JavaScriptCore/jit/JITArithmetic32_64.cpp
changeset 0 4f2f89ce4247
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/JavaScriptCore/jit/JITArithmetic32_64.cpp	Fri Sep 17 09:02:29 2010 +0300
@@ -0,0 +1,1409 @@
+/*
+* Copyright (C) 2008 Apple Inc. All rights reserved.
+*
+* Redistribution and use in source and binary forms, with or without
+* modification, are permitted provided that the following conditions
+* are met:
+* 1. Redistributions of source code must retain the above copyright
+*    notice, this list of conditions and the following disclaimer.
+* 2. Redistributions in binary form must reproduce the above copyright
+*    notice, this list of conditions and the following disclaimer in the
+*    documentation and/or other materials provided with the distribution.
+*
+* THIS SOFTWARE IS PROVIDED BY APPLE INC. ``AS IS'' AND ANY
+* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+* PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL APPLE INC. OR
+* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
+* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*/
+
+#include "config.h"
+
+#if ENABLE(JIT)
+#if USE(JSVALUE32_64)
+#include "JIT.h"
+
+#include "CodeBlock.h"
+#include "JITInlineMethods.h"
+#include "JITStubCall.h"
+#include "JITStubs.h"
+#include "JSArray.h"
+#include "JSFunction.h"
+#include "Interpreter.h"
+#include "ResultType.h"
+#include "SamplingTool.h"
+
+#ifndef NDEBUG
+#include <stdio.h>
+#endif
+
+using namespace std;
+
+namespace JSC {
+
+void JIT::emit_op_negate(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned src = currentInstruction[2].u.operand;
+
+    emitLoad(src, regT1, regT0);
+
+    Jump srcNotInt = branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag));
+    addSlowCase(branchTest32(Zero, regT0, Imm32(0x7fffffff)));
+    neg32(regT0);
+    emitStoreInt32(dst, regT0, (dst == src));
+
+    Jump end = jump();
+
+    srcNotInt.link(this);
+    addSlowCase(branch32(Above, regT1, Imm32(JSValue::LowestTag)));
+
+    xor32(Imm32(1 << 31), regT1);
+    store32(regT1, tagFor(dst));
+    if (dst != src)
+        store32(regT0, payloadFor(dst));
+
+    end.link(this);
+}
+
+void JIT::emitSlow_op_negate(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+
+    linkSlowCase(iter); // 0x7fffffff check
+    linkSlowCase(iter); // double check
+
+    JITStubCall stubCall(this, cti_op_negate);
+    stubCall.addArgument(regT1, regT0);
+    stubCall.call(dst);
+}
+
+void JIT::emit_op_jnless(Instruction* currentInstruction)
+{
+    unsigned op1 = currentInstruction[1].u.operand;
+    unsigned op2 = currentInstruction[2].u.operand;
+    unsigned target = currentInstruction[3].u.operand;
+
+    JumpList notInt32Op1;
+    JumpList notInt32Op2;
+
+    // Character less.
+    if (isOperandConstantImmediateChar(op1)) {
+        emitLoad(op2, regT1, regT0);
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::CellTag)));
+        JumpList failures;
+        emitLoadCharacterString(regT0, regT0, failures);
+        addSlowCase(failures);
+        addJump(branch32(LessThanOrEqual, regT0, Imm32(asString(getConstantOperand(op1))->tryGetValue()[0])), target);
+        return;
+    }
+    if (isOperandConstantImmediateChar(op2)) {
+        emitLoad(op1, regT1, regT0);
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::CellTag)));
+        JumpList failures;
+        emitLoadCharacterString(regT0, regT0, failures);
+        addSlowCase(failures);
+        addJump(branch32(GreaterThanOrEqual, regT0, Imm32(asString(getConstantOperand(op2))->tryGetValue()[0])), target);
+        return;
+    }
+    if (isOperandConstantImmediateInt(op1)) {
+        // Int32 less.
+        emitLoad(op2, regT3, regT2);
+        notInt32Op2.append(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+        addJump(branch32(LessThanOrEqual, regT2, Imm32(getConstantOperand(op1).asInt32())), target);
+    } else if (isOperandConstantImmediateInt(op2)) {
+        emitLoad(op1, regT1, regT0);
+        notInt32Op1.append(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+        addJump(branch32(GreaterThanOrEqual, regT0, Imm32(getConstantOperand(op2).asInt32())), target);
+    } else {
+        emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+        notInt32Op1.append(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+        notInt32Op2.append(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+        addJump(branch32(GreaterThanOrEqual, regT0, regT2), target);
+    }
+
+    if (!supportsFloatingPoint()) {
+        addSlowCase(notInt32Op1);
+        addSlowCase(notInt32Op2);
+        return;
+    }
+    Jump end = jump();
+
+    // Double less.
+    emitBinaryDoubleOp(op_jnless, target, op1, op2, OperandTypes(), notInt32Op1, notInt32Op2, !isOperandConstantImmediateInt(op1), isOperandConstantImmediateInt(op1) || !isOperandConstantImmediateInt(op2));
+    end.link(this);
+}
+
+void JIT::emitSlow_op_jnless(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned op1 = currentInstruction[1].u.operand;
+    unsigned op2 = currentInstruction[2].u.operand;
+    unsigned target = currentInstruction[3].u.operand;
+
+    if (isOperandConstantImmediateChar(op1) || isOperandConstantImmediateChar(op2)) {
+        linkSlowCase(iter);
+        linkSlowCase(iter);
+        linkSlowCase(iter);
+        linkSlowCase(iter);
+    } else {
+        if (!supportsFloatingPoint()) {
+            if (!isOperandConstantImmediateInt(op1) && !isOperandConstantImmediateInt(op2))
+                linkSlowCase(iter); // int32 check
+            linkSlowCase(iter); // int32 check
+        } else {
+            if (!isOperandConstantImmediateInt(op1)) {
+                linkSlowCase(iter); // double check
+                linkSlowCase(iter); // int32 check
+            }
+            if (isOperandConstantImmediateInt(op1) || !isOperandConstantImmediateInt(op2))
+                linkSlowCase(iter); // double check
+        }
+    }
+
+    JITStubCall stubCall(this, cti_op_jless);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call();
+    emitJumpSlowToHot(branchTest32(Zero, regT0), target);
+}
+
+void JIT::emit_op_jless(Instruction* currentInstruction)
+{
+    unsigned op1 = currentInstruction[1].u.operand;
+    unsigned op2 = currentInstruction[2].u.operand;
+    unsigned target = currentInstruction[3].u.operand;
+
+    JumpList notInt32Op1;
+    JumpList notInt32Op2;
+
+    // Character less.
