LLVM  14.0.0git
LoopVectorize.h
Go to the documentation of this file.
1 //===- LoopVectorize.h ------------------------------------------*- C++ -*-===//
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 is the LLVM loop vectorizer. This pass modifies 'vectorizable' loops
10 // and generates target-independent LLVM-IR.
11 // The vectorizer uses the TargetTransformInfo analysis to estimate the costs
12 // of instructions in order to estimate the profitability of vectorization.
13 //
14 // The loop vectorizer combines consecutive loop iterations into a single
15 // 'wide' iteration. After this transformation the index is incremented
16 // by the SIMD vector width, and not by one.
17 //
18 // This pass has three parts:
19 // 1. The main loop pass that drives the different parts.
20 // 2. LoopVectorizationLegality - A unit that checks for the legality
21 // of the vectorization.
22 // 3. InnerLoopVectorizer - A unit that performs the actual
23 // widening of instructions.
24 // 4. LoopVectorizationCostModel - A unit that checks for the profitability
25 // of vectorization. It decides on the optimal vector width, which
26 // can be one, if vectorization is not profitable.
27 //
28 // There is a development effort going on to migrate loop vectorizer to the
29 // VPlan infrastructure and to introduce outer loop vectorization support (see
30 // docs/Proposal/VectorizationPlan.rst and
31 // http://lists.llvm.org/pipermail/llvm-dev/2017-December/119523.html). For this
32 // purpose, we temporarily introduced the VPlan-native vectorization path: an
33 // alternative vectorization path that is natively implemented on top of the
34 // VPlan infrastructure. See EnableVPlanNativePath for enabling.
35 //
36 //===----------------------------------------------------------------------===//
37 //
38 // The reduction-variable vectorization is based on the paper:
39 // D. Nuzman and R. Henderson. Multi-platform Auto-vectorization.
40 //
41 // Variable uniformity checks are inspired by:
42 // Karrenberg, R. and Hack, S. Whole Function Vectorization.
43 //
44 // The interleaved access vectorization is based on the paper:
45 // Dorit Nuzman, Ira Rosen and Ayal Zaks. Auto-Vectorization of Interleaved
46 // Data for SIMD
47 //
48 // Other ideas/concepts are from:
49 // A. Zaks and D. Nuzman. Autovectorization in GCC-two years later.
50 //
51 // S. Maleki, Y. Gao, M. Garzaran, T. Wong and D. Padua. An Evaluation of
52 // Vectorizing Compilers.
53 //
54 //===----------------------------------------------------------------------===//
55 
56 #ifndef LLVM_TRANSFORMS_VECTORIZE_LOOPVECTORIZE_H
57 #define LLVM_TRANSFORMS_VECTORIZE_LOOPVECTORIZE_H
58 
59 #include "llvm/IR/PassManager.h"
61 #include <functional>
62 
63 namespace llvm {
64 
65 class AAResults;
66 class AssumptionCache;
67 class BlockFrequencyInfo;
68 class DemandedBits;
69 class DominatorTree;
70 class Function;
71 class Loop;
72 class LoopAccessInfo;
73 class LoopInfo;
74 class OptimizationRemarkEmitter;
75 class ProfileSummaryInfo;
76 class ScalarEvolution;
77 class TargetLibraryInfo;
78 class TargetTransformInfo;
79 
80 extern cl::opt<bool> EnableLoopInterleaving;
81 extern cl::opt<bool> EnableLoopVectorization;
82 
84  /// If false, consider all loops for interleaving.
85  /// If true, only loops that explicitly request interleaving are considered.
87 
88  /// If false, consider all loops for vectorization.
89  /// If true, only loops that explicitly request vectorization are considered.
91 
92  /// The current defaults when creating the pass with no arguments are:
93  /// EnableLoopInterleaving = true and EnableLoopVectorization = true. This
94  /// means that interleaving default is consistent with the cl::opt flag, while
95  /// vectorization is not.
96  /// FIXME: The default for EnableLoopVectorization in the cl::opt should be
97  /// set to true, and the corresponding change to account for this be made in
98  /// opt.cpp. The initializations below will become:
99  /// InterleaveOnlyWhenForced(!EnableLoopInterleaving)
100  /// VectorizeOnlyWhenForced(!EnableLoopVectorization).
107 
110  return *this;
111  }
112 
115  return *this;
116  }
117 };
118 
119 /// Storage for information about made changes.
123 
126 };
127 
128 /// The LoopVectorize Pass.
129 struct LoopVectorizePass : public PassInfoMixin<LoopVectorizePass> {
130 private:
131  /// If false, consider all loops for interleaving.
132  /// If true, only loops that explicitly request interleaving are considered.
133  bool InterleaveOnlyWhenForced;
134 
135  /// If false, consider all loops for vectorization.
136  /// If true, only loops that explicitly request vectorization are considered.
137  bool VectorizeOnlyWhenForced;
138 
139 public:
141 
154 
156  void printPipeline(raw_ostream &OS,
157  function_ref<StringRef(StringRef)> MapClassName2PassName);
158 
159  // Shim for old PM.
