ChangeLog

   1 2013-07-18  Andreas Kling  <akling@apple.com>
   2
   3         CodeBlock DFG entry list isn't getting shrunk-to-fit after linking.
   4         <http://webkit.org/b/118875>
   5         <rdar://problem/14488577>
   6
   7         Reviewed by Geoffrey Garen.
   8
   9         Move the CodeBlock::shrinkToFit() call out of JITCompiler::link() and to the call sites
  10         so SpeculativeJIT::linkOSREntries() can fill in CodeBlock::m_dfgData->osrEntry first.
  11
  12         886 kB progression on <http://twitter.com/awesomekling>
  13
  14         * dfg/DFGJITCompiler.cpp:
  15         (JSC::DFG::JITCompiler::link):
  16         (JSC::DFG::JITCompiler::compile):
  17         (JSC::DFG::JITCompiler::compileFunction):
  18
  19 2013-07-18  Andreas Kling  <akling@apple.com>
  20
  21         CodeBlock::m_argumentValueProfiles wastes a lot of memory.
  22         <http://webkit.org/b/118852>
  23         <rdar://problem/14481659>
  24
  25         Reviewed by Anders Carlsson.
  26
  27         Use Vector::resizeToFit() for CodeBlock::m_argumentValueProfiles. We don't need any padding
  28         for growth, since we won't be appending to it anyway.
  29
  30         921 KB progression on <http://twitter.com/awesomekling>
  31
  32         * bytecode/CodeBlock.cpp:
  33         (JSC::CodeBlock::setNumParameters):
  34
  35 2013-07-16  Mark Hahnenberg  <mhahnenberg@apple.com>
  36
  37         Remove reference to JSValueStructSupport.h from JSExport.h
  38         https://bugs.webkit.org/show_bug.cgi?id=118746
  39
  40         Reviewed by Filip Pizlo.
  41
  42         * API/JSExport.h: No such header exists, so it doesn't make sense to reference it.
  43
  44 2013-07-17  Lucas Forschler  <lforschler@apple.com>
  45
  46         Rollout r152600
  47
  48 2013-07-17  Lucas Forschler  <lforschler@apple.com>
  49
  50         Merge r152600
  51
  52     2013-07-12  Brent Fulgham  <bfulgham@apple.com>
  53
  54             [Windows] Build correction after r152573/r152577.
  55             https://bugs.webkit.org/show_bug.cgi?id=118610
  56
  57             Reviewed by Oliver Hunt.
  58
  59             * jit/JITThunks.cpp:
  60             (JSC::JITThunks::hostFunctionStub): Hand-feed MSVC++ the fact that we want the second
  61             argument of the make_pair to be a function pointer.
  62
  63 2013-07-17  Lucas Forschler  <lforschler@apple.com>
  64
  65         Merge r152742
  66
  67     2013-07-16  Filip Pizlo  <fpizlo@apple.com>
  68
  69             MakeRope fixup shouldn't lead to an Identity without kids
  70             https://bugs.webkit.org/show_bug.cgi?id=118745
  71
  72             Reviewed by Mark Hahnenberg.
  73
  74             Make the empty string pruning part of fixupMakeRope() stop if it's on the last child.
  75
  76             Make Node::convertToIdentity release-assert that it has exactly one kid.
  77
  78             * dfg/DFGFixupPhase.cpp:
  79             (JSC::DFG::FixupPhase::fixupMakeRope):
  80             * dfg/DFGNode.h:
  81             (JSC::DFG::Node::convertToIdentity):
  82
  83 2013-07-15  Lucas Forschler  <lforschler@apple.com>
  84
  85         Merge r152613
  86
  87     2013-07-13  Commit Queue  <commit-queue@webkit.org>
  88
  89             Unreviewed, rolling out r151978.
  90             http://trac.webkit.org/changeset/151978
  91             https://bugs.webkit.org/show_bug.cgi?id=118651
  92
  93             Caused regressions at least 3 websites (Requested by rniwa on
  94             #webkit).
  95
  96             * runtime/JSCJSValue.h:
  97             * runtime/JSString.h:
  98
  99 2013-07-15  Lucas Forschler  <lforschler@apple.com>
 100
 101         Merge r152577
 102
 103     2013-07-11  Oliver Hunt  <oliver@apple.com>
 104
 105             Attempt to fix the windows build.
 106
 107             * jit/JITThunks.cpp:
 108             (JSC::JITThunks::hostFunctionStub):
 109             * jit/JITThunks.h:
 110
 111 2013-07-15  Lucas Forschler  <lforschler@apple.com>
 112
 113         Merge r152573
 114
 115     2013-07-10  Oliver Hunt  <oliver@apple.com>
 116
 117             NativeExecutable cache needs to use both call and construct functions for key
 118             https://bugs.webkit.org/show_bug.cgi?id=118545
 119
 120             Reviewed by Geoffrey Garen.
 121
 122             Make the native executable cache make use a key pair so we don't decide to
 123             treat all subsequent functions as not being constructors.
 124
 125             * jit/JITThunks.cpp:
 126             (JSC::JITThunks::hostFunctionStub):
 127             * jit/JITThunks.h:
 128             * runtime/JSBoundFunction.cpp:
 129             (JSC::JSBoundFunction::create):
 130             * runtime/JSCell.cpp:
 131             (JSC::JSCell::getCallData):
 132             (JSC::JSCell::getConstructData):
 133
 134 2013-07-09  Lucas Forschler  <lforschler@apple.com>
 135
 136         Merge r152497
 137
 138     2013-07-09  Mark Lam  <mark.lam@apple.com>
 139
 140             Gardening to unbreak builds on the Windows bot.
 141
 142             Not reviewed.
 143
 144             * parser/ParserTokens.h:
 145
 146 2013-07-09  Lucas Forschler  <lforschler@apple.com>
 147
 148         Merge r152494
 149
 150     2013-07-09  Mark Lam  <mark.lam@apple.com>
 151
 152             Fix 30% JSBench regression (caused by adding column numbers to stack traces).
 153             https://bugs.webkit.org/show_bug.cgi?id=118481.
 154
 155             Reviewed by Mark Hahnenberg and Geoffrey Garen.
 156
 157             Previously, we already capture ExpressionRangeInfo that provides a divot for
 158             each bytecode that can potentially throw an exception (and therefore generate
 159             a stack trace). On first attempt to compute column numbers, we then do a walk
 160             of the source string to record all line start positions in a table associated
 161             with the SourceProvider. The column number can then be computed as
 162                 divot - lineStartFor(bytecodeOffset).
 163
 164             The computation of this lineStarts table is the source of the 30% JSBench
 165             performance regression.
 166
 167             The new code now records lineStarts as the lexer and parser scans the source
 168             code. These lineStarts are then used to compute the column number for the
 169             given divot, and stored in the ExpressionRangeInfo. Similarly, we also capture
 170             the line number at the divot point and store that in the ExpressionRangeInfo.
 171             Hence, to look up line and column numbers, we now lookup the ExpressionRangeInfo
 172             for the bytecodeOffset, and then compute the line and column from the values
 173             stored in the expression info.
 174
 175             The strategy:
 176             1. We want to minimize perturbations to the lexer and parser. Specifically,
 177                the changes added should not change how it scans code, and generate bytecode.
 178             2. We regard the divot as the source character position we are interested
 179                in. As such, we'll capture line and lineStart (for column) at the point
 180                when we capture the divot information. This ensures that the 3 values are
 181                consistent.
 182
 183             How the change is done:
 184             1. Change the lexer to track lineStarts.
 185             2. Change the parser to capture line and lineStarts at the point of capturing
 186                divots.
 187             3. Change the parser and associated code to plumb these values all the way to
 188                the point that the correspoinding ExpressionRangeInfo is emitted.
 189             4. Propagate and record SourceCode firstLine and firstLineColumnOffset to the
 190                the necessary places so that we can add them as needed when reifying
 191                UnlinkedCodeBlocks into CodeBlocks.
 192             5. Compress the line and column number values in the ExpressionRangeInfo. In
 193                practice, we seldom have both large line and column numbers. Hence, we can
 194                encode both in an uint32_t most of the time. For the times when we encounter
 195                both large line and column numbers, we have a fallback to store the "fat"
 196                position info.
 197             6. Emit an ExpressionRangeInfo for UnaryOp nodes to get more line and column
 198                number coverage.
 199             7. Change the interpreter to use the new way of computing line and column.
 200             8. Delete old line and column computation code that is now unused.
 201
 202             Misc details:
 203             - the old lexer was tracking both a startOffset and charPosition where
 204               charPosition equals startOffset - SourceCode.startOffset. We now use
 205               startOffset exclusively throughout the system for consistency.
 206               All offset values (including lineStart) are relative to the start of the
 207               SourceProvider string. These values will only be converted to be relative
 208               to the SourceCode.startOffset at the very last minute i.e. when the divot
 209               is stored into the ExpressionRangeInfo.
 210
 211               This change to use the same offset system everywhere reduces confusion
 212               from having to convert back and forth between the 2 systems. It also
 213               enables a lot of assertions to be used.
 214
 215             - Also fixed some bugs in the choice of divot positions to use. For example,
 216               both Eval and Function expressions previously used column numbers from
 217               the start of the expression but used the line number at the end of the
 218               expression. This is now fixed to use either the start or end positions
 219               as appropriate, but not a mix of line and columns from both.
 220
 221             - Why use ints instead of unsigneds for offsets and lineStarts inside the
 222               lexer and parser?
 223               Some tests (e.g. fast/js/call-base-resolution.html and
 224               fast/js/eval-cross-window.html) has shown that lineStart offsets can be
 225               prior to the SourceCode.startOffset. Keeping the lexer offsets as ints
 226               simplifies computations and makes it easier to maintain the assertions
 227               that (startOffset >= lineStartOffset).
 228
 229               However, column and line numbers are always unsigned when we publish
 230               them to the ExpressionRangeInfo. The ints are only used inside the
 231               lexer and parser ... well, and bytecode generator.
 232
 233             - For all cases, lineStart is always captured where the divot is captured.
 234               However, some sputnik conformance tests have shown that we cannot honor
 235               line breaks for assignment statements like the following:
 236
 237                   eval("x\u000A*=\u000A-1;");
 238
 239               In this case, the lineStart is expected to be captured at the start of
 240               the assignment expression instead of at the divot point in the middle.
 241               The assignment expression is the only special case for this.
 242
 243             This patch has been tested against the full layout tests both with release
 244             and debug builds with no regression.
 245
 246             * API/JSContextRef.cpp:
 247             (JSContextCreateBacktrace):
 248               - Updated to use the new StackFrame::computeLineAndColumn().
 249             * bytecode/CodeBlock.cpp:
 250             (JSC::CodeBlock::CodeBlock):
 251               - Added m_firstLineColumnOffset initialization.
 252               - Plumbed the firstLineColumnOffset into the SourceCode.
 253               - Initialized column for op_debug using the new way.
 254             (JSC::CodeBlock::lineNumberForBytecodeOffset):
 255               - Changed to compute line number using the ExpressionRangeInfo.
 256             (JSC::CodeBlock::columnNumberForBytecodeOffset): Added
 257               - Changed to compute column number using the ExpressionRangeInfo.
 258             (JSC::CodeBlock::expressionRangeForBytecodeOffset):
 259             * bytecode/CodeBlock.h:
 260             (JSC::CodeBlock::firstLineColumnOffset):
 261             (JSC::GlobalCodeBlock::GlobalCodeBlock):
 262               - Plumbed firstLineColumnOffset through to the super class.
 263             (JSC::ProgramCodeBlock::ProgramCodeBlock):
 264               - Plumbed firstLineColumnOffset through to the super class.
 265             (JSC::EvalCodeBlock::EvalCodeBlock):
 266               - Plumbed firstLineColumnOffset through to the super class.
 267                 But for EvalCodeBlocks, the firstLineColumnOffset is always 1
 268                 because we're starting with a new source string with no start
 269                 offset.
 270             (JSC::FunctionCodeBlock::FunctionCodeBlock):
 271               - Plumbed firstLineColumnOffset through to the super class.
 272
 273             * bytecode/ExpressionRangeInfo.h:
 274               - Added modes for encoding line and column into a single 30-bit
 275                 unsigned. The encoding is in 1 of 3 modes:
 276                 1. FatLineMode: 22-bit line, 8-bit column
 277                 2. FatColumnMode: 8-bit line, 22-bit column
 278                 3. FatLineAndColumnMode: 32-bit line, 32-bit column
 279             (JSC::ExpressionRangeInfo::encodeFatLineMode): Added.
 280               - Encodes line and column into the 30-bit position using FatLine mode.
 281             (JSC::ExpressionRangeInfo::encodeFatColumnMode): Added.
 282               - Encodes line and column into the 30-bit position using FatColumn mode.
 283             (JSC::ExpressionRangeInfo::decodeFatLineMode): Added.
 284               - Decodes the FatLine mode 30-bit position into line and column.
 285             (JSC::ExpressionRangeInfo::decodeFatColumnMode): Added.
 286               - Decodes the FatColumn mode 30-bit position into line and column.
 287
 288             * bytecode/UnlinkedCodeBlock.cpp:
 289             (JSC::UnlinkedFunctionExecutable::UnlinkedFunctionExecutable):
 290               - Plumbed startColumn through.
 291             (JSC::UnlinkedFunctionExecutable::link):
 292               - Plumbed startColumn through.
 293             (JSC::UnlinkedCodeBlock::lineNumberForBytecodeOffset):
 294               - Computes a line number using the new way.
 295             (JSC::UnlinkedCodeBlock::expressionRangeForBytecodeOffset):
 296               - Added decoding of line and column.
 297               - Added handling of the case when we do not find a fitting expression
 298                 range info for a specified bytecodeOffset. This only happens if the
 299                 bytecodeOffset is below the first expression range info. In that
 300                 case, we'll use the first expression range info entry.
 301             (JSC::UnlinkedCodeBlock::addExpressionInfo):
 302               - Added encoding of line and column.
 303
 304             * bytecode/UnlinkedCodeBlock.h:
 305               - Added m_expressionInfoFatPositions in RareData.
 306             (JSC::UnlinkedFunctionExecutable::functionStartColumn):
 307             (JSC::UnlinkedCodeBlock::shrinkToFit):
 308               - Removed obsoleted m_lineInfo.
 309             * bytecompiler/BytecodeGenerator.cpp:
 310             (JSC::BytecodeGenerator::emitCall): Plumbed line and lineStart through.
 311             (JSC::BytecodeGenerator::emitCallEval): Plumbed line and lineStart through.
 312             (JSC::BytecodeGenerator::emitCallVarargs): Plumbed line and lineStart through.
 313             (JSC::BytecodeGenerator::emitConstruct): Plumbed line and lineStart through.
 314             (JSC::BytecodeGenerator::emitDebugHook): Plumbed lineStart through.
 315             * bytecompiler/BytecodeGenerator.h:
 316             (JSC::BytecodeGenerator::emitNode):
 317             (JSC::BytecodeGenerator::emitNodeInConditionContext):
 318               - Removed obsoleted m_lineInfo.
 319             (JSC::BytecodeGenerator::emitExpressionInfo):
 320               - Plumbed line and lineStart through.
 321               - Compute the line and column to be added to the expression range info.
 322             * bytecompiler/NodesCodegen.cpp:
 323             (JSC::ThrowableExpressionData::emitThrowReferenceError):
 324             (JSC::ResolveNode::emitBytecode):
 325             (JSC::ArrayNode::toArgumentList):
 326             (JSC::BracketAccessorNode::emitBytecode):
 327             (JSC::DotAccessorNode::emitBytecode):
 328             (JSC::NewExprNode::emitBytecode):
 329             (JSC::EvalFunctionCallNode::emitBytecode):
 330             (JSC::FunctionCallValueNode::emitBytecode):
 331             (JSC::FunctionCallResolveNode::emitBytecode):
 332             (JSC::FunctionCallBracketNode::emitBytecode):
 333             (JSC::FunctionCallDotNode::emitBytecode):
 334             (JSC::CallFunctionCallDotNode::emitBytecode):
 335             (JSC::ApplyFunctionCallDotNode::emitBytecode):
 336             (JSC::PostfixNode::emitResolve):
 337             (JSC::PostfixNode::emitBracket):
 338             (JSC::PostfixNode::emitDot):
 339             (JSC::DeleteResolveNode::emitBytecode):
 340             (JSC::DeleteBracketNode::emitBytecode):
 341             (JSC::DeleteDotNode::emitBytecode):
 342             (JSC::PrefixNode::emitResolve):
 343             (JSC::PrefixNode::emitBracket):
 344             (JSC::PrefixNode::emitDot):
 345               - Plumbed line and lineStart through the above as needed.
 346
 347             (JSC::UnaryOpNode::emitBytecode):
 348               - Added emission of an ExpressionRangeInfo for the UnaryOp node.
 349
 350             (JSC::BinaryOpNode::emitStrcat):
 351             (JSC::ThrowableBinaryOpNode::emitBytecode):
 352             (JSC::InstanceOfNode::emitBytecode):
 353             (JSC::emitReadModifyAssignment):
 354             (JSC::ReadModifyResolveNode::emitBytecode):
 355             (JSC::AssignResolveNode::emitBytecode):
 356             (JSC::AssignDotNode::emitBytecode):
 357             (JSC::ReadModifyDotNode::emitBytecode):
 358             (JSC::AssignBracketNode::emitBytecode):
 359             (JSC::ReadModifyBracketNode::emitBytecode):
 360               - Plumbed line and lineStart through the above as needed.
 361
 362             (JSC::ConstStatementNode::emitBytecode):
 363             (JSC::EmptyStatementNode::emitBytecode):
 364             (JSC::DebuggerStatementNode::emitBytecode):
 365             (JSC::ExprStatementNode::emitBytecode):
 366             (JSC::VarStatementNode::emitBytecode):
 367             (JSC::IfElseNode::emitBytecode):
 368             (JSC::DoWhileNode::emitBytecode):
 369             (JSC::WhileNode::emitBytecode):
 370             (JSC::ForNode::emitBytecode):
 371             (JSC::ForInNode::emitBytecode):
 372             (JSC::ContinueNode::emitBytecode):
 373             (JSC::BreakNode::emitBytecode):
 374             (JSC::ReturnNode::emitBytecode):
 375             (JSC::WithNode::emitBytecode):
 376             (JSC::SwitchNode::emitBytecode):
 377             (JSC::LabelNode::emitBytecode):
 378             (JSC::ThrowNode::emitBytecode):
 379             (JSC::TryNode::emitBytecode):
 380             (JSC::ProgramNode::emitBytecode):
 381             (JSC::EvalNode::emitBytecode):
 382             (JSC::FunctionBodyNode::emitBytecode):
 383               - Plumbed line and lineStart through the above as needed.
 384
 385             * interpreter/Interpreter.cpp:
 386             (JSC::appendSourceToError):
 387               - Added line and column arguments for expressionRangeForBytecodeOffset().
 388             (JSC::StackFrame::computeLineAndColumn):
 389               - Replaces StackFrame::line() and StackFrame::column().
 390             (JSC::StackFrame::expressionInfo):
 391               - Added line and column arguments.
 392             (JSC::StackFrame::toString):
 393               - Changed to use the new StackFrame::computeLineAndColumn().
 394             (JSC::Interpreter::getStackTrace):
 395               - Added the needed firstLineColumnOffset arg for the StackFrame.
 396
 397             * interpreter/Interpreter.h:
 398             * parser/ASTBuilder.h:
 399             (JSC::ASTBuilder::BinaryOpInfo::BinaryOpInfo):
 400             (JSC::ASTBuilder::AssignmentInfo::AssignmentInfo):
 401             (JSC::ASTBuilder::createResolve):
 402             (JSC::ASTBuilder::createBracketAccess):
 403             (JSC::ASTBuilder::createDotAccess):
 404             (JSC::ASTBuilder::createRegExp):
 405             (JSC::ASTBuilder::createNewExpr):
 406             (JSC::ASTBuilder::createAssignResolve):
 407             (JSC::ASTBuilder::createFunctionExpr):
 408             (JSC::ASTBuilder::createFunctionBody):
 409             (JSC::ASTBuilder::createGetterOrSetterProperty):
 410             (JSC::ASTBuilder::createFuncDeclStatement):
 411             (JSC::ASTBuilder::createBlockStatement):
 412             (JSC::ASTBuilder::createExprStatement):
 413             (JSC::ASTBuilder::createIfStatement):
 414             (JSC::ASTBuilder::createForLoop):
 415             (JSC::ASTBuilder::createForInLoop):
 416             (JSC::ASTBuilder::createVarStatement):
 417             (JSC::ASTBuilder::createReturnStatement):
 418             (JSC::ASTBuilder::createBreakStatement):
 419             (JSC::ASTBuilder::createContinueStatement):
 420             (JSC::ASTBuilder::createTryStatement):
 421             (JSC::ASTBuilder::createSwitchStatement):
 422             (JSC::ASTBuilder::createWhileStatement):
 423             (JSC::ASTBuilder::createDoWhileStatement):
 424             (JSC::ASTBuilder::createLabelStatement):
 425             (JSC::ASTBuilder::createWithStatement):
 426             (JSC::ASTBuilder::createThrowStatement):
 427             (JSC::ASTBuilder::createDebugger):
 428             (JSC::ASTBuilder::createConstStatement):
 429             (JSC::ASTBuilder::appendBinaryExpressionInfo):
 430             (JSC::ASTBuilder::appendUnaryToken):
 431             (JSC::ASTBuilder::unaryTokenStackLastStart):
 432             (JSC::ASTBuilder::unaryTokenStackLastLineStartPosition): Added.
 433             (JSC::ASTBuilder::assignmentStackAppend):
 434             (JSC::ASTBuilder::createAssignment):
 435             (JSC::ASTBuilder::setExceptionLocation):
 436             (JSC::ASTBuilder::makeDeleteNode):
 437             (JSC::ASTBuilder::makeFunctionCallNode):
 438             (JSC::ASTBuilder::makeBinaryNode):
 439             (JSC::ASTBuilder::makeAssignNode):
 440             (JSC::ASTBuilder::makePrefixNode):
 441             (JSC::ASTBuilder::makePostfixNode):.
 442               - Plumbed line, lineStart, and startColumn through the above as needed.
 443
 444             * parser/Lexer.cpp:
 445             (JSC::::currentSourcePtr):
 446             (JSC::::setCode):
 447               - Added tracking for sourceoffset and lineStart.
 448             (JSC::::internalShift):
 449             (JSC::::parseIdentifier):
 450               - Added tracking for lineStart.
 451             (JSC::::parseIdentifierSlowCase):
 452             (JSC::::parseString):
 453               - Added tracking for lineStart.