+    if (isOperandConstantImmediateChar(op1)) {
+        emitLoad(op2, regT1, regT0);
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::CellTag)));
+        JumpList failures;
+        emitLoadCharacterString(regT0, regT0, failures);
+        addSlowCase(failures);
+        addJump(branch32(GreaterThan, regT0, Imm32(asString(getConstantOperand(op1))->tryGetValue()[0])), target);
+        return;
+    }
+    if (isOperandConstantImmediateChar(op2)) {
+        emitLoad(op1, regT1, regT0);
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::CellTag)));
+        JumpList failures;
+        emitLoadCharacterString(regT0, regT0, failures);
+        addSlowCase(failures);
+        addJump(branch32(LessThan, regT0, Imm32(asString(getConstantOperand(op2))->tryGetValue()[0])), target);
+        return;
+    } 
+    if (isOperandConstantImmediateInt(op1)) {
+        emitLoad(op2, regT3, regT2);
+        notInt32Op2.append(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+        addJump(branch32(GreaterThan, regT2, Imm32(getConstantOperand(op1).asInt32())), target);
+    } else if (isOperandConstantImmediateInt(op2)) {
+        emitLoad(op1, regT1, regT0);
+        notInt32Op1.append(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+        addJump(branch32(LessThan, regT0, Imm32(getConstantOperand(op2).asInt32())), target);
+    } else {
+        emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+        notInt32Op1.append(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+        notInt32Op2.append(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+        addJump(branch32(LessThan, regT0, regT2), target);
+    }
+
+    if (!supportsFloatingPoint()) {
+        addSlowCase(notInt32Op1);
+        addSlowCase(notInt32Op2);
+        return;
+    }
+    Jump end = jump();
+
+    // Double less.
+    emitBinaryDoubleOp(op_jless, target, op1, op2, OperandTypes(), notInt32Op1, notInt32Op2, !isOperandConstantImmediateInt(op1), isOperandConstantImmediateInt(op1) || !isOperandConstantImmediateInt(op2));
+    end.link(this);
+}
+
+void JIT::emitSlow_op_jless(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned op1 = currentInstruction[1].u.operand;
+    unsigned op2 = currentInstruction[2].u.operand;
+    unsigned target = currentInstruction[3].u.operand;
+    
+    if (isOperandConstantImmediateChar(op1) || isOperandConstantImmediateChar(op2)) {
+        linkSlowCase(iter);
+        linkSlowCase(iter);
+        linkSlowCase(iter);
+        linkSlowCase(iter);
+    } else {
+        if (!supportsFloatingPoint()) {
+            if (!isOperandConstantImmediateInt(op1) && !isOperandConstantImmediateInt(op2))
+                linkSlowCase(iter); // int32 check
+            linkSlowCase(iter); // int32 check
+        } else {
+            if (!isOperandConstantImmediateInt(op1)) {
+                linkSlowCase(iter); // double check
+                linkSlowCase(iter); // int32 check
+            }
+            if (isOperandConstantImmediateInt(op1) || !isOperandConstantImmediateInt(op2))
+                linkSlowCase(iter); // double check
+        }
+    }
+    JITStubCall stubCall(this, cti_op_jless);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call();
+    emitJumpSlowToHot(branchTest32(NonZero, regT0), target);
+}
+
+void JIT::emit_op_jlesseq(Instruction* currentInstruction, bool invert)
+{
+    unsigned op1 = currentInstruction[1].u.operand;
+    unsigned op2 = currentInstruction[2].u.operand;
+    unsigned target = currentInstruction[3].u.operand;
+
+    JumpList notInt32Op1;
+    JumpList notInt32Op2;
+
+    // Character less.
+    if (isOperandConstantImmediateChar(op1)) {
+        emitLoad(op2, regT1, regT0);
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::CellTag)));
+        JumpList failures;
+        emitLoadCharacterString(regT0, regT0, failures);
+        addSlowCase(failures);
+        addJump(branch32(invert ? LessThan : GreaterThanOrEqual, regT0, Imm32(asString(getConstantOperand(op1))->tryGetValue()[0])), target);
+        return;
+    }
+    if (isOperandConstantImmediateChar(op2)) {
+        emitLoad(op1, regT1, regT0);
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::CellTag)));
+        JumpList failures;
+        emitLoadCharacterString(regT0, regT0, failures);
+        addSlowCase(failures);
+        addJump(branch32(invert ? GreaterThan : LessThanOrEqual, regT0, Imm32(asString(getConstantOperand(op2))->tryGetValue()[0])), target);
+        return;
+    }
+    if (isOperandConstantImmediateInt(op1)) {
+        emitLoad(op2, regT3, regT2);
+        notInt32Op2.append(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+        addJump(branch32(invert ? LessThan : GreaterThanOrEqual, regT2, Imm32(getConstantOperand(op1).asInt32())), target);
+    } else if (isOperandConstantImmediateInt(op2)) {
+        emitLoad(op1, regT1, regT0);
+        notInt32Op1.append(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+        addJump(branch32(invert ? GreaterThan : LessThanOrEqual, regT0, Imm32(getConstantOperand(op2).asInt32())), target);
+    } else {
+        emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+        notInt32Op1.append(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+        notInt32Op2.append(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+        addJump(branch32(invert ? GreaterThan : LessThanOrEqual, regT0, regT2), target);
+    }
+
+    if (!supportsFloatingPoint()) {
+        addSlowCase(notInt32Op1);
+        addSlowCase(notInt32Op2);
+        return;
+    }
+    Jump end = jump();
+
+    // Double less.
