ChangeLog 1.09 MB
Newer Older
1 2 3 4 5 6 7 8 9 10
2011-09-23  Filip Pizlo  <fpizlo@apple.com>

        ValueToDouble handling in prediction propagation should be ASSERT_NOT_REACHED
        https://bugs.webkit.org/show_bug.cgi?id=68724

        Reviewed by Oliver Hunt.

        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNodePredictions):

oliver@apple.com's avatar
oliver@apple.com committed
11 12 13 14 15 16
2011-09-23  Oliver Hunt  <oliver@apple.com>

        Build fix.

        * JavaScriptCore.xcodeproj/project.pbxproj:

17 18 19 20 21 22 23 24 25 26
2011-09-23  Filip Pizlo  <fpizlo@apple.com>

        DFG implementation of PutScopedVar corrupts register allocation
        https://bugs.webkit.org/show_bug.cgi?id=68735

        Reviewed by Oliver Hunt.

        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):

27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85
2011-09-23  Oliver Hunt  <oliver@apple.com>

        Make write barriers actually do something when enabled
        https://bugs.webkit.org/show_bug.cgi?id=68717

        Reviewed by Geoffrey Garen.

        Add a basic card marking style write barrier to JSC (currently
        turned off).  This requires two scratch registers in the JIT
        so there was some register re-arranging to satisfy that requirement.
        Happily this produced a minor perf bump in sunspider (~0.5%).

        Turning the barriers on causes an overall regression of around 1.5%

        * JavaScriptCore.exp:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * assembler/MacroAssemblerX86Common.h:
        (JSC::MacroAssemblerX86Common::store8):
        * assembler/X86Assembler.h:
        (JSC::X86Assembler::movb_i8m):
        * dfg/DFGJITCodeGenerator.cpp:
        (JSC::DFG::JITCodeGenerator::isKnownNotCell):
        (JSC::DFG::JITCodeGenerator::writeBarrier):
        (JSC::DFG::JITCodeGenerator::markCellCard):
        (JSC::DFG::JITCodeGenerator::cachedPutById):
        * dfg/DFGJITCodeGenerator.h:
        * dfg/DFGRepatch.cpp:
        (JSC::DFG::tryCachePutByID):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):
        * heap/CardSet.h: Added.
        (JSC::CardSet::CardSet):
        (JSC::::cardForAtom):
        (JSC::::cardMarkedForAtom):
        (JSC::::markCardForAtom):
        * heap/Heap.cpp:
        * heap/Heap.h:
        (JSC::Heap::addressOfCardFor):
        (JSC::Heap::writeBarrierFastCase):
        * heap/MarkedBlock.h:
        (JSC::MarkedBlock::setDirtyObject):
        (JSC::MarkedBlock::addressOfCardFor):
        (JSC::MarkedBlock::offsetOfCards):
        * jit/JIT.h:
        * jit/JITPropertyAccess.cpp:
        (JSC::JIT::emit_op_put_by_val):
        (JSC::JIT::emit_op_put_by_id):
        (JSC::JIT::privateCompilePutByIdTransition):
        (JSC::JIT::emit_op_put_scoped_var):
        (JSC::JIT::emit_op_put_global_var):
        (JSC::JIT::emitWriteBarrier):
        * jit/JITPropertyAccess32_64.cpp:
        (JSC::JIT::emit_op_put_by_val):
        (JSC::JIT::emit_op_put_by_id):
        (JSC::JIT::emitSlow_op_put_by_id):
        (JSC::JIT::privateCompilePutByIdTransition):
        (JSC::JIT::emit_op_put_scoped_var):
        (JSC::JIT::emit_op_put_global_var):

86 87 88 89 90 91 92 93 94 95 96 97
2011-09-23  Thouraya ANDOLSI  <thouraya.andolsi@st.com>

        https://bugs.webkit.org/show_bug.cgi?id=68077
        SH4 assemblers doesn't refer to executable memory handle.

        Reviewed by Gavin Barraclough.

        * assembler/MacroAssemblerSH4.h:
        (JSC::MacroAssemblerSH4::branch8):
        * assembler/SH4Assembler.h:
        (JSC::SH4Assembler::executableCopy):

98 99 100 101 102 103 104 105 106 107 108 109
2011-09-23  Oliver Hunt  <oliver@apple.com>

        PutScopedVar nodes should report that it has a var number
        https://bugs.webkit.org/show_bug.cgi?id=68721

        Reviewed by Anders Carlsson.

        Another assertion fix.

        * dfg/DFGNode.h:
        (JSC::DFG::Node::hasVarNumber):

110 111 112 113 114 115 116 117 118 119 120 121 122 123 124
2011-09-23  Oliver Hunt  <oliver@apple.com>

        Add a bunch of unhandled node types to the propagator
        https://bugs.webkit.org/show_bug.cgi?id=68716

        Reviewed by Darin Adler.

        Remove the ASSERT_NOT_REACHED() default for debug builds in the
        prediction propagator, this way unhandled nodes will just cause
        compile time failures rather than failing at some point in the
        future.

        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNodePredictions):

125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229
2011-09-23  Mark Hahnenberg  <mhahnenberg@apple.com>

        Add static version of JSCell::visitChildren
        https://bugs.webkit.org/show_bug.cgi?id=68404

        Reviewed by Darin Adler.

        In this patch we just extract the bodies of the virtual visitChildren methods
        throughout the JSCell inheritance hierarchy out into static methods, which are 
        now called from the virtual methods.  This is an intermediate step in trying to 
        move the virtual-ness of visitChildren into our own custom vtable stored in 
        ClassInfo.  We need to convert the methods to static methods in order to be 
        able to more easily store and refer to them in our custom vtable since normal 
        member methods store some implicit information in their types, making it 
        impossible to store them generically in ClassInfo.

        * API/JSCallbackObject.h:
        (JSC::JSCallbackObject::visitChildrenVirtual):
        (JSC::JSCallbackObject::visitChildren):
        * JavaScriptCore.exp:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
        * debugger/DebuggerActivation.cpp:
        (JSC::DebuggerActivation::visitChildrenVirtual):
        (JSC::DebuggerActivation::visitChildren):
        * debugger/DebuggerActivation.h:
        * heap/MarkStack.cpp:
        (JSC::SlotVisitor::visitChildren):
        (JSC::SlotVisitor::drain):
        * runtime/Arguments.cpp:
        (JSC::Arguments::visitChildrenVirtual):
        (JSC::Arguments::visitChildren):
        * runtime/Arguments.h:
        * runtime/Executable.cpp:
        (JSC::EvalExecutable::visitChildrenVirtual):
        (JSC::EvalExecutable::visitChildren):
        (JSC::ProgramExecutable::visitChildrenVirtual):
        (JSC::ProgramExecutable::visitChildren):
        (JSC::FunctionExecutable::visitChildrenVirtual):
        (JSC::FunctionExecutable::visitChildren):
        * runtime/Executable.h:
        * runtime/GetterSetter.cpp:
        (JSC::GetterSetter::visitChildrenVirtual):
        (JSC::GetterSetter::visitChildren):
        * runtime/GetterSetter.h:
        * runtime/JSActivation.cpp:
        (JSC::JSActivation::visitChildrenVirtual):
        (JSC::JSActivation::visitChildren):
        * runtime/JSActivation.h:
        * runtime/JSArray.cpp:
        (JSC::JSArray::visitChildrenVirtual):
        (JSC::JSArray::visitChildren):
        * runtime/JSArray.h:
        * runtime/JSBoundFunction.cpp:
        (JSC::JSBoundFunction::visitChildrenVirtual):
        (JSC::JSBoundFunction::visitChildren):
        * runtime/JSBoundFunction.h:
        * runtime/JSCell.h:
        (JSC::JSCell::visitChildrenVirtual):
        (JSC::JSCell::visitChildren):
        * runtime/JSFunction.cpp:
        (JSC::JSFunction::visitChildrenVirtual):
        (JSC::JSFunction::visitChildren):
        * runtime/JSFunction.h:
        * runtime/JSGlobalObject.cpp:
        (JSC::JSGlobalObject::visitChildrenVirtual):
        (JSC::JSGlobalObject::visitChildren):
        * runtime/JSGlobalObject.h:
        * runtime/JSObject.cpp:
        (JSC::JSObject::visitChildrenVirtual):
        (JSC::JSObject::visitChildren):
        * runtime/JSObject.h:
        (JSC::JSObject::visitChildrenDirect):
        * runtime/JSPropertyNameIterator.cpp:
        (JSC::JSPropertyNameIterator::visitChildrenVirtual):
        (JSC::JSPropertyNameIterator::visitChildren):
        * runtime/JSPropertyNameIterator.h:
        * runtime/JSStaticScopeObject.cpp:
        (JSC::JSStaticScopeObject::visitChildrenVirtual):
        (JSC::JSStaticScopeObject::visitChildren):
        * runtime/JSStaticScopeObject.h:
        * runtime/JSWrapperObject.cpp:
        (JSC::JSWrapperObject::visitChildrenVirtual):
        (JSC::JSWrapperObject::visitChildren):
        * runtime/JSWrapperObject.h:
        * runtime/NativeErrorConstructor.cpp:
        (JSC::NativeErrorConstructor::visitChildrenVirtual):
        (JSC::NativeErrorConstructor::visitChildren):
        * runtime/NativeErrorConstructor.h:
        * runtime/RegExpObject.cpp:
        (JSC::RegExpObject::visitChildrenVirtual):
        (JSC::RegExpObject::visitChildren):
        * runtime/RegExpObject.h:
        * runtime/ScopeChain.cpp:
        (JSC::ScopeChainNode::visitChildrenVirtual):
        (JSC::ScopeChainNode::visitChildren):
        * runtime/ScopeChain.h:
        * runtime/Structure.cpp:
        (JSC::Structure::visitChildrenVirtual):
        (JSC::Structure::visitChildren):
        * runtime/Structure.h:
        * runtime/StructureChain.cpp:
        (JSC::StructureChain::visitChildrenVirtual):
        (JSC::StructureChain::visitChildren):
        * runtime/StructureChain.h:

230 231 232 233 234 235 236 237 238 239 240 241
2011-09-23  Oliver Hunt  <oliver@apple.com>

        Node propagation doesn't handle PutScopedVar
        https://bugs.webkit.org/show_bug.cgi?id=68713

        Reviewed by Sam Weinig.

        This was causing assertion failures.

        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNodePredictions):

242 243 244 245 246 247
2011-09-23  Anders Carlsson  <andersca@apple.com>

        Make sure to define OVERRIDE and FINAL for older builds of clang.

        * wtf/Compiler.h:

248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267
2011-09-23  Gavin Barraclough  <barraclough@apple.com>

        Implement op_resolve_global in the DFG JIT
        https://bugs.webkit.org/show_bug.cgi?id=68704

        Reviewed by Oliver Hunt.

        This is performance neutral, but increases coverage.

        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::ByteCodeParser):
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGNode.h:
        (JSC::DFG::Node::hasIdentifier):
        (JSC::DFG::Node::resolveInfoIndex):
        * dfg/DFGOperations.cpp:
        * dfg/DFGOperations.h:
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):

268 269 270 271 272 273
2011-09-23  Mark Rowe  <mrowe@apple.com>

        Define BUILDING_ON_LION / TARGETING_LION when appropriate in Platform.h.

        * wtf/Platform.h:

274 275 276 277 278 279 280 281 282 283 284
2011-09-22  Anders Carlsson  <andersca@apple.com>

        We should add support for OVERRIDE and FINAL annotations
        https://bugs.webkit.org/show_bug.cgi?id=68654

        Reviewed by David Hyatt.

        Add OVERRIDE and FINAL macros for compilers that support them.

        * wtf/Compiler.h:

285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307
2011-09-22  Filip Pizlo  <fpizlo@apple.com>

        GetScopedVar should have value profiling
        https://bugs.webkit.org/show_bug.cgi?id=68676

        Reviewed by Oliver Hunt.
        
        Added GetScopedVar value profiling and predictin propagation.
        Added GetScopeChain to CSE.

        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGGraph.h:
        (JSC::DFG::Graph::predict):
        * dfg/DFGNode.h:
        (JSC::DFG::Node::hasPrediction):
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNodePredictions):
        (JSC::DFG::Propagator::getScopeChainLoadElimination):
        (JSC::DFG::Propagator::performNodeCSE):
        * jit/JITPropertyAccess.cpp:
        (JSC::JIT::emit_op_get_scoped_var):

fpizlo@apple.com's avatar
fpizlo@apple.com committed
308 309 310 311 312 313 314
2011-09-22  Filip Pizlo  <fpizlo@apple.com>

        PPC build fix, part 3.

        * runtime/Executable.cpp:
        (JSC::FunctionExecutable::compileForConstructInternal):

fpizlo@apple.com's avatar
fpizlo@apple.com committed
315 316 317 318 319 320 321
2011-09-22  Filip Pizlo  <fpizlo@apple.com>

        Another PPC build fix.

        * runtime/Executable.cpp:
        * runtime/Executable.h:

dino@apple.com's avatar
dino@apple.com committed
322 323 324 325 326 327 328 329 330
2011-09-22  Dean Jackson  <dino@apple.com>

        Add ENABLE_CSS_FILTERS
        https://bugs.webkit.org/show_bug.cgi?id=68652

        Reviewed by Simon Fraser.

