src/3rdparty/javascriptcore/JavaScriptCore/ChangeLog

   1 2010-08-24  Oliver Hunt  <oliver@apple.com>
   2
   3         Reviewed by Geoff Garen.
   4
   5         Don't seed the JS random number generator from time()
   6         https://bugs.webkit.org/show_bug.cgi?id=41868
   7         <rdar://problem/8171025>
   8
   9         Switch to using the secure random number generator to
  10         seed the fast random generator, and make the generator
  11         be per global object.
  12
  13         * runtime/JSGlobalData.cpp:
  14         (JSC::JSGlobalData::JSGlobalData):
  15         * runtime/JSGlobalData.h:
  16         * runtime/JSGlobalObject.h:
  17         (JSC::JSGlobalObject::JSGlobalObjectData::JSGlobalObjectData):
  18         (JSC::JSGlobalObject::weakRandomNumber):
  19         * runtime/MathObject.cpp:
  20         (JSC::mathProtoFuncRandom):
  21
  22 2010-06-18  Tucker Jay  <jay.tucker@nokia.com>
  23
  24         Reviewed by NOBODY (OOPS!).
  25
  26         [Symbian] Lazy commit of memory required in JSC register file
  27         https://bugs.webkit.org/show_bug.cgi?id=34349
  28
  29         * JavaScriptCore.pro: Added 1 new Symbian source file
  30         * interpreter/RegisterFile.cpp:
  31         (JSC::RegisterFile::~RegisterFile):
  32         * interpreter/RegisterFile.h:
  33         (JSC::RegisterFile::):
  34         (JSC::RegisterFile::start):
  35         (JSC::RegisterFile::end):
  36         (JSC::RegisterFile::size):
  37         (JSC::RegisterFile::setNumGlobals):
  38         (JSC::RegisterFile::numGlobals):
  39         (JSC::RegisterFile::maxGlobals):
  40         (JSC::RegisterFile::lastGlobal):
  41         (JSC::RegisterFile::markGlobals):
  42         (JSC::RegisterFile::markCallFrames):
  43         (JSC::isPageAligned):
  44         (JSC::RegisterFile::RegisterFile):
  45         (JSC::RegisterFile::shrink):
  46         (JSC::RegisterFile::grow):
  47         * wtf/symbian/RegisterFileAllocatorSymbian.cpp: Added.
  48         (WTF::RegisterFileAllocator::RegisterFileAllocator):
  49         Helper class to allocate memory required by RegisterFile
  50         more efficiently.
  51         (WTF::RegisterFileAllocator::~RegisterFileAllocator):
  52         (WTF::RegisterFileAllocator::buffer):
  53         (WTF::RegisterFileAllocator::grow):
  54         (WTF::RegisterFileAllocator::shrink):
  55         * wtf/symbian/RegisterFileAllocatorSymbian.h: Added.
  56         * wtf/symbian/SymbianDefines.h: Added.
  57
  58 2010-06-19  Thiago Macieira <thiago.macieira@nokia.com>
  59
  60         Reviewed by Kenneth Rohde Christiansen.
  61
  62         Don't use __attribute__((may_alias)) with the Intel compiler,
  63         as it doesn't understand it.
  64
  65         * wtf/Vector.h:
  66
  67 2010-06-19  Thiago Macieira <thiago.macieira@nokia.com>
  68
  69         Reviewed by Kenneth Rohde Christiansen.
  70
  71         Fix compilation with the Intel C++ compiler (11.1.072).
  72
  73         Like RVCT, label pointers must be void*, not const void*.
  74
  75         * bytecode/Opcode.h:
  76
  77 2010-06-19  Thiago Macieira <thiago.macieira@nokia.com>
  78
  79         Reviewed by Kenneth Rohde Christiansen.
  80
  81         Add the WTF_COMPILER_INTEL for when the Intel compiler is used
  82         for building. Usually, the Intel compiler masquerades as
  83         another compiler in the system and gets away with it, but some
  84         times specific fixes are required (such as when using language
  85         extensions).
  86
  87         * wtf/Platform.h:
  88
  89 2010-06-07  Benjamin Poulain  <benjamin.poulain@nokia.com>
  90
  91         Reviewed by Simon Hausmann.
  92
  93         [Qt] Crash when compiling on Snow Leopard and running on Leopard
  94         https://bugs.webkit.org/show_bug.cgi?id=31403
  95
  96         Disable the use of pthread_setname_np and other symbols
  97         when targetting Leopard.
  98
  99         Use the defines TARGETING_XX instead of BUILDING_ON_XX
 100         for features that cannot be used before Snow Leopard.
 101
 102         * wtf/Platform.h:
 103
 104 2010-05-10  Laszlo Gombos  <laszlo.1.gombos@nokia.com>
 105
 106         Reviewed by Darin Adler.
 107
 108         [Qt] Disable JIT support for mingw-w64
 109         https://bugs.webkit.org/show_bug.cgi?id=38747
 110
 111         Disale JIT for mingw-w64 as it is reportedly
 112         unstable.
 113
 114         Thanks for Vanboxem Rruben for the investigation.
 115
 116         * wtf/Platform.h:
 117
 118 2010-05-06  Fumitoshi Ukai  <ukai@chromium.org>
 119
 120         Reviewed by Alexey Proskuryakov.
 121
 122         JavaScriptCore/wtf/RandomNumber.h should provide using WTF::*
 123         https://bugs.webkit.org/show_bug.cgi?id=38719
 124
 125         * wtf/RandomNumber.h:
 126           Add using directives.
 127
 128 2010-04-28  Simon Hausmann  <simon.hausmann@nokia.com>, Kent Hansen <kent.hansen@nokia.com>
 129
 130         Reviewed by Darin Adler.
 131
 132         JSC's currentThreadStackBase is not reentrant on some platforms
 133         https://bugs.webkit.org/show_bug.cgi?id=37195
 134
 135         This function needs to be reentrant to avoid memory corruption on platforms where
 136         the implementation uses global variables.
 137
 138         This patch adds a mutex lock where necessary and makes the Symbian implementation
 139         reentrant.
 140
 141         * runtime/Collector.cpp:
 142         (JSC::currentThreadStackBaseMutex):
 143         (JSC::currentThreadStackBase):
 144
 145 2010-04-14  Kent Hansen  <kent.hansen@nokia.com>
 146
 147         Reviewed by Maciej Stachowiak.
 148
 149         Mac OS X: Use deployment target to determine whether memory tagging should be enabled
 150         https://bugs.webkit.org/show_bug.cgi?id=34888
 151
 152         When building on (Snow) Leopard but targeting Tiger
 153         (TARGETING_TIGER defined, BUILDING_ON_TIGER not defined),
 154         WebKit would crash on Tiger because the tags passed to mmap
 155         caused those function calls to fail.
 156
 157         Conversely, when building on Tiger but targeting Leopard
 158         (BUILDING_ON_TIGER defined, TARGETING_LEOPARD defined), WebKit
 159         would crash on Leopard because the tags passed to vm_map and
 160         vm_allocate caused those function calls to fail.
 161
 162         Solution: Use TARGETING_TIGER rather than BUILDING_ON_TIGER to
 163         govern the tag definitions. Use the same tags for vm_map and
 164         vm_allocate regardless of target, since they work on
 165         both. Fall back to the mmap tags that work on Tiger (that is,
 166         "no tags") if targeting Tiger, since those tags also work on
 167         Leopard.
 168
 169         * wtf/VMTags.h:
 170
 171 2010-04-02  Ruben Van Boxem  <vanboxem.ruben@gmail.com>
 172
 173         Reviewed by Eric Seidel.
 174
 175         Mingw-w64 fixes for JavaScriptCore
 176         https://bugs.webkit.org/show_bug.cgi?id=35607
 177
 178         * runtime/Collector.cpp: use the msvc code for mingw-w64 (but not mingw-w32)
 179         (JSC::Heap::allocateBlock):
 180         (JSC::Heap::freeBlockPtr):
 181         (JSC::currentThreadStackBase):
 182         (JSC::currentThreadStackBase):
 183         * wtf/Platform.h: added COMPILER(MINGW64) check to differentiate between mingw.org and mingw-w64 functions
 184
 185 2010-03-29  Patrick Gansterer  <paroga@paroga.com>
 186
 187         Reviewed by Darin Adler.
 188
 189         Corrected name of (u)int64_t compile time assert.
 190         https://bugs.webkit.org/show_bug.cgi?id=36739
 191
 192         int64_t_is_four_bytes -> int64_t_is_eight_bytes
 193
 194         * os-win32/stdint.h:
 195
 196 2010-03-29  Thomas Zander  <t.zander@nokia.com>
 197
 198         Reviewed by Simon Hausmann.
 199
 200         https://bugs.webkit.org/show_bug.cgi?id=36742
 201
 202         gcc for Symbian doesn't support gcc extensions like atomicity.h - disable
 203
 204         * wtf/Threading.h: also detect os symbian
 205
 206 2010-03-23  Mark Rowe  <mrowe@apple.com>
 207
 208         Build fix.
 209
 210         * runtime/ArrayPrototype.cpp:
 211         (JSC::arrayProtoFuncSplice): Some versions of GCC emit a warning about the implicit 64- to 32-bit truncation
 212         that takes place here. An explicit cast is sufficient to silence it.
 213
 214 2010-03-23  Alexey Proskuryakov  <ap@apple.com>
 215
 216         Build fix.
 217
 218         * runtime/ArrayPrototype.cpp: (JSC::arrayProtoFuncSplice): Fixed a typo - length doesn't
 219         need to be converted with toInteger().
 220
 221 2010-03-23  Alexey Proskuryakov  <ap@apple.com>
 222
 223         Reviewed by Geoff Garen.
 224
 225         https://bugs.webkit.org/show_bug.cgi?id=36511
 226         <rdar://problem/7753498> Safari freezes when using SPUTNIK JavaScript conformance check
 227
 228         Test: fast/js/sputnik-S15.4.4.12_A3_T3.html
 229
 230         * runtime/ArrayPrototype.cpp: (JSC::arrayProtoFuncSplice): We were incorrectly computing
 231         the start offset, and iterated over (almost) all integers. Note that this can be fixed
 232         without using doubles, but the code would be much more complicated, and there is no important
 233         reason to stick to integers here.
 234
 235 2010-03-22  Siddharth Mathur  <siddharth.mathur@nokia.com>
 236
 237         Reviewed by Laszlo Gombos.
 238
 239         [Symbian] More efficient aligned memory allocation for JSC Collector
 240         https://bugs.webkit.org/show_bug.cgi?id=34350
 241
 242         * JavaScriptCore.pri: Added 2 new Symbian source files and HAL linkage
 243
 244         * runtime/Collector.cpp: Reduced port-specific code and added private data member
 245         (JSC::Heap::Heap):
 246         (JSC::Heap::~Heap):
 247         (JSC::Heap::destroy):
 248         (JSC::Heap::allocateBlock):
 249         (JSC::Heap::freeBlockPtr):
 250
 251         * runtime/Collector.h: Added private data member
 252
 253         * wtf/symbian: Added.
 254         * wtf/symbian/BlockAllocatorSymbian.cpp: Added.
 255         (WTF::AlignedBlockAllocator::AlignedBlockAllocator): Helper class to allocate
 256         aligned blocks more efficiently as required by Collector
 257         (WTF::AlignedBlockAllocator::alloc):
 258         (WTF::AlignedBlockAllocator::free):
 259         (WTF::AlignedBlockAllocator::destroy):
 260         (WTF::AlignedBlockAllocator::~AlignedBlockAllocator):
 261         * wtf/symbian/BlockAllocatorSymbian.h: Added.
 262
 263 2010-03-22  Geoffrey Garen  <ggaren@apple.com>
 264
 265         Reviewed by Sam Weinig.
 266
 267         Fixed <rdar://problem/7728196> REGRESSION (r46701): -(-2147483648)
 268         evaluates to -2147483648 on 32 bit (35842)
 269
 270         Two ways to fix the same bug:
 271
 272         1. Check for overflow when negating, since negating the largest negative
 273         int causes overflow.
 274
 275         2. Constant-fold even when negating a negative, since, like they say in
 276         high school, "math works."
 277
 278         * assembler/MacroAssemblerARM.h:
 279         (JSC::MacroAssemblerARM::branchNeg32):
 280         * assembler/MacroAssemblerX86Common.h:
 281         (JSC::MacroAssemblerX86Common::branchNeg32): Added a branching version
 282         of the negate operator.
 283
 284         * jit/JITArithmetic.cpp:
 285         (JSC::JIT::emit_op_negate): Use the branching version of the negate
 286         operator to check for overflow.
 287
 288         (JSC::JIT::emitSlow_op_negate): Link the check for overflow to a slow case.
 289         (We could emit inline code for this, since we know what the result would
 290         be, but that's probably just a waste of generated code.)
 291
 292         * parser/Grammar.y: Constant fold even when negating a negative.
 293
 294 2010-03-17  Mike Homey  <glandium@debian.org>
 295
 296         Reviewed by Gustavo Noronha.
 297
 298         Build fix for SPARC. Fix missing macro value.
 299
 300         * wtf/Platform.h:
 301
 302 2010-03-03  Mark Rowe  <mrowe@apple.com>
 303
 304         Reviewed by Geoff Garen.
 305
 306         Add virtual memory tags for TCMalloc and WebCore's purgeable buffers.
 307
 308         * wtf/TCSystemAlloc.cpp:
 309         (TryMmap): Use the VM tag.
 310         * wtf/VMTags.h: Make use of VM_MEMORY_TCMALLOC and VM_MEMORY_WEBCORE_PURGEABLE_BUFFERS.
 311
 312 2010-03-01  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
 313
 314         Reviewed by Simon Hausmann.
 315
 316         Fix the Qt build on Mac OS X/Cocoa 64-bit
 317
 318         * JavaScriptCore.pri: Add missing implementation file to resolve JSC symbols
 319
 320 2010-02-26  Janne Koskinen  <janne.p.koskinen@digia.com>
 321
 322         Reviewed by Simon Hausmann.
 323
 324         [Qt] Symbian specific getCPUTime implemetation
 325         https://bugs.webkit.org/show_bug.cgi?id=34742
 326
 327         Default implementation doesn't work on Symbian devices.
 328         This change adds a proper implementation by
 329         asking thread execution time from the current thread.
 330
 331         * runtime/TimeoutChecker.cpp:
 332         (JSC::getCPUTime):
 333
 334 2010-02-15  Gabor Loki  <loki@webkit.org>
 335
 336         Reviewed by Gavin Barraclough.
 337
 338         Fix the SP at ctiOpThrowNotCaught on Thumb2 (JSVALUE32)
 339         https://bugs.webkit.org/show_bug.cgi?id=34939
 340
 341         * jit/JITStubs.cpp:
 342
 343 2010-02-15  Gavin Barraclough  <barraclough@apple.com>
 344
 345         Reviewed by NOBODY (Build fix).
 346
 347         Add missing cast for !YARR (PPC) builds.
 348
 349         * runtime/RegExp.cpp:
 350         (JSC::RegExp::match):
 351
 352 2010-02-14  Laszlo Gombos  <laszlo.1.gombos@nokia.com>
 353
 354         Reviewed by Adam Barth.
 355
 356         Implement NEVER_INLINE and NO_RETURN for RVCT
 357         https://bugs.webkit.org/show_bug.cgi?id=34740
 358
 359         * wtf/AlwaysInline.h:
 360
 361 2010-02-12  Kwang Yul Seo  <skyul@company100.net>
 362
 363         Reviewed by Adam Barth.
 364
 365         Typedef both JSChar and UChar to wchar_t in RVCT.
 366         https://bugs.webkit.org/show_bug.cgi?id=34560
 367
 368         Define both JSChar and UChar to wchar_t as the size
 369         of wchar_t is 2 bytes in RVCT.
 370
 371         * API/JSStringRef.h:
 372         * wtf/unicode/qt4/UnicodeQt4.h:
 373
 374 2009-10-06  Yongjun Zhang  <yongjun.zhang@nokia.com>
 375
 376         Reviewed by Simon Hausmann.
 377
 378         Get rid of WINSCW hack for UnSpecifiedBoolType
 379
 380         Add parenthesis around (RefPtr::*UnspecifiedBoolType) to make the WINSCW
 381         compiler work with the default UnSpecifiedBoolType() operator.
 382
 383         https://bugs.webkit.org/show_bug.cgi?id=28054
 384
 385         * wtf/RefPtr.h:
 386
 387 2010-02-09  Janne Koskinen  <janne.p.koskinen@digia.com>
 388
 389         Reviewed by Laszlo Gombos.
 390
 391         [Qt] use nanval() for Symbian as nonInlineNaN
 392         https://bugs.webkit.org/show_bug.cgi?id=34170
 393
 394         numeric_limits<double>::quiet_NaN is broken in Symbian
 395         causing NaN to be evaluated as a number.
 396
 397         * runtime/JSValue.cpp:
 398         (JSC::nonInlineNaN):
 399
 400 2010-02-01  Kent Tamura  <tkent@chromium.org>
 401
 402         Reviewed by Darin Adler.
 403
 404         Date.UTC() should apply TimeClip operation.
 405         https://bugs.webkit.org/show_bug.cgi?id=34461
 406
 407         ECMAScript 5 15.9.4.3:
 408         > 9 Return TimeClip(MakeDate(MakeDay(yr, m, dt), MakeTime(h, min, s, milli))).
 409
 410         * runtime/DateConstructor.cpp:
 411         (JSC::dateUTC): Calls WTF::timeClip().
 412
 413 2010-02-01  Kent Tamura  <tkent@chromium.org>
 414
 415         Reviewed by Darin Adler.
 416
 417         Fix a bug that Math.round() retunrs incorrect results for huge integers
 418         https://bugs.webkit.org/show_bug.cgi?id=34462
 419
 420         * runtime/MathObject.cpp:
 421         (JSC::mathProtoFuncRound): Avoid "arg + 0.5".
 422
 423 2010-02-01  Patrick Gansterer  <paroga@paroga.com>
 424
 425         Reviewed by Darin Adler.
 426
 427         [Qt] WinCE buildfix after r52729 and fix for Q_BIG_ENDIAN typo.
 428         https://bugs.webkit.org/show_bug.cgi?id=34378
 429
 430         * wtf/Platform.h:
 431
 432 2010-01-31  Patrick Gansterer  <paroga@paroga.com>
 433
 434         Reviewed by Darin Adler.
 435
 436         Buildfix for WinCE + style fixes (TLS_OUT_OF_INDEXES is not defined).
 437         https://bugs.webkit.org/show_bug.cgi?id=34380
 438
 439         * wtf/ThreadSpecific.h:
 440
 441 2010-01-31  Kent Tamura  <tkent@chromium.org>
 442
 443         Reviewed by Darin Adler.
 444
 445         [Windows] Fix a bug of round() with huge integral numbers
 446         https://bugs.webkit.org/show_bug.cgi?id=34297
 447
 448         Fix a bug that round() for huge integral numbers returns incorrect
 449         results. For example, round(8639999913600001) returns
 450         8639999913600002 without this change though the double type can
 451         represent 8639999913600001 precisely.