164  AAResults &AA_, AssumptionCache &AC_,
165  std::function<const LoopAccessInfo &(Loop &)> &GetLAA_,
167 
168  bool processLoop(Loop *L);
169 };
170 
171 /// Reports a vectorization failure: print \p DebugMsg for debugging
172 /// purposes along with the corresponding optimization remark \p RemarkName.
173 /// If \p I is passed, it is an instruction that prevents vectorization.
174 /// Otherwise, the loop \p TheLoop is used for the location of the remark.
175 void reportVectorizationFailure(const StringRef DebugMsg,
176  const StringRef OREMsg, const StringRef ORETag,
177  OptimizationRemarkEmitter *ORE, Loop *TheLoop, Instruction *I = nullptr);
178 
179 /// Reports an informative message: print \p Msg for debugging purposes as well
180 /// as an optimization remark. Uses either \p I as location of the remark, or
181 /// otherwise \p TheLoop.
182 void reportVectorizationInfo(const StringRef OREMsg, const StringRef ORETag,
183  OptimizationRemarkEmitter *ORE, Loop *TheLoop,
184  Instruction *I = nullptr);
185 
186 } // end namespace llvm
187 
188 #endif // LLVM_TRANSFORMS_VECTORIZE_LOOPVECTORIZE_H
llvm::PreservedAnalyses
A set of analyses that are preserved following a run of a transformation pass.
Definition: PassManager.h:155
llvm::LoopVectorizeOptions::LoopVectorizeOptions
LoopVectorizeOptions()
The current defaults when creating the pass with no arguments are: EnableLoopInterleaving = true and ...
Definition: LoopVectorize.h:101
llvm::LoopVectorizeResult::MadeCFGChange
bool MadeCFGChange
Definition: LoopVectorize.h:122
llvm
This is an optimization pass for GlobalISel generic memory operations.
Definition: AllocatorList.h:23
llvm::LoopVectorizePass::printPipeline
void printPipeline(raw_ostream &OS, function_ref< StringRef(StringRef)> MapClassName2PassName)
Definition: LoopVectorize.cpp:10755
llvm::LoopVectorizePass::ORE
OptimizationRemarkEmitter * ORE
Definition: LoopVectorize.h:152
llvm::PassInfoMixin
A CRTP mix-in to automatically provide informational APIs needed for passes.
Definition: PassManager.h:374
llvm::Function
Definition: Function.h:62
llvm::Loop
Represents a single loop in the control flow graph.
Definition: LoopInfo.h:530
llvm::LoopVectorizeOptions::VectorizeOnlyWhenForced
bool VectorizeOnlyWhenForced
If false, consider all loops for vectorization.
Definition: LoopVectorize.h:90
llvm::TargetTransformInfo
This pass provides access to the codegen interfaces that are needed for IR-level transformations.
Definition: TargetTransformInfo.h:169
llvm::ScalarEvolution
The main scalar evolution driver.
Definition: ScalarEvolution.h:460
llvm::DominatorTree
Concrete subclass of DominatorTreeBase that is used to compute a normal dominator tree.
Definition: Dominators.h:151
llvm::reportVectorizationFailure
void reportVectorizationFailure(const StringRef DebugMsg, const StringRef OREMsg, const StringRef ORETag, OptimizationRemarkEmitter *ORE, Loop *TheLoop, Instruction *I=nullptr)
Reports a vectorization failure: print DebugMsg for debugging purposes along with the corresponding o...
Definition: LoopVectorize.cpp:1116
llvm::LoopVectorizeOptions::setInterleaveOnlyWhenForced
LoopVectorizeOptions & setInterleaveOnlyWhenForced(bool Value)
Definition: LoopVectorize.h:108
F
#define F(x, y, z)
Definition: MD5.cpp:56
llvm::LoopVectorizePass::AC
AssumptionCache * AC
Definition: LoopVectorize.h:150
llvm::LoopVectorizeOptions::LoopVectorizeOptions
LoopVectorizeOptions(bool InterleaveOnlyWhenForced, bool VectorizeOnlyWhenForced)
Definition: LoopVectorize.h:103
CommandLine.h
llvm::BlockFrequencyInfo
BlockFrequencyInfo pass uses BlockFrequencyInfoImpl implementation to estimate IR basic block frequen...
Definition: BlockFrequencyInfo.h:37
llvm::LoopVectorizePass::LI
LoopInfo * LI
Definition: LoopVectorize.h:143
llvm::AAResults
Definition: AliasAnalysis.h:508
llvm::reportVectorizationInfo
void reportVectorizationInfo(const StringRef OREMsg, const StringRef ORETag, OptimizationRemarkEmitter *ORE, Loop *TheLoop, Instruction *I=nullptr)
Reports an informative message: print Msg for debugging purposes as well as an optimization remark.