 454             (JSC::::parseStringSlowCase):
 455             (JSC::::lex):
 456               - Added tracking for sourceoffset.
 457             (JSC::::sourceCode):
 458             * parser/Lexer.h:
 459             (JSC::Lexer::currentOffset):
 460             (JSC::Lexer::currentLineStartOffset):
 461             (JSC::Lexer::setOffset):
 462               - Added tracking for lineStart.
 463             (JSC::Lexer::offsetFromSourcePtr): Added. conversion function.
 464             (JSC::Lexer::sourcePtrFromOffset): Added. conversion function.
 465             (JSC::Lexer::setOffsetFromSourcePtr):
 466             (JSC::::lexExpectIdentifier):
 467               - Added tracking for sourceoffset and lineStart.
 468
 469             * parser/NodeConstructors.h:
 470             (JSC::Node::Node):
 471             (JSC::ResolveNode::ResolveNode):
 472             (JSC::EvalFunctionCallNode::EvalFunctionCallNode):
 473             (JSC::FunctionCallValueNode::FunctionCallValueNode):
 474             (JSC::FunctionCallResolveNode::FunctionCallResolveNode):
 475             (JSC::FunctionCallBracketNode::FunctionCallBracketNode):
 476             (JSC::FunctionCallDotNode::FunctionCallDotNode):
 477             (JSC::CallFunctionCallDotNode::CallFunctionCallDotNode):
 478             (JSC::ApplyFunctionCallDotNode::ApplyFunctionCallDotNode):
 479             (JSC::PostfixNode::PostfixNode):
 480             (JSC::DeleteResolveNode::DeleteResolveNode):
 481             (JSC::DeleteBracketNode::DeleteBracketNode):
 482             (JSC::DeleteDotNode::DeleteDotNode):
 483             (JSC::PrefixNode::PrefixNode):
 484             (JSC::ReadModifyResolveNode::ReadModifyResolveNode):
 485             (JSC::ReadModifyBracketNode::ReadModifyBracketNode):
 486             (JSC::AssignBracketNode::AssignBracketNode):
 487             (JSC::AssignDotNode::AssignDotNode):
 488             (JSC::ReadModifyDotNode::ReadModifyDotNode):
 489             (JSC::AssignErrorNode::AssignErrorNode):
 490             (JSC::WithNode::WithNode):
 491             (JSC::ForInNode::ForInNode):
 492               - Plumbed line and lineStart through the above as needed.
 493             * parser/Nodes.cpp:
 494             (JSC::StatementNode::setLoc): Plumbed lineStart.
 495             (JSC::ScopeNode::ScopeNode): Plumbed lineStart.
 496             (JSC::ProgramNode::ProgramNode): Plumbed startColumn.
 497             (JSC::ProgramNode::create): Plumbed startColumn.
 498             (JSC::EvalNode::create):
 499             (JSC::FunctionBodyNode::FunctionBodyNode): Plumbed startColumn.
 500             (JSC::FunctionBodyNode::create): Plumbed startColumn.
 501             * parser/Nodes.h:
 502             (JSC::Node::startOffset):
 503             (JSC::Node::lineStartOffset): Added.
 504             (JSC::StatementNode::firstLine):
 505             (JSC::StatementNode::lastLine):
 506             (JSC::ThrowableExpressionData::ThrowableExpressionData):
 507             (JSC::ThrowableExpressionData::setExceptionSourceCode):
 508             (JSC::ThrowableExpressionData::divotStartOffset):
 509             (JSC::ThrowableExpressionData::divotEndOffset):
 510             (JSC::ThrowableExpressionData::divotLine):
 511             (JSC::ThrowableExpressionData::divotLineStart):
 512             (JSC::ThrowableSubExpressionData::ThrowableSubExpressionData):
 513             (JSC::ThrowableSubExpressionData::setSubexpressionInfo):
 514             (JSC::ThrowableSubExpressionData::subexpressionDivot):
 515             (JSC::ThrowableSubExpressionData::subexpressionStartOffset):
 516             (JSC::ThrowableSubExpressionData::subexpressionEndOffset):
 517             (JSC::ThrowableSubExpressionData::subexpressionLine):
 518             (JSC::ThrowableSubExpressionData::subexpressionLineStart):
 519             (JSC::ThrowablePrefixedSubExpressionData::ThrowablePrefixedSubExpressionData):
 520             (JSC::ThrowablePrefixedSubExpressionData::setSubexpressionInfo):
 521             (JSC::ThrowablePrefixedSubExpressionData::subexpressionDivot):
 522             (JSC::ThrowablePrefixedSubExpressionData::subexpressionStartOffset):
 523             (JSC::ThrowablePrefixedSubExpressionData::subexpressionEndOffset):
 524             (JSC::ThrowablePrefixedSubExpressionData::subexpressionLine):
 525             (JSC::ThrowablePrefixedSubExpressionData::subexpressionLineStart):
 526             (JSC::ScopeNode::startStartOffset):
 527             (JSC::ScopeNode::startLineStartOffset):
 528             (JSC::ProgramNode::startColumn):
 529             (JSC::EvalNode::startColumn):
 530             (JSC::FunctionBodyNode::startColumn):
 531               - Plumbed line and lineStart through the above as needed.
 532             * parser/Parser.cpp:
 533             (JSC::::Parser):
 534             (JSC::::parseSourceElements):
 535             (JSC::::parseVarDeclarationList):
 536             (JSC::::parseConstDeclarationList):
 537             (JSC::::parseForStatement):
 538             (JSC::::parseBreakStatement):
 539             (JSC::::parseContinueStatement):
 540             (JSC::::parseReturnStatement):
 541             (JSC::::parseThrowStatement):
 542             (JSC::::parseWithStatement):
 543               - Plumbed line and lineStart through the above as needed.
 544             (JSC::::parseFunctionBody):
 545               - Plumbed startColumn.
 546             (JSC::::parseFunctionInfo):
 547             (JSC::::parseFunctionDeclaration):
 548             (JSC::LabelInfo::LabelInfo):
 549             (JSC::::parseExpressionOrLabelStatement):
 550             (JSC::::parseAssignmentExpression):
 551             (JSC::::parseBinaryExpression):
 552             (JSC::::parseProperty):
 553             (JSC::::parseObjectLiteral):
 554             (JSC::::parsePrimaryExpression):
 555             (JSC::::parseMemberExpression):
 556             (JSC::::parseUnaryExpression):
 557               - Plumbed line, lineStart, startColumn through the above as needed.
 558             * parser/Parser.h:
 559             (JSC::Parser::next):
 560             (JSC::Parser::nextExpectIdentifier):
 561             (JSC::Parser::tokenStart):
 562             (JSC::Parser::tokenColumn):
 563             (JSC::Parser::tokenEnd):
 564             (JSC::Parser::tokenLineStart):
 565             (JSC::Parser::lastTokenLine):
 566             (JSC::Parser::lastTokenLineStart):
 567             (JSC::::parse):
 568             * parser/ParserTokens.h:
 569             (JSC::JSTokenLocation::JSTokenLocation):
 570               - Plumbed lineStart.
 571             (JSC::JSTokenLocation::lineStartPosition):
 572             (JSC::JSTokenLocation::startPosition):
 573             (JSC::JSTokenLocation::endPosition):
 574             * parser/SourceCode.h:
 575             (JSC::SourceCode::SourceCode):
 576             (JSC::SourceCode::startColumn):
 577             (JSC::makeSource):
 578             (JSC::SourceCode::subExpression):
 579             * parser/SourceProvider.cpp: delete old code.
 580             * parser/SourceProvider.h: delete old code.
 581             * parser/SourceProviderCacheItem.h:
 582             (JSC::SourceProviderCacheItem::closeBraceToken):
 583             (JSC::SourceProviderCacheItem::SourceProviderCacheItem):
 584               - Plumbed lineStart.
 585             * parser/SyntaxChecker.h:
 586             (JSC::SyntaxChecker::makeFunctionCallNode):
 587             (JSC::SyntaxChecker::makeAssignNode):
 588             (JSC::SyntaxChecker::makePrefixNode):
 589             (JSC::SyntaxChecker::makePostfixNode):
 590             (JSC::SyntaxChecker::makeDeleteNode):
 591             (JSC::SyntaxChecker::createResolve):
 592             (JSC::SyntaxChecker::createBracketAccess):
 593             (JSC::SyntaxChecker::createDotAccess):
 594             (JSC::SyntaxChecker::createRegExp):
 595             (JSC::SyntaxChecker::createNewExpr):
 596             (JSC::SyntaxChecker::createAssignResolve):
 597             (JSC::SyntaxChecker::createFunctionExpr):
 598             (JSC::SyntaxChecker::createFunctionBody):
 599             (JSC::SyntaxChecker::createFuncDeclStatement):
 600             (JSC::SyntaxChecker::createForInLoop):
 601             (JSC::SyntaxChecker::createReturnStatement):
 602             (JSC::SyntaxChecker::createBreakStatement):
 603             (JSC::SyntaxChecker::createContinueStatement):
 604             (JSC::SyntaxChecker::createWithStatement):
 605             (JSC::SyntaxChecker::createLabelStatement):
 606             (JSC::SyntaxChecker::createThrowStatement):
 607             (JSC::SyntaxChecker::createGetterOrSetterProperty):
 608             (JSC::SyntaxChecker::appendBinaryExpressionInfo):
 609             (JSC::SyntaxChecker::operatorStackPop):
 610               - Made SyntaxChecker prototype changes to match ASTBuilder due to new
 611                 args added for plumbing line, lineStart, and startColumn.
 612             * runtime/CodeCache.cpp:
 613             (JSC::CodeCache::generateBytecode):
 614             (JSC::CodeCache::getCodeBlock):
 615               - Plumbed startColumn.
 616             * runtime/Executable.cpp:
 617             (JSC::FunctionExecutable::FunctionExecutable):
 618             (JSC::ProgramExecutable::compileInternal):
 619             (JSC::FunctionExecutable::produceCodeBlockFor):
 620             (JSC::FunctionExecutable::fromGlobalCode):
 621               - Plumbed startColumn.
 622             * runtime/Executable.h:
 623             (JSC::ScriptExecutable::startColumn):
 624             (JSC::ScriptExecutable::recordParse):
 625             (JSC::FunctionExecutable::create):
 626               - Plumbed startColumn.
 627
 628 2013-07-08  Lucas Forschler  <lforschler@apple.com>
 629
 630         Merge r152052
 631
 632     2013-06-26  Anders Carlsson  <andersca@apple.com>
 633
 634             Add JSStringCreateWithCharactersNoCopy SPI
 635             https://bugs.webkit.org/show_bug.cgi?id=118074
 636             <rdar://problem/14279905>
 637
 638             Reviewed by Geoffrey Garen.
 639
 640             * API/JSStringRef.cpp:
 641             (JSStringCreateWithCharactersNoCopy):
 642             Create a new OpaqueJSString, using the newly added StringImpl::createWithoutCopying function.
 643
 644             * API/JSStringRefPrivate.h: Added.
 645             Add a home for the JSStringCreateWithCharactersNoCopy function.
 646
 647             * API/OpaqueJSString.h:
 648             (OpaqueJSString::OpaqueJSString):
 649             Just call isolatedCopy on the passed in string.
 650
 651             * API/tests/testapi.c:
 652             Add an API test for JSStringCreateWithCharactersNoCopy.
 653
 654             * JavaScriptCore.xcodeproj/project.pbxproj:
 655             Add new files.
 656
 657 2013-07-08  Lucas Forschler  <lforschler@apple.com>
 658
 659         Merge r152314
 660
 661     2013-07-02  Mark Hahnenberg  <mhahnenberg@apple.com>
 662
 663             Replace RELEASE_ASSERT with ASSERT in CodeBlock:: bytecodeOffsetForCallAtIndex
 664             https://bugs.webkit.org/show_bug.cgi?id=118316
 665
 666             Reviewed by Geoffrey Garen.
 667
 668             This is causing some crashiness in release builds. We should replace it with an ASSERT
 669             until we track down all the places that need fixing in bug 118315.
 670
 671             * bytecode/CodeBlock.h:
 672             (JSC::CodeBlock::bytecodeOffsetForCallAtIndex):
 673
 674 2013-07-01  Lucas Forschler  <lforschler@apple.com>
 675
 676         Merge r152091
 677
 678     2013-06-27  Timothy Hatcher  <timothy@apple.com>
 679
 680             Notify the debugger about functions created from source code via new Function() or WebCore::JSLazyEventListener.
 681
 682             https://bugs.webkit.org/show_bug.cgi?id=118063
 683
 684             Reviewed by Geoffrey Garen.
 685
 686             * bytecode/UnlinkedCodeBlock.cpp:
 687             (JSC::UnlinkedFunctionExecutable::fromGlobalCode): Call Debugger::sourceParsed.
 688
 689 2013-07-01  Lucas Forschler  <lforschler@apple.com>
 690
 691         Merge r151978
 692
 693     2013-06-25  Ryosuke Niwa  <rniwa@webkit.org>
 694
 695             JSString should remember AtomicString
 696             https://bugs.webkit.org/show_bug.cgi?id=117386
 697
 698             Reviewed by Geoffrey Garen.
 699
 700             Added JSValue::toAtomicString and JSString::atomicString. These two functions allow WebCore to update
 701             JSString's m_value to set isAtomic flag and avoid the AtomicStringTable lookups in subsequent attempts
 702             to obtain the AtomicString of the same value.
 703
 704             * runtime/JSCJSValue.h:
 705             * runtime/JSString.h:
 706             (JSC::JSString::atomicString):
 707             (JSC::JSValue::toAtomicString):
 708
 709 2013-06-25  Lucas Forschler  <lforschler@apple.com>
 710
 711         Merge r151786
 712
 713     2013-06-20  Mark Hahnenberg  <mhahnenberg@apple.com>
 714
 715             Improper deallocation of JSManagedValue causes crashes during autorelease pool draining
 716             https://bugs.webkit.org/show_bug.cgi?id=117840
 717
 718             Reviewed by Geoffrey Garen.
 719
 720             Improperly managing a JSManagedValue can cause a crash when the JSC::Weak inside the
 721             JSManagedValue is destroyed upon deallocation. We would rather have improperly maintained
 722             JSManagedValues cause memory leaks than take down the whole app.
 723
 724             The fix is to use the callback to the JSC::Weak on the destruction of the VM so that we
 725             can safely null it out. This will prevent ~Weak from crashing.
 726
 727             * API/JSManagedValue.mm:
 728             (-[JSManagedValue JSC::JSC::]):
 729             (JSManagedValueHandleOwner::finalize):
 730             * API/tests/testapi.mm: Added a test that crashed prior to this fix due to a leaked
 731             managed reference. Also fixed a small style nit I noticed in another test.
 732
 733 2013-06-25  Lucas Forschler  <lforschler@apple.com>
 734
 735         Merge r151923
 736
 737     2013-06-24  Roger Fong  <roger_fong@apple.com>
 738
 739             Unreviewed. Makefile build fix for AppleWindows.
 740
 741             * JavaScriptCore.vcxproj/JavaScriptCore.make:
 742
 743 2013-06-21  Lucas Forschler  <lforschler@apple.com>
 744
 745         Merge r151787
 746
 747     2013-06-20  Roger Fong  <roger_fong@apple.com>
 748
 749             Make Windows makefile copy build output to a different folder.
 750             <rdar://problem/14219184>.
 751
 752             * JavaScriptCore.vcxproj/JavaScriptCore.make:
 753
 754 2013-06-18  Roger Fong  <roger_fong@apple.com>
 755
 756         Disable some feature flags.
 757         <rdar://problem/14171207>.
 758
 759         Rubberstamped by Jon Lee.
 760
 761         * Configurations/FeatureDefines.xcconfig:
 762
 763 2013-06-18  Oliver Hunt  <oliver@apple.com>
 764
 765         Going to google.com/trends causes a crash
 766         https://bugs.webkit.org/show_bug.cgi?id=117602
 767
 768         Reviewed by Geoffrey Garen.
 769
 770         When handling op_throw, etc we need to flush the variables and arguments
 771         for the entire inline stack, not just the top frame.
 772
 773         * dfg/DFGByteCodeParser.cpp:
 774         (JSC::DFG::ByteCodeParser::flushAllArgumentsAndCapturedVariablesInInlineStack):
 775         (JSC::DFG::ByteCodeParser::parseBlock):
 776
 777 2013-06-18  Roger Fong  <roger_fong@apple.com>
 778
 779         Replace tools32 folder with tools and update WebKit Windows solution accordingly.
 780         <rdar://problem/14118143>.
 781
 782         Rubberstamped by Brent Fulgham.
 783
 784         * JavaScriptCore.vcxproj/JavaScriptCore.vcxproj:
 785         * JavaScriptCore.vcxproj/JavaScriptCoreDebug.props:
 786         * JavaScriptCore.vcxproj/JavaScriptCoreDebugCFLite.props:
 787         * JavaScriptCore.vcxproj/JavaScriptCoreGenerated.vcxproj:
 788         * JavaScriptCore.vcxproj/JavaScriptCoreGeneratedDebug.props:
 789         * JavaScriptCore.vcxproj/JavaScriptCoreGeneratedProduction.props:
 790         * JavaScriptCore.vcxproj/JavaScriptCoreGeneratedRelease.props:
 791         * JavaScriptCore.vcxproj/JavaScriptCorePostBuild.cmd:
 792         * JavaScriptCore.vcxproj/JavaScriptCoreProduction.props:
 793         * JavaScriptCore.vcxproj/JavaScriptCoreRelease.props:
 794         * JavaScriptCore.vcxproj/JavaScriptCoreReleaseCFLite.props:
 795         * JavaScriptCore.vcxproj/LLInt/LLIntAssembly/LLIntAssembly.vcxproj:
 796         * JavaScriptCore.vcxproj/LLInt/LLIntDesiredOffsets/LLIntDesiredOffsets.vcxproj:
 797         * JavaScriptCore.vcxproj/LLInt/LLIntOffsetsExtractor/LLIntOffsetsExtractor.vcxproj:
 798         * JavaScriptCore.vcxproj/LLInt/LLIntOffsetsExtractor/LLIntOffsetsExtractorDebug.props:
 799         * JavaScriptCore.vcxproj/LLInt/LLIntOffsetsExtractor/LLIntOffsetsExtractorProduction.props:
 800         * JavaScriptCore.vcxproj/LLInt/LLIntOffsetsExtractor/LLIntOffsetsExtractorRelease.props:
 801         * JavaScriptCore.vcxproj/jsc/jsc.vcxproj:
 802         * JavaScriptCore.vcxproj/jsc/jscDebug.props:
 803         * JavaScriptCore.vcxproj/jsc/jscProduction.props:
 804         * JavaScriptCore.vcxproj/jsc/jscRelease.props:
 805         * JavaScriptCore.vcxproj/testRegExp/testRegExp.vcxproj:
 806         * JavaScriptCore.vcxproj/testRegExp/testRegExpDebug.props:
 807         * JavaScriptCore.vcxproj/testRegExp/testRegExpProduction.props:
 808         * JavaScriptCore.vcxproj/testRegExp/testRegExpRelease.props:
 809         * JavaScriptCore.vcxproj/testapi/testapi.vcxproj:
 810         * JavaScriptCore.vcxproj/testapi/testapiDebug.props:
 811         * JavaScriptCore.vcxproj/testapi/testapiDebugCFLite.props:
 812         * JavaScriptCore.vcxproj/testapi/testapiProduction.props:
 813         * JavaScriptCore.vcxproj/testapi/testapiRelease.props:
 814         * JavaScriptCore.vcxproj/testapi/testapiReleaseCFLite.props:
 815
 816 2013-06-17  Roger Fong  <roger_fong@apple.com>
 817
 818         Modify Windows makefiles to copy some bin output into Program Files.
 819         https://bugs.webkit.org/show_bug.cgi?id=117714.
 820         <rdar://problem/14179054>
 821
 822         Reviewed by Brent Fulgham.
 823
 824         * JavaScriptCore.vcxproj/JavaScriptCore.make:
 825
 826 2013-06-14  Ryosuke Niwa  <rniwa@webkit.org>
 827
 828         Function names on Object.prototype should be common identifiers
 829         https://bugs.webkit.org/show_bug.cgi?id=117614
 830
 831         Reviewed by Darin Adler.
 832
 833         Patch written by Sam Weinig. Make Object's prototype function names common identififers since they're used frequently.
 834
 835         * runtime/CommonIdentifiers.h:
 836         * runtime/FunctionConstructor.cpp:
 837         (JSC::constructFunction):
 838         * runtime/JSGlobalObject.cpp:
 839         (JSC::JSGlobalObject::reset):
 840         * runtime/JSObject.h:
 841         * runtime/ObjectPrototype.cpp:
 842         (JSC::ObjectPrototype::finishCreation):
 843         * runtime/StringPrototype.cpp:
 844         (JSC::StringPrototype::finishCreation):
 845
 846 2013-06-13  Ryosuke Niwa  <rniwa@webkit.org>
 847
 848         Remove LiteralIdentifierTable
 849         https://bugs.webkit.org/show_bug.cgi?id=117613
 850
 851         Reviewed by Geoffrey Garen.
 852
 853         Removed LiteralIdentifierTable since it doesn't seem to have any perf. impact now.
 854
 855         * runtime/Identifier.cpp:
 856         (JSC::Identifier::add):
 857
 858 2013-06-12  Conrad Shultz  <conrad_shultz@apple.com>
 859
 860         JSExport header documentation substitutes "semicolon" for "colon"
 861         https://bugs.webkit.org/show_bug.cgi?id=117552
 862
 863         Reviewed by Mark Hahnenberg.
 864
 865         * API/JSExport.h:
 866         Fix a couple typos.
 867
 868 2013-06-10  Raphael Kubo da Costa  <raphael.kubo.da.costa@intel.com>
 869
 870         [JSC] Remove a vestige of wxWidgets support.
 871         https://bugs.webkit.org/show_bug.cgi?id=117419
 872
 873         Reviewed by Benjamin Poulain.
 874
 875         * runtime/JSExportMacros.h: Remove a check for BUILDING_WX__ that
 876         seems to have gone unnoticed when the wxWidgets port was removed.
 877
 878 2013-06-06  Roger Fong  <roger_fong@apple.com>
 879
 880         Stop copying AAS binaries into build folder.
 881         https://bugs.webkit.org/show_bug.cgi?id=117319.