 452
 453         Math.round() of JavaScript has a similar problem. But this change
 454         doesn't fix it because Math.round() doesn't use round() of
 455         MathExtra.h.
 456
 457         * wtf/MathExtras.h:
 458         (round): Avoid to do "num + 0.5" or "num - 0.5".
 459         (roundf): Fixed similarly.
 460         (llround): Calls round().
 461         (llroundf): Calls roundf().
 462         (lround): Calls round().
 463         (lroundf): Calls roundf().
 464
 465 2010-01-27  Anton Muhin  <antonm@chromium.org>
 466
 467         Reviewed by Darin Adler.
 468
 469         Remove trailing \ from inline function code
 470         https://bugs.webkit.org/show_bug.cgi?id=34223
 471
 472         * assembler/ARMv7Assembler.h:
 473         (JSC::ARMThumbImmediate::countLeadingZerosPartial):
 474
 475 2010-01-27  Kwang Yul Seo  <skyul@company100.net>
 476
 477         Reviewed by Oliver Hunt.
 478
 479         [BREWMP] Add MarkStack fastMalloc implementation for platforms without VirtualAlloc or mmap.
 480         https://bugs.webkit.org/show_bug.cgi?id=33582
 481
 482         Use fastMalloc and fastFree to implement MarkStack::allocateStack and
 483         MarkStack::releaseStack for platforms without page level allocation.
 484
 485         * runtime/MarkStack.h:
 486         (JSC::MarkStack::MarkStackArray::shrinkAllocation):
 487         * runtime/MarkStackNone.cpp: Added.
 488         (JSC::MarkStack::initializePagesize):
 489         (JSC::MarkStack::allocateStack):
 490         (JSC::MarkStack::releaseStack):
 491
 492 2010-01-27  Kwang Yul Seo  <skyul@company100.net>
 493
 494         Reviewed by Eric Seidel.
 495
 496         [BREWMP] Don't use time function
 497         https://bugs.webkit.org/show_bug.cgi?id=33577
 498
 499         Calling time(0) in BREW devices causes a crash because time
 500         is not properly ported in most devices. Cast currentTime() to
 501         time_t to get the same result as time(0).
 502
 503         * wtf/DateMath.cpp:
 504         (WTF::calculateUTCOffset):
 505
 506 2010-01-27  Alexey Proskuryakov  <ap@apple.com>
 507
 508         Revert r53899 (HashMap<AtomicStringImpl*, Value> key checks) and subsequent build fixes,
 509         because they make SVG tests crash in release builds.
 510
 511         * wtf/HashMap.h:
 512         (WTF::::remove):
 513         * wtf/HashSet.h:
 514         (WTF::::remove):
 515         * wtf/HashTable.h:
 516         (WTF::::add):
 517         (WTF::::addPassingHashCode):
 518         (WTF::::removeAndInvalidate):
 519         (WTF::::remove):
 520         (WTF::::rehash):
 521         (WTF::::checkTableConsistency):
 522         (WTF::::checkTableConsistencyExceptSize):
 523         * wtf/HashTraits.h:
 524         (WTF::GenericHashTraits::emptyValue):
 525         (WTF::):
 526         * wtf/RefPtrHashMap.h:
 527         (WTF::::remove):
 528
 529 2010-01-26  Alexey Proskuryakov  <ap@apple.com>
 530
 531         More Windows build fixing.
 532
 533         * wtf/HashTraits.h: _msize takes void*, remove const qualifier from type.
 534
 535 2010-01-26  Alexey Proskuryakov  <ap@apple.com>
 536
 537         Windows build fix.
 538
 539         * wtf/HashTraits.h: Include malloc.h for _msize().
 540
 541 2010-01-26  Alexey Proskuryakov  <ap@apple.com>
 542
 543         Build fix.
 544
 545         * wtf/HashTable.h: (WTF::HashTable::checkTableConsistencyExceptSize): Remove const from a
 546         static (empty) version of this function.
 547
 548 2010-01-26  Alexey Proskuryakov  <ap@apple.com>
 549
 550         Reviewed by Darin Adler.
 551
 552         https://bugs.webkit.org/show_bug.cgi?id=34150
 553         WebKit needs a mechanism to catch stale HashMap entries
 554
 555         It is very difficult to catch stale pointers that are HashMap keys - since a pointer's hash
 556         is just its value, it is very unlikely that any observable problem is reproducible.
 557
 558         This extends hash table consistency checks to check that pointers are referencing allocated
 559         memory blocks, and makes it possible to invoke the checks explicitly (it is not feasible
 560         to enable CHECK_HASHTABLE_CONSISTENCY by default, because that affects performance too much).
 561
 562         * wtf/HashMap.h: (WTF::::checkConsistency): Call through to HashTable implementation. We can
 563         add similar calls to HashSet and HashCountedSet, but I haven't seen hard to debug problems
 564         with those yet.
 565
 566         * wtf/HashSet.h: (WTF::::remove): The version of checkTableConsistency that's guarded by
 567         CHECK_HASHTABLE_CONSISTENCY is now called internalCheckTableConsistency().
 568
 569         * wtf/HashTable.h:
 570         (WTF::HashTable::internalCheckTableConsistency):
 571         (WTF::HashTable::internalCheckTableConsistencyExceptSize):
 572         (WTF::HashTable::checkTableConsistencyExceptSize):
 573         Expose checkTableConsistency() even if CHECK_HASHTABLE_CONSISTENCY is off.
 574         (WTF::::add): Updated for checkTableConsistency renaming.
 575         (WTF::::addPassingHashCode): Ditto.
 576         (WTF::::removeAndInvalidate): Ditto.
 577         (WTF::::remove): Ditto.
 578         (WTF::::rehash): Ditto.
 579         (WTF::::checkTableConsistency): The assertion for !shouldExpand() was not correct - this
 580         function returns true for tables with m_table == 0.
 581         (WTF::::checkTableConsistencyExceptSize): Call checkValueConsistency for key. Potentially,
 582         we could do the same for values.
 583
 584         * wtf/HashTraits.h:
 585         (WTF::GenericHashTraits::checkValueConsistency): An empty function that can be overridden
 586         to add checks. Currently, the only override is for pointer hashes.
 587
 588         * wtf/RefPtrHashMap.h: (WTF::::remove): Updated for checkTableConsistency renaming.
 589
 590 2010-01-26  Lyon Chen  <liachen@rim.com>
 591
 592         Reviewed by Maciej Stachowiak.
 593
 594         Opcode.h use const void* for Opcode cause error #1211 for RVCT compiler
 595         https://bugs.webkit.org/show_bug.cgi?id=33902
 596
 597         * bytecode/Opcode.h:
 598
 599 2010-01-26  Steve Falkenburg  <sfalken@apple.com>
 600
 601         Reviewed by Oliver Hunt.
 602
 603         Windows build references non-existent include paths
 604         https://bugs.webkit.org/show_bug.cgi?id=34175
 605
 606         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
 607         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreCommon.vsprops:
 608         * JavaScriptCore.vcproj/WTF/WTFCommon.vsprops:
 609         * JavaScriptCore.vcproj/jsc/jscCommon.vsprops:
 610         * JavaScriptCore.vcproj/testapi/testapi.vcproj:
 611         * JavaScriptCore.vcproj/testapi/testapiCommon.vsprops:
 612
 613 2010-01-26  Oliver Hunt  <oliver@apple.com>
 614
 615         Reviewed by Geoffrey Garen.
 616
 617         Using JavaScriptCore API with a webkit vended context can result in slow script dialog
 618         https://bugs.webkit.org/show_bug.cgi?id=34172
 619
 620         Make the APIShim correctly increment and decrement the timeout
 621         entry counter.
 622
 623         * API/APIShims.h:
 624         (JSC::APIEntryShimWithoutLock::APIEntryShimWithoutLock):
 625         (JSC::APIEntryShimWithoutLock::~APIEntryShimWithoutLock):
 626         (JSC::APICallbackShim::APICallbackShim):
 627         (JSC::APICallbackShim::~APICallbackShim):
 628
 629 2010-01-26  Simon Hausmann  <simon.hausmann@nokia.com>
 630
 631         [Qt] Fix compilation of QtScript with non-gcc compilers
 632
 633         Variable length stack arrays are a gcc extension. Use QVarLengthArray
 634         as a more portable solution that still tries to allocate on the stack
 635         first.
 636
 637         * qt/api/qscriptvalue_p.h:
 638         (QScriptValuePrivate::call):
 639
 640 2010-01-26  Simon Hausmann  <simon.hausmann@nokia.com>
 641
 642         Reviewed by Tor Arne Vestbø.
 643
 644         [Qt] Fix the build on platforms without JIT support.
 645
 646         The JIT support should be determined at compile-time via wtf/Platform.h
 647
 648         * qt/api/QtScript.pro:
 649
 650 2010-01-26  Jedrzej Nowacki  <jedrzej.nowacki@nokia.com>
 651
 652         Reviewed by Simon Hausmann.
 653
 654         First steps of the QtScript API.
 655
 656         Two new classes were created; QScriptEngine and QScriptValue.
 657         The first should encapsulate a javascript context and the second a script
 658         value.
 659
 660         This API is still in development, so it isn't compiled by default.
 661         To trigger compilation, pass --qmakearg="CONFIG+=build-qtscript" to
 662         build-webkit.
 663
 664         https://bugs.webkit.org/show_bug.cgi?id=32565
 665
 666         * qt/api/QtScript.pro: Added.
 667         * qt/api/qscriptconverter_p.h: Added.
 668         (QScriptConverter::toString):
 669         * qt/api/qscriptengine.cpp: Added.
 670         (QScriptEngine::QScriptEngine):
 671         (QScriptEngine::~QScriptEngine):
 672         (QScriptEngine::evaluate):
 673         (QScriptEngine::collectGarbage):
 674         * qt/api/qscriptengine.h: Added.
 675         * qt/api/qscriptengine_p.cpp: Added.
 676         (QScriptEnginePrivate::QScriptEnginePrivate):
 677         (QScriptEnginePrivate::~QScriptEnginePrivate):
 678         (QScriptEnginePrivate::evaluate):
 679         * qt/api/qscriptengine_p.h: Added.
 680         (QScriptEnginePrivate::get):
 681         (QScriptEnginePrivate::collectGarbage):
 682         (QScriptEnginePrivate::makeJSValue):
 683         (QScriptEnginePrivate::context):
 684         * qt/api/qscriptvalue.cpp: Added.
 685         (QScriptValue::QScriptValue):
 686         (QScriptValue::~QScriptValue):
 687         (QScriptValue::isValid):
 688         (QScriptValue::isBool):
 689         (QScriptValue::isBoolean):
 690         (QScriptValue::isNumber):
 691         (QScriptValue::isNull):
 692         (QScriptValue::isString):
 693         (QScriptValue::isUndefined):
 694         (QScriptValue::isError):
 695         (QScriptValue::isObject):
 696         (QScriptValue::isFunction):
 697         (QScriptValue::toString):
 698         (QScriptValue::toNumber):
 699         (QScriptValue::toBool):
 700         (QScriptValue::toBoolean):
 701         (QScriptValue::toInteger):
 702         (QScriptValue::toInt32):
 703         (QScriptValue::toUInt32):
 704         (QScriptValue::toUInt16):
 705         (QScriptValue::call):
 706         (QScriptValue::engine):
 707         (QScriptValue::operator=):
 708         (QScriptValue::equals):
 709         (QScriptValue::strictlyEquals):
 710         * qt/api/qscriptvalue.h: Added.
 711         (QScriptValue::):
 712         * qt/api/qscriptvalue_p.h: Added.
 713         (QScriptValuePrivate::):
 714         (QScriptValuePrivate::get):
 715         (QScriptValuePrivate::QScriptValuePrivate):
 716         (QScriptValuePrivate::isValid):
 717         (QScriptValuePrivate::isBool):
 718         (QScriptValuePrivate::isNumber):
 719         (QScriptValuePrivate::isNull):
 720         (QScriptValuePrivate::isString):
 721         (QScriptValuePrivate::isUndefined):
 722         (QScriptValuePrivate::isError):
 723         (QScriptValuePrivate::isObject):
 724         (QScriptValuePrivate::isFunction):
 725         (QScriptValuePrivate::toString):
 726         (QScriptValuePrivate::toNumber):
 727         (QScriptValuePrivate::toBool):
 728         (QScriptValuePrivate::toInteger):
 729         (QScriptValuePrivate::toInt32):
 730         (QScriptValuePrivate::toUInt32):
 731         (QScriptValuePrivate::toUInt16):
 732         (QScriptValuePrivate::equals):
 733         (QScriptValuePrivate::strictlyEquals):
 734         (QScriptValuePrivate::assignEngine):
 735         (QScriptValuePrivate::call):
 736         (QScriptValuePrivate::engine):
 737         (QScriptValuePrivate::context):
 738         (QScriptValuePrivate::value):
 739         (QScriptValuePrivate::object):
 740         (QScriptValuePrivate::inherits):
 741         (QScriptValuePrivate::isJSBased):
 742         (QScriptValuePrivate::isNumberBased):
 743         (QScriptValuePrivate::isStringBased):
 744         * qt/api/qtscriptglobal.h: Added.
 745         * qt/tests/qscriptengine/qscriptengine.pro: Added.
 746         * qt/tests/qscriptengine/tst_qscriptengine.cpp: Added.
 747         (tst_QScriptEngine::tst_QScriptEngine):
 748         (tst_QScriptEngine::~tst_QScriptEngine):
 749         (tst_QScriptEngine::init):
 750         (tst_QScriptEngine::cleanup):
 751         (tst_QScriptEngine::collectGarbage):
 752         (tst_QScriptEngine::evaluate):
 753         * qt/tests/qscriptvalue/qscriptvalue.pro: Added.
 754         * qt/tests/qscriptvalue/tst_qscriptvalue.cpp: Added.
 755         (tst_QScriptValue::tst_QScriptValue):
 756         (tst_QScriptValue::~tst_QScriptValue):
 757         (tst_QScriptValue::init):
 758         (tst_QScriptValue::cleanup):
 759         (tst_QScriptValue::ctor):
 760         (tst_QScriptValue::toString_data):
 761         (tst_QScriptValue::toString):
 762         (tst_QScriptValue::copyConstructor_data):
 763         (tst_QScriptValue::copyConstructor):
 764         (tst_QScriptValue::assignOperator_data):
 765         (tst_QScriptValue::assignOperator):
 766         (tst_QScriptValue::dataSharing):
 767         (tst_QScriptValue::constructors_data):
 768         (tst_QScriptValue::constructors):
 769         (tst_QScriptValue::call):
 770         * qt/tests/tests.pri: Added.
 771         * qt/tests/tests.pro: Added.
 772
 773 2010-01-25  Dmitry Titov  <dimich@chromium.org>
 774
 775         Reviewed by David Levin.
 776
 777         Fix Chromium Linux tests: the pthread functions on Linux produce segfault if they receive 0 thread handle.
 778         After r53714, we can have 0 thread handles passed to pthread_join and pthread_detach if corresponding threads
 779         were already terminated and their threadMap entries cleared.
 780         Add a 0 check.
 781
 782         * wtf/ThreadingPthreads.cpp:
 783         (WTF::waitForThreadCompletion):
 784         (WTF::detachThread):
 785
 786 2010-01-24  Laszlo Gombos  <laszlo.1.gombos@nokia.com>
 787
 788         Reviewed by Maciej Stachowiak.
 789
 790         Refactor JITStubs.cpp so that DEFINE_STUB_FUNCTION is only used once for each function
 791         https://bugs.webkit.org/show_bug.cgi?id=33866
 792
 793         Place the guard USE(JSVALUE32_64) inside the body of the DEFINE_STUB_FUNCTION
 794         macro for those functions that are always present.
 795
 796         * jit/JITStubs.cpp:
 797         (JSC::DEFINE_STUB_FUNCTION):
 798
 799 2010-01-22  Kevin Watters  <kevinwatters@gmail.com>
 800
 801         Reviewed by Kevin Ollivier.
 802
 803         [wx] Remove the Bakefile build system, which is no longer being used.
 804
 805         https://bugs.webkit.org/show_bug.cgi?id=34022
 806
 807         * JavaScriptCoreSources.bkl: Removed.
 808         * jscore.bkl: Removed.
 809
 810 2010-01-22  Steve Falkenburg  <sfalken@apple.com>
 811
 812         Reviewed by Darin Adler.
 813
 814         https://bugs.webkit.org/show_bug.cgi?id=34025
 815         Enable client-based Geolocation abstraction for Mac, Windows AppleWebKit targets.
 816
 817         * Configurations/FeatureDefines.xcconfig:
 818
 819 2010-01-22  Dmitry Titov  <dimich@chromium.org>
 820
 821         Not reviewed, attempted Snow Leopard build fix.
 822
 823         * wtf/ThreadingPthreads.cpp: Add a forward declaration of a function which is not 'static'.
 824
 825 2009-01-22  Dmitry Titov  <dimich@chromium.org>
 826
 827         Reviewed by Maciej Stachowiak.
 828
 829         Fix the leak of ThreadIdentifiers in threadMap across threads.
 830         https://bugs.webkit.org/show_bug.cgi?id=32689
 831
 832         Test is added to DumpRenderTree.mm.
 833
 834         * Android.mk: Added file ThreadIdentifierDataPthreads.(h|cpp) to build.
 835         * Android.v8.wtf.mk: Ditto.
 836         * GNUmakefile.am: Ditto.
 837         * JavaScriptCore.gyp/JavaScriptCore.gyp: Ditto.
 838         * JavaScriptCore.gypi: Ditto.
 839         * JavaScriptCore.xcodeproj/project.pbxproj: Ditto.
 840
 841         * wtf/ThreadIdentifierDataPthreads.cpp: Added. Contains custom implementation of thread-specific data that uses custom destructor.
 842         (WTF::ThreadIdentifierData::~ThreadIdentifierData): Removes the ThreadIdentifier from the threadMap.
 843         (WTF::ThreadIdentifierData::identifier):
 844         (WTF::ThreadIdentifierData::initialize):
 845         (WTF::ThreadIdentifierData::destruct): Custom thread-specific destructor. Resets the value for the key again to cause second invoke.
 846         (WTF::ThreadIdentifierData::initializeKeyOnceHelper):
 847         (WTF::ThreadIdentifierData::initializeKeyOnce): Need to use pthread_once since initialization may come on any thread(s).
 848         * wtf/ThreadIdentifierDataPthreads.h: Added.
 849         (WTF::ThreadIdentifierData::ThreadIdentifierData):
 850
 851         * wtf/Threading.cpp:
 852         (WTF::threadEntryPoint): Move initializeCurrentThreadInternal to after the lock to make
 853                                  sure it is invoked when ThreadIdentifier is already established.
