Ryujinx/ChocolArm64/Instructions/InstEmitMemoryHelper.cs
gdkchan a731ab3a2a Add a new JIT compiler for CPU code (#693)
* Start of the ARMeilleure project

* Refactoring around the old IRAdapter, now renamed to PreAllocator

* Optimize the LowestBitSet method

* Add CLZ support and fix CLS implementation

* Add missing Equals and GetHashCode overrides on some structs, misc small tweaks

* Implement the ByteSwap IR instruction, and some refactoring on the assembler

* Implement the DivideUI IR instruction and fix 64-bits IDIV

* Correct constant operand type on CSINC

* Move division instructions implementation to InstEmitDiv

* Fix destination type for the ConditionalSelect IR instruction

* Implement UMULH and SMULH, with new IR instructions

* Fix some issues with shift instructions

* Fix constant types for BFM instructions

* Fix up new tests using the new V128 struct

* Update tests

* Move DIV tests to a separate file

* Add support for calls, and some instructions that depends on them

* Start adding support for SIMD & FP types, along with some of the related ARM instructions

* Fix some typos and the divide instruction with FP operands

* Fix wrong method call on Clz_V

* Implement ARM FP & SIMD move instructions, Saddlv_V, and misc. fixes

* Implement SIMD logical instructions and more misc. fixes

* Fix PSRAD x86 instruction encoding, TRN, UABD and UABDL implementations

* Implement float conversion instruction, merge in LDj3SNuD fixes, and some other misc. fixes

* Implement SIMD shift instruction and fix Dup_V

* Add SCVTF and UCVTF (vector, fixed-point) variants to the opcode table

* Fix check with tolerance on tester

* Implement FP & SIMD comparison instructions, and some fixes

* Update FCVT (Scalar) encoding on the table to support the Half-float variants

* Support passing V128 structs, some cleanup on the register allocator, merge LDj3SNuD fixes

* Use old memory access methods, made a start on SIMD memory insts support, some fixes

* Fix float constant passed to functions, save and restore non-volatile XMM registers, other fixes

* Fix arguments count with struct return values, other fixes

* More instructions

* Misc. fixes and integrate LDj3SNuD fixes

* Update tests

* Add a faster linear scan allocator, unwinding support on windows, and other changes

* Update Ryujinx.HLE

* Update Ryujinx.Graphics

* Fix V128 return pointer passing, RCX is clobbered

* Update Ryujinx.Tests

* Update ITimeZoneService

* Stop using GetFunctionPointer as that can't be called from native code, misc. fixes and tweaks

* Use generic GetFunctionPointerForDelegate method and other tweaks

* Some refactoring on the code generator, assert on invalid operations and use a separate enum for intrinsics

* Remove some unused code on the assembler

* Fix REX.W prefix regression on float conversion instructions, add some sort of profiler

* Add hardware capability detection

* Fix regression on Sha1h and revert Fcm** changes

* Add SSE2-only paths on vector extract and insert, some refactoring on the pre-allocator

* Fix silly mistake introduced on last commit on CpuId

* Generate inline stack probes when the stack allocation is too large

* Initial support for the System-V ABI

* Support multiple destination operands

* Fix SSE2 VectorInsert8 path, and other fixes

* Change placement of XMM callee save and restore code to match other compilers

* Rename Dest to Destination and Inst to Instruction

* Fix a regression related to calls and the V128 type

* Add an extra space on comments to match code style

* Some refactoring

* Fix vector insert FP32 SSE2 path

* Port over the ARM32 instructions

* Avoid memory protection races on JIT Cache

* Another fix on VectorInsert FP32 (thanks to LDj3SNuD

* Float operands don't need to use the same register when VEX is supported

* Add a new register allocator, higher quality code for hot code (tier up), and other tweaks

* Some nits, small improvements on the pre allocator

* CpuThreadState is gone

* Allow changing CPU emulators with a config entry

* Add runtime identifiers on the ARMeilleure project

* Allow switching between CPUs through a config entry (pt. 2)

* Change win10-x64 to win-x64 on projects

* Update the Ryujinx project to use ARMeilleure

* Ensure that the selected register is valid on the hybrid allocator

* Allow exiting on returns to 0 (should fix test regression)

* Remove register assignments for most used variables on the hybrid allocator

* Do not use fixed registers as spill temp

* Add missing namespace and remove unneeded using

* Address PR feedback

* Fix types, etc

* Enable AssumeStrictAbiCompliance by default

* Ensure that Spill and Fill don't load or store any more than necessary
2019-08-08 21:56:22 +03:00

