TypePromotionAction(Inst), Idx(Idx) {
    LLVM_DEBUG(dbgs() << "Do: setOperand: " << Idx << "\n"do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: setOperand: " <<
 Idx << "\n" << "for:" << *Inst << "\n"
 << "with:" << *NewVal << "\n"; } } while (
false)
                      << "for:" << *Inst << "\n"do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: setOperand: " <<
 Idx << "\n" << "for:" << *Inst << "\n"
 << "with:" << *NewVal << "\n"; } } while (
false)
                      << "with:" << *NewVal << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: setOperand: " <<
 Idx << "\n" << "for:" << *Inst << "\n"
 << "with:" << *NewVal << "\n"; } } while (
false);
    Origin = Inst->getOperand(Idx);
    Inst->setOperand(Idx, NewVal);
  }

  /// Restore the original value of the instruction.
  void undo() override {
    LLVM_DEBUG(dbgs() << "Undo: setOperand:" << Idx << "\n"do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Undo: setOperand:" <<
 Idx << "\n" << "for: " << *Inst << "\n"
 << "with: " << *Origin << "\n"; } } while (
false)
                      << "for: " << *Inst << "\n"do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Undo: setOperand:" <<
 Idx << "\n" << "for: " << *Inst << "\n"
 << "with: " << *Origin << "\n"; } } while (
false)
                      << "with: " << *Origin << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Undo: setOperand:" <<
 Idx << "\n" << "for: " << *Inst << "\n"
 << "with: " << *Origin << "\n"; } } while (
false);
    Inst->setOperand(Idx, Origin);
  }
};

/// Hide the operands of an instruction.
/// Do as if this instruction was not using any of its operands.
class OperandsHider : public TypePromotionAction {
  /// The list of original operands.
  SmallVector<Value *, 4> OriginalValues;

public:
  /// Remove \p Inst from the uses of the operands of \p Inst.
  OperandsHider(Instruction *Inst) : TypePromotionAction(Inst) {
    LLVM_DEBUG(dbgs() << "Do: OperandsHider: " << *Inst << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: OperandsHider: " <<
 *Inst << "\n"; } } while (false);
    unsigned NumOpnds = Inst->getNumOperands();
    OriginalValues.reserve(NumOpnds);
    for (unsigned It = 0; It < NumOpnds; ++It) {
      // Save the current operand.
      Value *Val = Inst->getOperand(It);
      OriginalValues.push_back(Val);
      // Set a dummy one.
      // We could use OperandSetter here, but that would imply an overhead
      // that we are not willing to pay.
      Inst->setOperand(It, UndefValue::get(Val->getType()));
    }
  }

  /// Restore the original list of uses.
  void undo() override {
    LLVM_DEBUG(dbgs() << "Undo: OperandsHider: " << *Inst << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Undo: OperandsHider: "
 << *Inst << "\n"; } } while (false);
    for (unsigned It = 0, EndIt = OriginalValues.size(); It != EndIt; ++It)
      Inst->setOperand(It, OriginalValues[It]);
  }
};

/// Build a truncate instruction.
class TruncBuilder : public TypePromotionAction {
  Value *Val;

public:
  /// Build a truncate instruction of \p Opnd producing a \p Ty
  /// result.
  /// trunc Opnd to Ty.
  TruncBuilder(Instruction *Opnd, Type *Ty) : TypePromotionAction(Opnd) {
    IRBuilder<> Builder(Opnd);
    Builder.SetCurrentDebugLocation(DebugLoc());
    Val = Builder.CreateTrunc(Opnd, Ty, "promoted");
    LLVM_DEBUG(dbgs() << "Do: TruncBuilder: " << *Val << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: TruncBuilder: " <<
 *Val << "\n"; } } while (false);
  }

  /// Get the built value.
  Value *getBuiltValue() { return Val; }

  /// Remove the built instruction.
  void undo() override {
    LLVM_DEBUG(dbgs() << "Undo: TruncBuilder: " << *Val << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Undo: TruncBuilder: " <<
 *Val << "\n"; } } while (false);
    if (Instruction *IVal = dyn_cast<Instruction>(Val))
      IVal->eraseFromParent();
  }
};

/// Build a sign extension instruction.
class SExtBuilder : public TypePromotionAction {
  Value *Val;

public:
  /// Build a sign extension instruction of \p Opnd producing a \p Ty
  /// result.
  /// sext Opnd to Ty.
  SExtBuilder(Instruction *InsertPt, Value *Opnd, Type *Ty)
      : TypePromotionAction(InsertPt) {
    IRBuilder<> Builder(InsertPt);
    Val = Builder.CreateSExt(Opnd, Ty, "promoted");
    LLVM_DEBUG(dbgs() << "Do: SExtBuilder: " << *Val << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: SExtBuilder: " <<
 *Val << "\n"; } } while (false);
  }

  /// Get the built value.
  Value *getBuiltValue() { return Val; }

  /// Remove the built instruction.
  void undo() override {
    LLVM_DEBUG(dbgs() << "Undo: SExtBuilder: " << *Val << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Undo: SExtBuilder: " <<
 *Val << "\n"; } } while (false);
    if (Instruction *IVal = dyn_cast<Instruction>(Val))
      IVal->eraseFromParent();
  }
};

/// Build a zero extension instruction.
class ZExtBuilder : public TypePromotionAction {
  Value *Val;

public:
  /// Build a zero extension instruction of \p Opnd producing a \p Ty
  /// result.
  /// zext Opnd to Ty.
  ZExtBuilder(Instruction *InsertPt, Value *Opnd, Type *Ty)
      : TypePromotionAction(InsertPt) {
    IRBuilder<> Builder(InsertPt);
    Builder.SetCurrentDebugLocation(DebugLoc());
    Val = Builder.CreateZExt(Opnd, Ty, "promoted");
    LLVM_DEBUG(dbgs() << "Do: ZExtBuilder: " << *Val << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: ZExtBuilder: " <<
 *Val << "\n"; } } while (false);
  }

  /// Get the built value.
  Value *getBuiltValue() { return Val; }

  /// Remove the built instruction.
  void undo() override {
    LLVM_DEBUG(dbgs() << "Undo: ZExtBuilder: " << *Val << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Undo: ZExtBuilder: " <<
 *Val << "\n"; } } while (false);
    if (Instruction *IVal = dyn_cast<Instruction>(Val))
      IVal->eraseFromParent();
  }
};

/// Mutate an instruction to another type.
class TypeMutator : public TypePromotionAction {
  /// Record the original type.
  Type *OrigTy;

public:
  /// Mutate the type of \p Inst into \p NewTy.
  TypeMutator(Instruction *Inst, Type *NewTy)
      : TypePromotionAction(Inst), OrigTy(Inst->getType()) {
    LLVM_DEBUG(dbgs() << "Do: MutateType: " << *Inst << " with " << *NewTydo { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: MutateType: " <<
 *Inst << " with " << *NewTy << "\n"; } } while
 (false)
                      << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: MutateType: " <<
 *Inst << " with " << *NewTy << "\n"; } } while
 (false);
    Inst->mutateType(NewTy);
  }

  /// Mutate the instruction back to its original type.
  void undo() override {
    LLVM_DEBUG(dbgs() << "Undo: MutateType: " << *Inst << " with " << *OrigTydo { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Undo: MutateType: " <<
 *Inst << " with " << *OrigTy << "\n"; } } while
 (false)
                      << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Undo: MutateType: " <<
 *Inst << " with " << *OrigTy << "\n"; } } while
 (false);
    Inst->mutateType(OrigTy);
  }
};

/// Replace the uses of an instruction by another instruction.
class UsesReplacer : public TypePromotionAction {
  /// Helper structure to keep track of the replaced uses.
  struct InstructionAndIdx {
    /// The instruction using the instruction.
    Instruction *Inst;

    /// The index where this instruction is used for Inst.
    unsigned Idx;

    InstructionAndIdx(Instruction *Inst, unsigned Idx)
        : Inst(Inst), Idx(Idx) {}
  };

  /// Keep track of the original uses (pair Instruction, Index).
  SmallVector<InstructionAndIdx, 4> OriginalUses;
  /// Keep track of the debug users.
  SmallVector<DbgValueInst *, 1> DbgValues;

  /// Keep track of the new value so that we can undo it by replacing
  /// instances of the new value with the original value.
  Value *New;

  using use_iterator = SmallVectorImpl<InstructionAndIdx>::iterator;

public:
  /// Replace all the use of \p Inst by \p New.
  UsesReplacer(Instruction *Inst, Value *New)
      : TypePromotionAction(Inst), New(New) {
    LLVM_DEBUG(dbgs() << "Do: UsersReplacer: " << *Inst << " with " << *Newdo { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: UsersReplacer: " <<
 *Inst << " with " << *New << "\n"; } } while
 (false)
                      << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: UsersReplacer: " <<
 *Inst << " with " << *New << "\n"; } } while
 (false);
    // Record the original uses.
    for (Use &U : Inst->uses()) {
      Instruction *UserI = cast<Instruction>(U.getUser());
      OriginalUses.push_back(InstructionAndIdx(UserI, U.getOperandNo()));
    }
    // Record the debug uses separately. They are not in the instruction's
    // use list, but they are replaced by RAUW.
    findDbgValues(DbgValues, Inst);

    // Now, we can replace the uses.
    Inst->replaceAllUsesWith(New);
  }

  /// Reassign the original uses of Inst to Inst.
  void undo() override {
    LLVM_DEBUG(dbgs() << "Undo: UsersReplacer: " << *Inst << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Undo: UsersReplacer: "
 << *Inst << "\n"; } } while (false);
    for (InstructionAndIdx &Use : OriginalUses)
      Use.Inst->setOperand(Use.Idx, Inst);
    // RAUW has replaced all original uses with references to the new value,
    // including the debug uses. Since we are undoing the replacements,
    // the original debug uses must also be reinstated to maintain the
    // correctness and utility of debug value instructions.
    for (auto *DVI : DbgValues)
      DVI->replaceVariableLocationOp(New, Inst);
  }
};

/// Remove an instruction from the IR.
class InstructionRemover : public TypePromotionAction {
  /// Original position of the instruction.
  InsertionHandler Inserter;

  /// Helper structure to hide all the link to the instruction. In other
  /// words, this helps to do as if the instruction was removed.
  OperandsHider Hider;

  /// Keep track of the uses replaced, if any.
  UsesReplacer *Replacer = nullptr;

  /// Keep track of instructions removed.
  SetOfInstrs &RemovedInsts;

public:
  /// Remove all reference of \p Inst and optionally replace all its
  /// uses with New.
  /// \p RemovedInsts Keep track of the instructions removed by this Action.
  /// \pre If !Inst->use_empty(), then New != nullptr
  InstructionRemover(Instruction *Inst, SetOfInstrs &RemovedInsts,
                     Value *New = nullptr)
      : TypePromotionAction(Inst), Inserter(Inst), Hider(Inst),
        RemovedInsts(RemovedInsts) {
    if (New)
      Replacer = new UsesReplacer(Inst, New);
    LLVM_DEBUG(dbgs() << "Do: InstructionRemover: " << *Inst << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Do: InstructionRemover: "
 << *Inst << "\n"; } } while (false);
    RemovedInsts.insert(Inst);
    /// The instructions removed here will be freed after completing
    /// optimizeBlock() for all blocks as we need to keep track of the
    /// removed instructions during promotion.
    Inst->removeFromParent();
  }

  ~InstructionRemover() override { delete Replacer; }

  /// Resurrect the instruction and reassign it to the proper uses if
  /// new value was provided when build this action.
  void undo() override {
    LLVM_DEBUG(dbgs() << "Undo: InstructionRemover: " << *Inst << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Undo: InstructionRemover: "
 << *Inst << "\n"; } } while (false);
    Inserter.insert(Inst);
    if (Replacer)
      Replacer->undo();
    Hider.undo();
    RemovedInsts.erase(Inst);
  }
};

3071public:
/// Restoration point.
/// The restoration point is a pointer to an action instead of an iterator
/// because the iterator may be invalidated but not the pointer.
using ConstRestorationPt = const TypePromotionAction *;

TypePromotionTransaction(SetOfInstrs &RemovedInsts)
    : RemovedInsts(RemovedInsts) {}

/// Advocate every changes made in that transaction. Return true if any change
/// happen.
bool commit();

/// Undo all the changes made after the given point.
void rollback(ConstRestorationPt Point);

/// Get the current restoration point.
ConstRestorationPt getRestorationPoint() const;

/// \name API for IR modification with state keeping to support rollback.
/// @{
/// Same as Instruction::setOperand.
void setOperand(Instruction *Inst, unsigned Idx, Value *NewVal);

/// Same as Instruction::eraseFromParent.
void eraseInstruction(Instruction *Inst, Value *NewVal = nullptr);

/// Same as Value::replaceAllUsesWith.
void replaceAllUsesWith(Instruction *Inst, Value *New);

/// Same as Value::mutateType.
void mutateType(Instruction *Inst, Type *NewTy);

/// Same as IRBuilder::createTrunc.
Value *createTrunc(Instruction *Opnd, Type *Ty);

/// Same as IRBuilder::createSExt.
Value *createSExt(Instruction *Inst, Value *Opnd, Type *Ty);

/// Same as IRBuilder::createZExt.
Value *createZExt(Instruction *Inst, Value *Opnd, Type *Ty);

/// Same as Instruction::moveBefore.
void moveBefore(Instruction *Inst, Instruction *Before);
/// @}

3117private:
/// The ordered list of actions made so far.
SmallVector<std::unique_ptr<TypePromotionAction>, 16> Actions;

using CommitPt =
    SmallVectorImpl<std::unique_ptr<TypePromotionAction>>::iterator;

SetOfInstrs &RemovedInsts;
3125};

3127} // end anonymous namespace

3129void TypePromotionTransaction::setOperand(Instruction *Inst, unsigned Idx,
                                        Value *NewVal) {
Actions.push_back(std::make_unique<TypePromotionTransaction::OperandSetter>(
    Inst, Idx, NewVal));
3133}

3135void TypePromotionTransaction::eraseInstruction(Instruction *Inst,
                                              Value *NewVal) {
Actions.push_back(
    std::make_unique<TypePromotionTransaction::InstructionRemover>(
        Inst, RemovedInsts, NewVal));
3140}

3142void TypePromotionTransaction::replaceAllUsesWith(Instruction *Inst,
                                                Value *New) {
Actions.push_back(
    std::make_unique<TypePromotionTransaction::UsesReplacer>(Inst, New));
3146}

3148void TypePromotionTransaction::mutateType(Instruction *Inst, Type *NewTy) {
Actions.push_back(
    std::make_unique<TypePromotionTransaction::TypeMutator>(Inst, NewTy));
3151}

3153Value *TypePromotionTransaction::createTrunc(Instruction *Opnd, Type *Ty) {
std::unique_ptr<TruncBuilder> Ptr(new TruncBuilder(Opnd, Ty));
Value *Val = Ptr->getBuiltValue();
Actions.push_back(std::move(Ptr));
return Val;
3158}

3160Value *TypePromotionTransaction::createSExt(Instruction *Inst, Value *Opnd,
                                          Type *Ty) {
std::unique_ptr<SExtBuilder> Ptr(new SExtBuilder(Inst, Opnd, Ty));
Value *Val = Ptr->getBuiltValue();
Actions.push_back(std::move(Ptr));
return Val;
3166}

3168Value *TypePromotionTransaction::createZExt(Instruction *Inst, Value *Opnd,
                                          Type *Ty) {
std::unique_ptr<ZExtBuilder> Ptr(new ZExtBuilder(Inst, Opnd, Ty));
Value *Val = Ptr->getBuiltValue();
Actions.push_back(std::move(Ptr));
return Val;
3174}

3176void TypePromotionTransaction::moveBefore(Instruction *Inst,
                                        Instruction *Before) {
Actions.push_back(
    std::make_unique<TypePromotionTransaction::InstructionMoveBefore>(
        Inst, Before));
3181}

3183TypePromotionTransaction::ConstRestorationPt
3184TypePromotionTransaction::getRestorationPoint() const {
return !Actions.empty() ? Actions.back().get() : nullptr;
3186}

3188bool TypePromotionTransaction::commit() {
for (std::unique_ptr<TypePromotionAction> &Action : Actions)
  Action->commit();
bool Modified = !Actions.empty();
Actions.clear();
return Modified;
3194}

3196void TypePromotionTransaction::rollback(
  TypePromotionTransaction::ConstRestorationPt Point) {
while (!Actions.empty() && Point != Actions.back().get()) {
  std::unique_ptr<TypePromotionAction> Curr = Actions.pop_back_val();
  Curr->undo();
}
3202}

3204namespace {

3206/// A helper class for matching addressing modes.
3207///
3208/// This encapsulates the logic for matching the target-legal addressing modes.
3209class AddressingModeMatcher {
SmallVectorImpl<Instruction *> &AddrModeInsts;
const TargetLowering &TLI;
const TargetRegisterInfo &TRI;
const DataLayout &DL;
const LoopInfo &LI;
const std::function<const DominatorTree &()> getDTFn;

