1// Copyright 2010 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
31#include "x64/assembler-x64.h"
32#include "x64/assembler-x64-inl.h"
33#include "macro-assembler.h"
34#include "code.h"
35#include "x64/macro-assembler-x64.h"
36
37namespace v8 {
38namespace internal {
39
40#ifndef V8_INTERPRETED_REGEXP
41
42class RegExpMacroAssemblerX64: public NativeRegExpMacroAssembler {
43 public:
44  RegExpMacroAssemblerX64(Mode mode, int registers_to_save);
45  virtual ~RegExpMacroAssemblerX64();
46  virtual int stack_limit_slack();
47  virtual void AdvanceCurrentPosition(int by);
48  virtual void AdvanceRegister(int reg, int by);
49  virtual void Backtrack();
50  virtual void Bind(Label* label);
51  virtual void CheckAtStart(Label* on_at_start);
52  virtual void CheckCharacter(uint32_t c, Label* on_equal);
53  virtual void CheckCharacterAfterAnd(uint32_t c,
54                                      uint32_t mask,
55                                      Label* on_equal);
56  virtual void CheckCharacterGT(uc16 limit, Label* on_greater);
57  virtual void CheckCharacterLT(uc16 limit, Label* on_less);
58  virtual void CheckCharacters(Vector<const uc16> str,
59                               int cp_offset,
60                               Label* on_failure,
61                               bool check_end_of_string);
62  // A "greedy loop" is a loop that is both greedy and with a simple
63  // body. It has a particularly simple implementation.
64  virtual void CheckGreedyLoop(Label* on_tos_equals_current_position);
65  virtual void CheckNotAtStart(Label* on_not_at_start);
66  virtual void CheckNotBackReference(int start_reg, Label* on_no_match);
67  virtual void CheckNotBackReferenceIgnoreCase(int start_reg,
68                                               Label* on_no_match);
69  virtual void CheckNotRegistersEqual(int reg1, int reg2, Label* on_not_equal);
70  virtual void CheckNotCharacter(uint32_t c, Label* on_not_equal);
71  virtual void CheckNotCharacterAfterAnd(uint32_t c,
72                                         uint32_t mask,
73                                         Label* on_not_equal);
74  virtual void CheckNotCharacterAfterMinusAnd(uc16 c,
75                                              uc16 minus,
76                                              uc16 mask,
77                                              Label* on_not_equal);
78  // Checks whether the given offset from the current position is before
79  // the end of the string.
80  virtual void CheckPosition(int cp_offset, Label* on_outside_input);
81  virtual bool CheckSpecialCharacterClass(uc16 type,
82                                          Label* on_no_match);
83  virtual void Fail();
84  virtual Handle<HeapObject> GetCode(Handle<String> source);
85  virtual void GoTo(Label* label);
86  virtual void IfRegisterGE(int reg, int comparand, Label* if_ge);
87  virtual void IfRegisterLT(int reg, int comparand, Label* if_lt);
88  virtual void IfRegisterEqPos(int reg, Label* if_eq);
89  virtual IrregexpImplementation Implementation();
90  virtual void LoadCurrentCharacter(int cp_offset,
91                                    Label* on_end_of_input,
92                                    bool check_bounds = true,
93                                    int characters = 1);
94  virtual void PopCurrentPosition();
95  virtual void PopRegister(int register_index);
96  virtual void PushBacktrack(Label* label);
97  virtual void PushCurrentPosition();
98  virtual void PushRegister(int register_index,
99                            StackCheckFlag check_stack_limit);
100  virtual void ReadCurrentPositionFromRegister(int reg);
101  virtual void ReadStackPointerFromRegister(int reg);
102  virtual void SetCurrentPositionFromEnd(int by);
103  virtual void SetRegister(int register_index, int to);
104  virtual void Succeed();
105  virtual void WriteCurrentPositionToRegister(int reg, int cp_offset);
106  virtual void ClearRegisters(int reg_from, int reg_to);
107  virtual void WriteStackPointerToRegister(int reg);
108
109  static Result Match(Handle<Code> regexp,
110                      Handle<String> subject,
111                      int* offsets_vector,
112                      int offsets_vector_length,
113                      int previous_index,
114                      Isolate* isolate);
115
116  static Result Execute(Code* code,
117                        String* input,
118                        int start_offset,
119                        const byte* input_start,
120                        const byte* input_end,
121                        int* output,
122                        bool at_start);
123
124  // Called from RegExp if the stack-guard is triggered.
125  // If the code object is relocated, the return address is fixed before
126  // returning.
127  static int CheckStackGuardState(Address* return_address,
128                                  Code* re_code,
129                                  Address re_frame);
130
131 private:
132  // Offsets from rbp of function parameters and stored registers.
133  static const int kFramePointer = 0;
134  // Above the frame pointer - function parameters and return address.
135  static const int kReturn_eip = kFramePointer + kPointerSize;
136  static const int kFrameAlign = kReturn_eip + kPointerSize;
137
138#ifdef _WIN64
139  // Parameters (first four passed as registers, but with room on stack).
140  // In Microsoft 64-bit Calling Convention, there is room on the callers
141  // stack (before the return address) to spill parameter registers. We
142  // use this space to store the register passed parameters.
143  static const int kInputString = kFrameAlign;
144  // StartIndex is passed as 32 bit int.
145  static const int kStartIndex = kInputString + kPointerSize;
146  static const int kInputStart = kStartIndex + kPointerSize;
147  static const int kInputEnd = kInputStart + kPointerSize;
148  static const int kRegisterOutput = kInputEnd + kPointerSize;
149  static const int kStackHighEnd = kRegisterOutput + kPointerSize;
150  // DirectCall is passed as 32 bit int (values 0 or 1).