+    emitBinaryDoubleOp(invert ? op_jnlesseq : op_jlesseq, target, op1, op2, OperandTypes(), notInt32Op1, notInt32Op2, !isOperandConstantImmediateInt(op1), isOperandConstantImmediateInt(op1) || !isOperandConstantImmediateInt(op2));
+    end.link(this);
+}
+
+void JIT::emitSlow_op_jlesseq(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter, bool invert)
+{
+    unsigned op1 = currentInstruction[1].u.operand;
+    unsigned op2 = currentInstruction[2].u.operand;
+    unsigned target = currentInstruction[3].u.operand;
+
+    if (isOperandConstantImmediateChar(op1) || isOperandConstantImmediateChar(op2)) {
+        linkSlowCase(iter);
+        linkSlowCase(iter);
+        linkSlowCase(iter);
+        linkSlowCase(iter);
+    } else {
+        if (!supportsFloatingPoint()) {
+            if (!isOperandConstantImmediateInt(op1) && !isOperandConstantImmediateInt(op2))
+                linkSlowCase(iter); // int32 check
+            linkSlowCase(iter); // int32 check
+        } else {
+            if (!isOperandConstantImmediateInt(op1)) {
+                linkSlowCase(iter); // double check
+                linkSlowCase(iter); // int32 check
+            }
+            if (isOperandConstantImmediateInt(op1) || !isOperandConstantImmediateInt(op2))
+                linkSlowCase(iter); // double check
+        }
+    }
+
+    JITStubCall stubCall(this, cti_op_jlesseq);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call();
+    emitJumpSlowToHot(branchTest32(invert ? Zero : NonZero, regT0), target);
+}
+
+void JIT::emit_op_jnlesseq(Instruction* currentInstruction)
+{
+    emit_op_jlesseq(currentInstruction, true);
+}
+
+void JIT::emitSlow_op_jnlesseq(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    emitSlow_op_jlesseq(currentInstruction, iter, true);
+}
+
+// LeftShift (<<)
+
+void JIT::emit_op_lshift(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+
+    if (isOperandConstantImmediateInt(op2)) {
+        emitLoad(op1, regT1, regT0);
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+        lshift32(Imm32(getConstantOperand(op2).asInt32()), regT0);
+        emitStoreInt32(dst, regT0, dst == op1);
+        return;
+    }
+
+    emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+    if (!isOperandConstantImmediateInt(op1))
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+    addSlowCase(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+    lshift32(regT2, regT0);
+    emitStoreInt32(dst, regT0, dst == op1 || dst == op2);
+}
+
+void JIT::emitSlow_op_lshift(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+
+    if (!isOperandConstantImmediateInt(op1) && !isOperandConstantImmediateInt(op2))
+        linkSlowCase(iter); // int32 check
+    linkSlowCase(iter); // int32 check
+
+    JITStubCall stubCall(this, cti_op_lshift);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call(dst);
+}
+
+// RightShift (>>) and UnsignedRightShift (>>>) helper
+
+void JIT::emitRightShift(Instruction* currentInstruction, bool isUnsigned)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+
+    // Slow case of rshift makes assumptions about what registers hold the
+    // shift arguments, so any changes must be updated there as well.
+    if (isOperandConstantImmediateInt(op2)) {
+        emitLoad(op1, regT1, regT0);
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+        int shift = getConstantOperand(op2).asInt32();
+        if (isUnsigned) {
+            if (shift)
+                urshift32(Imm32(shift & 0x1f), regT0);
+            // unsigned shift < 0 or shift = k*2^32 may result in (essentially)
+            // a toUint conversion, which can result in a value we can represent
+            // as an immediate int.
+            if (shift < 0 || !(shift & 31))
+                addSlowCase(branch32(LessThan, regT0, Imm32(0)));
+        } else if (shift) { // signed right shift by zero is simply toInt conversion
+            rshift32(Imm32(shift & 0x1f), regT0);
+        }
+        emitStoreInt32(dst, regT0, dst == op1);
+        return;
+    }
+
+    emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+    if (!isOperandConstantImmediateInt(op1))
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+    addSlowCase(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+    if (isUnsigned) {
+        urshift32(regT2, regT0);
+        addSlowCase(branch32(LessThan, regT0, Imm32(0)));
+    } else
+        rshift32(regT2, regT0);
+    emitStoreInt32(dst, regT0, dst == op1 || dst == op2);
+}
+
+void JIT::emitRightShiftSlowCase(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter, bool isUnsigned)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+    if (isOperandConstantImmediateInt(op2)) {
+        int shift = getConstantOperand(op2).asInt32();
+        // op1 = regT1:regT0
+        linkSlowCase(iter); // int32 check
+        if (supportsFloatingPointTruncate()) {
+            JumpList failures;
+            failures.append(branch32(AboveOrEqual, regT1, Imm32(JSValue::LowestTag)));
+            emitLoadDouble(op1, fpRegT0);
+            failures.append(branchTruncateDoubleToInt32(fpRegT0, regT0));
+            if (isUnsigned) {
+                if (shift)
+                    urshift32(Imm32(shift & 0x1f), regT0);
+                if (shift < 0 || !(shift & 31))
+                    failures.append(branch32(LessThan, regT0, Imm32(0)));
+            } else if (shift)
+                rshift32(Imm32(shift & 0x1f), regT0);
+            emitStoreInt32(dst, regT0, false);
+            emitJumpSlowToHot(jump(), OPCODE_LENGTH(op_rshift));
+            failures.link(this);
+        }
+        if (isUnsigned && (shift < 0 || !(shift & 31)))
+            linkSlowCase(iter); // failed to box in hot path
+    } else {
+        // op1 = regT1:regT0
+        // op2 = regT3:regT2
+        if (!isOperandConstantImmediateInt(op1)) {