 882
 883         Rubberstamped by Darin Adler.
 884
 885         * JavaScriptCore.vcxproj/JavaScriptCorePreLink.cmd:
 886         * JavaScriptCore.vcxproj/jsc/jscPostBuild.cmd:
 887         * JavaScriptCore.vcxproj/jsc/jscPreLink.cmd:
 888         * JavaScriptCore.vcxproj/testRegExp/testRegExpPostBuild.cmd:
 889         * JavaScriptCore.vcxproj/testRegExp/testRegExpPreLink.cmd:
 890         * JavaScriptCore.vcxproj/testapi/testapiPreLink.cmd:
 891
 892 2013-06-05  Filip Pizlo  <fpizlo@apple.com>
 893
 894         DFG CFA shouldn't filter ArrayModes with ALL_NON_ARRAY_ARRAY_MODES if the speculated type is not SpecArray
 895         https://bugs.webkit.org/show_bug.cgi?id=117279
 896         <rdar://problem/14078025>
 897
 898         Reviewed by Mark Hahnenberg.
 899
 900         * dfg/DFGAbstractValue.h:
 901         (JSC::DFG::AbstractValue::filterArrayModesByType):
 902
 903 2013-06-05  Michael Saboff  <msaboff@apple.com>
 904
 905         JSC: Crash beneath cti_op_div @ http://gmailblog.blogspot.com
 906         https://bugs.webkit.org/show_bug.cgi?id=117280
 907
 908         Reviewed by Filip Pizlo.
 909
 910         Updated the merging of VariableAccessData nodes in ArgumentPosition lists
 911         to find the unified VariableAccessData node that is the root of the
 912         current node instead of using the current node directly when merging
 913         attributes.
 914         Added new dump code to dump the ArgumentPosition list.
 915
 916         * dfg/DFGArgumentPosition.h:
 917         (JSC::DFG::rgumentPosition::mergeArgumentPredictionAwareness):
 918         (JSC::DFG::ArgumentPosition::mergeArgumentUnboxingAwareness):
 919         (JSC::DFG::ArgumentPosition::dump):
 920         * dfg/DFGGraph.cpp:
 921         (JSC::DFG::Graph::dump):
 922
 923 2013-06-05  Bear Travis  <betravis@adobe.com>
 924
 925         [CSS Exclusions][CSS Shapes] Split CSS Exclusions & Shapes compile & runtime flags
 926         https://bugs.webkit.org/show_bug.cgi?id=117172
 927
 928         Reviewed by Alexandru Chiculita.
 929
 930         Adding the CSS_SHAPES compile flag.
 931
 932         * Configurations/FeatureDefines.xcconfig:
 933
 934 2013-06-05  Balazs Kilvady  <kilvadyb@homejinni.com>
 935
 936         JSC Assertion tests failures on MIPS.
 937         https://bugs.webkit.org/show_bug.cgi?id=116552
 938
 939         Reviewed by Geoffrey Garen.
 940
 941         Fix condition handlig in branchAdd32 implemetations.
 942
 943         * assembler/MacroAssemblerMIPS.h:
 944         (JSC::MacroAssemblerMIPS::branchAdd32):
 945
 946 2013-06-04  Julien Brianceau  <jbrianceau@nds.com>
 947
 948         [sh4] Add floating point absolute function support in baseline JIT.
 949         https://bugs.webkit.org/show_bug.cgi?id=117147
 950
 951         Reviewed by Geoffrey Garen.
 952
 953         * assembler/MacroAssemblerSH4.h:
 954         (JSC::MacroAssemblerSH4::supportsFloatingPointAbs):
 955         (JSC::MacroAssemblerSH4::absDouble):
 956         * assembler/SH4Assembler.h:
 957         (JSC::SH4Assembler::dabs):
 958         (JSC::SH4Assembler::printInstr):
 959
 960 2013-06-04  Zan Dobersek  <zdobersek@igalia.com>
 961
 962         [JSC] Test262 15.5.4.9_3 test is failing
 963         https://bugs.webkit.org/show_bug.cgi?id=116789
 964
 965         Reviewed by Geoffrey Garen.
 966
 967         Bring the String.prototype.localeCompare behavior in line wit ES5 15.9.4.9.
 968         If method is not given enough arguments, the minimal amount of arguments must be assumed, with their value being undefined.
 969         The first argument to localeCompare, in its string form, is used as the 'that' string that's used in the comparison.
 970         Therefor, when calling str.localeCompare() or str.localeCompare(undefined), the first argument is `undefined` and the
 971         string "undefined" is used as the string to which value of str is compared.
 972
 973         * runtime/StringPrototype.cpp:
 974         (JSC::stringProtoFuncLocaleCompare): Remove the early return in case of no given arguments to achieve the desired behavior.
 975
 976 2013-06-03  Hojong Han  <hojong.han@samsung.com>
 977
 978         [EFL] Implement GCActivityCallback
 979         https://bugs.webkit.org/show_bug.cgi?id=95923
 980
 981         Reviewed by Geoffrey Garen.
 982
 983         Implements the activity triggered garbage collector.
 984         Additional GCs can be triggered by platfrom timer.
 985         It has sort of compaction effect not to make JSC heap grow fast
 986         so that memory usage becomes lower than usual.
 987
 988         * PlatformEfl.cmake: Added.
 989         * heap/HeapTimer.cpp:
 990         (JSC):
 991         (JSC::HeapTimer::HeapTimer):
 992         (JSC::HeapTimer::~HeapTimer):
 993         (JSC::HeapTimer::add):
 994         (JSC::HeapTimer::stop):
 995         (JSC::HeapTimer::timerEvent):
 996         * heap/HeapTimer.h:
 997         (HeapTimer):
 998         * jsc.cpp:
 999         (main):
1000         * runtime/GCActivityCallback.cpp:
1001         (JSC):
1002         (JSC::DefaultGCActivityCallback::DefaultGCActivityCallback):
1003         (JSC::DefaultGCActivityCallback::scheduleTimer):
1004         (JSC::DefaultGCActivityCallback::cancelTimer):
1005         (JSC::DefaultGCActivityCallback::didAllocate):
1006         * runtime/GCActivityCallback.h:
1007         (GCActivityCallback):
1008         (JSC::GCActivityCallback::GCActivityCallback):
1009         (DefaultGCActivityCallback):
1010
1011 2013-06-03  Roger Fong  <roger_fong@apple.com>
1012
1013         Nuke VS2005 files from the tree.
1014         <rdar://problem/14042021>.
1015
1016         Rubberstamped by Brent Fulgham.
1017
1018         * JavaScriptCore.vcproj: Removed.
1019         * JavaScriptCore.vcproj/JavaScriptCore: Removed.
1020         * JavaScriptCore.vcproj/JavaScriptCore.make: Removed.
1021         * JavaScriptCore.vcproj/JavaScriptCore.resources: Removed.
1022         * JavaScriptCore.vcproj/JavaScriptCore.resources/Info.plist: Removed.
1023         * JavaScriptCore.vcproj/JavaScriptCore.sln: Removed.
1024         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj: Removed.
1025         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreCF.vsprops: Removed.
1026         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreCFLite.vsprops: Removed.
1027         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreCommon.vsprops: Removed.
1028         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreDebug.vsprops: Removed.
1029         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreDebugAll.vsprops: Removed.
1030         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreDebugCairoCFLite.vsprops: Removed.
1031         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreExports.def: Removed.
1032         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreGenerated.make: Removed.
1033         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreGenerated.vcproj: Removed.
1034         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreGeneratedCommon.vsprops: Removed.
1035         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreGeneratedDebug.vsprops: Removed.
1036         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreGeneratedDebugAll.vsprops: Removed.
1037         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreGeneratedDebugCairoCFLite.vsprops: Removed.
1038         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreGeneratedProduction.vsprops: Removed.
1039         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreGeneratedRelease.vsprops: Removed.
1040         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreGeneratedReleaseCairoCFLite.vsprops: Removed.
1041         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreGeneratedReleasePGO.vsprops: Removed.
1042         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCorePGOOptimize.vsprops: Removed.
1043         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCorePostBuild.cmd: Removed.
1044         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCorePreBuild.cmd: Removed.
1045         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCorePreLink.cmd: Removed.
1046         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreProduction.vsprops: Removed.
1047         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreRelease.vsprops: Removed.
1048         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreReleaseCairoCFLite.vsprops: Removed.
1049         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreReleasePGO.vsprops: Removed.
1050         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreReleasePGOOptimize.vsprops: Removed.
1051         * JavaScriptCore.vcproj/JavaScriptCore/build-generated-files.sh: Removed.
1052         * JavaScriptCore.vcproj/JavaScriptCore/copy-files.cmd: Removed.
1053         * JavaScriptCore.vcproj/JavaScriptCoreSubmit.sln: Removed.
1054         * JavaScriptCore.vcproj/LLIntAssembly: Removed.
1055         * JavaScriptCore.vcproj/LLIntAssembly/LLIntAssembly.make: Removed.
1056         * JavaScriptCore.vcproj/LLIntAssembly/LLIntAssembly.vcproj: Removed.
1057         * JavaScriptCore.vcproj/LLIntAssembly/build-LLIntAssembly.sh: Removed.
1058         * JavaScriptCore.vcproj/LLIntDesiredOffsets: Removed.
1059         * JavaScriptCore.vcproj/LLIntDesiredOffsets/LLIntDesiredOffsets.make: Removed.
1060         * JavaScriptCore.vcproj/LLIntDesiredOffsets/LLIntDesiredOffsets.vcproj: Removed.
1061         * JavaScriptCore.vcproj/LLIntDesiredOffsets/build-LLIntDesiredOffsets.sh: Removed.
1062         * JavaScriptCore.vcproj/LLIntOffsetsExtractor: Removed.
1063         * JavaScriptCore.vcproj/LLIntOffsetsExtractor/LLIntOffsetsExtractor.vcproj: Removed.
1064         * JavaScriptCore.vcproj/LLIntOffsetsExtractor/LLIntOffsetsExtractorCommon.vsprops: Removed.
1065         * JavaScriptCore.vcproj/LLIntOffsetsExtractor/LLIntOffsetsExtractorDebug.vsprops: Removed.
1066         * JavaScriptCore.vcproj/LLIntOffsetsExtractor/LLIntOffsetsExtractorDebugAll.vsprops: Removed.
1067         * JavaScriptCore.vcproj/LLIntOffsetsExtractor/LLIntOffsetsExtractorDebugCairoCFLite.vsprops: Removed.
1068         * JavaScriptCore.vcproj/LLIntOffsetsExtractor/LLIntOffsetsExtractorProduction.vsprops: Removed.
1069         * JavaScriptCore.vcproj/LLIntOffsetsExtractor/LLIntOffsetsExtractorRelease.vsprops: Removed.
1070         * JavaScriptCore.vcproj/LLIntOffsetsExtractor/LLIntOffsetsExtractorReleaseCairoCFLite.vsprops: Removed.
1071         * JavaScriptCore.vcproj/LLIntOffsetsExtractor/LLIntOffsetsExtractorReleasePGO.vsprops: Removed.
1072         * JavaScriptCore.vcproj/jsc: Removed.
1073         * JavaScriptCore.vcproj/jsc/jsc.vcproj: Removed.
1074         * JavaScriptCore.vcproj/jsc/jscCommon.vsprops: Removed.
1075         * JavaScriptCore.vcproj/jsc/jscDebug.vsprops: Removed.
1076         * JavaScriptCore.vcproj/jsc/jscDebugAll.vsprops: Removed.
1077         * JavaScriptCore.vcproj/jsc/jscDebugCairoCFLite.vsprops: Removed.
1078         * JavaScriptCore.vcproj/jsc/jscPostBuild.cmd: Removed.
1079         * JavaScriptCore.vcproj/jsc/jscPreBuild.cmd: Removed.
1080         * JavaScriptCore.vcproj/jsc/jscPreLink.cmd: Removed.
1081         * JavaScriptCore.vcproj/jsc/jscProduction.vsprops: Removed.
1082         * JavaScriptCore.vcproj/jsc/jscRelease.vsprops: Removed.
1083         * JavaScriptCore.vcproj/jsc/jscReleaseCairoCFLite.vsprops: Removed.
1084         * JavaScriptCore.vcproj/jsc/jscReleasePGO.vsprops: Removed.
1085         * JavaScriptCore.vcproj/testRegExp: Removed.
1086         * JavaScriptCore.vcproj/testRegExp/testRegExp.vcproj: Removed.
1087         * JavaScriptCore.vcproj/testRegExp/testRegExpCommon.vsprops: Removed.
1088         * JavaScriptCore.vcproj/testRegExp/testRegExpDebug.vsprops: Removed.
1089         * JavaScriptCore.vcproj/testRegExp/testRegExpDebugAll.vsprops: Removed.
1090         * JavaScriptCore.vcproj/testRegExp/testRegExpDebugCairoCFLite.vsprops: Removed.
1091         * JavaScriptCore.vcproj/testRegExp/testRegExpPostBuild.cmd: Removed.
1092         * JavaScriptCore.vcproj/testRegExp/testRegExpPreBuild.cmd: Removed.
1093         * JavaScriptCore.vcproj/testRegExp/testRegExpPreLink.cmd: Removed.
1094         * JavaScriptCore.vcproj/testRegExp/testRegExpProduction.vsprops: Removed.
1095         * JavaScriptCore.vcproj/testRegExp/testRegExpRelease.vsprops: Removed.
1096         * JavaScriptCore.vcproj/testRegExp/testRegExpReleaseCairoCFLite.vsprops: Removed.
1097         * JavaScriptCore.vcproj/testRegExp/testRegExpReleasePGO.vsprops: Removed.
1098         * JavaScriptCore.vcproj/testapi: Removed.
1099         * JavaScriptCore.vcproj/testapi/testapi.vcproj: Removed.
1100         * JavaScriptCore.vcproj/testapi/testapiCommon.vsprops: Removed.
1101         * JavaScriptCore.vcproj/testapi/testapiDebug.vsprops: Removed.
1102         * JavaScriptCore.vcproj/testapi/testapiDebugAll.vsprops: Removed.
1103         * JavaScriptCore.vcproj/testapi/testapiDebugCairoCFLite.vsprops: Removed.
1104         * JavaScriptCore.vcproj/testapi/testapiPostBuild.cmd: Removed.
1105         * JavaScriptCore.vcproj/testapi/testapiPreBuild.cmd: Removed.
1106         * JavaScriptCore.vcproj/testapi/testapiPreLink.cmd: Removed.
1107         * JavaScriptCore.vcproj/testapi/testapiProduction.vsprops: Removed.
1108         * JavaScriptCore.vcproj/testapi/testapiRelease.vsprops: Removed.
1109         * JavaScriptCore.vcproj/testapi/testapiReleaseCairoCFLite.vsprops: Removed.
1110
1111 2013-05-31  Filip Pizlo  <fpizlo@apple.com>
1112
1113         Incorrect assertion in DFG::Graph::uncheckedActivationRegisterFor()
1114         <rdar://problem/13989324>
1115
1116         Rubber stamped by Mark Hahnenberg.
1117
1118         This has a bogus assertion that checks that the passed CodeOrigin doesn't have
1119         an inline call frame. This was well intentioned in the sense that it is true
1120         that inlined call frames wouldn't have an activation register. But that doesn't
1121         mean that people won't ask. Removing the assertion fixes a debug-only crash and
1122         has no impact on production code. This change adds a comment to that effect.
1123
1124         * dfg/DFGGraph.h:
1125         (JSC::DFG::Graph::uncheckedActivationRegisterFor):
1126
1127 2013-05-31  Julien Brianceau  <jbrianceau@nds.com>
1128
1129         [sh4] Fix Overflow case of branchMul32 in baseline JIT.
1130         https://bugs.webkit.org/show_bug.cgi?id=117057
1131
1132         Reviewed by Oliver Hunt.
1133
1134         Current implementation of Overflow case in branchMul32 performs an
1135         unsigned multiplication whereas a signed multiplication is expected.
1136
1137         * assembler/MacroAssemblerSH4.h:
1138         (JSC::MacroAssemblerSH4::branchMul32):
1139
1140 2013-05-31  Julien Brianceau  <jbrianceau@nds.com>
1141
1142         [sh4] Fix floating point comparisons in baseline JIT.
1143         https://bugs.webkit.org/show_bug.cgi?id=117066.
1144
1145         Reviewed by Oliver Hunt.
1146
1147         Current implementation of branchDouble function in baseline JIT is wrong
1148         for some conditions and overkill for others. For instance:
1149         - With DoubleGreaterThanOrEqual condition, branch will be taken if either
1150           operand is NaN with current implementation whereras it should not.
1151         - With DoubleNotEqualOrUnordered condition, performed NaN checks are
1152           useless (because comparison result is false if either operand is NaN).
1153
1154         * assembler/MacroAssemblerSH4.h:
1155         (JSC::MacroAssemblerSH4::branchDouble):
1156
1157 2013-05-31  Julien Brianceau  <jbrianceau@nds.com>
1158
1159         [sh4] Fix double floating point transfer in baseline JIT.
1160         https://bugs.webkit.org/show_bug.cgi?id=117054
1161
1162         Reviewed by Oliver Hunt.
1163
1164         In current implementation, dmovRegReg function transfers only one single
1165         FPRegister as PR=1 and SZ=0 in floating point status/control register.
1166         Double transfers must be performed with two fmov.s opcodes.
1167
1168         * assembler/MacroAssemblerSH4.h:
1169         (JSC::MacroAssemblerSH4::moveDouble):
1170         (JSC::MacroAssemblerSH4::addDouble): Handle (op2==dest) case properly.
1171         (JSC::MacroAssemblerSH4::sqrtDouble):
1172         * assembler/SH4Assembler.h:
1173         (JSC::SH4Assembler::fmovsRegReg):
1174
1175 2013-05-31  Julien Brianceau  <jbrianceau@nds.com>
1176
1177         [sh4] Handle branchType properly in branchTruncateDoubleToInt32.
1178         https://bugs.webkit.org/show_bug.cgi?id=117062
1179
1180         Reviewed by Oliver Hunt.
1181
1182         Current implementation of branchTruncateDoubleToInt32 is incorrect
1183         when branchType == BranchIfTruncateSuccessful in sh4 baseline JIT.
1184
1185         * assembler/MacroAssemblerSH4.h:
1186         (JSC::MacroAssemblerSH4::branchTruncateDoubleToInt32):
1187
1188 2013-05-31  Brent Fulgham  <bfulgham@apple.com>
1189
1190         [Windows] Unreviewed build fix for VS2005 builders.
1191
1192         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreExports.def: Add
1193         missing export for WTF::SHA1::computeHexDigest
1194
1195 2013-05-30  David Farler  <dfarler@apple.com>
1196
1197         Fix jscore-test when not using --sdk option with jsDriver.pl
1198         https://bugs.webkit.org/show_bug.cgi?id=116339
1199
1200         Reviewed by Joe Pecoraro.
1201
1202         * tests/mozilla/jsDriver.pl:
1203         (execute_tests):
1204         With each test, the shell_command needs to be started from scratch.
1205
1206         This fix will clear the shell_command and start over as before with
1207         the opt_arch option when not using --sdk with jsDriver.pl.
1208
1209 2013-05-30  Roger Fong  <roger_fong@apple.com>
1210
1211         Get rid of JavaScript exports file on AppleWin port.
1212         https://bugs.webkit.org/show_bug.cgi?id=117050.
1213
1214         Reviewed by Darin Adler.
1215
1216         Delete the JavaScriptCoreExportGenerator folder and remove dependencies.
1217         Start linking in WTF.lib now that it's a shared library.
1218
1219         * JavaScriptCore.vcxproj/JavaScriptCore.submit.sln:
1220         * JavaScriptCore.vcxproj/JavaScriptCore.vcxproj.filters:
1221         * JavaScriptCore.vcxproj/JavaScriptCoreCommon.props:
1222         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator: Removed.
1223         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGenerator.vcxproj: Removed.
1224         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGenerator.vcxproj.filters: Removed.
1225         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorBuildCmd.cmd: Removed.
1226         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorCommon.props: Removed.
1227         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorDebug.props: Removed.
1228         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorPostBuild.cmd: Removed.
1229         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorPreBuild.cmd: Removed.
1230         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorProduction.props: Removed.
1231         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorRelease.props: Removed.
1232         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExports.def.in: Removed.
1233         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/make-export-file-generator: Removed.
1234         * JavaScriptCore.vcxproj/jsc/jscCommon.props:
1235         * JavaScriptCore.vcxproj/testRegExp/testRegExp.vcxproj:
1236         * JavaScriptCore.vcxproj/testRegExp/testRegExp.vcxproj.filters:
1237         * JavaScriptCore.vcxproj/testRegExp/testRegExpCommon.props:
1238         * JavaScriptCore.vcxproj/testapi/testapiCommon.props:
1239
1240 2013-05-22  David Farler  <dfarler@apple.com>
1241
1242         Add --sdk option to jsDriver.pl to run with iOS Simulator
1243         https://bugs.webkit.org/show_bug.cgi?id=116339
1244
1245         Reviewed by David Kilzer.
1246
1247         * tests/mozilla/jsDriver.pl:
1248         (execute_tests):
1249         Prefix shell command with the path to the "sim" tool.
1250         (parse_args):
1251         Add -d / --sdk option.
1252         (usage):
1253         Help message for -d / --sdk option.
1254
1255 2013-05-30  Julien Brianceau  <jbrianceau@nds.com>
1256
1257         [sh4] Optimize NaN checks in LLINT for floating point comparisons.
1258         https://bugs.webkit.org/show_bug.cgi?id=117049
1259
1260         Reviewed by Oliver Hunt.
1261
1262         Use the fcmp/eq opcode in sh4 LLINT to test if a double is NaN.
1263         This is more efficient, doesn't require two tmp registers and requires
1264         less code than current implementation (which converts double to float,
1265         then checks 'E = Emax + 1' and 'f != 0').
1266
1267         * offlineasm/sh4.rb:
1268
1269 2013-05-30  Oliver Hunt  <oliver@apple.com>
1270
1271         JSCallbackObject does not correctly initialise the PropertySlot for getOwnPropertyDescriptor
1272         https://bugs.webkit.org/show_bug.cgi?id=117053
1273
1274         Reviewed by Mark Hahnenberg.
1275
1276         Set appropriate thisValue on the PropertySlot
1277
1278         * API/JSCallbackObjectFunctions.h:
1279         (JSC::::getOwnPropertyDescriptor):
1280         * API/tests/testapi.mm:
1281
1282 2013-05-29  Jeffrey Pfau  <jpfau@apple.com>
1283
1284         [Mac] Enable cache partitioning and the public suffix list on 10.8
1285         <rdar://problem/13679019>
1286
1287         Rubber-stamped by David Kilzer.
1288
1289         * Configurations/FeatureDefines.xcconfig:
1290
1291 2013-05-28  Brent Fulgham  <bfulgham@apple.com>
1292
1293         [Windows] Put correct byteCompile symbol in file. Previous version
1294         had an extra 'i' appended to the end.
