Home | History | Annotate | Download | only in src
      1 // Copyright 2008-2009 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 #include "v8.h"
     29 #include "ast.h"
     30 #include "bytecodes-irregexp.h"
     31 #include "regexp-macro-assembler.h"
     32 #include "regexp-macro-assembler-irregexp.h"
     33 #include "regexp-macro-assembler-irregexp-inl.h"
     34 
     35 
     36 namespace v8 {
     37 namespace internal {
     38 
     39 #ifndef V8_NATIVE_REGEXP
     40 
     41 RegExpMacroAssemblerIrregexp::RegExpMacroAssemblerIrregexp(Vector<byte> buffer)
     42     : buffer_(buffer),
     43       pc_(0),
     44       own_buffer_(false),
     45       advance_current_end_(kInvalidPC) {
     46 }
     47 
     48 
     49 RegExpMacroAssemblerIrregexp::~RegExpMacroAssemblerIrregexp() {
     50   if (backtrack_.is_linked()) backtrack_.Unuse();
     51   if (own_buffer_) buffer_.Dispose();
     52 }
     53 
     54 
     55 RegExpMacroAssemblerIrregexp::IrregexpImplementation
     56 RegExpMacroAssemblerIrregexp::Implementation() {
     57   return kBytecodeImplementation;
     58 }
     59 
     60 
     61 void RegExpMacroAssemblerIrregexp::Bind(Label* l) {
     62   advance_current_end_ = kInvalidPC;
     63   ASSERT(!l->is_bound());
     64   if (l->is_linked()) {
     65     int pos = l->pos();
     66     while (pos != 0) {
     67       int fixup = pos;
     68       pos = *reinterpret_cast<int32_t*>(buffer_.start() + fixup);
     69       *reinterpret_cast<uint32_t*>(buffer_.start() + fixup) = pc_;
     70     }
     71   }
     72   l->bind_to(pc_);
     73 }
     74 
     75 
     76 void RegExpMacroAssemblerIrregexp::EmitOrLink(Label* l) {
     77   if (l == NULL) l = &backtrack_;
     78   if (l->is_bound()) {
     79     Emit32(l->pos());
     80   } else {
     81     int pos = 0;
     82     if (l->is_linked()) {
     83       pos = l->pos();
     84     }
     85     l->link_to(pc_);
