MCDisassembler.h revision 2c94d0faa0e1c268893d5e04dc77e8a35889db00 
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_H
10#define LLVM_MC_MCDISASSEMBLER_H
11
12#include "llvm-c/Disassembler.h"
13#include "llvm/ADT/OwningPtr.h"
14#include "llvm/MC/MCSymbolizer.h"
15#include "llvm/MC/MCRelocationInfo.h"
16#include "llvm/Support/DataTypes.h"
17
18namespace llvm {
19
20class MCInst;
21class MCSubtargetInfo;
22class MemoryObject;
23class raw_ostream;
24class MCContext;
25
26/// MCDisassembler - Superclass for all disassemblers.  Consumes a memory region
27///   and provides an array of assembly instructions.
28class MCDisassembler {
29public:
30  /// Ternary decode status. Most backends will just use Fail and
31  /// Success, however some have a concept of an instruction with
32  /// understandable semantics but which is architecturally
33  /// incorrect. An example of this is ARM UNPREDICTABLE instructions
34  /// which are disassemblable but cause undefined behaviour.
35  ///
36  /// Because it makes sense to disassemble these instructions, there
37  /// is a "soft fail" failure mode that indicates the MCInst& is
38  /// valid but architecturally incorrect.
39  ///
40  /// The enum numbers are deliberately chosen such that reduction
41  /// from Success->SoftFail ->Fail can be done with a simple
42  /// bitwise-AND:
43  ///
44  ///   LEFT & TOP =  | Success       Unpredictable   Fail
45  ///   --------------+-----------------------------------
46  ///   Success       | Success       Unpredictable   Fail
47  ///   Unpredictable | Unpredictable Unpredictable   Fail
48  ///   Fail          | Fail          Fail            Fail
49  ///
50  /// An easy way of encoding this is as 0b11, 0b01, 0b00 for
51  /// Success, SoftFail, Fail respectively.
52  enum DecodeStatus {
53    Fail = 0,
54    SoftFail = 1,
55    Success = 3
56  };
57
58  /// Constructor     - Performs initial setup for the disassembler.
59  MCDisassembler(const MCSubtargetInfo &STI) : STI(STI), Symbolizer(0),
60                                               CommentStream(0) {}
61
62  virtual ~MCDisassembler();
63
64  /// getInstruction  - Returns the disassembly of a single instruction.
65  ///
66  /// @param instr    - An MCInst to populate with the contents of the
67  ///                   instruction.
68  /// @param size     - A value to populate with the size of the instruction, or
69  ///                   the number of bytes consumed while attempting to decode
70  ///                   an invalid instruction.
71  /// @param region   - The memory object to use as a source for machine code.
72  /// @param address  - The address, in the memory space of region, of the first
73  ///                   byte of the instruction.
74  /// @param vStream  - The stream to print warnings and diagnostic messages on.
75  /// @param cStream  - The stream to print comments and annotations on.
76  /// @return         - MCDisassembler::Success if the instruction is valid,
77  ///                   MCDisassembler::SoftFail if the instruction was
78  ///                                            disassemblable but invalid,
79  ///                   MCDisassembler::Fail if the instruction was invalid.
80  virtual DecodeStatus  getInstruction(MCInst& instr,
81                                       uint64_t& size,
82                                       const MemoryObject &region,
83                                       uint64_t address,
84                                       raw_ostream &vStream,
85                                       raw_ostream &cStream) const = 0;
86
87protected:
88  // Subtarget information, for instruction decoding predicates if required.
89  const MCSubtargetInfo &STI;
90
91private:
92  OwningPtr<MCSymbolizer> Symbolizer;
93
94public:
95  // Helpers around MCSymbolizer
96  bool tryAddingSymbolicOperand(MCInst &Inst,
97                                int64_t Value,
98                                uint64_t Address, bool IsBranch,
99                                uint64_t Offset, uint64_t InstSize) const;
100
101  void tryAddingPcLoadReferenceComment(int64_t Value, uint64_t Address) const;
102
103  /// Set \p Symzer as the current symbolizer.
104  /// This takes ownership of \p Symzer, and deletes the previously set one.
105  void setSymbolizer(OwningPtr<MCSymbolizer> &Symzer);
106
107  /// Sets up an external symbolizer that uses the C API callbacks.
108  void setupForSymbolicDisassembly(LLVMOpInfoCallback GetOpInfo,
109                                   LLVMSymbolLookupCallback SymbolLookUp,
110                                   void *DisInfo,
111                                   MCContext *Ctx,
112                                   OwningPtr<MCRelocationInfo> &RelInfo);
113
114  // Marked mutable because we cache it inside the disassembler, rather than
115  // having to pass it around as an argument through all the autogenerated code.
116  mutable raw_ostream *CommentStream;
117};
118
119} // namespace llvm
120
121#endif
122