        * Configurations/FeatureDefines.xcconfig:

331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353
2011-09-22  Gavin Barraclough  <barraclough@apple.com>

        Incorrect this value passed to callbacks.
        https://bugs.webkit.org/show_bug.cgi?id=68668

        Reviewed by Oliver Hunt.

        From Array/String prototype function.  Should be undefined, but
        global object is passed instead (this is visible for strict callbacks).

        * runtime/ArrayPrototype.cpp:
        (JSC::arrayProtoFuncSort):
        (JSC::arrayProtoFuncFilter):
        (JSC::arrayProtoFuncMap):
        (JSC::arrayProtoFuncEvery):
        (JSC::arrayProtoFuncForEach):
        (JSC::arrayProtoFuncSome):
        * runtime/JSArray.cpp:
        (JSC::AVLTreeAbstractorForArrayCompare::compare_key_key):
        (JSC::JSArray::sort):
        * runtime/StringPrototype.cpp:
        (JSC::stringProtoFuncReplace):

354 355 356 357 358 359 360 361 362
2011-09-22  Gavin Barraclough  <barraclough@apple.com>

        Function.prototype.bind.length shoudl be 1.

        Rubber stamped by Olier Hunt.

        * runtime/FunctionPrototype.cpp:
        (JSC::FunctionPrototype::addFunctionProperties):

fpizlo@apple.com's avatar
fpizlo@apple.com committed
363 364 365 366 367 368
2011-09-22  Filip Pizlo  <fpizlo@apple.com>

        PPC build fix.

        * bytecode/CodeBlock.h:

369 370 371 372 373 374
2011-09-22  Gavin Barraclough  <barraclough@apple.com>

        Windows build fix pt. 2

        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:

375 376 377 378 379 380
2011-09-22  Gavin Barraclough  <barraclough@apple.com>

        Windows build fix pt. 1

        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:

381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430
2011-09-21  Filip Pizlo  <fpizlo@apple.com>

        DFG JIT does not support to_primitive or strcat
        https://bugs.webkit.org/show_bug.cgi?id=68582

        Reviewed by Darin Adler.
        
        This adds functional support for to_primitive and strcat. It focuses
        on minimizing the amount of code emitted on to_primitive (if we know
        that it is a primitive or can speculate cheaply, then we omit the
        slow path) and on keeping the implementation of strcat simple while
        leveraging whatever optimizations we have already. In particular,
        unlike the Call and Construct nodes which require extending the size
        of the DFG's callee registers, StrCat takes advantage of the fact
        that no JS code can run while StrCat is in progress and uses a
        scratch buffer, rather than the register file, to store the list of
        values to concatenate. This was done mainly to keep the code simple,
        but there are probably other benefits to keeping call frame sizes
        down. Essentially, this patch ensures that the presence of an
        op_strcat does not mess up any other optimizations we might do while
        ensuring that if you do execute it, it'll work about as well as you'd
        expect.
        
        When combined with the previous patch for integer division, this is a
        14% speed-up on Kraken. Without it, it would have been a 2% loss.

        * assembler/AbstractMacroAssembler.h:
        (JSC::AbstractMacroAssembler::TrustedImmPtr::TrustedImmPtr):
        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGCapabilities.h:
        (JSC::DFG::canCompileOpcode):
        * dfg/DFGJITCodeGenerator.h:
        (JSC::DFG::JITCodeGenerator::callOperation):
        * dfg/DFGJITCompiler.cpp:
        (JSC::DFG::JITCompiler::exitSpeculativeWithOSR):
        * dfg/DFGNode.h:
        * dfg/DFGOperations.cpp:
        * dfg/DFGOperations.h:
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNodePredictions):
        (JSC::DFG::Propagator::performNodeCSE):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):
        * runtime/JSGlobalData.cpp:
        (JSC::JSGlobalData::JSGlobalData):
        (JSC::JSGlobalData::~JSGlobalData):
        * runtime/JSGlobalData.h:
        (JSC::JSGlobalData::scratchBufferForSize):

431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457
2011-09-22  Filip Pizlo  <fpizlo@apple.com>

        DFG JIT should support integer division
        https://bugs.webkit.org/show_bug.cgi?id=68597

        Reviewed by Darin Adler.
        
        This adds support for ArithDiv speculating integer, and speculating
        that the result is integer (i.e. remainder = 0).
        
        This is a 4% win on Kraken and a 1% loss on V8.

        * bytecode/CodeBlock.h:
        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::makeDivSafe):
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGNode.h:
        (JSC::DFG::Node::hasArithNodeFlags):
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateArithNodeFlags):
        (JSC::DFG::Propagator::propagateNodePredictions):
        (JSC::DFG::Propagator::fixupNode):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):
        * jit/JITArithmetic.cpp:
        (JSC::JIT::emit_op_div):

458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482
2011-09-22  Oliver Hunt  <oliver@apple.com>

        Implement put_scoped_var in the DFG jit
        https://bugs.webkit.org/show_bug.cgi?id=68653

        Reviewed by Gavin Barraclough.

        Naive implementation of put_scoped_var.  Same story as the
        get_scoped_var implementation, although I've hoisted scope
        object acquisition into a separate dfg node.  Ideally in the
        future we would reuse the resolved scope chain object, but
        for now we don't.

        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGCapabilities.h:
        (JSC::DFG::canCompileOpcode):
        * dfg/DFGNode.h:
        (JSC::DFG::Node::hasScopeChainDepth):
        (JSC::DFG::Node::scopeChainDepth):
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNodePredictions):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):

483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545
2011-09-22  Gavin Barraclough  <barraclough@apple.com>

        Implement Function.prototype.bind
        https://bugs.webkit.org/show_bug.cgi?id=26382

        Reviewed by Sam Weinig.

        This patch provides a basic functional implementation
        for Function.bind. It should (hopefully!) be fully
        functionally correct, and the bound functions can be
        called to quickly (since they are a subclass of
        JSFunction, not InternalFunction), but we'll probably
        want to follow up with some optimization work to keep
        bound calls in JIT code.

        * JavaScriptCore.JSVALUE32_64only.exp:
        * JavaScriptCore.JSVALUE64only.exp:
        * JavaScriptCore.exp:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * jit/JITStubs.cpp:
        (JSC::JITThunks::hostFunctionStub):
        * jit/JITStubs.h:
        * jsc.cpp:
        (GlobalObject::addFunction):
        * runtime/CommonIdentifiers.h:
        * runtime/ConstructData.h:
        * runtime/Executable.h:
        (JSC::NativeExecutable::NativeExecutable):
        * runtime/FunctionPrototype.cpp:
        (JSC::FunctionPrototype::addFunctionProperties):
        (JSC::functionProtoFuncBind):
        * runtime/FunctionPrototype.h:
        * runtime/JSBoundFunction.cpp: Added.
        (JSC::boundFunctionCall):
        (JSC::boundFunctionConstruct):
        (JSC::JSBoundFunction::create):
        (JSC::JSBoundFunction::hasInstance):
        (JSC::JSBoundFunction::getOwnPropertySlot):
        (JSC::JSBoundFunction::getOwnPropertyDescriptor):
        (JSC::JSBoundFunction::JSBoundFunction):
        (JSC::JSBoundFunction::finishCreation):
        * runtime/JSBoundFunction.h: Added.
        (JSC::JSBoundFunction::targetFunction):
        (JSC::JSBoundFunction::boundThis):
        (JSC::JSBoundFunction::boundArgs):
        (JSC::JSBoundFunction::createStructure):
        * runtime/JSFunction.cpp:
        (JSC::JSFunction::create):
        (JSC::JSFunction::finishCreation):
        (JSC::createDescriptorForThrowingProperty):
        (JSC::JSFunction::getOwnPropertySlot):
        * runtime/JSFunction.h:
        * runtime/JSGlobalData.cpp:
        (JSC::JSGlobalData::getHostFunction):
        * runtime/JSGlobalData.h:
        * runtime/JSGlobalObject.cpp:
        (JSC::JSGlobalObject::reset):
        (JSC::JSGlobalObject::visitChildren):
        * runtime/JSGlobalObject.h:
        (JSC::JSGlobalObject::boundFunctionStructure):
        * runtime/Lookup.cpp:
        (JSC::setUpStaticFunctionSlot):

546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570
2011-09-22  Oliver Hunt  <oliver@apple.com>

        Implement get_scoped_var in the DFG
        https://bugs.webkit.org/show_bug.cgi?id=68640

        Reviewed by Gavin Barraclough.

        Naive implementation of get_scoped_var in the DFG.  Essentially this
        is the bare minimum required to get correct behaviour, so there's no
        load/store coalescing or type profiling involved, even though these
        would be wins.  No impact on SunSpider or V8.

        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGCapabilities.h:
        (JSC::DFG::canCompileOpcode):
        * dfg/DFGNode.h:
        (JSC::DFG::Node::hasVarNumber):
        (JSC::DFG::Node::hasScopeChainDepth):
        (JSC::DFG::Node::scopeChainDepth):
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNodePredictions):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):

571 572 573 574 575 576 577 578 579 580 581 582
2011-09-22  Adam Roben  <aroben@apple.com>

        Remove FindSafari from all our .sln files

        It isn't used anymore, so there's no point in building it.

        Part of <http://webkit.org/b/68628> Remove FindSafari

        Reviewed by Steve Falkenburg.

        * JavaScriptCore.vcproj/JavaScriptCore.sln:

583 584 585 586 587 588 589 590 591 592 593 594 595 596 597
2011-09-22  Filip Pizlo  <fpizlo@apple.com>

        32-bit call code clobbers the function cell tag
        https://bugs.webkit.org/show_bug.cgi?id=68606

        Reviewed by Csaba Osztrogonác.
        
        This is a minimalistic fix: it simply emits code to restore the
        cell tag on the slow path, if we know that we failed due to
        emitCallIfNotType.

        * jit/JITCall32_64.cpp:
        (JSC::JIT::compileOpCallVarargsSlowCase):
        (JSC::JIT::compileOpCallSlowCase):

598 599 600 601 602 603 604 605 606
2011-09-21  Gavin Barraclough  <barraclough@apple.com>

        Add missing addPtr->add32 mapping for X86.

        Rubber stamped by Sam Weinig.

        * assembler/MacroAssembler.h:
        (JSC::MacroAssembler::addPtr):

607 608 609 610 611 612 613 614 615 616 617 618 619
2011-09-21  Gavin Barraclough  <barraclough@apple.com>

        Add missing addDouble for AbsoluteAddress to X86

        Rubber stamped by Geoff Garen.

        * assembler/MacroAssemblerX86.h:
        (JSC::MacroAssemblerX86::addDouble):
        * assembler/X86Assembler.h:
        (JSC::X86Assembler::addsd_mr):
        (JSC::X86Assembler::cvtsi2sd_rr):
        (JSC::X86Assembler::cvtsi2sd_mr):

620 621 622 623 624 625 626
2011-09-21  Gavin Barraclough  <barraclough@apple.com>

        Build fix following fix for bug #68586.

        * jit/JIT.cpp:
        * jit/JITInlineMethods.h:

627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649
2011-09-21  Filip Pizlo  <fpizlo@apple.com>

        DFG JIT should be able to compile op_throw
        https://bugs.webkit.org/show_bug.cgi?id=68571

        Reviewed by Geoffrey Garen.
        
        This compiles op_throw in the simplest way possible: it's an OSR
        point back to the old JIT. This is a good step towards increasing
        coverage, particularly on Kraken, but it's neutral because the
        same functions that do throw also use some other unsupported
        opcodes.

        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGCapabilities.h:
        (JSC::DFG::canCompileOpcode):
        * dfg/DFGNode.h:
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNodePredictions):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):

650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748
2011-09-21  Filip Pizlo  <fpizlo@apple.com>

        DFG should support continuous optimization
        https://bugs.webkit.org/show_bug.cgi?id=68329

        Reviewed by Geoffrey Garen.
        
        This adds the ability to reoptimize a code block if speculation
        failures happen frequently. 6% speed-up on Kraken, 1% slow-down
        on V8, neutral on SunSpider.

