Subversion Repositories QNX 8.QNX8 LLVM/Clang compiler suite

Rev

Details | Last modification | View Log | RSS feed

Rev Author Line No. Line
14 pmbaty 1
//===- NaryReassociate.h - Reassociate n-ary expressions --------*- 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 pass reassociates n-ary add expressions and eliminates the redundancy
10
// exposed by the reassociation.
11
//
12
// A motivating example:
13
//
14
//   void foo(int a, int b) {
15
//     bar(a + b);
16
//     bar((a + 2) + b);
17
//   }
18
//
19
// An ideal compiler should reassociate (a + 2) + b to (a + b) + 2 and simplify
20
// the above code to
21
//
22
//   int t = a + b;
23
//   bar(t);
24
//   bar(t + 2);
25
//
26
// However, the Reassociate pass is unable to do that because it processes each
27
// instruction individually and believes (a + 2) + b is the best form according
28
// to its rank system.
29
//
30
// To address this limitation, NaryReassociate reassociates an expression in a
31
// form that reuses existing instructions. As a result, NaryReassociate can
32
// reassociate (a + 2) + b in the example to (a + b) + 2 because it detects that
33
// (a + b) is computed before.
34
//
35
// NaryReassociate works as follows. For every instruction in the form of (a +
36
// b) + c, it checks whether a + c or b + c is already computed by a dominating
37
// instruction. If so, it then reassociates (a + b) + c into (a + c) + b or (b +
38
// c) + a and removes the redundancy accordingly. To efficiently look up whether
39
// an expression is computed before, we store each instruction seen and its SCEV
40
// into an SCEV-to-instruction map.
41
//
42
// Although the algorithm pattern-matches only ternary additions, it
43
// automatically handles many >3-ary expressions by walking through the function
44
// in the depth-first order. For example, given
45
//
46
//   (a + c) + d
47
//   ((a + b) + c) + d
48
//
49
// NaryReassociate first rewrites (a + b) + c to (a + c) + b, and then rewrites
50
// ((a + c) + b) + d into ((a + c) + d) + b.
51
//
52
// Finally, the above dominator-based algorithm may need to be run multiple
53
// iterations before emitting optimal code. One source of this need is that we
54
// only split an operand when it is used only once. The above algorithm can
55
// eliminate an instruction and decrease the usage count of its operands. As a
56
// result, an instruction that previously had multiple uses may become a
57
// single-use instruction and thus eligible for split consideration. For
58
// example,
59
//
60
//   ac = a + c
61
//   ab = a + b
62
//   abc = ab + c
63
//   ab2 = ab + b
64
//   ab2c = ab2 + c
65
//
66
// In the first iteration, we cannot reassociate abc to ac+b because ab is used
67
// twice. However, we can reassociate ab2c to abc+b in the first iteration. As a
68
// result, ab2 becomes dead and ab will be used only once in the second
69
// iteration.
70
//
71
// Limitations and TODO items:
72
//
73
// 1) We only considers n-ary adds and muls for now. This should be extended
74
// and generalized.
75
//
76
//===----------------------------------------------------------------------===//
77
 
78
#ifndef LLVM_TRANSFORMS_SCALAR_NARYREASSOCIATE_H
79
#define LLVM_TRANSFORMS_SCALAR_NARYREASSOCIATE_H
80
 
81
#include "llvm/ADT/DenseMap.h"
82
#include "llvm/ADT/SmallVector.h"
83
#include "llvm/IR/PassManager.h"
84
#include "llvm/IR/ValueHandle.h"
85
 
86
namespace llvm {
87
 
88
class AssumptionCache;
89
class BinaryOperator;
90
class DataLayout;
91
class DominatorTree;
92
class Function;
93
class GetElementPtrInst;
94
class Instruction;
95
class ScalarEvolution;
96
class SCEV;
97
class TargetLibraryInfo;
98
class TargetTransformInfo;
99
class Type;
100
class Value;
101
 
102
class NaryReassociatePass : public PassInfoMixin<NaryReassociatePass> {
103
public:
104
  PreservedAnalyses run(Function &F, FunctionAnalysisManager &AM);
105
 
106
  // Glue for old PM.
107
  bool runImpl(Function &F, AssumptionCache *AC_, DominatorTree *DT_,
108
               ScalarEvolution *SE_, TargetLibraryInfo *TLI_,
109
               TargetTransformInfo *TTI_);
110
 
111
private:
112
  // Runs only one iteration of the dominator-based algorithm. See the header
113
  // comments for why we need multiple iterations.
114
  bool doOneIteration(Function &F);
115
 
