/* -*- Mode: C++; tab-width: 8; indent-tabs-mode: nil; c-basic-offset: 4 -*- * vim: set ts=8 sts=4 et sw=4 tw=99: * This Source Code Form is subject to the terms of the Mozilla Public * License, v. 2.0. If a copy of the MPL was not distributed with this * file, You can obtain one at http://mozilla.org/MPL/2.0/. */ #ifndef jit_MacroAssembler_inl_h #define jit_MacroAssembler_inl_h #include "jit/MacroAssembler.h" #include "mozilla/MathAlgorithms.h" #if defined(JS_CODEGEN_X86) # include "jit/x86/MacroAssembler-x86-inl.h" #elif defined(JS_CODEGEN_X64) # include "jit/x64/MacroAssembler-x64-inl.h" #elif defined(JS_CODEGEN_ARM) # include "jit/arm/MacroAssembler-arm-inl.h" #elif defined(JS_CODEGEN_ARM64) # include "jit/arm64/MacroAssembler-arm64-inl.h" #elif defined(JS_CODEGEN_MIPS32) # include "jit/mips32/MacroAssembler-mips32-inl.h" #elif defined(JS_CODEGEN_MIPS64) # include "jit/mips64/MacroAssembler-mips64-inl.h" #elif !defined(JS_CODEGEN_NONE) # error "Unknown architecture!" #endif namespace js { namespace jit { //{{{ check_macroassembler_style // =============================================================== // Frame manipulation functions. uint32_t MacroAssembler::framePushed() const { return framePushed_; } void MacroAssembler::setFramePushed(uint32_t framePushed) { framePushed_ = framePushed; } void MacroAssembler::adjustFrame(int32_t value) { MOZ_ASSERT_IF(value < 0, framePushed_ >= uint32_t(-value)); setFramePushed(framePushed_ + value); } void MacroAssembler::implicitPop(uint32_t bytes) { MOZ_ASSERT(bytes % sizeof(intptr_t) == 0); MOZ_ASSERT(bytes <= INT32_MAX); adjustFrame(-int32_t(bytes)); } // =============================================================== // Stack manipulation functions. CodeOffset MacroAssembler::PushWithPatch(ImmWord word) { framePushed_ += sizeof(word.value); return pushWithPatch(word); } CodeOffset MacroAssembler::PushWithPatch(ImmPtr imm) { return PushWithPatch(ImmWord(uintptr_t(imm.value))); } // =============================================================== // Simple call functions. void MacroAssembler::call(const wasm::CallSiteDesc& desc, const Register reg) { CodeOffset l = call(reg); append(desc, l, framePushed()); } void MacroAssembler::call(const wasm::CallSiteDesc& desc, uint32_t funcDefIndex) { CodeOffset l = callWithPatch(); append(desc, l, framePushed(), funcDefIndex); } void MacroAssembler::call(const wasm::CallSiteDesc& desc, wasm::Trap trap) { CodeOffset l = callWithPatch(); append(desc, l, framePushed(), trap); } // =============================================================== // ABI function calls. void MacroAssembler::passABIArg(Register reg) { passABIArg(MoveOperand(reg), MoveOp::GENERAL); } void MacroAssembler::passABIArg(FloatRegister reg, MoveOp::Type type) { passABIArg(MoveOperand(reg), type); } template void MacroAssembler::callWithABI(const T& fun, MoveOp::Type result) { AutoProfilerCallInstrumentation profiler(*this); callWithABINoProfiler(fun, result); } void MacroAssembler::appendSignatureType(MoveOp::Type type) { #ifdef JS_SIMULATOR signature_ <<= ArgType_Shift; switch (type) { case MoveOp::GENERAL: signature_ |= ArgType_General; break; case MoveOp::DOUBLE: signature_ |= ArgType_Double; break; case MoveOp::FLOAT32: signature_ |= ArgType_Float32; break; default: MOZ_CRASH("Invalid argument type"); } #endif } ABIFunctionType MacroAssembler::signature() const { #ifdef JS_SIMULATOR #ifdef DEBUG switch (signature_) { case Args_General0: case Args_General1: case Args_General2: case Args_General3: case Args_General4: case Args_General5: case Args_General6: case Args_General7: case Args_General8: case Args_Double_None: case Args_Int_Double: case Args_Float32_Float32: case Args_Double_Double: case Args_Double_Int: case Args_Double_DoubleInt: case Args_Double_DoubleDouble: case Args_Double_IntDouble: case Args_Int_IntDouble: case Args_Int_DoubleIntInt: case Args_Int_IntDoubleIntInt: case Args_Double_DoubleDoubleDouble: case Args_Double_DoubleDoubleDoubleDouble: break; default: MOZ_CRASH("Unexpected type"); } #endif // DEBUG return ABIFunctionType(signature_); #else // No simulator enabled. MOZ_CRASH("Only available for making calls within a simulator."); #endif } // =============================================================== // Jit Frames. uint32_t MacroAssembler::callJitNoProfiler(Register callee) { #ifdef JS_USE_LINK_REGISTER // The return address is pushed by the callee. call(callee); #else callAndPushReturnAddress(callee); #endif return currentOffset(); } uint32_t MacroAssembler::callJit(Register callee) { AutoProfilerCallInstrumentation profiler(*this); uint32_t ret = callJitNoProfiler(callee); return ret; } uint32_t MacroAssembler::callJit(JitCode* callee) { AutoProfilerCallInstrumentation profiler(*this); call(callee); return currentOffset(); } void MacroAssembler::makeFrameDescriptor(Register frameSizeReg, FrameType type, uint32_t headerSize) { // See JitFrames.h for a description of the frame descriptor format. // The saved-frame bit is zero for new frames. See js::SavedStacks. lshiftPtr(Imm32(FRAMESIZE_SHIFT), frameSizeReg); headerSize = EncodeFrameHeaderSize(headerSize); orPtr(Imm32((headerSize << FRAME_HEADER_SIZE_SHIFT) | type), frameSizeReg); } void MacroAssembler::pushStaticFrameDescriptor(FrameType type, uint32_t headerSize) { uint32_t descriptor = MakeFrameDescriptor(framePushed(), type, headerSize); Push(Imm32(descriptor)); } void MacroAssembler::PushCalleeToken(Register callee, bool constructing) { if (constructing) { orPtr(Imm32(CalleeToken_FunctionConstructing), callee); Push(callee); andPtr(Imm32(uint32_t(CalleeTokenMask)), callee); } else { static_assert(CalleeToken_Function == 0, "Non-constructing call requires no tagging"); Push(callee); } } void MacroAssembler::loadFunctionFromCalleeToken(Address token, Register dest) { #ifdef DEBUG Label ok; loadPtr(token, dest); andPtr(Imm32(uint32_t(~CalleeTokenMask)), dest); branchPtr(Assembler::Equal, dest, Imm32(CalleeToken_Function), &ok); branchPtr(Assembler::Equal, dest, Imm32(CalleeToken_FunctionConstructing), &ok); assumeUnreachable("Unexpected CalleeToken tag"); bind(&ok); #endif loadPtr(token, dest); andPtr(Imm32(uint32_t(CalleeTokenMask)), dest); } uint32_t MacroAssembler::buildFakeExitFrame(Register scratch) { mozilla::DebugOnly initialDepth = framePushed(); pushStaticFrameDescriptor(JitFrame_IonJS, ExitFrameLayout::Size()); uint32_t retAddr = pushFakeReturnAddress(scratch); MOZ_ASSERT(framePushed() == initialDepth + ExitFrameLayout::Size()); return retAddr; } // =============================================================== // Exit frame footer. void MacroAssembler::PushStubCode() { // Make sure that we do not erase an existing self-reference. MOZ_ASSERT(!hasSelfReference()); selfReferencePatch_ = PushWithPatch(ImmWord(-1)); } void MacroAssembler::enterExitFrame(const VMFunction* f) { linkExitFrame(); // Push the JitCode pointer. (Keep the code alive, when on the stack) PushStubCode(); // Push VMFunction pointer, to mark arguments. Push(ImmPtr(f)); } void MacroAssembler::enterFakeExitFrame(enum ExitFrameTokenValues token) { linkExitFrame(); Push(Imm32(token)); Push(ImmPtr(nullptr)); } void MacroAssembler::enterFakeExitFrameForNative(bool isConstructing) { enterFakeExitFrame(isConstructing ? ConstructNativeExitFrameLayoutToken : CallNativeExitFrameLayoutToken); } void MacroAssembler::leaveExitFrame(size_t extraFrame) { freeStack(ExitFooterFrame::Size() + extraFrame); } bool MacroAssembler::hasSelfReference() const { return selfReferencePatch_.bound(); } // =============================================================== // Arithmetic functions void MacroAssembler::addPtr(ImmPtr imm, Register dest) { addPtr(ImmWord(uintptr_t(imm.value)), dest); } void MacroAssembler::inc32(RegisterOrInt32Constant* key) { if (key->isRegister()) add32(Imm32(1), key->reg()); else key->bumpConstant(1); } void MacroAssembler::dec32(RegisterOrInt32Constant* key) { if (key->isRegister()) add32(Imm32(-1), key->reg()); else key->bumpConstant(-1); } // =============================================================== // Branch functions void MacroAssembler::branch32(Condition cond, Register length, const RegisterOrInt32Constant& key, Label* label) { branch32Impl(cond, length, key, label); } void MacroAssembler::branch32(Condition cond, const Address& length, const RegisterOrInt32Constant& key, Label* label) { branch32Impl(cond, length, key, label); } template void MacroAssembler::branch32Impl(Condition cond, const T& length, const RegisterOrInt32Constant& key, Label* label) { if (key.isRegister()) branch32(cond, length, key.reg(), label); else branch32(cond, length, Imm32(key.constant()), label); } template void MacroAssembler::branchIfFalseBool(Register reg, L label) { // Note that C++ bool is only 1 byte, so ignore the higher-order bits. branchTest32(Assembler::Zero, reg, Imm32(0xFF), label); } void MacroAssembler::branchIfTrueBool(Register reg, Label* label) { // Note that C++ bool is only 1 byte, so ignore the higher-order bits. branchTest32(Assembler::NonZero, reg, Imm32(0xFF), label); } void MacroAssembler::branchIfRope(Register str, Label* label) { Address flags(str, JSString::offsetOfFlags()); static_assert(JSString::ROPE_FLAGS == 0, "Rope type flags must be 0"); branchTest32(Assembler::Zero, flags, Imm32(JSString::TYPE_FLAGS_MASK), label); } void MacroAssembler::branchIfRopeOrExternal(Register str, Register temp, Label* label) { Address flags(str, JSString::offsetOfFlags()); move32(Imm32(JSString::TYPE_FLAGS_MASK), temp); and32(flags, temp); static_assert(JSString::ROPE_FLAGS == 0, "Rope type flags must be 0"); branchTest32(Assembler::Zero, temp, temp, label); branch32(Assembler::Equal, temp, Imm32(JSString::EXTERNAL_FLAGS), label); } void MacroAssembler::branchLatin1String(Register string, Label* label) { branchTest32(Assembler::NonZero, Address(string, JSString::offsetOfFlags()), Imm32(JSString::LATIN1_CHARS_BIT), label); } void MacroAssembler::branchTwoByteString(Register string, Label* label) { branchTest32(Assembler::Zero, Address(string, JSString::offsetOfFlags()), Imm32(JSString::LATIN1_CHARS_BIT), label); } void MacroAssembler::branchIfFunctionHasNoScript(Register fun, Label* label) { // 16-bit loads are slow and unaligned 32-bit loads may be too so // perform an aligned 32-bit load and adjust the bitmask accordingly. MOZ_ASSERT(JSFunction::offsetOfNargs() % sizeof(uint32_t) == 0); MOZ_ASSERT(JSFunction::offsetOfFlags() == JSFunction::offsetOfNargs() + 2); Address address(fun, JSFunction::offsetOfNargs()); int32_t bit = IMM32_16ADJ(JSFunction::INTERPRETED); branchTest32(Assembler::Zero, address, Imm32(bit), label); } void MacroAssembler::branchIfInterpreted(Register fun, Label* label) { // 16-bit loads are slow and unaligned 32-bit loads may be too so // perform an aligned 32-bit load and adjust the bitmask accordingly. MOZ_ASSERT(JSFunction::offsetOfNargs() % sizeof(uint32_t) == 0); MOZ_ASSERT(JSFunction::offsetOfFlags() == JSFunction::offsetOfNargs() + 