regexp-macro-assembler-x64.h revision f87a203d89e1bbb6708282e0b64dbd13d59b723d 
1// Copyright 2009 the V8 project authors. All rights reserved.
2// Redistribution and use in source and binary forms, with or without
3// modification, are permitted provided that the following conditions are
4// met:
5//
6//     * Redistributions of source code must retain the above copyright
7//       notice, this list of conditions and the following disclaimer.
8//     * Redistributions in binary form must reproduce the above
9//       copyright notice, this list of conditions and the following
10//       disclaimer in the documentation and/or other materials provided
11//       with the distribution.
12//     * Neither the name of Google Inc. nor the names of its
13//       contributors may be used to endorse or promote products derived
14//       from this software without specific prior written permission.
15//
16// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
17// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
18// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
19// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
20// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
21// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
22// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
23// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
24// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
25// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
26// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
27
28#ifndef V8_X64_REGEXP_MACRO_ASSEMBLER_X64_H_
29#define V8_X64_REGEXP_MACRO_ASSEMBLER_X64_H_
30
31namespace v8 {
32namespace internal {
33
34#ifndef V8_INTERPRETED_REGEXP
35
36class RegExpMacroAssemblerX64: public NativeRegExpMacroAssembler {
37 public:
38  RegExpMacroAssemblerX64(Mode mode, int registers_to_save);
39  virtual ~RegExpMacroAssemblerX64();
40  virtual int stack_limit_slack();
41  virtual void AdvanceCurrentPosition(int by);
42  virtual void AdvanceRegister(int reg, int by);
43  virtual void Backtrack();
44  virtual void Bind(Label* label);
45  virtual void CheckAtStart(Label* on_at_start);
46  virtual void CheckCharacter(uint32_t c, Label* on_equal);
47  virtual void CheckCharacterAfterAnd(uint32_t c,
48                                      uint32_t mask,
49                                      Label* on_equal);
50  virtual void CheckCharacterGT(uc16 limit, Label* on_greater);
51  virtual void CheckCharacterLT(uc16 limit, Label* on_less);
52  virtual void CheckCharacters(Vector<const uc16> str,
53                               int cp_offset,
54                               Label* on_failure,
55                               bool check_end_of_string);
56  // A "greedy loop" is a loop that is both greedy and with a simple
57  // body. It has a particularly simple implementation.
58  virtual void CheckGreedyLoop(Label* on_tos_equals_current_position);
59  virtual void CheckNotAtStart(Label* on_not_at_start);
60  virtual void CheckNotBackReference(int start_reg, Label* on_no_match);
61  virtual void CheckNotBackReferenceIgnoreCase(int start_reg,
62                                               Label* on_no_match);
63  virtual void CheckNotRegistersEqual(int reg1, int reg2, Label* on_not_equal);
64  virtual void CheckNotCharacter(uint32_t c, Label* on_not_equal);
65  virtual void CheckNotCharacterAfterAnd(uint32_t c,
66                                         uint32_t mask,
67                                         Label* on_not_equal);
68  virtual void CheckNotCharacterAfterMinusAnd(uc16 c,
69                                              uc16 minus,
70                                              uc16 mask,
71                                              Label* on_not_equal);
72  // Checks whether the given offset from the current position is before
73  // the end of the string.