1295
1296         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreExports.def:
1297         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExports.def.in:
1298
1299 2013-05-28  Brent Fulgham  <bfulgham@apple.com>
1300
1301         [Windows] Unreviewed build fix.  Remove ?byteCompile symbol that
1302         is no longer accessible during link.
1303
1304         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreExports.def:
1305         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExports.def.in:
1306
1307 2013-05-28  Gavin Barraclough  <barraclough@apple.com>
1308
1309         String(new Date(2010,10,1)) is wrong in KRAT, YAKT
1310         https://bugs.webkit.org/show_bug.cgi?id=106750
1311
1312         Reviewed by Darin Adler.
1313
1314         * runtime/JSDateMath.cpp:
1315         (JSC::msToGregorianDateTime):
1316             - Additional review comment fix.
1317
1318 2013-05-28  Brent Fulgham  <bfulgham@apple.com>
1319
1320         [Windows] Unreviewed build fix after r150833
1321
1322         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreExports.def:
1323         A CR/LF combination was lost in the file, combining two symbols.
1324
1325 2013-05-27  Gavin Barraclough  <barraclough@apple.com>
1326
1327         String(new Date(2010,10,1)) is wrong in KRAT, YAKT
1328         https://bugs.webkit.org/show_bug.cgi?id=106750
1329
1330         Reviewed by Darin Adler.
1331
1332         First part of a fix, simplfy date handling code, instead of operating separately
1333         on the UTC-standard and standard-DST offsets, just generate a combined UTC-local
1334         offset (this is what we actually need, and what the OS gives us).
1335
1336         * runtime/JSDateMath.cpp:
1337         (JSC::getLocalTimeOffset):
1338             - removed getUTCOffset, converted getDSTOffset -> getLocalTimeOffset
1339         (JSC::gregorianDateTimeToMS):
1340         (JSC::msToGregorianDateTime):
1341         (JSC::parseDateFromNullTerminatedCharacters):
1342             - call getLocalTimeOffset instead of getUTCOffset/getDSTOffset
1343         * runtime/VM.cpp:
1344         (JSC::VM::resetDateCache):
1345             - removed cachedUTCOffset, converted DSTOffsetCache -> LocalTimeOffsetCache
1346         * runtime/VM.h:
1347         (JSC::LocalTimeOffsetCache::LocalTimeOffsetCache):
1348         (JSC::LocalTimeOffsetCache::reset):
1349         (LocalTimeOffsetCache):
1350             - removed cachedUTCOffset, converted DSTOffsetCache -> LocalTimeOffsetCache
1351
1352 2013-05-28  Mark Hahnenberg  <mhahnenberg@apple.com>
1353
1354         r150199 is very wrong
1355         https://bugs.webkit.org/show_bug.cgi?id=116876
1356
1357         JSValue needs to protect its internal JSValueRef.
1358
1359         Reviewed by Darin Adler.
1360
1361         * API/JSValue.mm:
1362         (-[JSValue initWithValue:inContext:]):
1363         (-[JSValue dealloc]):
1364         * API/tests/testapi.mm: Added a simple test to make sure that we protect the
1365         underlying JavaScript value across garbage collections.
1366
1367 2013-05-27  Patrick Gansterer  <paroga@webkit.org>
1368
1369         Use ICU_INCLUDE_DIRS in BlackBerry CMake files
1370         https://bugs.webkit.org/show_bug.cgi?id=116210
1371
1372         Reviewed by Rob Buis.
1373
1374         Set and use the ICU_INCLUDE_DIRS variable to avoid
1375         duplicated adding of the ICU include directory.
1376
1377         * PlatformBlackBerry.cmake:
1378
1379 2013-05-27  Gabor Rapcsanyi  <rgabor@webkit.org>
1380
1381         MacroAssemblerARM should use xor to swap registers instead of move
1382         https://bugs.webkit.org/show_bug.cgi?id=116306
1383
1384         Reviewed by Zoltan Herczeg.
1385
1386         Change register swapping to xor from move and this way we don't need
1387         temporary register anymore.
1388
1389         * assembler/MacroAssemblerARM.h:
1390         (JSC::MacroAssemblerARM::swap):
1391
1392 2013-05-25  Filip Pizlo  <fpizlo@apple.com>
1393
1394         We broke (-2^31/-1)|0 in the DFG
1395         https://bugs.webkit.org/show_bug.cgi?id=116767
1396
1397         Reviewed by Andreas Kling.
1398
1399         The bug is that we were assuming that in the -2^31 case, we already had -2^31
1400         in the result register. This was a wrong assumption.
1401
1402         * dfg/DFGSpeculativeJIT.cpp:
1403         (JSC::DFG::SpeculativeJIT::compileIntegerArithDivForX86):
1404
1405 2013-05-24  Filip Pizlo  <fpizlo@apple.com>
1406
1407         We broke !(0/0)
1408         https://bugs.webkit.org/show_bug.cgi?id=116736
1409
1410         Reviewed by Gavin Barraclough.
1411
1412         * parser/ASTBuilder.h:
1413         (JSC::ASTBuilder::createLogicalNot):
1414         * runtime/JSCJSValueInlines.h:
1415         (JSC::JSValue::pureToBoolean):
1416
1417 2013-05-24  Julien Brianceau  <jbrianceau@nds.com>
1418
1419         [sh4] Optimize LLINT generated code and fix few bugs in baseline JIT.
1420         https://bugs.webkit.org/show_bug.cgi?id=116716
1421
1422         Reviewed by Geoffrey Garen.
1423
1424         * assembler/MacroAssemblerSH4.h:
1425         (JSC::MacroAssemblerSH4::mul32): Cosmetic changes.
1426         (JSC::MacroAssemblerSH4::convertInt32ToDouble): Absolute address was not dereferenced.
1427         (JSC::MacroAssemblerSH4::branch32): Absolute address was not dereferenced.
1428         (JSC::MacroAssemblerSH4::revertJumpReplacementToBranchPtrWithPatch): Use all 32 bits of pointer for revertJump call.
1429         * assembler/SH4Assembler.h:
1430         (JSC::SH4Assembler::revertJump): Use changePCrelativeAddress to patch the whole pointer.
1431         (JSC::SH4Assembler::linkJump): Cosmetic change.
1432         * offlineasm/sh4.rb: Optimize LLINT generated code.
1433
1434 2013-05-23  Peter Wang  <peter.wang@torchmobile.com.cn>
1435
1436         CLoop llint backend should not use the d8 register as scratch register
1437         https://bugs.webkit.org/show_bug.cgi?id=116019
1438
1439         Reviewed by Csaba Osztrogonác.
1440
1441         * offlineasm/cloop.rb:
1442
1443 2013-05-22  Peter Wang  <peter.wang@torchmobile.com.cn>
1444
1445         Use uninitialized register in "JIT::emit_op_neq_null" and "emit_op_eq_null"
1446         https://bugs.webkit.org/show_bug.cgi?id=116593
1447
1448         Reviewed by Filip Pizlo.
1449
1450         Generated instructions using uninitialized register. It's caused by a mistake of r126494.
1451
1452         * jit/JITOpcodes32_64.cpp:
1453         (JSC::JIT::emit_op_eq_null):
1454         (JSC::JIT::emit_op_neq_null):
1455
1456 2013-05-22  Filip Pizlo  <fpizlo@apple.com>
1457
1458         Fix indentation of CodeBlock.h
1459
1460         Rubber stampted by Mark Hahnenberg.
1461
1462         * bytecode/CodeBlock.h:
1463
1464 2013-05-22  Julien Brianceau  <jbrianceau@nds.com>
1465
1466         [sh4] Remove MacroAssemblerSH4.cpp file.
1467         https://bugs.webkit.org/show_bug.cgi?id=116596.
1468
1469         Reviewed by Geoffrey Garen.
1470
1471         Move linkCall and repatchCall implementations from MacroAssemblerSH4.cpp
1472         to MacroAssemblerSH4.h and remove MacroAssemblerSH4.cpp, as it is done
1473         for other architectures.
1474
1475         * GNUmakefile.list.am:
1476         * JavaScriptCore.xcodeproj/project.pbxproj:
1477         * Target.pri:
1478         * assembler/MacroAssemblerSH4.cpp: Removed.
1479         * assembler/MacroAssemblerSH4.h:
1480         (JSC::MacroAssemblerSH4::linkCall):
1481         (MacroAssemblerSH4):
1482         (JSC::MacroAssemblerSH4::repatchCall):
1483
1484 2013-05-21  Brent Fulgham  <bfulgham@apple.com>
1485
1486         [Windows] Unreviewed speculative fix for test-bots.
1487
1488         Add export declaration for WTFInvokeCrashHook to avoid runtime
1489         load error on test bots.
1490
1491         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreExports.def:
1492         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExports.def.in:
1493
1494 2013-05-21  Mark Lam  <mark.lam@apple.com>
1495
1496         Added missing assert condition for PositiveOrZero in ARM branch32().
1497         https://bugs.webkit.org/show_bug.cgi?id=116538.
1498
1499         Reviewed by Geoffrey Garen.
1500
1501         * assembler/MacroAssemblerARM.h:
1502         (JSC::MacroAssemblerARM::branchAdd32):
1503
1504 2013-05-20  Mark Hahnenberg  <mhahnenberg@apple.com>
1505
1506         Disable SuperRegion
1507         https://bugs.webkit.org/show_bug.cgi?id=116362
1508
1509         Rubber stamped by Geoff Garen.
1510
1511         * heap/Region.h:
1512
1513 2013-05-20  Oliver Hunt  <oliver@apple.com>
1514
1515         Make C API more robust against null contexts
1516         https://bugs.webkit.org/show_bug.cgi?id=116462
1517
1518         Reviewed by Anders Carlsson.
1519
1520         Handle null contexts in a non-crashy way.  It's a bug to ever call the
1521         API with a null context, and the absence of a context means we can't
1522         produce a meaningful result, so we still assert in debug builds.
1523
1524         Now where possible we detect and early return, returning null for any
1525         pointer type, NaN for doubles, and false for any boolean result.
1526
1527         * API/JSBase.cpp:
1528         (JSEvaluateScript):
1529         (JSCheckScriptSyntax):
1530         (JSReportExtraMemoryCost):
1531         * API/JSContextRef.cpp:
1532         (JSContextGetGlobalObject):
1533         (JSContextGetGroup):
1534         (JSContextGetGlobalContext):
1535         (JSContextCreateBacktrace):
1536         * API/JSObjectRef.cpp:
1537         (JSObjectMake):
1538         (JSObjectMakeFunctionWithCallback):
1539         (JSObjectMakeConstructor):
1540         (JSObjectMakeFunction):
1541         (JSObjectMakeArray):
1542         (JSObjectMakeDate):
1543         (JSObjectMakeError):
1544         (JSObjectMakeRegExp):
1545         (JSObjectGetPrototype):
1546         (JSObjectSetPrototype):
1547         (JSObjectHasProperty):
1548         (JSObjectGetProperty):
1549         (JSObjectSetProperty):
1550         (JSObjectGetPropertyAtIndex):
1551         (JSObjectSetPropertyAtIndex):
1552         (JSObjectDeleteProperty):
1553         (JSObjectCopyPropertyNames):
1554         * API/JSValueRef.cpp:
1555         (JSValueGetType):
1556         (JSValueIsUndefined):
1557         (JSValueIsNull):
1558         (JSValueIsBoolean):
1559         (JSValueIsNumber):
1560         (JSValueIsString):
1561         (JSValueIsObject):
1562         (JSValueIsObjectOfClass):
1563         (JSValueIsEqual):
1564         (JSValueIsStrictEqual):
1565         (JSValueIsInstanceOfConstructor):
1566         (JSValueMakeUndefined):
1567         (JSValueMakeNull):
1568         (JSValueMakeBoolean):
1569         (JSValueMakeNumber):
1570         (JSValueMakeString):
1571         (JSValueMakeFromJSONString):
1572         (JSValueCreateJSONString):
1573         (JSValueToBoolean):
1574         (JSValueToNumber):
1575         (JSValueToStringCopy):
1576         (JSValueToObject):
1577         (JSValueProtect):
1578         * API/JSWeakObjectMapRefPrivate.cpp:
1579
1580 2013-05-20  David Kilzer  <ddkilzer@apple.com>
1581
1582         Synchronize FeatureDefines.xcconfig
1583
1584         * Configurations/FeatureDefines.xcconfig: Remove
1585         ENABLE_LINK_PRERENDER.  This was missed in r150356.
1586
1587 2013-05-19  Anders Carlsson  <andersca@apple.com>
1588
1589         Remove link prerendering code
1590         https://bugs.webkit.org/show_bug.cgi?id=116415
1591
1592         Reviewed by Darin Adler.
1593
1594         This code was only used by Chromium and is dead now.
1595
1596         * Configurations/FeatureDefines.xcconfig:
1597
1598 2013-05-18  Patrick Gansterer  <paroga@webkit.org>
1599
1600         [CMake] Replace *_LIBRARY_NAME with *_OUTPUT_NAME
1601         https://bugs.webkit.org/show_bug.cgi?id=114554
1602
1603         Reviewed by Gyuyoung Kim.
1604
1605         Using variables as target names is very uncommon in CMake.
1606         The usual way to specify the name of the resulting binary
1607         is to set the OUTPUT_NAME target property.
1608
1609         * CMakeLists.txt:
1610         * shell/CMakeLists.txt:
1611
1612 2013-05-17  Patrick Gansterer  <paroga@webkit.org>
1613
1614         [CMake] Remove invalid include paths
1615         https://bugs.webkit.org/show_bug.cgi?id=116213
1616
1617         Reviewed by Gyuyoung Kim.
1618
1619         Since "${JAVASCRIPTCORE_DIR}/wtf" does not exist, it is safe
1620         to remove them from the list of include directories.
1621
1622         * PlatformEfl.cmake: Removed.
1623         * PlatformGTK.cmake: Removed.
1624
1625 2013-05-16  Patrick Gansterer  <paroga@webkit.org>
1626
1627         Consolidate lists in JavaScriptCore CMake files
1628         https://bugs.webkit.org/show_bug.cgi?id=115992
1629
1630         Reviewed by Gyuyoung Kim.
1631
1632         Move common files into the CMakeLists.txt to avoid duplicating the list of files.
1633         Also rebase the recently added GTK files to match the other CMake ports, since
1634         the submitted patch was based on an older version of the source tree.
1635
1636         * CMakeLists.txt:
1637         * PlatformEfl.cmake:
1638         * PlatformGTK.cmake:
1639         * shell/CMakeLists.txt:
1640         * shell/PlatformEfl.cmake:
1641         * shell/PlatformGTK.cmake:
1642
1643 2013-05-16  Geoffrey Garen  <ggaren@apple.com>
1644
1645         JSValue shouldn't protect/unprotect its context
1646         https://bugs.webkit.org/show_bug.cgi?id=116234
1647
1648         Reviewed by Mark Hahnenberg.
1649
1650         Our retain on _context is sufficient.
1651
1652         * API/JSValue.mm:
1653         (-[JSValue initWithValue:inContext:]):
1654         (-[JSValue dealloc]):
1655
1656 2013-05-15  Ryosuke Niwa  <rniwa@webkit.org>
1657
1658         Another Windows build fix attempt after r150160.
1659
1660         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreExports.def:
1661         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExports.def.in:
1662
1663 2013-05-15  Oliver Hunt  <oliver@apple.com>
1664
1665         RefCountedArray needs to use vector initialisers for its backing store
1666         https://bugs.webkit.org/show_bug.cgi?id=116194
1667
1668         Reviewed by Gavin Barraclough.
1669
1670         Use an out of line function to clear the exception stack to avoid
1671         needing to include otherwise unnecessary headers all over the place.
1672
1673         Everything else is just being updated to use that.
1674
1675         * bytecompiler/BytecodeGenerator.cpp:
1676         * interpreter/CallFrame.h:
1677         (JSC::ExecState::clearSupplementaryExceptionInfo):
1678         * interpreter/Interpreter.cpp:
1679         (JSC::Interpreter::addStackTraceIfNecessary):
1680         (JSC::Interpreter::throwException):
1681         * runtime/JSGlobalObject.cpp:
1682         (JSC::DynamicGlobalObjectScope::DynamicGlobalObjectScope):
1683         * runtime/VM.cpp:
1684         (JSC):
1685         (JSC::VM::clearExceptionStack):
1686         * runtime/VM.h:
1687         (VM):
1688         (JSC::VM::exceptionStack):
1689
1690 2013-05-15  Commit Queue  <commit-queue@webkit.org>
1691
1692         Unreviewed, rolling out r150051.
1693         http://trac.webkit.org/changeset/150051
1694         https://bugs.webkit.org/show_bug.cgi?id=116186
1695
1696         Broke all JSC tests on Mac and the author is unresponsive
1697         (Requested by rniwa on #webkit).
1698
1699         * JavaScriptCore.xcodeproj/project.pbxproj:
1700
1701 2013-05-15  Julien Brianceau  <jbrianceau@nds.com>
1702
1703         Remove savedTimeoutReg from JITStackFrame for sh4 base JIT.
1704         https://bugs.webkit.org/show_bug.cgi?id=116143
1705
1706         Reviewed by Geoffrey Garen.
1707
1708         Since r148119, timeoutCheckRegister is removed from baseline JIT.
1709         So we don't need to save r8 register in JITStackFrame anymore for sh4.
1710
1711         * jit/JITStubs.cpp:
1712         * jit/JITStubs.h:
1713         (JITStackFrame):
1714
1715 2013-05-15  Nico Weber  <thakis@chromium.org>
1716
1717         WebKit doesn't support MSVS2003 any more, remove preprocessor checks for older versions.
1718         https://bugs.webkit.org/show_bug.cgi?id=116157
1719
1720         Reviewed by Anders Carlsson.
1721
1722         Also remove a gcc3.2 workaround.
1723
1724         Merges parts of these two commits by the talented Nico Weber:
1725         https://chromium.googlesource.com/chromium/blink/+/3677e2f47348daeff405a40b6f90fbdf0654c2f5
1726         https://chromium.googlesource.com/chromium/blink/+/0fcd96c448dc30be1416dcc15713c53710c1a312
1727
1728         * os-win32/inttypes.h:
1729
1730 2013-05-13  Alvaro Lopez Ortega  <alvaro@alobbs.com>
1731
1732         Nightly build's jsc doesn't work without DYLD_FRAMEWORK...
1733         https://bugs.webkit.org/show_bug.cgi?id=79065
1734
1735         Reviewed by Darin Adler.
1736
1737         Fixes the build process so the depencencies of the jsc binary are
1738         modified before its copied to its target directory. In this way
1739         jsc should always use relative reference to the JavaScriptCore
1740         libraries.
1741
1742         * JavaScriptCore.xcodeproj/project.pbxproj: Fixes the commands in
1743         the "Copy Into Framework" target.
1744
1745 2013-05-13  Mark Hahnenberg  <mhahnenberg@apple.com>
1746
1747         Objective-C API: scanExternalObjectGraph should not create new JSVirtualMachine wrappers
1748         https://bugs.webkit.org/show_bug.cgi?id=116074
1749
1750         If scanExternalObjectGraph creates a new JSVirtualMachine wrapper during collection, when the
1751         scanExternalObjectGraph call finishes and the autorelease pool is drained we will dealloc the
1752         JSVirtualMachine which will cause us to try to take the API lock for the corresponding VM.
1753         If this happens on a GC thread other than the "main" thread, we will deadlock. The solution
1754         is to just check the VM cache, and if there is no JSVirtualMachine wrapper, return early.
1755
1756         Reviewed by Darin Adler.
1757
1758         * API/JSVirtualMachine.mm:
1759         (scanExternalObjectGraph):
1760
1761 2013-05-13  Benjamin Poulain  <benjamin@webkit.org>
1762
1763         Improve stringProtoFuncLastIndexOf for the prefix case
1764         https://bugs.webkit.org/show_bug.cgi?id=115952
1765
1766         Reviewed by Geoffrey Garen.
1767
1768         * runtime/StringPrototype.cpp:
1769         (JSC::stringProtoFuncLastIndexOf):
1770         Use the optimized string search when possible.
1771
1772         On Joseph Pecoraro's tests, this gives a ~30% speed improvement.
1773
1774 2013-05-13  Zalan Bujtas  <zalan@apple.com>
1775
1776         WebProcess consuming very high CPU on linkedin.com
1777         https://bugs.webkit.org/show_bug.cgi?id=115601
1778
1779         Reviewed by Andreas Kling.
1780
1781         Disable WEB_TIMING_MINIMAL.
1782         Turn off window.performance and performance.now(). Some JS frameworks expect
1783         additional Web Timing APIs, when performance.now() is available.
1784
1785         * Configurations/FeatureDefines.xcconfig:
1786
1787 2013-05-12  Anders Carlsson  <andersca@apple.com>
1788
1789         Stop including UnusedParam.h
1790         https://bugs.webkit.org/show_bug.cgi?id=116003
1791
1792         Reviewed by Sam Weinig.
1793
1794         UnusedParam.h is empty now so there's no need to include it anymore.
1795
1796         * API/APICast.h:
1797         * API/tests/JSNode.c:
1798         * API/tests/JSNodeList.c:
1799         * API/tests/minidom.c:
1800         * API/tests/testapi.c:
1801         * assembler/AbstractMacroAssembler.h:
1802         * assembler/MacroAssemblerCodeRef.h:
1803         * bytecode/CodeBlock.cpp:
1804         * heap/HandleStack.h:
1805         * interpreter/JSStackInlines.h:
1806         * jit/CompactJITCodeMap.h:
1807         * jit/ExecutableAllocator.h:
1808         * parser/SourceProvider.h:
1809         * runtime/DatePrototype.cpp:
1810         * runtime/JSNotAnObject.cpp:
1811         * runtime/JSSegmentedVariableObject.h:
1812         * runtime/JSVariableObject.h:
1813         * runtime/Options.cpp:
1814         * runtime/PropertyOffset.h:
1815
1816 2013-05-11  Martin Robinson  <mrobinson@igalia.com>
1817
1818         [GTK] Add a basic cmake build for WTF and JavaScriptCore
1819         https://bugs.webkit.org/show_bug.cgi?id=115967
1820
1821         Reviewed by Laszlo Gombos.
1822
1823         * PlatformGTK.cmake: Added.
1824         * shell/PlatformGTK.cmake: Added.
