MCDisassembler.h revision de7cbbfcce5c068f0699bdcb6dac093c0c91ba6f 
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) : GetOpInfo(0), SymbolLookUp(0),
60                                               DisInfo(0), Ctx(0),
61                                               STI(STI), Symbolizer(0),
62                                               CommentStream(0) {}
63
64  virtual ~MCDisassembler();
65
66  /// getInstruction  - Returns the disassembly of a single instruction.
67  ///
68  /// @param instr    - An MCInst to populate with the contents of the
69  ///                   instruction.
70  /// @param size     - A value to populate with the size of the instruction, or
71  ///                   the number of bytes consumed while attempting to decode
72  ///                   an invalid instruction.
73  /// @param region   - The memory object to use as a source for machine code.
74  /// @param address  - The address, in the memory space of region, of the first
75  ///                   byte of the instruction.
76  /// @param vStream  - The stream to print warnings and diagnostic messages on.
77  /// @param cStream  - The stream to print comments and annotations on.
78  /// @return         - MCDisassembler::Success if the instruction is valid,
79  ///                   MCDisassembler::SoftFail if the instruction was
80  ///                                            disassemblable but invalid,
81  ///                   MCDisassembler::Fail if the instruction was invalid.
82  virtual DecodeStatus  getInstruction(MCInst& instr,
83                                       uint64_t& size,
84                                       const MemoryObject &region,
85                                       uint64_t address,
86                                       raw_ostream &vStream,
87                                       raw_ostream &cStream) const = 0;
88private:
89  //
90  // Hooks for symbolic disassembly via the public 'C' interface.
91  //
92  // The function to get the symbolic information for operands.
93  LLVMOpInfoCallback GetOpInfo;
94  // The function to lookup a symbol name.
95  LLVMSymbolLookupCallback SymbolLookUp;
96  // The pointer to the block of symbolic information for above call back.
97  void *DisInfo;
98  // The assembly context for creating symbols and MCExprs in place of
99  // immediate operands when there is symbolic information.
100  MCContext *Ctx;
101
102protected:
103  // Subtarget information, for instruction decoding predicates if required.
104  const MCSubtargetInfo &STI;
105  OwningPtr<MCSymbolizer> Symbolizer;
106
107public:
108  // Helpers around MCSymbolizer
109  bool tryAddingSymbolicOperand(MCInst &Inst,
110                                int64_t Value,
111                                uint64_t Address, bool IsBranch,
112                                uint64_t Offset, uint64_t InstSize) const;
113
114  void tryAddingPcLoadReferenceComment(int64_t Value, uint64_t Address) const;
115
116  /// Set \p Symzer as the current symbolizer.
117  /// This takes ownership of \p Symzer, and deletes the previously set one.
118  void setSymbolizer(OwningPtr<MCSymbolizer> &Symzer);
119
120  /// Sets up an external symbolizer that uses the C API callbacks.
121  void setupForSymbolicDisassembly(LLVMOpInfoCallback GetOpInfo,
122                                   LLVMSymbolLookupCallback SymbolLookUp,
123                                   void *DisInfo,
124                                   MCContext *Ctx,
125                                   OwningPtr<MCRelocationInfo> &RelInfo);
126
127  LLVMOpInfoCallback getLLVMOpInfoCallback() const { return GetOpInfo; }
128  LLVMSymbolLookupCallback getLLVMSymbolLookupCallback() const {
129    return SymbolLookUp;
130  }
131  void *getDisInfoBlock() const { return DisInfo; }
132  MCContext *getMCContext() const { return Ctx; }
133
134  // Marked mutable because we cache it inside the disassembler, rather than
135  // having to pass it around as an argument through all the autogenerated code.
136  mutable raw_ostream *CommentStream;
137};
138
139} // namespace llvm
140
141#endif
142