        * CMakeLists.txt:
        * GNUmakefile.list.am:
        * JavaScriptCore.pro:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
        * JavaScriptCore.vcproj/WTF/WTF.vcproj:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * bytecode/CodeBlock.cpp:
        (JSC::CodeBlock::CodeBlock):
        (JSC::ProgramCodeBlock::jettison):
        (JSC::EvalCodeBlock::jettison):
        (JSC::FunctionCodeBlock::jettison):
        (JSC::CodeBlock::shouldOptimizeNow):
        (JSC::CodeBlock::dumpValueProfiles):
        * bytecode/CodeBlock.h:
        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::getStrongPrediction):
        * dfg/DFGJITCompiler.cpp:
        (JSC::DFG::JITCompiler::exitSpeculativeWithOSR):
        (JSC::DFG::JITCompiler::compileEntry):
        (JSC::DFG::JITCompiler::compileBody):
        * dfg/DFGJITCompiler.h:
        (JSC::DFG::JITCompiler::noticeOSREntry):
        * dfg/DFGOSREntry.cpp:
        (JSC::DFG::prepareOSREntry):
        * dfg/DFGOSREntry.h:
        (JSC::DFG::getOSREntryDataBytecodeIndex):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):
        * heap/ConservativeRoots.cpp:
        (JSC::ConservativeRoots::ConservativeRoots):
        (JSC::ConservativeRoots::~ConservativeRoots):
        (JSC::DummyMarkHook::mark):
        (JSC::ConservativeRoots::genericAddPointer):
        (JSC::ConservativeRoots::genericAddSpan):
        (JSC::ConservativeRoots::add):
        * heap/ConservativeRoots.h:
        * heap/Heap.cpp:
        (JSC::Heap::addJettisonCodeBlock):
        (JSC::Heap::markRoots):
        * heap/Heap.h:
        * heap/JettisonedCodeBlocks.cpp: Added.
        (JSC::JettisonedCodeBlocks::JettisonedCodeBlocks):
        (JSC::JettisonedCodeBlocks::~JettisonedCodeBlocks):
        (JSC::JettisonedCodeBlocks::addCodeBlock):
        (JSC::JettisonedCodeBlocks::clearMarks):
        (JSC::JettisonedCodeBlocks::deleteUnmarkedCodeBlocks):
        (JSC::JettisonedCodeBlocks::traceCodeBlocks):
        * heap/JettisonedCodeBlocks.h: Added.
        (JSC::JettisonedCodeBlocks::mark):
        * interpreter/RegisterFile.cpp:
        (JSC::RegisterFile::gatherConservativeRoots):
        * interpreter/RegisterFile.h:
        * jit/JITStubs.cpp:
        (JSC::DEFINE_STUB_FUNCTION):
        * runtime/Executable.cpp:
        (JSC::jettisonCodeBlock):
        (JSC::EvalExecutable::jettisonOptimizedCode):
        (JSC::ProgramExecutable::jettisonOptimizedCode):
        (JSC::FunctionExecutable::jettisonOptimizedCodeForCall):
        (JSC::FunctionExecutable::jettisonOptimizedCodeForConstruct):
        * runtime/Executable.h:
        (JSC::FunctionExecutable::jettisonOptimizedCodeFor):
        * wtf/BitVector.h: Added.
        (WTF::BitVector::BitVector):
        (WTF::BitVector::~BitVector):
        (WTF::BitVector::operator=):
        (WTF::BitVector::size):
        (WTF::BitVector::ensureSize):
        (WTF::BitVector::resize):
        (WTF::BitVector::clearAll):
        (WTF::BitVector::get):
        (WTF::BitVector::set):
        (WTF::BitVector::clear):
        (WTF::BitVector::bitsInPointer):
        (WTF::BitVector::maxInlineBits):
        (WTF::BitVector::byteCount):
        (WTF::BitVector::makeInlineBits):
        (WTF::BitVector::OutOfLineBits::numBits):
        (WTF::BitVector::OutOfLineBits::numWords):
        (WTF::BitVector::OutOfLineBits::bits):
        (WTF::BitVector::OutOfLineBits::create):
        (WTF::BitVector::OutOfLineBits::destroy):
        (WTF::BitVector::OutOfLineBits::OutOfLineBits):
        (WTF::BitVector::isInline):
        (WTF::BitVector::outOfLineBits):
        (WTF::BitVector::resizeOutOfLine):
        (WTF::BitVector::bits):

749 750 751 752 753 754 755 756 757 758 759 760
2011-09-21  Gavin Barraclough  <barraclough@apple.com>

        Add X86 GPRInfo for DFG JIT.
        https://bugs.webkit.org/show_bug.cgi?id=68586

        Reviewed by Geoff Garen.

        * dfg/DFGGPRInfo.h:
        (JSC::DFG::GPRInfo::toRegister):
        (JSC::DFG::GPRInfo::toIndex):
        (JSC::DFG::GPRInfo::debugName):

761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785
2011-09-21  Gavin Barraclough  <barraclough@apple.com>

        Should support value profiling on CPU(X86)
        https://bugs.webkit.org/show_bug.cgi?id=68575

        Reviewed by Sam Weinig.

        Fix verbose profiling in ToT (SlowCaseProfile had been
        partially renamed to RareCaseProfile), add in-memory
        bucket counter for CPU(X86), move JIT::m_canBeOptimized
        out of the DFG_JIT ifdef.

        * bytecode/CodeBlock.cpp:
        (JSC::CodeBlock::resetRareCaseProfiles):
        (JSC::CodeBlock::dumpValueProfiles):
        * bytecode/CodeBlock.h:
        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::makeSafe):
        * jit/JIT.cpp:
        (JSC::JIT::privateCompileSlowCases):
        (JSC::JIT::privateCompile):
        * jit/JIT.h:
        * jit/JITInlineMethods.h:
        (JSC::JIT::emitValueProfilingSite):

786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818
2011-09-21  Filip Pizlo  <fpizlo@apple.com>

        DFG does not support compiling functions as constructors
        https://bugs.webkit.org/show_bug.cgi?id=68500

        Reviewed by Oliver Hunt.
        
        This adds support for compiling constructors to the DFG. It's a
        1% speed-up on V8, mostly due to a 6% speed-up on early-boyer.
        It's also a 13% win on access-binary-trees, but it's neutral in
        the SunSpider and Kraken averages.

        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGCapabilities.h:
        (JSC::DFG::mightCompileFunctionForConstruct):
        (JSC::DFG::canCompileOpcode):
        * dfg/DFGNode.h:
        * dfg/DFGOperations.cpp:
        * dfg/DFGOperations.h:
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNodePredictions):
        (JSC::DFG::Propagator::performNodeCSE):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):
        * runtime/Executable.cpp:
        (JSC::FunctionExecutable::compileOptimizedForConstruct):
        (JSC::FunctionExecutable::compileForConstructInternal):
        * runtime/Executable.h:
        (JSC::FunctionExecutable::compileForConstruct):
        (JSC::FunctionExecutable::compileFor):
        (JSC::FunctionExecutable::compileOptimizedFor):

819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865
2011-09-21  Gavin Barraclough  <barraclough@apple.com>

        Replace jsFunctionVPtr compares with a type check on the Structure.
        https://bugs.webkit.org/show_bug.cgi?id=68557

        Reviewed by Oliver Hunt.

        This will permit calls to still optimize to subclasses of JSFunction
        that have the correct type (but a different C++ vptr).

        This patch stops passing the globalData into numerous functions.

        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGGraph.h:
        (JSC::DFG::Graph::isFunctionConstant):
        (JSC::DFG::Graph::valueOfFunctionConstant):
        * dfg/DFGJITCompiler.h:
        (JSC::DFG::JITCompiler::isFunctionConstant):
        (JSC::DFG::JITCompiler::valueOfFunctionConstant):
        * dfg/DFGOperations.cpp:
        * interpreter/Interpreter.cpp:
        (JSC::Interpreter::privateExecute):
        * jit/JIT.h:
        * jit/JITCall.cpp:
        (JSC::JIT::compileOpCallVarargs):
        (JSC::JIT::compileOpCallSlowCase):
        * jit/JITCall32_64.cpp:
        (JSC::JIT::compileOpCallVarargs):
        (JSC::JIT::compileOpCallSlowCase):
        * jit/JITInlineMethods.h:
        (JSC::JIT::emitJumpIfNotType):
        * jit/JITStubs.cpp:
        (JSC::DEFINE_STUB_FUNCTION):
        * runtime/Executable.h:
        (JSC::isHostFunction):
        * runtime/JSFunction.h:
        (JSC::JSFunction::createStructure):
        * runtime/JSObject.cpp:
        (JSC::JSObject::put):
        (JSC::JSObject::putWithAttributes):
        * runtime/JSObject.h:
        (JSC::getJSFunction):
        (JSC::JSObject::putDirect):
        (JSC::JSObject::putDirectWithoutTransition):
        * runtime/JSType.h:

866 867 868 869 870 871 872 873 874 875 876 877 878
2011-09-21  Geoffrey Garen  <ggaren@apple.com>

        Removed WTFTHREADDATA_MULTITHREADED, making it always true
        https://bugs.webkit.org/show_bug.cgi?id=68549

        Reviewed by Darin Adler.
        
        Another part of making threads exist in WebKit.

        * wtf/WTFThreadData.cpp:
        * wtf/WTFThreadData.h:
        (WTF::wtfThreadData):

879 880 881 882 883 884 885 886 887 888
2011-09-21  Dan Bernstein  <mitz@apple.com>

        JavaScriptCore Part of: Prevent the WebKit frameworks from defining inappropriately-named Objective-C classes
        https://bugs.webkit.org/show_bug.cgi?id=68451

        Reviewed by Darin Adler.

        * JavaScriptCore.xcodeproj/project.pbxproj: Added a script build phase that invokes
        check-for-inappropriate-objc-class-names, allowing only class names prefixed with "JS".

barraclough@apple.com's avatar
barraclough@apple.com committed
889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914
2011-09-20  Gavin Barraclough  <barraclough@apple.com>

        MacroAssembler fixes.
        https://bugs.webkit.org/show_bug.cgi?id=68494

        Reviewed by Sam Weinig.

        Add X86-64's 3 operand or32 to other MacroAssembler, fix load32's [const] void* mismatch

        * assembler/MacroAssembler.h:
        (JSC::MacroAssembler::orPtr):
        (JSC::MacroAssembler::loadPtr):
        * assembler/MacroAssemblerARM.h:
        (JSC::MacroAssemblerARM::or32):
        * assembler/MacroAssemblerARMv7.h:
        (JSC::MacroAssemblerARMv7::or32):
        * assembler/MacroAssemblerMIPS.h:
        (JSC::MacroAssemblerMIPS::or32):
        * assembler/MacroAssemblerSH4.h:
        (JSC::MacroAssemblerSH4::or32):
        (JSC::MacroAssemblerSH4::load32):
        * assembler/MacroAssemblerX86.h:
        (JSC::MacroAssemblerX86::load32):
        * assembler/MacroAssemblerX86_64.h:
        (JSC::MacroAssemblerX86_64::load32):

ggaren@apple.com's avatar
ggaren@apple.com committed
915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932
2011-09-20  Geoffrey Garen  <ggaren@apple.com>

        Some Heap cleanup.

        Reviewed by Beth Dakin.

        * heap/MarkedBlock.cpp:
        (JSC::MarkedBlock::blessNewBlock): Removed blessNewBlockForSlowPath()
        because it was unused; renamed blessNewBlockForFastPath() to blessNewBlock()
        since there is only one now.

        * heap/MarkedBlock.h: Removed ownerSet-related stuff since it was unused.
        Updated mark bit overhead calculation. Deployed atomsPerBlock in one
        place where we were recalculating it.

        * heap/MarkedSpace.cpp:
        (JSC::MarkedSpace::addBlock): Updated for rename.

933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957
2011-09-20  Filip Pizlo  <fpizlo@apple.com>

        DFG JIT always speculates integer on modulo
        https://bugs.webkit.org/show_bug.cgi?id=68485

        Reviewed by Oliver Hunt.
        
        Added support for double modulo, which is a call to fmod().
        Also added support for recording the old JIT's statistics
        on op_mod and propagating them along the graph. Finally,
        fixed a goof in the ArithNodeFlags propagation logic that
        was made obvious when I started testing ArithMod.

        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::makeSafe):
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGNode.h:
        (JSC::DFG::Node::hasArithNodeFlags):
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateArithNodeFlags):
        (JSC::DFG::Propagator::propagateNodePredictions):
        (JSC::DFG::Propagator::fixupNode):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):

958 959 960 961 962 963 964 965 966 967 968
2011-09-20  ChangSeok Oh  <shivamidow@gmail.com>

        [GTK] requestAnimationFrame support for gtk port
        https://bugs.webkit.org/show_bug.cgi?id=66280

        Reviewed by Martin Robinson.

        Let GTK port use REQUEST_ANIMATION_FRAME_TIMER.

        * wtf/Platform.h:

969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039
2011-09-20  Filip Pizlo  <fpizlo@apple.com>

        DFG JIT performs too many negative zero checks, and too many
        overflow checks
        https://bugs.webkit.org/show_bug.cgi?id=68430

        Reviewed by Oliver Hunt.
        
        This adds comprehensive support for deciding how to perform an
        arithmetic operations based on a combination of overflow profiling,
        negative zero profiling, value profiling, and a static analysis of
        how the results of these operations get used.
        
        This is a 72% speed-up on stanford-crypto-sha256-iterative, and a
        2.5% speed-up on the Kraken average, a 1.4% speed-up on the V8
        geomean, and neutral on SunSpider. It's also an 8.5% speed-up on
        V8-crypto, because apparenty everything we do speeds up crypto.