 854
 855         * wtf/Threading.h: Rename setThreadNameInternal -> initializeCurrentThreadInternal since it does more then only set the name now.
 856         * wtf/ThreadingNone.cpp:
 857         (WTF::initializeCurrentThreadInternal): Ditto.
 858         * wtf/ThreadingWin.cpp:
 859         (WTF::initializeCurrentThreadInternal): Ditto.
 860         (WTF::initializeThreading): Ditto.
 861         * wtf/gtk/ThreadingGtk.cpp:
 862         (WTF::initializeCurrentThreadInternal): Ditto.
 863         * wtf/qt/ThreadingQt.cpp:
 864         (WTF::initializeCurrentThreadInternal): Ditto.
 865
 866         * wtf/ThreadingPthreads.cpp:
 867         (WTF::establishIdentifierForPthreadHandle):
 868         (WTF::clearPthreadHandleForIdentifier): Make it not 'static' so the ~ThreadIdentifierData() in another file can call it.
 869         (WTF::initializeCurrentThreadInternal): Set the thread-specific data. The ThreadIdentifier is already established by creating thread.
 870         (WTF::waitForThreadCompletion): Remove call to clearPthreadHandleForIdentifier(threadID) since it is now done in ~ThreadIdentifierData().
 871         (WTF::detachThread): Ditto.
 872         (WTF::currentThread): Use the thread-specific data to get the ThreadIdentifier. It's many times faster then Mutex-protected iteration through the map.
 873                               Also, set the thread-specific data if called first time on the thread.
 874
 875 2010-01-21  Kwang Yul Seo  <skyul@company100.net>
 876
 877         Reviewed by Alexey Proskuryakov.
 878
 879         Add ThreadSpecific for ENABLE(SINGLE_THREADED)
 880         https://bugs.webkit.org/show_bug.cgi?id=33878
 881
 882         Implement ThreadSpecific with a simple getter/setter
 883         when ENABLE(SINGLE_THREADED) is true.
 884
 885         Due to the change in https://bugs.webkit.org/show_bug.cgi?id=33236,
 886         an implementation of ThreadSpecific must be available to build WebKit.
 887         This causes a build failure for platforms without a proper
 888         ThreadSpecific implementation.
 889
 890         * wtf/ThreadSpecific.h:
 891         (WTF::::ThreadSpecific):
 892         (WTF::::~ThreadSpecific):
 893         (WTF::::get):
 894         (WTF::::set):
 895         (WTF::::destroy):
 896
 897 2010-01-21  Kwang Yul Seo  <skyul@company100.net>
 898
 899         Reviewed by Maciej Stachowiak.
 900
 901         Add fastStrDup to FastMalloc
 902         https://bugs.webkit.org/show_bug.cgi?id=33937
 903
 904         The new string returned by fastStrDup is obtained with fastMalloc,
 905         and can be freed with fastFree. This makes the memory management
 906         more consistent because we don't need to keep strdup allocated pointers
 907         and free them with free(). Instead we can use fastFree everywhere.
 908
 909         * wtf/FastMalloc.cpp:
 910         (WTF::fastStrDup):
 911         * wtf/FastMalloc.h:
 912
 913 2010-01-21  Brady Eidson  <beidson@apple.com>
 914
 915         Reviewed by Maciej Stachowiak.
 916
 917         history.back() for same-document history traversals isn't synchronous as the specification states.
 918         <rdar://problem/7535011> and https://bugs.webkit.org/show_bug.cgi?id=33538
 919
 920         * wtf/Platform.h: Add a "HISTORY_ALWAYS_ASYNC" enable and turn it on for Chromium.
 921
 922 2010-01-21  Geoffrey Garen  <ggaren@apple.com>
 923
 924         Reviewed by Oliver Hunt.
 925
 926         Always create a prototype for automatically managed classes.
 927
 928         This fixes some errors where prototype chains were not correctly hooked
 929         up, and also ensures that API classes work correctly with features like
 930         instanceof.
 931
 932         * API/JSClassRef.cpp:
 933         (OpaqueJSClass::create): Cleaned up some of this code. Also changed it
 934         to always create a prototype class.
 935
 936         * API/tests/testapi.c:
 937         (Derived2_class):
 938         (main): Fixed a null value crash in the exception checking code.
 939         * API/tests/testapi.js: Added some tests for the case where a prototype
 940         chain would not be hooked up correctly.
 941
 942 2010-01-21  Oliver Hunt  <oliver@apple.com>
 943
 944         Reviewed by Geoff Garen.
 945
 946         Force JSC to create a prototype chain for API classes with a
 947         parent class but no static functions.
 948
 949         * API/JSClassRef.cpp:
 950         (OpaqueJSClass::create):
 951
 952 2010-01-21  Kent Hansen  <kent.hansen@nokia.com>
 953
 954         Reviewed by Geoffrey Garen.
 955
 956         Object.getOwnPropertyDescriptor always returns undefined for JS API objects
 957         https://bugs.webkit.org/show_bug.cgi?id=33946
 958
 959         Ideally the getOwnPropertyDescriptor() reimplementation should return an
 960         access descriptor that wraps the property getter and setter callbacks, but
 961         that approach is much more involved than returning a value descriptor.
 962         Keep it simple for now.
 963
 964         * API/JSCallbackObject.h:
 965         * API/JSCallbackObjectFunctions.h:
 966         (JSC::::getOwnPropertyDescriptor):
 967         * API/tests/testapi.js:
 968
 969 2010-01-20  Mark Rowe  <mrowe@apple.com>
 970
 971         Build fix.
 972
 973         * wtf/FastMalloc.cpp:
 974         (WTF::TCMalloc_PageHeap::initializeScavenger): Remove unnecessary function call.
 975
 976 2010-01-20  Mark Rowe  <mrowe@apple.com>
 977
 978         Reviewed by Oliver Hunt.
 979
 980         Use the inline i386 assembly for x86_64 as well rather than falling back to using pthread mutexes.
 981
 982         * wtf/TCSpinLock.h:
 983         (TCMalloc_SpinLock::Lock):
 984         (TCMalloc_SpinLock::Unlock):
 985         (TCMalloc_SlowLock):
 986
 987 2010-01-20  Mark Rowe  <mrowe@apple.com>
 988
 989         Reviewed by Oliver Hunt.
 990
 991         <rdar://problem/7215063> Use GCD instead of an extra thread for FastMalloc scavenging on platforms where it is supported
 992
 993         Abstract the background scavenging slightly so that an alternate implementation that uses GCD can be used on platforms
 994         where it is supported.
 995
 996         * wtf/FastMalloc.cpp:
 997         (WTF::TCMalloc_PageHeap::init):
 998         (WTF::TCMalloc_PageHeap::initializeScavenger):
 999         (WTF::TCMalloc_PageHeap::signalScavenger):
1000         (WTF::TCMalloc_PageHeap::shouldContinueScavenging):
1001         (WTF::TCMalloc_PageHeap::Delete):
1002         (WTF::TCMalloc_PageHeap::periodicScavenge):
1003         * wtf/Platform.h:
1004
1005 2010-01-20  Geoffrey Garen  <ggaren@apple.com>
1006
1007         Reviewed by Oliver Hunt.
1008
1009         <rdar://problem/7562708> REGRESSION(53460): Heap::destroy may not run
1010         all destructors
1011
1012         * runtime/Collector.cpp:
1013         (JSC::Heap::freeBlocks): Instead of fully marking protected objects,
1014         just set their mark bits. This prevents protected objects from keeping
1015         unprotected objects alive. Destructor order is not guaranteed, so it's
1016         OK to destroy objects pointed to by protected objects before destroying
1017         protected objects.
1018
1019 2010-01-19  David Levin  <levin@chromium.org>
1020
1021         Reviewed by Oliver Hunt.
1022
1023         CrossThreadCopier needs to support ThreadSafeShared better.
1024         https://bugs.webkit.org/show_bug.cgi?id=33698
1025
1026         * wtf/TypeTraits.cpp: Added tests for the new type traits.
1027         * wtf/TypeTraits.h:
1028         (WTF::IsSubclass): Determines if a class is a derived from another class.
1029         (WTF::IsSubclassOfTemplate): Determines if a class is a derived from a
1030         template class (with one parameter that is unknown).
1031         (WTF::RemoveTemplate): Reveals the type for a template parameter.
1032
1033 2010-01-20  Steve Falkenburg  <sfalken@apple.com>
1034
1035         Reviewed by Darin Adler and Adam Roben.
1036
1037         Feature defines are difficult to maintain on Windows builds
1038         https://bugs.webkit.org/show_bug.cgi?id=33883
1039
1040         FeatureDefines.vsprops are now maintained in a way similar to
1041         Configurations/FeatureDefines.xcconfig, with the added advantage
1042         of having a single FeatureDefines file across all projects.
1043
1044         * Configurations/FeatureDefines.xcconfig: Add comments about keeping feature definitions in sync.
1045         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj: Add FeatureDefines.vsprops inherited property sheet.
1046         * JavaScriptCore.vcproj/WTF/WTF.vcproj: Add FeatureDefines.vsprops inherited property sheet.
1047
1048 2010-01-20  Csaba Osztrogonác  <ossy@webkit.org>
1049
1050         [Qt] Unreviewed buildfix for r53547.
1051
1052         * DerivedSources.pro:
1053
1054 2010-01-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1055
1056         Reviewed by Simon Hausmann.
1057
1058         [Qt] Make extraCompilers for generated sources depend on their scripts
1059
1060         * DerivedSources.pro:
1061
1062 2010-01-19  Brian Weinstein  <bweinstein@apple.com>
1063
1064         Reviewed by Tim Hatcher.
1065
1066         When JavaScriptCore calls Debugger::Exception, have it pass a
1067         hasHandler variable that represents if exception is being handled
1068         in the same function (not in a parent on the call stack).
1069
1070         This just adds a new parameter, no behavior is changed.
1071
1072         * debugger/Debugger.h:
1073         * interpreter/Interpreter.cpp:
1074         (JSC::Interpreter::throwException):
1075
1076 2010-01-18  Maciej Stachowiak  <mjs@apple.com>
1077
1078         Reviewed by Adam Barth.
1079
1080         Inline functions that are hot in DOM manipulation
1081         https://bugs.webkit.org/show_bug.cgi?id=33820
1082
1083         (3% speedup on Dromaeo DOM Core tests)
1084
1085         * runtime/WeakGCMap.h:
1086         (JSC::::get): inline
1087
1088 2010-01-19  Laszlo Gombos  <laszlo.1.gombos@nokia.com>
1089
1090         Unreviewed build fix for JIT with RVCT.
1091
1092         Remove IMPORT statement; cti_vm_throw is already defined in JITStubs.h.
1093         Remove extra ')'.
1094
1095         * jit/JITStubs.cpp:
1096         (JSC::ctiVMThrowTrampoline):
1097
1098 2010-01-19  Geoffrey Garen  <ggaren@apple.com>
1099
1100         Reviewed by Oliver Hunt.
1101
1102         REGRESSION (52082): Crash on worker thread when reloading http://radnan.public.iastate.edu/procedural/
1103         https://bugs.webkit.org/show_bug.cgi?id=33826
1104
1105         This bug was caused by a GC-protected object being destroyed early by
1106         Heap::destroy. Clients of the GC protect APIs (reasonably) expect pointers
1107         to GC-protected memory to be valid.
1108
1109         The solution is to do two passes of tear-down in Heap::destroy. The first
1110         pass tears down all unprotected objects. The second pass ASSERTs that all
1111         previously protected objects are now unprotected, and then tears down
1112         all perviously protected objects. These two passes simulate the two passes
1113         that would have been required to free a protected object during normal GC.
1114
1115         * API/JSContextRef.cpp: Removed some ASSERTs that have moved into Heap.
1116
1117         * runtime/Collector.cpp:
1118         (JSC::Heap::destroy): Moved ASSERTs to here.
1119         (JSC::Heap::freeBlock): Tidied up the use of didShrink by moving its
1120         setter to the function that does the shrinking.
1121         (JSC::Heap::freeBlocks): Implemented above algorithm.
1122         (JSC::Heap::shrinkBlocks): Tidied up the use of didShrink.
1123
1124 2010-01-19  Gavin Barraclough  <barraclough@apple.com>
1125
1126         Reviewed by NOBODY (build fix).
1127
1128         Reverting r53455, breaks 2 javascriptcore tests.
1129
1130         * API/JSContextRef.cpp:
1131         * runtime/Collector.cpp:
1132         (JSC::Heap::destroy):
1133         (JSC::Heap::freeBlock):
1134         (JSC::Heap::freeBlocks):
1135         (JSC::Heap::shrinkBlocks):
1136
1137 2010-01-18  Gavin Barraclough  <barraclough@apple.com>
1138
1139         Reviewed by NOBODY (build fix).
1140
1141         Revert r53454, since it causes much sadness in this world.
1142
1143         * runtime/UString.cpp:
1144         (JSC::UString::spliceSubstringsWithSeparators):
1145         (JSC::UString::replaceRange):
1146         * runtime/UStringImpl.cpp:
1147         (JSC::UStringImpl::baseSharedBuffer):
1148         (JSC::UStringImpl::sharedBuffer):
1149         (JSC::UStringImpl::~UStringImpl):
1150         * runtime/UStringImpl.h:
1151         (JSC::UntypedPtrAndBitfield::UntypedPtrAndBitfield):
1152         (JSC::UntypedPtrAndBitfield::asPtr):
1153         (JSC::UntypedPtrAndBitfield::operator&=):
1154         (JSC::UntypedPtrAndBitfield::operator|=):
1155         (JSC::UntypedPtrAndBitfield::operator&):
1156         (JSC::UStringImpl::create):
1157         (JSC::UStringImpl::cost):
1158         (JSC::UStringImpl::isIdentifier):
1159         (JSC::UStringImpl::setIsIdentifier):
1160         (JSC::UStringImpl::ref):
1161         (JSC::UStringImpl::deref):
1162         (JSC::UStringImpl::checkConsistency):
1163         (JSC::UStringImpl::UStringImpl):
1164         (JSC::UStringImpl::bufferOwnerString):
1165         (JSC::UStringImpl::bufferOwnership):
1166         (JSC::UStringImpl::isStatic):
1167         * wtf/StringHashFunctions.h:
1168         (WTF::stringHash):
1169
1170 2010-01-18  Geoffrey Garen  <ggaren@apple.com>
1171
1172         Reviewed by Oliver Hunt.
1173
1174         REGRESSION (52082): Crash on worker thread when reloading http://radnan.public.iastate.edu/procedural/
1175         https://bugs.webkit.org/show_bug.cgi?id=33826
1176
1177         This bug was caused by a GC-protected object being destroyed early by
1178         Heap::destroy. Clients of the GC protect APIs (reasonably) expect pointers
1179         to GC-protected memory to be valid.
1180
1181         The solution is to do two passes of tear-down in Heap::destroy. The first
1182         pass tears down all unprotected objects. The second pass ASSERTs that all
1183         previously protected objects are now unprotected, and then tears down
1184         all perviously protected objects. These two passes simulate the two passes
1185         that would have been required to free a protected object during normal GC.
1186
1187         * API/JSContextRef.cpp: Removed some ASSERTs that have moved into Heap.
1188
1189         * runtime/Collector.cpp:
1190         (JSC::Heap::destroy): Moved ASSERTs to here.
1191         (JSC::Heap::freeBlock): Tidied up the use of didShrink by moving its
1192         setter to the function that does the shrinking.
1193         (JSC::Heap::freeBlocks): Implemented above algorithm.
1194         (JSC::Heap::shrinkBlocks): Tidied up the use of didShrink.
1195
1196 2010-01-18  Gavin Barraclough  <barraclough@apple.com>
1197
1198         Reviewed by Oliver Hunt.
1199
1200         https://bugs.webkit.org/show_bug.cgi?id=33731
1201         Remove UntypedPtrAndBitfield from UStringImpl (akin to PtrAndFlags).
1202
1203         This break the OS X Leaks tool.  Instead, free up some more bits from the refCount.
1204
1205         * runtime/UStringImpl.cpp:
1206         (JSC::UStringImpl::sharedBuffer):
1207         (JSC::UStringImpl::~UStringImpl):
1208         * runtime/UStringImpl.h:
1209         (JSC::UStringImpl::cost):
1210         (JSC::UStringImpl::checkConsistency):
1211         (JSC::UStringImpl::UStringImpl):
1212         (JSC::UStringImpl::bufferOwnerString):
1213         (JSC::UStringImpl::):
1214         * wtf/StringHashFunctions.h:
1215         (WTF::stringHash):
1216
1217 2010-01-18  Kent Tamura  <tkent@chromium.org>
1218
1219         Reviewed by Darin Adler.
1220
1221         HTMLInputElement::valueAsDate setter support for type=month.
1222         https://bugs.webkit.org/show_bug.cgi?id=33021
1223
1224         Expose the following functions to be used by WebCore:
1225          - WTF::msToyear()
1226          - WTF::dayInYear()
1227          - WTF::monthFromDayInYear()
1228          - WTF::dayInMonthFromDayInYear()
1229
1230         * JavaScriptCore.exp:
1231         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
1232         * wtf/DateMath.cpp:
1233         (WTF::msToYear): Remove "static inline".
1234         (WTF::dayInYear): Remove "static inline".
1235         (WTF::monthFromDayInYear): Remove "static inline".
1236         (WTF::dayInMonthFromDayInYear): Remove "static inline".
1237         * wtf/DateMath.h: Declare the above functions.
1238
1239 2010-01-18  Darin Adler  <darin@apple.com>
1240
1241         Fix build by reverting the previous change.
1242
1243         * runtime/UString.h: Rolled out the FastAllocBase base class.
1244         It was making UString larger, and therefore JSString larger,
1245         and too big for a garbage collection cell.
1246
1247         This raises the unpleasant possibility that many classes became
1248         larger because we added the FastAllocBase base class. I am
1249         worried about this, and it needs to be investigated.
1250
1251 2010-01-18  Zoltan Horvath  <zoltan@webkit.org>
1252
1253         Reviewed by Darin Adler.
1254
1255         Allow custom memory allocation control for UString class
1256         https://bugs.webkit.org/show_bug.cgi?id=27831
1257
1258         Inherits the following class from FastAllocBase because it is
1259         instantiated by 'new' and no need to be copyable:
1260
1261         class name     - instantiated at:
1262         classs UString - JavaScriptCore/runtime/UString.cpp:160
1263
1264         * runtime/UString.h:
1265
1266 2010-01-18  Evan Cheng  <evan.cheng@apple.com>
1267
1268         Reviewed by Darin Adler.
1269
1270         Add some ALWAYS_INLINE for key functions not inlined by some versions of GCC.
1271         rdar://problem/7553780
1272
1273         * runtime/JSObject.h:
1274         (JSC::JSObject::getPropertySlot): ALWAYS_INLINE both overloads.
1275         * runtime/JSString.h:
1276         (JSC::JSString::JSString): ALWAYS_INLINE the version that takes a UString.