151  static const int kDirectCall = kStackHighEnd + kPointerSize;
152  static const int kIsolate = kDirectCall + kPointerSize;
153#else
154  // In AMD64 ABI Calling Convention, the first six integer parameters
155  // are passed as registers, and caller must allocate space on the stack
156  // if it wants them stored. We push the parameters after the frame pointer.
157  static const int kInputString = kFramePointer - kPointerSize;
158  static const int kStartIndex = kInputString - kPointerSize;
159  static const int kInputStart = kStartIndex - kPointerSize;
160  static const int kInputEnd = kInputStart - kPointerSize;
161  static const int kRegisterOutput = kInputEnd - kPointerSize;
162  static const int kStackHighEnd = kRegisterOutput - kPointerSize;
163  static const int kDirectCall = kFrameAlign;
164  static const int kIsolate = kDirectCall + kPointerSize;
165#endif
166
167#ifdef _WIN64
168  // Microsoft calling convention has three callee-saved registers
169  // (that we are using). We push these after the frame pointer.
170  static const int kBackup_rsi = kFramePointer - kPointerSize;
171  static const int kBackup_rdi = kBackup_rsi - kPointerSize;
172  static const int kBackup_rbx = kBackup_rdi - kPointerSize;
173  static const int kLastCalleeSaveRegister = kBackup_rbx;
174#else
175  // AMD64 Calling Convention has only one callee-save register that
176  // we use. We push this after the frame pointer (and after the
177  // parameters).
178  static const int kBackup_rbx = kStackHighEnd - kPointerSize;
179  static const int kLastCalleeSaveRegister = kBackup_rbx;
180#endif
181
182  // When adding local variables remember to push space for them in
183  // the frame in GetCode.
184  static const int kInputStartMinusOne =
185      kLastCalleeSaveRegister - kPointerSize;
186
187  // First register address. Following registers are below it on the stack.
188  static const int kRegisterZero = kInputStartMinusOne - kPointerSize;
189
190  // Initial size of code buffer.
191  static const size_t kRegExpCodeSize = 1024;
192
193  // Load a number of characters at the given offset from the
194  // current position, into the current-character register.
195  void LoadCurrentCharacterUnchecked(int cp_offset, int character_count);
196
197  // Check whether preemption has been requested.
198  void CheckPreemption();
199
200  // Check whether we are exceeding the stack limit on the backtrack stack.
201  void CheckStackLimit();
202
203  // Generate a call to CheckStackGuardState.
204  void CallCheckStackGuardState();
205
206  // The rbp-relative location of a regexp register.
207  Operand register_location(int register_index);
208
209  // The register containing the current character after LoadCurrentCharacter.
210  inline Register current_character() { return rdx; }
211
212  // The register containing the backtrack stack top. Provides a meaningful
213  // name to the register.
214  inline Register backtrack_stackpointer() { return rcx; }
215
216  // The registers containing a self pointer to this code's Code object.
217  inline Register code_object_pointer() { return r8; }
218
219  // Byte size of chars in the string to match (decided by the Mode argument)
220  inline int char_size() { return static_cast<int>(mode_); }
221
222  // Equivalent to a conditional branch to the label, unless the label
223  // is NULL, in which case it is a conditional Backtrack.
224  void BranchOrBacktrack(Condition condition, Label* to);
225
226  void MarkPositionForCodeRelativeFixup() {
227    code_relative_fixup_positions_.Add(masm_.pc_offset());
228  }
229
230  void FixupCodeRelativePositions();
231
232  // Call and return internally in the generated code in a way that
233  // is GC-safe (i.e., doesn't leave absolute code addresses on the stack)
234  inline void SafeCall(Label* to);
235  inline void SafeCallTarget(Label* label);
236  inline void SafeReturn();
237
238  // Pushes the value of a register on the backtrack stack. Decrements the
239  // stack pointer (rcx) by a word size and stores the register's value there.
240  inline void Push(Register source);
241
242  // Pushes a value on the backtrack stack. Decrements the stack pointer (rcx)
243  // by a word size and stores the value there.
244  inline void Push(Immediate value);
245
246  // Pushes the Code object relative offset of a label on the backtrack stack
247  // (i.e., a backtrack target). Decrements the stack pointer (rcx)
248  // by a word size and stores the value there.
249  inline void Push(Label* label);
250
251  // Pops a value from the backtrack stack. Reads the word at the stack pointer
252  // (rcx) and increments it by a word size.
253  inline void Pop(Register target);
254
255  // Drops the top value from the backtrack stack without reading it.
256  // Increments the stack pointer (rcx) by a word size.
257  inline void Drop();
258
259  MacroAssembler masm_;
260  MacroAssembler::NoRootArrayScope no_root_array_scope_;
261
262  ZoneList<int> code_relative_fixup_positions_;
263
264  // Which mode to generate code for (ASCII or UC16).
265  Mode mode_;
266
267  // One greater than maximal register index actually used.
268  int num_registers_;
269
270  // Number of registers to output at the end (the saved registers
271  // are always 0..num_saved_registers_-1)
272  int num_saved_registers_;
273
274  // Labels used internally.
275  Label entry_label_;
276  Label start_label_;
277  Label success_label_;
278  Label backtrack_label_;
279  Label exit_label_;
280  Label check_preempt_label_;
281  Label stack_overflow_label_;
282};
283
284#endif  // V8_INTERPRETED_REGEXP
285
286}}  // namespace v8::internal
287
288#endif  // V8_X64_REGEXP_MACRO_ASSEMBLER_X64_H_
289