+            linkSlowCase(iter); // int32 check -- op1 is not an int
+            if (supportsFloatingPointTruncate()) {
+                Jump notDouble = branch32(Above, regT1, Imm32(JSValue::LowestTag)); // op1 is not a double
+                emitLoadDouble(op1, fpRegT0);
+                Jump notInt = branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)); // op2 is not an int
+                Jump cantTruncate = branchTruncateDoubleToInt32(fpRegT0, regT0);
+                if (isUnsigned)
+                    urshift32(regT2, regT0);
+                else
+                    rshift32(regT2, regT0);
+                emitStoreInt32(dst, regT0, false);
+                emitJumpSlowToHot(jump(), OPCODE_LENGTH(op_rshift));
+                notDouble.link(this);
+                notInt.link(this);
+                cantTruncate.link(this);
+            }
+        }
+
+        linkSlowCase(iter); // int32 check - op2 is not an int
+        if (isUnsigned)
+            linkSlowCase(iter); // Can't represent unsigned result as an immediate
+    }
+
+    JITStubCall stubCall(this, isUnsigned ? cti_op_urshift : cti_op_rshift);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call(dst);
+}
+
+// RightShift (>>)
+
+void JIT::emit_op_rshift(Instruction* currentInstruction)
+{
+    emitRightShift(currentInstruction, false);
+}
+
+void JIT::emitSlow_op_rshift(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    emitRightShiftSlowCase(currentInstruction, iter, false);
+}
+
+// UnsignedRightShift (>>>)
+
+void JIT::emit_op_urshift(Instruction* currentInstruction)
+{
+    emitRightShift(currentInstruction, true);
+}
+
+void JIT::emitSlow_op_urshift(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    emitRightShiftSlowCase(currentInstruction, iter, true);
+}
+
+// BitAnd (&)
+
+void JIT::emit_op_bitand(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+
+    unsigned op;
+    int32_t constant;
+    if (getOperandConstantImmediateInt(op1, op2, op, constant)) {
+        emitLoad(op, regT1, regT0);
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+        and32(Imm32(constant), regT0);
+        emitStoreInt32(dst, regT0, (op == dst));
+        return;
+    }
+
+    emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+    addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+    addSlowCase(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+    and32(regT2, regT0);
+    emitStoreInt32(dst, regT0, (op1 == dst || op2 == dst));
+}
+
+void JIT::emitSlow_op_bitand(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+
+    if (!isOperandConstantImmediateInt(op1) && !isOperandConstantImmediateInt(op2))
+        linkSlowCase(iter); // int32 check
+    linkSlowCase(iter); // int32 check
+
+    JITStubCall stubCall(this, cti_op_bitand);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call(dst);
+}
+
+// BitOr (|)
+
+void JIT::emit_op_bitor(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+
+    unsigned op;
+    int32_t constant;
+    if (getOperandConstantImmediateInt(op1, op2, op, constant)) {
+        emitLoad(op, regT1, regT0);
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+        or32(Imm32(constant), regT0);
+        emitStoreInt32(dst, regT0, (op == dst));
+        return;
+    }
+
+    emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+    addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+    addSlowCase(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+    or32(regT2, regT0);
+    emitStoreInt32(dst, regT0, (op1 == dst || op2 == dst));
+}
+
+void JIT::emitSlow_op_bitor(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+
+    if (!isOperandConstantImmediateInt(op1) && !isOperandConstantImmediateInt(op2))
+        linkSlowCase(iter); // int32 check
+    linkSlowCase(iter); // int32 check
+
+    JITStubCall stubCall(this, cti_op_bitor);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call(dst);
+}
+
+// BitXor (^)
+
+void JIT::emit_op_bitxor(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+
+    unsigned op;
+    int32_t constant;
+    if (getOperandConstantImmediateInt(op1, op2, op, constant)) {
+        emitLoad(op, regT1, regT0);
+        addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+        xor32(Imm32(constant), regT0);
+        emitStoreInt32(dst, regT0, (op == dst));
+        return;
+    }
+
+    emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+    addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+    addSlowCase(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+    xor32(regT2, regT0);
+    emitStoreInt32(dst, regT0, (op1 == dst || op2 == dst));
+}
+
+void JIT::emitSlow_op_bitxor(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+
+    if (!isOperandConstantImmediateInt(op1) && !isOperandConstantImmediateInt(op2))
+        linkSlowCase(iter); // int32 check
+    linkSlowCase(iter); // int32 check
+
+    JITStubCall stubCall(this, cti_op_bitxor);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call(dst);
+}
+
+// BitNot (~)
+
+void JIT::emit_op_bitnot(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned src = currentInstruction[2].u.operand;
+
+    emitLoad(src, regT1, regT0);
+    addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+
+    not32(regT0);
+    emitStoreInt32(dst, regT0, (dst == src));
+}
+
+void JIT::emitSlow_op_bitnot(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+
+    linkSlowCase(iter); // int32 check
+
+    JITStubCall stubCall(this, cti_op_bitnot);
+    stubCall.addArgument(regT1, regT0);
+    stubCall.call(dst);
+}
+
+// PostInc (i++)
+
+void JIT::emit_op_post_inc(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned srcDst = currentInstruction[2].u.operand;
+
+    emitLoad(srcDst, regT1, regT0);
+    addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+
+    if (dst == srcDst) // x = x++ is a noop for ints.