/// AccessTy/MemoryInst - This is the type for the access (e.g. double) and
/// the memory instruction that we're computing this address for.
Type *AccessTy;
unsigned AddrSpace;
Instruction *MemoryInst;

/// This is the addressing mode that we're building up. This is
/// part of the return value of this addressing mode matching stuff.
ExtAddrMode &AddrMode;

/// The instructions inserted by other CodeGenPrepare optimizations.
const SetOfInstrs &InsertedInsts;

/// A map from the instructions to their type before promotion.
InstrToOrigTy &PromotedInsts;

/// The ongoing transaction where every action should be registered.
TypePromotionTransaction &TPT;

// A GEP which has too large offset to be folded into the addressing mode.
std::pair<AssertingVH<GetElementPtrInst>, int64_t> &LargeOffsetGEP;

/// This is set to true when we should not do profitability checks.
/// When true, IsProfitableToFoldIntoAddressingMode always returns true.
bool IgnoreProfitability;

/// True if we are optimizing for size.
bool OptSize;

ProfileSummaryInfo *PSI;
BlockFrequencyInfo *BFI;

AddressingModeMatcher(
    SmallVectorImpl<Instruction *> &AMI, const TargetLowering &TLI,
    const TargetRegisterInfo &TRI, const LoopInfo &LI,
    const std::function<const DominatorTree &()> getDTFn, Type *AT,
    unsigned AS, Instruction *MI, ExtAddrMode &AM,
    const SetOfInstrs &InsertedInsts, InstrToOrigTy &PromotedInsts,
    TypePromotionTransaction &TPT,
    std::pair<AssertingVH<GetElementPtrInst>, int64_t> &LargeOffsetGEP,
    bool OptSize, ProfileSummaryInfo *PSI, BlockFrequencyInfo *BFI)
    : AddrModeInsts(AMI), TLI(TLI), TRI(TRI),
      DL(MI->getModule()->getDataLayout()), LI(LI), getDTFn(getDTFn),
      AccessTy(AT), AddrSpace(AS), MemoryInst(MI), AddrMode(AM),
      InsertedInsts(InsertedInsts), PromotedInsts(PromotedInsts), TPT(TPT),
      LargeOffsetGEP(LargeOffsetGEP), OptSize(OptSize), PSI(PSI), BFI(BFI) {
  IgnoreProfitability = false;
}

3266public:
/// Find the maximal addressing mode that a load/store of V can fold,
/// give an access type of AccessTy.  This returns a list of involved
/// instructions in AddrModeInsts.
/// \p InsertedInsts The instructions inserted by other CodeGenPrepare
/// optimizations.
/// \p PromotedInsts maps the instructions to their type before promotion.
/// \p The ongoing transaction where every action should be registered.
static ExtAddrMode
Match(Value *V, Type *AccessTy, unsigned AS, Instruction *MemoryInst,
      SmallVectorImpl<Instruction *> &AddrModeInsts,
      const TargetLowering &TLI, const LoopInfo &LI,
      const std::function<const DominatorTree &()> getDTFn,
      const TargetRegisterInfo &TRI, const SetOfInstrs &InsertedInsts,
      InstrToOrigTy &PromotedInsts, TypePromotionTransaction &TPT,
      std::pair<AssertingVH<GetElementPtrInst>, int64_t> &LargeOffsetGEP,
      bool OptSize, ProfileSummaryInfo *PSI, BlockFrequencyInfo *BFI) {
  ExtAddrMode Result;

  bool Success = AddressingModeMatcher(AddrModeInsts, TLI, TRI, LI, getDTFn,
                                       AccessTy, AS, MemoryInst, Result,
                                       InsertedInsts, PromotedInsts, TPT,
                                       LargeOffsetGEP, OptSize, PSI, BFI)
                     .matchAddr(V, 0);
  (void)Success;
  assert(Success && "Couldn't select *anything*?")(static_cast <bool> (Success && "Couldn't select *anything*?"
) ? void (0) : __assert_fail ("Success && \"Couldn't select *anything*?\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3291, __extension__ __PRETTY_FUNCTION__
));
  return Result;
}

3295private:
bool matchScaledValue(Value *ScaleReg, int64_t Scale, unsigned Depth);
bool matchAddr(Value *Addr, unsigned Depth);
bool matchOperationAddr(User *AddrInst, unsigned Opcode, unsigned Depth,
                        bool *MovedAway = nullptr);
bool isProfitableToFoldIntoAddressingMode(Instruction *I,
                                          ExtAddrMode &AMBefore,
                                          ExtAddrMode &AMAfter);
bool valueAlreadyLiveAtInst(Value *Val, Value *KnownLive1, Value *KnownLive2);
bool isPromotionProfitable(unsigned NewCost, unsigned OldCost,
                           Value *PromotedOperand) const;
3306};

3308class PhiNodeSet;

3310/// An iterator for PhiNodeSet.
3311class PhiNodeSetIterator {
PhiNodeSet *const Set;
size_t CurrentIndex = 0;

3315public:
/// The constructor. Start should point to either a valid element, or be equal
/// to the size of the underlying SmallVector of the PhiNodeSet.
PhiNodeSetIterator(PhiNodeSet *const Set, size_t Start);
PHINode *operator*() const;
PhiNodeSetIterator &operator++();
bool operator==(const PhiNodeSetIterator &RHS) const;
bool operator!=(const PhiNodeSetIterator &RHS) const;
3323};

3325/// Keeps a set of PHINodes.
3326///
3327/// This is a minimal set implementation for a specific use case:
3328/// It is very fast when there are very few elements, but also provides good
3329/// performance when there are many. It is similar to SmallPtrSet, but also
3330/// provides iteration by insertion order, which is deterministic and stable
3331/// across runs. It is also similar to SmallSetVector, but provides removing
3332/// elements in O(1) time. This is achieved by not actually removing the element
3333/// from the underlying vector, so comes at the cost of using more memory, but
3334/// that is fine, since PhiNodeSets are used as short lived objects.
3335class PhiNodeSet {
friend class PhiNodeSetIterator;

using MapType = SmallDenseMap<PHINode *, size_t, 32>;
using iterator = PhiNodeSetIterator;

/// Keeps the elements in the order of their insertion in the underlying
/// vector. To achieve constant time removal, it never deletes any element.
SmallVector<PHINode *, 32> NodeList;

/// Keeps the elements in the underlying set implementation. This (and not the
/// NodeList defined above) is the source of truth on whether an element
/// is actually in the collection.
MapType NodeMap;

/// Points to the first valid (not deleted) element when the set is not empty
/// and the value is not zero. Equals to the size of the underlying vector
/// when the set is empty. When the value is 0, as in the beginning, the
/// first element may or may not be valid.
size_t FirstValidElement = 0;

3356public:
/// Inserts a new element to the collection.
/// \returns true if the element is actually added, i.e. was not in the
/// collection before the operation.
bool insert(PHINode *Ptr) {
  if (NodeMap.insert(std::make_pair(Ptr, NodeList.size())).second) {
    NodeList.push_back(Ptr);
    return true;
  }
  return false;
}

/// Removes the element from the collection.
/// \returns whether the element is actually removed, i.e. was in the
/// collection before the operation.
bool erase(PHINode *Ptr) {
  if (NodeMap.erase(Ptr)) {
    SkipRemovedElements(FirstValidElement);
    return true;
  }
  return false;
}

/// Removes all elements and clears the collection.
void clear() {
  NodeMap.clear();
  NodeList.clear();
  FirstValidElement = 0;
}

/// \returns an iterator that will iterate the elements in the order of
/// insertion.
iterator begin() {
  if (FirstValidElement == 0)
    SkipRemovedElements(FirstValidElement);
  return PhiNodeSetIterator(this, FirstValidElement);
}

/// \returns an iterator that points to the end of the collection.
iterator end() { return PhiNodeSetIterator(this, NodeList.size()); }

/// Returns the number of elements in the collection.
size_t size() const { return NodeMap.size(); }

/// \returns 1 if the given element is in the collection, and 0 if otherwise.
size_t count(PHINode *Ptr) const { return NodeMap.count(Ptr); }

3403private:
/// Updates the CurrentIndex so that it will point to a valid element.
///
/// If the element of NodeList at CurrentIndex is valid, it does not
/// change it. If there are no more valid elements, it updates CurrentIndex
/// to point to the end of the NodeList.
void SkipRemovedElements(size_t &CurrentIndex) {
  while (CurrentIndex < NodeList.size()) {
    auto it = NodeMap.find(NodeList[CurrentIndex]);
    // If the element has been deleted and added again later, NodeMap will
    // point to a different index, so CurrentIndex will still be invalid.
    if (it != NodeMap.end() && it->second == CurrentIndex)
      break;
    ++CurrentIndex;
  }
}
3419};

3421PhiNodeSetIterator::PhiNodeSetIterator(PhiNodeSet *const Set, size_t Start)
  : Set(Set), CurrentIndex(Start) {}

3424PHINode *PhiNodeSetIterator::operator*() const {
assert(CurrentIndex < Set->NodeList.size() &&(static_cast <bool> (CurrentIndex < Set->NodeList
.size() && "PhiNodeSet access out of range") ? void (
0) : __assert_fail ("CurrentIndex < Set->NodeList.size() && \"PhiNodeSet access out of range\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3426, __extension__ __PRETTY_FUNCTION__
))
       "PhiNodeSet access out of range")(static_cast <bool> (CurrentIndex < Set->NodeList
.size() && "PhiNodeSet access out of range") ? void (
0) : __assert_fail ("CurrentIndex < Set->NodeList.size() && \"PhiNodeSet access out of range\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3426, __extension__ __PRETTY_FUNCTION__
));
return Set->NodeList[CurrentIndex];
3428}

3430PhiNodeSetIterator &PhiNodeSetIterator::operator++() {
assert(CurrentIndex < Set->NodeList.size() &&(static_cast <bool> (CurrentIndex < Set->NodeList
.size() && "PhiNodeSet access out of range") ? void (
0) : __assert_fail ("CurrentIndex < Set->NodeList.size() && \"PhiNodeSet access out of range\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3432, __extension__ __PRETTY_FUNCTION__
))
       "PhiNodeSet access out of range")(static_cast <bool> (CurrentIndex < Set->NodeList
.size() && "PhiNodeSet access out of range") ? void (
0) : __assert_fail ("CurrentIndex < Set->NodeList.size() && \"PhiNodeSet access out of range\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3432, __extension__ __PRETTY_FUNCTION__
));
++CurrentIndex;
Set->SkipRemovedElements(CurrentIndex);
return *this;
3436}

3438bool PhiNodeSetIterator::operator==(const PhiNodeSetIterator &RHS) const {
return CurrentIndex == RHS.CurrentIndex;
3440}

3442bool PhiNodeSetIterator::operator!=(const PhiNodeSetIterator &RHS) const {
return !((*this) == RHS);
3444}

3446/// Keep track of simplification of Phi nodes.
3447/// Accept the set of all phi nodes and erase phi node from this set
3448/// if it is simplified.
3449class SimplificationTracker {
DenseMap<Value *, Value *> Storage;
const SimplifyQuery &SQ;
// Tracks newly created Phi nodes. The elements are iterated by insertion
// order.
PhiNodeSet AllPhiNodes;
// Tracks newly created Select nodes.
SmallPtrSet<SelectInst *, 32> AllSelectNodes;

3458public:
SimplificationTracker(const SimplifyQuery &sq) : SQ(sq) {}

Value *Get(Value *V) {
  do {
    auto SV = Storage.find(V);
    if (SV == Storage.end())
      return V;
    V = SV->second;
  } while (true);
}

Value *Simplify(Value *Val) {
  SmallVector<Value *, 32> WorkList;
  SmallPtrSet<Value *, 32> Visited;
  WorkList.push_back(Val);
  while (!WorkList.empty()) {
    auto *P = WorkList.pop_back_val();
    if (!Visited.insert(P).second)
      continue;
    if (auto *PI = dyn_cast<Instruction>(P))
      if (Value *V = simplifyInstruction(cast<Instruction>(PI), SQ)) {
        for (auto *U : PI->users())
          WorkList.push_back(cast<Value>(U));
        Put(PI, V);
        PI->replaceAllUsesWith(V);
        if (auto *PHI = dyn_cast<PHINode>(PI))
          AllPhiNodes.erase(PHI);
        if (auto *Select = dyn_cast<SelectInst>(PI))
          AllSelectNodes.erase(Select);
        PI->eraseFromParent();
      }
  }
  return Get(Val);
}

void Put(Value *From, Value *To) { Storage.insert({From, To}); }

void ReplacePhi(PHINode *From, PHINode *To) {
  Value *OldReplacement = Get(From);
  while (OldReplacement != From) {
    From = To;
    To = dyn_cast<PHINode>(OldReplacement);
    OldReplacement = Get(From);
  }
  assert(To && Get(To) == To && "Replacement PHI node is already replaced.")(static_cast <bool> (To && Get(To) == To &&
 "Replacement PHI node is already replaced.") ? void (0) : __assert_fail
 ("To && Get(To) == To && \"Replacement PHI node is already replaced.\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3503, __extension__ __PRETTY_FUNCTION__
));
  Put(From, To);
  From->replaceAllUsesWith(To);
  AllPhiNodes.erase(From);
  From->eraseFromParent();
}

PhiNodeSet &newPhiNodes() { return AllPhiNodes; }

void insertNewPhi(PHINode *PN) { AllPhiNodes.insert(PN); }

void insertNewSelect(SelectInst *SI) { AllSelectNodes.insert(SI); }

unsigned countNewPhiNodes() const { return AllPhiNodes.size(); }

unsigned countNewSelectNodes() const { return AllSelectNodes.size(); }

void destroyNewNodes(Type *CommonType) {
  // For safe erasing, replace the uses with dummy value first.
  auto *Dummy = PoisonValue::get(CommonType);
  for (auto *I : AllPhiNodes) {
    I->replaceAllUsesWith(Dummy);
    I->eraseFromParent();
  }
  AllPhiNodes.clear();
  for (auto *I : AllSelectNodes) {
    I->replaceAllUsesWith(Dummy);
    I->eraseFromParent();
  }
  AllSelectNodes.clear();
}
3534};

3536/// A helper class for combining addressing modes.
3537class AddressingModeCombiner {
typedef DenseMap<Value *, Value *> FoldAddrToValueMapping;
typedef std::pair<PHINode *, PHINode *> PHIPair;

3541private:
/// The addressing modes we've collected.
SmallVector<ExtAddrMode, 16> AddrModes;

/// The field in which the AddrModes differ, when we have more than one.
ExtAddrMode::FieldName DifferentField = ExtAddrMode::NoField;

/// Are the AddrModes that we have all just equal to their original values?
bool AllAddrModesTrivial = true;

/// Common Type for all different fields in addressing modes.
Type *CommonType = nullptr;

/// SimplifyQuery for simplifyInstruction utility.
const SimplifyQuery &SQ;

/// Original Address.
Value *Original;

3560public:
AddressingModeCombiner(const SimplifyQuery &_SQ, Value *OriginalValue)
    : SQ(_SQ), Original(OriginalValue) {}

/// Get the combined AddrMode
const ExtAddrMode &getAddrMode() const { return AddrModes[0]; }

/// Add a new AddrMode if it's compatible with the AddrModes we already
/// have.
/// \return True iff we succeeded in doing so.
bool addNewAddrMode(ExtAddrMode &NewAddrMode) {
  // Take note of if we have any non-trivial AddrModes, as we need to detect
  // when all AddrModes are trivial as then we would introduce a phi or select
  // which just duplicates what's already there.
  AllAddrModesTrivial = AllAddrModesTrivial && NewAddrMode.isTrivial();

  // If this is the first addrmode then everything is fine.
  if (AddrModes.empty()) {
    AddrModes.emplace_back(NewAddrMode);
    return true;
  }

  // Figure out how different this is from the other address modes, which we
  // can do just by comparing against the first one given that we only care
  // about the cumulative difference.
  ExtAddrMode::FieldName ThisDifferentField =
      AddrModes[0].compare(NewAddrMode);
  if (DifferentField == ExtAddrMode::NoField)
    DifferentField = ThisDifferentField;
  else if (DifferentField != ThisDifferentField)
    DifferentField = ExtAddrMode::MultipleFields;

  // If NewAddrMode differs in more than one dimension we cannot handle it.
  bool CanHandle = DifferentField != ExtAddrMode::MultipleFields;

  // If Scale Field is different then we reject.
  CanHandle = CanHandle && DifferentField != ExtAddrMode::ScaleField;

  // We also must reject the case when base offset is different and
  // scale reg is not null, we cannot handle this case due to merge of
  // different offsets will be used as ScaleReg.
  CanHandle = CanHandle && (DifferentField != ExtAddrMode::BaseOffsField ||
                            !NewAddrMode.ScaledReg);