        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::toInt32):
        (JSC::DFG::ByteCodeParser::toNumber):
        (JSC::DFG::ByteCodeParser::isSmallInt32Constant):
        (JSC::DFG::ByteCodeParser::valueOfInt32Constant):
        (JSC::DFG::ByteCodeParser::weaklyPredictInt32):
        (JSC::DFG::ByteCodeParser::makeSafe):
        (JSC::DFG::ByteCodeParser::handleMinMax):
        (JSC::DFG::ByteCodeParser::handleIntrinsic):
        (JSC::DFG::ByteCodeParser::parseBlock):
        (JSC::DFG::ByteCodeParser::processPhiStack):
        (JSC::DFG::ByteCodeParser::parse):
        * dfg/DFGGraph.cpp:
        (JSC::DFG::Graph::dump):
        * dfg/DFGJITCodeGenerator.cpp:
        (JSC::DFG::JITCodeGenerator::nonSpeculativeBasicArithOp):
        * dfg/DFGNode.h:
        (JSC::DFG::nodeUsedAsNumber):
        (JSC::DFG::nodeCanTruncateInteger):
        (JSC::DFG::nodeCanIgnoreNegativeZero):
        (JSC::DFG::nodeCanSpeculateInteger):
        (JSC::DFG::arithNodeFlagsAsString):
        (JSC::DFG::Node::Node):
        (JSC::DFG::Node::hasArithNodeFlags):
        (JSC::DFG::Node::rawArithNodeFlags):
        (JSC::DFG::Node::arithNodeFlags):
        (JSC::DFG::Node::arithNodeFlagsForCompare):
        (JSC::DFG::Node::setArithNodeFlag):
        (JSC::DFG::Node::mergeArithNodeFlags):
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::fixpoint):
        (JSC::DFG::Propagator::isNotNegZero):
        (JSC::DFG::Propagator::isNotZero):
        (JSC::DFG::Propagator::propagateArithNodeFlags):
        (JSC::DFG::Propagator::propagateArithNodeFlagsForward):
        (JSC::DFG::Propagator::propagateArithNodeFlagsBackward):
        (JSC::DFG::Propagator::propagateNodePredictions):
        (JSC::DFG::Propagator::propagatePredictionsForward):
        (JSC::DFG::Propagator::propagatePredictionsBackward):
        (JSC::DFG::Propagator::toDouble):
        (JSC::DFG::Propagator::fixupNode):
        (JSC::DFG::Propagator::fixup):
        (JSC::DFG::Propagator::startIndexForChildren):
        (JSC::DFG::Propagator::endIndexForPureCSE):
        (JSC::DFG::Propagator::pureCSE):
        (JSC::DFG::Propagator::clobbersWorld):
        (JSC::DFG::Propagator::setReplacement):
        (JSC::DFG::Propagator::performNodeCSE):
        (JSC::DFG::Propagator::localCSE):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):
        (JSC::DFG::SpeculativeJIT::computeValueRecoveryFor):

1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108
2011-09-19  Oliver Hunt  <oliver@apple.com>

        Refactor Heap allocation logic into separate AllocationSpace class
        https://bugs.webkit.org/show_bug.cgi?id=68409

        Reviewed by Gavin Barraclough.

        This patch hoists direct manipulation of the MarkedSpace and related
        data out of Heap and into a separate class.  This will allow us to
        have multiple allocation spaces in future, so easing the way towards
        having GC'd backing stores for objects.

        * CMakeLists.txt:
        * GNUmakefile.list.am:
        * JavaScriptCore.exp:
        * JavaScriptCore.gypi:
        * JavaScriptCore.pro:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * debugger/Debugger.cpp:
        (JSC::Debugger::recompileAllJSFunctions):
        * heap/AllocationSpace.cpp: Added.
        (JSC::AllocationSpace::tryAllocate):
        (JSC::AllocationSpace::allocateSlowCase):
        (JSC::AllocationSpace::allocateBlock):
        (JSC::AllocationSpace::freeBlocks):
        (JSC::TakeIfEmpty::TakeIfEmpty):
        (JSC::TakeIfEmpty::operator()):
        (JSC::TakeIfEmpty::returnValue):
        (JSC::AllocationSpace::shrink):
        * heap/AllocationSpace.h: Added.
        (JSC::AllocationSpace::AllocationSpace):
        (JSC::AllocationSpace::blocks):
        (JSC::AllocationSpace::sizeClassFor):
        (JSC::AllocationSpace::setHighWaterMark):
        (JSC::AllocationSpace::highWaterMark):
        (JSC::AllocationSpace::canonicalizeBlocks):
        (JSC::AllocationSpace::resetAllocator):
        (JSC::AllocationSpace::forEachCell):
        (JSC::AllocationSpace::forEachBlock):
        (JSC::AllocationSpace::allocate):
        * heap/Heap.cpp:
        (JSC::Heap::Heap):
        (JSC::Heap::reportExtraMemoryCostSlowCase):
        (JSC::Heap::getConservativeRegisterRoots):
        (JSC::Heap::markRoots):
        (JSC::Heap::clearMarks):
        (JSC::Heap::sweep):
        (JSC::Heap::objectCount):
        (JSC::Heap::size):
        (JSC::Heap::capacity):
        (JSC::Heap::globalObjectCount):
        (JSC::Heap::objectTypeCounts):
        (JSC::Heap::collect):
        (JSC::Heap::canonicalizeBlocks):
        (JSC::Heap::resetAllocator):
        (JSC::Heap::freeBlocks):
        (JSC::Heap::shrink):
        * heap/Heap.h:
        (JSC::Heap::objectSpace):
        (JSC::Heap::sizeClassForObject):
        (JSC::Heap::allocate):
        * jit/JITInlineMethods.h:
        (JSC::JIT::emitAllocateBasicJSObject):
        * runtime/JSGlobalData.cpp:
        (JSC::JSGlobalData::recompileAllJSFunctions):
        (JSC::JSGlobalData::releaseExecutableMemory):

1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144
2011-09-19  Geoffrey Garen  <ggaren@apple.com>

        Removed BREWMP* platform #ifdefs
        https://bugs.webkit.org/show_bug.cgi?id=68425
        
        BREWMP* has no maintainer, and this is dead code.

        Reviewed by Darin Adler.

        * heap/MarkStack.h:
        (JSC::::shrinkAllocation):
        * jit/ExecutableAllocator.h:
        (JSC::ExecutableAllocator::cacheFlush):
        * runtime/TimeoutChecker.cpp:
        (JSC::getCPUTime):
        * wtf/Assertions.cpp:
        * wtf/Assertions.h:
        * wtf/CurrentTime.cpp:
        * wtf/DateMath.cpp:
        (WTF::calculateUTCOffset):
        * wtf/FastMalloc.cpp:
        (WTF::fastMalloc):
        (WTF::fastCalloc):
        (WTF::fastMallocSize):
        * wtf/FastMalloc.h:
        * wtf/MainThread.cpp:
        * wtf/MathExtras.h:
        * wtf/OwnPtrCommon.h:
        * wtf/Platform.h:
        * wtf/RandomNumber.cpp:
        (WTF::randomNumber):
        * wtf/RandomNumberSeed.h:
        (WTF::initializeRandomNumberGenerator):
        * wtf/text/WTFString.h:
        * wtf/unicode/Unicode.h:

1145 1146 1147 1148 1149 1150
2011-09-20  Adam Roben  <aroben@apple.com>

        Windows build fix after r95523

        * wtf/CheckedArithmetic.h: Added stdint.h so we can have int64_t defined.

1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227
2011-09-18  Filip Pizlo  <fpizlo@apple.com>

        DFG JIT does not speculate aggressively enough on GetById
        https://bugs.webkit.org/show_bug.cgi?id=68320

        Reviewed by Oliver Hunt.
        
        This adds the ability to access properties directly, by offset.
        This optimization kicks in when at the time of DFG compilation,
        it appears that the given get_by_id is self-cached by the old JIT.
        Two new opcodes get introduced: CheckStructure and GetByOffset.
        CheckStructure performs a speculation check on the object's
        structure, and returns the storage pointer. GetByOffset performs
        a direct read of the field from the storage pointer. Both
        CheckStructure and GetByOffset can be CSE'd, so that we can
        eliminate redundant structure checks, and redundant reads of the
        same field.
        
        This is a 4% speed-up on V8, a 2% slow-down on Kraken, and
        neutral on SunSpider.

        * bytecode/PredictedType.cpp:
        (JSC::predictionFromClassInfo):
        (JSC::predictionFromStructure):
        (JSC::predictionFromCell):
        * bytecode/PredictedType.h:
        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGGenerationInfo.h:
        (JSC::DFG::dataFormatToString):
        (JSC::DFG::needDataFormatConversion):
        (JSC::DFG::GenerationInfo::initStorage):
        (JSC::DFG::GenerationInfo::spill):
        (JSC::DFG::GenerationInfo::fillStorage):
        * dfg/DFGGraph.h:
        (JSC::DFG::Graph::predict):
        (JSC::DFG::Graph::getPrediction):
        * dfg/DFGJITCodeGenerator.cpp:
        (JSC::DFG::JITCodeGenerator::fillInteger):
        (JSC::DFG::JITCodeGenerator::fillDouble):
        (JSC::DFG::JITCodeGenerator::fillJSValue):
        (JSC::DFG::JITCodeGenerator::fillStorage):
        (JSC::DFG::GPRTemporary::GPRTemporary):
        * dfg/DFGJITCodeGenerator.h:
        (JSC::DFG::JITCodeGenerator::silentSpillGPR):
        (JSC::DFG::JITCodeGenerator::silentFillGPR):
        (JSC::DFG::JITCodeGenerator::spill):
        (JSC::DFG::JITCodeGenerator::storageResult):
        (JSC::DFG::StorageOperand::StorageOperand):
        (JSC::DFG::StorageOperand::~StorageOperand):
        (JSC::DFG::StorageOperand::index):
        (JSC::DFG::StorageOperand::gpr):
        (JSC::DFG::StorageOperand::use):
        * dfg/DFGNode.h:
        (JSC::DFG::OpInfo::OpInfo):
        (JSC::DFG::Node::Node):
        (JSC::DFG::Node::hasPrediction):
        (JSC::DFG::Node::hasStructure):
        (JSC::DFG::Node::structure):
        (JSC::DFG::Node::hasStorageAccessData):
        (JSC::DFG::Node::storageAccessDataIndex):
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNode):
        (JSC::DFG::Propagator::globalVarLoadElimination):
        (JSC::DFG::Propagator::getMethodLoadElimination):
        (JSC::DFG::Propagator::checkStructureLoadElimination):
        (JSC::DFG::Propagator::getByOffsetLoadElimination):
        (JSC::DFG::Propagator::performNodeCSE):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::fillSpeculateIntInternal):
        (JSC::DFG::SpeculativeJIT::fillSpeculateDouble):
        (JSC::DFG::SpeculativeJIT::fillSpeculateCell):
        (JSC::DFG::SpeculativeJIT::fillSpeculateBoolean):
        (JSC::DFG::SpeculativeJIT::compile):
        * wtf/StdLibExtras.h:
        (WTF::safeCast):

1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254
2011-09-19  Mark Hahnenberg  <mhahnenberg@apple.com>

        Remove toPrimitive from JSCell
        https://bugs.webkit.org/show_bug.cgi?id=67875

        Reviewed by Darin Adler.

        Part of the refactoring process to un-virtualize JSCell.  We move 
        all of the implicit functionality provided by the virtual toPrimitive method 
        in JSCell to be explicit in JSValue::toPrimitive and JSCell:toPrimitive while 
        also de-virtualizing JSCell::toPrimitive.

        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
        * runtime/JSCell.cpp:
        (JSC::JSCell::toPrimitive):
        * runtime/JSCell.h:

        We replace JSNotAnObject::toPrimitive with defaultValue, which it overrides from 
        JSObject.  This pushes the virtual method further down, enabling us to get rid 
        of the virtual call in JSCell.  Eventually we'll probably have to deal with this
        again, but we'll cross that bridge when we come to it.
        * runtime/JSNotAnObject.cpp:
        (JSC::JSNotAnObject::defaultValue):
        * runtime/JSNotAnObject.h:
        * runtime/JSObject.h:
        * runtime/JSString.h:

1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278
2011-09-19  Geoffrey Garen  <ggaren@apple.com>

        Removed ENABLE_LAZY_BLOCK_FREEING and related #ifdefs
        https://bugs.webkit.org/show_bug.cgi?id=68424

        As discussed on webkit-dev. All ports build with threads enabled in JSC now.
        
        This may break WinCE and other ports that have not built and tested with
        this configuration. I've filed bugs for port maintainers. It's time for
        WebKit to move forward.

        Reviewed by Mark Rowe.

        * heap/Heap.cpp:
        (JSC::Heap::Heap):
        (JSC::Heap::~Heap):
        (JSC::Heap::destroy):
        (JSC::Heap::blockFreeingThreadMain):
        (JSC::Heap::allocateBlock):
        (JSC::Heap::freeBlocks):
        (JSC::Heap::releaseFreeBlocks):
        * heap/Heap.h:
        * wtf/Platform.h:

1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312
2011-09-19  Geoffrey Garen  <ggaren@apple.com>

        Removed ENABLE_WTF_MULTIPLE_THREADS and related #ifdefs
        https://bugs.webkit.org/show_bug.cgi?id=68423

        As discussed on webkit-dev. All ports build with threads enabled in WTF now.
        
        This may break WinCE and other ports that have not built and tested with
        this configuration. I've filed bugs for port maintainers. It's time for
        WebKit to move forward.

        Reviewed by Mark Rowe.