1277         * runtime/UString.h:
1278         (JSC::operator==): ALWAYS_INLINE the version that compares two UString objects.
1279
1280 2010-01-18  Csaba Osztrogonác  <ossy@webkit.org>
1281
1282         Reviewed by Darin Adler.
1283
1284         Delete dftables-xxxxxxxx.in files automatically.
1285         https://bugs.webkit.org/show_bug.cgi?id=33796
1286
1287         * pcre/dftables: unlink unnecessary temporary file.
1288
1289 2010-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1290
1291         Reviewed by Simon Hausmann.
1292
1293         [Qt] Force qmake to generate a single makefile for DerivedSources.pro
1294
1295         * DerivedSources.pro:
1296
1297 2010-01-18  Csaba Osztrogonác  <ossy@webkit.org>
1298
1299         Rubber-stamped by Gustavo Noronha Silva.
1300
1301         Rolling out r53391 and r53392 because of random crashes on buildbots.
1302         https://bugs.webkit.org/show_bug.cgi?id=33731
1303
1304         * bytecode/CodeBlock.h:
1305         (JSC::CallLinkInfo::seenOnce):
1306         (JSC::CallLinkInfo::setSeen):
1307         (JSC::MethodCallLinkInfo::MethodCallLinkInfo):
1308         (JSC::MethodCallLinkInfo::seenOnce):
1309         (JSC::MethodCallLinkInfo::setSeen):
1310         * jit/JIT.cpp:
1311         (JSC::JIT::unlinkCall):
1312         * jit/JITPropertyAccess.cpp:
1313         (JSC::JIT::patchMethodCallProto):
1314         * runtime/UString.cpp:
1315         (JSC::UString::spliceSubstringsWithSeparators):
1316         (JSC::UString::replaceRange):
1317         * runtime/UString.h:
1318         * runtime/UStringImpl.cpp:
1319         (JSC::UStringImpl::baseSharedBuffer):
1320         (JSC::UStringImpl::sharedBuffer):
1321         (JSC::UStringImpl::~UStringImpl):
1322         * runtime/UStringImpl.h:
1323         (JSC::UntypedPtrAndBitfield::UntypedPtrAndBitfield):
1324         (JSC::UntypedPtrAndBitfield::asPtr):
1325         (JSC::UntypedPtrAndBitfield::operator&=):
1326         (JSC::UntypedPtrAndBitfield::operator|=):
1327         (JSC::UntypedPtrAndBitfield::operator&):
1328         (JSC::UStringImpl::create):
1329         (JSC::UStringImpl::cost):
1330         (JSC::UStringImpl::isIdentifier):
1331         (JSC::UStringImpl::setIsIdentifier):
1332         (JSC::UStringImpl::ref):
1333         (JSC::UStringImpl::deref):
1334         (JSC::UStringImpl::checkConsistency):
1335         (JSC::UStringImpl::UStringImpl):
1336         (JSC::UStringImpl::bufferOwnerString):
1337         (JSC::UStringImpl::bufferOwnership):
1338         (JSC::UStringImpl::isStatic):
1339         * wtf/StringHashFunctions.h:
1340         (WTF::stringHash):
1341
1342 2010-01-18  Simon Hausmann  <simon.hausmann@nokia.com>
1343
1344         Reviewed by Kenneth Rohde Christiansen.
1345
1346         Fix the build with strict gcc and RVCT versions: It's not legal to cast a
1347         pointer to a function to a void* without an intermediate cast to a non-pointer
1348         type. A cast to a ptrdiff_t inbetween fixes it.
1349
1350         * runtime/JSString.h:
1351         (JSC::Fiber::JSString):
1352
1353 2010-01-15  Gavin Barraclough  <barraclough@apple.com>
1354
1355         Reviewed by Oliver Hunt.
1356
1357         https://bugs.webkit.org/show_bug.cgi?id=33731
1358         Remove UntypedPtrAndBitfield from UStringImpl (akin to PtrAndFlags).
1359
1360         This break the OS X Leaks tool.  Instead, free up some more bits from the refCount.
1361
1362         * runtime/UStringImpl.cpp:
1363         (JSC::UStringImpl::sharedBuffer):
1364         (JSC::UStringImpl::~UStringImpl):
1365         * runtime/UStringImpl.h:
1366         (JSC::UStringImpl::cost):
1367         (JSC::UStringImpl::checkConsistency):
1368         (JSC::UStringImpl::UStringImpl):
1369         (JSC::UStringImpl::bufferOwnerString):
1370         (JSC::UStringImpl::):
1371         * wtf/StringHashFunctions.h:
1372         (WTF::stringHash):
1373
1374 2010-01-15  Gavin Barraclough  <barraclough@apple.com>
1375
1376         Reviewed by Oliver Hunt.
1377
1378         https://bugs.webkit.org/show_bug.cgi?id=33731
1379         Remove uses of PtrAndFlags from JIT data stuctures.
1380
1381         These break the OS X Leaks tool.  Free up a bit in CallLinkInfo, and invalid
1382         permutation of pointer states in MethodCallLinkInfo to represent the removed bits.
1383
1384         * bytecode/CodeBlock.h:
1385         (JSC::CallLinkInfo::seenOnce):
1386         (JSC::CallLinkInfo::setSeen):
1387         (JSC::MethodCallLinkInfo::MethodCallLinkInfo):
1388         (JSC::MethodCallLinkInfo::seenOnce):
1389         (JSC::MethodCallLinkInfo::setSeen):
1390         * jit/JIT.cpp:
1391         (JSC::JIT::unlinkCall):
1392         * jit/JITPropertyAccess.cpp:
1393         (JSC::JIT::patchMethodCallProto):
1394         * runtime/UString.h:
1395
1396 2010-01-16  Maciej Stachowiak  <mjs@apple.com>
1397
1398         Reviewed by Oliver Hunt.
1399
1400         Cache JS string values made from DOM strings (Dromaeo speedup)
1401         https://bugs.webkit.org/show_bug.cgi?id=33768
1402         <rdar://problem/7353576>
1403
1404         * runtime/JSString.h:
1405         (JSC::jsStringWithFinalizer): Added new mechanism for a string to have an optional
1406         finalizer callback, for the benefit of weak-referencing caches.
1407         (JSC::):
1408         (JSC::Fiber::JSString):
1409         (JSC::Fiber::~JSString):
1410         * runtime/JSString.cpp:
1411         (JSC::JSString::resolveRope): Clear fibers so this doesn't look like a string with a finalizer.
1412         * runtime/WeakGCMap.h: Include "Collector.h" to make this header includable by itself.
1413
1414 2010-01-15  Sam Weinig  <sam@webkit.org>
1415
1416         Reviewed by Maciej Stachowiak.
1417
1418         Fix for <rdar://problem/7548432>
1419         Add ALWAYS_INLINE to jsLess for a 1% speedup on llvm-gcc.
1420
1421         * runtime/Operations.h:
1422         (JSC::jsLess):
1423
1424 2010-01-14  Geoffrey Garen  <ggaren@apple.com>
1425
1426         Reviewed by Oliver Hunt.
1427
1428         REGRESISON: Google maps buttons not working properly
1429         https://bugs.webkit.org/show_bug.cgi?id=31871
1430
1431         REGRESSION(r52948): JavaScript exceptions thrown on Google Maps when
1432         getting directions for a second time
1433         https://bugs.webkit.org/show_bug.cgi?id=33446
1434
1435         SunSpider and v8 report no change.
1436
1437         * interpreter/Interpreter.cpp:
1438         (JSC::Interpreter::tryCacheGetByID): Update our cached offset in case
1439         flattening the dictionary changed any of its offsets.
1440
1441         * jit/JITStubs.cpp:
1442         (JSC::JITThunks::tryCacheGetByID):
1443         (JSC::DEFINE_STUB_FUNCTION):
1444         * runtime/Operations.h:
1445         (JSC::normalizePrototypeChain): ditto
1446
1447 2010-01-14  Gavin Barraclough  <barraclough@apple.com>
1448
1449         Reviewed by Oliver Hunt.
1450
1451         https://bugs.webkit.org/show_bug.cgi?id=33705
1452         UStringImpl::create() should use internal storage
1453
1454         When creating a UStringImpl copying of a UChar*, we can use an internal buffer,
1455         by calling UStringImpl::tryCreateUninitialized().
1456
1457         Also, remove duplicate of copyChars from JSString, call UStringImpl's version.
1458
1459         Small (max 0.5%) progression on Sunspidey.
1460
1461         * runtime/JSString.cpp:
1462         (JSC::JSString::resolveRope):
1463         * runtime/UStringImpl.h:
1464         (JSC::UStringImpl::create):
1465
1466 2010-01-14  Gavin Barraclough  <barraclough@apple.com>
1467
1468         Reviewed by Sam Weinig.
1469
1470         Make naming & behaviour of UString[Impl] methods more consistent.
1471         https://bugs.webkit.org/show_bug.cgi?id=33702
1472
1473         UString::create() creates a copy of the UChar* passed, but UStringImpl::create() assumes
1474         that it should assume ownership of the provided buffer (with UString::createNonCopying()
1475         and UStringImpl::createCopying() providing the alternate behaviours).  Unify on create()
1476         taking a copy of the provided buffer.  For non-copying cases, use the name 'adopt', and
1477         make this method take a Vector<UChar>&.  For cases where non-copying construction was being
1478         used, other than from a Vector<UChar>, change the code to allocate the storage along with
1479         the UStringImpl using UStringImpl::createUninitialized().  (The adopt() method also more
1480         closely matches that of WebCore::StringImpl).
1481
1482         Also, UString::createUninitialized() and UStringImpl::createUninitialized() have incompatible
1483         behaviours, in that the UString form sets the provided UChar* to a null or non-null value to
1484         indicate success or failure, but UStringImpl uses the returned PassRefPtr<UStringImpl> to
1485         indicate when allocation has failed (potentially leaving the output Char* uninitialized).
1486         This is also incompatible with WebCore::StringImpl's behaviour, in that
1487         StringImpl::createUninitialized() will CRASH() if unable to allocate.  Some uses of
1488         createUninitialized() in JSC are unsafe, since they do not test the result for null.
1489         UStringImpl's indication is preferable, since we may want a successful call to set the result
1490         buffer to 0 (specifically, StringImpl returns 0 for the buffer where createUninitialized()
1491         returns the empty string, which seems reasonable to catch bugs early).  UString's method
1492         cannot support UStringImpl's behaviour directly, since it returns an object rather than a
1493         pointer.
1494             - remove UString::createUninitialized(), replace with calls to UStringImpl::createUninitialized()
1495             - create a UStringImpl::tryCreateUninitialized() form UStringImpl::createUninitialized(),
1496               with current behaviour, make createUninitialized() crash on failure to allocate.
1497             - make cases in JSC that do not check the result call createUninitialized(), and cases that do
1498               check call tryCreateUninitialized().
1499
1500         Rename computedHash() to existingHash(), to bring this in line wih WebCore::StringImpl.
1501
1502         * API/JSClassRef.cpp:
1503         (OpaqueJSClassContextData::OpaqueJSClassContextData):
1504         * JavaScriptCore.exp:
1505         * runtime/ArrayPrototype.cpp:
1506         (JSC::arrayProtoFuncToString):
1507         * runtime/Identifier.cpp:
1508         (JSC::CStringTranslator::translate):
1509         (JSC::UCharBufferTranslator::translate):
1510         * runtime/JSString.cpp:
1511         (JSC::JSString::resolveRope):
1512         * runtime/Lookup.cpp:
1513         (JSC::HashTable::createTable):
1514         * runtime/Lookup.h:
1515         (JSC::HashTable::entry):
1516         * runtime/StringBuilder.h:
1517         (JSC::StringBuilder::release):
1518         * runtime/StringConstructor.cpp:
1519         (JSC::stringFromCharCodeSlowCase):
1520         * runtime/StringPrototype.cpp:
1521         (JSC::substituteBackreferencesSlow):
1522         (JSC::stringProtoFuncToLowerCase):
1523         (JSC::stringProtoFuncToUpperCase):
1524         (JSC::stringProtoFuncFontsize):
1525         (JSC::stringProtoFuncLink):
1526         * runtime/Structure.cpp:
1527         (JSC::Structure::despecifyDictionaryFunction):
1528         (JSC::Structure::get):
1529         (JSC::Structure::despecifyFunction):
1530         (JSC::Structure::put):
1531         (JSC::Structure::remove):
1532         (JSC::Structure::insertIntoPropertyMapHashTable):
1533         (JSC::Structure::checkConsistency):
1534         * runtime/Structure.h:
1535         (JSC::Structure::get):
1536         * runtime/StructureTransitionTable.h:
1537         (JSC::StructureTransitionTableHash::hash):
1538         * runtime/UString.cpp:
1539         (JSC::createRep):
1540         (JSC::UString::UString):
1541         (JSC::UString::spliceSubstringsWithSeparators):
1542         (JSC::UString::replaceRange):
1543         (JSC::UString::operator=):
1544         * runtime/UString.h:
1545         (JSC::UString::adopt):
1546         (JSC::IdentifierRepHash::hash):
1547         (JSC::makeString):
1548         * runtime/UStringImpl.h:
1549         (JSC::UStringImpl::adopt):
1550         (JSC::UStringImpl::create):
1551         (JSC::UStringImpl::createUninitialized):
1552         (JSC::UStringImpl::tryCreateUninitialized):
1553         (JSC::UStringImpl::existingHash):
1554
1555 2010-01-13  Kent Hansen  <kent.hansen@nokia.com>
1556
1557         Reviewed by Oliver Hunt.
1558
1559         JSON.stringify and JSON.parse needlessly process properties in the prototype chain
1560         https://bugs.webkit.org/show_bug.cgi?id=33053
1561
1562         * runtime/JSONObject.cpp:
1563         (JSC::Stringifier::Holder::appendNextProperty):
1564         (JSC::Walker::walk):
1565
1566 2010-01-13  Gavin Barraclough  <barraclough@apple.com>
1567
1568         Reviewed by NOBODY (buildfix).
1569
1570         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
1571
1572 2010-01-13  Alexey Proskuryakov  <ap@apple.com>
1573
1574         Reviewed by Darin Adler.
1575
1576         https://bugs.webkit.org/show_bug.cgi?id=33641
1577         Assertion failure in Lexer.cpp if input stream ends while in string escape
1578
1579         Test: fast/js/end-in-string-escape.html
1580
1581         * parser/Lexer.cpp: (JSC::Lexer::lex): Bail out quickly on end of stream, not giving the
1582         assertion a chance to fire.
1583
1584 2010-01-13  Gavin Barraclough  <barraclough@apple.com>
1585
1586         Reviewed by NOBODY (buildfix).
1587
1588         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
1589
1590 2010-01-13  Gavin Barraclough  <barraclough@apple.com>
1591
1592         Rubber stamped by Sam Weinig & Darin Adler.
1593
1594         Three quick fixes to UStringImpl.
1595             - The destroy() method can be switched back to a normal destructor; since we've switched
1596               the way we protect static strings to be using an odd ref-count the destroy() won't abort.
1597             - The cost() calculation logic was wrong.  If you have multiple JSStrings wrapping substrings
1598               of a base string, they would each report the full cost of the base string to the heap.
1599               Instead we should only be reporting once for the base string.
1600             - Remove the overloaded new operator calling fastMalloc, replace this with a 'using' to pick
1601               up the implementation from the parent class.
1602
1603         * JavaScriptCore.exp:
1604         * runtime/UStringImpl.cpp:
1605         (JSC::UStringImpl::~UStringImpl):
1606         * runtime/UStringImpl.h:
1607         (JSC::UStringImpl::cost):
1608         (JSC::UStringImpl::deref):
1609
1610 2010-01-13  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
1611
1612         Reviewed by Simon Hausmann.
1613
1614         [Qt] Split the build process in two different .pro files.
1615         This allows qmake to be run once all source files are available.
1616
1617         * DerivedSources.pro: Added.
1618         * JavaScriptCore.pri: Moved source generation to DerivedSources.pro
1619         * pcre/pcre.pri: Moved source generation to DerivedSources.pro
1620
1621 2010-01-12  Kent Hansen  <kent.hansen@nokia.com>
1622
1623         Reviewed by Geoffrey Garen.
1624
1625         [ES5] Implement Object.getOwnPropertyNames
1626         https://bugs.webkit.org/show_bug.cgi?id=32242
1627
1628         Add an extra argument to getPropertyNames() and getOwnPropertyNames()
1629         (and all reimplementations thereof) that indicates whether non-enumerable
1630         properties should be added.
1631
1632         * API/JSCallbackObject.h:
1633         * API/JSCallbackObjectFunctions.h:
1634         (JSC::::getOwnPropertyNames):
1635         * JavaScriptCore.exp:
1636         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
1637         * debugger/DebuggerActivation.cpp:
1638         (JSC::DebuggerActivation::getOwnPropertyNames):
1639         * debugger/DebuggerActivation.h:
1640         * runtime/Arguments.cpp:
1641         (JSC::Arguments::getOwnPropertyNames):
1642         * runtime/Arguments.h:
1643         * runtime/CommonIdentifiers.h:
1644         * runtime/JSArray.cpp:
1645         (JSC::JSArray::getOwnPropertyNames):
1646         * runtime/JSArray.h:
1647         * runtime/JSByteArray.cpp:
1648         (JSC::JSByteArray::getOwnPropertyNames):
1649         * runtime/JSByteArray.h:
1650         * runtime/JSFunction.cpp:
1651         (JSC::JSFunction::getOwnPropertyNames):
1652         * runtime/JSFunction.h:
1653         * runtime/JSNotAnObject.cpp:
1654         (JSC::JSNotAnObject::getOwnPropertyNames):
1655         * runtime/JSNotAnObject.h:
1656         * runtime/JSObject.cpp:
1657         (JSC::getClassPropertyNames):
1658         (JSC::JSObject::getPropertyNames):
1659         (JSC::JSObject::getOwnPropertyNames):
1660         * runtime/JSObject.h:
1661         * runtime/JSVariableObject.cpp:
1662         (JSC::JSVariableObject::getOwnPropertyNames):
1663         * runtime/JSVariableObject.h:
1664         * runtime/ObjectConstructor.cpp:
1665         (JSC::ObjectConstructor::ObjectConstructor):
1666         (JSC::objectConstructorGetOwnPropertyNames):
1667         * runtime/RegExpMatchesArray.h:
1668         (JSC::RegExpMatchesArray::getOwnPropertyNames):
1669         * runtime/StringObject.cpp:
1670         (JSC::StringObject::getOwnPropertyNames):
1671         * runtime/StringObject.h:
1672         * runtime/Structure.cpp: Rename getEnumerablePropertyNames() to getPropertyNames(), which takes an extra argument.