1825
1826 2013-05-10  Laszlo Gombos  <l.gombos@samsung.com>
1827
1828         Remove USE(OS_RANDOMNESS)
1829         https://bugs.webkit.org/show_bug.cgi?id=108095
1830
1831         Reviewed by Darin Adler.
1832
1833         Remove the USE(OS_RANDOMNESS) guard as it is turned on for all
1834         ports.
1835
1836         * jit/JIT.cpp:
1837         (JSC::JIT::JIT):
1838
1839 2013-05-10  Mark Hahnenberg  <mhahnenberg@apple.com>
1840
1841         Rename StructureCheckHoistingPhase to TypeCheckHoistingPhase
1842         https://bugs.webkit.org/show_bug.cgi?id=115938
1843
1844         We're going to add some more types of check hoisting soon, so let's have the right name here.
1845
1846         Rubber stamped by Filip Pizlo.
1847
1848         * CMakeLists.txt:
1849         * GNUmakefile.list.am:
1850         * JavaScriptCore.xcodeproj/project.pbxproj:
1851         * Target.pri:
1852         * dfg/DFGDriver.cpp:
1853         (JSC::DFG::compile):
1854         * dfg/DFGStructureCheckHoistingPhase.cpp: Removed.
1855         * dfg/DFGStructureCheckHoistingPhase.h: Removed.
1856         * dfg/DFGTypeCheckHoistingPhase.cpp: Copied from Source/JavaScriptCore/dfg/DFGStructureCheckHoistingPhase.cpp.
1857         (JSC::DFG::TypeCheckHoistingPhase::TypeCheckHoistingPhase):
1858         (JSC::DFG::performTypeCheckHoisting):
1859         * dfg/DFGTypeCheckHoistingPhase.h: Copied from Source/JavaScriptCore/dfg/DFGStructureCheckHoistingPhase.h.
1860
1861 2013-05-09  Christophe Dumez  <ch.dumez@sisa.samsung.com>
1862
1863         Unreviewed build fix after r149836.
1864
1865         It broke at least EFL and GTK builds. Move new static members initialization
1866         outside the class. Those need to have a definition outside the class because
1867         their address is used (e.g. CodeCacheMap::nonGlobalWorkingSetMaxEntries).
1868
1869         * runtime/CodeCache.cpp:
1870         (JSC):
1871         * runtime/CodeCache.h:
1872         (CodeCacheMap):
1873
1874 2013-05-08  Oliver Hunt  <oliver@apple.com>
1875
1876         Code cache stores bogus var references for functions in eval code
1877         https://bugs.webkit.org/show_bug.cgi?id=115747
1878
1879         Reviewed by Mark Hahnenberg.
1880
1881         Non-global eval now uses a per-CodeBlock cache, and only use it
1882         when we're at the top of a function's scope.  This means that we
1883         will no longer cache the parsing of a single string across
1884         multiple functions, and we won't cache when we're nested inside
1885         constructs like |with| and |catch| where previously we would, which
1886         is good because caching in those cases is unsound.
1887
1888         * bytecode/EvalCodeCache.h:
1889         (JSC):
1890         (JSC::EvalCodeCache::getSlow):
1891         (JSC::EvalCodeCache::get):
1892         * bytecode/UnlinkedCodeBlock.h:
1893         (JSC::UnlinkedCodeBlock::codeCacheForEval):
1894         (UnlinkedCodeBlock):
1895         (RareData):
1896         * debugger/Debugger.cpp:
1897         (JSC::evaluateInGlobalCallFrame):
1898         * debugger/DebuggerCallFrame.cpp:
1899         (JSC::DebuggerCallFrame::evaluate):
1900         * interpreter/Interpreter.cpp:
1901         (JSC::eval):
1902         * runtime/CodeCache.cpp:
1903         (JSC::CodeCache::CodeCache):
1904         (JSC::CodeCache::generateBytecode):
1905         (JSC):
1906         (JSC::CodeCache::getCodeBlock):
1907         * runtime/CodeCache.h:
1908         (JSC::CodeCacheMap::CodeCacheMap):
1909         (CodeCacheMap):
1910         (JSC::CodeCacheMap::canPruneQuickly):
1911         (JSC::CodeCacheMap::prune):
1912         (JSC::CodeCache::create):
1913         (CodeCache):
1914         * runtime/Executable.cpp:
1915         (JSC::EvalExecutable::EvalExecutable):
1916         (JSC::EvalExecutable::compileInternal):
1917         * runtime/Executable.h:
1918         (JSC::EvalExecutable::create):
1919         (EvalExecutable):
1920         * runtime/JSGlobalObject.cpp:
1921         (JSC::JSGlobalObject::createEvalCodeBlock):
1922         * runtime/JSGlobalObject.h:
1923         (JSGlobalObject):
1924         * runtime/JSGlobalObjectFunctions.cpp:
1925         (JSC::globalFuncEval):
1926         * runtime/VM.cpp:
1927         (JSC::VM::VM):
1928         * runtime/VM.h:
1929         (VM):
1930
1931 2013-05-08  Mark Hahnenberg  <mhahnenberg@apple.com>
1932
1933         DFGArrayMode::fromObserved is too liberal when it sees different Array and NonArray shapes
1934         https://bugs.webkit.org/show_bug.cgi?id=115805
1935
1936         Reviewed by Geoffrey Garen.
1937
1938         It checks the observed ArrayModes to see if we have seen any ArrayWith* first. If so, it assumes it's
1939         an Array::Array, even if we've also observed any NonArrayWith* in the ArrayProfile. This leads to the
1940         code generated by jumpSlowForUnwantedArrayMode to check the indexing type against (shape | IsArray)
1941         instead of just shape, which can cause us to exit a lot in the case that we saw a NonArray.
1942
1943         To fix this we need to add a case that checks for both ArrayWith* and NonArrayWith* cases first, which
1944         should then use Array::PossiblyArray, then do the checks we were already doing.
1945
1946         * bytecode/ArrayProfile.h:
1947         (JSC::hasSeenArray):
1948         (JSC::hasSeenNonArray):
1949         * dfg/DFGArrayMode.cpp:
1950         (JSC::DFG::ArrayMode::fromObserved):
1951
1952 2013-05-09  Joe Mason  <jmason@blackberry.com>
1953
1954         [BlackBerry] Set up logging buffer on start of jsc executable
1955         https://bugs.webkit.org/show_bug.cgi?id=114688
1956
1957         Reviewed by Rob Buis.
1958
1959         Internal PR: 322715
1960         Internally Reviewed By: Jeff Rogers
1961
1962         * jsc.cpp:
1963         (main): call BB::Platform::setupApplicationLogging
1964
1965 2013-05-08  Michael Saboff  <msaboff@apple.com>
1966
1967         JSC: There should be a disassembler for ARM Thumb 2
1968         https://bugs.webkit.org/show_bug.cgi?id=115827
1969
1970         Reviewed by Filip Pizlo.
1971
1972         Added a new disassembler for ARMv7 Thumb2 instructions for use by the JSC debugging
1973         and profiling code.  The opcode coverage is currently not complete.  It covers all
1974         of the integer instructions JSC currently emits, but only a limited number of
1975         floating point opcodes.  Currently that is just the 64 bit vmov and vmsr instructions.
1976
1977         The disassembler is structured as a base opcode class ARMv7DOpcode with sub-classes
1978         for each instruction group.  There is a public format method that does the bulk of
1979         the disassembly work.  There are two broad sub-classes, ARMv7D16BitOpcode and
1980         ARMv7D32BitOpcode, for the 16 bit and 32 bit opcodes.  There are sub-classes under
1981         those two classes for individual and related groups of opcodes.  Instructions are
1982         "dispatched" to the right subclass via two arrays of linked lists in the inner classes
1983         OpcodeGroup.  There is one such inner class for each ARMv7D16BitOpcode and ARMv7D32BitOpcode.
1984         Each OpcodeGroup has a mask and a pattern that it applies to the instruction to determine
1985         that it matches a particular group.  OpcodeGroup uses a static method to reinterpret_cast
1986         the Opcode object to the right base class for the instruction group for formatting.
1987         The cast eliminates the need of allocating an object for each decoded instruction.
1988         Unknown instructions are formatted as ".word 1234" or ".long 12345678" depending whether
1989         the instruction is 16 or 32 bit.
1990
1991         * JavaScriptCore.xcodeproj/project.pbxproj:
1992         * disassembler/ARMv7: Added.
1993         * disassembler/ARMv7/ARMv7DOpcode.cpp: Added.
1994         (ARMv7Disassembler):
1995         (OpcodeGroupInitializer):
1996         (JSC::ARMv7Disassembler::ARMv7DOpcode::init):
1997         (JSC::ARMv7Disassembler::ARMv7DOpcode::startITBlock):
1998         (JSC::ARMv7Disassembler::ARMv7DOpcode::saveITConditionAt):
1999         (JSC::ARMv7Disassembler::ARMv7DOpcode::fetchOpcode):
2000         (JSC::ARMv7Disassembler::ARMv7DOpcode::disassemble):
2001         (JSC::ARMv7Disassembler::ARMv7DOpcode::bufferPrintf):
2002         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendInstructionName):
2003         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendRegisterName):
2004         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendRegisterList):
2005         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendFPRegisterName):
2006         (JSC::ARMv7Disassembler::ARMv7D16BitOpcode::init):
2007         (JSC::ARMv7Disassembler::ARMv7D16BitOpcode::doDisassemble):
2008         (JSC::ARMv7Disassembler::ARMv7D16BitOpcode::defaultFormat):
2009         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddRegisterT2::format):
2010         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSPPlusImmediate::format):
2011         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractT1::format):
2012         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractImmediate3::format):
2013         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractImmediate8::format):
2014         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchConditionalT1::format):
2015         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchExchangeT1::format):
2016         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchT2::format):
2017         (JSC::ARMv7Disassembler::ARMv7DOpcodeCompareImmediateT1::format):
2018         (JSC::ARMv7Disassembler::ARMv7DOpcodeCompareRegisterT1::format):
2019         (JSC::ARMv7Disassembler::ARMv7DOpcodeCompareRegisterT2::format):
2020         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegisterT1::format):
2021         (JSC::ARMv7Disassembler::ARMv7DOpcodeGeneratePCRelativeAddress::format):
2022         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadFromLiteralPool::format):
2023         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterImmediate::format):
2024         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterOffsetT1::format):
2025         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterSPRelative::format):
2026         (JSC::ARMv7Disassembler::ARMv7DOpcodeLogicalImmediateT1::format):
2027         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscAddSubSP::format):
2028         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscBreakpointT1::format):
2029         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscByteHalfwordOps::format):
2030         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscCompareAndBranch::format):
2031         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscHint16::format):
2032         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscIfThenT1::format):
2033         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscPushPop::format):
2034         (JSC::ARMv7Disassembler::ARMv7DOpcodeMoveImmediateT1::format):
2035         (JSC::ARMv7Disassembler::ARMv7DOpcodeMoveRegisterT1::format):
2036         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::init):
2037         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::doDisassemble):
2038         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::defaultFormat):
2039         (JSC::ARMv7Disassembler::ARMv7DOpcodeConditionalBranchT3::format):
2040         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchOrBranchLink::format):
2041         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingModifiedImmediate::appendModifiedImmediate):
2042         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingModifiedImmediate::format):
2043         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingShiftedReg::appendImmShift):
2044         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingShiftedReg::format):
2045         (JSC::ARMv7Disassembler::ARMv7DOpcodeFPTransfer::format):
2046         (JSC::ARMv7Disassembler::ARMv7DOpcodeFPTransfer::appendFPRegister):
2047         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegShift::format):
2048         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegExtend::format):
2049         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegParallel::format):
2050         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegMisc::format):
2051         (JSC::ARMv7Disassembler::ARMv7DOpcodeHint32::format):
2052         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadRegister::format):
2053         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadSignedImmediate::format):
2054         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadUnsignedImmediate::format):
2055         (JSC::ARMv7Disassembler::ARMv7DOpcodeLongMultipleDivide::format):
2056         (JSC::ARMv7Disassembler::ARMv7DOpcodeUnmodifiedImmediate::format):
2057         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataPushPopSingle::format):
2058         (JSC::ARMv7Disassembler::ARMv7DOpcodeStoreSingleImmediate12::format):
2059         (JSC::ARMv7Disassembler::ARMv7DOpcodeStoreSingleImmediate8::format):
2060         (JSC::ARMv7Disassembler::ARMv7DOpcodeStoreSingleRegister::format):
2061         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMOVDoublePrecision::format):
2062         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMOVSinglePrecision::format):
2063         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMSR::format):
2064         * disassembler/ARMv7/ARMv7DOpcode.h: Added.
2065         (ARMv7Disassembler):
2066         (ARMv7DOpcode):
2067         (JSC::ARMv7Disassembler::ARMv7DOpcode::ARMv7DOpcode):
2068         (JSC::ARMv7Disassembler::ARMv7DOpcode::is32BitInstruction):
2069         (JSC::ARMv7Disassembler::ARMv7DOpcode::isFPInstruction):
2070         (JSC::ARMv7Disassembler::ARMv7DOpcode::conditionName):
2071         (JSC::ARMv7Disassembler::ARMv7DOpcode::shiftName):
2072         (JSC::ARMv7Disassembler::ARMv7DOpcode::inITBlock):
2073         (JSC::ARMv7Disassembler::ARMv7DOpcode::startingITBlock):
2074         (JSC::ARMv7Disassembler::ARMv7DOpcode::endITBlock):
2075         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendInstructionNameNoITBlock):
2076         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendSeparator):
2077         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendCharacter):
2078         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendString):
2079         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendShiftType):
2080         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendSignedImmediate):
2081         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendUnsignedImmediate):
2082         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendPCRelativeOffset):
2083         (JSC::ARMv7Disassembler::ARMv7DOpcode::appendShiftAmount):
2084         (ARMv7D16BitOpcode):
2085         (OpcodeGroup):
2086         (JSC::ARMv7Disassembler::ARMv7D16BitOpcode::OpcodeGroup::OpcodeGroup):
2087         (JSC::ARMv7Disassembler::ARMv7D16BitOpcode::OpcodeGroup::setNext):
2088         (JSC::ARMv7Disassembler::ARMv7D16BitOpcode::OpcodeGroup::next):
2089         (JSC::ARMv7Disassembler::ARMv7D16BitOpcode::OpcodeGroup::matches):
2090         (JSC::ARMv7Disassembler::ARMv7D16BitOpcode::OpcodeGroup::format):
2091         (JSC::ARMv7Disassembler::ARMv7D16BitOpcode::rm):
2092         (JSC::ARMv7Disassembler::ARMv7D16BitOpcode::rd):
2093         (JSC::ARMv7Disassembler::ARMv7D16BitOpcode::opcodeGroupNumber):
2094         (ARMv7DOpcodeAddRegisterT2):
2095         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddRegisterT2::rdn):
2096         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddRegisterT2::rm):
2097         (ARMv7DOpcodeAddSPPlusImmediate):
2098         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSPPlusImmediate::rd):
2099         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSPPlusImmediate::immediate8):
2100         (ARMv7DOpcodeAddSubtract):
2101         (ARMv7DOpcodeAddSubtractT1):
2102         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractT1::opName):
2103         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractT1::op):
2104         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractT1::rm):
2105         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractT1::rn):
2106         (ARMv7DOpcodeAddSubtractImmediate3):
2107         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractImmediate3::opName):
2108         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractImmediate3::op):
2109         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractImmediate3::immediate3):
2110         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractImmediate3::rn):
2111         (ARMv7DOpcodeAddSubtractImmediate8):
2112         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractImmediate8::opName):
2113         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractImmediate8::op):
2114         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractImmediate8::rdn):
2115         (JSC::ARMv7Disassembler::ARMv7DOpcodeAddSubtractImmediate8::immediate8):
2116         (ARMv7DOpcodeBranchConditionalT1):
2117         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchConditionalT1::condition):
2118         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchConditionalT1::offset):
2119         (ARMv7DOpcodeBranchExchangeT1):
2120         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchExchangeT1::opName):
2121         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchExchangeT1::rm):
2122         (ARMv7DOpcodeBranchT2):
2123         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchT2::immediate11):
2124         (ARMv7DOpcodeCompareImmediateT1):
2125         (JSC::ARMv7Disassembler::ARMv7DOpcodeCompareImmediateT1::rn):
2126         (JSC::ARMv7Disassembler::ARMv7DOpcodeCompareImmediateT1::immediate8):
2127         (ARMv7DOpcodeCompareRegisterT1):
2128         (JSC::ARMv7Disassembler::ARMv7DOpcodeCompareRegisterT1::rn):
2129         (ARMv7DOpcodeCompareRegisterT2):
2130         (JSC::ARMv7Disassembler::ARMv7DOpcodeCompareRegisterT2::rn):
2131         (JSC::ARMv7Disassembler::ARMv7DOpcodeCompareRegisterT2::rm):
2132         (ARMv7DOpcodeDataProcessingRegisterT1):
2133         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegisterT1::opName):
2134         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegisterT1::op):
2135         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegisterT1::rm):
2136         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegisterT1::rdn):
2137         (ARMv7DOpcodeGeneratePCRelativeAddress):
2138         (JSC::ARMv7Disassembler::ARMv7DOpcodeGeneratePCRelativeAddress::rd):
2139         (JSC::ARMv7Disassembler::ARMv7DOpcodeGeneratePCRelativeAddress::immediate8):
2140         (ARMv7DOpcodeLoadFromLiteralPool):
2141         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadFromLiteralPool::rt):
2142         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadFromLiteralPool::immediate8):
2143         (ARMv7DOpcodeLoadStoreRegisterImmediate):
2144         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterImmediate::opName):
2145         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterImmediate::op):
2146         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterImmediate::immediate5):
2147         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterImmediate::rn):
2148         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterImmediate::rt):
2149         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterImmediate::scale):
2150         (ARMv7DOpcodeLoadStoreRegisterImmediateWordAndByte):
2151         (ARMv7DOpcodeLoadStoreRegisterImmediateHalfWord):
2152         (ARMv7DOpcodeLoadStoreRegisterOffsetT1):
2153         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterOffsetT1::opName):
2154         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterOffsetT1::opB):
2155         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterOffsetT1::rm):
2156         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterOffsetT1::rn):
2157         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterOffsetT1::rt):
2158         (ARMv7DOpcodeLoadStoreRegisterSPRelative):
2159         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterSPRelative::opName):
2160         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterSPRelative::op):
2161         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterSPRelative::rt):
2162         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadStoreRegisterSPRelative::immediate8):
2163         (ARMv7DOpcodeLogicalImmediateT1):
2164         (JSC::ARMv7Disassembler::ARMv7DOpcodeLogicalImmediateT1::opName):
2165         (JSC::ARMv7Disassembler::ARMv7DOpcodeLogicalImmediateT1::op):
2166         (JSC::ARMv7Disassembler::ARMv7DOpcodeLogicalImmediateT1::immediate5):
2167         (ARMv7DOpcodeMiscAddSubSP):
2168         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscAddSubSP::opName):
2169         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscAddSubSP::op):
2170         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscAddSubSP::immediate7):
2171         (ARMv7DOpcodeMiscByteHalfwordOps):
2172         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscByteHalfwordOps::opName):
2173         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscByteHalfwordOps::op):
2174         (ARMv7DOpcodeMiscBreakpointT1):
2175         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscBreakpointT1::immediate8):
2176         (ARMv7DOpcodeMiscCompareAndBranch):
2177         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscCompareAndBranch::opName):
2178         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscCompareAndBranch::op):
2179         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscCompareAndBranch::immediate6):
2180         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscCompareAndBranch::rn):
2181         (ARMv7DOpcodeMiscHint16):
2182         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscHint16::opName):
2183         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscHint16::opA):
2184         (ARMv7DOpcodeMiscIfThenT1):
2185         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscIfThenT1::firstCondition):
2186         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscIfThenT1::mask):
2187         (ARMv7DOpcodeMiscPushPop):
2188         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscPushPop::opName):
2189         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscPushPop::op):
2190         (JSC::ARMv7Disassembler::ARMv7DOpcodeMiscPushPop::registerMask):
2191         (ARMv7DOpcodeMoveImmediateT1):
2192         (JSC::ARMv7Disassembler::ARMv7DOpcodeMoveImmediateT1::rd):
2193         (JSC::ARMv7Disassembler::ARMv7DOpcodeMoveImmediateT1::immediate8):
2194         (ARMv7DOpcodeMoveRegisterT1):
2195         (JSC::ARMv7Disassembler::ARMv7DOpcodeMoveRegisterT1::rd):
2196         (JSC::ARMv7Disassembler::ARMv7DOpcodeMoveRegisterT1::rm):
2197         (ARMv7D32BitOpcode):
2198         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::OpcodeGroup::OpcodeGroup):
2199         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::OpcodeGroup::setNext):
2200         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::OpcodeGroup::next):
2201         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::OpcodeGroup::matches):
2202         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::OpcodeGroup::format):
2203         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::rd):
2204         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::rm):
2205         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::rn):
2206         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::rt):
2207         (JSC::ARMv7Disassembler::ARMv7D32BitOpcode::opcodeGroupNumber):
2208         (ARMv7DOpcodeBranchRelative):
2209         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchRelative::sBit):
2210         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchRelative::j1):
2211         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchRelative::j2):
2212         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchRelative::immediate11):
2213         (ARMv7DOpcodeConditionalBranchT3):
2214         (JSC::ARMv7Disassembler::ARMv7DOpcodeConditionalBranchT3::offset):
2215         (JSC::ARMv7Disassembler::ARMv7DOpcodeConditionalBranchT3::condition):
2216         (JSC::ARMv7Disassembler::ARMv7DOpcodeConditionalBranchT3::immediate6):
2217         (ARMv7DOpcodeBranchOrBranchLink):
2218         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchOrBranchLink::offset):
2219         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchOrBranchLink::immediate10):
2220         (JSC::ARMv7Disassembler::ARMv7DOpcodeBranchOrBranchLink::isBL):
2221         (ARMv7DOpcodeDataProcessingLogicalAndRithmetic):
2222         (ARMv7DOpcodeDataProcessingModifiedImmediate):
2223         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingModifiedImmediate::opName):
2224         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingModifiedImmediate::op):
2225         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingModifiedImmediate::sBit):
2226         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingModifiedImmediate::immediate12):
2227         (ARMv7DOpcodeDataProcessingShiftedReg):
2228         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingShiftedReg::opName):
2229         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingShiftedReg::sBit):
2230         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingShiftedReg::op):
2231         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingShiftedReg::immediate5):
2232         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingShiftedReg::type):
2233         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingShiftedReg::tbBit):
2234         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingShiftedReg::tBit):
2235         (ARMv7DOpcodeDataProcessingReg):
2236         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingReg::op1):
2237         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingReg::op2):
2238         (ARMv7DOpcodeDataProcessingRegShift):
2239         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegShift::opName):
2240         (ARMv7DOpcodeDataProcessingRegExtend):
2241         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegExtend::opExtendName):
2242         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegExtend::opExtendAndAddName):
2243         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegExtend::rotate):
2244         (ARMv7DOpcodeDataProcessingRegParallel):
2245         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegParallel::opName):
2246         (ARMv7DOpcodeDataProcessingRegMisc):
2247         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataProcessingRegMisc::opName):
2248         (ARMv7DOpcodeHint32):
2249         (JSC::ARMv7Disassembler::ARMv7DOpcodeHint32::opName):
2250         (JSC::ARMv7Disassembler::ARMv7DOpcodeHint32::isDebugHint):
2251         (JSC::ARMv7Disassembler::ARMv7DOpcodeHint32::debugOption):
2252         (JSC::ARMv7Disassembler::ARMv7DOpcodeHint32::op):
2253         (ARMv7DOpcodeFPTransfer):
2254         (JSC::ARMv7Disassembler::ARMv7DOpcodeFPTransfer::opH):
2255         (JSC::ARMv7Disassembler::ARMv7DOpcodeFPTransfer::opL):
2256         (JSC::ARMv7Disassembler::ARMv7DOpcodeFPTransfer::rt):
2257         (JSC::ARMv7Disassembler::ARMv7DOpcodeFPTransfer::opC):
2258         (JSC::ARMv7Disassembler::ARMv7DOpcodeFPTransfer::opB):
2259         (JSC::ARMv7Disassembler::ARMv7DOpcodeFPTransfer::vd):
2260         (JSC::ARMv7Disassembler::ARMv7DOpcodeFPTransfer::vn):
2261         (ARMv7DOpcodeDataLoad):
2262         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataLoad::opName):
2263         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataLoad::op):
2264         (ARMv7DOpcodeLoadRegister):
2265         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadRegister::immediate2):
2266         (ARMv7DOpcodeLoadSignedImmediate):
2267         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadSignedImmediate::pBit):
2268         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadSignedImmediate::uBit):
2269         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadSignedImmediate::wBit):
2270         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadSignedImmediate::immediate8):
2271         (ARMv7DOpcodeLoadUnsignedImmediate):
2272         (JSC::ARMv7Disassembler::ARMv7DOpcodeLoadUnsignedImmediate::immediate12):
2273         (ARMv7DOpcodeLongMultipleDivide):
2274         (JSC::ARMv7Disassembler::ARMv7DOpcodeLongMultipleDivide::opName):
2275         (JSC::ARMv7Disassembler::ARMv7DOpcodeLongMultipleDivide::smlalOpName):
2276         (JSC::ARMv7Disassembler::ARMv7DOpcodeLongMultipleDivide::smlaldOpName):
2277         (JSC::ARMv7Disassembler::ARMv7DOpcodeLongMultipleDivide::smlsldOpName):
2278         (JSC::ARMv7Disassembler::ARMv7DOpcodeLongMultipleDivide::rdLo):
2279         (JSC::ARMv7Disassembler::ARMv7DOpcodeLongMultipleDivide::rdHi):
2280         (JSC::ARMv7Disassembler::ARMv7DOpcodeLongMultipleDivide::op1):
2281         (JSC::ARMv7Disassembler::ARMv7DOpcodeLongMultipleDivide::op2):
2282         (JSC::ARMv7Disassembler::ARMv7DOpcodeLongMultipleDivide::nBit):
2283         (JSC::ARMv7Disassembler::ARMv7DOpcodeLongMultipleDivide::mBit):
2284         (ARMv7DOpcodeDataPushPopSingle):
2285         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataPushPopSingle::opName):
2286         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataPushPopSingle::op):
2287         (ARMv7DOpcodeDataStoreSingle):
2288         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataStoreSingle::opName):
2289         (JSC::ARMv7Disassembler::ARMv7DOpcodeDataStoreSingle::op):
2290         (ARMv7DOpcodeStoreSingleImmediate12):
2291         (JSC::ARMv7Disassembler::ARMv7DOpcodeStoreSingleImmediate12::immediate12):
2292         (ARMv7DOpcodeStoreSingleImmediate8):
2293         (JSC::ARMv7Disassembler::ARMv7DOpcodeStoreSingleImmediate8::pBit):
2294         (JSC::ARMv7Disassembler::ARMv7DOpcodeStoreSingleImmediate8::uBit):
2295         (JSC::ARMv7Disassembler::ARMv7DOpcodeStoreSingleImmediate8::wBit):
2296         (JSC::ARMv7Disassembler::ARMv7DOpcodeStoreSingleImmediate8::immediate8):
2297         (ARMv7DOpcodeStoreSingleRegister):
2298         (JSC::ARMv7Disassembler::ARMv7DOpcodeStoreSingleRegister::immediate2):
2299         (ARMv7DOpcodeUnmodifiedImmediate):
2300         (JSC::ARMv7Disassembler::ARMv7DOpcodeUnmodifiedImmediate::opName):
2301         (JSC::ARMv7Disassembler::ARMv7DOpcodeUnmodifiedImmediate::op):
2302         (JSC::ARMv7Disassembler::ARMv7DOpcodeUnmodifiedImmediate::shBit):
2303         (JSC::ARMv7Disassembler::ARMv7DOpcodeUnmodifiedImmediate::bitNumOrSatImmediate):
2304         (JSC::ARMv7Disassembler::ARMv7DOpcodeUnmodifiedImmediate::immediate5):
2305         (JSC::ARMv7Disassembler::ARMv7DOpcodeUnmodifiedImmediate::immediate12):
2306         (JSC::ARMv7Disassembler::ARMv7DOpcodeUnmodifiedImmediate::immediate16):
2307         (ARMv7DOpcodeVMOVDoublePrecision):
2308         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMOVDoublePrecision::op):
2309         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMOVDoublePrecision::rt2):
2310         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMOVDoublePrecision::rt):
2311         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMOVDoublePrecision::vm):
2312         (ARMv7DOpcodeVMOVSinglePrecision):
2313         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMOVSinglePrecision::op):
2314         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMOVSinglePrecision::rt2):
2315         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMOVSinglePrecision::rt):
2316         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMOVSinglePrecision::vm):
2317         (ARMv7DOpcodeVMSR):
2318         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMSR::opL):
2319         (JSC::ARMv7Disassembler::ARMv7DOpcodeVMSR::rt):
2320         * disassembler/ARMv7Disassembler.cpp: Added.