        * wtf/CryptographicallyRandomNumber.cpp:
        (WTF::ARC4Stream::ARC4RandomNumberGenerator::randomNumber):
        (WTF::ARC4Stream::ARC4RandomNumberGenerator::randomValues):
        * wtf/FastMalloc.cpp:
        * wtf/Platform.h:
        * wtf/RandomNumber.cpp:
        (WTF::randomNumber):
        * wtf/RefCountedLeakCounter.cpp:
        (WTF::RefCountedLeakCounter::increment):
        (WTF::RefCountedLeakCounter::decrement):
        * wtf/ThreadingPthreads.cpp:
        (WTF::initializeThreading):
        * wtf/ThreadingWin.cpp:
        (WTF::initializeThreading):
        * wtf/dtoa.cpp:
        (WTF::pow5mult):
        * wtf/gtk/ThreadingGtk.cpp:
        (WTF::initializeThreading):
        * wtf/qt/ThreadingQt.cpp:
        (WTF::initializeThreading):

1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344
2011-09-19  Geoffrey Garen  <ggaren@apple.com>

        Removed ENABLE_JSC_MULTIPLE_THREADS and related #ifdefs.
        https://bugs.webkit.org/show_bug.cgi?id=68422
        
        As discussed on webkit-dev. All ports build with threads enabled in JSC now.
        
        This may break WinCE and other ports that have not built and tested with
        this configuration. I've filed bugs for port maintainers. It's time for
        WebKit to move forward.

        Reviewed by Sam Weinig.

        * API/APIShims.h:
        (JSC::APIEntryShimWithoutLock::APIEntryShimWithoutLock):
        * API/JSContextRef.cpp:
        * heap/MachineStackMarker.cpp:
        (JSC::MachineThreads::MachineThreads):
        (JSC::MachineThreads::~MachineThreads):
        (JSC::MachineThreads::gatherConservativeRoots):
        * heap/MachineStackMarker.h:
        * runtime/InitializeThreading.cpp:
        (JSC::initializeThreadingOnce):
        (JSC::initializeThreading):
        * runtime/JSGlobalData.cpp:
        (JSC::JSGlobalData::sharedInstance):
        * runtime/JSGlobalData.h:
        (JSC::JSGlobalData::makeUsableFromMultipleThreads):
        * runtime/JSLock.cpp:
        * runtime/Structure.cpp:
        * wtf/Platform.h:

1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400
2011-09-19  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r95493 and r95496.
        http://trac.webkit.org/changeset/95493
        http://trac.webkit.org/changeset/95496
        https://bugs.webkit.org/show_bug.cgi?id=68418

        Broke Windows build (Requested by rniwa on #webkit).

        * CMakeLists.txt:
        * GNUmakefile.list.am:
        * JavaScriptCore.exp:
        * JavaScriptCore.gypi:
        * JavaScriptCore.pro:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * debugger/Debugger.cpp:
        (JSC::Debugger::recompileAllJSFunctions):
        * heap/AllocationSpace.cpp: Removed.
        * heap/AllocationSpace.h: Removed.
        * heap/Heap.cpp:
        (JSC::CountFunctor::TakeIfEmpty::TakeIfEmpty):
        (JSC::CountFunctor::TakeIfEmpty::operator()):
        (JSC::CountFunctor::TakeIfEmpty::returnValue):
        (JSC::Heap::Heap):
        (JSC::Heap::reportExtraMemoryCostSlowCase):
        (JSC::Heap::tryAllocate):
        (JSC::Heap::allocateSlowCase):
        (JSC::Heap::getConservativeRegisterRoots):
        (JSC::Heap::markRoots):
        (JSC::Heap::clearMarks):
        (JSC::Heap::sweep):
        (JSC::Heap::objectCount):
        (JSC::Heap::size):
        (JSC::Heap::capacity):
        (JSC::Heap::globalObjectCount):
        (JSC::Heap::objectTypeCounts):
        (JSC::Heap::collect):
        (JSC::Heap::canonicalizeBlocks):
        (JSC::Heap::resetAllocator):
        (JSC::Heap::allocateBlock):
        (JSC::Heap::freeBlocks):
        (JSC::Heap::shrink):
        * heap/Heap.h:
        (JSC::Heap::markedSpace):
        (JSC::Heap::forEachCell):
        (JSC::Heap::forEachBlock):
        (JSC::Heap::sizeClassFor):
        (JSC::Heap::allocate):
        * jit/JITInlineMethods.h:
        (JSC::JIT::emitAllocateBasicJSObject):
        * runtime/JSGlobalData.cpp:
        (JSC::JSGlobalData::recompileAllJSFunctions):
        (JSC::JSGlobalData::releaseExecutableMemory):

1401 1402 1403 1404 1405 1406 1407
2011-09-19  Gavin Barraclough  <barraclough@apple.com>

        Errrk, missed stylebot comments in last commit.

        * runtime/StringPrototype.cpp:
        (JSC::stringProtoFuncSplit):

barraclough@apple.com's avatar
barraclough@apple.com committed
1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423
2011-09-19  Gavin Barraclough  <barraclough@apple.com>

        String#split is buggy
        https://bugs.webkit.org/show_bug.cgi?id=68348

        Reviewed by Sam Weinig.

        * runtime/StringPrototype.cpp:
        (JSC::jsStringWithReuse):
            - added helper function to reuse original JSString value.
        (JSC::stringProtoFuncSplit):
            - Rewritten from the spec.
        * tests/mozilla/ecma/String/15.5.4.8-2.js:
        (getTestCases):
            - This test is not ES5 compliant.

1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473
2011-09-19  Geoffrey Garen  <ggaren@apple.com>

        Removed lots of friend declarations from JSCell, so we can more
        effectively make use of private and protected.

        Reviewed by Sam Weinig.

        * runtime/JSCell.h: Removed MSVCBugWorkaround because it was a lot of
        confusion for not much safety.
        (JSC::JSCell::operator new): Made this public because it is used by a
        few clients, and not really dangerous.

        * runtime/JSObject.cpp:
        (JSC::JSObject::put):
        (JSC::JSObject::deleteProperty):
        (JSC::JSObject::defineGetter):
        (JSC::JSObject::defineSetter):
        (JSC::JSObject::getPropertySpecificValue):
        (JSC::JSObject::getOwnPropertyNames):
        (JSC::JSObject::seal):
        (JSC::JSObject::freeze):
        (JSC::JSObject::preventExtensions):
        (JSC::JSObject::removeDirect):
        (JSC::JSObject::createInheritorID):
        (JSC::JSObject::allocatePropertyStorage):
        (JSC::JSObject::getOwnPropertyDescriptor):
        * runtime/JSObject.h:
        (JSC::JSObject::getDirect):
        (JSC::JSObject::getDirectLocation):
        (JSC::JSObject::hasCustomProperties):
        (JSC::JSObject::hasGetterSetterProperties):
        (JSC::JSObject::isSealed):
        (JSC::JSObject::isFrozen):
        (JSC::JSObject::isExtensible):
        (JSC::JSObject::flattenDictionaryObject):
        (JSC::JSObject::finishCreation):
        (JSC::JSObject::prototype):
        (JSC::JSObject::setPrototype):
        (JSC::JSObject::inlineGetOwnPropertySlot):
        (JSC::JSCell::fastGetOwnProperty):
        (JSC::JSObject::putDirectInternal):
        (JSC::JSObject::putDirectWithoutTransition):
        (JSC::JSObject::transitionTo):
        (JSC::JSObject::visitChildrenDirect): Changed all use of m_structure to
        structure() / setStructure(), so we don't have to be a friend of JSCell.

        * runtime/Structure.h:
        (JSC::JSCell::setStructure): Added, to avoid direct access by JSObject
        to JSCell::m_structure.

1474 1475 1476 1477 1478 1479 1480 1481 1482
2011-09-19  Adam Barth  <abarth@webkit.org>

        Always enable ENABLE(EVENTSOURCE)
        https://bugs.webkit.org/show_bug.cgi?id=68414

        Reviewed by Eric Seidel.

        * Configurations/FeatureDefines.xcconfig:

1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500
2011-09-19  Eli Fidler  <efidler@rim.com>

        Enable JSC_MULTIPLE_THREADS for OS(QNX).
        https://bugs.webkit.org/show_bug.cgi?id=68047

        Reviewed by Daniel Bates.

        SA_RESTART was required for SIGUSR2-based debugging, but is not
        present on QNX. This debugging doesn't seem critical to
        JSC_MULTIPLE_THREADS, so allow it to proceed.

        * heap/MachineStackMarker.cpp:
        (JSC::MachineThreads::Thread::Thread):
        (JSC::getPlatformThreadRegisters):
        (JSC::otherThreadStackPointer):
        (JSC::freePlatformThreadRegisters):
        * wtf/Platform.h: enable PTHREADS for OS(QNX)

oliver@apple.com's avatar
oliver@apple.com committed
1501 1502 1503 1504 1505 1506
2011-09-19  Oliver Hunt  <oliver@apple.com>

        Windows build fix.

        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:

1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574
2011-09-19  Oliver Hunt  <oliver@apple.com>

        Refactor Heap allocation logic into separate AllocationSpace class
        https://bugs.webkit.org/show_bug.cgi?id=68409

        Reviewed by Gavin Barraclough.

        This patch hoists direct manipulation of the MarkedSpace and related
        data out of Heap and into a separate class.  This will allow us to
        have multiple allocation spaces in future, so easing the way towards
        having GC'd backing stores for objects.

        * CMakeLists.txt:
        * GNUmakefile.list.am:
        * JavaScriptCore.exp:
        * JavaScriptCore.gypi:
        * JavaScriptCore.pro:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * debugger/Debugger.cpp:
        (JSC::Debugger::recompileAllJSFunctions):
        * heap/AllocationSpace.cpp: Added.
        (JSC::AllocationSpace::tryAllocate):
        (JSC::AllocationSpace::allocateSlowCase):
        (JSC::AllocationSpace::allocateBlock):
        (JSC::AllocationSpace::freeBlocks):
        (JSC::TakeIfEmpty::TakeIfEmpty):
        (JSC::TakeIfEmpty::operator()):
        (JSC::TakeIfEmpty::returnValue):
        (JSC::AllocationSpace::shrink):
        * heap/AllocationSpace.h: Added.
        (JSC::AllocationSpace::AllocationSpace):
        (JSC::AllocationSpace::blocks):
        (JSC::AllocationSpace::sizeClassFor):
        (JSC::AllocationSpace::setHighWaterMark):
        (JSC::AllocationSpace::highWaterMark):
        (JSC::AllocationSpace::canonicalizeBlocks):
        (JSC::AllocationSpace::resetAllocator):
        (JSC::AllocationSpace::forEachCell):
        (JSC::AllocationSpace::forEachBlock):
        (JSC::AllocationSpace::allocate):
        * heap/Heap.cpp:
        (JSC::Heap::Heap):
        (JSC::Heap::reportExtraMemoryCostSlowCase):
        (JSC::Heap::getConservativeRegisterRoots):
        (JSC::Heap::markRoots):
        (JSC::Heap::clearMarks):
        (JSC::Heap::sweep):
        (JSC::Heap::objectCount):
        (JSC::Heap::size):
        (JSC::Heap::capacity):
        (JSC::Heap::globalObjectCount):
        (JSC::Heap::objectTypeCounts):
        (JSC::Heap::collect):
        (JSC::Heap::canonicalizeBlocks):
        (JSC::Heap::resetAllocator):
        (JSC::Heap::freeBlocks):
        (JSC::Heap::shrink):
        * heap/Heap.h:
        (JSC::Heap::objectSpace):
        (JSC::Heap::sizeClassForObject):
        (JSC::Heap::allocate):
        * jit/JITInlineMethods.h:
        (JSC::JIT::emitAllocateBasicJSObject):
        * runtime/JSGlobalData.cpp:
        (JSC::JSGlobalData::recompileAllJSFunctions):
        (JSC::JSGlobalData::releaseExecutableMemory):

1575 1576 1577 1578 1579 1580 1581
2011-09-19  Adam Roben  <aroben@apple.com>

        Windows build fix after r95310

        * JavaScriptCore.vcproj/testRegExp/testRegExpCommon.vsprops: Added
        include\private\JavaScriptCore to the include path so DFGIntrinsic.h can be found.

1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644
2011-09-19  Filip Pizlo  <fpizlo@apple.com>

        DFG speculation failures should act as additional value profiles
        https://bugs.webkit.org/show_bug.cgi?id=68335

        Reviewed by Oliver Hunt.
        
        This adds slow-case counters to the old JIT. It also ensures that
        negative zero in multiply is handled carefully. The old JIT
        previously took slow path if the result of a multiply was zero,
        which, without any changes, would cause the DFG to think that
        every such multiply produced a double result.
        
        This also fixes a bug in the old JIT's handling of decrements. It
        would take the slow path if the result was zero, but not if it
        underflowed.
        
        By itself, this would be a 1% slow-down on V8 and Kraken. But then
        I wrote optimizations in the DFG that take advantage of this new
        information. It's no longer the case that every multiply needs to
        do a check for negative zero; it only happens if the negative
        zero is ignored.
        
        This results in a 12% speed-up on v8-crypto, for a 1.4% geomean
        speed-up in V8. It's mostly neutral on Kraken. I can see an
        0.5% slow-down and it appears to be significant.