  // We also must reject the case when GV is different and BaseReg installed
  // due to we want to use base reg as a merge of GV values.
  CanHandle = CanHandle && (DifferentField != ExtAddrMode::BaseGVField ||
                            !NewAddrMode.HasBaseReg);

  // Even if NewAddMode is the same we still need to collect it due to
  // original value is different. And later we will need all original values
  // as anchors during finding the common Phi node.
  if (CanHandle)
    AddrModes.emplace_back(NewAddrMode);
  else
    AddrModes.clear();

  return CanHandle;
}

/// Combine the addressing modes we've collected into a single
/// addressing mode.
/// \return True iff we successfully combined them or we only had one so
/// didn't need to combine them anyway.
bool combineAddrModes() {
  // If we have no AddrModes then they can't be combined.
  if (AddrModes.size() == 0)
    return false;

  // A single AddrMode can trivially be combined.
  if (AddrModes.size() == 1 || DifferentField == ExtAddrMode::NoField)
    return true;

  // If the AddrModes we collected are all just equal to the value they are
  // derived from then combining them wouldn't do anything useful.
  if (AllAddrModesTrivial)
    return false;

  if (!addrModeCombiningAllowed())
    return false;

  // Build a map between <original value, basic block where we saw it> to
  // value of base register.
  // Bail out if there is no common type.
  FoldAddrToValueMapping Map;
  if (!initializeMap(Map))
    return false;

  Value *CommonValue = findCommon(Map);
  if (CommonValue)
    AddrModes[0].SetCombinedField(DifferentField, CommonValue, AddrModes);
  return CommonValue != nullptr;
}

3654private:
/// Initialize Map with anchor values. For address seen
/// we set the value of different field saw in this address.
/// At the same time we find a common type for different field we will
/// use to create new Phi/Select nodes. Keep it in CommonType field.
/// Return false if there is no common type found.
bool initializeMap(FoldAddrToValueMapping &Map) {
  // Keep track of keys where the value is null. We will need to replace it
  // with constant null when we know the common type.
  SmallVector<Value *, 2> NullValue;
  Type *IntPtrTy = SQ.DL.getIntPtrType(AddrModes[0].OriginalValue->getType());
  for (auto &AM : AddrModes) {
    Value *DV = AM.GetFieldAsValue(DifferentField, IntPtrTy);
    if (DV) {
      auto *Type = DV->getType();
      if (CommonType && CommonType != Type)
        return false;
      CommonType = Type;
      Map[AM.OriginalValue] = DV;
    } else {
      NullValue.push_back(AM.OriginalValue);
    }
  }
  assert(CommonType && "At least one non-null value must be!")(static_cast <bool> (CommonType && "At least one non-null value must be!"
) ? void (0) : __assert_fail ("CommonType && \"At least one non-null value must be!\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3677, __extension__ __PRETTY_FUNCTION__
));
  for (auto *V : NullValue)
    Map[V] = Constant::getNullValue(CommonType);
  return true;
}

/// We have mapping between value A and other value B where B was a field in
/// addressing mode represented by A. Also we have an original value C
/// representing an address we start with. Traversing from C through phi and
/// selects we ended up with A's in a map. This utility function tries to find
/// a value V which is a field in addressing mode C and traversing through phi
/// nodes and selects we will end up in corresponded values B in a map.
/// The utility will create a new Phi/Selects if needed.
// The simple example looks as follows:
// BB1:
//   p1 = b1 + 40
//   br cond BB2, BB3
// BB2:
//   p2 = b2 + 40
//   br BB3
// BB3:
//   p = phi [p1, BB1], [p2, BB2]
//   v = load p
// Map is
//   p1 -> b1
//   p2 -> b2
// Request is
//   p -> ?
// The function tries to find or build phi [b1, BB1], [b2, BB2] in BB3.
Value *findCommon(FoldAddrToValueMapping &Map) {
  // Tracks the simplification of newly created phi nodes. The reason we use
  // this mapping is because we will add new created Phi nodes in AddrToBase.
  // Simplification of Phi nodes is recursive, so some Phi node may
  // be simplified after we added it to AddrToBase. In reality this
  // simplification is possible only if original phi/selects were not
  // simplified yet.
  // Using this mapping we can find the current value in AddrToBase.
  SimplificationTracker ST(SQ);

  // First step, DFS to create PHI nodes for all intermediate blocks.
  // Also fill traverse order for the second step.
  SmallVector<Value *, 32> TraverseOrder;
  InsertPlaceholders(Map, TraverseOrder, ST);

  // Second Step, fill new nodes by merged values and simplify if possible.
  FillPlaceholders(Map, TraverseOrder, ST);

  if (!AddrSinkNewSelects && ST.countNewSelectNodes() > 0) {
    ST.destroyNewNodes(CommonType);
    return nullptr;
  }

  // Now we'd like to match New Phi nodes to existed ones.
  unsigned PhiNotMatchedCount = 0;
  if (!MatchPhiSet(ST, AddrSinkNewPhis, PhiNotMatchedCount)) {
    ST.destroyNewNodes(CommonType);
    return nullptr;
  }

  auto *Result = ST.Get(Map.find(Original)->second);
  if (Result) {
    NumMemoryInstsPhiCreated += ST.countNewPhiNodes() + PhiNotMatchedCount;
    NumMemoryInstsSelectCreated += ST.countNewSelectNodes();
  }
  return Result;
}

/// Try to match PHI node to Candidate.
/// Matcher tracks the matched Phi nodes.
bool MatchPhiNode(PHINode *PHI, PHINode *Candidate,
                  SmallSetVector<PHIPair, 8> &Matcher,
                  PhiNodeSet &PhiNodesToMatch) {
  SmallVector<PHIPair, 8> WorkList;
  Matcher.insert({PHI, Candidate});
  SmallSet<PHINode *, 8> MatchedPHIs;
  MatchedPHIs.insert(PHI);
  WorkList.push_back({PHI, Candidate});
  SmallSet<PHIPair, 8> Visited;
  while (!WorkList.empty()) {
    auto Item = WorkList.pop_back_val();
    if (!Visited.insert(Item).second)
      continue;
    // We iterate over all incoming values to Phi to compare them.
    // If values are different and both of them Phi and the first one is a
    // Phi we added (subject to match) and both of them is in the same basic
    // block then we can match our pair if values match. So we state that
    // these values match and add it to work list to verify that.
    for (auto *B : Item.first->blocks()) {
      Value *FirstValue = Item.first->getIncomingValueForBlock(B);
      Value *SecondValue = Item.second->getIncomingValueForBlock(B);
      if (FirstValue == SecondValue)
        continue;

      PHINode *FirstPhi = dyn_cast<PHINode>(FirstValue);
      PHINode *SecondPhi = dyn_cast<PHINode>(SecondValue);

      // One of them is not Phi or
      // The first one is not Phi node from the set we'd like to match or
      // Phi nodes from different basic blocks then
      // we will not be able to match.
      if (!FirstPhi || !SecondPhi || !PhiNodesToMatch.count(FirstPhi) ||
          FirstPhi->getParent() != SecondPhi->getParent())
        return false;

      // If we already matched them then continue.
      if (Matcher.count({FirstPhi, SecondPhi}))
        continue;
      // So the values are different and does not match. So we need them to
      // match. (But we register no more than one match per PHI node, so that
      // we won't later try to replace them twice.)
      if (MatchedPHIs.insert(FirstPhi).second)
        Matcher.insert({FirstPhi, SecondPhi});
      // But me must check it.
      WorkList.push_back({FirstPhi, SecondPhi});
    }
  }
  return true;
}

/// For the given set of PHI nodes (in the SimplificationTracker) try
/// to find their equivalents.
/// Returns false if this matching fails and creation of new Phi is disabled.
bool MatchPhiSet(SimplificationTracker &ST, bool AllowNewPhiNodes,
                 unsigned &PhiNotMatchedCount) {
  // Matched and PhiNodesToMatch iterate their elements in a deterministic
  // order, so the replacements (ReplacePhi) are also done in a deterministic
  // order.
  SmallSetVector<PHIPair, 8> Matched;
  SmallPtrSet<PHINode *, 8> WillNotMatch;
  PhiNodeSet &PhiNodesToMatch = ST.newPhiNodes();
  while (PhiNodesToMatch.size()) {
    PHINode *PHI = *PhiNodesToMatch.begin();

    // Add us, if no Phi nodes in the basic block we do not match.
    WillNotMatch.clear();
    WillNotMatch.insert(PHI);

    // Traverse all Phis until we found equivalent or fail to do that.
    bool IsMatched = false;
    for (auto &P : PHI->getParent()->phis()) {
      // Skip new Phi nodes.
      if (PhiNodesToMatch.count(&P))
        continue;
      if ((IsMatched = MatchPhiNode(PHI, &P, Matched, PhiNodesToMatch)))
        break;
      // If it does not match, collect all Phi nodes from matcher.
      // if we end up with no match, them all these Phi nodes will not match
      // later.
      for (auto M : Matched)
        WillNotMatch.insert(M.first);
      Matched.clear();
    }
    if (IsMatched) {
      // Replace all matched values and erase them.
      for (auto MV : Matched)
        ST.ReplacePhi(MV.first, MV.second);
      Matched.clear();
      continue;
    }
    // If we are not allowed to create new nodes then bail out.
    if (!AllowNewPhiNodes)
      return false;
    // Just remove all seen values in matcher. They will not match anything.
    PhiNotMatchedCount += WillNotMatch.size();
    for (auto *P : WillNotMatch)
      PhiNodesToMatch.erase(P);
  }
  return true;
}
/// Fill the placeholders with values from predecessors and simplify them.
void FillPlaceholders(FoldAddrToValueMapping &Map,
                      SmallVectorImpl<Value *> &TraverseOrder,
                      SimplificationTracker &ST) {
  while (!TraverseOrder.empty()) {
    Value *Current = TraverseOrder.pop_back_val();
    assert(Map.contains(Current) && "No node to fill!!!")(static_cast <bool> (Map.contains(Current) && "No node to fill!!!"
) ? void (0) : __assert_fail ("Map.contains(Current) && \"No node to fill!!!\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3852, __extension__ __PRETTY_FUNCTION__
));
    Value *V = Map[Current];

    if (SelectInst *Select = dyn_cast<SelectInst>(V)) {
      // CurrentValue also must be Select.
      auto *CurrentSelect = cast<SelectInst>(Current);
      auto *TrueValue = CurrentSelect->getTrueValue();
      assert(Map.contains(TrueValue) && "No True Value!")(static_cast <bool> (Map.contains(TrueValue) &&
 "No True Value!") ? void (0) : __assert_fail ("Map.contains(TrueValue) && \"No True Value!\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3859, __extension__ __PRETTY_FUNCTION__
));
      Select->setTrueValue(ST.Get(Map[TrueValue]));
      auto *FalseValue = CurrentSelect->getFalseValue();
      assert(Map.contains(FalseValue) && "No False Value!")(static_cast <bool> (Map.contains(FalseValue) &&
 "No False Value!") ? void (0) : __assert_fail ("Map.contains(FalseValue) && \"No False Value!\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3862, __extension__ __PRETTY_FUNCTION__
));
      Select->setFalseValue(ST.Get(Map[FalseValue]));
    } else {
      // Must be a Phi node then.
      auto *PHI = cast<PHINode>(V);
      // Fill the Phi node with values from predecessors.
      for (auto *B : predecessors(PHI->getParent())) {
        Value *PV = cast<PHINode>(Current)->getIncomingValueForBlock(B);
        assert(Map.contains(PV) && "No predecessor Value!")(static_cast <bool> (Map.contains(PV) && "No predecessor Value!"
) ? void (0) : __assert_fail ("Map.contains(PV) && \"No predecessor Value!\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3870, __extension__ __PRETTY_FUNCTION__
));
        PHI->addIncoming(ST.Get(Map[PV]), B);
      }
    }
    Map[Current] = ST.Simplify(V);
  }
}

/// Starting from original value recursively iterates over def-use chain up to
/// known ending values represented in a map. For each traversed phi/select
/// inserts a placeholder Phi or Select.
/// Reports all new created Phi/Select nodes by adding them to set.
/// Also reports and order in what values have been traversed.
void InsertPlaceholders(FoldAddrToValueMapping &Map,
                        SmallVectorImpl<Value *> &TraverseOrder,
                        SimplificationTracker &ST) {
  SmallVector<Value *, 32> Worklist;
  assert((isa<PHINode>(Original) || isa<SelectInst>(Original)) &&(static_cast <bool> ((isa<PHINode>(Original) || isa
<SelectInst>(Original)) && "Address must be a Phi or Select node"
) ? void (0) : __assert_fail ("(isa<PHINode>(Original) || isa<SelectInst>(Original)) && \"Address must be a Phi or Select node\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3888, __extension__ __PRETTY_FUNCTION__
))
         "Address must be a Phi or Select node")(static_cast <bool> ((isa<PHINode>(Original) || isa
<SelectInst>(Original)) && "Address must be a Phi or Select node"
) ? void (0) : __assert_fail ("(isa<PHINode>(Original) || isa<SelectInst>(Original)) && \"Address must be a Phi or Select node\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 3888, __extension__ __PRETTY_FUNCTION__
));
  auto *Dummy = PoisonValue::get(CommonType);
  Worklist.push_back(Original);
  while (!Worklist.empty()) {
    Value *Current = Worklist.pop_back_val();
    // if it is already visited or it is an ending value then skip it.
    if (Map.contains(Current))
      continue;
    TraverseOrder.push_back(Current);

    // CurrentValue must be a Phi node or select. All others must be covered
    // by anchors.
    if (SelectInst *CurrentSelect = dyn_cast<SelectInst>(Current)) {
      // Is it OK to get metadata from OrigSelect?!
      // Create a Select placeholder with dummy value.
      SelectInst *Select = SelectInst::Create(
          CurrentSelect->getCondition(), Dummy, Dummy,
          CurrentSelect->getName(), CurrentSelect, CurrentSelect);
      Map[Current] = Select;
      ST.insertNewSelect(Select);
      // We are interested in True and False values.
      Worklist.push_back(CurrentSelect->getTrueValue());
      Worklist.push_back(CurrentSelect->getFalseValue());
    } else {
      // It must be a Phi node then.
      PHINode *CurrentPhi = cast<PHINode>(Current);
      unsigned PredCount = CurrentPhi->getNumIncomingValues();
      PHINode *PHI =
          PHINode::Create(CommonType, PredCount, "sunk_phi", CurrentPhi);
      Map[Current] = PHI;
      ST.insertNewPhi(PHI);
      append_range(Worklist, CurrentPhi->incoming_values());
    }
  }
}

bool addrModeCombiningAllowed() {
  if (DisableComplexAddrModes)
    return false;
  switch (DifferentField) {
  default:
    return false;
  case ExtAddrMode::BaseRegField:
    return AddrSinkCombineBaseReg;
  case ExtAddrMode::BaseGVField:
    return AddrSinkCombineBaseGV;
  case ExtAddrMode::BaseOffsField:
    return AddrSinkCombineBaseOffs;
  case ExtAddrMode::ScaledRegField:
    return AddrSinkCombineScaledReg;
  }
}
3940};
3941} // end anonymous namespace

3943/// Try adding ScaleReg*Scale to the current addressing mode.
3944/// Return true and update AddrMode if this addr mode is legal for the target,
3945/// false if not.
3946bool AddressingModeMatcher::matchScaledValue(Value *ScaleReg, int64_t Scale,
                                           unsigned Depth) {
// If Scale is 1, then this is the same as adding ScaleReg to the addressing
// mode.  Just process that directly.
if (Scale == 1)
  return matchAddr(ScaleReg, Depth);

// If the scale is 0, it takes nothing to add this.
if (Scale == 0)
  return true;

// If we already have a scale of this value, we can add to it, otherwise, we
// need an available scale field.
if (AddrMode.Scale != 0 && AddrMode.ScaledReg != ScaleReg)
  return false;

ExtAddrMode TestAddrMode = AddrMode;

// Add scale to turn X*4+X*3 -> X*7.  This could also do things like
// [A+B + A*7] -> [B+A*8].
TestAddrMode.Scale += Scale;
TestAddrMode.ScaledReg = ScaleReg;

// If the new address isn't legal, bail out.
if (!TLI.isLegalAddressingMode(DL, TestAddrMode, AccessTy, AddrSpace))
  return false;

// It was legal, so commit it.
AddrMode = TestAddrMode;

// Okay, we decided that we can add ScaleReg+Scale to AddrMode.  Check now
// to see if ScaleReg is actually X+C.  If so, we can turn this into adding
// X*Scale + C*Scale to addr mode. If we found available IV increment, do not
// go any further: we can reuse it and cannot eliminate it.
ConstantInt *CI = nullptr;
Value *AddLHS = nullptr;
if (isa<Instruction>(ScaleReg) && // not a constant expr.
    match(ScaleReg, m_Add(m_Value(AddLHS), m_ConstantInt(CI))) &&
    !isIVIncrement(ScaleReg, &LI) && CI->getValue().isSignedIntN(64)) {
  TestAddrMode.InBounds = false;
  TestAddrMode.ScaledReg = AddLHS;
  TestAddrMode.BaseOffs += CI->getSExtValue() * TestAddrMode.Scale;