+        return;
+
+    emitStoreInt32(dst, regT0);
+
+    addSlowCase(branchAdd32(Overflow, Imm32(1), regT0));
+    emitStoreInt32(srcDst, regT0, true);
+}
+
+void JIT::emitSlow_op_post_inc(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned srcDst = currentInstruction[2].u.operand;
+
+    linkSlowCase(iter); // int32 check
+    if (dst != srcDst)
+        linkSlowCase(iter); // overflow check
+
+    JITStubCall stubCall(this, cti_op_post_inc);
+    stubCall.addArgument(srcDst);
+    stubCall.addArgument(Imm32(srcDst));
+    stubCall.call(dst);
+}
+
+// PostDec (i--)
+
+void JIT::emit_op_post_dec(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned srcDst = currentInstruction[2].u.operand;
+
+    emitLoad(srcDst, regT1, regT0);
+    addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+
+    if (dst == srcDst) // x = x-- is a noop for ints.
+        return;
+
+    emitStoreInt32(dst, regT0);
+
+    addSlowCase(branchSub32(Overflow, Imm32(1), regT0));
+    emitStoreInt32(srcDst, regT0, true);
+}
+
+void JIT::emitSlow_op_post_dec(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned srcDst = currentInstruction[2].u.operand;
+
+    linkSlowCase(iter); // int32 check
+    if (dst != srcDst)
+        linkSlowCase(iter); // overflow check
+
+    JITStubCall stubCall(this, cti_op_post_dec);
+    stubCall.addArgument(srcDst);
+    stubCall.addArgument(Imm32(srcDst));
+    stubCall.call(dst);
+}
+
+// PreInc (++i)
+
+void JIT::emit_op_pre_inc(Instruction* currentInstruction)
+{
+    unsigned srcDst = currentInstruction[1].u.operand;
+
+    emitLoad(srcDst, regT1, regT0);
+
+    addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+    addSlowCase(branchAdd32(Overflow, Imm32(1), regT0));
+    emitStoreInt32(srcDst, regT0, true);
+}
+
+void JIT::emitSlow_op_pre_inc(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned srcDst = currentInstruction[1].u.operand;
+
+    linkSlowCase(iter); // int32 check
+    linkSlowCase(iter); // overflow check
+
+    JITStubCall stubCall(this, cti_op_pre_inc);
+    stubCall.addArgument(srcDst);
+    stubCall.call(srcDst);
+}
+
+// PreDec (--i)
+
+void JIT::emit_op_pre_dec(Instruction* currentInstruction)
+{
+    unsigned srcDst = currentInstruction[1].u.operand;
+
+    emitLoad(srcDst, regT1, regT0);
+
+    addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+    addSlowCase(branchSub32(Overflow, Imm32(1), regT0));
+    emitStoreInt32(srcDst, regT0, true);
+}
+
+void JIT::emitSlow_op_pre_dec(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned srcDst = currentInstruction[1].u.operand;
+
+    linkSlowCase(iter); // int32 check
+    linkSlowCase(iter); // overflow check
+
+    JITStubCall stubCall(this, cti_op_pre_dec);
+    stubCall.addArgument(srcDst);
+    stubCall.call(srcDst);
+}
+
+// Addition (+)
+
+void JIT::emit_op_add(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+    OperandTypes types = OperandTypes::fromInt(currentInstruction[4].u.operand);
+
+    if (!types.first().mightBeNumber() || !types.second().mightBeNumber()) {
+        JITStubCall stubCall(this, cti_op_add);
+        stubCall.addArgument(op1);
+        stubCall.addArgument(op2);
+        stubCall.call(dst);
+        return;
+    }
+
+    JumpList notInt32Op1;
+    JumpList notInt32Op2;
+
+    unsigned op;
+    int32_t constant;
+    if (getOperandConstantImmediateInt(op1, op2, op, constant)) {
+        emitAdd32Constant(dst, op, constant, op == op1 ? types.first() : types.second());
+        return;
+    }
+
+    emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+    notInt32Op1.append(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+    notInt32Op2.append(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+
+    // Int32 case.
+    addSlowCase(branchAdd32(Overflow, regT2, regT0));
+    emitStoreInt32(dst, regT0, (op1 == dst || op2 == dst));
+
+    if (!supportsFloatingPoint()) {
+        addSlowCase(notInt32Op1);
+        addSlowCase(notInt32Op2);
+        return;
+    }
+    Jump end = jump();
+
+    // Double case.
+    emitBinaryDoubleOp(op_add, dst, op1, op2, types, notInt32Op1, notInt32Op2);
+    end.link(this);
+}
+
+void JIT::emitAdd32Constant(unsigned dst, unsigned op, int32_t constant, ResultType opType)
+{
+    // Int32 case.
+    emitLoad(op, regT1, regT0);
+    Jump notInt32 = branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag));
+    addSlowCase(branchAdd32(Overflow, Imm32(constant), regT0));
+    emitStoreInt32(dst, regT0, (op == dst));
+
+    // Double case.
+    if (!supportsFloatingPoint()) {
+        addSlowCase(notInt32);
+        return;
+    }
+    Jump end = jump();
+
+    notInt32.link(this);
+    if (!opType.definitelyIsNumber())
+        addSlowCase(branch32(Above, regT1, Imm32(JSValue::LowestTag)));
+    move(Imm32(constant), regT2);
+    convertInt32ToDouble(regT2, fpRegT0);
+    emitLoadDouble(op, fpRegT1);
+    addDouble(fpRegT1, fpRegT0);
+    emitStoreDouble(dst, fpRegT0);
+
+    end.link(this);
+}
+
+void JIT::emitSlow_op_add(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+    OperandTypes types = OperandTypes::fromInt(currentInstruction[4].u.operand);
+
+    if (!types.first().mightBeNumber() || !types.second().mightBeNumber())
+        return;
+
+    unsigned op;
+    int32_t constant;
+    if (getOperandConstantImmediateInt(op1, op2, op, constant)) {
+        linkSlowCase(iter); // overflow check
+
+        if (!supportsFloatingPoint())
+            linkSlowCase(iter); // non-sse case
+        else {
+            ResultType opType = op == op1 ? types.first() : types.second();
+            if (!opType.definitelyIsNumber())
+                linkSlowCase(iter); // double check
+        }
+    } else {
+        linkSlowCase(iter); // overflow check
+
+        if (!supportsFloatingPoint()) {
+            linkSlowCase(iter); // int32 check
+            linkSlowCase(iter); // int32 check
+        } else {
+            if (!types.first().definitelyIsNumber())
+                linkSlowCase(iter); // double check
+
+            if (!types.second().definitelyIsNumber()) {
+                linkSlowCase(iter); // int32 check
+                linkSlowCase(iter); // double check
+            }
+        }
+    }
+
+    JITStubCall stubCall(this, cti_op_add);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call(dst);
+}
+
+// Subtraction (-)
+
+void JIT::emit_op_sub(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+    OperandTypes types = OperandTypes::fromInt(currentInstruction[4].u.operand);
+
+    JumpList notInt32Op1;
+    JumpList notInt32Op2;
+
+    if (isOperandConstantImmediateInt(op2)) {
+        emitSub32Constant(dst, op1, getConstantOperand(op2).asInt32(), types.first());
+        return;
+    }
+
+    emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+    notInt32Op1.append(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+    notInt32Op2.append(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+
+    // Int32 case.