        * bytecode/CodeBlock.cpp:
        (JSC::CodeBlock::resetRareCaseProfiles):
        (JSC::CodeBlock::dumpValueProfiles):
        * bytecode/CodeBlock.h:
        * bytecode/ValueProfile.h:
        (JSC::RareCaseProfile::RareCaseProfile):
        (JSC::getRareCaseProfileBytecodeOffset):
        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::toInt32):
        (JSC::DFG::ByteCodeParser::makeSafe):
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGJITCodeGenerator.cpp:
        (JSC::DFG::GPRTemporary::GPRTemporary):
        * dfg/DFGJITCodeGenerator.h:
        * dfg/DFGNode.h:
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNode):
        (JSC::DFG::Propagator::fixupNode):
        (JSC::DFG::Propagator::clobbersWorld):
        (JSC::DFG::Propagator::performNodeCSE):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):
        (JSC::DFG::SpeculativeJIT::computeValueRecoveryFor):
        * jit/JIT.cpp:
        (JSC::JIT::privateCompileSlowCases):
        * jit/JIT.h:
        (JSC::JIT::linkDummySlowCase):
        * jit/JITArithmetic.cpp:
        (JSC::JIT::emit_op_post_dec):
        (JSC::JIT::emit_op_pre_dec):
        (JSC::JIT::compileBinaryArithOp):
        (JSC::JIT::emit_op_add):
        (JSC::JIT::emitSlow_op_add):
        * jit/JITInlineMethods.h:
        (JSC::JIT::addSlowCase):

1645 1646 1647 1648 1649 1650 1651
2011-09-19  Adam Roben  <aroben@apple.com>

        Windows build fix after r94575

        * JavaScriptCore.vcproj/JavaScriptCore.sln: Relinearized project dependencies. testRegExp
        now builds just before FindSafari.

1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671
2011-09-19  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r95466.
        http://trac.webkit.org/changeset/95466
        https://bugs.webkit.org/show_bug.cgi?id=68389

        Incorrect version of the patch. (Requested by mhahnenberg on
        #webkit).

        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
        * runtime/JSCell.cpp:
        (JSC::JSCell::toPrimitive):
        * runtime/JSCell.h:
        (JSC::JSCell::JSValue::toPrimitive):
        * runtime/JSNotAnObject.cpp:
        (JSC::JSNotAnObject::toPrimitive):
        * runtime/JSNotAnObject.h:
        * runtime/JSObject.h:
        * runtime/JSString.h:

1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699
2011-09-19  Mark Hahnenberg  <mhahnenberg@apple.com>

        Remove toPrimitive from JSCell
        https://bugs.webkit.org/show_bug.cgi?id=67875

        Reviewed by Geoffrey Garen.

        Part of the refactoring process to un-virtualize JSCell.  We move 
        all of the implicit functionality provided by the virtual toPrimitive method 
        in JSCell to be explicit in JSValue::toPrimitive and JSCell:toPrimitive while 
        also de-virtualizing JSCell::toPrimitive.

        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
        * runtime/JSCell.cpp:
        (JSC::JSCell::toPrimitive):
        * runtime/JSCell.h:

        We replace JSNotAnObject::toPrimitive with defaultValue, which it overrides from 
        JSObject.  This pushes the virtual method further down, enabling us to get rid 
        of the virtual call in JSCell.  Eventually we'll probably have to deal with this
        again, but we'll cross that bridge when we come to it.
        * runtime/JSNotAnObject.cpp:
        (JSC::JSNotAnObject::defaultValue):
        * runtime/JSNotAnObject.h:
        * runtime/JSObject.h:
        * runtime/JSString.h:
        (JSC::JSValue::toPrimitive):

oliver@apple.com's avatar
oliver@apple.com committed
1700 1701 1702 1703 1704 1705 1706
2011-09-19  Oliver Hunt  <oliver@apple.com>

        Build fix.

        * jit/JITPropertyAccess32_64.cpp:
        (JSC::JIT::compileGetDirectOffset):

1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741
2011-09-19  Oliver Hunt  <oliver@apple.com>

        Rename NewSpace.{h,cpp} to MarkedSpace.{h,cpp}
        https://bugs.webkit.org/show_bug.cgi?id=68376

        Reviewed by Gavin Barraclough.

        Renamed the the MarkedSpace files to match new name, and
        updated the relevant references.

        * CMakeLists.txt:
        * GNUmakefile.list.am:
        * JavaScriptCore.gypi:
        * JavaScriptCore.pro:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * heap/Heap.h:
        * heap/MarkedSpace.cpp: Renamed from Source/JavaScriptCore/heap/NewSpace.cpp.
        (JSC::MarkedSpace::MarkedSpace):
        (JSC::MarkedSpace::addBlock):
        (JSC::MarkedSpace::removeBlock):
        (JSC::MarkedSpace::resetAllocator):
        (JSC::MarkedSpace::canonicalizeBlocks):
        * heap/MarkedSpace.h: Renamed from Source/JavaScriptCore/heap/NewSpace.h.
        (JSC::MarkedSpace::waterMark):
        (JSC::MarkedSpace::highWaterMark):
        (JSC::MarkedSpace::setHighWaterMark):
        (JSC::MarkedSpace::sizeClassFor):
        (JSC::MarkedSpace::allocate):
        (JSC::MarkedSpace::forEachBlock):
        (JSC::MarkedSpace::SizeClass::SizeClass):
        (JSC::MarkedSpace::SizeClass::resetAllocator):
        (JSC::MarkedSpace::SizeClass::canonicalizeBlock):
        * runtime/JSCell.h:

1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789
2011-09-19  Oliver Hunt  <oliver@apple.com>

        Rename NewSpace to MarkedSpace
        https://bugs.webkit.org/show_bug.cgi?id=68375

        Reviewed by Gavin Barraclough.

        Rename NewSpace to a more accurate name, and update all uses.
        This patch doesn't rename the files themselves as that will
        just make the patch appear bigger than it is.

        * JavaScriptCore.exp:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
        * heap/Heap.cpp:
        (JSC::CountFunctor::TakeIfEmpty::TakeIfEmpty):
        (JSC::CountFunctor::TakeIfEmpty::operator()):
        (JSC::Heap::Heap):
        (JSC::Heap::reportExtraMemoryCostSlowCase):
        (JSC::Heap::tryAllocate):
        (JSC::Heap::allocateSlowCase):
        (JSC::Heap::collect):
        (JSC::Heap::canonicalizeBlocks):
        (JSC::Heap::resetAllocator):
        (JSC::Heap::isValidAllocation):
        (JSC::Heap::shrink):
        * heap/Heap.h:
        (JSC::Heap::markedSpace):
        (JSC::Heap::sizeClassFor):
        (JSC::Heap::allocate):
        * heap/NewSpace.cpp:
        (JSC::MarkedSpace::MarkedSpace):
        (JSC::MarkedSpace::addBlock):
        (JSC::MarkedSpace::removeBlock):
        (JSC::MarkedSpace::resetAllocator):
        (JSC::MarkedSpace::canonicalizeBlocks):
        * heap/NewSpace.h:
        (JSC::MarkedSpace::waterMark):
        (JSC::MarkedSpace::highWaterMark):
        (JSC::MarkedSpace::setHighWaterMark):
        (JSC::MarkedSpace::sizeClassFor):
        (JSC::MarkedSpace::allocate):
        (JSC::MarkedSpace::forEachBlock):
        (JSC::MarkedSpace::SizeClass::SizeClass):
        (JSC::MarkedSpace::SizeClass::resetAllocator):
        (JSC::MarkedSpace::SizeClass::canonicalizeBlock):
        * jit/JITInlineMethods.h:
        (JSC::JIT::emitAllocateBasicJSObject):

1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812
2011-09-19  Peter Rybin  <peter.rybin@gmail.com>

        TextPosition refactoring: Merge ZeroBasedNumber and OneBasedNumber classes
        https://bugs.webkit.org/show_bug.cgi?id=63541

        Reviewed by Adam Barth.

        * parser/SourceProvider.h:
        (JSC::SourceProvider::startPosition):
        * wtf/text/TextPosition.h:
        (WTF::OrdinalNumber::fromZeroBasedInt):
        (WTF::OrdinalNumber::fromOneBasedInt):
        (WTF::OrdinalNumber::OrdinalNumber):
        (WTF::OrdinalNumber::zeroBasedInt):
        (WTF::OrdinalNumber::oneBasedInt):
        (WTF::OrdinalNumber::operator==):
        (WTF::OrdinalNumber::operator!=):
        (WTF::OrdinalNumber::first):
        (WTF::OrdinalNumber::beforeFirst):
        (WTF::TextPosition::TextPosition):
        (WTF::TextPosition::minimumPosition):
        (WTF::TextPosition::belowRangePosition):

1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825
2011-09-19  Dan Bernstein  <mitz@apple.com>

        JavaScriptCore part of [mac] WebKit contains Objective-C classes that are not prefixed with its standard prefixes
        https://bugs.webkit.org/show_bug.cgi?id=68323

        Reviewed by Sam Weinig.

        Renamed WTFMainThreadCaller to JSWTFMainThreadCaller.

        * wtf/mac/MainThreadMac.mm:
        (WTF::initializeMainThreadPlatform):
        (WTF::initializeMainThreadToProcessMainThreadPlatform):

1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845
2011-09-19  Oliver Hunt  <oliver@apple.com>

        Remove direct property slot pointers from the instruction stream
        https://bugs.webkit.org/show_bug.cgi?id=68373

        Reviewed by Gavin Barraclough.

        Use an indirect load to access prototype properties rather than directly
        storing the property address in the instruction stream.  This should allow
        further optimisations in future, and also provides a 0.5% win to sunspider.

        * dfg/DFGRepatch.cpp:
        (JSC::DFG::generateProtoChainAccessStub):
        * jit/JITPropertyAccess.cpp:
        (JSC::JIT::compileGetDirectOffset):
        * jit/JITPropertyAccess32_64.cpp:
        (JSC::JIT::compileGetDirectOffset):
        * runtime/JSObject.h:
        (JSC::JSObject::addressOfPropertyStorage):

oliver@apple.com's avatar
oliver@apple.com committed
1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872
2011-09-19  Oliver Hunt  <oliver@apple.com>

        Remove bump allocator
        https://bugs.webkit.org/show_bug.cgi?id=68370

        Reviewed by Sam Weinig.

        Can't do anything with this allocator currently, and it's
        increasing the complexity of the GC code.  Slight progression
        on SunSpider, slight regression (undoing the original progression)
        in V8.

        * heap/Heap.cpp:
        (JSC::Heap::collect):
        * heap/Heap.h:
        * heap/NewSpace.cpp:
        (JSC::NewSpace::NewSpace):
        * heap/NewSpace.h:
        (JSC::NewSpace::allocate):
        * runtime/JSObject.cpp:
        (JSC::JSObject::allocatePropertyStorage):
        * runtime/JSObject.h:
        (JSC::JSObject::~JSObject):
        (JSC::JSObject::visitChildrenDirect):
        * runtime/StorageBarrier.h:
        (JSC::StorageBarrier::set):

1873 1874 1875 1876 1877 1878 1879 1880 1881
2011-09-19  Carlos Garcia Campos  <cgarcia@igalia.com>

        [GTK] Fix distcheck build
        https://bugs.webkit.org/show_bug.cgi?id=68346

        Reviewed by Philippe Normand.

        * GNUmakefile.list.am:

1882 1883 1884 1885 1886 1887 1888 1889 1890
2011-09-19  Carlos Garcia Campos  <cgarcia@igalia.com>

        [GTK] Fix distcheck build
        https://bugs.webkit.org/show_bug.cgi?id=68241

        Reviewed by Martin Robinson.

        * GNUmakefile.list.am:

mitz@apple.com's avatar
mitz@apple.com committed
1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904
2011-09-18  Dan Bernstein  <mitz@apple.com>

        Removed ProfilerServer.

        Reviewed by Mark Rowe.

        * JavaScriptCore.gypi:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * profiler/ProfilerServer.h: Removed.
        * profiler/ProfilerServer.mm: Removed.
        * runtime/JSGlobalData.cpp:
        (JSC::JSGlobalData::JSGlobalData):
        * wscript:

1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933
2011-09-17  Filip Pizlo  <fpizlo@apple.com>

        DFG JIT should inline Math.min, Math.max, and Math.sqrt
        https://bugs.webkit.org/show_bug.cgi?id=68318

        Reviewed by Gavin Barraclough.
        
        Adds Math.min, Math.max, and Math.sqrt intrinsics. Adds support for
        a function to have an intrinsic but not a thunk generator. This is
        a 7% speed-up on access-nbody, and neutral elsewhere, mainly because
        we're still not DFG compiling the bulk of the hot code in Kraken audio
        benchmarks.

        * create_hash_table:
        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::handleMinMax):
        (JSC::DFG::ByteCodeParser::handleIntrinsic):
        * dfg/DFGIntrinsic.h:
        * dfg/DFGNode.h:
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNode):
        (JSC::DFG::Propagator::fixupNode):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):
        * jit/JITStubs.cpp:
        (JSC::JITThunks::hostFunctionStub):
        * runtime/Lookup.cpp:
        (JSC::setUpStaticFunctionSlot):

1934 1935 1936 1937 1938 1939 1940 1941 1942
2011-09-18  Nico Weber  <thakis@chromium.org>

        Remove two files from JavaScriptCore.gypi that were removed in r95240
        https://bugs.webkit.org/show_bug.cgi?id=68327

        Unreviewed, build warning fix.

