regexp-macro-assembler-x64.h revision eab96aab0834f21954b5d6aa6366bcfb348ed811 
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#ifdef V8_NATIVE_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 SetRegister(int register_index, int to);
97  virtual void Succeed();
98  virtual void WriteCurrentPositionToRegister(int reg, int cp_offset);
99  virtual void ClearRegisters(int reg_from, int reg_to);
100  virtual void WriteStackPointerToRegister(int reg);
101
102  static Result Match(Handle<Code> regexp,
103                      Handle<String> subject,
104                      int* offsets_vector,
105                      int offsets_vector_length,
106                      int previous_index);
107
108  static Result Execute(Code* code,
109                        String* input,
110                        int start_offset,
111                        const byte* input_start,
112                        const byte* input_end,
113                        int* output,
114                        bool at_start);
115
116  // Called from RegExp if the stack-guard is triggered.
117  // If the code object is relocated, the return address is fixed before
118  // returning.
119  static int CheckStackGuardState(Address* return_address,
120                                  Code* re_code,
121                                  Address re_frame);
122
123 private:
124  // Offsets from rbp of function parameters and stored registers.
125  static const int kFramePointer = 0;
126  // Above the frame pointer - function parameters and return address.
127  static const int kReturn_eip = kFramePointer + kPointerSize;
128  static const int kFrameAlign = kReturn_eip + kPointerSize;
129
130#ifdef _WIN64
131  // Parameters (first four passed as registers, but with room on stack).
132  // In Microsoft 64-bit Calling Convention, there is room on the callers
133  // stack (before the return address) to spill parameter registers. We
134  // use this space to store the register passed parameters.
135  static const int kInputString = kFrameAlign;
136  // StartIndex is passed as 32 bit int.
137  static const int kStartIndex = kInputString + kPointerSize;
138  static const int kInputStart = kStartIndex + kPointerSize;
139  static const int kInputEnd = kInputStart + kPointerSize;
140  static const int kRegisterOutput = kInputEnd + kPointerSize;
141  // AtStart is passed as 32 bit int (values 0 or 1).
142  static const int kAtStart = kRegisterOutput + kPointerSize;
143  static const int kStackHighEnd = kAtStart + kPointerSize;
144  // DirectCall is passed as 32 bit int (values 0 or 1).
145  static const int kDirectCall = kStackHighEnd + kPointerSize;
146#else
147  // In AMD64 ABI Calling Convention, the first six integer parameters
148  // are passed as registers, and caller must allocate space on the stack
149  // if it wants them stored. We push the parameters after the frame pointer.
150  static const int kInputString = kFramePointer - kPointerSize;
151  static const int kStartIndex = kInputString - kPointerSize;
152  static const int kInputStart = kStartIndex - kPointerSize;
153  static const int kInputEnd = kInputStart - kPointerSize;
154  static const int kRegisterOutput = kInputEnd - kPointerSize;
155  static const int kAtStart = kRegisterOutput - kPointerSize;
156  static const int kStackHighEnd = kFrameAlign;
157  static const int kDirectCall = kStackHighEnd + kPointerSize;
158#endif
159
160#ifdef _WIN64
161  // Microsoft calling convention has three callee-saved registers
162  // (that we are using). We push these after the frame pointer.
163  static const int kBackup_rsi = kFramePointer - kPointerSize;
164  static const int kBackup_rdi = kBackup_rsi - kPointerSize;
165  static const int kBackup_rbx = kBackup_rdi - kPointerSize;
166  static const int kLastCalleeSaveRegister = kBackup_rbx;
167#else
168  // AMD64 Calling Convention has only one callee-save register that
169  // we use. We push this after the frame pointer (and after the
170  // parameters).
171  static const int kBackup_rbx = kAtStart - kPointerSize;
172  static const int kLastCalleeSaveRegister = kBackup_rbx;
173#endif
174
175  // When adding local variables remember to push space for them in
176  // the frame in GetCode.
177  static const int kInputStartMinusOne =
178      kLastCalleeSaveRegister - kPointerSize;
179
180  // First register address. Following registers are below it on the stack.
181  static const int kRegisterZero = kInputStartMinusOne - kPointerSize;
182
183  // Initial size of code buffer.