116
  // Reassociates I for better CSE.
117
  Instruction *tryReassociate(Instruction *I, const SCEV *&OrigSCEV);
118
 
119
  // Reassociate GEP for better CSE.
120
  Instruction *tryReassociateGEP(GetElementPtrInst *GEP);
121
 
122
  // Try splitting GEP at the I-th index and see whether either part can be
123
  // CSE'ed. This is a helper function for tryReassociateGEP.
124
  //
125
  // \p IndexedType The element type indexed by GEP's I-th index. This is
126
  //                equivalent to
127
  //                  GEP->getIndexedType(GEP->getPointerOperand(), 0-th index,
128
  //                                      ..., i-th index).
129
  GetElementPtrInst *tryReassociateGEPAtIndex(GetElementPtrInst *GEP,
130
                                              unsigned I, Type *IndexedType);
131
 
132
  // Given GEP's I-th index = LHS + RHS, see whether &Base[..][LHS][..] or
133
  // &Base[..][RHS][..] can be CSE'ed and rewrite GEP accordingly.
134
  GetElementPtrInst *tryReassociateGEPAtIndex(GetElementPtrInst *GEP,
135
                                              unsigned I, Value *LHS,
136
                                              Value *RHS, Type *IndexedType);
137
 
138
  // Reassociate binary operators for better CSE.
139
  Instruction *tryReassociateBinaryOp(BinaryOperator *I);
140
 
141
  // A helper function for tryReassociateBinaryOp. LHS and RHS are explicitly
142
  // passed.
143
  Instruction *tryReassociateBinaryOp(Value *LHS, Value *RHS,
144
                                      BinaryOperator *I);
145
  // Rewrites I to (LHS op RHS) if LHS is computed already.
146
  Instruction *tryReassociatedBinaryOp(const SCEV *LHS, Value *RHS,
147
                                       BinaryOperator *I);
148
 
149
  // Tries to match Op1 and Op2 by using V.
150
  bool matchTernaryOp(BinaryOperator *I, Value *V, Value *&Op1, Value *&Op2);
151
 
152
  // Gets SCEV for (LHS op RHS).
153
  const SCEV *getBinarySCEV(BinaryOperator *I, const SCEV *LHS,
154
                            const SCEV *RHS);
155
 
156
  // Returns the closest dominator of \c Dominatee that computes
157
  // \c CandidateExpr. Returns null if not found.
158
  Instruction *findClosestMatchingDominator(const SCEV *CandidateExpr,
159
                                            Instruction *Dominatee);
160
 
161
  // Try to match \p I as signed/unsigned Min/Max and reassociate it. \p
162
  // OrigSCEV is set if \I matches Min/Max regardless whether resassociation is
163
  // done or not. If reassociation was successful newly generated instruction is
164
  // returned, otherwise nullptr.
165
  template <typename PredT>
166
  Instruction *matchAndReassociateMinOrMax(Instruction *I,
167
                                           const SCEV *&OrigSCEV);
168
 
169
  // Reassociate Min/Max.
170
  template <typename MaxMinT>
171
  Value *tryReassociateMinOrMax(Instruction *I, MaxMinT MaxMinMatch, Value *LHS,
172
                                Value *RHS);
173
 
174
  // GetElementPtrInst implicitly sign-extends an index if the index is shorter
175
  // than the pointer size. This function returns whether Index is shorter than
176
  // GEP's pointer size, i.e., whether Index needs to be sign-extended in order
177
  // to be an index of GEP.
178
  bool requiresSignExtension(Value *Index, GetElementPtrInst *GEP);
179
 
180
  AssumptionCache *AC;
181
  const DataLayout *DL;
182
  DominatorTree *DT;
183
  ScalarEvolution *SE;
184
  TargetLibraryInfo *TLI;
185
  TargetTransformInfo *TTI;
186
 
187
  // A lookup table quickly telling which instructions compute the given SCEV.
188
  // Note that there can be multiple instructions at different locations
189
  // computing to the same SCEV, so we map a SCEV to an instruction list.  For
190
  // example,
191
  //
192
  //   if (p1)
193
  //     foo(a + b);
194
  //   if (p2)
195
  //     bar(a + b);
196
  DenseMap<const SCEV *, SmallVector<WeakTrackingVH, 2>> SeenExprs;
197
};
198
 
199
} // end namespace llvm
200
 
201
#endif // LLVM_TRANSFORMS_SCALAR_NARYREASSOCIATE_H