Details | Last modification | View Log | RSS feed
| Rev | Author | Line No. | Line |
|---|---|---|---|
| 14 | pmbaty | 1 | //===-- IRMutator.h - Mutation engine for fuzzing IR ------------*- 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 | // Provides the IRMutator class, which drives mutations on IR based on a |
||
| 10 | // configurable set of strategies. Some common strategies are also included |
||
| 11 | // here. |
||
| 12 | // |
||
| 13 | // Fuzzer-friendly (de)serialization functions are also provided, as these |
||
| 14 | // are usually needed when mutating IR. |
||
| 15 | // |
||
| 16 | //===----------------------------------------------------------------------===// |
||
| 17 | |||
| 18 | #ifndef LLVM_FUZZMUTATE_IRMUTATOR_H |
||
| 19 | #define LLVM_FUZZMUTATE_IRMUTATOR_H |
||
| 20 | |||
| 21 | #include "llvm/FuzzMutate/OpDescriptor.h" |
||
| 22 | #include "llvm/Support/ErrorHandling.h" |
||
| 23 | #include <optional> |
||
| 24 | |||
| 25 | namespace llvm { |
||
| 26 | class BasicBlock; |
||
| 27 | class Function; |
||
| 28 | class Instruction; |
||
| 29 | class Module; |
||
| 30 | |||
| 31 | struct RandomIRBuilder; |
||
| 32 | |||
| 33 | /// Base class for describing how to mutate a module. mutation functions for |
||
| 34 | /// each IR unit forward to the contained unit. |
||
| 35 | class IRMutationStrategy { |
||
| 36 | public: |
||
| 37 | virtual ~IRMutationStrategy() = default; |
||
| 38 | |||
| 39 | /// Provide a weight to bias towards choosing this strategy for a mutation. |
||
| 40 | /// |
||
| 41 | /// The value of the weight is arbitrary, but a good default is "the number of |
||
| 42 | /// distinct ways in which this strategy can mutate a unit". This can also be |
||
| 43 | /// used to prefer strategies that shrink the overall size of the result when |
||
| 44 | /// we start getting close to \c MaxSize. |
||
| 45 | virtual uint64_t getWeight(size_t CurrentSize, size_t MaxSize, |
||
| 46 | uint64_t CurrentWeight) = 0; |
||
| 47 | |||
| 48 | /// @{ |
||
| 49 | /// Mutators for each IR unit. By default these forward to a contained |
||
| 50 | /// instance of the next smaller unit. |
||
| 51 | virtual void mutate(Module &M, RandomIRBuilder &IB); |
||
| 52 | virtual void mutate(Function &F, RandomIRBuilder &IB); |
||
| 53 | virtual void mutate(BasicBlock &BB, RandomIRBuilder &IB); |
||
| 54 | virtual void mutate(Instruction &I, RandomIRBuilder &IB) { |
||
| 55 | llvm_unreachable("Strategy does not implement any mutators"); |
||
| 56 | } |
||
| 57 | /// @} |
||
| 58 | }; |
||
| 59 | |||
| 60 | using TypeGetter = std::function<Type *(LLVMContext &)>; |
||
| 61 | |||
| 62 | /// Entry point for configuring and running IR mutations. |
||
| 63 | class IRMutator { |
||
| 64 | std::vector<TypeGetter> AllowedTypes; |
||
| 65 | std::vector<std::unique_ptr<IRMutationStrategy>> Strategies; |
||
| 66 | |||
| 67 | public: |
||
| 68 | IRMutator(std::vector<TypeGetter> &&AllowedTypes, |
||
| 69 | std::vector<std::unique_ptr<IRMutationStrategy>> &&Strategies) |
||
| 70 | : AllowedTypes(std::move(AllowedTypes)), |
||
| 71 | Strategies(std::move(Strategies)) {} |
||
| 72 | |||
| 73 | void mutateModule(Module &M, int Seed, size_t CurSize, size_t MaxSize); |
||
| 74 | }; |
||
| 75 | |||
| 76 | /// Strategy that injects operations into the function. |
||
| 77 | class InjectorIRStrategy : public IRMutationStrategy { |
||
| 78 | std::vector<fuzzerop::OpDescriptor> Operations; |
||
| 79 | |||
| 80 | std::optional<fuzzerop::OpDescriptor> chooseOperation(Value *Src, |
||
| 81 | RandomIRBuilder &IB); |
||
| 82 | |||
| 83 | public: |
||
| 84 | InjectorIRStrategy(std::vector<fuzzerop::OpDescriptor> &&Operations) |
||
| 85 | : Operations(std::move(Operations)) {} |
||
| 86 | static std::vector<fuzzerop::OpDescriptor> getDefaultOps(); |
||
| 87 | |||
| 88 | uint64_t getWeight(size_t CurrentSize, size_t MaxSize, |
||
| 89 | uint64_t CurrentWeight) override { |
||
| 90 | return Operations.size(); |
||
| 91 | } |
||
| 92 | |||
| 93 | using IRMutationStrategy::mutate; |
||
| 94 | void mutate(Function &F, RandomIRBuilder &IB) override; |
||
| 95 | void mutate(BasicBlock &BB, RandomIRBuilder &IB) override; |
||
| 96 | }; |
||
| 97 | |||
| 98 | /// Strategy that deletes instructions when the Module is too large. |
||
| 99 | class InstDeleterIRStrategy : public IRMutationStrategy { |
||
| 100 | public: |
||
| 101 | uint64_t getWeight(size_t CurrentSize, size_t MaxSize, |