     86     Emit32(pos);
     87   }
     88 }
     89 
     90 
     91 void RegExpMacroAssemblerIrregexp::PopRegister(int register_index) {
     92   ASSERT(register_index >= 0);
     93   ASSERT(register_index <= kMaxRegister);
     94   Emit(BC_POP_REGISTER, register_index);
     95 }
     96 
     97 
     98 void RegExpMacroAssemblerIrregexp::PushRegister(
     99     int register_index,
    100     StackCheckFlag check_stack_limit) {
    101   ASSERT(register_index >= 0);
    102   ASSERT(register_index <= kMaxRegister);
    103   Emit(BC_PUSH_REGISTER, register_index);
    104 }
    105 
    106 
    107 void RegExpMacroAssemblerIrregexp::WriteCurrentPositionToRegister(
    108     int register_index, int cp_offset) {
    109   ASSERT(register_index >= 0);
    110   ASSERT(register_index <= kMaxRegister);
    111   Emit(BC_SET_REGISTER_TO_CP, register_index);
    112   Emit32(cp_offset);  // Current position offset.
    113 }
    114 
    115 
    116 void RegExpMacroAssemblerIrregexp::ClearRegisters(int reg_from, int reg_to) {
    117   ASSERT(reg_from <= reg_to);
    118   for (int reg = reg_from; reg <= reg_to; reg++) {
    119     SetRegister(reg, -1);
    120   }
    121 }
    122 
    123 
    124 void RegExpMacroAssemblerIrregexp::ReadCurrentPositionFromRegister(
    125     int register_index) {
    126   ASSERT(register_index >= 0);
    127   ASSERT(register_index <= kMaxRegister);
    128   Emit(BC_SET_CP_TO_REGISTER, register_index);
    129 }
    130 
    131 
    132 void RegExpMacroAssemblerIrregexp::WriteStackPointerToRegister(
    133     int register_index) {
    134   ASSERT(register_index >= 0);
    135   ASSERT(register_index <= kMaxRegister);
    136   Emit(BC_SET_REGISTER_TO_SP, register_index);
    137 }
    138 
    139 
    140 void RegExpMacroAssemblerIrregexp::ReadStackPointerFromRegister(
    141     int register_index) {
    142   ASSERT(register_index >= 0);
    143   ASSERT(register_index <= kMaxRegister);
    144   Emit(BC_SET_SP_TO_REGISTER, register_index);
    145 }
    146 
    147 
    148 void RegExpMacroAssemblerIrregexp::SetRegister(int register_index, int to) {
    149   ASSERT(register_index >= 0);
    150   ASSERT(register_index <= kMaxRegister);
    151   Emit(BC_SET_REGISTER, register_index);
    152   Emit32(to);
    153 }
    154 
    155 
    156 void RegExpMacroAssemblerIrregexp::AdvanceRegister(int register_index, int by) {
    157   ASSERT(register_index >= 0);
    158   ASSERT(register_index <= kMaxRegister);
    159   Emit(BC_ADVANCE_REGISTER, register_index);
    160   Emit32(by);
    161 }
    162 
    163 
    164 void RegExpMacroAssemblerIrregexp::PopCurrentPosition() {
    165   Emit(BC_POP_CP, 0);
    166 }
    167 
    168 
    169 void RegExpMacroAssemblerIrregexp::PushCurrentPosition() {
    170   Emit(BC_PUSH_CP, 0);
    171 }
    172 
    173 
    174 void RegExpMacroAssemblerIrregexp::Backtrack() {
    175   Emit(BC_POP_BT, 0);
    176 }
    177 
    178 
    179 void RegExpMacroAssemblerIrregexp::GoTo(Label* l) {
    180   if (advance_current_end_ == pc_) {
    181     // Combine advance current and goto.
    182     pc_ = advance_current_start_;
    183     Emit(BC_ADVANCE_CP_AND_GOTO, advance_current_offset_);
    184     EmitOrLink(l);
    185     advance_current_end_ = kInvalidPC;
    186   } else {
    187     // Regular goto.
    188     Emit(BC_GOTO, 0);
    189     EmitOrLink(l);
    190   }
    191 }
    192 
    193 
    194 void RegExpMacroAssemblerIrregexp::PushBacktrack(Label* l) {
    195   Emit(BC_PUSH_BT, 0);
    196   EmitOrLink(l);
    197 }
    198 
    199 
    200 void RegExpMacroAssemblerIrregexp::Succeed() {
    201   Emit(BC_SUCCEED, 0);
    202 }
    203 
    204 
    205 void RegExpMacroAssemblerIrregexp::Fail() {
    206   Emit(BC_FAIL, 0);
    207 }
    208 
    209 
    210 void RegExpMacroAssemblerIrregexp::AdvanceCurrentPosition(int by) {
    211   ASSERT(by >= kMinCPOffset);
    212   ASSERT(by <= kMaxCPOffset);
    213   advance_current_start_ = pc_;
    214   advance_current_offset_ = by;
    215   Emit(BC_ADVANCE_CP, by);
    216   advance_current_end_ = pc_;
    217 }
    218 
    219 
    220 void RegExpMacroAssemblerIrregexp::CheckGreedyLoop(
    221       Label* on_tos_equals_current_position) {
    222   Emit(BC_CHECK_GREEDY, 0);
    223   EmitOrLink(on_tos_equals_current_position);
    224 }
    225 
    226 
    227 void RegExpMacroAssemblerIrregexp::LoadCurrentCharacter(int cp_offset,
    228                                                         Label* on_failure,
    229                                                         bool check_bounds,
    230                                                         int characters) {
    231   ASSERT(cp_offset >= kMinCPOffset);
    232   ASSERT(cp_offset <= kMaxCPOffset);
    233   int bytecode;
    234   if (check_bounds) {
    235     if (characters == 4) {
    236       bytecode = BC_LOAD_4_CURRENT_CHARS;