  // If this addressing mode is legal, commit it and remember that we folded
  // this instruction.
  if (TLI.isLegalAddressingMode(DL, TestAddrMode, AccessTy, AddrSpace)) {
    AddrModeInsts.push_back(cast<Instruction>(ScaleReg));
    AddrMode = TestAddrMode;
    return true;
  }
  // Restore status quo.
  TestAddrMode = AddrMode;
}

// If this is an add recurrence with a constant step, return the increment
// instruction and the canonicalized step.
auto GetConstantStep =
    [this](const Value *V) -> std::optional<std::pair<Instruction *, APInt>> {
  auto *PN = dyn_cast<PHINode>(V);
  if (!PN)
    return std::nullopt;
  auto IVInc = getIVIncrement(PN, &LI);
  if (!IVInc)
    return std::nullopt;
  // TODO: The result of the intrinsics above is two-complement. However when
  // IV inc is expressed as add or sub, iv.next is potentially a poison value.
  // If it has nuw or nsw flags, we need to make sure that these flags are
  // inferrable at the point of memory instruction. Otherwise we are replacing
  // well-defined two-complement computation with poison. Currently, to avoid
  // potentially complex analysis needed to prove this, we reject such cases.
  if (auto *OIVInc = dyn_cast<OverflowingBinaryOperator>(IVInc->first))
    if (OIVInc->hasNoSignedWrap() || OIVInc->hasNoUnsignedWrap())
      return std::nullopt;
  if (auto *ConstantStep = dyn_cast<ConstantInt>(IVInc->second))
    return std::make_pair(IVInc->first, ConstantStep->getValue());
  return std::nullopt;
};

// Try to account for the following special case:
// 1. ScaleReg is an inductive variable;
// 2. We use it with non-zero offset;
// 3. IV's increment is available at the point of memory instruction.
//
// In this case, we may reuse the IV increment instead of the IV Phi to
// achieve the following advantages:
// 1. If IV step matches the offset, we will have no need in the offset;
// 2. Even if they don't match, we will reduce the overlap of living IV
//    and IV increment, that will potentially lead to better register
//    assignment.
if (AddrMode.BaseOffs) {
  if (auto IVStep = GetConstantStep(ScaleReg)) {
    Instruction *IVInc = IVStep->first;
    // The following assert is important to ensure a lack of infinite loops.
    // This transforms is (intentionally) the inverse of the one just above.
    // If they don't agree on the definition of an increment, we'd alternate
    // back and forth indefinitely.
    assert(isIVIncrement(IVInc, &LI) && "implied by GetConstantStep")(static_cast <bool> (isIVIncrement(IVInc, &LI) &&
 "implied by GetConstantStep") ? void (0) : __assert_fail ("isIVIncrement(IVInc, &LI) && \"implied by GetConstantStep\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 4042, __extension__ __PRETTY_FUNCTION__
));
    APInt Step = IVStep->second;
    APInt Offset = Step * AddrMode.Scale;
    if (Offset.isSignedIntN(64)) {
      TestAddrMode.InBounds = false;
      TestAddrMode.ScaledReg = IVInc;
      TestAddrMode.BaseOffs -= Offset.getLimitedValue();
      // If this addressing mode is legal, commit it..
      // (Note that we defer the (expensive) domtree base legality check
      // to the very last possible point.)
      if (TLI.isLegalAddressingMode(DL, TestAddrMode, AccessTy, AddrSpace) &&
          getDTFn().dominates(IVInc, MemoryInst)) {
        AddrModeInsts.push_back(cast<Instruction>(IVInc));
        AddrMode = TestAddrMode;
        return true;
      }
      // Restore status quo.
      TestAddrMode = AddrMode;
    }
  }
}

// Otherwise, just return what we have.
return true;
4066}

4068/// This is a little filter, which returns true if an addressing computation
4069/// involving I might be folded into a load/store accessing it.
4070/// This doesn't need to be perfect, but needs to accept at least
4071/// the set of instructions that MatchOperationAddr can.
4072static bool MightBeFoldableInst(Instruction *I) {
switch (I->getOpcode()) {
case Instruction::BitCast:
case Instruction::AddrSpaceCast:
  // Don't touch identity bitcasts.
  if (I->getType() == I->getOperand(0)->getType())
    return false;
  return I->getType()->isIntOrPtrTy();
case Instruction::PtrToInt:
  // PtrToInt is always a noop, as we know that the int type is pointer sized.
  return true;
case Instruction::IntToPtr:
  // We know the input is intptr_t, so this is foldable.
  return true;
case Instruction::Add:
  return true;
case Instruction::Mul:
case Instruction::Shl:
  // Can only handle X*C and X << C.
  return isa<ConstantInt>(I->getOperand(1));
case Instruction::GetElementPtr:
  return true;
default:
  return false;
}
4097}

4099/// Check whether or not \p Val is a legal instruction for \p TLI.
4100/// \note \p Val is assumed to be the product of some type promotion.
4101/// Therefore if \p Val has an undefined state in \p TLI, this is assumed
4102/// to be legal, as the non-promoted value would have had the same state.
4103static bool isPromotedInstructionLegal(const TargetLowering &TLI,
                                     const DataLayout &DL, Value *Val) {
Instruction *PromotedInst = dyn_cast<Instruction>(Val);
if (!PromotedInst)
  return false;
int ISDOpcode = TLI.InstructionOpcodeToISD(PromotedInst->getOpcode());
// If the ISDOpcode is undefined, it was undefined before the promotion.
if (!ISDOpcode)
  return true;
// Otherwise, check if the promoted instruction is legal or not.
return TLI.isOperationLegalOrCustom(
    ISDOpcode, TLI.getValueType(DL, PromotedInst->getType()));
4115}

4117namespace {

4119/// Hepler class to perform type promotion.
4120class TypePromotionHelper {
/// Utility function to add a promoted instruction \p ExtOpnd to
/// \p PromotedInsts and record the type of extension we have seen.
static void addPromotedInst(InstrToOrigTy &PromotedInsts,
                            Instruction *ExtOpnd, bool IsSExt) {
  ExtType ExtTy = IsSExt ? SignExtension : ZeroExtension;
  InstrToOrigTy::iterator It = PromotedInsts.find(ExtOpnd);
  if (It != PromotedInsts.end()) {
    // If the new extension is same as original, the information in
    // PromotedInsts[ExtOpnd] is still correct.
    if (It->second.getInt() == ExtTy)
      return;

    // Now the new extension is different from old extension, we make
    // the type information invalid by setting extension type to
    // BothExtension.
    ExtTy = BothExtension;
  }
  PromotedInsts[ExtOpnd] = TypeIsSExt(ExtOpnd->getType(), ExtTy);
}

/// Utility function to query the original type of instruction \p Opnd
/// with a matched extension type. If the extension doesn't match, we
/// cannot use the information we had on the original type.
/// BothExtension doesn't match any extension type.
static const Type *getOrigType(const InstrToOrigTy &PromotedInsts,
                               Instruction *Opnd, bool IsSExt) {
  ExtType ExtTy = IsSExt ? SignExtension : ZeroExtension;
  InstrToOrigTy::const_iterator It = PromotedInsts.find(Opnd);
  if (It != PromotedInsts.end() && It->second.getInt() == ExtTy)
    return It->second.getPointer();
  return nullptr;
}

/// Utility function to check whether or not a sign or zero extension
/// of \p Inst with \p ConsideredExtType can be moved through \p Inst by
/// either using the operands of \p Inst or promoting \p Inst.
/// The type of the extension is defined by \p IsSExt.
/// In other words, check if:
/// ext (Ty Inst opnd1 opnd2 ... opndN) to ConsideredExtType.
/// #1 Promotion applies:
/// ConsideredExtType Inst (ext opnd1 to ConsideredExtType, ...).
/// #2 Operand reuses:
/// ext opnd1 to ConsideredExtType.
/// \p PromotedInsts maps the instructions to their type before promotion.
static bool canGetThrough(const Instruction *Inst, Type *ConsideredExtType,
                          const InstrToOrigTy &PromotedInsts, bool IsSExt);

/// Utility function to determine if \p OpIdx should be promoted when
/// promoting \p Inst.
static bool shouldExtOperand(const Instruction *Inst, int OpIdx) {
  return !(isa<SelectInst>(Inst) && OpIdx == 0);
}

/// Utility function to promote the operand of \p Ext when this
/// operand is a promotable trunc or sext or zext.
/// \p PromotedInsts maps the instructions to their type before promotion.
/// \p CreatedInstsCost[out] contains the cost of all instructions
/// created to promote the operand of Ext.
/// Newly added extensions are inserted in \p Exts.
/// Newly added truncates are inserted in \p Truncs.
/// Should never be called directly.
/// \return The promoted value which is used instead of Ext.
static Value *promoteOperandForTruncAndAnyExt(
    Instruction *Ext, TypePromotionTransaction &TPT,
    InstrToOrigTy &PromotedInsts, unsigned &CreatedInstsCost,
    SmallVectorImpl<Instruction *> *Exts,
    SmallVectorImpl<Instruction *> *Truncs, const TargetLowering &TLI);

/// Utility function to promote the operand of \p Ext when this
/// operand is promotable and is not a supported trunc or sext.
/// \p PromotedInsts maps the instructions to their type before promotion.
/// \p CreatedInstsCost[out] contains the cost of all the instructions
/// created to promote the operand of Ext.
/// Newly added extensions are inserted in \p Exts.
/// Newly added truncates are inserted in \p Truncs.
/// Should never be called directly.
/// \return The promoted value which is used instead of Ext.
static Value *promoteOperandForOther(Instruction *Ext,
                                     TypePromotionTransaction &TPT,
                                     InstrToOrigTy &PromotedInsts,
                                     unsigned &CreatedInstsCost,
                                     SmallVectorImpl<Instruction *> *Exts,
                                     SmallVectorImpl<Instruction *> *Truncs,
                                     const TargetLowering &TLI, bool IsSExt);

/// \see promoteOperandForOther.
static Value *signExtendOperandForOther(
    Instruction *Ext, TypePromotionTransaction &TPT,
    InstrToOrigTy &PromotedInsts, unsigned &CreatedInstsCost,
    SmallVectorImpl<Instruction *> *Exts,
    SmallVectorImpl<Instruction *> *Truncs, const TargetLowering &TLI) {
  return promoteOperandForOther(Ext, TPT, PromotedInsts, CreatedInstsCost,
                                Exts, Truncs, TLI, true);
}

/// \see promoteOperandForOther.
static Value *zeroExtendOperandForOther(
    Instruction *Ext, TypePromotionTransaction &TPT,
    InstrToOrigTy &PromotedInsts, unsigned &CreatedInstsCost,
    SmallVectorImpl<Instruction *> *Exts,
    SmallVectorImpl<Instruction *> *Truncs, const TargetLowering &TLI) {
  return promoteOperandForOther(Ext, TPT, PromotedInsts, CreatedInstsCost,
                                Exts, Truncs, TLI, false);
}

4226public:
/// Type for the utility function that promotes the operand of Ext.
using Action = Value *(*)(Instruction *Ext, TypePromotionTransaction &TPT,
                          InstrToOrigTy &PromotedInsts,
                          unsigned &CreatedInstsCost,
                          SmallVectorImpl<Instruction *> *Exts,
                          SmallVectorImpl<Instruction *> *Truncs,
                          const TargetLowering &TLI);

/// Given a sign/zero extend instruction \p Ext, return the appropriate
/// action to promote the operand of \p Ext instead of using Ext.
/// \return NULL if no promotable action is possible with the current
/// sign extension.
/// \p InsertedInsts keeps track of all the instructions inserted by the
/// other CodeGenPrepare optimizations. This information is important
/// because we do not want to promote these instructions as CodeGenPrepare
/// will reinsert them later. Thus creating an infinite loop: create/remove.
/// \p PromotedInsts maps the instructions to their type before promotion.
static Action getAction(Instruction *Ext, const SetOfInstrs &InsertedInsts,
                        const TargetLowering &TLI,
                        const InstrToOrigTy &PromotedInsts);
4247};

4249} // end anonymous namespace

4251bool TypePromotionHelper::canGetThrough(const Instruction *Inst,
                                      Type *ConsideredExtType,
                                      const InstrToOrigTy &PromotedInsts,
                                      bool IsSExt) {
// The promotion helper does not know how to deal with vector types yet.
// To be able to fix that, we would need to fix the places where we
// statically extend, e.g., constants and such.
if (Inst->getType()->isVectorTy())
  return false;

// We can always get through zext.
if (isa<ZExtInst>(Inst))
  return true;

// sext(sext) is ok too.
if (IsSExt && isa<SExtInst>(Inst))
  return true;

// We can get through binary operator, if it is legal. In other words, the
// binary operator must have a nuw or nsw flag.
if (const auto *BinOp = dyn_cast<BinaryOperator>(Inst))
  if (isa<OverflowingBinaryOperator>(BinOp) &&
      ((!IsSExt && BinOp->hasNoUnsignedWrap()) ||
       (IsSExt && BinOp->hasNoSignedWrap())))
    return true;

// ext(and(opnd, cst)) --> and(ext(opnd), ext(cst))
if ((Inst->getOpcode() == Instruction::And ||
     Inst->getOpcode() == Instruction::Or))
  return true;

// ext(xor(opnd, cst)) --> xor(ext(opnd), ext(cst))
if (Inst->getOpcode() == Instruction::Xor) {
  // Make sure it is not a NOT.
  if (const auto *Cst = dyn_cast<ConstantInt>(Inst->getOperand(1)))
    if (!Cst->getValue().isAllOnes())
      return true;
}

// zext(shrl(opnd, cst)) --> shrl(zext(opnd), zext(cst))
// It may change a poisoned value into a regular value, like
//     zext i32 (shrl i8 %val, 12)  -->  shrl i32 (zext i8 %val), 12
//          poisoned value                    regular value
// It should be OK since undef covers valid value.
if (Inst->getOpcode() == Instruction::LShr && !IsSExt)
  return true;

// and(ext(shl(opnd, cst)), cst) --> and(shl(ext(opnd), ext(cst)), cst)
// It may change a poisoned value into a regular value, like
//     zext i32 (shl i8 %val, 12)  -->  shl i32 (zext i8 %val), 12
//          poisoned value                    regular value
// It should be OK since undef covers valid value.
if (Inst->getOpcode() == Instruction::Shl && Inst->hasOneUse()) {
  const auto *ExtInst = cast<const Instruction>(*Inst->user_begin());
  if (ExtInst->hasOneUse()) {
    const auto *AndInst = dyn_cast<const Instruction>(*ExtInst->user_begin());
    if (AndInst && AndInst->getOpcode() == Instruction::And) {
      const auto *Cst = dyn_cast<ConstantInt>(AndInst->getOperand(1));
      if (Cst &&
          Cst->getValue().isIntN(Inst->getType()->getIntegerBitWidth()))
        return true;
    }
  }
}

// Check if we can do the following simplification.
// ext(trunc(opnd)) --> ext(opnd)
if (!isa<TruncInst>(Inst))
  return false;

Value *OpndVal = Inst->getOperand(0);
// Check if we can use this operand in the extension.
// If the type is larger than the result type of the extension, we cannot.
if (!OpndVal->getType()->isIntegerTy() ||
    OpndVal->getType()->getIntegerBitWidth() >
        ConsideredExtType->getIntegerBitWidth())
  return false;

// If the operand of the truncate is not an instruction, we will not have
// any information on the dropped bits.
// (Actually we could for constant but it is not worth the extra logic).
Instruction *Opnd = dyn_cast<Instruction>(OpndVal);
if (!Opnd)
  return false;

// Check if the source of the type is narrow enough.
// I.e., check that trunc just drops extended bits of the same kind of
// the extension.
// #1 get the type of the operand and check the kind of the extended bits.
const Type *OpndType = getOrigType(PromotedInsts, Opnd, IsSExt);
if (OpndType)
  ;
else if ((IsSExt && isa<SExtInst>(Opnd)) || (!IsSExt && isa<ZExtInst>(Opnd)))
  OpndType = Opnd->getOperand(0)->getType();
else
  return false;

// #2 check that the truncate just drops extended bits.
return Inst->getType()->getIntegerBitWidth() >=
       OpndType->getIntegerBitWidth();
4351}