||
| 102 | uint64_t CurrentWeight) override; |
||
| 103 | |||
| 104 | using IRMutationStrategy::mutate; |
||
| 105 | void mutate(Function &F, RandomIRBuilder &IB) override; |
||
| 106 | void mutate(Instruction &Inst, RandomIRBuilder &IB) override; |
||
| 107 | }; |
||
| 108 | |||
| 109 | /// Strategy that modifies instruction attributes and operands. |
||
| 110 | class InstModificationIRStrategy : public IRMutationStrategy { |
||
| 111 | public: |
||
| 112 | uint64_t getWeight(size_t CurrentSize, size_t MaxSize, |
||
| 113 | uint64_t CurrentWeight) override { |
||
| 114 | return 4; |
||
| 115 | } |
||
| 116 | |||
| 117 | using IRMutationStrategy::mutate; |
||
| 118 | void mutate(Instruction &Inst, RandomIRBuilder &IB) override; |
||
| 119 | }; |
||
| 120 | |||
| 121 | /// Strategy to split a random block and insert a random CFG in between. |
||
| 122 | class InsertCFGStrategy : public IRMutationStrategy { |
||
| 123 | private: |
||
| 124 | uint64_t MaxNumCases; |
||
| 125 | enum CFGToSink { Return, DirectSink, SinkOrSelfLoop, EndOfCFGToLink }; |
||
| 126 | |||
| 127 | public: |
||
| 128 | InsertCFGStrategy(uint64_t MNC = 8) : MaxNumCases(MNC){}; |
||
| 129 | uint64_t getWeight(size_t CurrentSize, size_t MaxSize, |
||
| 130 | uint64_t CurrentWeight) override { |
||
| 131 | return 5; |
||
| 132 | } |
||
| 133 | |||
| 134 | void mutate(BasicBlock &BB, RandomIRBuilder &IB) override; |
||
| 135 | |||
| 136 | private: |
||
| 137 | void connectBlocksToSink(ArrayRef<BasicBlock *> Blocks, BasicBlock *Sink, |
||
| 138 | RandomIRBuilder &IB); |
||
| 139 | }; |
||
| 140 | |||
| 141 | /// Strategy to insert PHI Nodes at the head of each basic block. |
||
| 142 | class InsertPHIStrategy : public IRMutationStrategy { |
||
| 143 | public: |
||
| 144 | uint64_t getWeight(size_t CurrentSize, size_t MaxSize, |
||
| 145 | uint64_t CurrentWeight) override { |
||
| 146 | return 2; |
||
| 147 | } |
||
| 148 | |||
| 149 | void mutate(BasicBlock &BB, RandomIRBuilder &IB) override; |
||
| 150 | }; |
||
| 151 | |||
| 152 | /// Strategy to select a random instruction and add a new sink (user) to it to |
||
| 153 | /// increate data dependency. |
||
| 154 | class SinkInstructionStrategy : public IRMutationStrategy { |
||
| 155 | public: |
||
| 156 | uint64_t getWeight(size_t CurrentSize, size_t MaxSize, |
||
| 157 | uint64_t CurrentWeight) override { |
||
| 158 | return 2; |
||
| 159 | } |
||
| 160 | |||
| 161 | void mutate(Function &F, RandomIRBuilder &IB) override; |
||
| 162 | void mutate(BasicBlock &BB, RandomIRBuilder &IB) override; |
||
| 163 | }; |
||
| 164 | |||
| 165 | /// Strategy to randomly select a block and shuffle the operations without |
||
| 166 | /// affecting data dependency. |
||
| 167 | class ShuffleBlockStrategy : public IRMutationStrategy { |
||
| 168 | public: |
||
| 169 | uint64_t getWeight(size_t CurrentSize, size_t MaxSize, |
||
| 170 | uint64_t CurrentWeight) override { |
||
| 171 | return 2; |
||
| 172 | } |
||
| 173 | |||
| 174 | void mutate(BasicBlock &BB, RandomIRBuilder &IB) override; |
||
| 175 | }; |
||
| 176 | |||
| 177 | /// Fuzzer friendly interface for the llvm bitcode parser. |
||
| 178 | /// |
||
| 179 | /// \param Data Bitcode we are going to parse |
||
| 180 | /// \param Size Size of the 'Data' in bytes |
||
| 181 | /// \return New module or nullptr in case of error |
||
| 182 | std::unique_ptr<Module> parseModule(const uint8_t *Data, size_t Size, |
||
| 183 | LLVMContext &Context); |
||
| 184 | |||
| 185 | /// Fuzzer friendly interface for the llvm bitcode printer. |
||
| 186 | /// |
||
| 187 | /// \param M Module to print |
||
| 188 | /// \param Dest Location to store serialized module |
||
| 189 | /// \param MaxSize Size of the destination buffer |
||
| 190 | /// \return Number of bytes that were written. When module size exceeds MaxSize |
||
| 191 | /// returns 0 and leaves Dest unchanged. |
||
| 192 | size_t writeModule(const Module &M, uint8_t *Dest, size_t MaxSize); |
||
| 193 | |||
| 194 | /// Try to parse module and verify it. May output verification errors to the |
||
| 195 | /// errs(). |
||
| 196 | /// \return New module or nullptr in case of error. |
||
| 197 | std::unique_ptr<Module> parseAndVerify(const uint8_t *Data, size_t Size, |
||
| 198 | LLVMContext &Context); |
||
| 199 | |||
| 200 | } // namespace llvm |
||
| 201 | |||
| 202 | #endif // LLVM_FUZZMUTATE_IRMUTATOR_H |