74  virtual void CheckPosition(int cp_offset, Label* on_outside_input);
75  virtual bool CheckSpecialCharacterClass(uc16 type,
76                                          Label* on_no_match);
77  virtual void Fail();
78  virtual Handle<Object> GetCode(Handle<String> source);
79  virtual void GoTo(Label* label);
80  virtual void IfRegisterGE(int reg, int comparand, Label* if_ge);
81  virtual void IfRegisterLT(int reg, int comparand, Label* if_lt);
82  virtual void IfRegisterEqPos(int reg, Label* if_eq);
83  virtual IrregexpImplementation Implementation();
84  virtual void LoadCurrentCharacter(int cp_offset,
85                                    Label* on_end_of_input,
86                                    bool check_bounds = true,
87                                    int characters = 1);
88  virtual void PopCurrentPosition();
89  virtual void PopRegister(int register_index);
90  virtual void PushBacktrack(Label* label);
91  virtual void PushCurrentPosition();
92  virtual void PushRegister(int register_index,
93                            StackCheckFlag check_stack_limit);
94  virtual void ReadCurrentPositionFromRegister(int reg);
95  virtual void ReadStackPointerFromRegister(int reg);
96  virtual void SetCurrentPositionFromEnd(int by);
97  virtual void SetRegister(int register_index, int to);
98  virtual void Succeed();
99  virtual void WriteCurrentPositionToRegister(int reg, int cp_offset);
100  virtual void ClearRegisters(int reg_from, int reg_to);
101  virtual void WriteStackPointerToRegister(int reg);
102
103  static Result Match(Handle<Code> regexp,
104                      Handle<String> subject,
105                      int* offsets_vector,
106                      int offsets_vector_length,
107                      int previous_index);
108
109  static Result Execute(Code* code,
110                        String* input,
111                        int start_offset,
112                        const byte* input_start,
113                        const byte* input_end,
114                        int* output,
115                        bool at_start);
116
117  // Called from RegExp if the stack-guard is triggered.
118  // If the code object is relocated, the return address is fixed before
119  // returning.
120  static int CheckStackGuardState(Address* return_address,
121                                  Code* re_code,
122                                  Address re_frame);
123
124 private:
125  // Offsets from rbp of function parameters and stored registers.
126  static const int kFramePointer = 0;
127  // Above the frame pointer - function parameters and return address.
128  static const int kReturn_eip = kFramePointer + kPointerSize;
129  static const int kFrameAlign = kReturn_eip + kPointerSize;
130
131#ifdef _WIN64
132  // Parameters (first four passed as registers, but with room on stack).
133  // In Microsoft 64-bit Calling Convention, there is room on the callers
134  // stack (before the return address) to spill parameter registers. We
135  // use this space to store the register passed parameters.
136  static const int kInputString = kFrameAlign;
137  // StartIndex is passed as 32 bit int.
138  static const int kStartIndex = kInputString + kPointerSize;
139  static const int kInputStart = kStartIndex + kPointerSize;
140  static const int kInputEnd = kInputStart + kPointerSize;
141  static const int kRegisterOutput = kInputEnd + kPointerSize;
142  static const int kStackHighEnd = kRegisterOutput + kPointerSize;
143  // DirectCall is passed as 32 bit int (values 0 or 1).
144  static const int kDirectCall = kStackHighEnd + kPointerSize;
145#else
146  // In AMD64 ABI Calling Convention, the first six integer parameters
147  // are passed as registers, and caller must allocate space on the stack
148  // if it wants them stored. We push the parameters after the frame pointer.
149  static const int kInputString = kFramePointer - kPointerSize;
150  static const int kStartIndex = kInputString - kPointerSize;
151  static const int kInputStart = kStartIndex - kPointerSize;
152  static const int kInputEnd = kInputStart - kPointerSize;
153  static const int kRegisterOutput = kInputEnd - kPointerSize;
154  static const int kStackHighEnd = kRegisterOutput - kPointerSize;
155  static const int kDirectCall = kFrameAlign;
156#endif
157
158#ifdef _WIN64
159  // Microsoft calling convention has three callee-saved registers
160  // (that we are using). We push these after the frame pointer.
161  static const int kBackup_rsi = kFramePointer - kPointerSize;
162  static const int kBackup_rdi = kBackup_rsi - kPointerSize;
163  static const int kBackup_rbx = kBackup_rdi - kPointerSize;
164  static const int kLastCalleeSaveRegister = kBackup_rbx;
165#else
166  // AMD64 Calling Convention has only one callee-save register that
167  // we use. We push this after the frame pointer (and after the
168  // parameters).