4353TypePromotionHelper::Action TypePromotionHelper::getAction(
  Instruction *Ext, const SetOfInstrs &InsertedInsts,
  const TargetLowering &TLI, const InstrToOrigTy &PromotedInsts) {
assert((isa<SExtInst>(Ext) || isa<ZExtInst>(Ext)) &&(static_cast <bool> ((isa<SExtInst>(Ext) || isa<
ZExtInst>(Ext)) && "Unexpected instruction type") ?
 void (0) : __assert_fail ("(isa<SExtInst>(Ext) || isa<ZExtInst>(Ext)) && \"Unexpected instruction type\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 4357, __extension__ __PRETTY_FUNCTION__
))
       "Unexpected instruction type")(static_cast <bool> ((isa<SExtInst>(Ext) || isa<
ZExtInst>(Ext)) && "Unexpected instruction type") ?
 void (0) : __assert_fail ("(isa<SExtInst>(Ext) || isa<ZExtInst>(Ext)) && \"Unexpected instruction type\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 4357, __extension__ __PRETTY_FUNCTION__
));
Instruction *ExtOpnd = dyn_cast<Instruction>(Ext->getOperand(0));
Type *ExtTy = Ext->getType();
bool IsSExt = isa<SExtInst>(Ext);
// If the operand of the extension is not an instruction, we cannot
// get through.
// If it, check we can get through.
if (!ExtOpnd || !canGetThrough(ExtOpnd, ExtTy, PromotedInsts, IsSExt))
  return nullptr;

// Do not promote if the operand has been added by codegenprepare.
// Otherwise, it means we are undoing an optimization that is likely to be
// redone, thus causing potential infinite loop.
if (isa<TruncInst>(ExtOpnd) && InsertedInsts.count(ExtOpnd))
  return nullptr;

// SExt or Trunc instructions.
// Return the related handler.
if (isa<SExtInst>(ExtOpnd) || isa<TruncInst>(ExtOpnd) ||
    isa<ZExtInst>(ExtOpnd))
  return promoteOperandForTruncAndAnyExt;

// Regular instruction.
// Abort early if we will have to insert non-free instructions.
if (!ExtOpnd->hasOneUse() && !TLI.isTruncateFree(ExtTy, ExtOpnd->getType()))
  return nullptr;
return IsSExt ? signExtendOperandForOther : zeroExtendOperandForOther;
4384}

4386Value *TypePromotionHelper::promoteOperandForTruncAndAnyExt(
  Instruction *SExt, TypePromotionTransaction &TPT,
  InstrToOrigTy &PromotedInsts, unsigned &CreatedInstsCost,
  SmallVectorImpl<Instruction *> *Exts,
  SmallVectorImpl<Instruction *> *Truncs, const TargetLowering &TLI) {
// By construction, the operand of SExt is an instruction. Otherwise we cannot
// get through it and this method should not be called.
Instruction *SExtOpnd = cast<Instruction>(SExt->getOperand(0));
Value *ExtVal = SExt;
bool HasMergedNonFreeExt = false;
if (isa<ZExtInst>(SExtOpnd)) {
  // Replace s|zext(zext(opnd))
  // => zext(opnd).
  HasMergedNonFreeExt = !TLI.isExtFree(SExtOpnd);
  Value *ZExt =
      TPT.createZExt(SExt, SExtOpnd->getOperand(0), SExt->getType());
  TPT.replaceAllUsesWith(SExt, ZExt);
  TPT.eraseInstruction(SExt);
  ExtVal = ZExt;
} else {
  // Replace z|sext(trunc(opnd)) or sext(sext(opnd))
  // => z|sext(opnd).
  TPT.setOperand(SExt, 0, SExtOpnd->getOperand(0));
}
CreatedInstsCost = 0;

// Remove dead code.
if (SExtOpnd->use_empty())
  TPT.eraseInstruction(SExtOpnd);

// Check if the extension is still needed.
Instruction *ExtInst = dyn_cast<Instruction>(ExtVal);
if (!ExtInst || ExtInst->getType() != ExtInst->getOperand(0)->getType()) {
  if (ExtInst) {
    if (Exts)
      Exts->push_back(ExtInst);
    CreatedInstsCost = !TLI.isExtFree(ExtInst) && !HasMergedNonFreeExt;
  }
  return ExtVal;
}

// At this point we have: ext ty opnd to ty.
// Reassign the uses of ExtInst to the opnd and remove ExtInst.
Value *NextVal = ExtInst->getOperand(0);
TPT.eraseInstruction(ExtInst, NextVal);
return NextVal;
4432}

4434Value *TypePromotionHelper::promoteOperandForOther(
  Instruction *Ext, TypePromotionTransaction &TPT,
  InstrToOrigTy &PromotedInsts, unsigned &CreatedInstsCost,
  SmallVectorImpl<Instruction *> *Exts,
  SmallVectorImpl<Instruction *> *Truncs, const TargetLowering &TLI,
  bool IsSExt) {
// By construction, the operand of Ext is an instruction. Otherwise we cannot
// get through it and this method should not be called.
Instruction *ExtOpnd = cast<Instruction>(Ext->getOperand(0));
CreatedInstsCost = 0;
if (!ExtOpnd->hasOneUse()) {
  // ExtOpnd will be promoted.
  // All its uses, but Ext, will need to use a truncated value of the
  // promoted version.
  // Create the truncate now.
  Value *Trunc = TPT.createTrunc(Ext, ExtOpnd->getType());
  if (Instruction *ITrunc = dyn_cast<Instruction>(Trunc)) {
    // Insert it just after the definition.
    ITrunc->moveAfter(ExtOpnd);
    if (Truncs)
      Truncs->push_back(ITrunc);
  }

  TPT.replaceAllUsesWith(ExtOpnd, Trunc);
  // Restore the operand of Ext (which has been replaced by the previous call
  // to replaceAllUsesWith) to avoid creating a cycle trunc <-> sext.
  TPT.setOperand(Ext, 0, ExtOpnd);
}

// Get through the Instruction:
// 1. Update its type.
// 2. Replace the uses of Ext by Inst.
// 3. Extend each operand that needs to be extended.

// Remember the original type of the instruction before promotion.
// This is useful to know that the high bits are sign extended bits.
addPromotedInst(PromotedInsts, ExtOpnd, IsSExt);
// Step #1.
TPT.mutateType(ExtOpnd, Ext->getType());
// Step #2.
TPT.replaceAllUsesWith(Ext, ExtOpnd);
// Step #3.
Instruction *ExtForOpnd = Ext;

LLVM_DEBUG(dbgs() << "Propagate Ext to operands\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Propagate Ext to operands\n"
; } } while (false);
for (int OpIdx = 0, EndOpIdx = ExtOpnd->getNumOperands(); OpIdx != EndOpIdx;
     ++OpIdx) {
  LLVM_DEBUG(dbgs() << "Operand:\n" << *(ExtOpnd->getOperand(OpIdx)) << '\n')do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Operand:\n" << *
(ExtOpnd->getOperand(OpIdx)) << '\n'; } } while (false
);
  if (ExtOpnd->getOperand(OpIdx)->getType() == Ext->getType() ||
      !shouldExtOperand(ExtOpnd, OpIdx)) {
    LLVM_DEBUG(dbgs() << "No need to propagate\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "No need to propagate\n"
; } } while (false);
    continue;
  }
  // Check if we can statically extend the operand.
  Value *Opnd = ExtOpnd->getOperand(OpIdx);
  if (const ConstantInt *Cst = dyn_cast<ConstantInt>(Opnd)) {
    LLVM_DEBUG(dbgs() << "Statically extend\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Statically extend\n"; }
 } while (false);
    unsigned BitWidth = Ext->getType()->getIntegerBitWidth();
    APInt CstVal = IsSExt ? Cst->getValue().sext(BitWidth)
                          : Cst->getValue().zext(BitWidth);
    TPT.setOperand(ExtOpnd, OpIdx, ConstantInt::get(Ext->getType(), CstVal));
    continue;
  }
  // UndefValue are typed, so we have to statically sign extend them.
  if (isa<UndefValue>(Opnd)) {
    LLVM_DEBUG(dbgs() << "Statically extend\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Statically extend\n"; }
 } while (false);
    TPT.setOperand(ExtOpnd, OpIdx, UndefValue::get(Ext->getType()));
    continue;
  }

  // Otherwise we have to explicitly sign extend the operand.
  // Check if Ext was reused to extend an operand.
  if (!ExtForOpnd) {
    // If yes, create a new one.
    LLVM_DEBUG(dbgs() << "More operands to ext\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "More operands to ext\n"
; } } while (false);
    Value *ValForExtOpnd = IsSExt ? TPT.createSExt(Ext, Opnd, Ext->getType())
                                  : TPT.createZExt(Ext, Opnd, Ext->getType());
    if (!isa<Instruction>(ValForExtOpnd)) {
      TPT.setOperand(ExtOpnd, OpIdx, ValForExtOpnd);
      continue;
    }
    ExtForOpnd = cast<Instruction>(ValForExtOpnd);
  }
  if (Exts)
    Exts->push_back(ExtForOpnd);
  TPT.setOperand(ExtForOpnd, 0, Opnd);

  // Move the sign extension before the insertion point.
  TPT.moveBefore(ExtForOpnd, ExtOpnd);
  TPT.setOperand(ExtOpnd, OpIdx, ExtForOpnd);
  CreatedInstsCost += !TLI.isExtFree(ExtForOpnd);
  // If more sext are required, new instructions will have to be created.
  ExtForOpnd = nullptr;
}
if (ExtForOpnd == Ext) {
  LLVM_DEBUG(dbgs() << "Extension is useless now\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Extension is useless now\n"
; } } while (false);
  TPT.eraseInstruction(Ext);
}
return ExtOpnd;
4533}

4535/// Check whether or not promoting an instruction to a wider type is profitable.
4536/// \p NewCost gives the cost of extension instructions created by the
4537/// promotion.
4538/// \p OldCost gives the cost of extension instructions before the promotion
4539/// plus the number of instructions that have been
4540/// matched in the addressing mode the promotion.
4541/// \p PromotedOperand is the value that has been promoted.
4542/// \return True if the promotion is profitable, false otherwise.
4543bool AddressingModeMatcher::isPromotionProfitable(
  unsigned NewCost, unsigned OldCost, Value *PromotedOperand) const {
LLVM_DEBUG(dbgs() << "OldCost: " << OldCost << "\tNewCost: " << NewCostdo { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "OldCost: " << OldCost
 << "\tNewCost: " << NewCost << '\n'; } } while
 (false)
                  << '\n')do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "OldCost: " << OldCost
 << "\tNewCost: " << NewCost << '\n'; } } while
 (false);
// The cost of the new extensions is greater than the cost of the
// old extension plus what we folded.
// This is not profitable.
if (NewCost > OldCost)
  return false;
if (NewCost < OldCost)
  return true;
// The promotion is neutral but it may help folding the sign extension in
// loads for instance.
// Check that we did not create an illegal instruction.
return isPromotedInstructionLegal(TLI, DL, PromotedOperand);
4558}

4560/// Given an instruction or constant expr, see if we can fold the operation
4561/// into the addressing mode. If so, update the addressing mode and return
4562/// true, otherwise return false without modifying AddrMode.
4563/// If \p MovedAway is not NULL, it contains the information of whether or
4564/// not AddrInst has to be folded into the addressing mode on success.
4565/// If \p MovedAway == true, \p AddrInst will not be part of the addressing
4566/// because it has been moved away.
4567/// Thus AddrInst must not be added in the matched instructions.
4568/// This state can happen when AddrInst is a sext, since it may be moved away.
4569/// Therefore, AddrInst may not be valid when MovedAway is true and it must
4570/// not be referenced anymore.
4571bool AddressingModeMatcher::matchOperationAddr(User *AddrInst, unsigned Opcode,
                                             unsigned Depth,
                                             bool *MovedAway) {
// Avoid exponential behavior on extremely deep expression trees.
if (Depth >= 5)
  return false;

// By default, all matched instructions stay in place.
if (MovedAway)
  *MovedAway = false;

switch (Opcode) {
case Instruction::PtrToInt:
  // PtrToInt is always a noop, as we know that the int type is pointer sized.
  return matchAddr(AddrInst->getOperand(0), Depth);
case Instruction::IntToPtr: {
  auto AS = AddrInst->getType()->getPointerAddressSpace();
  auto PtrTy = MVT::getIntegerVT(DL.getPointerSizeInBits(AS));
  // This inttoptr is a no-op if the integer type is pointer sized.
  if (TLI.getValueType(DL, AddrInst->getOperand(0)->getType()) == PtrTy)
    return matchAddr(AddrInst->getOperand(0), Depth);
  return false;
}
case Instruction::BitCast:
  // BitCast is always a noop, and we can handle it as long as it is
  // int->int or pointer->pointer (we don't want int<->fp or something).
  if (AddrInst->getOperand(0)->getType()->isIntOrPtrTy() &&
      // Don't touch identity bitcasts.  These were probably put here by LSR,
      // and we don't want to mess around with them.  Assume it knows what it
      // is doing.
      AddrInst->getOperand(0)->getType() != AddrInst->getType())
    return matchAddr(AddrInst->getOperand(0), Depth);
  return false;
case Instruction::AddrSpaceCast: {
  unsigned SrcAS =
      AddrInst->getOperand(0)->getType()->getPointerAddressSpace();
  unsigned DestAS = AddrInst->getType()->getPointerAddressSpace();
  if (TLI.getTargetMachine().isNoopAddrSpaceCast(SrcAS, DestAS))
    return matchAddr(AddrInst->getOperand(0), Depth);
  return false;
}
case Instruction::Add: {
  // Check to see if we can merge in the RHS then the LHS.  If so, we win.
  ExtAddrMode BackupAddrMode = AddrMode;
  unsigned OldSize = AddrModeInsts.size();
  // Start a transaction at this point.
  // The LHS may match but not the RHS.
  // Therefore, we need a higher level restoration point to undo partially
  // matched operation.
  TypePromotionTransaction::ConstRestorationPt LastKnownGood =
      TPT.getRestorationPoint();

  AddrMode.InBounds = false;
  if (matchAddr(AddrInst->getOperand(1), Depth + 1) &&
      matchAddr(AddrInst->getOperand(0), Depth + 1))
    return true;

  // Restore the old addr mode info.
  AddrMode = BackupAddrMode;
  AddrModeInsts.resize(OldSize);
  TPT.rollback(LastKnownGood);

  // Otherwise this was over-aggressive.  Try merging in the LHS then the RHS.
  if (matchAddr(AddrInst->getOperand(0), Depth + 1) &&
      matchAddr(AddrInst->getOperand(1), Depth + 1))
    return true;

  // Otherwise we definitely can't merge the ADD in.
  AddrMode = BackupAddrMode;
  AddrModeInsts.resize(OldSize);
  TPT.rollback(LastKnownGood);
  break;
}
// case Instruction::Or:
//  TODO: We can handle "Or Val, Imm" iff this OR is equivalent to an ADD.
// break;
case Instruction::Mul:
case Instruction::Shl: {
  // Can only handle X*C and X << C.
  AddrMode.InBounds = false;
  ConstantInt *RHS = dyn_cast<ConstantInt>(AddrInst->getOperand(1));
  if (!RHS || RHS->getBitWidth() > 64)
    return false;
  int64_t Scale = Opcode == Instruction::Shl
                      ? 1LL << RHS->getLimitedValue(RHS->getBitWidth() - 1)
                      : RHS->getSExtValue();

  return matchScaledValue(AddrInst->getOperand(0), Scale, Depth);
}
case Instruction::GetElementPtr: {
  // Scan the GEP.  We check it if it contains constant offsets and at most
  // one variable offset.
  int VariableOperand = -1;
  unsigned VariableScale = 0;

  int64_t ConstantOffset = 0;
  gep_type_iterator GTI = gep_type_begin(AddrInst);
  for (unsigned i = 1, e = AddrInst->getNumOperands(); i != e; ++i, ++GTI) {
    if (StructType *STy = GTI.getStructTypeOrNull()) {
      const StructLayout *SL = DL.getStructLayout(STy);
      unsigned Idx =
          cast<ConstantInt>(AddrInst->getOperand(i))->getZExtValue();
      ConstantOffset += SL->getElementOffset(Idx);
    } else {
      TypeSize TS = DL.getTypeAllocSize(GTI.getIndexedType());
      if (TS.isNonZero()) {
        // The optimisations below currently only work for fixed offsets.
        if (TS.isScalable())
          return false;
        int64_t TypeSize = TS.getFixedValue();
        if (ConstantInt *CI =
                dyn_cast<ConstantInt>(AddrInst->getOperand(i))) {
          const APInt &CVal = CI->getValue();
          if (CVal.getSignificantBits() <= 64) {
            ConstantOffset += CVal.getSExtValue() * TypeSize;
            continue;
          }
        }
        // We only allow one variable index at the moment.
        if (VariableOperand != -1)
          return false;

        // Remember the variable index.
        VariableOperand = i;
        VariableScale = TypeSize;
      }
    }
  }