2); Address address(fun, JSFunction::offsetOfNargs()); int32_t bit = IMM32_16ADJ(JSFunction::INTERPRETED); branchTest32(Assembler::NonZero, address, Imm32(bit), label); } void MacroAssembler::branchFunctionKind(Condition cond, JSFunction::FunctionKind kind, Register fun, Register scratch, Label* label) { // 16-bit loads are slow and unaligned 32-bit loads may be too so // perform an aligned 32-bit load and adjust the bitmask accordingly. MOZ_ASSERT(JSFunction::offsetOfNargs() % sizeof(uint32_t) == 0); MOZ_ASSERT(JSFunction::offsetOfFlags() == JSFunction::offsetOfNargs() + 2); Address address(fun, JSFunction::offsetOfNargs()); int32_t mask = IMM32_16ADJ(JSFunction::FUNCTION_KIND_MASK); int32_t bit = IMM32_16ADJ(kind << JSFunction::FUNCTION_KIND_SHIFT); load32(address, scratch); and32(Imm32(mask), scratch); branch32(cond, scratch, Imm32(bit), label); } void MacroAssembler::branchTestObjClass(Condition cond, Register obj, Register scratch, const js::Class* clasp, Label* label) { loadObjGroup(obj, scratch); branchPtr(cond, Address(scratch, ObjectGroup::offsetOfClasp()), ImmPtr(clasp), label); } void MacroAssembler::branchTestObjShape(Condition cond, Register obj, const Shape* shape, Label* label) { branchPtr(cond, Address(obj, ShapedObject::offsetOfShape()), ImmGCPtr(shape), label); } void MacroAssembler::branchTestObjShape(Condition cond, Register obj, Register shape, Label* label) { branchPtr(cond, Address(obj, ShapedObject::offsetOfShape()), shape, label); } void MacroAssembler::branchTestObjGroup(Condition cond, Register obj, ObjectGroup* group, Label* label) { branchPtr(cond, Address(obj, JSObject::offsetOfGroup()), ImmGCPtr(group), label); } void MacroAssembler::branchTestObjGroup(Condition cond, Register obj, Register group, Label* label) { branchPtr(cond, Address(obj, JSObject::offsetOfGroup()), group, label); } void MacroAssembler::branchTestObjectTruthy(bool truthy, Register objReg, Register scratch, Label* slowCheck, Label* checked) { // The branches to out-of-line code here implement a conservative version // of the JSObject::isWrapper test performed in EmulatesUndefined. If none // of the branches are taken, we can check class flags directly. loadObjClass(objReg, scratch); Address flags(scratch, Class::offsetOfFlags()); branchTestClassIsProxy(true, scratch, slowCheck); Condition cond = truthy ? Assembler::Zero : Assembler::NonZero; branchTest32(cond, flags, Imm32(JSCLASS_EMULATES_UNDEFINED), checked); } void MacroAssembler::branchTestClassIsProxy(bool proxy, Register clasp, Label* label) { branchTest32(proxy ? Assembler::NonZero : Assembler::Zero, Address(clasp, Class::offsetOfFlags()), Imm32(JSCLASS_IS_PROXY), label); } void MacroAssembler::branchTestObjectIsProxy(bool proxy, Register object, Register scratch, Label* label) { loadObjClass(object, scratch); branchTestClassIsProxy(proxy, scratch, label); } void MacroAssembler::branchTestProxyHandlerFamily(Condition cond, Register proxy, Register scratch, const void* handlerp, Label* label) { Address handlerAddr(proxy, ProxyObject::offsetOfHandler()); loadPtr(handlerAddr, scratch); Address familyAddr(scratch, BaseProxyHandler::offsetOfFamily()); branchPtr(cond, familyAddr, ImmPtr(handlerp), label); } template void MacroAssembler::branchTestMIRType(Condition cond, const Value& val, MIRType type, Label* label) { switch (type) { case MIRType::Null: return branchTestNull(cond, val, label); case MIRType::Undefined: return branchTestUndefined(cond, val, label); case MIRType::Boolean: return branchTestBoolean(cond, val, label); case MIRType::Int32: return branchTestInt32(cond, val, label); case MIRType::String: return