169  static const int kBackup_rbx = kStackHighEnd - kPointerSize;
170  static const int kLastCalleeSaveRegister = kBackup_rbx;
171#endif
172
173  // When adding local variables remember to push space for them in
174  // the frame in GetCode.
175  static const int kInputStartMinusOne =
176      kLastCalleeSaveRegister - kPointerSize;
177
178  // First register address. Following registers are below it on the stack.
179  static const int kRegisterZero = kInputStartMinusOne - kPointerSize;
180
181  // Initial size of code buffer.
182  static const size_t kRegExpCodeSize = 1024;
183
184  // Load a number of characters at the given offset from the
185  // current position, into the current-character register.
186  void LoadCurrentCharacterUnchecked(int cp_offset, int character_count);
187
188  // Check whether preemption has been requested.
189  void CheckPreemption();
190
191  // Check whether we are exceeding the stack limit on the backtrack stack.
192  void CheckStackLimit();
193
194  // Generate a call to CheckStackGuardState.
195  void CallCheckStackGuardState();
196
197  // The rbp-relative location of a regexp register.
198  Operand register_location(int register_index);
199
200  // The register containing the current character after LoadCurrentCharacter.
201  inline Register current_character() { return rdx; }
202
203  // The register containing the backtrack stack top. Provides a meaningful
204  // name to the register.
205  inline Register backtrack_stackpointer() { return rcx; }
206
207  // The registers containing a self pointer to this code's Code object.
208  inline Register code_object_pointer() { return r8; }
209
210  // Byte size of chars in the string to match (decided by the Mode argument)
211  inline int char_size() { return static_cast<int>(mode_); }
212
213  // Equivalent to a conditional branch to the label, unless the label
214  // is NULL, in which case it is a conditional Backtrack.
215  void BranchOrBacktrack(Condition condition, Label* to);
216
217  void MarkPositionForCodeRelativeFixup() {
218    code_relative_fixup_positions_.Add(masm_->pc_offset());
219  }
220
221  void FixupCodeRelativePositions();
222
223  // Call and return internally in the generated code in a way that
224  // is GC-safe (i.e., doesn't leave absolute code addresses on the stack)
225  inline void SafeCall(Label* to);
226  inline void SafeCallTarget(Label* label);
227  inline void SafeReturn();
228
229  // Pushes the value of a register on the backtrack stack. Decrements the
230  // stack pointer (rcx) by a word size and stores the register's value there.
231  inline void Push(Register source);
232
233  // Pushes a value on the backtrack stack. Decrements the stack pointer (rcx)
234  // by a word size and stores the value there.
235  inline void Push(Immediate value);
236
237  // Pushes the Code object relative offset of a label on the backtrack stack
238  // (i.e., a backtrack target). Decrements the stack pointer (rcx)
239  // by a word size and stores the value there.
240  inline void Push(Label* label);
241
242  // Pops a value from the backtrack stack. Reads the word at the stack pointer
243  // (rcx) and increments it by a word size.
244  inline void Pop(Register target);
245
246  // Drops the top value from the backtrack stack without reading it.
247  // Increments the stack pointer (rcx) by a word size.
248  inline void Drop();
249
250  MacroAssembler* masm_;
251
252  ZoneList<int> code_relative_fixup_positions_;
253
254  // Which mode to generate code for (ASCII or UC16).
255  Mode mode_;
256
257  // One greater than maximal register index actually used.
258  int num_registers_;
259
260  // Number of registers to output at the end (the saved registers
261  // are always 0..num_saved_registers_-1)
262  int num_saved_registers_;
263
264  // Labels used internally.
265  Label entry_label_;
266  Label start_label_;
267  Label success_label_;
268  Label backtrack_label_;
269  Label exit_label_;
270  Label check_preempt_label_;
271  Label stack_overflow_label_;
272};
273
274#endif  // V8_INTERPRETED_REGEXP
275
276}}  // namespace v8::internal
277
278#endif  // V8_X64_REGEXP_MACRO_ASSEMBLER_X64_H_
279