        * JavaScriptCore.gypi:

1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968
2011-09-17  Oliver Hunt  <oliver@apple.com>

        Remove special case handling of inline storage from the JIT
        https://bugs.webkit.org/show_bug.cgi?id=68319

        Reviewed by Gavin Barraclough.

        Simplify logic used for reading and writing to property storage
        by removing the special cases for inline storage.  This has no
        perf impact.

        * dfg/DFGRepatch.cpp:
        (JSC::DFG::generateProtoChainAccessStub):
        (JSC::DFG::tryBuildGetByIDList):
        * jit/JIT.h:
        * jit/JITPropertyAccess.cpp:
        (JSC::JIT::compilePutDirectOffset):
        (JSC::JIT::compileGetDirectOffset):
        (JSC::JIT::privateCompilePutByIdTransition):
        (JSC::JIT::privateCompileGetByIdSelfList):
        * jit/JITPropertyAccess32_64.cpp:
        (JSC::JIT::compilePutDirectOffset):
        (JSC::JIT::compileGetDirectOffset):
        (JSC::JIT::privateCompilePutByIdTransition):
        (JSC::JIT::privateCompileGetByIdSelfList):

1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
2011-09-17  Filip Pizlo  <fpizlo@apple.com>

        DFG JIT does not have full block-local CSE
        https://bugs.webkit.org/show_bug.cgi?id=68316

        Reviewed by Oliver Hunt.
        
        This adds block-local CSE to the DFG. CSE runs in the propagator just after
        type propagation. It is part of the propagator itself because it needs to
        use the propagator's internal data structures to determine which operations
        may have side effects. Because it changes the live-ranges of nodes, the
        virtual register allocator had to be moved into the propagator so that it
        runs after CSE. To ensure that the back-end knows to keep the inputs to
        any eliminated node alive for OSR, a new node type, Phantom, was introduced.
        It is a no-op but prolonges the live-range of its inputs.
        
        This is an 80% speed-up on imaging-gaussian-blur, and a 10% speed-up on
        Kraken.
        
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * dfg/DFGAliasTracker.h: Removed.
        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::parseBlock):
        (JSC::DFG::ByteCodeParser::parse):
        * dfg/DFGGraph.cpp:
        (JSC::DFG::Graph::dump):
        * dfg/DFGGraph.h:
        (JSC::DFG::MethodCheckData::operator==):
        (JSC::DFG::MethodCheckData::operator!=):
        * dfg/DFGNode.h:
        (JSC::DFG::Node::hasVirtualRegister):
        (JSC::DFG::Node::setRefCount):
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::Propagator):
        (JSC::DFG::Propagator::fixpoint):
        (JSC::DFG::Propagator::propagateNode):
        (JSC::DFG::Propagator::canonicalize):
        (JSC::DFG::Propagator::computeStartIndex):
        (JSC::DFG::Propagator::startIndex):
        (JSC::DFG::Propagator::pureCSE):
        (JSC::DFG::Propagator::globalVarLoadElimination):
        (JSC::DFG::Propagator::getByValLoadElimination):
        (JSC::DFG::Propagator::getMethodLoadElimination):
        (JSC::DFG::Propagator::performSubstitution):
        (JSC::DFG::Propagator::setReplacement):
        (JSC::DFG::Propagator::performNodeCSE):
        (JSC::DFG::Propagator::performBlockCSE):
        (JSC::DFG::Propagator::localCSE):
        (JSC::DFG::Propagator::allocateVirtualRegisters):
        (JSC::DFG::propagate):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):

2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044
2011-09-16  Filip Pizlo  <fpizlo@apple.com>

        method_check should repatch itself if it finds that the new structure(s)
        are the result of transitions from the old structure(s)
        https://bugs.webkit.org/show_bug.cgi?id=68294

        Reviewed by Gavin Barraclough.
        
        Previously a patched method_check would slow-path to get_by_id. Now it
        slow-paths to method_check_update, which attempts to correct the
        method_check due to structure transitions before bailing to get_by_id.
        
        This is a 1-2% speed-up on some benchmarks and is not a slow-down
        anywhere, leading to a 0.6% speed-up on the Kraken geomean.

        * jit/JITPropertyAccess.cpp:
        (JSC::JIT::patchMethodCallProto):
        * jit/JITStubs.cpp:
        (JSC::DEFINE_STUB_FUNCTION):
        * jit/JITStubs.h:
        * runtime/Structure.h:
        (JSC::Structure::transitivelyTransitionedFrom):

2045 2046 2047 2048 2049 2050
2011-09-16  Ryosuke Niwa  <rniwa@webkit.org>

        Touch Platform.h in the hope to fix SnowLeopard Intel Release (WebKit2 Tests).

        * wtf/Platform.h:

2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068
2011-09-16  Sam Weinig  <sam@webkit.org>

        Rename APIValueWrapper type to APIValueWrapperType for consistency
        https://bugs.webkit.org/show_bug.cgi?id=68306

        Reviewed by Anders Carlsson.

        * runtime/JSAPIValueWrapper.h:
        (JSC::JSAPIValueWrapper::createStructure):
        Update name.

        * runtime/JSType.h:
        Update name and un-indent.

        * runtime/Structure.h:
        (JSC::JSCell::isAPIValueWrapper):
        Update name.

2069 2070 2071 2072 2073 2074 2075 2076 2077 2078
2011-09-16  Sam Weinig  <sam@webkit.org>

        Remove unused isStrictModeFunction function
        https://bugs.webkit.org/show_bug.cgi?id=68305

        Reviewed by Anders Carlsson.

        * runtime/JSObject.h:
        (JSC::JSObject::isStrictModeFunction):

weinig@apple.com's avatar
weinig@apple.com committed
2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118
2011-09-16  Sam Weinig  <sam@webkit.org>

        Cleanup JSTypeInfo a bit
        https://bugs.webkit.org/show_bug.cgi?id=68289

        Reviewed by Anders Carlsson.

        * dfg/DFGOperations.cpp:
        * jit/JITStubs.cpp:
        (JSC::DEFINE_STUB_FUNCTION):
        Replace direct access to flags() with predicate.

        * runtime/JSObject.h:
        (JSC::JSFinalObject::createStructure):
        Pass FinalObjectType instead of using special IsJSFinalObject.

        * runtime/JSTypeInfo.h:
        (JSC::TypeInfo::TypeInfo):
        Add additional assert that you should no object should OverridesHasInstance but not have ImplementsHasInstance set.

        (JSC::TypeInfo::isFinalObject):
        Added.

        (JSC::TypeInfo::masqueradesAsUndefined):
        (JSC::TypeInfo::implementsHasInstance):
        (JSC::TypeInfo::isEnvironmentRecord):
        (JSC::TypeInfo::overridesHasInstance):
        (JSC::TypeInfo::implementsDefaultHasInstance):
        (JSC::TypeInfo::overridesGetOwnPropertySlot):
        (JSC::TypeInfo::overridesVisitChildren):
        (JSC::TypeInfo::overridesGetPropertyNames):
        (JSC::TypeInfo::prohibitsPropertyCaching):
        (JSC::TypeInfo::isSetOnFlags1):
        (JSC::TypeInfo::isSetOnFlags2):
        Replace direct bit twiddling with helper functions.

        * runtime/Structure.cpp:
        (JSC::Structure::Structure):
        Use new isFinalObject() predicate.

2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135
2011-09-16  Gavin Barraclough  <barraclough@apple.com>

        Unsigned bit shift fails under certain conditions in 32 bit builds
        https://bugs.webkit.org/show_bug.cgi?id=68166

        Reviewed by Geoff Garen.

        The major bug here is that the slow case (which handles shifts of
        doubles) doesn't check for negative results from an unsigned shift
        (which should be unsigned, and as such can't be represented by a
        signed integer immediate).  The implementation is also flawed for
        shifts by negative shift amounts (treats as shift by zero).

        * jit/JITArithmetic32_64.cpp:
        (JSC::JIT::emitRightShift):
        (JSC::JIT::emitRightShiftSlowCase):

2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153
2011-09-16  Geoffrey Garen  <ggaren@apple.com>

        Removed undetectable style.filter.

        Reviewed by Sam Weinig.
        
        This feature was added in http://trac.webkit.org/changeset/15557 to
        support housingmaps.com. But housingmaps.com no longer needs this hack,
        we don't know of other websites that need it, and we don't know of
        any other browsers that have implemented this feature.

        * GNUmakefile.list.am:
        * JavaScriptCore.gypi:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * runtime/JSTypeInfo.h:
        * runtime/StringObjectThatMasqueradesAsUndefined.h: Removed.

2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185
2011-09-15  Sam Weinig  <sam@webkit.org>

        Prepare JSTypes for more Object subtypes
        https://bugs.webkit.org/show_bug.cgi?id=68200

        Reviewed by Gavin Barraclough.

        * dfg/DFGJITCompiler.h:
        (JSC::DFG::JITCompiler::branchIfNotObject):
        * jit/JITInlineMethods.h:
        (JSC::JIT::emitJumpIfNotObject):
        * runtime/JSGlobalObject.h:
        (JSC::Structure::prototypeForLookup):
        * runtime/JSObject.h:
        (JSC::JSObject::finishCreation):
        * runtime/JSType.h:
        * runtime/JSTypeInfo.h:
        (JSC::TypeInfo::type):
        (JSC::TypeInfo::isObject):
        (JSC::TypeInfo::isFinal):
        (JSC::TypeInfo::prohibitsPropertyCaching):
        * runtime/NativeErrorConstructor.h:
        (JSC::NativeErrorConstructor::finishCreation):
        * runtime/Operations.cpp:
        (JSC::jsIsObjectType):
        * runtime/Structure.cpp:
        (JSC::Structure::addPropertyTransitionToExistingStructure):
        (JSC::Structure::addPropertyTransition):
        * runtime/Structure.h:
        (JSC::Structure::isObject):
        (JSC::JSCell::isObject):

2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209
2011-09-16  Geoffrey Garen  <ggaren@apple.com>

        Rolled back in r95201 with test failure fixed.
        
        I missed two cases of jumpSlowToHot in rshift -- these cases need to be
        sure to initialize regT1 to the int tag, since it will otherwise hold
        the top 32 bits of a double.

        * jit/JIT.h:
        * jit/JITArithmetic32_64.cpp:
        (JSC::JIT::emit_op_lshift):
        (JSC::JIT::emitRightShift):
        (JSC::JIT::emitRightShiftSlowCase):
        (JSC::JIT::emit_op_bitand):
        (JSC::JIT::emit_op_bitor):
        (JSC::JIT::emit_op_bitxor):
        (JSC::JIT::emit_op_bitnot):
        (JSC::JIT::emit_op_post_inc):
        (JSC::JIT::emit_op_post_dec):
        (JSC::JIT::emit_op_pre_inc):
        (JSC::JIT::emit_op_pre_dec):
        * jit/JITInlineMethods.h:
        (JSC::JIT::emitStoreAndMapInt32):

2210 2211 2212 2213 2214 2215
2011-09-16  Filip Pizlo  <fpizlo@apple.com>

        Unreviewed Windows build fix after 95318.

        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:

2216 2217 2218 2219 2220 2221 2222
2011-09-16  Adam Roben  <aroben@apple.com>

        Windows build fix after r95310

        * JavaScriptCore.vcproj/jsc/jscCommon.vsprops: Added include\private\JavaScriptCore to the
        include path so DFGIntrinsic.h can be found.

2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254
2011-09-16  Gavin Barraclough  <barraclough@apple.com>

        Rationalize JSObject::putDirect* methods
        https://bugs.webkit.org/show_bug.cgi?id=68274

        Reviewed by Sam Weinig.
        
        Delete the *Function variants. These are overall inefficient,
        in the way they get the name back from the function rather
        than just passing it in.

        * JavaScriptCore.exp:
        * jsc.cpp:
        (GlobalObject::finishCreation):
        (GlobalObject::addFunction):
        * runtime/FunctionPrototype.cpp:
        (JSC::FunctionPrototype::addFunctionProperties):
        * runtime/JSGlobalObject.cpp:
        (JSC::JSGlobalObject::reset):
        * runtime/JSObject.cpp:
        (JSC::JSObject::put):
        (JSC::JSObject::putWithAttributes):
        (JSC::JSObject::defineGetter):
        (JSC::JSObject::defineSetter):
        * runtime/JSObject.h:
        (JSC::JSObject::putDirect):
        (JSC::JSObject::putDirectWithoutTransition):
        * runtime/Lookup.cpp:
        (JSC::setUpStaticFunctionSlot):
        * runtime/Lookup.h:
        (JSC::lookupPut):

2255 2256 2257 2258 2259 2260
2011-09-16  Filip Pizlo  <fpizlo@apple.com>

        Unreviewed build fix for Windows.

        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:

2261 2262 2263 2264 2265 2266 2267
2011-09-16  Filip Pizlo  <fpizlo@apple.com>

        Unreviewed build fix for non-DFG builds.

        * runtime/Executable.h:
        (JSC::NativeExecutable::finishCreation):

2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329
2011-09-16  Filip Pizlo  <fpizlo@apple.com>

        DFG JIT should inline Math.abs
        https://bugs.webkit.org/show_bug.cgi?id=68227

        Reviewed by Oliver Hunt.
        