+    addSlowCase(branchSub32(Overflow, regT2, regT0));
+    emitStoreInt32(dst, regT0, (op1 == dst || op2 == dst));
+
+    if (!supportsFloatingPoint()) {
+        addSlowCase(notInt32Op1);
+        addSlowCase(notInt32Op2);
+        return;
+    }
+    Jump end = jump();
+
+    // Double case.
+    emitBinaryDoubleOp(op_sub, dst, op1, op2, types, notInt32Op1, notInt32Op2);
+    end.link(this);
+}
+
+void JIT::emitSub32Constant(unsigned dst, unsigned op, int32_t constant, ResultType opType)
+{
+    // Int32 case.
+    emitLoad(op, regT1, regT0);
+    Jump notInt32 = branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag));
+    addSlowCase(branchSub32(Overflow, Imm32(constant), regT0));
+    emitStoreInt32(dst, regT0, (op == dst));
+
+    // Double case.
+    if (!supportsFloatingPoint()) {
+        addSlowCase(notInt32);
+        return;
+    }
+    Jump end = jump();
+
+    notInt32.link(this);
+    if (!opType.definitelyIsNumber())
+        addSlowCase(branch32(Above, regT1, Imm32(JSValue::LowestTag)));
+    move(Imm32(constant), regT2);
+    convertInt32ToDouble(regT2, fpRegT0);
+    emitLoadDouble(op, fpRegT1);
+    subDouble(fpRegT0, fpRegT1);
+    emitStoreDouble(dst, fpRegT1);
+
+    end.link(this);
+}
+
+void JIT::emitSlow_op_sub(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+    OperandTypes types = OperandTypes::fromInt(currentInstruction[4].u.operand);
+
+    if (isOperandConstantImmediateInt(op2)) {
+        linkSlowCase(iter); // overflow check
+
+        if (!supportsFloatingPoint() || !types.first().definitelyIsNumber())
+            linkSlowCase(iter); // int32 or double check
+    } else {
+        linkSlowCase(iter); // overflow check
+
+        if (!supportsFloatingPoint()) {
+            linkSlowCase(iter); // int32 check
+            linkSlowCase(iter); // int32 check
+        } else {
+            if (!types.first().definitelyIsNumber())
+                linkSlowCase(iter); // double check
+
+            if (!types.second().definitelyIsNumber()) {
+                linkSlowCase(iter); // int32 check
+                linkSlowCase(iter); // double check
+            }
+        }
+    }
+
+    JITStubCall stubCall(this, cti_op_sub);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call(dst);
+}
+
+void JIT::emitBinaryDoubleOp(OpcodeID opcodeID, unsigned dst, unsigned op1, unsigned op2, OperandTypes types, JumpList& notInt32Op1, JumpList& notInt32Op2, bool op1IsInRegisters, bool op2IsInRegisters)
+{
+    JumpList end;
+
+    if (!notInt32Op1.empty()) {
+        // Double case 1: Op1 is not int32; Op2 is unknown.
+        notInt32Op1.link(this);
+
+        ASSERT(op1IsInRegisters);
+
+        // Verify Op1 is double.
+        if (!types.first().definitelyIsNumber())
+            addSlowCase(branch32(Above, regT1, Imm32(JSValue::LowestTag)));
+
+        if (!op2IsInRegisters)
+            emitLoad(op2, regT3, regT2);
+
+        Jump doubleOp2 = branch32(Below, regT3, Imm32(JSValue::LowestTag));
+
+        if (!types.second().definitelyIsNumber())
+            addSlowCase(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+
+        convertInt32ToDouble(regT2, fpRegT0);
+        Jump doTheMath = jump();
+
+        // Load Op2 as double into double register.
+        doubleOp2.link(this);
+        emitLoadDouble(op2, fpRegT0);
+
+        // Do the math.
+        doTheMath.link(this);
+        switch (opcodeID) {
+            case op_mul:
+                emitLoadDouble(op1, fpRegT2);
+                mulDouble(fpRegT2, fpRegT0);
+                emitStoreDouble(dst, fpRegT0);
+                break;
+            case op_add:
+                emitLoadDouble(op1, fpRegT2);
+                addDouble(fpRegT2, fpRegT0);
+                emitStoreDouble(dst, fpRegT0);
+                break;
+            case op_sub:
+                emitLoadDouble(op1, fpRegT1);
+                subDouble(fpRegT0, fpRegT1);
+                emitStoreDouble(dst, fpRegT1);
+                break;
+            case op_div:
+                emitLoadDouble(op1, fpRegT1);
+                divDouble(fpRegT0, fpRegT1);
+                emitStoreDouble(dst, fpRegT1);
+                break;
+            case op_jnless:
+                emitLoadDouble(op1, fpRegT2);
+                addJump(branchDouble(DoubleLessThanOrEqualOrUnordered, fpRegT0, fpRegT2), dst);
+                break;
+            case op_jless:
+                emitLoadDouble(op1, fpRegT2);
+                addJump(branchDouble(DoubleLessThan, fpRegT2, fpRegT0), dst);
+                break;
+            case op_jlesseq:
+                emitLoadDouble(op1, fpRegT2);
+                addJump(branchDouble(DoubleLessThanOrEqual, fpRegT2, fpRegT0), dst);
+                break;
+            case op_jnlesseq:
+                emitLoadDouble(op1, fpRegT2);
+                addJump(branchDouble(DoubleLessThanOrUnordered, fpRegT0, fpRegT2), dst);
+                break;
+            default:
+                ASSERT_NOT_REACHED();
+        }
+
+        if (!notInt32Op2.empty())
+            end.append(jump());
+    }
+
+    if (!notInt32Op2.empty()) {
+        // Double case 2: Op1 is int32; Op2 is not int32.