branchTestString(cond, val, label); case MIRType::Symbol: return branchTestSymbol(cond, val, label); case MIRType::Object: return branchTestObject(cond, val, label); case MIRType::Double: return branchTestDouble(cond, val, label); case MIRType::MagicOptimizedArguments: // Fall through. case MIRType::MagicIsConstructing: case MIRType::MagicHole: return branchTestMagic(cond, val, label); default: MOZ_CRASH("Bad MIRType"); } } void MacroAssembler::branchTestNeedsIncrementalBarrier(Condition cond, Label* label) { MOZ_ASSERT(cond == Zero || cond == NonZero); CompileZone* zone = GetJitContext()->compartment->zone(); AbsoluteAddress needsBarrierAddr(zone->addressOfNeedsIncrementalBarrier()); branchTest32(cond, needsBarrierAddr, Imm32(0x1), label); } void MacroAssembler::branchTestMagicValue(Condition cond, const ValueOperand& val, JSWhyMagic why, Label* label) { MOZ_ASSERT(cond == Equal || cond == NotEqual); branchTestValue(cond, val, MagicValue(why), label); } void MacroAssembler::branchDoubleNotInInt64Range(Address src, Register temp, Label* fail) { // Tests if double is in [INT64_MIN; INT64_MAX] range uint32_t EXPONENT_MASK = 0x7ff00000; uint32_t EXPONENT_SHIFT = FloatingPoint::kExponentShift - 32; uint32_t TOO_BIG_EXPONENT = (FloatingPoint::kExponentBias + 63) << EXPONENT_SHIFT; load32(Address(src.base, src.offset + sizeof(int32_t)), temp); and32(Imm32(EXPONENT_MASK), temp); branch32(Assembler::GreaterThanOrEqual, temp, Imm32(TOO_BIG_EXPONENT), fail); } void MacroAssembler::branchDoubleNotInUInt64Range(Address src, Register temp, Label* fail) { // Note: returns failure on -0.0 // Tests if double is in [0; UINT64_MAX] range // Take the sign also in the equation. That way we can compare in one test? uint32_t EXPONENT_MASK = 0xfff00000; uint32_t EXPONENT_SHIFT = FloatingPoint::kExponentShift - 32; uint32_t TOO_BIG_EXPONENT = (FloatingPoint::kExponentBias + 64) << EXPONENT_SHIFT; load32(Address(src.base, src.offset + sizeof(int32_t)), temp); and32(Imm32(EXPONENT_MASK), temp); branch32(Assembler::AboveOrEqual, temp, Imm32(TOO_BIG_EXPONENT), fail); } void MacroAssembler::branchFloat32NotInInt64Range(Address src, Register temp, Label* fail) { // Tests if float is in [INT64_MIN; INT64_MAX] range uint32_t EXPONENT_MASK = 0x7f800000; uint32_t EXPONENT_SHIFT = FloatingPoint::kExponentShift; uint32_t TOO_BIG_EXPONENT = (FloatingPoint::kExponentBias + 63) << EXPONENT_SHIFT; load32(src, temp); and32(Imm32(EXPONENT_MASK), temp); branch32(Assembler::GreaterThanOrEqual, temp, Imm32(TOO_BIG_EXPONENT), fail); } void MacroAssembler::branchFloat32NotInUInt64Range(Address src, Register temp, Label* fail) { // Note: returns failure on -0.0 // Tests if float is in [0; UINT64_MAX] range // Take the sign also in the equation. That way we can compare in one test? uint32_t EXPONENT_MASK = 0xff800000; uint32_t EXPONENT_SHIFT = FloatingPoint::kExponentShift; uint32_t TOO_BIG_EXPONENT = (FloatingPoint::kExponentBias + 64) << EXPONENT_SHIFT; load32(src, temp); and32(Imm32(EXPONENT_MASK), temp); branch32(Assembler::AboveOrEqual, temp, Imm32(TOO_BIG_EXPONENT), fail); } // ======================================================================== // Canonicalization primitives. void MacroAssembler::canonicalizeFloat(FloatRegister reg) { Label notNaN; branchFloat(DoubleOrdered, reg, reg, ¬NaN); loadConstantFloat32(float(JS::GenericNaN()), reg); bind(¬NaN); } void MacroAssembler::canonicalizeFloatIfDeterministic(FloatRegister reg) { #ifdef JS_MORE_DETERMINISTIC // See the comment in TypedArrayObjectTemplate::getIndexValue. canonicalizeFloat(reg); #endif // JS_MORE_DETERMINISTIC } void MacroAssembler::canonicalizeDouble(FloatRegister