  // A common case is for the GEP to only do a constant offset.  In this case,
  // just add it to the disp field and check validity.
  if (VariableOperand == -1) {
    AddrMode.BaseOffs += ConstantOffset;
    if (ConstantOffset == 0 ||
        TLI.isLegalAddressingMode(DL, AddrMode, AccessTy, AddrSpace)) {
      // Check to see if we can fold the base pointer in too.
      if (matchAddr(AddrInst->getOperand(0), Depth + 1)) {
        if (!cast<GEPOperator>(AddrInst)->isInBounds())
          AddrMode.InBounds = false;
        return true;
      }
    } else if (EnableGEPOffsetSplit && isa<GetElementPtrInst>(AddrInst) &&
               TLI.shouldConsiderGEPOffsetSplit() && Depth == 0 &&
               ConstantOffset > 0) {
      // Record GEPs with non-zero offsets as candidates for splitting in the
      // event that the offset cannot fit into the r+i addressing mode.
      // Simple and common case that only one GEP is used in calculating the
      // address for the memory access.
      Value *Base = AddrInst->getOperand(0);
      auto *BaseI = dyn_cast<Instruction>(Base);
      auto *GEP = cast<GetElementPtrInst>(AddrInst);
      if (isa<Argument>(Base) || isa<GlobalValue>(Base) ||
          (BaseI && !isa<CastInst>(BaseI) &&
           !isa<GetElementPtrInst>(BaseI))) {
        // Make sure the parent block allows inserting non-PHI instructions
        // before the terminator.
        BasicBlock *Parent =
            BaseI ? BaseI->getParent() : &GEP->getFunction()->getEntryBlock();
        if (!Parent->getTerminator()->isEHPad())
          LargeOffsetGEP = std::make_pair(GEP, ConstantOffset);
      }
    }
    AddrMode.BaseOffs -= ConstantOffset;
    return false;
  }

  // Save the valid addressing mode in case we can't match.
  ExtAddrMode BackupAddrMode = AddrMode;
  unsigned OldSize = AddrModeInsts.size();

  // See if the scale and offset amount is valid for this target.
  AddrMode.BaseOffs += ConstantOffset;
  if (!cast<GEPOperator>(AddrInst)->isInBounds())
    AddrMode.InBounds = false;

  // Match the base operand of the GEP.
  if (!matchAddr(AddrInst->getOperand(0), Depth + 1)) {
    // If it couldn't be matched, just stuff the value in a register.
    if (AddrMode.HasBaseReg) {
      AddrMode = BackupAddrMode;
      AddrModeInsts.resize(OldSize);
      return false;
    }
    AddrMode.HasBaseReg = true;
    AddrMode.BaseReg = AddrInst->getOperand(0);
  }

  // Match the remaining variable portion of the GEP.
  if (!matchScaledValue(AddrInst->getOperand(VariableOperand), VariableScale,
                        Depth)) {
    // If it couldn't be matched, try stuffing the base into a register
    // instead of matching it, and retrying the match of the scale.
    AddrMode = BackupAddrMode;
    AddrModeInsts.resize(OldSize);
    if (AddrMode.HasBaseReg)
      return false;
    AddrMode.HasBaseReg = true;
    AddrMode.BaseReg = AddrInst->getOperand(0);
    AddrMode.BaseOffs += ConstantOffset;
    if (!matchScaledValue(AddrInst->getOperand(VariableOperand),
                          VariableScale, Depth)) {
      // If even that didn't work, bail.
      AddrMode = BackupAddrMode;
      AddrModeInsts.resize(OldSize);
      return false;
    }
  }

  return true;
}
case Instruction::SExt:
case Instruction::ZExt: {
  Instruction *Ext = dyn_cast<Instruction>(AddrInst);
  if (!Ext)
    return false;

  // Try to move this ext out of the way of the addressing mode.
  // Ask for a method for doing so.
  TypePromotionHelper::Action TPH =
      TypePromotionHelper::getAction(Ext, InsertedInsts, TLI, PromotedInsts);
  if (!TPH)
    return false;

  TypePromotionTransaction::ConstRestorationPt LastKnownGood =
      TPT.getRestorationPoint();
  unsigned CreatedInstsCost = 0;
  unsigned ExtCost = !TLI.isExtFree(Ext);
  Value *PromotedOperand =
      TPH(Ext, TPT, PromotedInsts, CreatedInstsCost, nullptr, nullptr, TLI);
  // SExt has been moved away.
  // Thus either it will be rematched later in the recursive calls or it is
  // gone. Anyway, we must not fold it into the addressing mode at this point.
  // E.g.,
  // op = add opnd, 1
  // idx = ext op
  // addr = gep base, idx
  // is now:
  // promotedOpnd = ext opnd            <- no match here
  // op = promoted_add promotedOpnd, 1  <- match (later in recursive calls)
  // addr = gep base, op                <- match
  if (MovedAway)
    *MovedAway = true;

  assert(PromotedOperand &&(static_cast <bool> (PromotedOperand && "TypePromotionHelper should have filtered out those cases"
) ? void (0) : __assert_fail ("PromotedOperand && \"TypePromotionHelper should have filtered out those cases\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 4815, __extension__ __PRETTY_FUNCTION__
))
         "TypePromotionHelper should have filtered out those cases")(static_cast <bool> (PromotedOperand && "TypePromotionHelper should have filtered out those cases"
) ? void (0) : __assert_fail ("PromotedOperand && \"TypePromotionHelper should have filtered out those cases\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 4815, __extension__ __PRETTY_FUNCTION__
));

  ExtAddrMode BackupAddrMode = AddrMode;
  unsigned OldSize = AddrModeInsts.size();

  if (!matchAddr(PromotedOperand, Depth) ||
      // The total of the new cost is equal to the cost of the created
      // instructions.
      // The total of the old cost is equal to the cost of the extension plus
      // what we have saved in the addressing mode.
      !isPromotionProfitable(CreatedInstsCost,
                             ExtCost + (AddrModeInsts.size() - OldSize),
                             PromotedOperand)) {
    AddrMode = BackupAddrMode;
    AddrModeInsts.resize(OldSize);
    LLVM_DEBUG(dbgs() << "Sign extension does not pay off: rollback\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "Sign extension does not pay off: rollback\n"
; } } while (false);
    TPT.rollback(LastKnownGood);
    return false;
  }
  return true;
}
}
return false;
4838}

4840/// If we can, try to add the value of 'Addr' into the current addressing mode.
4841/// If Addr can't be added to AddrMode this returns false and leaves AddrMode
4842/// unmodified. This assumes that Addr is either a pointer type or intptr_t
4843/// for the target.
4844///
4845bool AddressingModeMatcher::matchAddr(Value *Addr, unsigned Depth) {
// Start a transaction at this point that we will rollback if the matching
// fails.
TypePromotionTransaction::ConstRestorationPt LastKnownGood =
    TPT.getRestorationPoint();
if (ConstantInt *CI = dyn_cast<ConstantInt>(Addr)) {
  if (CI->getValue().isSignedIntN(64)) {
    // Fold in immediates if legal for the target.
    AddrMode.BaseOffs += CI->getSExtValue();
    if (TLI.isLegalAddressingMode(DL, AddrMode, AccessTy, AddrSpace))
      return true;
    AddrMode.BaseOffs -= CI->getSExtValue();
  }
} else if (GlobalValue *GV = dyn_cast<GlobalValue>(Addr)) {
  // If this is a global variable, try to fold it into the addressing mode.
  if (!AddrMode.BaseGV) {
    AddrMode.BaseGV = GV;
    if (TLI.isLegalAddressingMode(DL, AddrMode, AccessTy, AddrSpace))
      return true;
    AddrMode.BaseGV = nullptr;
  }
} else if (Instruction *I = dyn_cast<Instruction>(Addr)) {
  ExtAddrMode BackupAddrMode = AddrMode;
  unsigned OldSize = AddrModeInsts.size();

  // Check to see if it is possible to fold this operation.
  bool MovedAway = false;
  if (matchOperationAddr(I, I->getOpcode(), Depth, &MovedAway)) {
    // This instruction may have been moved away. If so, there is nothing
    // to check here.
    if (MovedAway)
      return true;
    // Okay, it's possible to fold this.  Check to see if it is actually
    // *profitable* to do so.  We use a simple cost model to avoid increasing
    // register pressure too much.
    if (I->hasOneUse() ||
        isProfitableToFoldIntoAddressingMode(I, BackupAddrMode, AddrMode)) {
      AddrModeInsts.push_back(I);
      return true;
    }

    // It isn't profitable to do this, roll back.
    AddrMode = BackupAddrMode;
    AddrModeInsts.resize(OldSize);
    TPT.rollback(LastKnownGood);
  }
} else if (ConstantExpr *CE = dyn_cast<ConstantExpr>(Addr)) {
  if (matchOperationAddr(CE, CE->getOpcode(), Depth))
    return true;
  TPT.rollback(LastKnownGood);
} else if (isa<ConstantPointerNull>(Addr)) {
  // Null pointer gets folded without affecting the addressing mode.
  return true;
}

// Worse case, the target should support [reg] addressing modes. :)
if (!AddrMode.HasBaseReg) {
  AddrMode.HasBaseReg = true;
  AddrMode.BaseReg = Addr;
  // Still check for legality in case the target supports [imm] but not [i+r].
  if (TLI.isLegalAddressingMode(DL, AddrMode, AccessTy, AddrSpace))
    return true;
  AddrMode.HasBaseReg = false;
  AddrMode.BaseReg = nullptr;
}

// If the base register is already taken, see if we can do [r+r].
if (AddrMode.Scale == 0) {
  AddrMode.Scale = 1;
  AddrMode.ScaledReg = Addr;
  if (TLI.isLegalAddressingMode(DL, AddrMode, AccessTy, AddrSpace))
    return true;
  AddrMode.Scale = 0;
  AddrMode.ScaledReg = nullptr;
}
// Couldn't match.
TPT.rollback(LastKnownGood);
return false;
4923}

4925/// Check to see if all uses of OpVal by the specified inline asm call are due
4926/// to memory operands. If so, return true, otherwise return false.
4927static bool IsOperandAMemoryOperand(CallInst *CI, InlineAsm *IA, Value *OpVal,
                                  const TargetLowering &TLI,
                                  const TargetRegisterInfo &TRI) {
const Function *F = CI->getFunction();
TargetLowering::AsmOperandInfoVector TargetConstraints =
    TLI.ParseConstraints(F->getParent()->getDataLayout(), &TRI, *CI);

for (TargetLowering::AsmOperandInfo &OpInfo : TargetConstraints) {
  // Compute the constraint code and ConstraintType to use.
  TLI.ComputeConstraintToUse(OpInfo, SDValue());

  // If this asm operand is our Value*, and if it isn't an indirect memory
  // operand, we can't fold it!  TODO: Also handle C_Address?
  if (OpInfo.CallOperandVal == OpVal &&
      (OpInfo.ConstraintType != TargetLowering::C_Memory ||
       !OpInfo.isIndirect))
    return false;
}

return true;
4947}

4949// Max number of memory uses to look at before aborting the search to conserve
4950// compile time.
4951static constexpr int MaxMemoryUsesToScan = 20;

4953/// Recursively walk all the uses of I until we find a memory use.
4954/// If we find an obviously non-foldable instruction, return true.
4955/// Add accessed addresses and types to MemoryUses.
4956static bool FindAllMemoryUses(
  Instruction *I, SmallVectorImpl<std::pair<Value *, Type *>> &MemoryUses,
  SmallPtrSetImpl<Instruction *> &ConsideredInsts, const TargetLowering &TLI,
  const TargetRegisterInfo &TRI, bool OptSize, ProfileSummaryInfo *PSI,
  BlockFrequencyInfo *BFI, int SeenInsts = 0) {
// If we already considered this instruction, we're done.
if (!ConsideredInsts.insert(I).second)
  return false;

// If this is an obviously unfoldable instruction, bail out.
if (!MightBeFoldableInst(I))
  return true;

// Loop over all the uses, recursively processing them.
for (Use &U : I->uses()) {
  // Conservatively return true if we're seeing a large number or a deep chain
  // of users. This avoids excessive compilation times in pathological cases.
  if (SeenInsts++ >= MaxMemoryUsesToScan)
    return true;

  Instruction *UserI = cast<Instruction>(U.getUser());
  if (LoadInst *LI = dyn_cast<LoadInst>(UserI)) {
    MemoryUses.push_back({U.get(), LI->getType()});
    continue;
  }

  if (StoreInst *SI = dyn_cast<StoreInst>(UserI)) {
    if (U.getOperandNo() != StoreInst::getPointerOperandIndex())
      return true; // Storing addr, not into addr.
    MemoryUses.push_back({U.get(), SI->getValueOperand()->getType()});
    continue;
  }

  if (AtomicRMWInst *RMW = dyn_cast<AtomicRMWInst>(UserI)) {
    if (U.getOperandNo() != AtomicRMWInst::getPointerOperandIndex())
      return true; // Storing addr, not into addr.
    MemoryUses.push_back({U.get(), RMW->getValOperand()->getType()});
    continue;
  }

  if (AtomicCmpXchgInst *CmpX = dyn_cast<AtomicCmpXchgInst>(UserI)) {
    if (U.getOperandNo() != AtomicCmpXchgInst::getPointerOperandIndex())
      return true; // Storing addr, not into addr.
    MemoryUses.push_back({U.get(), CmpX->getCompareOperand()->getType()});
    continue;
  }

  if (CallInst *CI = dyn_cast<CallInst>(UserI)) {
    if (CI->hasFnAttr(Attribute::Cold)) {
      // If this is a cold call, we can sink the addressing calculation into
      // the cold path.  See optimizeCallInst
      bool OptForSize =
          OptSize || llvm::shouldOptimizeForSize(CI->getParent(), PSI, BFI);
      if (!OptForSize)
        continue;
    }

    InlineAsm *IA = dyn_cast<InlineAsm>(CI->getCalledOperand());
    if (!IA)
      return true;

    // If this is a memory operand, we're cool, otherwise bail out.
    if (!IsOperandAMemoryOperand(CI, IA, I, TLI, TRI))
      return true;
    continue;
  }

  if (FindAllMemoryUses(UserI, MemoryUses, ConsideredInsts, TLI, TRI, OptSize,
                        PSI, BFI, SeenInsts))
    return true;
}

return false;
5029}

5031/// Return true if Val is already known to be live at the use site that we're
5032/// folding it into. If so, there is no cost to include it in the addressing
5033/// mode. KnownLive1 and KnownLive2 are two values that we know are live at the
5034/// instruction already.
5035bool AddressingModeMatcher::valueAlreadyLiveAtInst(Value *Val,
                                                 Value *KnownLive1,
                                                 Value *KnownLive2) {
// If Val is either of the known-live values, we know it is live!
if (Val == nullptr || Val == KnownLive1 || Val == KnownLive2)
  return true;

// All values other than instructions and arguments (e.g. constants) are live.
if (!isa<Instruction>(Val) && !isa<Argument>(Val))
  return true;

// If Val is a constant sized alloca in the entry block, it is live, this is
// true because it is just a reference to the stack/frame pointer, which is
// live for the whole function.
if (AllocaInst *AI = dyn_cast<AllocaInst>(Val))
  if (AI->isStaticAlloca())
    return true;

// Check to see if this value is already used in the memory instruction's
// block.  If so, it's already live into the block at the very least, so we
// can reasonably fold it.
return Val->isUsedInBasicBlock(MemoryInst->getParent());
5057}

5059/// It is possible for the addressing mode of the machine to fold the specified
5060/// instruction into a load or store that ultimately uses it.
5061/// However, the specified instruction has multiple uses.
5062/// Given this, it may actually increase register pressure to fold it
5063/// into the load. For example, consider this code:
5064///
5065///     X = ...
5066///     Y = X+1
5067///     use(Y)   -> nonload/store
5068///     Z = Y+1
5069///     load Z
5070///
5071/// In this case, Y has multiple uses, and can be folded into the load of Z
5072/// (yielding load [X+2]).  However, doing this will cause both "X" and "X+1" to
5073/// be live at the use(Y) line.  If we don't fold Y into load Z, we use one
5074/// fewer register.  Since Y can't be folded into "use(Y)" we don't increase the
5075/// number of computations either.
5076///
5077/// Note that this (like most of CodeGenPrepare) is just a rough heuristic.  If
5078/// X was live across 'load Z' for other reasons, we actually *would* want to
5079/// fold the addressing mode in the Z case.  This would make Y die earlier.
5080bool AddressingModeMatcher::isProfitableToFoldIntoAddressingMode(
  Instruction *I, ExtAddrMode &AMBefore, ExtAddrMode &AMAfter) {
if (IgnoreProfitability)
  return true;

// AMBefore is the addressing mode before this instruction was folded into it,
// and AMAfter is the addressing mode after the instruction was folded.  Get
// the set of registers referenced by AMAfter and subtract out those
// referenced by AMBefore: this is the set of values which folding in this
// address extends the lifetime of.
//
// Note that there are only two potential values being referenced here,
// BaseReg and ScaleReg (global addresses are always available, as are any
// folded immediates).
Value *BaseReg = AMAfter.BaseReg, *ScaledReg = AMAfter.ScaledReg;

// If the BaseReg or ScaledReg was referenced by the previous addrmode, their
// lifetime wasn't extended by adding this instruction.
if (valueAlreadyLiveAtInst(BaseReg, AMBefore.BaseReg, AMBefore.ScaledReg))
  BaseReg = nullptr;
if (valueAlreadyLiveAtInst(ScaledReg, AMBefore.BaseReg, AMBefore.ScaledReg))
  ScaledReg = nullptr;

// If folding this instruction (and it's subexprs) didn't extend any live
// ranges, we're ok with it.
if (!BaseReg && !ScaledReg)
  return true;

// If all uses of this instruction can have the address mode sunk into them,
// we can remove the addressing mode and effectively trade one live register
// for another (at worst.)  In this context, folding an addressing mode into
// the use is just a particularly nice way of sinking it.
SmallVector<std::pair<Value *, Type *>, 16> MemoryUses;
SmallPtrSet<Instruction *, 16> ConsideredInsts;
if (FindAllMemoryUses(I, MemoryUses, ConsideredInsts, TLI, TRI, OptSize, PSI,
                      BFI))
  return false; // Has a non-memory, non-foldable use!