+        notInt32Op2.link(this);
+
+        ASSERT(op2IsInRegisters);
+
+        if (!op1IsInRegisters)
+            emitLoadPayload(op1, regT0);
+
+        convertInt32ToDouble(regT0, fpRegT0);
+
+        // Verify op2 is double.
+        if (!types.second().definitelyIsNumber())
+            addSlowCase(branch32(Above, regT3, Imm32(JSValue::LowestTag)));
+
+        // Do the math.
+        switch (opcodeID) {
+            case op_mul:
+                emitLoadDouble(op2, fpRegT2);
+                mulDouble(fpRegT2, fpRegT0);
+                emitStoreDouble(dst, fpRegT0);
+                break;
+            case op_add:
+                emitLoadDouble(op2, fpRegT2);
+                addDouble(fpRegT2, fpRegT0);
+                emitStoreDouble(dst, fpRegT0);
+                break;
+            case op_sub:
+                emitLoadDouble(op2, fpRegT2);
+                subDouble(fpRegT2, fpRegT0);
+                emitStoreDouble(dst, fpRegT0);
+                break;
+            case op_div:
+                emitLoadDouble(op2, fpRegT2);
+                divDouble(fpRegT2, fpRegT0);
+                emitStoreDouble(dst, fpRegT0);
+                break;
+            case op_jnless:
+                emitLoadDouble(op2, fpRegT1);
+                addJump(branchDouble(DoubleLessThanOrEqualOrUnordered, fpRegT1, fpRegT0), dst);
+                break;
+            case op_jless:
+                emitLoadDouble(op2, fpRegT1);
+                addJump(branchDouble(DoubleLessThan, fpRegT0, fpRegT1), dst);
+                break;
+            case op_jnlesseq:
+                emitLoadDouble(op2, fpRegT1);
+                addJump(branchDouble(DoubleLessThanOrUnordered, fpRegT1, fpRegT0), dst);
+                break;
+            case op_jlesseq:
+                emitLoadDouble(op2, fpRegT1);
+                addJump(branchDouble(DoubleLessThanOrEqual, fpRegT0, fpRegT1), dst);
+                break;
+            default:
+                ASSERT_NOT_REACHED();
+        }
+    }
+
+    end.link(this);
+}
+
+// Multiplication (*)
+
+void JIT::emit_op_mul(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+    OperandTypes types = OperandTypes::fromInt(currentInstruction[4].u.operand);
+
+    JumpList notInt32Op1;
+    JumpList notInt32Op2;
+
+    emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+    notInt32Op1.append(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+    notInt32Op2.append(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+
+    // Int32 case.
+    move(regT0, regT3);
+    addSlowCase(branchMul32(Overflow, regT2, regT0));
+    addSlowCase(branchTest32(Zero, regT0));
+    emitStoreInt32(dst, regT0, (op1 == dst || op2 == dst));
+
+    if (!supportsFloatingPoint()) {
+        addSlowCase(notInt32Op1);
+        addSlowCase(notInt32Op2);
+        return;
+    }
+    Jump end = jump();
+
+    // Double case.
+    emitBinaryDoubleOp(op_mul, dst, op1, op2, types, notInt32Op1, notInt32Op2);
+    end.link(this);
+}
+
+void JIT::emitSlow_op_mul(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+    OperandTypes types = OperandTypes::fromInt(currentInstruction[4].u.operand);
+
+    Jump overflow = getSlowCase(iter); // overflow check
+    linkSlowCase(iter); // zero result check
+
+    Jump negZero = branchOr32(Signed, regT2, regT3);
+    emitStoreInt32(dst, Imm32(0), (op1 == dst || op2 == dst));
+
+    emitJumpSlowToHot(jump(), OPCODE_LENGTH(op_mul));
+
+    negZero.link(this);
+    overflow.link(this);
+
+    if (!supportsFloatingPoint()) {
+        linkSlowCase(iter); // int32 check
+        linkSlowCase(iter); // int32 check
+    }
+
+    if (supportsFloatingPoint()) {
+        if (!types.first().definitelyIsNumber())
+            linkSlowCase(iter); // double check
+
+        if (!types.second().definitelyIsNumber()) {
+            linkSlowCase(iter); // int32 check
+            linkSlowCase(iter); // double check
+        }
+    }
+
+    Label jitStubCall(this);
+    JITStubCall stubCall(this, cti_op_mul);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call(dst);
+}
+
+// Division (/)
+
+void JIT::emit_op_div(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+    OperandTypes types = OperandTypes::fromInt(currentInstruction[4].u.operand);
+
+    if (!supportsFloatingPoint()) {
+        addSlowCase(jump());
+        return;
+    }
+
+    // Int32 divide.
+    JumpList notInt32Op1;
+    JumpList notInt32Op2;
+
+    JumpList end;
+
+    emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+
+    notInt32Op1.append(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+    notInt32Op2.append(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+
+    convertInt32ToDouble(regT0, fpRegT0);
+    convertInt32ToDouble(regT2, fpRegT1);
+    divDouble(fpRegT1, fpRegT0);
+
+    JumpList doubleResult;
+    branchConvertDoubleToInt32(fpRegT0, regT0, doubleResult, fpRegT1);
+
+    // Int32 result.