reg) { Label notNaN; branchDouble(DoubleOrdered, reg, reg, ¬NaN); loadConstantDouble(JS::GenericNaN(), reg); bind(¬NaN); } void MacroAssembler::canonicalizeDoubleIfDeterministic(FloatRegister reg) { #ifdef JS_MORE_DETERMINISTIC // See the comment in TypedArrayObjectTemplate::getIndexValue. canonicalizeDouble(reg); #endif // JS_MORE_DETERMINISTIC } // ======================================================================== // Memory access primitives. template void MacroAssembler::storeDouble(FloatRegister src, const T& dest) { canonicalizeDoubleIfDeterministic(src); storeUncanonicalizedDouble(src, dest); } template void MacroAssembler::storeDouble(FloatRegister src, const Address& dest); template void MacroAssembler::storeDouble(FloatRegister src, const BaseIndex& dest); template void MacroAssembler::storeFloat32(FloatRegister src, const T& dest) { canonicalizeFloatIfDeterministic(src); storeUncanonicalizedFloat32(src, dest); } template void MacroAssembler::storeFloat32(FloatRegister src, const Address& dest); template void MacroAssembler::storeFloat32(FloatRegister src, const BaseIndex& dest); //}}} check_macroassembler_style // =============================================================== #ifndef JS_CODEGEN_ARM64 template void MacroAssembler::branchTestStackPtr(Condition cond, T t, Label* label) { branchTestPtr(cond, getStackPointer(), t, label); } template void MacroAssembler::branchStackPtr(Condition cond, T rhs, Label* label) { branchPtr(cond, getStackPointer(), rhs, label); } template void MacroAssembler::branchStackPtrRhs(Condition cond, T lhs, Label* label) { branchPtr(cond, lhs, getStackPointer(), label); } template void MacroAssembler::addToStackPtr(T t) { addPtr(t, getStackPointer()); } template void MacroAssembler::addStackPtrTo(T t) { addPtr(getStackPointer(), t); } void MacroAssembler::reserveStack(uint32_t amount) { subFromStackPtr(Imm32(amount)); adjustFrame(amount); } #endif // !JS_CODEGEN_ARM64 template void MacroAssembler::storeObjectOrNull(Register src, const T& dest) { Label notNull, done; branchTestPtr(Assembler::NonZero, src, src, ¬Null); storeValue(NullValue(), dest); jump(&done); bind(¬Null); storeValue(JSVAL_TYPE_OBJECT, src, dest); bind(&done); } void MacroAssembler::assertStackAlignment(uint32_t alignment, int32_t offset /* = 0 */) { #ifdef DEBUG Label ok, bad; MOZ_ASSERT(mozilla::IsPowerOfTwo(alignment)); // Wrap around the offset to be a non-negative number. offset %= alignment; if (offset < 0) offset += alignment; // Test if each bit from offset is set. uint32_t off = offset; while (off) { uint32_t lowestBit = 1 << mozilla::CountTrailingZeroes32(off); branchTestStackPtr(Assembler::Zero, Imm32(lowestBit), &bad); off ^= lowestBit; } // Check that all remaining bits are zero. branchTestStackPtr(Assembler::Zero, Imm32((alignment - 1) ^ offset), &ok); bind(&bad); breakpoint(); bind(&ok); #endif } void MacroAssembler::storeCallBoolResult(Register reg) { if (reg != ReturnReg) mov(ReturnReg, reg); // C++ compilers like to only use the bottom byte for bools, but we // need to maintain the entire register. and32(Imm32(0xFF), reg); } void MacroAssembler::storeCallInt32Result(Register reg) { #if JS_BITS_PER_WORD == 32 storeCallPointerResult(reg); #else // Ensure the upper 32 bits are cleared. move32(ReturnReg, reg); #endif } void MacroAssembler::storeCallResultValue(AnyRegister dest) { unboxValue(JSReturnOperand, dest); } void MacroAssembler::storeCallResultValue(TypedOrValueRegister dest) { if (dest.hasValue()) storeCallResultValue(dest.valueReg()); else storeCallResultValue(dest.typedReg()); } } // namespace jit } // namespace js #endif /* jit_MacroAssembler_inl_h */