2321         (JSC::tryToDisassemble):
2322
2323 2013-05-07  Julien Brianceau  <jbrianceau@nds.com>
2324
2325         Take advantage of pre-decrement and post-increment opcodes for sh4 base JIT.
2326         https://bugs.webkit.org/show_bug.cgi?id=115722
2327
2328         Reviewed by Oliver Hunt.
2329
2330         * assembler/MacroAssemblerSH4.h:
2331         (JSC::MacroAssemblerSH4::load8PostInc):
2332         (MacroAssemblerSH4):
2333         (JSC::MacroAssemblerSH4::load16Unaligned):
2334         (JSC::MacroAssemblerSH4::load16PostInc):
2335         (JSC::MacroAssemblerSH4::storeDouble):
2336         (JSC::MacroAssemblerSH4::load32WithUnalignedHalfWords):
2337         * assembler/SH4Assembler.h:
2338         (JSC::SH4Assembler::movwMemRegIn):
2339         (SH4Assembler):
2340         (JSC::SH4Assembler::movbMemRegIn):
2341         (JSC::SH4Assembler::printInstr):
2342
2343 2013-05-07  Anders Carlsson  <andersca@apple.com>
2344
2345         Remove AlwaysInline.h from WTF
2346         https://bugs.webkit.org/show_bug.cgi?id=115727
2347
2348         Reviewed by Brent Fulgham.
2349
2350         The macro that used to be in AlwaysInline.h is now in Compiler.h so there's no reason
2351         to keep AlwaysInline.h around anymore.
2352
2353         * jit/JSInterfaceJIT.h:
2354         * parser/Lexer.h:
2355         * runtime/JSCJSValue.h:
2356         * runtime/SymbolTable.h:
2357
2358 2013-05-07  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
2359
2360         HashTraits<RefPtr<P> >::PeekType should be raw pointer for better performance
2361         https://bugs.webkit.org/show_bug.cgi?id=115646
2362
2363         Reviewed by Darin Adler.
2364
2365         * bytecompiler/StaticPropertyAnalyzer.h:
2366         (JSC::StaticPropertyAnalyzer::putById):
2367             Updated accordingly to new HashMap<.., RefPtr>::get() semantics.
2368
2369 2013-05-06  Julien Brianceau  <jbrianceau@nds.com>
2370
2371         Misc bugfix and cleaning in sh4 base JIT.
2372         https://bugs.webkit.org/show_bug.cgi?id=115627
2373
2374         Reviewed by Oliver Hunt.
2375
2376         Get rid of loadX(RegisterID r0, RegisterID src, RegisterID dest) functions.
2377         Remove misplaced extuw() implementation from MacroAssemblerSH4.
2378         Add movbRegMemr0 and movwRegMemr0 functions in SH4Assembler.
2379
2380         * assembler/MacroAssemblerSH4.h:
2381         (JSC::MacroAssemblerSH4::add32): Skip operation when first operand is a zero immediate.
2382         (JSC::MacroAssemblerSH4::sub32): Skip operation when first operand is a zero immediate.
2383         (JSC::MacroAssemblerSH4::load32): Fix wrong usage of r0 register.
2384         (JSC::MacroAssemblerSH4::load8Signed): Handle "base == r0" case.
2385         (MacroAssemblerSH4):
2386         (JSC::MacroAssemblerSH4::load16): Handle "base == r0" case.
2387         (JSC::MacroAssemblerSH4::load16Unaligned): Use extuw() implementation from SH4Assembler.
2388         (JSC::MacroAssemblerSH4::load16Signed): Cosmetic change.
2389         (JSC::MacroAssemblerSH4::store8): Fix unhandled BaseIndex offset and handle (base == r0) case.
2390         (JSC::MacroAssemblerSH4::store16): Fix unhandled BaseIndex offset and handle (base == r0) case.
2391         (JSC::MacroAssemblerSH4::store32):
2392         * assembler/SH4Assembler.h:
2393         (JSC::SH4Assembler::movwRegMemr0):
2394         (SH4Assembler):
2395         (JSC::SH4Assembler::movbRegMemr0):
2396         (JSC::SH4Assembler::placeConstantPoolBarrier): Cosmetic change.
2397         (JSC::SH4Assembler::maxJumpReplacementSize):
2398         (JSC::SH4Assembler::replaceWithJump): Correct branch range and save an opcode.
2399         (JSC::SH4Assembler::printInstr):
2400
2401 2013-05-06  Anders Carlsson  <andersca@apple.com>
2402
2403         Stop using WTF::deleteAllValues in JavaScriptCore
2404         https://bugs.webkit.org/show_bug.cgi?id=115670
2405
2406         Reviewed by Oliver Hunt.
2407
2408         Change the Vectors used to Vectors of OwnPtrs instead.
2409
2410         * heap/DFGCodeBlocks.cpp:
2411         (JSC::DFGCodeBlocks::~DFGCodeBlocks):
2412         (JSC::DFGCodeBlocks::deleteUnmarkedJettisonedCodeBlocks):
2413
2414 2013-05-06  Andras Becsi  <andras.becsi@digia.com>
2415
2416         Build with GCC 4.8 fails because of -Wmaybe-uninitialized
2417         https://bugs.webkit.org/show_bug.cgi?id=115648
2418
2419         Reviewed by Michael Saboff.
2420
2421         Initialize values in Options::setOption since from
2422         there we end up calling OptionRange::init with
2423         uninitialized members.
2424
2425         * runtime/Options.cpp:
2426
2427 2013-05-06  Gabor Rapcsanyi  <rgabor@webkit.org>
2428
2429         JSC ARM traditional failing on Octane NavierStokes test
2430         https://bugs.webkit.org/show_bug.cgi?id=115626
2431
2432         Reviewed by Zoltan Herczeg.
2433
2434         Change the ARM traditional assembler to use double precision on value
2435         conversions.
2436
2437         * assembler/ARMAssembler.h:
2438
2439 2013-05-03  Michael Saboff  <msaboff@apple.com>
2440
2441         There should be a runtime option to constrain what functions get DFG compiled
2442         https://bugs.webkit.org/show_bug.cgi?id=115576
2443
2444         Reviewed by Mark Hahnenberg.
2445
2446         Added OptionRange to Options to allow checking that something is within an option
2447         or not.  The new OptionClass supports range strings in the form of [!]<low>[:<high>].
2448         If only one value is given, then it will be used for both low and high.  A leading
2449         '!' inverts the check.  If no range is given, then checking for a value within a range
2450         will always return true.  Added the option "bytecodeRangeToDFGCompile" that takes an
2451         OptionRange string to select the bytecode range of code blocks to DFG compile.
2452
2453         * dfg/DFGDriver.cpp:
2454         (JSC::DFG::compile): Added new check for bytecode count within bytecodeRangeToDFGCompile
2455         range.
2456         * runtime/Options.cpp:
2457         (JSC::parse): Added overloaded parse() for OptionRange.
2458         (JSC::OptionRange::init): Parse range string and then initialize the range.
2459         (JSC::OptionRange::isInRange): Function used by consumer to check if a value is within
2460         the specified range.
2461         (JSC::Options::dumpOption): Added code to dump OptionRange options.
2462         * runtime/Options.h:
2463         (OptionRange): New class.
2464         (JSC::OptionRange::operator= ): This is really used as a default ctor for use within
2465         the Option static array initialization.
2466         (JSC::OptionRange::rangeString): This is used for debug.  It assumes that the char*
2467         passed into OptionRange::init is valid when this function is called.
2468
2469 2013-05-02  Oliver Hunt  <oliver@apple.com>
2470
2471         Fix potential bug in lookup logic
2472         https://bugs.webkit.org/show_bug.cgi?id=115522
2473
2474         Reviewed by Mark Hahnenberg.
2475
2476         Though not a problem in practise, it is technically possible
2477         to inject an un-proxied global object into the scope chain
2478         via the C API.  This change makes sure that the scope walk
2479         in BytecodeGenerator actually limits itself to scopes that
2480         are statically bindable.
2481
2482         * bytecompiler/BytecodeGenerator.cpp:
2483         (JSC::BytecodeGenerator::resolve):
2484         * runtime/JSObject.h:
2485         (JSObject):
2486         (JSC):
2487         (JSC::JSObject::isStaticScopeObject):
2488
2489 2013-05-01  Roger Fong  <roger_fong@apple.com>
2490
2491         Set Path in makefile for AppleWin.
2492
2493         * JavaScriptCore.vcxproj/JavaScriptCore.make:
2494
2495 2013-05-01  Benjamin Poulain  <benjamin@webkit.org>
2496
2497         Remove the remaining wscript
2498         https://bugs.webkit.org/show_bug.cgi?id=115459
2499
2500         Reviewed by Andreas Kling.
2501
2502         * wscript: Removed.
2503
2504 2013-04-30  Mark Lam  <mark.lam@apple.com>
2505
2506         JSContextGroupSetExecutionTimeLimit() should not pass a callback to the
2507         VM watchdog if its client did not pass one in.
2508         https://bugs.webkit.org/show_bug.cgi?id=115461.
2509
2510         Reviewed by Geoffrey Garen.
2511
2512         * API/JSContextRef.cpp:
2513         (internalScriptTimeoutCallback):
2514         (JSContextGroupSetExecutionTimeLimit):
2515         * API/tests/testapi.c:
2516         (main):
2517         - Added test case when the time limit callback is 0.
2518         - Also updated a check to verify that a TerminatedExecutionException is
2519           thrown when the time out is cancelled.
2520         - Also fixed some cosmetic typos.
2521
2522 2013-04-30  Geoffrey Garen  <ggaren@apple.com>
2523
2524         Removed op_ensure_property_exists
2525         https://bugs.webkit.org/show_bug.cgi?id=115460
2526
2527         Reviewed by Mark Hahnenberg.
2528
2529         It was unused, and whatever it was once used for was not optimized.
2530
2531         * JavaScriptCore.order:
2532         * bytecode/CodeBlock.cpp:
2533         (JSC::CodeBlock::dumpBytecode):
2534         * bytecode/Opcode.h:
2535         (JSC::padOpcodeName):
2536         * jit/JIT.cpp:
2537         (JSC::JIT::privateCompileMainPass):
2538         * jit/JIT.h:
2539         * jit/JITOpcodes.cpp:
2540         * jit/JITOpcodes32_64.cpp:
2541         * jit/JITStubs.cpp:
2542         * jit/JITStubs.h:
2543         * llint/LLIntSlowPaths.cpp:
2544         * llint/LLIntSlowPaths.h:
2545         * llint/LowLevelInterpreter.asm:
2546
2547 2013-04-30  Oliver Hunt  <oliver@apple.com>
2548
2549         JSC Stack walking logic craches in the face of inlined functions triggering VM re-entry
2550         https://bugs.webkit.org/show_bug.cgi?id=115449
2551
2552         Reviewed by Geoffrey Garen.
2553
2554         Rename callframeishost to something that makes sense, and fix
2555         getCallerInfo to correctly handle inline functions calling into
2556         the VM.
2557
2558         * bytecode/CodeBlock.cpp:
2559         (JSC::CodeBlock::codeOriginForReturn):
2560           Make this more robust in the face of incorrect stack walking
2561         * interpreter/CallFrame.cpp:
2562         (JSC::CallFrame::trueCallerFrame):
2563           Everyone has to perform a codeblock() check before calling this
2564           so we might as well just do it here.
2565         * interpreter/Interpreter.cpp:
2566         (JSC::getCallerInfo):
2567
2568 2013-04-30  Julien Brianceau  <jbrianceau@nds.com>
2569
2570         Bug fixing in sh4 base JIT and LLINT.
2571         https://bugs.webkit.org/show_bug.cgi?id=115420
2572
2573         Reviewed by Oliver Hunt.
2574
2575         * assembler/MacroAssemblerSH4.h:
2576         (JSC::MacroAssemblerSH4::lshift32):
2577         (JSC::MacroAssemblerSH4::rshift32):
2578         (JSC::MacroAssemblerSH4::branchMul32):
2579         (JSC::MacroAssemblerSH4::urshift32):
2580         (JSC::MacroAssemblerSH4::replaceWithJump):
2581         (JSC::MacroAssemblerSH4::maxJumpReplacementSize):
2582         * assembler/SH4Assembler.h:
2583         (JSC::SH4Assembler::shldRegReg):
2584         (JSC::SH4Assembler::shadRegReg):
2585         (JSC::SH4Assembler::shalImm8r):
2586         (SH4Assembler):
2587         (JSC::SH4Assembler::sharImm8r):
2588         (JSC::SH4Assembler::maxJumpReplacementSize):
2589         (JSC::SH4Assembler::replaceWithJump):
2590         * offlineasm/sh4.rb:
2591
2592 2013-04-30  Geoffrey Garen  <ggaren@apple.com>
2593
2594         Objective-C JavaScriptCore API should publicly support bridging to C
2595         https://bugs.webkit.org/show_bug.cgi?id=115447
2596
2597         Reviewed by Mark Hahnenberg.
2598
2599         For consistency, I renamed
2600
2601             +[JSValue valueWithValue:] => +[JSValue valueWithJSValueRef]
2602             +[JSContext contextWithGlobalContextRef] => +[JSContext contextWithJSGlobalContextRef]
2603             -[JSContext globalContext] => -[JSContext JSGlobalContextRef]
2604
2605         I searched svn to verify that these functions don't have clients yet,
2606         so we won't break anything.
2607
2608         I also exported as public API
2609
2610             +[JSValue valueWithJSValueRef:]
2611             +[JSContext contextWithJSGlobalContextRef:]
2612
2613         It's hard to integrate with the C API without these.
2614
2615 2013-04-30  Commit Queue  <rniwa@webkit.org>
2616
2617         Unreviewed, rolling out r149349 and r149354.
2618         http://trac.webkit.org/changeset/149349
2619         http://trac.webkit.org/changeset/149354
2620         https://bugs.webkit.org/show_bug.cgi?id=115444
2621
2622          The Thumb version of compileSoftModulo make invalid use of
2623         registers (Requested by benjaminp on #webkit).
2624
2625         * CMakeLists.txt:
2626         * GNUmakefile.list.am:
2627         * JavaScriptCore.xcodeproj/project.pbxproj:
2628         * assembler/ARMv7Assembler.h:
2629         (ARMv7Assembler):
2630         * assembler/AbstractMacroAssembler.h:
2631         (JSC::isARMv7s):
2632         (JSC):
2633         * assembler/MacroAssemblerARMv7.cpp: Removed.
2634         * assembler/MacroAssemblerARMv7.h:
2635         (MacroAssemblerARMv7):
2636         * dfg/DFGFixupPhase.cpp:
2637         (JSC::DFG::FixupPhase::fixupNode):
2638         * dfg/DFGOperations.cpp:
2639         * dfg/DFGOperations.h:
2640         * dfg/DFGSpeculativeJIT.cpp:
2641         (JSC::DFG::SpeculativeJIT::compileSoftModulo):
2642         (DFG):
2643         (JSC::DFG::SpeculativeJIT::compileIntegerArithDivForARMv7s):
2644         * dfg/DFGSpeculativeJIT.h:
2645         (JSC::DFG::SpeculativeJIT::callOperation):
2646         (SpeculativeJIT):
2647         * dfg/DFGSpeculativeJIT32_64.cpp:
2648         (JSC::DFG::SpeculativeJIT::compile):
2649
2650 2013-04-30  Zalan Bujtas  <zalan@apple.com>
2651
2652         Animations fail to start on http://www.google.com/insidesearch/howsearchworks/thestory/
2653         https://bugs.webkit.org/show_bug.cgi?id=111244
2654
2655         Reviewed by David Kilzer.
2656
2657         Enable performance.now() as a minimal subset of Web Timing API.
2658         It returns DOMHighResTimeStamp, a monotonically increasing value representing the
2659         number of milliseconds from the start of the navigation of the current document.