1673         (JSC::Structure::getPropertyNames):
1674         * runtime/Structure.h:
1675         (JSC::):
1676
1677 2010-01-12  Alexey Proskuryakov  <ap@apple.com>
1678
1679         Reviewed by Darin Adler.
1680
1681         https://bugs.webkit.org/show_bug.cgi?id=33540
1682         Make it possible to build in debug mode with assertions disabled
1683
1684         * jit/JITStubs.cpp: (JSC::DEFINE_STUB_FUNCTION):
1685         * runtime/Identifier.cpp: (JSC::Identifier::checkSameIdentifierTable):
1686         * wtf/FastMalloc.cpp:
1687         * wtf/HashTable.h: (WTF::HashTableConstIterator::checkValidity):
1688         * yarr/RegexCompiler.cpp: (JSC::Yarr::compileRegex):
1689
1690 2009-11-23  Yong Li  <yoli@rim.com>
1691
1692         Reviewed by Adam Treat.
1693
1694         Make GIF decoder support down-sampling
1695         https://bugs.webkit.org/show_bug.cgi?id=31806
1696
1697         * platform/image-decoders/ImageDecoder.cpp:
1698         (WebCore::ImageDecoder::upperBoundScaledY):
1699         (WebCore::ImageDecoder::lowerBoundScaledY):
1700         * platform/image-decoders/ImageDecoder.h:
1701         (WebCore::RGBA32Buffer::scaledRect):
1702         (WebCore::RGBA32Buffer::setScaledRect):
1703         (WebCore::ImageDecoder::scaledSize):
1704         * platform/image-decoders/gif/GIFImageDecoder.cpp:
1705         (WebCore::GIFImageDecoder::sizeNowAvailable):
1706         (WebCore::GIFImageDecoder::initFrameBuffer):
1707         (WebCore::copyOnePixel):
1708         (WebCore::GIFImageDecoder::haveDecodedRow):
1709         (WebCore::GIFImageDecoder::frameComplete):
1710
1711 2010-01-12  Adam Barth  <abarth@webkit.org>
1712
1713         Reviewed by Eric Seidel.
1714
1715         ecma/Date/15.9.5.12-1.js fails every night at midnight
1716         https://bugs.webkit.org/show_bug.cgi?id=28041
1717
1718         Change the test to use a concrete time instead of "now".
1719
1720         * tests/mozilla/ecma/Date/15.9.5.10-1.js:
1721         * tests/mozilla/ecma/Date/15.9.5.12-1.js:
1722
1723 2010-01-11  Csaba Osztrogonác  <ossy@webkit.org>
1724
1725         Reviewed by Ariya Hidayat.
1726
1727         [Qt] Enable JIT and YARR_JIT if (CPU(X86_64) && OS(LINUX) && GCC_VERSION >= 40100)
1728
1729         * wtf/Platform.h:
1730
1731 2010-01-11  Geoffrey Garen  <ggaren@apple.com>
1732
1733         Reviewed by Alexey Proskuryakov.
1734
1735         https://bugs.webkit.org/show_bug.cgi?id=33481
1736         Uninitialized data members in ArrayStorage
1737
1738         SunSpider reports no change.
1739
1740         * runtime/JSArray.cpp:
1741         (JSC::JSArray::JSArray): Initialize missing data members in the two cases
1742         where we don't use fastZeroedMalloc, so it doesn't happen automatically.
1743
1744 2010-01-11  Steve Falkenburg  <sfalken@apple.com>
1745
1746         Reviewed by Sam Weinig.
1747
1748         https://bugs.webkit.org/show_bug.cgi?id=33480
1749
1750         Improve debugging reliability for WTF on Windows.
1751         Store WTF static library's PDB file into a better location.
1752
1753         * JavaScriptCore.vcproj/WTF/WTF.vcproj:
1754
1755 2010-01-11  Steve Falkenburg  <sfalken@apple.com>
1756
1757         Windows build fix.
1758         Remove extraneous entries from def file causing build warning.
1759
1760         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
1761
1762 2010-01-10  Kent Hansen  <kent.hansen@nokia.com>
1763
1764         Reviewed by Darin Adler.
1765
1766         RegExp.prototype.toString returns "//" for empty regular expressions
1767         https://bugs.webkit.org/show_bug.cgi?id=33319
1768
1769         "//" starts a single-line comment, hence "/(?:)/" should be used, according to ECMA.
1770
1771         * runtime/RegExpPrototype.cpp:
1772         (JSC::regExpProtoFuncToString):
1773
1774         * tests/mozilla/ecma_2/RegExp/properties-001.js:
1775         (AddRegExpCases):
1776         * tests/mozilla/js1_2/regexp/toString.js:
1777         Update relevant Mozilla tests (Mozilla has had this behavior since November 2003).
1778
1779 2010-01-10  Darin Adler  <darin@apple.com>
1780
1781         * tests/mozilla/ecma/Array/15.4.1.1.js: Added property allow-tabs.
1782         * tests/mozilla/ecma/Array/15.4.1.2.js: Added property allow-tabs.
1783         * tests/mozilla/ecma/Array/15.4.2.1-1.js: Added property allow-tabs.
1784         * tests/mozilla/ecma/Array/15.4.2.2-1.js: Added property allow-tabs.
1785         * tests/mozilla/ecma/Array/15.4.2.2-2.js: Added property allow-tabs.
1786         * tests/mozilla/ecma/Array/15.4.2.3.js: Added property allow-tabs.
1787         * tests/mozilla/ecma/Array/15.4.3.2.js: Added property allow-tabs.
1788         * tests/mozilla/ecma/Array/15.4.3.js: Added property allow-tabs.
1789         * tests/mozilla/ecma/Array/15.4.4.1.js: Added property allow-tabs.
1790         * tests/mozilla/ecma/Array/15.4.4.js: Added property allow-tabs.
1791         * tests/mozilla/ecma/LexicalConventions/7.7.4.js: Added property allow-tabs.
1792         * tests/mozilla/ecma/Math/15.8.2.13.js: Added property allow-tabs.
1793         * tests/mozilla/ecma/Math/15.8.2.16.js: Added property allow-tabs.
1794         * tests/mozilla/ecma/Math/15.8.2.18.js: Added property allow-tabs.
1795         * tests/mozilla/ecma/Math/15.8.2.2.js: Added property allow-tabs.
1796         * tests/mozilla/ecma/Math/15.8.2.4.js: Added property allow-tabs.
1797         * tests/mozilla/ecma/Math/15.8.2.5.js: Added property allow-tabs.
1798         * tests/mozilla/ecma/Math/15.8.2.7.js: Added property allow-tabs.
1799         * tests/mozilla/ecma/String/15.5.1.js: Added property allow-tabs.
1800         * tests/mozilla/ecma/String/15.5.2.js: Added property allow-tabs.
1801         * tests/mozilla/ecma/String/15.5.3.1-3.js: Added property allow-tabs.
1802         * tests/mozilla/ecma/String/15.5.3.1-4.js: Added property allow-tabs.
1803         * tests/mozilla/ecma/String/15.5.3.js: Added property allow-tabs.
1804         * tests/mozilla/ecma/TypeConversion/9.5-2.js: Added property allow-tabs.
1805         * tests/mozilla/ecma/jsref.js: Modified property allow-tabs.
1806         * tests/mozilla/ecma/shell.js: Modified property allow-tabs.
1807         * tests/mozilla/ecma_2/LexicalConventions/keywords-001.js: Added property allow-tabs.
1808         * tests/mozilla/ecma_2/RegExp/exec-001.js: Added property allow-tabs.
1809         * tests/mozilla/ecma_2/String/match-004.js: Added property allow-tabs.
1810         * tests/mozilla/ecma_2/String/replace-001.js: Added property allow-tabs.
1811         * tests/mozilla/ecma_2/String/split-002.js: Added property allow-tabs.
1812         * tests/mozilla/ecma_2/jsref.js: Modified property allow-tabs.
1813         * tests/mozilla/ecma_2/shell.js: Added property allow-tabs.
1814         * tests/mozilla/ecma_3/Date/shell.js: Modified property allow-tabs.
1815         * tests/mozilla/ecma_3/Exceptions/regress-181654.js: Added property allow-tabs.
1816         * tests/mozilla/ecma_3/RegExp/regress-209067.js: Added property allow-tabs.
1817         * tests/mozilla/ecma_3/RegExp/regress-85721.js: Added property allow-tabs.
1818         * tests/mozilla/importList.html: Added property allow-tabs.
1819         * tests/mozilla/js1_1/shell.js: Added property allow-tabs.
1820         * tests/mozilla/js1_2/Array/general1.js: Added property allow-tabs.
1821         * tests/mozilla/js1_2/Array/general2.js: Added property allow-tabs.
1822         * tests/mozilla/js1_2/Array/slice.js: Added property allow-tabs.
1823         * tests/mozilla/js1_2/Array/splice1.js: Added property allow-tabs.
1824         * tests/mozilla/js1_2/Array/splice2.js: Added property allow-tabs.
1825         * tests/mozilla/js1_2/Objects/toString-001.js: Added property allow-tabs.
1826         * tests/mozilla/js1_2/String/charCodeAt.js: Added property allow-tabs.
1827         * tests/mozilla/js1_2/String/concat.js: Modified property allow-tabs.
1828         * tests/mozilla/js1_2/String/match.js: Added property allow-tabs.
1829         * tests/mozilla/js1_2/String/slice.js: Added property allow-tabs.
1830         * tests/mozilla/js1_2/function/Function_object.js: Added property allow-tabs.
1831         * tests/mozilla/js1_2/function/Number.js: Modified property allow-tabs.
1832         * tests/mozilla/js1_2/function/String.js: Modified property allow-tabs.
1833         * tests/mozilla/js1_2/function/nesting.js: Added property allow-tabs.
1834         * tests/mozilla/js1_2/function/regexparg-1.js: Added property allow-tabs.
1835         * tests/mozilla/js1_2/function/regexparg-2-n.js: Added property allow-tabs.
1836         * tests/mozilla/js1_2/jsref.js: Added property allow-tabs.
1837         * tests/mozilla/js1_2/operator/equality.js: Added property allow-tabs.
1838         * tests/mozilla/js1_2/operator/strictEquality.js: Added property allow-tabs.
1839         * tests/mozilla/js1_2/regexp/RegExp_dollar_number.js: Added property allow-tabs.
1840         * tests/mozilla/js1_2/regexp/RegExp_input.js: Added property allow-tabs.
1841         * tests/mozilla/js1_2/regexp/RegExp_input_as_array.js: Added property allow-tabs.
1842         * tests/mozilla/js1_2/regexp/RegExp_lastIndex.js: Added property allow-tabs.
1843         * tests/mozilla/js1_2/regexp/RegExp_lastMatch.js: Added property allow-tabs.
1844         * tests/mozilla/js1_2/regexp/RegExp_lastMatch_as_array.js: Added property allow-tabs.
1845         * tests/mozilla/js1_2/regexp/RegExp_lastParen.js: Added property allow-tabs.
1846         * tests/mozilla/js1_2/regexp/RegExp_lastParen_as_array.js: Added property allow-tabs.
1847         * tests/mozilla/js1_2/regexp/RegExp_leftContext.js: Added property allow-tabs.
1848         * tests/mozilla/js1_2/regexp/RegExp_leftContext_as_array.js: Added property allow-tabs.
1849         * tests/mozilla/js1_2/regexp/RegExp_multiline.js: Added property allow-tabs.
1850         * tests/mozilla/js1_2/regexp/RegExp_multiline_as_array.js: Added property allow-tabs.
1851         * tests/mozilla/js1_2/regexp/RegExp_object.js: Added property allow-tabs.
1852         * tests/mozilla/js1_2/regexp/RegExp_rightContext.js: Added property allow-tabs.
1853         * tests/mozilla/js1_2/regexp/RegExp_rightContext_as_array.js: Added property allow-tabs.
1854         * tests/mozilla/js1_2/regexp/alphanumeric.js: Added property allow-tabs.
1855         * tests/mozilla/js1_2/regexp/asterisk.js: Added property allow-tabs.
1856         * tests/mozilla/js1_2/regexp/backslash.js: Added property allow-tabs.
1857         * tests/mozilla/js1_2/regexp/backspace.js: Added property allow-tabs.
1858         * tests/mozilla/js1_2/regexp/beginLine.js: Added property allow-tabs.
1859         * tests/mozilla/js1_2/regexp/character_class.js: Added property allow-tabs.
1860         * tests/mozilla/js1_2/regexp/compile.js: Added property allow-tabs.
1861         * tests/mozilla/js1_2/regexp/control_characters.js: Added property allow-tabs.
1862         * tests/mozilla/js1_2/regexp/digit.js: Added property allow-tabs.
1863         * tests/mozilla/js1_2/regexp/dot.js: Added property allow-tabs.
1864         * tests/mozilla/js1_2/regexp/endLine.js: Added property allow-tabs.
1865         * tests/mozilla/js1_2/regexp/everything.js: Added property allow-tabs.
1866         * tests/mozilla/js1_2/regexp/exec.js: Added property allow-tabs.
1867         * tests/mozilla/js1_2/regexp/flags.js: Added property allow-tabs.
1868         * tests/mozilla/js1_2/regexp/global.js: Added property allow-tabs.
1869         * tests/mozilla/js1_2/regexp/hexadecimal.js: Added property allow-tabs.
1870         * tests/mozilla/js1_2/regexp/ignoreCase.js: Added property allow-tabs.
1871         * tests/mozilla/js1_2/regexp/interval.js: Added property allow-tabs.
1872         * tests/mozilla/js1_2/regexp/octal.js: Added property allow-tabs.
1873         * tests/mozilla/js1_2/regexp/parentheses.js: Added property allow-tabs.
1874         * tests/mozilla/js1_2/regexp/plus.js: Added property allow-tabs.
1875         * tests/mozilla/js1_2/regexp/question_mark.js: Added property allow-tabs.
1876         * tests/mozilla/js1_2/regexp/simple_form.js: Added property allow-tabs.
1877         * tests/mozilla/js1_2/regexp/source.js: Added property allow-tabs.
1878         * tests/mozilla/js1_2/regexp/special_characters.js: Added property allow-tabs.
1879         * tests/mozilla/js1_2/regexp/string_replace.js: Added property allow-tabs.
1880         * tests/mozilla/js1_2/regexp/string_search.js: Added property allow-tabs.
1881         * tests/mozilla/js1_2/regexp/string_split.js: Added property allow-tabs.
1882         * tests/mozilla/js1_2/regexp/test.js: Added property allow-tabs.
1883         * tests/mozilla/js1_2/regexp/toString.js: Added property allow-tabs.
1884         * tests/mozilla/js1_2/regexp/vertical_bar.js: Added property allow-tabs.
1885         * tests/mozilla/js1_2/regexp/whitespace.js: Added property allow-tabs.
1886         * tests/mozilla/js1_2/regexp/word_boundary.js: Added property allow-tabs.
1887         * tests/mozilla/js1_2/shell.js: Added property allow-tabs.
1888         * tests/mozilla/js1_2/statements/break.js: Added property allow-tabs.
1889         * tests/mozilla/js1_2/statements/continue.js: Added property allow-tabs.
1890         * tests/mozilla/js1_2/statements/do_while.js: Added property allow-tabs.
1891         * tests/mozilla/js1_2/statements/switch.js: Added property allow-tabs.
1892         * tests/mozilla/js1_2/statements/switch2.js: Added property allow-tabs.
1893         * tests/mozilla/js1_3/shell.js: Added property allow-tabs.
1894         * tests/mozilla/js1_4/shell.js: Added property allow-tabs.
1895         * tests/mozilla/js1_5/Regress/regress-111557.js: Added property allow-tabs.
1896         * tests/mozilla/js1_5/Regress/regress-216320.js: Added property allow-tabs.
1897         * tests/mozilla/menuhead.html: Added property allow-tabs.
1898         * tests/mozilla/mklistpage.pl: Added property allow-tabs.
1899         * tests/mozilla/runtests.pl: Added property allow-tabs.
1900
1901 2010-01-08  Daniel Bates  <dbates@webkit.org>
1902
1903         Reviewed by Adam Barth.
1904
1905         https://bugs.webkit.org/show_bug.cgi?id=33417
1906
1907         Cleans up style errors exposed by the patch for bug #33198.
1908         Moreover, fixes all "Weird number of spaces at line-start. Are you using a 4-space indent?"
1909         errors reported by check-webkit-style.
1910
1911         No functionality was changed. So, no new tests.
1912
1913         * wtf/Platform.h:
1914
1915 2010-01-08  Kent Hansen  <kent.hansen@nokia.com>
1916
1917         Reviewed by Eric Seidel.
1918
1919         Don't store RegExp flags string representation
1920         https://bugs.webkit.org/show_bug.cgi?id=33321
1921
1922         It's unused; the string representation is reconstructed from flags.
1923
1924         * runtime/RegExp.cpp:
1925         (JSC::RegExp::RegExp):
1926         * runtime/RegExp.h:
1927
1928 2010-01-08  Geoffrey Garen  <ggaren@apple.com>
1929
1930         Reviewed by Oliver Hunt.
1931
1932         Memory use grows grows possibly unbounded in this JavaScript Array test case
1933         https://bugs.webkit.org/show_bug.cgi?id=31675
1934
1935         This fixes one observed bug in this test case, which is that
1936         arrays don't report extra cost for the sparse value maps.
1937
1938         SunSpider reports a small speedup.
1939
1940         * runtime/JSArray.cpp:
1941         (JSC::JSArray::putSlowCase): Report extra memory cost for
1942         the sparse value map.
1943         * runtime/JSArray.h:
1944
1945 2010-01-08  Yong Li  <yoli@rim.com>
1946
1947         Reviewed by Darin Adler.
1948
1949         Remove unnecessary #include from FastMalloc.cpp
1950         https://bugs.webkit.org/show_bug.cgi?id=33393
1951
1952         * wtf/FastMalloc.cpp:
1953
1954 2010-01-08  Eric Seidel  <eric@webkit.org>
1955
1956         No review, rolling out r52983.
1957         http://trac.webkit.org/changeset/52983
1958         https://bugs.webkit.org/show_bug.cgi?id=33321
1959
1960         Broke 59 JavaScriptCore tests.  I don't think Kent knew about
1961         run-javascriptcore-tests.  Sadly neither does the commit-bot,
1962         yet.
1963
1964         * runtime/RegExp.cpp:
1965         (JSC::RegExp::RegExp):
1966         * runtime/RegExp.h:
1967         (JSC::RegExp::flags):
1968
1969 2010-01-08  Eric Seidel  <eric@webkit.org>
1970
1971         No review, rolling out r52981.
1972         http://trac.webkit.org/changeset/52981
1973         https://bugs.webkit.org/show_bug.cgi?id=33319
1974
1975         Caused two JS tests to start failing:
1976         ecma_2/RegExp/properties-001.js and js1_2/regexp/toString.js
1977
1978         * runtime/RegExpPrototype.cpp:
1979         (JSC::regExpProtoFuncToString):
1980
1981 2010-01-08  Kent Hansen  <kent.hansen@nokia.com>
1982
1983         Reviewed by Darin Adler.
1984
1985         Don't store RegExp flags string representation
1986         https://bugs.webkit.org/show_bug.cgi?id=33321
1987
1988         It's unused; the string representation is reconstructed from flags.
1989
1990         * runtime/RegExp.cpp:
1991         (JSC::RegExp::RegExp):
1992         * runtime/RegExp.h:
1993
1994 2010-01-08  Kent Hansen  <kent.hansen@nokia.com>
1995
1996         Reviewed by Darin Adler.
1997
1998         RegExp.prototype.toString returns "//" for empty regular expressions
1999         https://bugs.webkit.org/show_bug.cgi?id=33319
2000
2001         "//" starts a single-line comment, hence "/(?:)/" should be used, according to ECMA.