    237     } else if (characters == 2) {
    238       bytecode = BC_LOAD_2_CURRENT_CHARS;
    239     } else {
    240       ASSERT(characters == 1);
    241       bytecode = BC_LOAD_CURRENT_CHAR;
    242     }
    243   } else {
    244     if (characters == 4) {
    245       bytecode = BC_LOAD_4_CURRENT_CHARS_UNCHECKED;
    246     } else if (characters == 2) {
    247       bytecode = BC_LOAD_2_CURRENT_CHARS_UNCHECKED;
    248     } else {
    249       ASSERT(characters == 1);
    250       bytecode = BC_LOAD_CURRENT_CHAR_UNCHECKED;
    251     }
    252   }
    253   Emit(bytecode, cp_offset);
    254   if (check_bounds) EmitOrLink(on_failure);
    255 }
    256 
    257 
    258 void RegExpMacroAssemblerIrregexp::CheckCharacterLT(uc16 limit,
    259                                                     Label* on_less) {
    260   Emit(BC_CHECK_LT, limit);
    261   EmitOrLink(on_less);
    262 }
    263 
    264 
    265 void RegExpMacroAssemblerIrregexp::CheckCharacterGT(uc16 limit,
    266                                                     Label* on_greater) {
    267   Emit(BC_CHECK_GT, limit);
    268   EmitOrLink(on_greater);
    269 }
    270 
    271 
    272 void RegExpMacroAssemblerIrregexp::CheckCharacter(uint32_t c, Label* on_equal) {
    273   if (c > MAX_FIRST_ARG) {
    274     Emit(BC_CHECK_4_CHARS, 0);
    275     Emit32(c);
    276   } else {
    277     Emit(BC_CHECK_CHAR, c);
    278   }
    279   EmitOrLink(on_equal);
    280 }
    281 
    282 
    283 void RegExpMacroAssemblerIrregexp::CheckAtStart(Label* on_at_start) {
    284   Emit(BC_CHECK_AT_START, 0);
    285   EmitOrLink(on_at_start);
    286 }
    287 
    288 
    289 void RegExpMacroAssemblerIrregexp::CheckNotAtStart(Label* on_not_at_start) {
    290   Emit(BC_CHECK_NOT_AT_START, 0);
    291   EmitOrLink(on_not_at_start);
    292 }
    293 
    294 
    295 void RegExpMacroAssemblerIrregexp::CheckNotCharacter(uint32_t c,
    296                                                      Label* on_not_equal) {
    297   if (c > MAX_FIRST_ARG) {
    298     Emit(BC_CHECK_NOT_4_CHARS, 0);
    299     Emit32(c);
    300   } else {
    301     Emit(BC_CHECK_NOT_CHAR, c);
    302   }
    303   EmitOrLink(on_not_equal);
    304 }
    305 
    306 
    307 void RegExpMacroAssemblerIrregexp::CheckCharacterAfterAnd(
    308     uint32_t c,
    309     uint32_t mask,
    310     Label* on_equal) {
    311   if (c > MAX_FIRST_ARG) {
    312     Emit(BC_AND_CHECK_4_CHARS, 0);
    313     Emit32(c);
    314   } else {
    315     Emit(BC_AND_CHECK_CHAR, c);
    316   }
    317   Emit32(mask);
    318   EmitOrLink(on_equal);
    319 }
    320 
    321 
    322 void RegExpMacroAssemblerIrregexp::CheckNotCharacterAfterAnd(
    323     uint32_t c,
    324     uint32_t mask,
    325     Label* on_not_equal) {
    326   if (c > MAX_FIRST_ARG) {
    327     Emit(BC_AND_CHECK_NOT_4_CHARS, 0);
    328     Emit32(c);
    329   } else {
    330     Emit(BC_AND_CHECK_NOT_CHAR, c);
    331   }
    332   Emit32(mask);
    333   EmitOrLink(on_not_equal);
    334 }
    335 
    336 
    337 void RegExpMacroAssemblerIrregexp::CheckNotCharacterAfterMinusAnd(
    338     uc16 c,
    339     uc16 minus,
    340     uc16 mask,
    341     Label* on_not_equal) {
    342   Emit(BC_MINUS_AND_CHECK_NOT_CHAR, c);
    343   Emit16(minus);
    344   Emit16(mask);
    345   EmitOrLink(on_not_equal);
    346 }
    347 
    348 
    349 void RegExpMacroAssemblerIrregexp::CheckNotBackReference(int start_reg,
    350                                                          Label* on_not_equal) {
    351   ASSERT(start_reg >= 0);
    352   ASSERT(start_reg <= kMaxRegister);
    353   Emit(BC_CHECK_NOT_BACK_REF, start_reg);
    354   EmitOrLink(on_not_equal);
    355 }
    356 
    357 
    358 void RegExpMacroAssemblerIrregexp::CheckNotBackReferenceIgnoreCase(
    359     int start_reg,
    360     Label* on_not_equal) {
    361   ASSERT(start_reg >= 0);
    362   ASSERT(start_reg <= kMaxRegister);
    363   Emit(BC_CHECK_NOT_BACK_REF_NO_CASE, start_reg);
    364   EmitOrLink(on_not_equal);
    365 }
    366 
    367 
    368 void RegExpMacroAssemblerIrregexp::CheckNotRegistersEqual(int reg1,
    369                                                           int reg2,
    370                                                           Label* on_not_equal) {
    371   ASSERT(reg1 >= 0);
    372   ASSERT(reg1 <= kMaxRegister);
    373   Emit(BC_CHECK_NOT_REGS_EQUAL, reg1);
    374   Emit32(reg2);
    375   EmitOrLink(on_not_equal);
    376 }
    377 
    378 
    379 void RegExpMacroAssemblerIrregexp::CheckCharacters(
    380   Vector<const uc16> str,
    381   int cp_offset,
    382   Label* on_failure,
    383   bool check_end_of_string) {
    384   ASSERT(cp_offset >= kMinCPOffset);
    385   ASSERT(cp_offset + str.length() - 1 <= kMaxCPOffset);
    386   // It is vital that this loop is backwards due to the unchecked character
    387   // load below.