        This adds the ability to track intrinsic functions throughout the
        host function infrastructure, so that the DFG can easily query
        whether or not a call's target is intrinsic, and if so, which
        intrinsic it is.
        
        On top of this, it adds Math.abs intrinsics to DFG. Call(Math.abs)
        is transformed into ValueToNumber<-ArithAbs nodes. These nodes
        then get optimized using the usual tricks.
        
        Also had to make a completely unrelated change to
        DateInstanceCache.h in order to fix a preexisting alphabetical
        sorting problem in JSGlobalData.h
        
        This results in a big win in imaging-gaussian-blur: 61% faster
        than before. The net win on Kraken is around 13%.

        * JavaScriptCore.xcodeproj/project.pbxproj:
        * create_hash_table:
        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGGraph.h:
        (JSC::DFG::Graph::isFunctionConstant):
        (JSC::DFG::Graph::valueOfFunctionConstant):
        * dfg/DFGIntrinsic.h: Added.
        * dfg/DFGJITCodeGenerator.h:
        (JSC::DFG::JITCodeGenerator::isFunctionConstant):
        (JSC::DFG::JITCodeGenerator::valueOfFunctionConstant):
        * dfg/DFGJITCompiler.h:
        (JSC::DFG::JITCompiler::isFunctionConstant):
        (JSC::DFG::JITCompiler::valueOfFunctionConstant):
        * dfg/DFGNode.h:
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNode):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):
        * jit/JITStubs.cpp:
        (JSC::JITThunks::hostFunctionStub):
        * jit/JITStubs.h:
        * runtime/DateInstanceCache.h:
        * runtime/Executable.cpp:
        (JSC::ExecutableBase::intrinsic):
        (JSC::NativeExecutable::intrinsic):
        * runtime/Executable.h:
        (JSC::NativeExecutable::create):
        (JSC::NativeExecutable::finishCreation):
        * runtime/JSGlobalData.cpp:
        (JSC::JSGlobalData::getHostFunction):
        * runtime/JSGlobalData.h:
        * runtime/Lookup.cpp:
        (JSC::HashTable::createTable):
        (JSC::setUpStaticFunctionSlot):
        * runtime/Lookup.h:
        (JSC::HashEntry::initialize):
        (JSC::HashEntry::intrinsic):

2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346
2011-09-16  Filip Pizlo  <fpizlo@apple.com>

        REGRESSION: Reproducible crash below SlotVisitor::harvestWeakReferences
        using Domino's online ordering
        https://bugs.webkit.org/show_bug.cgi?id=68220

        Reviewed by Oliver Hunt.
        
        Weak handle processing can result in new objects being marked, which
        results in new WeakReferencesHarvesters being added. But weak
        reference harvesters are only processed before weak handle processing,
        so there's the risk that a weak reference harvester will persist
        until the next collection, by which time it may have been deleted.

        * heap/Heap.cpp:
        (JSC::Heap::markRoots):

2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367
2011-09-16  Csaba Osztrogonác  <ossy@webkit.org>

        REGRESSION(r95201): It made two tests fail
        https://bugs.webkit.org/show_bug.cgi?id=68230

        Unreviewed rolling out r95201.

        * jit/JIT.h:
        * jit/JITArithmetic32_64.cpp:
        (JSC::JIT::emit_op_lshift):
        (JSC::JIT::emitRightShift):
        (JSC::JIT::emit_op_bitand):
        (JSC::JIT::emit_op_bitor):
        (JSC::JIT::emit_op_bitxor):
        (JSC::JIT::emit_op_bitnot):
        (JSC::JIT::emit_op_post_inc):
        (JSC::JIT::emit_op_post_dec):
        (JSC::JIT::emit_op_pre_inc):
        (JSC::JIT::emit_op_pre_dec):
        * jit/JITInlineMethods.h:

2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459
2011-09-15  Filip Pizlo  <fpizlo@apple.com>

        DFG JIT does not optimize method_check
        https://bugs.webkit.org/show_bug.cgi?id=68215

        Reviewed by Oliver Hunt.
        
        MethodCallLinkInfo and StructureStubInfo are now searchable by
        bytecodeIndex, so that DFG::ByteCodeParser can use that information
        to determine how to optimize GetMethod.
        
        A new node op has been added to DFG: CheckMethod. This is a variant
        of GetMethod that has been optimized for the case that GetMethod
        always takes the fast path. CheckMethod results in only a very
        small amount of code (two loads and two branches in the worst case,
        one load and one branch in the best case). CheckMethod behaves as
        if it were a constant.  
        
        Introduced the notion that a DFG node that is not JSConstant
        behaves as a constant. CheckMethod uses this functionality.
        
        This is a 3% speed-up on Kraken, and a small speed-up on V8.
        Appears to be neutral on SunSpider.

        * bytecode/CodeBlock.h:
        (JSC::getStructureStubInfoBytecodeIndex):
        (JSC::getMethodCallLinkInfoBytecodeIndex):
        * bytecode/PredictedType.cpp:
        (JSC::predictionFromCell):
        (JSC::predictionFromValue):
        * bytecode/PredictedType.h:
        * bytecode/StructureStubInfo.h:
        * dfg/DFGAliasTracker.h:
        (JSC::DFG::AliasTracker::recordGetMethod):
        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGGraph.cpp:
        (JSC::DFG::Graph::dump):
        * dfg/DFGGraph.h:
        (JSC::DFG::Graph::getMethodCheckPrediction):
        (JSC::DFG::Graph::getPrediction):
        (JSC::DFG::Graph::isConstant):
        (JSC::DFG::Graph::isJSConstant):
        (JSC::DFG::Graph::valueOfJSConstant):
        (JSC::DFG::Graph::valueOfInt32Constant):
        (JSC::DFG::Graph::valueOfNumberConstant):
        (JSC::DFG::Graph::valueOfBooleanConstant):
        (JSC::DFG::Graph::valueOfJSConstantNode):
        * dfg/DFGJITCodeGenerator.cpp:
        (JSC::DFG::JITCodeGenerator::fillInteger):
        (JSC::DFG::JITCodeGenerator::fillDouble):
        (JSC::DFG::JITCodeGenerator::fillJSValue):
        (JSC::DFG::JITCodeGenerator::isKnownNotInteger):
        (JSC::DFG::JITCodeGenerator::isKnownNotNumber):
        * dfg/DFGJITCodeGenerator.h:
        (JSC::DFG::JITCodeGenerator::silentSpillFPR):
        (JSC::DFG::JITCodeGenerator::silentFillGPR):
        (JSC::DFG::JITCodeGenerator::silentFillFPR):
        * dfg/DFGJITCompiler.cpp:
        (JSC::DFG::JITCompiler::fillNumericToDouble):
        (JSC::DFG::JITCompiler::fillInt32ToInteger):
        (JSC::DFG::JITCompiler::fillToJS):
        * dfg/DFGNode.h:
        (JSC::DFG::Node::hasConstant):
        (JSC::DFG::Node::hasIdentifier):
        (JSC::DFG::Node::hasMethodCheckData):
        (JSC::DFG::Node::methodCheckDataIndex):
        (JSC::DFG::Node::valueOfJSConstant):
        * dfg/DFGPropagator.cpp:
        (JSC::DFG::Propagator::propagateNode):
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::fillSpeculateIntInternal):
        (JSC::DFG::SpeculativeJIT::fillSpeculateDouble):
        (JSC::DFG::SpeculativeJIT::fillSpeculateCell):
        (JSC::DFG::SpeculativeJIT::fillSpeculateBoolean):
        (JSC::DFG::SpeculativeJIT::compile):
        * jit/JIT.cpp:
        (JSC::JIT::privateCompile):
        * jit/JIT.h:
        (JSC::PropertyStubCompilationInfo::PropertyStubCompilationInfo):
        (JSC::MethodCallCompilationInfo::MethodCallCompilationInfo):
        * jit/JITPropertyAccess.cpp:
        (JSC::JIT::emit_op_method_check):
        (JSC::JIT::compileGetByIdHotPath):
        (JSC::JIT::emit_op_put_by_id):
        * jit/JITPropertyAccess32_64.cpp:
        (JSC::JIT::emit_op_method_check):
        (JSC::JIT::compileGetByIdHotPath):
        (JSC::JIT::emit_op_put_by_id):
        * runtime/JSCell.h:
        (JSC::JSCell::JSCell::structureAddress):

2460 2461 2462 2463 2464 2465 2466 2467 2468 2469
2011-09-15  Adam Barth  <abarth@webkit.org>

        Rename ENABLE(DATABASE) to ENABLE(SQL_DATABASE)
        https://bugs.webkit.org/show_bug.cgi?id=68205

        Reviewed by Eric Seidel.

        * Configurations/FeatureDefines.xcconfig:
        * wtf/Platform.h:

2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489
2011-09-15  Mark Hahnenberg  <mhahnenberg@apple.com>

        Unzip initialization lists and constructors in JSCell hierarchy (7/7)
        https://bugs.webkit.org/show_bug.cgi?id=68122

        Reviewed by Geoffrey Garen.

        Completed the seventh and final level of the refactoring to add finishCreation() 
        methods to all classes within the JSCell hierarchy with non-trivial 
        constructor bodies.

        JSCallbackObject was missed in previous patches due to the fact that 
        it's non-obvious (at least to my script) that it is in the JSCell hierarchy, so 
        this is just a bit of retroactive cleanup.

        * API/JSCallbackObject.h:
        (JSC::JSCallbackObject::create):
        * API/JSCallbackObjectFunctions.h:
        (JSC::::JSCallbackObject):

2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547
2011-09-15  Filip Pizlo  <fpizlo@apple.com>

        The DFG non-speculative JIT is no longer used and should be removed.
        https://bugs.webkit.org/show_bug.cgi?id=68177

        Reviewed by Geoffrey Garen.
        
        This removes the non-speculative JIT and everything that relied on it,
        including the ability to turn on DFG but not tiered compilation the,
        ability to perform speculation failure into non-speculative JIT code,
        and the ability to statically terminate speculation.

        * GNUmakefile.list.am:
        * JavaScriptCore.pro:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * bytecode/CodeBlock.h:
        * bytecompiler/BytecodeGenerator.cpp:
        (JSC::BytecodeGenerator::emitLoopHint):
        * dfg/DFGByteCodeParser.cpp:
        (JSC::DFG::ByteCodeParser::ByteCodeParser):
        (JSC::DFG::ByteCodeParser::getStrongPrediction):
        (JSC::DFG::ByteCodeParser::parseBlock):
        * dfg/DFGDriver.cpp:
        (JSC::DFG::compile):
        * dfg/DFGGenerationInfo.h:
        * dfg/DFGGraph.cpp:
        (JSC::DFG::Graph::predictArgumentTypes):
        * dfg/DFGJITCodeGenerator.cpp:
        * dfg/DFGJITCompiler.cpp:
        (JSC::DFG::JITCompiler::linkOSRExits):
        (JSC::DFG::JITCompiler::compileBody):
        * dfg/DFGJITCompiler.h:
        * dfg/DFGNode.h:
        * dfg/DFGNonSpeculativeJIT.cpp: Removed.
        * dfg/DFGNonSpeculativeJIT.h: Removed.
        * dfg/DFGOSREntry.cpp:
        (JSC::DFG::prepareOSREntry):
        * dfg/DFGPropagator.cpp:
        * dfg/DFGPropagator.h:
        * dfg/DFGSpeculativeJIT.cpp:
        (JSC::DFG::SpeculativeJIT::compile):
        * dfg/DFGSpeculativeJIT.h:
        (JSC::DFG::SpeculativeJIT::osrExits):
        (JSC::DFG::SpeculativeJIT::speculationRecovery):
        (JSC::DFG::SpeculativeJIT::speculationCheck):
        (JSC::DFG::SpeculativeJIT::terminateSpeculativeExecution):
        * jit/JIT.cpp:
        (JSC::JIT::privateCompileMainPass):
        (JSC::JIT::privateCompile):
        * jit/JIT.h:
        * jit/JITCode.h:
        (JSC::JITCode::bottomTierJIT):
        * runtime/JSGlobalData.cpp:
        (JSC::JSGlobalData::JSGlobalData):
        (JSC::JSGlobalData::~JSGlobalData):
        * runtime/JSGlobalData.h:
        * wtf/Platform.h:

2548 2549 2550 2551 2552 2553 2554 2555 2556
2011-09-15  Eric Seidel  <eric@webkit.org>

        Remove ENABLE(SVG_AS_IMAGE) since all major ports have it on by default
        https://bugs.webkit.org/show_bug.cgi?id=68182

        Reviewed by Adam Barth.

        * Configurations/FeatureDefines.xcconfig:

2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569
2011-09-15  Filip Pizlo  <fpizlo@apple.com>

        DFG speculative JIT sometimes asserts that a value is not a number
        even when it doesn't know anything about the number
        https://bugs.webkit.org/show_bug.cgi?id=68189

        Reviewed by Oliver Hunt.

        * dfg/DFGGenerationInfo.h:
        (JSC::DFG::GenerationInfo::isUnknownJS):
        * dfg/DFGJITCodeGenerator.cpp:
        (JSC::DFG::JITCodeGenerator::isKnownNotNumber):

2570 2571 2572 2573 2574