2002
2003         * runtime/RegExpPrototype.cpp:
2004         (JSC::regExpProtoFuncToString):
2005
2006 2010-01-08  Norbert Leser  <norbert.leser@nokia.com>
2007
2008         Reviewed by Darin Adler.
2009
2010         RVCT compiler with "-Otime -O3" optimization tries to optimize out
2011         inline new'ed pointers that are passed as arguments.
2012         Proposed patch assigns new'ed pointer explicitly outside function call.
2013
2014         https://bugs.webkit.org/show_bug.cgi?id=33084
2015
2016         * API/JSClassRef.cpp:
2017         (OpaqueJSClass::OpaqueJSClass):
2018         (OpaqueJSClassContextData::OpaqueJSClassContextData):
2019
2020 2010-01-08  Gabor Loki  <loki@webkit.org>
2021
2022         Reviewed by Gavin Barraclough.
2023
2024         Remove an unnecessary cacheFlush from ARM_TRADITIONAL JIT
2025         https://bugs.webkit.org/show_bug.cgi?id=33203
2026
2027         * assembler/ARMAssembler.cpp: Remove obsolete linkBranch function.
2028         (JSC::ARMAssembler::executableCopy): Inline a clean linkBranch code.
2029         * assembler/ARMAssembler.h:
2030         (JSC::ARMAssembler::getLdrImmAddress): Use inline function.
2031         (JSC::ARMAssembler::getLdrImmAddressOnPool): Ditto.
2032         (JSC::ARMAssembler::patchPointerInternal): Remove an unnecessary cacheFlush.
2033         (JSC::ARMAssembler::linkJump): Use patchPointerInternal instead of linkBranch.
2034         (JSC::ARMAssembler::linkCall): Ditto.
2035         (JSC::ARMAssembler::relinkCall): Ditto.
2036
2037 2010-01-07  Gabor Loki  <loki@webkit.org>
2038
2039         Reviewed by Gavin Barraclough.
2040
2041         Build fix for JSVALUE32 when ENABLE_JIT_OPTIMIZE* are disabled
2042         https://bugs.webkit.org/show_bug.cgi?id=33311
2043
2044         Move compileGetDirectOffset function to common part of JSVALUE32
2045
2046         * jit/JITPropertyAccess.cpp:
2047         (JSC::JIT::compileGetDirectOffset):
2048
2049 2010-01-07  Laszlo Gombos  <laszlo.1.gombos@nokia.com>
2050
2051         Reviewed by Maciej Stachowiak.
2052
2053         Allow call sites to determine if ASSERT_* and LOG_* macros are operational
2054         https://bugs.webkit.org/show_bug.cgi?id=33020
2055
2056         * wtf/Assertions.h: Set ASSERT_MSG_DISABLED, FATAL_DISABLED,
2057         ERROR_DISABLED, LOG_DISABLED to 1 if the compiler does not support
2058         variadic macros. Refactor for better readibility.
2059
2060 2010-01-07  Daniel Bates  <dbates@rim.com>
2061
2062         Reviewed by Eric Seidel.
2063
2064         https://bugs.webkit.org/show_bug.cgi?id=32987
2065
2066         Added ENABLE_XHTMLMP flag. Disabled by default.
2067
2068         * Configurations/FeatureDefines.xcconfig:
2069
2070 2010-01-07  Laszlo Gombos  <laszlo.1.gombos@nokia.com>
2071
2072         Reviewed by Gavin Barraclough.
2073
2074         [Symbian] Port ARM traditional JIT Trampolines to RVCT
2075         https://bugs.webkit.org/show_bug.cgi?id=30552
2076
2077         Take the GCC implementation and mechanically convert
2078         it to RVCT syntax.
2079
2080         Use 'bx rX' instead of 'mov pc, rX' when it is available.
2081
2082         Developed in cooperation with Iain Campbell and Gabor Loki.
2083
2084         * JavaScriptCore.pri: Extra step to generate RVCT stubs. The
2085         script generation intentionally executed all the time not just
2086         for RVCT targets.
2087
2088         * create_rvct_stubs: Added. Perl script to expand precompiler macros
2089         for RVCT assembler - the template is defined in JITStubs.cpp.
2090
2091         * jit/JITStubs.cpp:
2092         (JSC::ctiTrampoline):
2093         (JSC::ctiVMThrowTrampoline):
2094         (JSC::ctiOpThrowNotCaught):
2095
2096 2010-01-07  Geoffrey Garen  <ggaren@apple.com>
2097
2098         Reviewed by Sam Weinig.
2099
2100         Fix a crash seen on the buildbots.
2101
2102         * runtime/JSGlobalObject.cpp:
2103         (JSC::JSGlobalObject::init): Disable specific function tracking here,
2104         instead of in WebCore, to ensure that the disabling happens before a
2105         specific function can be registered.
2106
2107 2010-01-07  Alexey Proskuryakov  <ap@apple.com>
2108
2109         Mac build fix.
2110
2111         * JavaScriptCore.exp: Export new JSGlobalData static data members.
2112
2113 2010-01-07  Alexey Proskuryakov  <ap@apple.com>
2114
2115         Reviewed by Geoffrey Garen.
2116
2117         https://bugs.webkit.org/show_bug.cgi?id=33057
2118         REGRESSION(r49365): typeof(xhr.responseText) != "string" in Windows
2119
2120         <rdar://problem/7296920> REGRESSION: WebKit fails to start PeaceKeeper benchmark
2121
2122         Test: fast/js/webcore-string-comparison.html
2123
2124         In r49365, some code was moved from JSString.cpp to JSString.h, and as a result, WebCore
2125         got a way to directly instantiate JSStrings over DLL borders. Since vftable for JSString was
2126         not exported, objects created from WebCore got a different vptr, and JavaScriptCore
2127         optimizations that relied on vptr of all JSString objects being equal failed.
2128
2129         * config.h: Added a JS_EXPORTCLASS macro for exporting classes. It's currently the same as
2130         JS_EXPORTDATA, but it clearly needed a new name.
2131
2132         * runtime/InitializeThreading.cpp:
2133         (JSC::initializeThreadingOnce):
2134         * runtime/JSGlobalData.cpp:
2135         (JSC::JSGlobalData::storeVPtrs):
2136         (JSC::JSGlobalData::JSGlobalData):
2137         (JSC::JSGlobalData::createNonDefault):
2138         (JSC::JSGlobalData::create):
2139         (JSC::JSGlobalData::sharedInstance):
2140         * runtime/JSGlobalData.h:
2141         Store vptrs just once, no need to repeatedly pick and copy them. This makes it possible to
2142         assert vptr correctness in object destructors (which don't have access to JSGlobalData,
2143         and even Heap::heap(this) will fail for fake objects created from storeVPtrs()).
2144
2145         * runtime/JSArray.cpp: (JSC::JSArray::~JSArray): Assert that vptr is what we expect it to be.
2146         It's important to assert in destructor, because MSVC changes the vptr after constructor
2147         is invoked.
2148         * runtime/JSByteArray.cpp: (JSC::JSByteArray::~JSByteArray): Ditto.
2149         * runtime/JSByteArray.h: Ditto.
2150         * runtime/JSFunction.h: Ditto.
2151         * runtime/JSFunction.cpp: (JSC::JSFunction::~JSFunction): Ditto.
2152
2153         * runtime/JSCell.h: (JSC::JSCell::setVPtr): Added a method to substitute vptr for another
2154         one.
2155
2156         * runtime/JSString.h: Export JSString class together with its vftable, and tell other
2157         libraries tp import it. This is needed on platforms that have a separate JavaScriptCore
2158         dynamic library - and on Mac, we already did the export via JavaScriptCore.exp.
2159         (JSC::JSString::~JSString): Assert tha vptr is what we expect it to be.
2160         (JSC::fixupVPtr): Store a previously saved primary vftable pointer (do nothing if building
2161         JavaScriptCore itself).
2162         (JSC::jsSingleCharacterString): Call fixupVPtr in case this is call across DLL boundary.
2163         (JSC::jsSingleCharacterSubstring): Ditto.
2164         (JSC::jsNontrivialString): Ditto.
2165         (JSC::jsString): Ditto.
2166         (JSC::jsSubstring): Ditto.
2167         (JSC::jsOwnedString): Ditto.
2168
2169         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def: Export the new static
2170         JSGlobalData members that are used in WebCore via inline functions.
2171
2172 2010-01-07  Geoffrey Garen  <ggaren@apple.com>
2173
2174         Reviewed by Sam Weinig.
2175
2176         Safari memory usage skyrockets using new Google AdWords interface
2177         https://bugs.webkit.org/show_bug.cgi?id=33343
2178
2179         The memory use was caused by the global object creating too many structures
2180         as it thrashed between different specific functions.
2181
2182         * runtime/Structure.cpp:
2183         (JSC::Structure::Structure):
2184         (JSC::Structure::addPropertyTransition):
2185         (JSC::Structure::changePrototypeTransition):
2186         (JSC::Structure::despecifyFunctionTransition):
2187         (JSC::Structure::addAnonymousSlotsTransition):
2188         (JSC::Structure::getterSetterTransition):
2189         (JSC::Structure::toDictionaryTransition):
2190         (JSC::Structure::addPropertyWithoutTransition):
2191         (JSC::Structure::despecifyAllFunctions):
2192         * runtime/Structure.h:
2193         (JSC::Structure::disableSpecificFunctionTracking): Track a thrash count
2194         for specific functions. Disable specific function tracking once the
2195         thrash count has been hit.
2196
2197 2010-01-07  Csaba Osztrogonác  <ossy@webkit.org>
2198
2199         Reviewed by Simon Hausmann.
2200
2201         [Qt] Enable JIT in debug mode on win32 after r51141 fixed the crashes.
2202
2203         * JavaScriptCore.pri:
2204
2205 2010-01-07  Zoltan Horvath  <zoltan@webkit.org>
2206
2207         Reviewed by Holger Freyther.
2208
2209         [Mac] Build fix when FAST_MALLOC_MATCH_VALIDATION=1
2210         https://bugs.webkit.org/show_bug.cgi?id=33312
2211
2212         Using of operator += cause compile error on Mac, so it is changed to
2213         "= static_cast<AllocAlignmentInteger*>(old_ptr) + 1".
2214
2215         * wtf/FastMalloc.cpp:
2216         (WTF::TCMallocStats::realloc):
2217
2218 2010-01-07  Zoltan Horvath  <zoltan@webkit.org>
2219
2220         Reviewed by Holger Freyther.
2221
2222         [Qt] Build fix when FAST_MALLOC_MATCH_VALIDATION=1
2223         https://bugs.webkit.org/show_bug.cgi?id=33312
2224
2225         Remove pByte (committed in r42344 from #20422), because pByte doesn't
2226         exist and it is unnecessary.
2227
2228         * wtf/FastMalloc.cpp:
2229         (WTF::TCMallocStats::realloc):
2230
2231 2010-01-06  Gavin Barraclough  <barraclough@apple.com>
2232
2233         QT build fix.
2234
2235         * runtime/Identifier.cpp:
2236         (JSC::createIdentifierTableSpecific):
2237
2238 2010-01-06  Gavin Barraclough  <barraclough@apple.com>
2239
2240         Windows build fix part I.
2241
2242         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
2243
2244 2010-01-06  Dan Bernstein  <mitz@apple.com>
2245
2246         Build fix
2247
2248         * runtime/Identifier.cpp:
2249         (JSC::createIdentifierTableSpecificCallback):
2250
2251 2010-01-05  Gavin Barraclough  <barraclough@apple.com>
2252
2253         Reviewed by Sam Weinig.
2254
2255         https://bugs.webkit.org/show_bug.cgi?id=33236
2256         Remove m_identifierTable pointer from UString
2257
2258         Currently every string holds a pointer so that during destruction,
2259         if a string has been used as an identifier, it can remove itself
2260         from the table.  By instead accessing the identifierTable via a
2261         thread specific tracking the table associated with the current
2262         globaldata, we can save the memory cost of this pointer.
2263
2264         * API/APIShims.h:
2265         (JSC::APIEntryShimWithoutLock::APIEntryShimWithoutLock):
2266         (JSC::APIEntryShimWithoutLock::~APIEntryShimWithoutLock):
2267         (JSC::APICallbackShim::APICallbackShim):
2268         (JSC::APICallbackShim::~APICallbackShim):
2269
2270             - change the API shims to track the identifierTable of the current JSGlobalData.
2271
2272         * API/JSContextRef.cpp:
2273         (JSContextGroupCreate):
2274
2275             - update creation of JSGlobalData for API usage to use new create method.
2276             - fix shim instanciation bug in JSGlobalContextCreateInGroup.
2277
2278         * JavaScriptCore.exp:
2279         * runtime/Completion.cpp:
2280         (JSC::checkSyntax):
2281         (JSC::evaluate):
2282
2283             - add asserts to check the identifierTable is being tracked correctly.
2284
2285         * runtime/Identifier.cpp:
2286         (JSC::IdentifierTable::~IdentifierTable):
2287         (JSC::IdentifierTable::add):
2288         (JSC::Identifier::remove):
2289         (JSC::Identifier::checkSameIdentifierTable):
2290         (JSC::createIdentifierTableSpecificCallback):
2291         (JSC::createIdentifierTableSpecific):
2292         (JSC::createDefaultDataSpecific):
2293
2294             - Use currentIdentifierTable() instead of UStringImpl::m_identifierTable.
2295             - Define methods to access the thread specific identifier tables.
2296
2297         * runtime/Identifier.h:
2298         (JSC::ThreadIdentifierTableData::ThreadIdentifierTableData):
2299         (JSC::defaultIdentifierTable):
2300         (JSC::setDefaultIdentifierTable):
2301         (JSC::currentIdentifierTable):
2302         (JSC::setCurrentIdentifierTable):
2303         (JSC::resetCurrentIdentifierTable):
2304
2305             - Declare methods to access the thread specific identifier tables.
2306
2307         * runtime/JSGlobalData.cpp:
2308         (JSC::JSGlobalData::createNonDefault):
2309         (JSC::JSGlobalData::create):
2310         (JSC::JSGlobalData::sharedInstance):
2311
2312             - creation of JSGlobalData objects, other than for API usage, associate themselves with the current thread.
2313
2314         * runtime/JSGlobalData.h:
2315         * runtime/UStringImpl.cpp:
2316         (JSC::UStringImpl::destroy):
2317
2318             - destroy() method should be using isIdentifier().
2319
2320         * runtime/UStringImpl.h:
2321         (JSC::UStringImpl::isIdentifier):
2322         (JSC::UStringImpl::setIsIdentifier):
2323         (JSC::UStringImpl::checkConsistency):
2324         (JSC::UStringImpl::UStringImpl):
2325
2326             - replace m_identifierTable with a single m_isIdentifier bit.
2327
2328         * wtf/StringHashFunctions.h:
2329         (WTF::stringHash):
2330
2331             - change string hash result from 32-bit to 31-bit, to free a bit in UStringImpl for m_isIdentifier.
2332
2333 2009-12-25 Patrick Gansterer <paroga@paroga.com>
2334
2335         Reviewed by Eric Seidel.
2336
2337         Buildfix for WinCE + style fixes.
2338         https://bugs.webkit.org/show_bug.cgi?id=32939
2339
2340         * jsc.cpp:
2341         (functionPrint):
2342         (functionQuit):
2343         (parseArguments):
2344         (fillBufferWithContentsOfFile):
2345
2346 2010-01-05  Patrick Gansterer  <paroga@paroga.com>
2347
2348         Reviewed by Eric Seidel.
2349
2350         WinCE buildfix after r52791 (renamed PLATFORM(WINCE) to OS(WINCE)).
2351         https://bugs.webkit.org/show_bug.cgi?id=33205
2352
2353         * jit/ExecutableAllocator.h:
2354
2355 2010-01-05  Patrick Gansterer  <paroga@paroga.com>
2356
2357         Reviewed by Darin Adler.
2358
2359         Added compiler error for unsupported platforms.
2360         https://bugs.webkit.org/show_bug.cgi?id=33112
2361
2362         * jit/JITStubs.cpp:
2363
2364 2010-01-05  Gabor Loki  <loki@webkit.org>
2365
2366         Reviewed by Maciej Stachowiak.
2367
2368         Follow r52729 in ARMAssembler.
2369         https://bugs.webkit.org/show_bug.cgi?id=33208
2370
2371         Use WTF_ARM_ARCH_AT_LEAST instead of ARM_ARCH_VERSION
2372
2373         * assembler/ARMAssembler.cpp:
2374         (JSC::ARMAssembler::encodeComplexImm): Move tmp declaration to ARMv7
2375         * assembler/ARMAssembler.h:
2376         (JSC::ARMAssembler::):
2377         (JSC::ARMAssembler::bkpt):
2378
2379 2010-01-05  Maciej Stachowiak  <mjs@apple.com>
2380
2381         Unreviewed build fix for Gtk+
2382
2383         Don't use // comments in Platform.h, at least some of them seem to make the version of GCC
2384         used on the Gtk buildbot unhappy.
2385
2386         * wtf/Platform.h:
2387
2388 2010-01-04  Maciej Stachowiak  <mjs@apple.com>
2389
2390         Reviewed by Darin Fisher.
2391
2392         Reorganize, document and rename OS() platform macros.
2393         https://bugs.webkit.org/show_bug.cgi?id=33198
2394
2395         * wtf/Platform.h: Rename, reorganize and document OS() macros.
