1//===-- llvm/MC/MCDisassembler.h - Disassembler interface -------*- C++ -*-===//
2//
3//                     The LLVM Compiler Infrastructure
4//
5// This file is distributed under the University of Illinois Open Source
6// License. See LICENSE.TXT for details.
7//
8//===----------------------------------------------------------------------===//
9#ifndef LLVM_MC_MCDISASSEMBLER_MCDISASSEMBLER_H
10#define LLVM_MC_MCDISASSEMBLER_MCDISASSEMBLER_H
11
12#include "llvm-c/Disassembler.h"
13#include "llvm/MC/MCDisassembler/MCSymbolizer.h"
14#include "llvm/Support/DataTypes.h"
15
16namespace llvm {
17
18template <typename T> class ArrayRef;
19class MCInst;
20class MCSubtargetInfo;
21class raw_ostream;
22class MCContext;
23
24/// Superclass for all disassemblers. Consumes a memory region and provides an
25/// array of assembly instructions.
26class MCDisassembler {
27public:
28  /// Ternary decode status. Most backends will just use Fail and
29  /// Success, however some have a concept of an instruction with
30  /// understandable semantics but which is architecturally
31  /// incorrect. An example of this is ARM UNPREDICTABLE instructions
32  /// which are disassemblable but cause undefined behaviour.
33  ///
34  /// Because it makes sense to disassemble these instructions, there
35  /// is a "soft fail" failure mode that indicates the MCInst& is
36  /// valid but architecturally incorrect.
37  ///
38  /// The enum numbers are deliberately chosen such that reduction
39  /// from Success->SoftFail ->Fail can be done with a simple
40  /// bitwise-AND:
41  ///
42  ///   LEFT & TOP =  | Success       Unpredictable   Fail
43  ///   --------------+-----------------------------------
44  ///   Success       | Success       Unpredictable   Fail
45  ///   Unpredictable | Unpredictable Unpredictable   Fail
46  ///   Fail          | Fail          Fail            Fail
47  ///
48  /// An easy way of encoding this is as 0b11, 0b01, 0b00 for
49  /// Success, SoftFail, Fail respectively.
50  enum DecodeStatus {
51    Fail = 0,
52    SoftFail = 1,
53    Success = 3
54  };
55
56  MCDisassembler(const MCSubtargetInfo &STI, MCContext &Ctx)
57    : Ctx(Ctx), STI(STI), Symbolizer(), CommentStream(nullptr) {}
58
59  virtual ~MCDisassembler();
60
61  /// Returns the disassembly of a single instruction.
62  ///
63  /// \param Instr    - An MCInst to populate with the contents of the
64  ///                   instruction.
65  /// \param Size     - A value to populate with the size of the instruction, or
66  ///                   the number of bytes consumed while attempting to decode
67  ///                   an invalid instruction.
68  /// \param Address  - The address, in the memory space of region, of the first
69  ///                   byte of the instruction.
70  /// \param VStream  - The stream to print warnings and diagnostic messages on.
71  /// \param CStream  - The stream to print comments and annotations on.
72  /// \return         - MCDisassembler::Success if the instruction is valid,
73  ///                   MCDisassembler::SoftFail if the instruction was
74  ///                                            disassemblable but invalid,
75  ///                   MCDisassembler::Fail if the instruction was invalid.
76  virtual DecodeStatus getInstruction(MCInst &Instr, uint64_t &Size,
77                                      ArrayRef<uint8_t> Bytes, uint64_t Address,
78                                      raw_ostream &VStream,
79                                      raw_ostream &CStream) const = 0;
80
81private:
82  MCContext &Ctx;
83
84protected:
85  // Subtarget information, for instruction decoding predicates if required.
86  const MCSubtargetInfo &STI;
87  std::unique_ptr<MCSymbolizer> Symbolizer;
88
89public:
90  // Helpers around MCSymbolizer
91  bool tryAddingSymbolicOperand(MCInst &Inst,
92                                int64_t Value,
93                                uint64_t Address, bool IsBranch,
94                                uint64_t Offset, uint64_t InstSize) const;
95
96  void tryAddingPcLoadReferenceComment(int64_t Value, uint64_t Address) const;
97
98  /// Set \p Symzer as the current symbolizer.
99  /// This takes ownership of \p Symzer, and deletes the previously set one.
100  void setSymbolizer(std::unique_ptr<MCSymbolizer> Symzer);
101
102  MCContext& getContext() const { return Ctx; }
103
104  const MCSubtargetInfo& getSubtargetInfo() const { return STI; }
105
106  // Marked mutable because we cache it inside the disassembler, rather than
107  // having to pass it around as an argument through all the autogenerated code.
108  mutable raw_ostream *CommentStream;
109};
110
111} // namespace llvm
112
113#endif
114