// Now that we know that all uses of this instruction are part of a chain of
// computation involving only operations that could theoretically be folded
// into a memory use, loop over each of these memory operation uses and see
// if they could  *actually* fold the instruction.  The assumption is that
// addressing modes are cheap and that duplicating the computation involved
// many times is worthwhile, even on a fastpath. For sinking candidates
// (i.e. cold call sites), this serves as a way to prevent excessive code
// growth since most architectures have some reasonable small and fast way to
// compute an effective address.  (i.e LEA on x86)
SmallVector<Instruction *, 32> MatchedAddrModeInsts;
for (const std::pair<Value *, Type *> &Pair : MemoryUses) {
  Value *Address = Pair.first;
  Type *AddressAccessTy = Pair.second;
  unsigned AS = Address->getType()->getPointerAddressSpace();

  // Do a match against the root of this address, ignoring profitability. This
  // will tell us if the addressing mode for the memory operation will
  // *actually* cover the shared instruction.
  ExtAddrMode Result;
  std::pair<AssertingVH<GetElementPtrInst>, int64_t> LargeOffsetGEP(nullptr,
                                                                    0);
  TypePromotionTransaction::ConstRestorationPt LastKnownGood =
      TPT.getRestorationPoint();
  AddressingModeMatcher Matcher(MatchedAddrModeInsts, TLI, TRI, LI, getDTFn,
                                AddressAccessTy, AS, MemoryInst, Result,
                                InsertedInsts, PromotedInsts, TPT,
                                LargeOffsetGEP, OptSize, PSI, BFI);
  Matcher.IgnoreProfitability = true;
  bool Success = Matcher.matchAddr(Address, 0);
  (void)Success;
  assert(Success && "Couldn't select *anything*?")(static_cast <bool> (Success && "Couldn't select *anything*?"
) ? void (0) : __assert_fail ("Success && \"Couldn't select *anything*?\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 5148, __extension__ __PRETTY_FUNCTION__
));

  // The match was to check the profitability, the changes made are not
  // part of the original matcher. Therefore, they should be dropped
  // otherwise the original matcher will not present the right state.
  TPT.rollback(LastKnownGood);

  // If the match didn't cover I, then it won't be shared by it.
  if (!is_contained(MatchedAddrModeInsts, I))
    return false;

  MatchedAddrModeInsts.clear();
}

return true;
5163}

5165/// Return true if the specified values are defined in a
5166/// different basic block than BB.
5167static bool IsNonLocalValue(Value *V, BasicBlock *BB) {
if (Instruction *I = dyn_cast<Instruction>(V))
  return I->getParent() != BB;
return false;
5171}

5173/// Sink addressing mode computation immediate before MemoryInst if doing so
5174/// can be done without increasing register pressure.  The need for the
5175/// register pressure constraint means this can end up being an all or nothing
5176/// decision for all uses of the same addressing computation.
5177///
5178/// Load and Store Instructions often have addressing modes that can do
5179/// significant amounts of computation. As such, instruction selection will try
5180/// to get the load or store to do as much computation as possible for the
5181/// program. The problem is that isel can only see within a single block. As
5182/// such, we sink as much legal addressing mode work into the block as possible.
5183///
5184/// This method is used to optimize both load/store and inline asms with memory
5185/// operands.  It's also used to sink addressing computations feeding into cold
5186/// call sites into their (cold) basic block.
5187///
5188/// The motivation for handling sinking into cold blocks is that doing so can
5189/// both enable other address mode sinking (by satisfying the register pressure
5190/// constraint above), and reduce register pressure globally (by removing the
5191/// addressing mode computation from the fast path entirely.).
5192bool CodeGenPrepare::optimizeMemoryInst(Instruction *MemoryInst, Value *Addr,
                                      Type *AccessTy, unsigned AddrSpace) {
Value *Repl = Addr;

// Try to collapse single-value PHI nodes.  This is necessary to undo
// unprofitable PRE transformations.
SmallVector<Value *, 8> worklist;
SmallPtrSet<Value *, 16> Visited;
worklist.push_back(Addr);

// Use a worklist to iteratively look through PHI and select nodes, and
// ensure that the addressing mode obtained from the non-PHI/select roots of
// the graph are compatible.
bool PhiOrSelectSeen = false;
SmallVector<Instruction *, 16> AddrModeInsts;
const SimplifyQuery SQ(*DL, TLInfo);
AddressingModeCombiner AddrModes(SQ, Addr);
TypePromotionTransaction TPT(RemovedInsts);
TypePromotionTransaction::ConstRestorationPt LastKnownGood =
    TPT.getRestorationPoint();
while (!worklist.empty()) {
  Value *V = worklist.pop_back_val();

  // We allow traversing cyclic Phi nodes.
  // In case of success after this loop we ensure that traversing through
  // Phi nodes ends up with all cases to compute address of the form
  //    BaseGV + Base + Scale * Index + Offset
  // where Scale and Offset are constans and BaseGV, Base and Index
  // are exactly the same Values in all cases.
  // It means that BaseGV, Scale and Offset dominate our memory instruction
  // and have the same value as they had in address computation represented
  // as Phi. So we can safely sink address computation to memory instruction.
  if (!Visited.insert(V).second)
    continue;

  // For a PHI node, push all of its incoming values.
  if (PHINode *P = dyn_cast<PHINode>(V)) {
    append_range(worklist, P->incoming_values());
    PhiOrSelectSeen = true;
    continue;
  }
  // Similar for select.
  if (SelectInst *SI = dyn_cast<SelectInst>(V)) {
    worklist.push_back(SI->getFalseValue());
    worklist.push_back(SI->getTrueValue());
    PhiOrSelectSeen = true;
    continue;
  }

  // For non-PHIs, determine the addressing mode being computed.  Note that
  // the result may differ depending on what other uses our candidate
  // addressing instructions might have.
  AddrModeInsts.clear();
  std::pair<AssertingVH<GetElementPtrInst>, int64_t> LargeOffsetGEP(nullptr,
                                                                    0);
  // Defer the query (and possible computation of) the dom tree to point of
  // actual use.  It's expected that most address matches don't actually need
  // the domtree.
  auto getDTFn = [MemoryInst, this]() -> const DominatorTree & {
    Function *F = MemoryInst->getParent()->getParent();
    return this->getDT(*F);
  };
  ExtAddrMode NewAddrMode = AddressingModeMatcher::Match(
      V, AccessTy, AddrSpace, MemoryInst, AddrModeInsts, *TLI, *LI, getDTFn,
      *TRI, InsertedInsts, PromotedInsts, TPT, LargeOffsetGEP, OptSize, PSI,
      BFI.get());

  GetElementPtrInst *GEP = LargeOffsetGEP.first;
  if (GEP && !NewGEPBases.count(GEP)) {
    // If splitting the underlying data structure can reduce the offset of a
    // GEP, collect the GEP.  Skip the GEPs that are the new bases of
    // previously split data structures.
    LargeOffsetGEPMap[GEP->getPointerOperand()].push_back(LargeOffsetGEP);
    LargeOffsetGEPID.insert(std::make_pair(GEP, LargeOffsetGEPID.size()));
  }

  NewAddrMode.OriginalValue = V;
  if (!AddrModes.addNewAddrMode(NewAddrMode))
    break;
}

// Try to combine the AddrModes we've collected. If we couldn't collect any,
// or we have multiple but either couldn't combine them or combining them
// wouldn't do anything useful, bail out now.
if (!AddrModes.combineAddrModes()) {
  TPT.rollback(LastKnownGood);
  return false;
}
bool Modified = TPT.commit();

// Get the combined AddrMode (or the only AddrMode, if we only had one).
ExtAddrMode AddrMode = AddrModes.getAddrMode();

// If all the instructions matched are already in this BB, don't do anything.
// If we saw a Phi node then it is not local definitely, and if we saw a
// select then we want to push the address calculation past it even if it's
// already in this BB.
if (!PhiOrSelectSeen && none_of(AddrModeInsts, [&](Value *V) {
      return IsNonLocalValue(V, MemoryInst->getParent());
    })) {
  LLVM_DEBUG(dbgs() << "CGP: Found      local addrmode: " << AddrModedo { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "CGP: Found      local addrmode: "
 << AddrMode << "\n"; } } while (false)
                    << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "CGP: Found      local addrmode: "
 << AddrMode << "\n"; } } while (false);
  return Modified;
}

// Insert this computation right after this user.  Since our caller is
// scanning from the top of the BB to the bottom, reuse of the expr are
// guaranteed to happen later.
IRBuilder<> Builder(MemoryInst);

// Now that we determined the addressing expression we want to use and know
// that we have to sink it into this block.  Check to see if we have already
// done this for some other load/store instr in this block.  If so, reuse
// the computation.  Before attempting reuse, check if the address is valid
// as it may have been erased.

WeakTrackingVH SunkAddrVH = SunkAddrs[Addr];

Value *SunkAddr = SunkAddrVH.pointsToAliveValue() ? SunkAddrVH : nullptr;
Type *IntPtrTy = DL->getIntPtrType(Addr->getType());
if (SunkAddr) {
  LLVM_DEBUG(dbgs() << "CGP: Reusing nonlocal addrmode: " << AddrModedo { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "CGP: Reusing nonlocal addrmode: "
 << AddrMode << " for " << *MemoryInst <<
 "\n"; } } while (false)
                    << " for " << *MemoryInst << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "CGP: Reusing nonlocal addrmode: "
 << AddrMode << " for " << *MemoryInst <<
 "\n"; } } while (false);
  if (SunkAddr->getType() != Addr->getType()) {
    if (SunkAddr->getType()->getPointerAddressSpace() !=
            Addr->getType()->getPointerAddressSpace() &&
        !DL->isNonIntegralPointerType(Addr->getType())) {
      // There are two reasons the address spaces might not match: a no-op
      // addrspacecast, or a ptrtoint/inttoptr pair. Either way, we emit a
      // ptrtoint/inttoptr pair to ensure we match the original semantics.
      // TODO: allow bitcast between different address space pointers with the
      // same size.
      SunkAddr = Builder.CreatePtrToInt(SunkAddr, IntPtrTy, "sunkaddr");
      SunkAddr =
          Builder.CreateIntToPtr(SunkAddr, Addr->getType(), "sunkaddr");
    } else
      SunkAddr = Builder.CreatePointerCast(SunkAddr, Addr->getType());
  }
} else if (AddrSinkUsingGEPs || (!AddrSinkUsingGEPs.getNumOccurrences() &&
                                 SubtargetInfo->addrSinkUsingGEPs())) {
  // By default, we use the GEP-based method when AA is used later. This
  // prevents new inttoptr/ptrtoint pairs from degrading AA capabilities.
  LLVM_DEBUG(dbgs() << "CGP: SINKING nonlocal addrmode: " << AddrModedo { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "CGP: SINKING nonlocal addrmode: "
 << AddrMode << " for " << *MemoryInst <<
 "\n"; } } while (false)
                    << " for " << *MemoryInst << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "CGP: SINKING nonlocal addrmode: "
 << AddrMode << " for " << *MemoryInst <<
 "\n"; } } while (false);
  Value *ResultPtr = nullptr, *ResultIndex = nullptr;

  // First, find the pointer.
  if (AddrMode.BaseReg && AddrMode.BaseReg->getType()->isPointerTy()) {
    ResultPtr = AddrMode.BaseReg;
    AddrMode.BaseReg = nullptr;
  }

  if (AddrMode.Scale && AddrMode.ScaledReg->getType()->isPointerTy()) {
    // We can't add more than one pointer together, nor can we scale a
    // pointer (both of which seem meaningless).
    if (ResultPtr || AddrMode.Scale != 1)
      return Modified;

    ResultPtr = AddrMode.ScaledReg;
    AddrMode.Scale = 0;
  }

  // It is only safe to sign extend the BaseReg if we know that the math
  // required to create it did not overflow before we extend it. Since
  // the original IR value was tossed in favor of a constant back when
  // the AddrMode was created we need to bail out gracefully if widths
  // do not match instead of extending it.
  //
  // (See below for code to add the scale.)
  if (AddrMode.Scale) {
    Type *ScaledRegTy = AddrMode.ScaledReg->getType();
    if (cast<IntegerType>(IntPtrTy)->getBitWidth() >
        cast<IntegerType>(ScaledRegTy)->getBitWidth())
      return Modified;
  }

  if (AddrMode.BaseGV) {
    if (ResultPtr)
      return Modified;

    ResultPtr = AddrMode.BaseGV;
  }

  // If the real base value actually came from an inttoptr, then the matcher
  // will look through it and provide only the integer value. In that case,
  // use it here.
  if (!DL->isNonIntegralPointerType(Addr->getType())) {
    if (!ResultPtr && AddrMode.BaseReg) {
      ResultPtr = Builder.CreateIntToPtr(AddrMode.BaseReg, Addr->getType(),
                                         "sunkaddr");
      AddrMode.BaseReg = nullptr;
    } else if (!ResultPtr && AddrMode.Scale == 1) {
      ResultPtr = Builder.CreateIntToPtr(AddrMode.ScaledReg, Addr->getType(),
                                         "sunkaddr");
      AddrMode.Scale = 0;
    }
  }

  if (!ResultPtr && !AddrMode.BaseReg && !AddrMode.Scale &&
      !AddrMode.BaseOffs) {
    SunkAddr = Constant::getNullValue(Addr->getType());
  } else if (!ResultPtr) {
    return Modified;
  } else {
    Type *I8PtrTy =
        Builder.getInt8PtrTy(Addr->getType()->getPointerAddressSpace());
    Type *I8Ty = Builder.getInt8Ty();

    // Start with the base register. Do this first so that subsequent address
    // matching finds it last, which will prevent it from trying to match it
    // as the scaled value in case it happens to be a mul. That would be
    // problematic if we've sunk a different mul for the scale, because then
    // we'd end up sinking both muls.
    if (AddrMode.BaseReg) {
      Value *V = AddrMode.BaseReg;
      if (V->getType() != IntPtrTy)
        V = Builder.CreateIntCast(V, IntPtrTy, /*isSigned=*/true, "sunkaddr");

      ResultIndex = V;
    }

    // Add the scale value.
    if (AddrMode.Scale) {
      Value *V = AddrMode.ScaledReg;
      if (V->getType() == IntPtrTy) {
        // done.
      } else {
        assert(cast<IntegerType>(IntPtrTy)->getBitWidth() <(static_cast <bool> (cast<IntegerType>(IntPtrTy)->
getBitWidth() < cast<IntegerType>(V->getType())->
getBitWidth() && "We can't transform if ScaledReg is too narrow"
) ? void (0) : __assert_fail ("cast<IntegerType>(IntPtrTy)->getBitWidth() < cast<IntegerType>(V->getType())->getBitWidth() && \"We can't transform if ScaledReg is too narrow\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 5421, __extension__ __PRETTY_FUNCTION__
))
                   cast<IntegerType>(V->getType())->getBitWidth() &&(static_cast <bool> (cast<IntegerType>(IntPtrTy)->
getBitWidth() < cast<IntegerType>(V->getType())->
getBitWidth() && "We can't transform if ScaledReg is too narrow"
) ? void (0) : __assert_fail ("cast<IntegerType>(IntPtrTy)->getBitWidth() < cast<IntegerType>(V->getType())->getBitWidth() && \"We can't transform if ScaledReg is too narrow\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 5421, __extension__ __PRETTY_FUNCTION__
))
               "We can't transform if ScaledReg is too narrow")(static_cast <bool> (cast<IntegerType>(IntPtrTy)->
getBitWidth() < cast<IntegerType>(V->getType())->
getBitWidth() && "We can't transform if ScaledReg is too narrow"
) ? void (0) : __assert_fail ("cast<IntegerType>(IntPtrTy)->getBitWidth() < cast<IntegerType>(V->getType())->getBitWidth() && \"We can't transform if ScaledReg is too narrow\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 5421, __extension__ __PRETTY_FUNCTION__
));
        V = Builder.CreateTrunc(V, IntPtrTy, "sunkaddr");
      }

      if (AddrMode.Scale != 1)
        V = Builder.CreateMul(V, ConstantInt::get(IntPtrTy, AddrMode.Scale),
                              "sunkaddr");
      if (ResultIndex)
        ResultIndex = Builder.CreateAdd(ResultIndex, V, "sunkaddr");
      else
        ResultIndex = V;
    }