2396
2397         Adapt to name changes. Also fixed a few incorrect OS checks.
2398
2399         * API/JSContextRef.cpp:
2400         * assembler/MacroAssemblerARM.cpp:
2401         (JSC::isVFPPresent):
2402         * assembler/MacroAssemblerX86Common.h:
2403         * bytecode/SamplingTool.cpp:
2404         * config.h:
2405         * interpreter/RegisterFile.cpp:
2406         (JSC::RegisterFile::~RegisterFile):
2407         * interpreter/RegisterFile.h:
2408         (JSC::RegisterFile::RegisterFile):
2409         (JSC::RegisterFile::grow):
2410         * jit/ExecutableAllocator.h:
2411         * jit/ExecutableAllocatorFixedVMPool.cpp:
2412         * jit/ExecutableAllocatorPosix.cpp:
2413         * jit/ExecutableAllocatorSymbian.cpp:
2414         * jit/ExecutableAllocatorWin.cpp:
2415         * jit/JITOpcodes.cpp:
2416         (JSC::JIT::privateCompileCTIMachineTrampolines):
2417         * jit/JITStubs.cpp:
2418         * jsc.cpp:
2419         (main):
2420         * parser/Grammar.y:
2421         * profiler/ProfileNode.cpp:
2422         (JSC::getCount):
2423         * runtime/Collector.cpp:
2424         (JSC::Heap::Heap):
2425         (JSC::Heap::allocateBlock):
2426         (JSC::Heap::freeBlockPtr):
2427         (JSC::currentThreadStackBase):
2428         (JSC::getCurrentPlatformThread):
2429         (JSC::suspendThread):
2430         (JSC::resumeThread):
2431         (JSC::getPlatformThreadRegisters):
2432         (JSC::otherThreadStackPointer):
2433         * runtime/Collector.h:
2434         * runtime/DateConstructor.cpp:
2435         * runtime/DatePrototype.cpp:
2436         (JSC::formatLocaleDate):
2437         * runtime/InitializeThreading.cpp:
2438         (JSC::initializeThreading):
2439         * runtime/MarkStack.h:
2440         (JSC::MarkStack::MarkStackArray::shrinkAllocation):
2441         * runtime/MarkStackPosix.cpp:
2442         * runtime/MarkStackSymbian.cpp:
2443         * runtime/MarkStackWin.cpp:
2444         * runtime/StringPrototype.cpp:
2445         (JSC::stringProtoFuncLastIndexOf):
2446         * runtime/TimeoutChecker.cpp:
2447         (JSC::getCPUTime):
2448         * runtime/UString.cpp:
2449         (JSC::UString::from):
2450         * wtf/Assertions.cpp:
2451         * wtf/Assertions.h:
2452         * wtf/CurrentTime.cpp:
2453         (WTF::lowResUTCTime):
2454         * wtf/CurrentTime.h:
2455         (WTF::getLocalTime):
2456         * wtf/DateMath.cpp:
2457         * wtf/FastMalloc.cpp:
2458         (WTF::TCMalloc_ThreadCache::InitModule):
2459         (WTF::TCMallocStats::):
2460         * wtf/FastMalloc.h:
2461         * wtf/MathExtras.h:
2462         * wtf/RandomNumber.cpp:
2463         (WTF::randomNumber):
2464         * wtf/RandomNumberSeed.h:
2465         (WTF::initializeRandomNumberGenerator):
2466         * wtf/StringExtras.h:
2467         * wtf/TCSpinLock.h:
2468         (TCMalloc_SpinLock::Unlock):
2469         (TCMalloc_SlowLock):
2470         * wtf/TCSystemAlloc.cpp:
2471         * wtf/ThreadSpecific.h:
2472         (WTF::::destroy):
2473         * wtf/Threading.h:
2474         * wtf/ThreadingPthreads.cpp:
2475         (WTF::initializeThreading):
2476         (WTF::isMainThread):
2477         * wtf/ThreadingWin.cpp:
2478         (WTF::wtfThreadEntryPoint):
2479         (WTF::createThreadInternal):
2480         * wtf/VMTags.h:
2481         * wtf/unicode/icu/CollatorICU.cpp:
2482         (WTF::Collator::userDefault):
2483         * wtf/win/MainThreadWin.cpp:
2484         (WTF::initializeMainThreadPlatform):
2485
2486 2010-01-04  Gustavo Noronha Silva  <gns@gnome.org>
2487
2488         Add missing files to the build system - make distcheck build fix.
2489
2490         * GNUmakefile.am:
2491
2492 2010-01-04  Gavin Barraclough <barraclough@apple.com>
2493
2494         Reviewed by Sam Weinig, additional coding by Mark Rowe.
2495
2496         https://bugs.webkit.org/show_bug.cgi?id=33163
2497         Add string hashing functions to WTF.
2498         Use WTF's string hashing functions from UStringImpl.
2499
2500         * GNUmakefile.am:
2501         * JavaScriptCore.exp:
2502         * JavaScriptCore.gypi:
2503         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
2504         * JavaScriptCore.xcodeproj/project.pbxproj:
2505         * runtime/UStringImpl.cpp:
2506         * runtime/UStringImpl.h:
2507         (JSC::UStringImpl::computeHash):
2508         * wtf/HashFunctions.h:
2509         * wtf/StringHashFunctions.h: Added.
2510         (WTF::stringHash):
2511
2512 2010-01-04  Dmitry Titov  <dimich@chromium.org>
2513
2514         Not reviewed, attempt to fix ARM bulid.
2515
2516         * wtf/Platform.h:
2517
2518 2010-01-04  Gavin Barraclough  <barraclough@apple.com>
2519
2520         Rubber stamped by Geoff Garen.
2521
2522         Add an 'isIdentifier' to UStringImpl, use this where appropriate
2523         (where previously 'identifierTable' was being tested).
2524
2525         * API/JSClassRef.cpp:
2526         (OpaqueJSClass::~OpaqueJSClass):
2527         (OpaqueJSClassContextData::OpaqueJSClassContextData):
2528         * runtime/Identifier.cpp:
2529         (JSC::Identifier::addSlowCase):
2530         * runtime/Identifier.h:
2531         (JSC::Identifier::add):
2532         * runtime/PropertyNameArray.cpp:
2533         (JSC::PropertyNameArray::add):
2534         * runtime/UStringImpl.h:
2535         (JSC::UStringImpl::isIdentifier):
2536
2537 2010-01-04  Gavin Barraclough  <barraclough@apple.com>
2538
2539         Reviewed by Sam "Shimmey Shimmey" Weinig.
2540
2541         https://bugs.webkit.org/show_bug.cgi?id=33158
2542         Refactor JSC API entry/exit to use RAII instead of copy/pasting code.
2543         Make it easier to change set of actions taken when passing across the API boundary.
2544
2545         * API/APIShims.h: Added.
2546         (JSC::APIEntryShimWithoutLock::APIEntryShimWithoutLock):
2547         (JSC::APIEntryShimWithoutLock::~APIEntryShimWithoutLock):
2548         (JSC::APIEntryShim::APIEntryShim):
2549         (JSC::APICallbackShim::APICallbackShim):
2550         (JSC::APICallbackShim::~APICallbackShim):
2551         * API/JSBase.cpp:
2552         (JSEvaluateScript):
2553         (JSCheckScriptSyntax):
2554         (JSGarbageCollect):
2555         (JSReportExtraMemoryCost):
2556         * API/JSCallbackConstructor.cpp:
2557         (JSC::constructJSCallback):
2558         * API/JSCallbackFunction.cpp:
2559         (JSC::JSCallbackFunction::call):
2560         * API/JSCallbackObjectFunctions.h:
2561         (JSC::::init):
2562         (JSC::::getOwnPropertySlot):
2563         (JSC::::put):
2564         (JSC::::deleteProperty):
2565         (JSC::::construct):
2566         (JSC::::hasInstance):
2567         (JSC::::call):
2568         (JSC::::getOwnPropertyNames):
2569         (JSC::::toNumber):
2570         (JSC::::toString):
2571         (JSC::::staticValueGetter):
2572         (JSC::::callbackGetter):
2573         * API/JSContextRef.cpp:
2574         * API/JSObjectRef.cpp:
2575         (JSObjectMake):
2576         (JSObjectMakeFunctionWithCallback):
2577         (JSObjectMakeConstructor):
2578         (JSObjectMakeFunction):
2579         (JSObjectMakeArray):
2580         (JSObjectMakeDate):
2581         (JSObjectMakeError):
2582         (JSObjectMakeRegExp):
2583         (JSObjectGetPrototype):
2584         (JSObjectSetPrototype):
2585         (JSObjectHasProperty):
2586         (JSObjectGetProperty):
2587         (JSObjectSetProperty):
2588         (JSObjectGetPropertyAtIndex):
2589         (JSObjectSetPropertyAtIndex):
2590         (JSObjectDeleteProperty):
2591         (JSObjectCallAsFunction):
2592         (JSObjectCallAsConstructor):
2593         (JSObjectCopyPropertyNames):
2594         (JSPropertyNameArrayRelease):
2595         (JSPropertyNameAccumulatorAddName):
2596         * API/JSValueRef.cpp:
2597         (JSValueGetType):
2598         (JSValueIsUndefined):
2599         (JSValueIsNull):
2600         (JSValueIsBoolean):
2601         (JSValueIsNumber):
2602         (JSValueIsString):
2603         (JSValueIsObject):
2604         (JSValueIsObjectOfClass):
2605         (JSValueIsEqual):
2606         (JSValueIsStrictEqual):
2607         (JSValueIsInstanceOfConstructor):
2608         (JSValueMakeUndefined):
2609         (JSValueMakeNull):
2610         (JSValueMakeBoolean):
2611         (JSValueMakeNumber):
2612         (JSValueMakeString):
2613         (JSValueToBoolean):
2614         (JSValueToNumber):
2615         (JSValueToStringCopy):
2616         (JSValueToObject):
2617         (JSValueProtect):
2618         (JSValueUnprotect):
2619         * JavaScriptCore.xcodeproj/project.pbxproj:
2620
2621 2010-01-04  Dan Bernstein  <mitz@apple.com>
2622
2623         Reviewed by Ada Chan and Mark Rowe.
2624
2625         Updated copyright string
2626
2627         * Info.plist:
2628         * JavaScriptCore.vcproj/JavaScriptCore.resources/Info.plist:
2629         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.rc:
2630
2631 2010-01-04  Adam Roben  <aroben@apple.com>
2632
2633         No review, rolling out r52741.
2634         http://trac.webkit.org/changeset/52741
2635         https://bugs.webkit.org/show_bug.cgi?id=33056
2636
2637         * wtf/AlwaysInline.h:
2638
2639 2010-01-04  Patrick Gansterer  <paroga@paroga.com>
2640
2641         Reviewed by Darin Adler.
2642
2643         Add cacheFlush support for WinCE
2644         https://bugs.webkit.org/show_bug.cgi?id=33110
2645
2646         * jit/ExecutableAllocator.h:
2647         (JSC::ExecutableAllocator::cacheFlush):
2648
2649 2010-01-04  Patrick Gansterer  <paroga@paroga.com>
2650
2651         Reviewed by Adam Roben.
2652
2653         Implement NO_RETURN for COMPILER(MSVC).
2654         https://bugs.webkit.org/show_bug.cgi?id=33056
2655
2656         * wtf/AlwaysInline.h:
2657
2658 2010-01-04  Maciej Stachowiak  <mjs@apple.com>
2659
2660         Reviewed by Simon Hausmann.
2661
2662         Fix some PLATFORM(*_ENDIAN) uses to CPU()
2663         https://bugs.webkit.org/show_bug.cgi?id=33148
2664
2665         * runtime/JSCell.cpp:
2666         (JSC::):
2667         * runtime/JSValue.h:
2668         (JSC::JSValue::):
2669
2670 2010-01-04  Maciej Stachowiak  <mjs@apple.com>
2671
2672         Reviewed by Adam Barth.
2673
2674         Document CPU() macros in comments.
2675         https://bugs.webkit.org/show_bug.cgi?id=33147
2676
2677         * wtf/Platform.h:
2678
2679 2010-01-04  Maciej Stachowiak  <mjs@apple.com>
2680
2681         Reviewed by Adam Barth.
2682
2683         Reorganize, document and rename CPU() platform macros.
2684         https://bugs.webkit.org/show_bug.cgi?id=33145
2685         ExecutableAllocatorSymbian appears to have buggy ARM version check
2686         https://bugs.webkit.org/show_bug.cgi?id=33138
2687
2688         * wtf/Platform.h:
2689         Rename all macros related to detection of particular CPUs or
2690         classes of CPUs to CPU(), reorganize and document them.
2691
2692         All remaining changes are adapting to the renames, plus fixing the
2693         second bug cited above.
2694
2695         * assembler/ARMAssembler.cpp:
2696         * assembler/ARMAssembler.h:
2697         * assembler/ARMv7Assembler.h:
2698         * assembler/AbstractMacroAssembler.h:
2699         (JSC::AbstractMacroAssembler::Imm32::Imm32):
2700         * assembler/MacroAssembler.h:
2701         * assembler/MacroAssemblerARM.cpp:
2702         * assembler/MacroAssemblerARM.h:
2703         * assembler/MacroAssemblerCodeRef.h:
2704         (JSC::MacroAssemblerCodePtr::MacroAssemblerCodePtr):
2705         * assembler/MacroAssemblerX86.h:
2706         * assembler/MacroAssemblerX86Common.h:
2707         * assembler/MacroAssemblerX86_64.h:
2708         * assembler/X86Assembler.h:
2709         (JSC::X86Registers::):
2710         (JSC::X86Assembler::):
2711         (JSC::X86Assembler::movl_mEAX):
2712         (JSC::X86Assembler::movl_EAXm):
2713         (JSC::X86Assembler::repatchLoadPtrToLEA):
2714         (JSC::X86Assembler::X86InstructionFormatter::memoryModRM):
2715         * jit/ExecutableAllocator.h:
2716         * jit/ExecutableAllocatorFixedVMPool.cpp:
2717         * jit/ExecutableAllocatorPosix.cpp:
2718         * jit/ExecutableAllocatorSymbian.cpp:
2719         (JSC::ExecutableAllocator::intializePageSize):
2720         * jit/JIT.cpp:
2721         * jit/JIT.h:
2722         * jit/JITArithmetic.cpp:
2723         * jit/JITInlineMethods.h:
2724         (JSC::JIT::beginUninterruptedSequence):
2725         (JSC::JIT::restoreArgumentReferenceForTrampoline):
2726         (JSC::JIT::emitCount):
2727         * jit/JITOpcodes.cpp:
2728         (JSC::JIT::privateCompileCTIMachineTrampolines):
2729         * jit/JITPropertyAccess.cpp:
2730         (JSC::JIT::privateCompileGetByIdProto):
2731         (JSC::JIT::privateCompileGetByIdProtoList):
2732         (JSC::JIT::privateCompileGetByIdChainList):
2733         (JSC::JIT::privateCompileGetByIdChain):
2734         * jit/JITStubs.cpp:
2735         (JSC::JITThunks::JITThunks):
2736         * jit/JITStubs.h:
2737         * runtime/Collector.cpp:
2738         (JSC::currentThreadStackBase):
2739         (JSC::getPlatformThreadRegisters):
2740         (JSC::otherThreadStackPointer):
2741         * wrec/WREC.h:
2742         * wrec/WRECGenerator.cpp:
2743         (JSC::WREC::Generator::generateEnter):
2744         (JSC::WREC::Generator::generateReturnSuccess):
2745         (JSC::WREC::Generator::generateReturnFailure):
2746         * wrec/WRECGenerator.h:
2747         * wtf/FastMalloc.cpp:
2748         * wtf/TCSpinLock.h:
2749         (TCMalloc_SpinLock::Lock):
2750         (TCMalloc_SpinLock::Unlock):
2751         (TCMalloc_SlowLock):
2752         * wtf/Threading.h:
2753         * wtf/dtoa.cpp:
2754         * yarr/RegexJIT.cpp:
2755         (JSC::Yarr::RegexGenerator::generateEnter):
2756         (JSC::Yarr::RegexGenerator::generateReturn):
2757         * yarr/RegexJIT.h:
2758
2759 2010-01-04  Maciej Stachowiak  <mjs@apple.com>
2760
2761         Reviewed by Adam Barth.
2762
2763         Clean up COMPILER macros and remove unused ones.
2764         https://bugs.webkit.org/show_bug.cgi?id=33132
2765
2766         Removed values are COMPILER(BORLAND) and COMPILER(CYGWIN) - they were
2767         not used anywhere.
2768
2769         * wtf/Platform.h:
2770
2771 2010-01-03  Maciej Stachowiak  <mjs@apple.com>
2772
2773         Reviewed by Eric Seidel.
2774
2775         Update wtf/Platform.h to document the new system for porting macros.
2776         https://bugs.webkit.org/show_bug.cgi?id=33130
2777
2778         * wtf/Platform.h:
2779
2780 2009-12-29  Laszlo Gombos  <laszlo.1.gombos@nokia.com>
2781
2782         Reviewed by Maciej Stachowiak.
2783
2784         PLATFORM(CAIRO) should be defined by WIN_CAIRO define
2785         https://bugs.webkit.org/show_bug.cgi?id=22250
2786
2787         * wtf/Platform.h: Define WTF_PLATFORM_CAIRO for GTK port only
2788         For the WinCairo port WTF_PLATFORM_CAIRO is already defined in config.h
2789
2790 2009-12-28  Shu Chang  <Chang.Shu@nokia.com>
2791
2792         Reviewed by Laszlo Gombos.
2793
2794         [Qt] Delete ThreadPrivate instance after it is finished.
2795         https://bugs.webkit.org/show_bug.cgi?id=32614
2796
2797         * wtf/qt/ThreadingQt.cpp:
2798         (WTF::ThreadMonitor::instance):
2799         (WTF::ThreadMonitor::threadFinished):
2800         (WTF::createThreadInternal):
2801         (WTF::detachThread):
2802
2803 2009-12-28  Patrick Gansterer  <paroga@paroga.com>
2804
2805         Reviewed by Maciej Stachowiak.
2806
2807         Cleanup of #define JS_EXPORT.
2808
2809         * API/JSBase.h:
2810
2811 2009-12-27  Patrick Gansterer  <paroga@paroga.com>
2812
2813         Reviewed by Adam Barth.
2814
2815         WinCE buildfix (HWND_MESSAGE isn't supported there)
2816
2817         * wtf/win/MainThreadWin.cpp:
2818         (WTF::initializeMainThreadPlatform):
2819
2820 2009-12-27  Patrick Gansterer  <paroga@paroga.com>
2821
2822         Reviewed by Adam Barth.
2823
2824         Added a file with WinMain function to link agains in WinCE.
2825
2826         * os-win32/WinMain.cpp: Added.
2827         (convertToUtf8):
2828         (WinMain):
2829
2830 2009-12-24  Laszlo Gombos  <laszlo.1.gombos@nokia.com>
2831
2832         Unreviewed; revert of r52550.
2833
2834         The change regressed the following LayoutTests for QtWebKit.
2835
2836         fast/workers/worker-call.html -> crashed
2837         fast/workers/worker-close.html -> crashed
2838
2839         * wtf/qt/ThreadingQt.cpp:
2840         (WTF::waitForThreadCompletion):
2841         (WTF::detachThread):
2842
2843 2009-12-24  Shu Chang  <Chang.Shu@nokia.com>
2844
2845         Reviewed by Laszlo Gombos.
2846
2847         [Qt] Fix memory leak by deleting instance of ThreadPrivate
2848         in function waitForThreadCompletion(), synchronously, or in
2849         detachThread(), asynchronously.