Definition: LoopVectorize.cpp:1127
false
Definition: StackSlotColoring.cpp:142
llvm::LoopVectorizeResult::MadeAnyChange
bool MadeAnyChange
Definition: LoopVectorize.h:121
llvm::Instruction
Definition: Instruction.h:45
llvm::raw_ostream
This class implements an extremely fast bulk output stream that can only output to a stream.
Definition: raw_ostream.h:53
llvm::LoopVectorizePass::TTI
TargetTransformInfo * TTI
Definition: LoopVectorize.h:144
llvm::EnableLoopVectorization
cl::opt< bool > EnableLoopVectorization
llvm::ProfileSummaryInfo
Analysis providing profile information.
Definition: ProfileSummaryInfo.h:39
llvm::function_ref
An efficient, type-erasing, non-owning reference to a callable.
Definition: STLExtras.h:168
llvm::LoopVectorizePass::runImpl
LoopVectorizeResult runImpl(Function &F, ScalarEvolution &SE_, LoopInfo &LI_, TargetTransformInfo &TTI_, DominatorTree &DT_, BlockFrequencyInfo &BFI_, TargetLibraryInfo *TLI_, DemandedBits &DB_, AAResults &AA_, AssumptionCache &AC_, std::function< const LoopAccessInfo &(Loop &)> &GetLAA_, OptimizationRemarkEmitter &ORE_, ProfileSummaryInfo *PSI_)
Definition: LoopVectorize.cpp:10647
llvm::LoopVectorizePass::AA
AAResults * AA
Definition: LoopVectorize.h:149
llvm::DemandedBits
Definition: DemandedBits.h:40
I
#define I(x, y, z)
Definition: MD5.cpp:59
llvm::LoopAccessInfo
Drive the analysis of memory accesses in the loop.
Definition: LoopAccessAnalysis.h:515
function
print Print MemDeps of function
Definition: MemDepPrinter.cpp:83
llvm::LoopInfo
Definition: LoopInfo.h:1083
llvm::OptimizationRemarkEmitter
The optimization diagnostic interface.
Definition: OptimizationRemarkEmitter.h:33
llvm::StringRef
StringRef - Represent a constant reference to a string, i.e.
Definition: StringRef.h:57
llvm::AssumptionCache
A cache of @llvm.assume calls within a function.
Definition: AssumptionCache.h:42
llvm::LoopVectorizeOptions::setVectorizeOnlyWhenForced
LoopVectorizeOptions & setVectorizeOnlyWhenForced(bool Value)
Definition: LoopVectorize.h:113
llvm::LoopVectorizeOptions
Definition: LoopVectorize.h:83
llvm::LoopVectorizePass::processLoop
bool processLoop(Loop *L)
Definition: LoopVectorize.cpp:10284
llvm::LoopVectorizePass::TLI
TargetLibraryInfo * TLI
Definition: LoopVectorize.h:147
llvm::TargetStackID::Value
Value
Definition: TargetFrameLowering.h:27
PassManager.h
llvm::TargetLibraryInfo
Provides information about what library functions are available for the current target.
Definition: TargetLibraryInfo.h:221
llvm::LoopVectorizePass::run
PreservedAnalyses run(Function &F, FunctionAnalysisManager &AM)
Definition: LoopVectorize.cpp:10713
llvm::LoopVectorizeResult
Storage for information about made changes.
Definition: LoopVectorize.h:120
llvm::LoopVectorizePass::DB
DemandedBits * DB
Definition: LoopVectorize.h:148
llvm::LoopVectorizePass::BFI
BlockFrequencyInfo * BFI
Definition: LoopVectorize.h:146
llvm::LoopVectorizePass::GetLAA
std::function< const LoopAccessInfo &(Loop &)> * GetLAA
Definition: LoopVectorize.h:151
llvm::LoopVectorizeResult::LoopVectorizeResult
LoopVectorizeResult(bool MadeAnyChange, bool MadeCFGChange)
Definition: LoopVectorize.h:124
llvm::LoopVectorizePass::PSI
ProfileSummaryInfo * PSI
Definition: LoopVectorize.h:153
llvm::AnalysisManager
A container for analyses that lazily runs them and caches their results.
Definition: InstructionSimplify.h:44
llvm::LoopVectorizeOptions::InterleaveOnlyWhenForced
bool InterleaveOnlyWhenForced
If false, consider all loops for interleaving.
Definition: LoopVectorize.h:86
llvm::LoopVectorizePass::LoopVectorizePass
LoopVectorizePass(LoopVectorizeOptions Opts={})
Definition: LoopVectorize.cpp:10278
llvm::Value
LLVM Value Representation.
Definition: Value.h:74
llvm::LoopVectorizePass
The LoopVectorize Pass.
Definition: LoopVectorize.h:129
llvm::codeview::PublicSymFlags::Function
@ Function
llvm::LoopVectorizePass::DT
DominatorTree * DT
Definition: LoopVectorize.h:145
llvm::LoopVectorizePass::SE
ScalarEvolution * SE
Definition: LoopVectorize.h:142
llvm::EnableLoopInterleaving
cl::opt< bool > EnableLoopInterleaving