2660         JS libraries use this API to check against the requestAnimationFrame() timestamp.
2661
2662         * Configurations/FeatureDefines.xcconfig:
2663
2664 2013-04-30  Zoltan Arvai  <zarvai@inf.u-szeged.hu>
2665
2666         Unreviewed. Speculative build fix on Qt Arm and Mips after r149349.
2667
2668         * dfg/DFGSpeculativeJIT.cpp:
2669         (JSC::DFG::SpeculativeJIT::compileSoftModulo):
2670
2671 2013-04-29  Cosmin Truta  <ctruta@blackberry.com>
2672
2673         [ARM] Expand the use of integer division
2674         https://bugs.webkit.org/show_bug.cgi?id=115138
2675
2676         Reviewed by Benjamin Poulain.
2677
2678         If availability of hardware integer division isn't known at compile
2679         time, check the CPU flags and decide at runtime whether to fall back
2680         to software. Currently, this OS-specific check is implemented on QNX.
2681
2682         Moreover, use operator % instead of fmod() in the calculation of the
2683         software modulo. Even when it's software-emulated, operator % is faster
2684         than fmod(): on ARM v7 QNX, without hardware division, we noticed
2685         >3% speedup on SunSpider.
2686
2687         * CMakeLists.txt:
2688         * GNUmakefile.list.am:
2689         * JavaScriptCore.xcodeproj/project.pbxproj:
2690         * assembler/ARMv7Assembler.h:
2691         (JSC::ARMv7Assembler::sdiv): Did not compile conditionally.
2692         (JSC::ARMv7Assembler::udiv): Ditto.
2693         * assembler/AbstractMacroAssembler.h:
2694         (JSC::isARMv7s): Removed.
2695         * assembler/MacroAssemblerARMv7.cpp: Added.
2696         (JSC::isIntegerDivSupported): Added.
2697         * assembler/MacroAssemblerARMv7.h:
2698         (JSC::MacroAssemblerARMv7::supportsIntegerDiv): Added.
2699         * dfg/DFGFixupPhase.cpp:
2700         (JSC::DFG::FixupPhase::fixupNode): Checked MacroAssembler::supportsIntegerDiv() in ArithDiv case.
2701         * dfg/DFGOperations.cpp:
2702         (JSC::DFG::operationModOnInts): Added.
2703         * dfg/DFGOperations.h:
2704         (JSC::DFG::Z_DFGOperation_ZZ): Added.
2705         * dfg/DFGSpeculativeJIT.cpp:
2706         (JSC::DFG::SpeculativeJIT::compileSoftModulo): Separated the X86-specific and ARM-specific codegen
2707         from the common implementation; used operationModOnInts on ARM.
2708         (JSC::DFG::SpeculativeJIT::compileIntegerArithDivForARM): Renamed from compileIntegerArithDivForARMv7.
2709         (JSC::DFG::SpeculativeJIT::compileArithMod): Allowed run-time detection of integer div on ARM.
2710         * dfg/DFGSpeculativeJIT.h:
2711         (JSC::DFG::SpeculativeJIT::callOperation): Added overloads with Z_DFGOperation_ZZ arguments.
2712         * dfg/DFGSpeculativeJIT32_64.cpp:
2713         (JSC::DFG::SpeculativeJIT::compile): Used compileIntegerArithDivForARM.
2714
2715 2013-04-29  Benjamin Poulain  <benjamin@webkit.org>
2716
2717         Unify the data access of StringImpl members from JavaScriptCore
2718         https://bugs.webkit.org/show_bug.cgi?id=115320
2719
2720         Reviewed by Andreas Kling.
2721
2722         DFG accesses the member infos by directly calling the methods on StringImpl,
2723         while the baseline JIT was using helper methods on ThunkHelpers.
2724
2725         Cut the middle man, and use StringImpl directly everywhere.
2726
2727         * jit/JITInlines.h:
2728         (JSC::JIT::emitLoadCharacterString):
2729         * jit/JITPropertyAccess.cpp:
2730         (JSC::JIT::stringGetByValStubGenerator):
2731         * jit/JITPropertyAccess32_64.cpp:
2732         (JSC::JIT::stringGetByValStubGenerator):
2733         * jit/JSInterfaceJIT.h:
2734         * jit/ThunkGenerators.cpp:
2735         (JSC::stringCharLoad):
2736
2737 2013-04-29  Benjamin Poulain  <bpoulain@apple.com>
2738
2739         Use push and pop for iOS math function thunks
2740         https://bugs.webkit.org/show_bug.cgi?id=115215
2741
2742         Reviewed by Filip Pizlo.
2743
2744         The iOS ABI is a little different than regular ARM ABI regarding stack alignment.
2745         The requirement is 4 bytes:
2746         "The ARM environment uses a stack that—at the point of function calls—is 4-byte aligned,
2747          grows downward, and contains local variables and a function’s parameters."
2748
2749         Subsequently, we can just use push and pop to preserve the link register.
2750
2751         * jit/ThunkGenerators.cpp:
2752
2753 2013-04-29  Brent Fulgham  <bfulgham@webkit.org>
2754
2755         [Windows, WinCairo] Get rid of last few pthread include/link references.
2756         https://bugs.webkit.org/show_bug.cgi?id=115375
2757
2758         Reviewed by Tim Horton.
2759
2760         * JavaScriptCore.vcproj/jsc/jscPostBuild.cmd:
2761         * JavaScriptCore.vcxproj/JavaScriptCoreCommon.props:
2762         * JavaScriptCore.vcxproj/LLInt/LLIntOffsetsExtractor/LLIntOffsetsExtractorCommon.props:
2763         * JavaScriptCore.vcxproj/jsc/jscCommon.props:
2764         * JavaScriptCore.vcxproj/testRegExp/testRegExpCommon.props:
2765         * JavaScriptCore.vcxproj/testapi/testapiCommon.props:
2766
2767 2013-04-29  Roger Fong  <roger_fong@apple.com>
2768
2769         Unreviewed. AppleWin VS2010 build fix.
2770
2771         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExports.def.in:
2772
2773 2013-04-26  Mark Hahnenberg  <mhahnenberg@apple.com>
2774
2775         ~BlockAllocator should ASSERT that it has no more Regions left
2776         https://bugs.webkit.org/show_bug.cgi?id=115287
2777
2778         Reviewed by Andreas Kling.
2779
2780         * heap/BlockAllocator.cpp:
2781         (JSC::BlockAllocator::~BlockAllocator):
2782         (JSC::BlockAllocator::allRegionSetsAreEmpty):
2783         * heap/BlockAllocator.h:
2784         (RegionSet):
2785         (JSC::BlockAllocator::RegionSet::isEmpty):
2786         (BlockAllocator):
2787
2788 2013-04-29  Mark Hahnenberg  <mhahnenberg@apple.com>
2789
2790         IndexingTypes should use hex
2791         https://bugs.webkit.org/show_bug.cgi?id=115286
2792
2793         Decimal is kind of confusing/hard to read because they're used as bit masks. Hex seems more appropriate.
2794
2795         Reviewed by Geoffrey Garen.
2796
2797         * runtime/IndexingType.h:
2798
2799 2013-04-29  Carlos Garcia Campos  <cgarcia@igalia.com>
2800
2801         Unreviewed. Fix make distcheck.
2802
2803         * GNUmakefile.list.am: Add missing headers files to compilation
2804         and offlineasm/sh4.rb script.
2805
2806 2013-04-28  Dean Jackson  <dino@apple.com>
2807
2808         [Mac] Disable canvas backing store scaling (HIGH_DPI_CANVAS)
2809         https://bugs.webkit.org/show_bug.cgi?id=115310
2810
2811         Reviewed by Simon Fraser.
2812
2813         Remove ENABLE_HIGH_DPI_CANVAS_macosx.
2814
2815         * Configurations/FeatureDefines.xcconfig:
2816
2817 2013-04-27  Darin Adler  <darin@apple.com>
2818
2819         Move from constructor and member function adoptCF/NS to free function adoptCF/NS.
2820         https://bugs.webkit.org/show_bug.cgi?id=115307
2821
2822         Reviewed by Geoffrey Garen.
2823
2824         * heap/HeapTimer.cpp:
2825         (JSC::HeapTimer::HeapTimer):
2826         * runtime/VM.cpp:
2827         (JSC::enableAssembler):
2828         Use adoptCF free function.
2829
2830 2013-04-27  Anders Carlsson  <andersca@apple.com>
2831
2832         Try to fix the Windows build.
2833
2834         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreExports.def:
2835
2836 2013-04-25  Geoffrey Garen  <ggaren@apple.com>
2837
2838         Cleaned up pre/post inc/dec in bytecode
2839         https://bugs.webkit.org/show_bug.cgi?id=115222
2840
2841         Reviewed by Filip Pizlo.
2842
2843         A few related changes here:
2844
2845         (*) Removed post_inc and post_dec. The two-result form was awkward to
2846         reason about. Being explicit about the intermediate mov and to_number
2847         reduces DFG overhead, removes some fragile ASSERTs from the DFG, and
2848         fixes a const bug. Plus, we get to blow away 262 lines of code.
2849
2850         (*) Renamed pre_inc and pre_dec to inc and dec, since there's only one
2851         version now.
2852
2853         (*) Renamed to_jsnumber to to_number, to match the ECMA name.
2854
2855         (*) Tightened up the codegen and runtime support for to_number.
2856
2857
2858         * JavaScriptCore.order: Order!
2859
2860         * bytecode/CodeBlock.cpp:
2861         (JSC::CodeBlock::dumpBytecode):
2862         * bytecode/Opcode.h:
2863         (JSC::padOpcodeName):
2864         * bytecompiler/BytecodeGenerator.cpp:
2865         (JSC::BytecodeGenerator::emitInc):
2866         (JSC::BytecodeGenerator::emitDec):
2867         * bytecompiler/BytecodeGenerator.h:
2868         (JSC::BytecodeGenerator::emitToNumber):
2869         (BytecodeGenerator): Removed post_inc and post_dec.
2870
2871         * bytecompiler/NodesCodegen.cpp:
2872         (JSC::emitPreIncOrDec): Updated for rename.
2873
2874         (JSC::emitPostIncOrDec): Issue an explicit mov and to_number when needed.
2875         These are rare, and they boil away in the DFG.
2876
2877         (JSC::PostfixNode::emitResolve):
2878         (JSC::PrefixNode::emitResolve): For const, use an explicit mov instead
2879         of any special forms. This fixes a bug where we would do string
2880         add/subtract instead of number.
2881
2882         * dfg/DFGByteCodeParser.cpp:
2883         (JSC::DFG::ByteCodeParser::parseBlock):
2884         * dfg/DFGCapabilities.h:
2885         (JSC::DFG::canCompileOpcode):
2886         * jit/JIT.cpp:
2887         (JSC::JIT::privateCompileMainPass):
2888         (JSC::JIT::privateCompileSlowCases):
2889         * jit/JIT.h:
2890         * jit/JITArithmetic.cpp:
2891         (JSC::JIT::emit_op_inc):
2892         (JSC::JIT::emitSlow_op_inc):
2893         (JSC::JIT::emit_op_dec):
2894         (JSC::JIT::emitSlow_op_dec):
2895         * jit/JITArithmetic32_64.cpp:
2896         (JSC::JIT::emit_op_inc):
2897         (JSC::JIT::emitSlow_op_inc):
2898         (JSC::JIT::emit_op_dec):
2899         (JSC::JIT::emitSlow_op_dec): Removed post_inc/dec, and updated for renames.
2900
2901         * jit/JITOpcodes.cpp:
2902         (JSC::JIT::emit_op_to_number):
2903         (JSC::JIT::emitSlow_op_to_number): Removed a test for number cells. There's
2904         no such thing!
2905
2906         * jit/JITOpcodes32_64.cpp:
2907         (JSC::JIT::emit_op_to_number): Use LowestTag to avoid making assumptions
2908         about the lowest valued tag.
2909
2910         (JSC::JIT::emitSlow_op_to_number): Updated for renames.
2911
2912         * jit/JITStubs.cpp:
2913         (JSC::DEFINE_STUB_FUNCTION):
2914         * jit/JITStubs.h:
2915         * llint/LLIntSlowPaths.cpp:
2916         (JSC::LLInt::LLINT_SLOW_PATH_DECL):
2917         * llint/LLIntSlowPaths.h:
2918         * llint/LowLevelInterpreter32_64.asm:
2919         * llint/LowLevelInterpreter64.asm:
2920         * parser/NodeConstructors.h:
2921         (JSC::UnaryPlusNode::UnaryPlusNode): Removed post_inc/dec, and updated for renames.
2922
2923         * runtime/Operations.cpp:
2924         (JSC::jsIsObjectType): Removed a test for number cells. There's
2925         no such thing!
2926
2927 2013-04-27  Julien Brianceau  <jbrianceau@nds.com>
2928
2929         REGRESSION(r149114): cache flush for SH4 arch may flush an extra page.
2930         https://bugs.webkit.org/show_bug.cgi?id=115305
2931
2932         Reviewed by Andreas Kling.
2933
2934         * assembler/SH4Assembler.h:
2935         (JSC::SH4Assembler::cacheFlush):
2936
2937 2013-04-26  Geoffrey Garen  <ggaren@apple.com>
2938
2939         Re-landing <http://trac.webkit.org/changeset/148999>
2940
2941             Filled out more cases of branch folding in bytecode when emitting
2942             expressions into a branching context
2943             https://bugs.webkit.org/show_bug.cgi?id=115057
2944
2945             Reviewed by Phil Pizlo.
2946
2947         We can't fold the number == 1 case to boolean because all non-zero numbers
2948         down-cast to true, but only 1 is == to true.
2949
2950 2013-04-26  Filip Pizlo  <fpizlo@apple.com>
2951
2952         Correct indentation of SymbolTable.h
2953
2954         Rubber stamped by Mark Hahnenberg.
2955
2956         * runtime/SymbolTable.h:
2957
2958 2013-04-26  Roger Fong  <roger_fong@apple.com>
2959
2960         Make Apple Windows VS2010 build results into and get dependencies from __32 suffixed folders.
2961         Make the DebugSuffix configuration use _debug dependencies.
2962
2963         * JavaScriptCore.vcxproj/JavaScriptCore.make:
2964         * JavaScriptCore.vcxproj/JavaScriptCore.vcxproj:
2965         * JavaScriptCore.vcxproj/JavaScriptCore.vcxproj.filters:
2966         * JavaScriptCore.vcxproj/JavaScriptCoreCF.props:
2967         * JavaScriptCore.vcxproj/JavaScriptCoreCommon.props:
2968         * JavaScriptCore.vcxproj/JavaScriptCoreDebug.props:
2969         * JavaScriptCore.vcxproj/JavaScriptCoreDebugCFLite.props:
2970         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGenerator.vcxproj:
2971         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGenerator.vcxproj.filters:
2972         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorBuildCmd.cmd:
2973         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorCommon.props:
2974         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorDebug.props:
2975         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorPostBuild.cmd:
2976         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorPreBuild.cmd:
2977         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorProduction.props:
2978         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorRelease.props:
2979         * JavaScriptCore.vcxproj/JavaScriptCoreGenerated.make:
2980         * JavaScriptCore.vcxproj/JavaScriptCoreGenerated.vcxproj:
2981         * JavaScriptCore.vcxproj/JavaScriptCoreGeneratedCommon.props:
2982         * JavaScriptCore.vcxproj/JavaScriptCoreGeneratedDebug.props:
2983         * JavaScriptCore.vcxproj/JavaScriptCoreGeneratedProduction.props:
2984         * JavaScriptCore.vcxproj/JavaScriptCoreGeneratedRelease.props:
2985         * JavaScriptCore.vcxproj/JavaScriptCorePostBuild.cmd:
2986         * JavaScriptCore.vcxproj/JavaScriptCorePreLink.cmd:
2987         * JavaScriptCore.vcxproj/JavaScriptCoreProduction.props:
2988         * JavaScriptCore.vcxproj/JavaScriptCoreRelease.props:
2989         * JavaScriptCore.vcxproj/JavaScriptCoreReleaseCFLite.props:
2990         * JavaScriptCore.vcxproj/LLInt/LLIntAssembly/LLIntAssembly.make:
2991         * JavaScriptCore.vcxproj/LLInt/LLIntAssembly/LLIntAssembly.vcxproj:
2992         * JavaScriptCore.vcxproj/LLInt/LLIntAssembly/build-LLIntAssembly.sh:
2993         * JavaScriptCore.vcxproj/LLInt/LLIntDesiredOffsets/LLIntDesiredOffsets.make:
2994         * JavaScriptCore.vcxproj/LLInt/LLIntDesiredOffsets/LLIntDesiredOffsets.vcxproj:
2995         * JavaScriptCore.vcxproj/LLInt/LLIntDesiredOffsets/build-LLIntDesiredOffsets.sh:
2996         * JavaScriptCore.vcxproj/LLInt/LLIntOffsetsExtractor/LLIntOffsetsExtractor.vcxproj:
2997         * JavaScriptCore.vcxproj/LLInt/LLIntOffsetsExtractor/LLIntOffsetsExtractorCommon.props:
2998         * JavaScriptCore.vcxproj/LLInt/LLIntOffsetsExtractor/LLIntOffsetsExtractorDebug.props:
2999         * JavaScriptCore.vcxproj/LLInt/LLIntOffsetsExtractor/LLIntOffsetsExtractorProduction.props:
3000         * JavaScriptCore.vcxproj/LLInt/LLIntOffsetsExtractor/LLIntOffsetsExtractorRelease.props:
3001         * JavaScriptCore.vcxproj/build-generated-files.sh:
3002         * JavaScriptCore.vcxproj/copy-files.cmd:
3003         * JavaScriptCore.vcxproj/jsc/jsc.vcxproj:
3004         * JavaScriptCore.vcxproj/jsc/jscCommon.props:
3005         * JavaScriptCore.vcxproj/jsc/jscDebug.props:
3006         * JavaScriptCore.vcxproj/jsc/jscPostBuild.cmd:
3007         * JavaScriptCore.vcxproj/jsc/jscPreLink.cmd:
3008         * JavaScriptCore.vcxproj/jsc/jscProduction.props:
3009         * JavaScriptCore.vcxproj/jsc/jscRelease.props:
3010         * JavaScriptCore.vcxproj/testRegExp/testRegExp.vcxproj:
3011         * JavaScriptCore.vcxproj/testRegExp/testRegExp.vcxproj.filters:
3012         * JavaScriptCore.vcxproj/testRegExp/testRegExpCommon.props:
3013         * JavaScriptCore.vcxproj/testRegExp/testRegExpDebug.props:
3014         * JavaScriptCore.vcxproj/testRegExp/testRegExpPostBuild.cmd:
3015         * JavaScriptCore.vcxproj/testRegExp/testRegExpPreLink.cmd:
3016         * JavaScriptCore.vcxproj/testRegExp/testRegExpProduction.props:
3017         * JavaScriptCore.vcxproj/testRegExp/testRegExpRelease.props:
3018         * JavaScriptCore.vcxproj/testapi/testapi.vcxproj:
3019         * JavaScriptCore.vcxproj/testapi/testapiCommon.props:
3020         * JavaScriptCore.vcxproj/testapi/testapiCommonCFLite.props:
3021         * JavaScriptCore.vcxproj/testapi/testapiDebug.props:
3022         * JavaScriptCore.vcxproj/testapi/testapiDebugCFLite.props:
3023         * JavaScriptCore.vcxproj/testapi/testapiPreLink.cmd:
3024         * JavaScriptCore.vcxproj/testapi/testapiProduction.props:
3025         * JavaScriptCore.vcxproj/testapi/testapiRelease.props:
3026         * JavaScriptCore.vcxproj/testapi/testapiReleaseCFLite.props:
3027
3028 2013-04-26  Roger Fong  <roger_fong@apple.com>
3029
3030         Disable sub-pixel layout on mac.
3031         https://bugs.webkit.org/show_bug.cgi?id=114999.
3032
3033         Reviewed by Simon Fraser.
3034
3035         * Configurations/FeatureDefines.xcconfig:
3036
3037 2013-04-26  Oliver Hunt  <oliver@apple.com>
3038
3039         Make stack tracing more robust
3040         https://bugs.webkit.org/show_bug.cgi?id=115272
3041
3042         Reviewed by Geoffrey Garen.
3043
3044         CallFrame already handles stack walking confusion robustly,
3045         so we should make sure that the actual walk handles that as well.
3046
3047         * interpreter/Interpreter.cpp:
3048         (JSC::getCallerInfo):
3049
3050 2013-04-26  Mark Hahnenberg  <mhahnenberg@apple.com>
3051
3052         REGRESSION(r149165): It made many tests crash on 32 bit
3053         https://bugs.webkit.org/show_bug.cgi?id=115227
3054
3055         Reviewed by Csaba Osztrogonác.
3056
3057         m_reservation is uninitialized when ENABLE(SUPER_REGION) is false.
3058
3059         * heap/SuperRegion.cpp:
3060         (JSC::SuperRegion::~SuperRegion):
3061
3062 2013-04-26  Julien Brianceau  <jbrianceau@nds.com>
3063
3064         Fix SH4 build broken since r149159.
3065         https://bugs.webkit.org/show_bug.cgi?id=115229
3066
3067         Add BranchTruncateType enum in SH4 port and handle it in branchTruncateDoubleToInt32.
3068
3069         Reviewed by Allan Sandfeld Jensen.
3070
3071         * assembler/MacroAssemblerSH4.h:
3072         (JSC::MacroAssemblerSH4::branchTruncateDoubleToInt32):
3073
3074 2013-04-25  Mark Hahnenberg  <mhahnenberg@apple.com>
3075
3076         SuperRegion doesn't call deallocate() on its PageReservation
3077         https://bugs.webkit.org/show_bug.cgi?id=115208
3078
3079         Reviewed by Geoffrey Garen.
3080
3081         It should. This doesn't cause us to leak physical memory, but it does cause us to leak virtual
3082         address space (and probably mach ports), which is also bad :-( FixedVMPoolExecutableAllocator
3083         also has this bug, but it doesn't matter much because there's only one instance of that class
3084         throughout the entire lifetime of the process, whereas each VM has its own SuperRegion.
3085
3086         * heap/SuperRegion.cpp:
3087         (JSC::SuperRegion::~SuperRegion):
3088         * heap/SuperRegion.h:
3089         (SuperRegion):
3090         * jit/ExecutableAllocatorFixedVMPool.cpp:
3091         (FixedVMPoolExecutableAllocator):
3092         (JSC::FixedVMPoolExecutableAllocator::~FixedVMPoolExecutableAllocator):
3093
3094 2013-04-25  Filip Pizlo  <fpizlo@apple.com>
3095
3096         DFG doesn't support to_jsnumber
3097         https://bugs.webkit.org/show_bug.cgi?id=115129
3098
3099         Reviewed by Geoffrey Garen.
3100
3101         Based on Oliver's patch. Implements to_jsnumber as Identity(Number:@thingy), and then does
3102         an optimization in Fixup to turn Identity(Number:) into Identity(Int32:) if the predictions
3103         tell us to. Identity is later turned into Phantom.