    // Add in the Base Offset if present.
    if (AddrMode.BaseOffs) {
      Value *V = ConstantInt::get(IntPtrTy, AddrMode.BaseOffs);
      if (ResultIndex) {
        // We need to add this separately from the scale above to help with
        // SDAG consecutive load/store merging.
        if (ResultPtr->getType() != I8PtrTy)
          ResultPtr = Builder.CreatePointerCast(ResultPtr, I8PtrTy);
        ResultPtr = Builder.CreateGEP(I8Ty, ResultPtr, ResultIndex,
                                      "sunkaddr", AddrMode.InBounds);
      }

      ResultIndex = V;
    }

    if (!ResultIndex) {
      SunkAddr = ResultPtr;
    } else {
      if (ResultPtr->getType() != I8PtrTy)
        ResultPtr = Builder.CreatePointerCast(ResultPtr, I8PtrTy);
      SunkAddr = Builder.CreateGEP(I8Ty, ResultPtr, ResultIndex, "sunkaddr",
                                   AddrMode.InBounds);
    }

    if (SunkAddr->getType() != Addr->getType()) {
      if (SunkAddr->getType()->getPointerAddressSpace() !=
              Addr->getType()->getPointerAddressSpace() &&
          !DL->isNonIntegralPointerType(Addr->getType())) {
        // There are two reasons the address spaces might not match: a no-op
        // addrspacecast, or a ptrtoint/inttoptr pair. Either way, we emit a
        // ptrtoint/inttoptr pair to ensure we match the original semantics.
        // TODO: allow bitcast between different address space pointers with
        // the same size.
        SunkAddr = Builder.CreatePtrToInt(SunkAddr, IntPtrTy, "sunkaddr");
        SunkAddr =
            Builder.CreateIntToPtr(SunkAddr, Addr->getType(), "sunkaddr");
      } else
        SunkAddr = Builder.CreatePointerCast(SunkAddr, Addr->getType());
    }
  }
} else {
  // We'd require a ptrtoint/inttoptr down the line, which we can't do for
  // non-integral pointers, so in that case bail out now.
  Type *BaseTy = AddrMode.BaseReg ? AddrMode.BaseReg->getType() : nullptr;
  Type *ScaleTy = AddrMode.Scale ? AddrMode.ScaledReg->getType() : nullptr;
  PointerType *BasePtrTy = dyn_cast_or_null<PointerType>(BaseTy);
  PointerType *ScalePtrTy = dyn_cast_or_null<PointerType>(ScaleTy);
  if (DL->isNonIntegralPointerType(Addr->getType()) ||
      (BasePtrTy && DL->isNonIntegralPointerType(BasePtrTy)) ||
      (ScalePtrTy && DL->isNonIntegralPointerType(ScalePtrTy)) ||
      (AddrMode.BaseGV &&
       DL->isNonIntegralPointerType(AddrMode.BaseGV->getType())))
    return Modified;

  LLVM_DEBUG(dbgs() << "CGP: SINKING nonlocal addrmode: " << AddrModedo { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "CGP: SINKING nonlocal addrmode: "
 << AddrMode << " for " << *MemoryInst <<
 "\n"; } } while (false)
                    << " for " << *MemoryInst << "\n")do { if (::llvm::DebugFlag && ::llvm::isCurrentDebugType
("codegenprepare")) { dbgs() << "CGP: SINKING nonlocal addrmode: "
 << AddrMode << " for " << *MemoryInst <<
 "\n"; } } while (false);
  Type *IntPtrTy = DL->getIntPtrType(Addr->getType());
  Value *Result = nullptr;

  // Start with the base register. Do this first so that subsequent address
  // matching finds it last, which will prevent it from trying to match it
  // as the scaled value in case it happens to be a mul. That would be
  // problematic if we've sunk a different mul for the scale, because then
  // we'd end up sinking both muls.
  if (AddrMode.BaseReg) {
    Value *V = AddrMode.BaseReg;
    if (V->getType()->isPointerTy())
      V = Builder.CreatePtrToInt(V, IntPtrTy, "sunkaddr");
    if (V->getType() != IntPtrTy)
      V = Builder.CreateIntCast(V, IntPtrTy, /*isSigned=*/true, "sunkaddr");
    Result = V;
  }

  // Add the scale value.
  if (AddrMode.Scale) {
    Value *V = AddrMode.ScaledReg;
    if (V->getType() == IntPtrTy) {
      // done.
    } else if (V->getType()->isPointerTy()) {
      V = Builder.CreatePtrToInt(V, IntPtrTy, "sunkaddr");
    } else if (cast<IntegerType>(IntPtrTy)->getBitWidth() <
               cast<IntegerType>(V->getType())->getBitWidth()) {
      V = Builder.CreateTrunc(V, IntPtrTy, "sunkaddr");
    } else {
      // It is only safe to sign extend the BaseReg if we know that the math
      // required to create it did not overflow before we extend it. Since
      // the original IR value was tossed in favor of a constant back when
      // the AddrMode was created we need to bail out gracefully if widths
      // do not match instead of extending it.
      Instruction *I = dyn_cast_or_null<Instruction>(Result);
      if (I && (Result != AddrMode.BaseReg))
        I->eraseFromParent();
      return Modified;
    }
    if (AddrMode.Scale != 1)
      V = Builder.CreateMul(V, ConstantInt::get(IntPtrTy, AddrMode.Scale),
                            "sunkaddr");
    if (Result)
      Result = Builder.CreateAdd(Result, V, "sunkaddr");
    else
      Result = V;
  }

  // Add in the BaseGV if present.
  if (AddrMode.BaseGV) {
    Value *V = Builder.CreatePtrToInt(AddrMode.BaseGV, IntPtrTy, "sunkaddr");
    if (Result)
      Result = Builder.CreateAdd(Result, V, "sunkaddr");
    else
      Result = V;
  }

  // Add in the Base Offset if present.
  if (AddrMode.BaseOffs) {
    Value *V = ConstantInt::get(IntPtrTy, AddrMode.BaseOffs);
    if (Result)
      Result = Builder.CreateAdd(Result, V, "sunkaddr");
    else
      Result = V;
  }

  if (!Result)
    SunkAddr = Constant::getNullValue(Addr->getType());
  else
    SunkAddr = Builder.CreateIntToPtr(Result, Addr->getType(), "sunkaddr");
}

MemoryInst->replaceUsesOfWith(Repl, SunkAddr);
// Store the newly computed address into the cache. In the case we reused a
// value, this should be idempotent.
SunkAddrs[Addr] = WeakTrackingVH(SunkAddr);

// If we have no uses, recursively delete the value and all dead instructions
// using it.
if (Repl->use_empty()) {
  resetIteratorIfInvalidatedWhileCalling(CurInstIterator->getParent(), [&]() {
    RecursivelyDeleteTriviallyDeadInstructions(
        Repl, TLInfo, nullptr,
        [&](Value *V) { removeAllAssertingVHReferences(V); });
  });
}
++NumMemoryInsts;
return true;
5577}

5579/// Rewrite GEP input to gather/scatter to enable SelectionDAGBuilder to find
5580/// a uniform base to use for ISD::MGATHER/MSCATTER. SelectionDAGBuilder can
5581/// only handle a 2 operand GEP in the same basic block or a splat constant
5582/// vector. The 2 operands to the GEP must have a scalar pointer and a vector
5583/// index.
5584///
5585/// If the existing GEP has a vector base pointer that is splat, we can look
5586/// through the splat to find the scalar pointer. If we can't find a scalar
5587/// pointer there's nothing we can do.
5588///
5589/// If we have a GEP with more than 2 indices where the middle indices are all
5590/// zeroes, we can replace it with 2 GEPs where the second has 2 operands.
5591///
5592/// If the final index isn't a vector or is a splat, we can emit a scalar GEP
5593/// followed by a GEP with an all zeroes vector index. This will enable
5594/// SelectionDAGBuilder to use the scalar GEP as the uniform base and have a
5595/// zero index.
5596bool CodeGenPrepare::optimizeGatherScatterInst(Instruction *MemoryInst,
                                             Value *Ptr) {
Value *NewAddr;

if (const auto *GEP = dyn_cast<GetElementPtrInst>(Ptr)) {
  // Don't optimize GEPs that don't have indices.
  if (!GEP->hasIndices())
    return false;

  // If the GEP and the gather/scatter aren't in the same BB, don't optimize.
  // FIXME: We should support this by sinking the GEP.
  if (MemoryInst->getParent() != GEP->getParent())
    return false;

  SmallVector<Value *, 2> Ops(GEP->operands());

  bool RewriteGEP = false;

  if (Ops[0]->getType()->isVectorTy()) {
    Ops[0] = getSplatValue(Ops[0]);
    if (!Ops[0])
      return false;
    RewriteGEP = true;
  }

  unsigned FinalIndex = Ops.size() - 1;

  // Ensure all but the last index is 0.
  // FIXME: This isn't strictly required. All that's required is that they are
  // all scalars or splats.
  for (unsigned i = 1; i < FinalIndex; ++i) {
    auto *C = dyn_cast<Constant>(Ops[i]);
    if (!C)
      return false;
    if (isa<VectorType>(C->getType()))
      C = C->getSplatValue();
    auto *CI = dyn_cast_or_null<ConstantInt>(C);
    if (!CI || !CI->isZero())
      return false;
    // Scalarize the index if needed.
    Ops[i] = CI;
  }

  // Try to scalarize the final index.
  if (Ops[FinalIndex]->getType()->isVectorTy()) {
    if (Value *V = getSplatValue(Ops[FinalIndex])) {
      auto *C = dyn_cast<ConstantInt>(V);
      // Don't scalarize all zeros vector.
      if (!C || !C->isZero()) {
        Ops[FinalIndex] = V;
        RewriteGEP = true;
      }
    }
  }

  // If we made any changes or the we have extra operands, we need to generate
  // new instructions.
  if (!RewriteGEP && Ops.size() == 2)
    return false;

  auto NumElts = cast<VectorType>(Ptr->getType())->getElementCount();

  IRBuilder<> Builder(MemoryInst);

  Type *SourceTy = GEP->getSourceElementType();
  Type *ScalarIndexTy = DL->getIndexType(Ops[0]->getType()->getScalarType());

  // If the final index isn't a vector, emit a scalar GEP containing all ops
  // and a vector GEP with all zeroes final index.
  if (!Ops[FinalIndex]->getType()->isVectorTy()) {
    NewAddr = Builder.CreateGEP(SourceTy, Ops[0], ArrayRef(Ops).drop_front());
    auto *IndexTy = VectorType::get(ScalarIndexTy, NumElts);
    auto *SecondTy = GetElementPtrInst::getIndexedType(
        SourceTy, ArrayRef(Ops).drop_front());
    NewAddr =
        Builder.CreateGEP(SecondTy, NewAddr, Constant::getNullValue(IndexTy));
  } else {
    Value *Base = Ops[0];
    Value *Index = Ops[FinalIndex];

    // Create a scalar GEP if there are more than 2 operands.
    if (Ops.size() != 2) {
      // Replace the last index with 0.
      Ops[FinalIndex] = Constant::getNullValue(ScalarIndexTy);
      Base = Builder.CreateGEP(SourceTy, Base, ArrayRef(Ops).drop_front());
      SourceTy = GetElementPtrInst::getIndexedType(
          SourceTy, ArrayRef(Ops).drop_front());
    }

    // Now create the GEP with scalar pointer and vector index.
    NewAddr = Builder.CreateGEP(SourceTy, Base, Index);
  }
} else if (!isa<Constant>(Ptr)) {
  // Not a GEP, maybe its a splat and we can create a GEP to enable
  // SelectionDAGBuilder to use it as a uniform base.
  Value *V = getSplatValue(Ptr);
  if (!V)
    return false;

  auto NumElts = cast<VectorType>(Ptr->getType())->getElementCount();

  IRBuilder<> Builder(MemoryInst);

  // Emit a vector GEP with a scalar pointer and all 0s vector index.
  Type *ScalarIndexTy = DL->getIndexType(V->getType()->getScalarType());
  auto *IndexTy = VectorType::get(ScalarIndexTy, NumElts);
  Type *ScalarTy;
  if (cast<IntrinsicInst>(MemoryInst)->getIntrinsicID() ==
      Intrinsic::masked_gather) {
    ScalarTy = MemoryInst->getType()->getScalarType();
  } else {
    assert(cast<IntrinsicInst>(MemoryInst)->getIntrinsicID() ==(static_cast <bool> (cast<IntrinsicInst>(MemoryInst
)->getIntrinsicID() == Intrinsic::masked_scatter) ? void (
0) : __assert_fail ("cast<IntrinsicInst>(MemoryInst)->getIntrinsicID() == Intrinsic::masked_scatter"
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 5708, __extension__ __PRETTY_FUNCTION__
))
           Intrinsic::masked_scatter)(static_cast <bool> (cast<IntrinsicInst>(MemoryInst
)->getIntrinsicID() == Intrinsic::masked_scatter) ? void (
0) : __assert_fail ("cast<IntrinsicInst>(MemoryInst)->getIntrinsicID() == Intrinsic::masked_scatter"
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 5708, __extension__ __PRETTY_FUNCTION__
));
    ScalarTy = MemoryInst->getOperand(0)->getType()->getScalarType();
  }
  NewAddr = Builder.CreateGEP(ScalarTy, V, Constant::getNullValue(IndexTy));
} else {
  // Constant, SelectionDAGBuilder knows to check if its a splat.
  return false;
}

MemoryInst->replaceUsesOfWith(Ptr, NewAddr);

// If we have no uses, recursively delete the value and all dead instructions
// using it.
if (Ptr->use_empty())
  RecursivelyDeleteTriviallyDeadInstructions(
      Ptr, TLInfo, nullptr,
      [&](Value *V) { removeAllAssertingVHReferences(V); });

return true;
5727}

5729/// If there are any memory operands, use OptimizeMemoryInst to sink their
5730/// address computing into the block when possible / profitable.
5731bool CodeGenPrepare::optimizeInlineAsmInst(CallInst *CS) {
bool MadeChange = false;

const TargetRegisterInfo *TRI =
    TM->getSubtargetImpl(*CS->getFunction())->getRegisterInfo();
TargetLowering::AsmOperandInfoVector TargetConstraints =
    TLI->ParseConstraints(*DL, TRI, *CS);
unsigned ArgNo = 0;
for (TargetLowering::AsmOperandInfo &OpInfo : TargetConstraints) {
  // Compute the constraint code and ConstraintType to use.
  TLI->ComputeConstraintToUse(OpInfo, SDValue());

  // TODO: Also handle C_Address?
  if (OpInfo.ConstraintType == TargetLowering::C_Memory &&
      OpInfo.isIndirect) {
    Value *OpVal = CS->getArgOperand(ArgNo++);
    MadeChange |= optimizeMemoryInst(CS, OpVal, OpVal->getType(), ~0u);
  } else if (OpInfo.Type == InlineAsm::isInput)
    ArgNo++;
}

return MadeChange;
5753}

5755/// Check if all the uses of \p Val are equivalent (or free) zero or
5756/// sign extensions.
5757static bool hasSameExtUse(Value *Val, const TargetLowering &TLI) {
assert(!Val->use_empty() && "Input must have at least one use")(static_cast <bool> (!Val->use_empty() && "Input must have at least one use"
) ? void (0) : __assert_fail ("!Val->use_empty() && \"Input must have at least one use\""
, "llvm/lib/CodeGen/CodeGenPrepare.cpp", 5758, __extension__ __PRETTY_FUNCTION__
));
const Instruction *FirstUser = cast<Instruction>(*Val->user_begin());
bool IsSExt = isa<SExtInst>(FirstUser);
Type *ExtTy = FirstUser->getType();
for (const User *U : Val->users()) {
  const Instruction *UI = cast<Instruction>(U);
  if ((IsSExt && !isa<SExtInst>(UI)) || (!IsSExt && !isa<ZExtInst>(UI)))
    return false;
  Type *CurTy = UI->getType();
  // Same input and output types: Same instruction after CSE.
  if (CurTy == ExtTy)
    continue;

  // If IsSExt is true, we are in this situation:
  // a = Val
  // b = sext ty1 a to ty2
  // c = sext ty1 a to ty3
  // Assuming ty2 is shorter than ty3, this could be turned into:
  // a = Val
  // b = sext ty1 a to ty2
  // c = sext ty2 b to ty3
  // However, the last sext is not free.
  if (IsSExt)
    return false;

  // This is a ZExt, maybe this is free to extend from one type to another.
  // In that case, we would not account for a different use.
  Type *NarrowTy;
  Type *LargeTy;
  if (ExtTy->getScalarType()->getIntegerBitWidth() >
      CurTy->getScalarType()->getIntegerBitWidth()) {
    NarrowTy = CurTy;
    LargeTy = ExtTy;
  } else {
    NarrowTy = ExtTy;