2850         https://bugs.webkit.org/show_bug.cgi?id=32614
2851
2852         * wtf/qt/ThreadingQt.cpp:
2853         (WTF::waitForThreadCompletion):
2854         (WTF::detachThread):
2855
2856 2009-12-23  Kwang Yul Seo  <skyul@company100.net>
2857
2858         Reviewed by Laszlo Gombos.
2859
2860         Include stddef.h for ptrdiff_t
2861         https://bugs.webkit.org/show_bug.cgi?id=32891
2862
2863         ptrdiff_t is typedef-ed in stddef.h.
2864         Include stddef.h in jit/ExecutableAllocator.h.
2865
2866         * jit/ExecutableAllocator.h:
2867
2868 2009-12-23  Patrick Gansterer  <paroga@paroga.com>
2869
2870         Reviewed by Eric Seidel.
2871
2872         Buildfix after r47092.
2873
2874         * wtf/wince/MemoryManager.cpp:
2875         (WTF::tryFastMalloc):
2876         (WTF::tryFastZeroedMalloc):
2877         (WTF::tryFastCalloc):
2878         (WTF::tryFastRealloc):
2879
2880 2009-12-23  Kent Tamura  <tkent@chromium.org>
2881
2882         Reviewed by Darin Adler.
2883
2884         HTMLInputElement::valueAsDate getter support.
2885         https://bugs.webkit.org/show_bug.cgi?id=32876
2886
2887         Expose dateToDaysFrom1970().
2888
2889         * JavaScriptCore.exp:
2890         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
2891         * wtf/DateMath.cpp:
2892         (WTF::dateToDaysFrom1970):
2893         * wtf/DateMath.h:
2894
2895 2009-12-22  Darin Adler  <darin@apple.com>
2896
2897         Reviewed by Mark Rowe.
2898
2899         Turn off datagrid by default, at least for all platforms Apple ships.
2900         The datagrid implementation isn't ready for general web use yet.
2901
2902         * Configurations/FeatureDefines.xcconfig: Turn off datagrid by default.
2903
2904 2009-12-22  Steve Block  <steveblock@google.com>
2905
2906         Reviewed by David Levin.
2907
2908         Updates Android's scheduleDispatchFunctionsOnMainThread() to use new
2909         AndroidThreading class, rather than using JavaSharedClient directly.
2910         This fixes the current layering violation.
2911         https://bugs.webkit.org/show_bug.cgi?id=32651
2912
2913         The pattern is copied from Chromium, which uses the ChromiumThreading
2914         class. This patch also fixes the style in ChromiumThreading.h.
2915
2916         * wtf/android/AndroidThreading.h: Added. Declares AndroidThreading.
2917         * wtf/android/MainThreadAndroid.cpp: Modified
2918         (WTF::scheduleDispatchFunctionsOnMainThread): Uses AndroidThreading.
2919         * wtf/chromium/ChromiumThreading.h: Modified. Fixes style.
2920
2921 2009-12-22  Gavin Barraclough  <barraclough@apple.com>
2922
2923         Reviewed by Sam Weinig.
2924
2925         Fix a couple of problems with UntypedPtrAndBitfield.
2926
2927         Add a m_leaksPtr to reduce false positives from leaks in debug builds
2928         (this isn't perfect because we'd like a solution for release builds,
2929         but this is now at least as good as a PtrAndFlags would be).
2930
2931         Switch SmallStringsto use a regular string for the base, rather than
2932         a static one.  UntypedPtrAndBitfield assumes all strings are at least
2933         8 byte aligned; this migt not be true of static strings.  Shared buffers
2934         are heap allocated, as are all UStringImpls other than static strings.
2935         Static strings cannot end up being the owner string of substrings,
2936         since the only static strings are length 0.
2937
2938         * runtime/SmallStrings.cpp:
2939         (JSC::SmallStringsStorage::SmallStringsStorage):
2940         * runtime/UStringImpl.h:
2941         (JSC::UntypedPtrAndBitfield::UntypedPtrAndBitfield):
2942         (JSC::UStringImpl::UStringImpl):
2943
2944 2009-12-22  Kwang Yul Seo  <skyul@company100.net>
2945
2946         Reviewed by Darin Adler.
2947
2948         RVCT (__ARMCC_VERSION < 400000) does not provide strcasecmp and strncasecmp
2949         https://bugs.webkit.org/show_bug.cgi?id=32857
2950
2951         Add implementation of strcasecmp and strncasecmp for RVCT < 4.0
2952         because earlier versions of RVCT 4.0 does not provide these functions.
2953
2954         * wtf/StringExtras.cpp: Added.
2955         (strcasecmp):
2956         (strncasecmp):
2957         * wtf/StringExtras.h:
2958
2959 2009-12-22  Kwang Yul Seo  <skyul@company100.net>
2960
2961         Reviewed by Darin Adler.
2962
2963         Define ALWAYS_INLINE and WTF_PRIVATE_INLINE to __forceinline for RVCT
2964         https://bugs.webkit.org/show_bug.cgi?id=32853
2965
2966         Use __forceinline forces RVCT to compile a C or C++ function
2967         inline. The compiler attempts to inline the function, regardless of
2968         the characteristics of the function.
2969
2970         * wtf/AlwaysInline.h:
2971         * wtf/FastMalloc.h:
2972
2973 2009-12-21  Simon Hausmann  <simon.hausmann@nokia.com>
2974
2975         Prospective GTK build fix: Add UStringImpl.cpp/h to the build.
2976
2977         * GNUmakefile.am:
2978
2979 2009-12-21  Simon Hausmann  <simon.hausmann@nokia.com>
2980
2981         Fix the Qt build, add UStringImpl.cpp to the build.
2982
2983         * JavaScriptCore.pri:
2984
2985 2009-12-21  Gavin Barraclough  <barraclough@apple.com>
2986
2987         Windows Build fix part 5.
2988         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
2989
2990 2009-12-21  Gavin Barraclough  <barraclough@apple.com>
2991
2992         Reviewed by NOBODY (build fix).
2993         Fix breakage of world introduced in build fix to r52463.
2994
2995         * runtime/UStringImpl.h:
2996
2997 2009-12-21  Gavin Barraclough  <barraclough@apple.com>
2998
2999         Reviewed by Darin Adler.
3000
3001         https://bugs.webkit.org/show_bug.cgi?id=32831
3002         Replace UString::Rep implementation, following introduction of ropes to JSC.
3003
3004             * Remove redundant overcapacity mechanisms.
3005             * Reduce memory cost of Rep's.
3006             * Add an inline storage mechanism akin to that in WebCore's StringImpl.
3007
3008         ~1% Sunspider progression.
3009
3010         * JavaScriptCore.exp:
3011         * JavaScriptCore.xcodeproj/project.pbxproj:
3012         * runtime/JSString.cpp:
3013         (JSC::JSString::resolveRope):
3014         * runtime/SmallStrings.cpp:
3015         (JSC::SmallStringsStorage::SmallStringsStorage):
3016         * runtime/UString.cpp:
3017         (JSC::initializeUString):
3018         (JSC::createRep):
3019         (JSC::UString::createFromUTF8):
3020         (JSC::UString::createUninitialized):
3021         (JSC::UString::spliceSubstringsWithSeparators):
3022         (JSC::UString::replaceRange):
3023         (JSC::UString::ascii):
3024         (JSC::UString::operator=):
3025         (JSC::UString::toStrictUInt32):
3026         (JSC::equal):
3027         * runtime/UString.h:
3028         (JSC::UString::isEmpty):
3029         (JSC::UString::cost):
3030         (JSC::makeString):
3031         * runtime/UStringImpl.cpp: Added.
3032         (JSC::UStringImpl::baseSharedBuffer):
3033         (JSC::UStringImpl::sharedBuffer):
3034         (JSC::UStringImpl::destroy):
3035         (JSC::UStringImpl::computeHash):
3036         * runtime/UStringImpl.h: Added.
3037         (JSC::UntypedPtrAndBitfield::UntypedPtrAndBitfield):
3038         (JSC::UntypedPtrAndBitfield::asPtr):
3039         (JSC::UntypedPtrAndBitfield::operator&=):
3040         (JSC::UntypedPtrAndBitfield::operator|=):
3041         (JSC::UntypedPtrAndBitfield::operator&):
3042         (JSC::UStringImpl::create):
3043         (JSC::UStringImpl::createCopying):
3044         (JSC::UStringImpl::createUninitialized):
3045         (JSC::UStringImpl::data):
3046         (JSC::UStringImpl::size):
3047         (JSC::UStringImpl::cost):
3048         (JSC::UStringImpl::hash):
3049         (JSC::UStringImpl::computedHash):
3050         (JSC::UStringImpl::setHash):
3051         (JSC::UStringImpl::identifierTable):
3052         (JSC::UStringImpl::setIdentifierTable):
3053         (JSC::UStringImpl::ref):
3054         (JSC::UStringImpl::deref):
3055         (JSC::UStringImpl::allocChars):
3056         (JSC::UStringImpl::copyChars):
3057         (JSC::UStringImpl::computeHash):
3058         (JSC::UStringImpl::null):
3059         (JSC::UStringImpl::empty):
3060         (JSC::UStringImpl::checkConsistency):
3061         (JSC::UStringImpl::):
3062         (JSC::UStringImpl::UStringImpl):
3063         (JSC::UStringImpl::operator new):
3064         (JSC::UStringImpl::bufferOwnerString):
3065         (JSC::UStringImpl::bufferOwnership):
3066         (JSC::UStringImpl::isStatic):
3067
3068 2009-12-18  Laszlo Gombos  <laszlo.1.gombos@nokia.com>
3069
3070         Reviewed by Kenneth Rohde Christiansen.
3071
3072         Move some build decisions from Qt build system into source files
3073         https://bugs.webkit.org/show_bug.cgi?id=31956
3074
3075         * JavaScriptCore.pri: Compile files unconditionally
3076         * jit/ExecutableAllocatorPosix.cpp: Guard with PLATFORM(UNIX) && !PLATFORM(SYMBIAN)
3077         * jit/ExecutableAllocatorWin.cpp: Guard with PLATFORM(WIN_OS)
3078         * runtime/MarkStackPosix.cpp: Guard with PLATFORM(UNIX) && !PLATFORM(SYMBIAN)
3079         * runtime/MarkStackSymbian.cpp: Guard with PLATFORM(SYMBIAN)
3080         * runtime/MarkStackWin.cpp: Guard with PLATFORM(WIN_OS)
3081         * wtf/Platform.h: Guard ENABLE_JSC_MULTIPLE_THREADS with ENABLE_SINGLE_THREADED for the Qt port
3082         * wtf/ThreadingNone.cpp: Guard with ENABLE(SINGLE_THREADED)
3083         * wtf/qt/ThreadingQt.cpp: Guard with !ENABLE(SINGLE_THREADED)
3084
3085 2009-12-18  Gavin Barraclough  <barraclough@apple.com>
3086
3087         Reviewed by Sam Weinig.
3088
3089         Add createNonCopying method to UString to make replace constructor passed bool,
3090         to make behaviour more explicit.  Add createFromUTF8 to UString (wrapping method
3091         on UString::Rep), since other cases of transliteration (e.g.  from ascii) are
3092         performed in UString constructors.  Add/use setHash & size() accessors on Rep,
3093         rather than accessing _hash/len directly.
3094
3095         * API/JSClassRef.cpp:
3096         (OpaqueJSClass::OpaqueJSClass):
3097         * API/OpaqueJSString.cpp:
3098         (OpaqueJSString::ustring):
3099         * JavaScriptCore.exp:
3100         * runtime/ArrayPrototype.cpp:
3101         (JSC::arrayProtoFuncToString):
3102         * runtime/Identifier.cpp:
3103         (JSC::Identifier::equal):
3104         (JSC::CStringTranslator::translate):
3105         (JSC::UCharBufferTranslator::translate):
3106         (JSC::Identifier::addSlowCase):
3107         * runtime/JSString.cpp:
3108         (JSC::JSString::resolveRope):
3109         * runtime/JSString.h:
3110         (JSC::JSString::Rope::Fiber::refAndGetLength):
3111         (JSC::JSString::Rope::append):
3112         * runtime/StringBuilder.h:
3113         (JSC::StringBuilder::release):
3114         * runtime/StringConstructor.cpp:
3115         (JSC::stringFromCharCodeSlowCase):
3116         * runtime/StringPrototype.cpp:
3117         (JSC::substituteBackreferencesSlow):
3118         (JSC::stringProtoFuncToLowerCase):
3119         (JSC::stringProtoFuncToUpperCase):
3120         (JSC::stringProtoFuncFontsize):
3121         (JSC::stringProtoFuncLink):
3122         * runtime/UString.cpp:
3123         (JSC::UString::UString):
3124         (JSC::UString::createNonCopying):
3125         (JSC::UString::createFromUTF8):
3126         * runtime/UString.h:
3127         (JSC::UString::Rep::setHash):
3128         (JSC::UString::~UString):
3129         (JSC::makeString):
3130
3131 2009-12-18  Geoffrey Garen  <ggaren@apple.com>
3132
3133         Reviewed by Cameron Zwarich and Gavin Barraclough.
3134
3135         Changed Register constructors to assignment operators, to streamline
3136         moving values into registers. (In theory, there's no difference between
3137         the two, since the constructor should just inline away, but there seems
3138         to be a big difference in the addled mind of the GCC optimizer.)
3139
3140         In the interpreter, this is a 3.5% SunSpider speedup and a 1K-2K
3141         reduction in stack usage per privateExecute stack frame.
3142
3143         * interpreter/CallFrame.h:
3144         (JSC::ExecState::setCalleeArguments):
3145         (JSC::ExecState::setCallerFrame):
3146         (JSC::ExecState::setScopeChain):
3147         (JSC::ExecState::init):
3148         (JSC::ExecState::setArgumentCount):
3149         (JSC::ExecState::setCallee):
3150         (JSC::ExecState::setCodeBlock): Added a little bit of casting so these
3151         functions could use the new Register assignment operators.
3152
3153         * interpreter/Register.h:
3154         (JSC::Register::withInt):
3155         (JSC::Register::Register):
3156         (JSC::Register::operator=): Swapped in assignment operators for constructors.
3157
3158 2009-12-18  Yongjun Zhang  <yongjun.zhang@nokia.com>
3159
3160         Reviewed by Simon Hausmann.
3161
3162         https://bugs.webkit.org/show_bug.cgi?id=32713
3163         [Qt] make wtf/Assertions.h compile in winscw compiler.
3164
3165         Add string arg before ellipsis to help winscw compiler resolve variadic
3166         macro definitions in wtf/Assertions.h.
3167
3168         * wtf/Assertions.h:
3169
3170 2009-12-18  Geoffrey Garen  <ggaren@apple.com>
3171
3172         Reviewed by Adam Roben.
3173
3174         Fixed intermittent failure seen on Windows buildbot, and in other JSC
3175         API clients.
3176
3177         Added a WeakGCPtr class and changed OpaqueJSClass::cachedPrototype to
3178         use it, to avoid vending a stale object as a prototype.
3179
3180         * API/JSClassRef.cpp:
3181         (OpaqueJSClassContextData::OpaqueJSClassContextData):
3182         (OpaqueJSClass::prototype):
3183         * API/JSClassRef.h: Use WeakGCPtr.
3184
3185         * JavaScriptCore.xcodeproj/project.pbxproj:
3186         * runtime/WeakGCPtr.h: Added.
3187         (JSC::WeakGCPtr::WeakGCPtr):
3188         (JSC::WeakGCPtr::get):
3189         (JSC::WeakGCPtr::clear):
3190         (JSC::WeakGCPtr::operator*):
3191         (JSC::WeakGCPtr::operator->):
3192         (JSC::WeakGCPtr::operator!):
3193         (JSC::WeakGCPtr::operator bool):
3194         (JSC::WeakGCPtr::operator UnspecifiedBoolType):
3195         (JSC::WeakGCPtr::assign):
3196         (JSC::::operator):
3197         (JSC::operator==):
3198         (JSC::operator!=):
3199         (JSC::static_pointer_cast):
3200         (JSC::const_pointer_cast):
3201         (JSC::getPtr): Added WeakGCPtr to the project.
3202
3203 2009-12-18  Gavin Barraclough  <barraclough@apple.com>
3204
3205         Reviewed by Sam Weinig.
3206
3207         https://bugs.webkit.org/show_bug.cgi?id=32720
3208
3209         * JavaScriptCore.exp:
3210             - Remove exports for UString::append
3211         * JavaScriptCore.xcodeproj/project.pbxproj:
3212             - Make StringBuilder a private header (was project).
3213
3214 2009-12-18  Martin Robinson  <martin.james.robinson@gmail.com>
3215
3216         Reviewed by Gustavo Noronha Silva.
3217
3218         [GTK] GRefPtr does not take a reference when assigned a raw pointer
3219         https://bugs.webkit.org/show_bug.cgi?id=32709
3220
3221         Ensure that when assigning a raw pointer to a GRefPtr, the reference
3222         count is incremented. Also remove the GRefPtr conversion overload as
3223         GRefPtr types have necessarily incompatible reference counting.
3224
3225         * wtf/gtk/GRefPtr.h:
3226         (WTF::GRefPtr::operator=):
3227
3228 2009-12-18  Simon Hausmann  <simon.hausmann@nokia.com>
3229
3230         Reviewed by Tor Arne Vestbø.
3231
3232         [Qt] Clean up the qmake build system to distinguish between trunk builds and package builds
3233
3234         https://bugs.webkit.org/show_bug.cgi?id=32716
3235
3236         * pcre/pcre.pri: Use standalone_package instead of QTDIR_build
3237
3238 2009-12-18  Martin Robinson  <martin.james.robinson@gmail.com>
3239
3240         Reviewed by Gustavo Noronha Silva.
3241
3242         [GTK] Compile warning from line 29 of GRefPtr.cpp
3243         https://bugs.webkit.org/show_bug.cgi?id=32703
3244
3245         Fix memory leak and compiler warning in GRefPtr GHashTable template
3246         specialization.
3247
3248         * wtf/gtk/GRefPtr.cpp:
3249         (WTF::refGPtr):
3250
3251 2009-12-17  Sam Weinig  <sam@webkit.org>
3252
3253         Reviewed by Mark Rowe.
3254
3255         Add BUILDING_ON_SNOW_LEOPARD and TARGETING_SNOW_LEOPARD #defines.
3256
3257         * wtf/Platform.h:
3258
3259 2009-12-17  Adam Roben  <aroben@apple.com>
3260
3261         Sync JavaScriptCore.vcproj with JavaScriptCore.xcodeproj and the
3262         source tree
3263
3264         Fixes <http://webkit.org/b/32665>.
3265
3266         Reviewed by Ada Chan.
3267
3268         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj: Moved
3269         around files and filters so that the structure matches
3270         JavaScriptCore.xcodeproj and the source tree. A few headers that were
3271         previously omitted have been added, as well as JSZombie.{cpp,h}.
3272
3273 2009-12-17  Adam Roben  <aroben@apple.com>
3274