    388   for (int i = str.length() - 1; i >= 0; i--) {
    389     if (check_end_of_string && i == str.length() - 1) {
    390       Emit(BC_LOAD_CURRENT_CHAR, cp_offset + i);
    391       EmitOrLink(on_failure);
    392     } else {
    393       Emit(BC_LOAD_CURRENT_CHAR_UNCHECKED, cp_offset + i);
    394     }
    395     Emit(BC_CHECK_NOT_CHAR, str[i]);
    396     EmitOrLink(on_failure);
    397   }
    398 }
    399 
    400 
    401 void RegExpMacroAssemblerIrregexp::IfRegisterLT(int register_index,
    402                                                 int comparand,
    403                                                 Label* on_less_than) {
    404   ASSERT(register_index >= 0);
    405   ASSERT(register_index <= kMaxRegister);
    406   Emit(BC_CHECK_REGISTER_LT, register_index);
    407   Emit32(comparand);
    408   EmitOrLink(on_less_than);
    409 }
    410 
    411 
    412 void RegExpMacroAssemblerIrregexp::IfRegisterGE(int register_index,
    413                                                 int comparand,
    414                                                 Label* on_greater_or_equal) {
    415   ASSERT(register_index >= 0);
    416   ASSERT(register_index <= kMaxRegister);
    417   Emit(BC_CHECK_REGISTER_GE, register_index);
    418   Emit32(comparand);
    419   EmitOrLink(on_greater_or_equal);
    420 }
    421 
    422 
    423 void RegExpMacroAssemblerIrregexp::IfRegisterEqPos(int register_index,
    424                                                    Label* on_eq) {
    425   ASSERT(register_index >= 0);
    426   ASSERT(register_index <= kMaxRegister);
    427   Emit(BC_CHECK_REGISTER_EQ_POS, register_index);
    428   EmitOrLink(on_eq);
    429 }
    430 
    431 
    432 Handle<Object> RegExpMacroAssemblerIrregexp::GetCode(Handle<String> source) {
    433   Bind(&backtrack_);
    434   Emit(BC_POP_BT, 0);
    435   Handle<ByteArray> array = Factory::NewByteArray(length());
    436   Copy(array->GetDataStartAddress());
    437   return array;
    438 }
    439 
    440 
    441 int RegExpMacroAssemblerIrregexp::length() {
    442   return pc_;
    443 }
    444 
    445 
    446 void RegExpMacroAssemblerIrregexp::Copy(Address a) {
    447   memcpy(a, buffer_.start(), length());
    448 }
    449 
    450 
    451 void RegExpMacroAssemblerIrregexp::Expand() {
    452   bool old_buffer_was_our_own = own_buffer_;
    453   Vector<byte> old_buffer = buffer_;
    454   buffer_ = Vector<byte>::New(old_buffer.length() * 2);
    455   own_buffer_ = true;
    456   memcpy(buffer_.start(), old_buffer.start(), old_buffer.length());
    457   if (old_buffer_was_our_own) {
    458     old_buffer.Dispose();
    459   }
    460 }
    461 
    462 #endif  // !V8_NATIVE_REGEXP
    463 
    464 } }  // namespace v8::internal
    465