+    emitStoreInt32(dst, regT0, (op1 == dst || op2 == dst));
+    end.append(jump());
+
+    // Double result.
+    doubleResult.link(this);
+    emitStoreDouble(dst, fpRegT0);
+    end.append(jump());
+
+    // Double divide.
+    emitBinaryDoubleOp(op_div, dst, op1, op2, types, notInt32Op1, notInt32Op2);
+    end.link(this);
+}
+
+void JIT::emitSlow_op_div(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+    OperandTypes types = OperandTypes::fromInt(currentInstruction[4].u.operand);
+
+    if (!supportsFloatingPoint())
+        linkSlowCase(iter);
+    else {
+        if (!types.first().definitelyIsNumber())
+            linkSlowCase(iter); // double check
+
+        if (!types.second().definitelyIsNumber()) {
+            linkSlowCase(iter); // int32 check
+            linkSlowCase(iter); // double check
+        }
+    }
+
+    JITStubCall stubCall(this, cti_op_div);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call(dst);
+}
+
+// Mod (%)
+
+/* ------------------------------ BEGIN: OP_MOD ------------------------------ */
+
+#if CPU(X86) || CPU(X86_64)
+
+void JIT::emit_op_mod(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+
+    if (isOperandConstantImmediateInt(op2) && getConstantOperand(op2).asInt32() != 0) {
+        emitLoad(op1, X86Registers::edx, X86Registers::eax);
+        move(Imm32(getConstantOperand(op2).asInt32()), X86Registers::ecx);
+        addSlowCase(branch32(NotEqual, X86Registers::edx, Imm32(JSValue::Int32Tag)));
+        if (getConstantOperand(op2).asInt32() == -1)
+            addSlowCase(branch32(Equal, X86Registers::eax, Imm32(0x80000000))); // -2147483648 / -1 => EXC_ARITHMETIC
+    } else {
+        emitLoad2(op1, X86Registers::edx, X86Registers::eax, op2, X86Registers::ebx, X86Registers::ecx);
+        addSlowCase(branch32(NotEqual, X86Registers::edx, Imm32(JSValue::Int32Tag)));
+        addSlowCase(branch32(NotEqual, X86Registers::ebx, Imm32(JSValue::Int32Tag)));
+
+        addSlowCase(branch32(Equal, X86Registers::eax, Imm32(0x80000000))); // -2147483648 / -1 => EXC_ARITHMETIC
+        addSlowCase(branch32(Equal, X86Registers::ecx, Imm32(0))); // divide by 0
+    }
+
+    move(X86Registers::eax, X86Registers::ebx); // Save dividend payload, in case of 0.
+    m_assembler.cdq();
+    m_assembler.idivl_r(X86Registers::ecx);
+
+    // If the remainder is zero and the dividend is negative, the result is -0.
+    Jump storeResult1 = branchTest32(NonZero, X86Registers::edx);
+    Jump storeResult2 = branchTest32(Zero, X86Registers::ebx, Imm32(0x80000000)); // not negative
+    emitStore(dst, jsNumber(m_globalData, -0.0));
+    Jump end = jump();
+
+    storeResult1.link(this);
+    storeResult2.link(this);
+    emitStoreInt32(dst, X86Registers::edx, (op1 == dst || op2 == dst));
+    end.link(this);
+}
+
+void JIT::emitSlow_op_mod(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+
+    if (isOperandConstantImmediateInt(op2) && getConstantOperand(op2).asInt32() != 0) {
+        linkSlowCase(iter); // int32 check
+        if (getConstantOperand(op2).asInt32() == -1)
+            linkSlowCase(iter); // 0x80000000 check
+    } else {
+        linkSlowCase(iter); // int32 check
+        linkSlowCase(iter); // int32 check
+        linkSlowCase(iter); // 0 check
+        linkSlowCase(iter); // 0x80000000 check
+    }
+
+    JITStubCall stubCall(this, cti_op_mod);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call(dst);
+}
+
+#else // CPU(X86) || CPU(X86_64)
+
+void JIT::emit_op_mod(Instruction* currentInstruction)
+{
+    unsigned dst = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+
+#if ENABLE(JIT_OPTIMIZE_MOD)
+    emitLoad2(op1, regT1, regT0, op2, regT3, regT2);
+    addSlowCase(branch32(NotEqual, regT1, Imm32(JSValue::Int32Tag)));
+    addSlowCase(branch32(NotEqual, regT3, Imm32(JSValue::Int32Tag)));
+
+    addSlowCase(branch32(Equal, regT2, Imm32(0)));
+
+    emitNakedCall(m_globalData->jitStubs->ctiSoftModulo());
+
+    emitStoreInt32(dst, regT0, (op1 == dst || op2 == dst));
+#else
+    JITStubCall stubCall(this, cti_op_mod);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call(dst);
+#endif
+}
+
+void JIT::emitSlow_op_mod(Instruction* currentInstruction, Vector<SlowCaseEntry>::iterator& iter)
+{
+#if ENABLE(JIT_OPTIMIZE_MOD)
+    unsigned result = currentInstruction[1].u.operand;
+    unsigned op1 = currentInstruction[2].u.operand;
+    unsigned op2 = currentInstruction[3].u.operand;
+    linkSlowCase(iter);
+    linkSlowCase(iter);
+    linkSlowCase(iter);
+    JITStubCall stubCall(this, cti_op_mod);
+    stubCall.addArgument(op1);
+    stubCall.addArgument(op2);
+    stubCall.call(result);
+#else
+    ASSERT_NOT_REACHED();
+#endif
+}
+
+#endif // CPU(X86) || CPU(X86_64)
+
+/* ------------------------------ END: OP_MOD ------------------------------ */
+
+} // namespace JSC
+
+#endif // USE(JSVALUE32_64)
+#endif // ENABLE(JIT)