3104
3105         Also fixed BackPropMask, which appeared to have NodeDoesNotExit included in it. That's
3106         wrong; NodeDoesNotExit is not a backward propagation property.
3107
3108         Also fixed Identity to be marked as CanExit (i.e. not NodeDoesNotExit).
3109
3110         This more than doubles the FPS on ammo.
3111
3112         * dfg/DFGByteCodeParser.cpp:
3113         (JSC::DFG::ByteCodeParser::parseBlock):
3114         * dfg/DFGCapabilities.h:
3115         (JSC::DFG::canCompileOpcode):
3116         * dfg/DFGFixupPhase.cpp:
3117         (JSC::DFG::FixupPhase::fixupNode):
3118         (FixupPhase):
3119         (JSC::DFG::FixupPhase::observeUseKindOnNode):
3120         (JSC::DFG::FixupPhase::observeUseKindOnEdge):
3121         * dfg/DFGNodeFlags.h:
3122         (DFG):
3123         * dfg/DFGNodeType.h:
3124         (DFG):
3125         * dfg/DFGPredictionPropagationPhase.cpp:
3126         (JSC::DFG::PredictionPropagationPhase::propagate):
3127
3128 2013-04-24  Oliver Hunt  <oliver@apple.com>
3129
3130         Add support for Math.imul
3131         https://bugs.webkit.org/show_bug.cgi?id=115143
3132
3133         Reviewed by Filip Pizlo.
3134
3135         Add support for Math.imul, a thunk generator for Math.imul,
3136         and an intrinsic.
3137
3138         Fairly self explanatory set of changes, DFG intrinsics simply
3139         leverages the existing ValueToInt32 nodes.
3140
3141         * create_hash_table:
3142         * dfg/DFGAbstractState.cpp:
3143         (JSC::DFG::AbstractState::executeEffects):
3144         * dfg/DFGBackwardsPropagationPhase.cpp:
3145         (JSC::DFG::BackwardsPropagationPhase::propagate):
3146         * dfg/DFGByteCodeParser.cpp:
3147         (JSC::DFG::ByteCodeParser::handleIntrinsic):
3148         * dfg/DFGCSEPhase.cpp:
3149         (JSC::DFG::CSEPhase::performNodeCSE):
3150         * dfg/DFGFixupPhase.cpp:
3151         (JSC::DFG::FixupPhase::fixupNode):
3152         * dfg/DFGNodeType.h:
3153         (DFG):
3154         * dfg/DFGPredictionPropagationPhase.cpp:
3155         (JSC::DFG::PredictionPropagationPhase::propagate):
3156         * dfg/DFGSpeculativeJIT.cpp:
3157         (JSC::DFG::SpeculativeJIT::compileArithIMul):
3158         * dfg/DFGSpeculativeJIT.h:
3159         (SpeculativeJIT):
3160         * dfg/DFGSpeculativeJIT32_64.cpp:
3161         (JSC::DFG::SpeculativeJIT::compile):
3162         * dfg/DFGSpeculativeJIT64.cpp:
3163         (JSC::DFG::SpeculativeJIT::compile):
3164         * jit/ThunkGenerators.cpp:
3165         (JSC::imulThunkGenerator):
3166         (JSC):
3167         * jit/ThunkGenerators.h:
3168         (JSC):
3169         * runtime/Intrinsic.h:
3170         * runtime/MathObject.cpp:
3171         (JSC):
3172         (JSC::mathProtoFuncIMul):
3173         * runtime/VM.cpp:
3174         (JSC::thunkGeneratorForIntrinsic):
3175
3176 2013-04-25  Filip Pizlo  <fpizlo@apple.com>
3177
3178         Unreviewed, roll out http://trac.webkit.org/changeset/148999
3179         It broke http://kripken.github.io/ammo.js/examples/new/ammo.html
3180
3181         * JavaScriptCore.order:
3182         * bytecompiler/BytecodeGenerator.cpp:
3183         (JSC::BytecodeGenerator::emitNewArray):
3184         (JSC::BytecodeGenerator::emitThrowReferenceError):
3185         (JSC::BytecodeGenerator::emitReadOnlyExceptionIfNeeded):
3186         * bytecompiler/BytecodeGenerator.h:
3187         (JSC::BytecodeGenerator::shouldEmitProfileHooks):
3188         (BytecodeGenerator):
3189         * bytecompiler/NodesCodegen.cpp:
3190         (JSC):
3191         (JSC::NullNode::emitBytecode):
3192         (JSC::BooleanNode::emitBytecode):
3193         (JSC::NumberNode::emitBytecode):
3194         (JSC::StringNode::emitBytecode):
3195         (JSC::IfNode::emitBytecode):
3196         (JSC::IfElseNode::emitBytecode):
3197         * parser/ASTBuilder.h:
3198         (JSC::ASTBuilder::createIfStatement):
3199         (ASTBuilder):
3200         * parser/NodeConstructors.h:
3201         (JSC):
3202         (JSC::NullNode::NullNode):
3203         (JSC::BooleanNode::BooleanNode):
3204         (JSC::NumberNode::NumberNode):
3205         (JSC::StringNode::StringNode):
3206         (JSC::IfNode::IfNode):
3207         (JSC::IfElseNode::IfElseNode):
3208         * parser/Nodes.h:
3209         (JSC::ExpressionNode::isPure):
3210         (JSC::ExpressionNode::isSubtract):
3211         (StatementNode):
3212         (NullNode):
3213         (JSC::NullNode::isNull):
3214         (BooleanNode):
3215         (JSC::BooleanNode::isPure):
3216         (NumberNode):
3217         (JSC::NumberNode::value):
3218         (JSC::NumberNode::isPure):
3219         (StringNode):
3220         (JSC::StringNode::isPure):
3221         (JSC::StringNode::isString):
3222         (BinaryOpNode):
3223         (IfNode):
3224         (JSC):
3225         (IfElseNode):
3226         (ContinueNode):
3227         (BreakNode):
3228         * parser/Parser.cpp:
3229         (JSC::::parseIfStatement):
3230         * parser/ResultType.h:
3231         (ResultType):
3232         * runtime/JSCJSValueInlines.h:
3233         (JSC::JSValue::pureToBoolean):
3234         * runtime/JSCell.h:
3235         (JSCell):
3236         * runtime/JSCellInlines.h:
3237         (JSC):
3238
3239 2013-04-25  Filip Pizlo  <fpizlo@apple.com>
3240
3241         PreciseJumpTargets should treat loop_hint as a jump target
3242         https://bugs.webkit.org/show_bug.cgi?id=115209
3243
3244         Reviewed by Mark Hahnenberg.
3245
3246         I didn't add a test but I turned this into a release assertion. Running Octane is enough
3247         to trigger it.
3248
3249         * bytecode/PreciseJumpTargets.cpp:
3250         (JSC::computePreciseJumpTargets):
3251         * dfg/DFGByteCodeParser.cpp:
3252         (JSC::DFG::ByteCodeParser::parseBlock):
3253
3254 2013-04-25  Roman Zhuykov  <zhroma@ispras.ru>
3255
3256         Fix problems with processing negative zero on DFG.
3257         https://bugs.webkit.org/show_bug.cgi?id=113862
3258
3259         Reviewed by Filip Pizlo.
3260
3261         Fix NodeNeedsNegZero flag propagation in BackwardPropagationPhase.
3262         Function arithNodeFlags should not mask NodeNeedsNegZero flag for ArithNegate and DoubleAsInt32
3263         nodes and this flag should be always used to decide where we need to generate nezative-zero checks.
3264         Remove unnecessary negative-zero checks from integer ArithDiv on ARM.
3265         Also remove such checks from integer ArithMod on ARM and X86, and make them always to
3266         check not only "modulo_result == 0" but also "dividend < 0".
3267         Generate faster code for case when ArithMod operation divisor is constant power of 2 on ARMv7
3268         in the same way as on ARMv7s, and add negative-zero checks into this code when needed.
3269         Change speculationCheck ExitKind from Overflow to NegativeZero where applicable.
3270
3271         This shows 30% speedup of math-spectral-norm, and 5% speedup
3272         on SunSpider overall on ARMv7 Linux.
3273
3274         * assembler/MacroAssemblerARM.h:
3275         (JSC::MacroAssemblerARM::branchConvertDoubleToInt32):
3276         * assembler/MacroAssemblerARMv7.h:
3277         (JSC::MacroAssemblerARMv7::branchConvertDoubleToInt32):
3278         * assembler/MacroAssemblerMIPS.h:
3279         (JSC::MacroAssemblerMIPS::branchConvertDoubleToInt32):
3280         * assembler/MacroAssemblerSH4.h:
3281         (JSC::MacroAssemblerSH4::branchConvertDoubleToInt32):
3282         * assembler/MacroAssemblerX86Common.h:
3283         (JSC::MacroAssemblerX86Common::branchConvertDoubleToInt32):
3284         * dfg/DFGBackwardsPropagationPhase.cpp:
3285         (JSC::DFG::BackwardsPropagationPhase::isNotNegZero):
3286         (JSC::DFG::BackwardsPropagationPhase::isNotPosZero):
3287         (JSC::DFG::BackwardsPropagationPhase::propagate):
3288         * dfg/DFGNode.h:
3289         (JSC::DFG::Node::arithNodeFlags):
3290         * dfg/DFGSpeculativeJIT.cpp:
3291         (JSC::DFG::SpeculativeJIT::compileDoubleAsInt32):
3292         (JSC::DFG::SpeculativeJIT::compileSoftModulo):
3293         (JSC::DFG::SpeculativeJIT::compileArithNegate):
3294
3295 2013-04-25  Oliver Hunt  <oliver@apple.com>
3296
3297         Stack guards are too conservative
3298         https://bugs.webkit.org/show_bug.cgi?id=115147
3299
3300         Reviewed by Mark Hahnenberg.
3301
3302         Increase stack guard to closer to old size.
3303
3304         * interpreter/Interpreter.cpp:
3305         (JSC::Interpreter::StackPolicy::StackPolicy):
3306
3307 2013-04-25  Oliver Hunt  <oliver@apple.com>
3308
3309         Stack guards are too conservative
3310         https://bugs.webkit.org/show_bug.cgi?id=115147
3311
3312         Reviewed by Geoffrey Garen.
3313
3314         Reduce the limits and simplify the decision making.
3315
3316         * interpreter/Interpreter.cpp:
3317         (JSC::Interpreter::StackPolicy::StackPolicy):
3318
3319 2013-04-25  Nick Diego Yamane  <nick.yamane@openbossa.org>
3320
3321         JSC: Fix interpreter misbehavior in builds with JIT disabled
3322         https://bugs.webkit.org/show_bug.cgi?id=115190
3323
3324         Reviewed by Oliver Hunt.
3325
3326         Commit http://trac.webkit.org/changeset/147858 modified
3327         some details on how JS stack traces are built. The method
3328         "getLineNumberForCallFrame", renamed in that changeset to
3329         "getBytecodeOffsetForCallFrame" is always returning `0' when
3330         JIT is disabled
3331
3332         How to reproduce:
3333          - Build webkit with JIT disabled
3334          - Open MiniBrowser, for example, with http://google.com
3335          - In a debug build, WebProcess will hit the following ASSERT:
3336            Source/JavaScriptCore/bytecode/UnlinkedCodeBlock.cpp:279 ASSERT(low);
3337
3338         * interpreter/Interpreter.cpp:
3339         (JSC::getBytecodeOffsetForCallFrame):
3340
3341 2013-04-25  Oliver Hunt  <oliver@apple.com>
3342
3343         Make checkSyntax take a VM instead of an ExecState
3344
3345         RS=Tim
3346
3347         * jsc.cpp:
3348         (runInteractive):
3349         * runtime/Completion.cpp:
3350         (JSC::checkSyntax):
3351         * runtime/Completion.h:
3352         (JSC):
3353
3354 2013-04-25  Michael Saboff  <msaboff@apple.com>
3355
3356         32 Bit: Crash due to RegExpTest nodes not setting result type to Boolean
3357         https://bugs.webkit.org/show_bug.cgi?id=115188
3358
3359         Reviewed by Geoffrey Garen.
3360
3361         Changed the RegExpTest node to set the AbstractValue to boolean, since that
3362         what it is.
3363
3364         * dfg/DFGAbstractState.cpp:
3365         (JSC::DFG::AbstractState::executeEffects):
3366
3367 2013-04-25  Julien Brianceau  <jbrianceau@nds.com>
3368
3369         REGRESSION(r137994): Random crashes occur with SH4 JSC.
3370         https://bugs.webkit.org/show_bug.cgi?id=115167.
3371
3372         Reviewed by Oliver Hunt.
3373
3374         Since r137994, uncommited pages could be inside the area of memory in
3375         parameter of the cacheFlush function. That's why we have to flush each
3376         page separately to avoid a fail of the whole flush, if an uncommited page
3377         is in the area.
3378
3379         This patch is very similar to changeset 145194 made for ARMv7 architecture,
3380         see https://bugs.webkit.org/show_bug.cgi?id=111441 for further information.
3381
3382         * assembler/SH4Assembler.h:
3383         (JSC::SH4Assembler::cacheFlush):
3384
3385 2013-04-24  Mark Lam  <mark.lam@apple.com>
3386
3387         Add watchdog timer polling for the DFG.
3388         https://bugs.webkit.org/show_bug.cgi?id=115134.
3389
3390         Reviewed by Geoffrey Garen.
3391
3392         The strategy is to add a speculation check to the DFG generated code to
3393         test if the watchdog timer has fired or not. If the watchdog timer has
3394         fired, the generated code will do an OSR exit to the baseline JIT, and
3395         let it handle servicing the watchdog timer.
3396
3397         If the watchdog is not enabled, this speculation check will not be
3398         emitted.
3399
3400         * API/tests/testapi.c:
3401         (currentCPUTime_callAsFunction):
3402         (extendTerminateCallback):
3403         (main):
3404         - removed try/catch statements so that we can test the watchdog on the DFG.
3405         - added JS bindings to a native currentCPUTime() function so that the timeout
3406           tests can be more accurate.
3407         - also shortened the time values so that the tests can complete sooner.
3408
3409         * bytecode/ExitKind.h:
3410         * dfg/DFGAbstractState.cpp:
3411         (JSC::DFG::AbstractState::executeEffects):
3412         * dfg/DFGByteCodeParser.cpp:
3413         (JSC::DFG::ByteCodeParser::parseBlock):
3414         * dfg/DFGFixupPhase.cpp:
3415         (JSC::DFG::FixupPhase::fixupNode):
3416         * dfg/DFGNodeType.h:
3417         * dfg/DFGPredictionPropagationPhase.cpp:
3418         (JSC::DFG::PredictionPropagationPhase::propagate):
3419         * dfg/DFGSpeculativeJIT32_64.cpp:
3420         (JSC::DFG::SpeculativeJIT::compile):
3421         * dfg/DFGSpeculativeJIT64.cpp:
3422         (JSC::DFG::SpeculativeJIT::compile):
3423         * runtime/Watchdog.cpp:
3424         (JSC::Watchdog::setTimeLimit):
3425
3426 2013-04-24  Filip Pizlo  <fpizlo@apple.com>
3427
3428         Special thunks for math functions should work on ARMv7
3429         https://bugs.webkit.org/show_bug.cgi?id=115144
3430
3431         Reviewed by Gavin Barraclough and Oliver Hunt.
3432
3433         The only hard bit here was ensuring that we implemented the very special
3434         "cheap C call" convention on ARMv7.
3435
3436         * assembler/AbstractMacroAssembler.h:
3437         (JSC::isARMv7s):
3438         (JSC):
3439         (JSC::isX86):
3440         * dfg/DFGCommon.h:
3441         * jit/SpecializedThunkJIT.h:
3442         (SpecializedThunkJIT):
3443         (JSC::SpecializedThunkJIT::callDoubleToDoublePreservingReturn):
3444         * jit/ThunkGenerators.cpp:
3445         (JSC::floorThunkGenerator):
3446         (JSC::ceilThunkGenerator):
3447         (JSC::roundThunkGenerator):
3448         (JSC::expThunkGenerator):
3449         (JSC::logThunkGenerator):
3450
3451 2013-04-24  Julien Brianceau  <jbrianceau@nds.com>
3452
3453         Misc bugfix and cleaning in sh4 base JIT.
3454         https://bugs.webkit.org/show_bug.cgi?id=115022.
3455
3456         Reviewed by Oliver Hunt.
3457
3458         Remove unused add32() and sub32() with scratchreg parameter to avoid
3459         confusion as this function prototype means another behaviour.
3460         Remove unused "void push(Address)" function which seems quite buggy.
3461
3462         * assembler/MacroAssemblerSH4.h:
3463         (JSC::MacroAssemblerSH4::and32): Cosmetic change.
3464         (JSC::MacroAssemblerSH4::lshift32): Cosmetic change.
3465         (JSC::MacroAssemblerSH4::or32): Cosmetic change.
3466         (JSC::MacroAssemblerSH4::xor32): Cosmetic change.
3467         (MacroAssemblerSH4):
3468         (JSC::MacroAssemblerSH4::load32): Cosmetic change.
3469         (JSC::MacroAssemblerSH4::load8Signed): Fix invalid offset upper limit
3470         when using r0 register and cosmetic changes.
3471         (JSC::MacroAssemblerSH4::load8): Reuse load8Signed to avoid duplication.
3472         (JSC::MacroAssemblerSH4::load16): Fix invalid offset upper limit when
3473         using r0 register, fix missing offset shift and cosmetic changes.
3474         (JSC::MacroAssemblerSH4::store32): Cosmetic change.
3475         (JSC::MacroAssemblerSH4::branchAdd32): Store result value before branch.
3476
3477 2013-04-24  Patrick Gansterer  <paroga@webkit.org>
3478
3479         [WIN] Remove pthread from Visual Studio files in JavaScriptCore
3480         https://bugs.webkit.org/show_bug.cgi?id=114864
3481
3482         Reviewed by Brent Fulgham.
3483
3484         * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreCommon.vsprops:
3485         * JavaScriptCore.vcproj/LLIntOffsetsExtractor/LLIntOffsetsExtractorCommon.vsprops:
3486         * JavaScriptCore.vcproj/jsc/jscCommon.vsprops:
3487         * JavaScriptCore.vcproj/testRegExp/testRegExpCommon.vsprops:
3488         * JavaScriptCore.vcproj/testapi/testapiCommon.vsprops:
3489         * JavaScriptCore.vcxproj/JavaScriptCoreCommon.props:
3490         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorCommon.props:
3491         * JavaScriptCore.vcxproj/LLInt/LLIntOffsetsExtractor/LLIntOffsetsExtractorCommon.props:
3492         * JavaScriptCore.vcxproj/jsc/jscCommon.props:
3493         * JavaScriptCore.vcxproj/testRegExp/testRegExpCommon.props:
3494         * JavaScriptCore.vcxproj/testapi/testapiCommon.props:
3495         * JavaScriptCore.vcxproj/testapi/testapiCommonCFLite.props:
3496
3497 2013-04-24  Filip Pizlo  <fpizlo@apple.com>
3498
3499         DFG should keep the operand to create_this alive if it's emitting code for create_this
3500         https://bugs.webkit.org/show_bug.cgi?id=115133
3501
3502         Reviewed by Mark Hahnenberg.
3503
3504         The DFG must model bytecode liveness, or else OSR exit is going to have a really bad time.
3505
3506         * dfg/DFGByteCodeParser.cpp:
3507         (JSC::DFG::ByteCodeParser::parseBlock):
3508
3509 2013-04-24  Roger Fong  <roger_fong@apple.com>
3510
3511         Have VS2010 WebKit solution look in WebKit_Libraries/lib32 for dependencies.
3512
3513         * JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExportGeneratorPostBuild.cmd:
3514         * JavaScriptCore.vcxproj/JavaScriptCorePreLink.cmd:
3515         * JavaScriptCore.vcxproj/jsc/jscPostBuild.cmd:
3516         * JavaScriptCore.vcxproj/jsc/jscPreLink.cmd:
3517         * JavaScriptCore.vcxproj/testRegExp/testRegExp.vcxproj.filters:
3518         * JavaScriptCore.vcxproj/testRegExp/testRegExpPostBuild.cmd:
3519         * JavaScriptCore.vcxproj/testRegExp/testRegExpPreLink.cmd:
3520         * JavaScriptCore.vcxproj/testapi/testapiPreLink.cmd:
3521
3522 2013-04-24  Geoffrey Garen  <ggaren@apple.com>
3523
3524         32-bit build fix.
3525
3526         Unreviewed.
3527
3528         * dfg/DFGSpeculativeJIT.cpp:
3529         (JSC::DFG::SpeculativeJIT::compilePeepHoleBooleanBranch): Explicitly
3530         truncate to 32-bit to avoid compiler warnings. It's safe to truncate
3531         because the payload of a boolean is the low bits on both 64-bit and 32-bit.
3532
3533 2013-04-23  Geoffrey Garen  <ggaren@apple.com>
3534
3535         Filled out more cases of branch folding in the DFG
3536         https://bugs.webkit.org/show_bug.cgi?id=115088
3537
3538         Reviewed by Oliver Hunt.
3539
3540         No change on the benchmarks we track, but a 3X speedup on a
3541         microbenchmark that uses these techniques.
3542
3543         * dfg/DFGByteCodeParser.cpp:
3544         (JSC::DFG::ByteCodeParser::parseBlock): (!/=)= and (!/=)== can constant
3545         fold all types, not just numbers, because true constants have no
3546         side effects when type-converted at runtime.
3547
3548         * dfg/DFGFixupPhase.cpp:
3549         (JSC::DFG::FixupPhase::fixupNode):
3550         * dfg/DFGNode.h:
3551         (JSC::DFG::Node::shouldSpeculateBoolean): Added support for fixing up
3552         boolean uses, like we do for other types like number.
3553
3554         * dfg/DFGSpeculativeJIT.cpp:
3555         (JSC::DFG::SpeculativeJIT::compilePeepHoleBooleanBranch):
3556         (JSC::DFG::SpeculativeJIT::compilePeepHoleBranch):
3557         (JSC::DFG::SpeculativeJIT::compare):
3558         (JSC::DFG::SpeculativeJIT::compileStrictEq):
3559         (JSC::DFG::SpeculativeJIT::compileBooleanCompare): Peephole fuse
3560         boolean compare and/or compare-branch, now that we have the types for
3561         them.
3562
3563         * dfg/DFGSpeculativeJIT.h: Updated declarations.
3564
3565 == Rolled over to ChangeLog-2013-04-24 ==