184  static const size_t kRegExpCodeSize = 1024;
185
186  // Load a number of characters at the given offset from the
187  // current position, into the current-character register.
188  void LoadCurrentCharacterUnchecked(int cp_offset, int character_count);
189
190  // Check whether preemption has been requested.
191  void CheckPreemption();
192
193  // Check whether we are exceeding the stack limit on the backtrack stack.
194  void CheckStackLimit();
195
196  // Generate a call to CheckStackGuardState.
197  void CallCheckStackGuardState();
198
199  // The rbp-relative location of a regexp register.
200  Operand register_location(int register_index);
201
202  // The register containing the current character after LoadCurrentCharacter.
203  inline Register current_character() { return rdx; }
204
205  // The register containing the backtrack stack top. Provides a meaningful
206  // name to the register.
207  inline Register backtrack_stackpointer() { return rcx; }
208
209  // The registers containing a self pointer to this code's Code object.
210  inline Register code_object_pointer() { return r8; }
211
212  // Byte size of chars in the string to match (decided by the Mode argument)
213  inline int char_size() { return static_cast<int>(mode_); }
214
215  // Equivalent to a conditional branch to the label, unless the label
216  // is NULL, in which case it is a conditional Backtrack.
217  void BranchOrBacktrack(Condition condition, Label* to);
218
219  void MarkPositionForCodeRelativeFixup() {
220    code_relative_fixup_positions_.Add(masm_->pc_offset());
221  }
222
223  void FixupCodeRelativePositions();
224
225  // Call and return internally in the generated code in a way that
226  // is GC-safe (i.e., doesn't leave absolute code addresses on the stack)
227  inline void SafeCall(Label* to);
228  inline void SafeCallTarget(Label* label);
229  inline void SafeReturn();
230
231  // Pushes the value of a register on the backtrack stack. Decrements the
232  // stack pointer (rcx) by a word size and stores the register's value there.
233  inline void Push(Register source);
234
235  // Pushes a value on the backtrack stack. Decrements the stack pointer (rcx)
236  // by a word size and stores the value there.
237  inline void Push(Immediate value);
238
239  // Pushes the Code object relative offset of a label on the backtrack stack
240  // (i.e., a backtrack target). Decrements the stack pointer (rcx)
241  // by a word size and stores the value there.
242  inline void Push(Label* label);
243
244  // Pops a value from the backtrack stack. Reads the word at the stack pointer
245  // (rcx) and increments it by a word size.
246  inline void Pop(Register target);
247
248  // Drops the top value from the backtrack stack without reading it.
249  // Increments the stack pointer (rcx) by a word size.
250  inline void Drop();
251
252  // Before calling a C-function from generated code, align arguments on stack.
253  // After aligning the frame, arguments must be stored in esp[0], esp[4],
254  // etc., not pushed. The argument count assumes all arguments are word sized.
255  // Some compilers/platforms require the stack to be aligned when calling
256  // C++ code.
257  // Needs a scratch register to do some arithmetic. This register will be
258  // trashed.
259  inline void FrameAlign(int num_arguments);
260
261  // Calls a C function and cleans up the space for arguments allocated
262  // by FrameAlign. The called function is not allowed to trigger a garbage
263  // collection, since that might move the code and invalidate the return
264  // address (unless this is somehow accounted for by the called function).
265  inline void CallCFunction(ExternalReference function, int num_arguments);
266
267  MacroAssembler* masm_;
268
269  ZoneList<int> code_relative_fixup_positions_;
270
271  // Which mode to generate code for (ASCII or UC16).
272  Mode mode_;
273
274  // One greater than maximal register index actually used.
275  int num_registers_;
276
277  // Number of registers to output at the end (the saved registers
278  // are always 0..num_saved_registers_-1)
279  int num_saved_registers_;
280
281  // Labels used internally.
282  Label entry_label_;
283  Label start_label_;
284  Label success_label_;
285  Label backtrack_label_;
286  Label exit_label_;
287  Label check_preempt_label_;
288  Label stack_overflow_label_;
289};
290
291#endif  // V8_NATIVE_REGEXP
292
293}}  // namespace v8::internal
294
295#endif  // V8_X64_REGEXP_MACRO_ASSEMBLER_X64_H_
296