475 lines
15 KiB
C#

using ChocolArm64.Decoders;
using ChocolArm64.IntermediateRepresentation;
using ChocolArm64.Memory;
using ChocolArm64.State;
using ChocolArm64.Translation;
using System;
using System.Reflection.Emit;
using System.Runtime.Intrinsics.X86;
namespace ChocolArm64.Instructions
{
static class InstEmitMemoryHelper
{
private static int _tempIntAddress = ILEmitterCtx.GetIntTempIndex();
private static int _tempIntValue = ILEmitterCtx.GetIntTempIndex();
private static int _tempIntPtAddr = ILEmitterCtx.GetIntTempIndex();
private static int _tempVecValue = ILEmitterCtx.GetVecTempIndex();
private enum Extension
{
Zx,
Sx32,
Sx64
}
public static void EmitReadZxCall(ILEmitterCtx context, int size)
{
EmitReadCall(context, Extension.Zx, size);
}
public static void EmitReadSx32Call(ILEmitterCtx context, int size)
{
EmitReadCall(context, Extension.Sx32, size);
}
public static void EmitReadSx64Call(ILEmitterCtx context, int size)
{
EmitReadCall(context, Extension.Sx64, size);
}
private static void EmitReadCall(ILEmitterCtx context, Extension ext, int size)
{
// Save the address into a temp.
context.EmitStint(_tempIntAddress);
bool isSimd = IsSimd(context);
if (size < 0 || size > (isSimd ? 4 : 3))
{
throw new ArgumentOutOfRangeException(nameof(size));
}
if (isSimd)
{
if (context.Tier == TranslationTier.Tier0 || !Sse2.IsSupported || size < 2)
{
EmitReadVectorFallback(context, size);
}
else
{
EmitReadVector(context, size);
}
}
else
{
if (context.Tier == TranslationTier.Tier0)
{
EmitReadIntFallback(context, size);
}
else
{
EmitReadInt(context, size);
}
}
if (!isSimd)
{
if (ext == Extension.Sx32 ||
ext == Extension.Sx64)
{
switch (size)
{
case 0: context.Emit(OpCodes.Conv_I1); break;
case 1: context.Emit(OpCodes.Conv_I2); break;
case 2: context.Emit(OpCodes.Conv_I4); break;
}
}
if (size < 3)
{
context.Emit(ext == Extension.Sx64
? OpCodes.Conv_I8
: OpCodes.Conv_U8);
}
}
}
public static void EmitWriteCall(ILEmitterCtx context, int size)
{
bool isSimd = IsSimd(context);
// Save the value into a temp.
if (isSimd)
{
context.EmitStvec(_tempVecValue);
}
else
{
context.EmitStint(_tempIntValue);
}
// Save the address into a temp.
context.EmitStint(_tempIntAddress);
if (size < 0 || size > (isSimd ? 4 : 3))
{
throw new ArgumentOutOfRangeException(nameof(size));
}
if (isSimd)
{
if (context.Tier == TranslationTier.Tier0 || !Sse2.IsSupported || size < 2)
{
EmitWriteVectorFallback(context, size);
}
else
{
EmitWriteVector(context, size);
}
}
else
{
if (context.Tier == TranslationTier.Tier0)
{
EmitWriteIntFallback(context, size);
}
else
{
EmitWriteInt(context, size);
}
}
}
private static bool IsSimd(ILEmitterCtx context)
{
return context.CurrOp is IOpCodeSimd64 &&
!(context.CurrOp is OpCodeSimdMemMs64 ||
context.CurrOp is OpCodeSimdMemSs64);
}
private static void EmitReadInt(ILEmitterCtx context, int size)
{
EmitAddressCheck(context, size);
ILLabel lblFastPath = new ILLabel();
ILLabel lblSlowPath = new ILLabel();
ILLabel lblEnd = new ILLabel();
context.Emit(OpCodes.Brfalse_S, lblFastPath);
context.MarkLabel(lblSlowPath);
EmitReadIntFallback(context, size);
context.Emit(OpCodes.Br, lblEnd);
context.MarkLabel(lblFastPath);
EmitPtPointerLoad(context, lblSlowPath);
switch (size)
{
case 0: context.Emit(OpCodes.Ldind_U1); break;
case 1: context.Emit(OpCodes.Ldind_U2); break;
case 2: context.Emit(OpCodes.Ldind_U4); break;
case 3: context.Emit(OpCodes.Ldind_I8); break;
}
context.MarkLabel(lblEnd);
}
private static void EmitReadVector(ILEmitterCtx context, int size)
{
EmitAddressCheck(context, size);
ILLabel lblFastPath = new ILLabel();
ILLabel lblSlowPath = new ILLabel();
ILLabel lblEnd = new ILLabel();
context.Emit(OpCodes.Brfalse_S, lblFastPath);
context.MarkLabel(lblSlowPath);
EmitReadVectorFallback(context, size);
context.Emit(OpCodes.Br, lblEnd);
context.MarkLabel(lblFastPath);
EmitPtPointerLoad(context, lblSlowPath);
switch (size)
{
case 2: context.EmitCall(typeof(Sse), nameof(Sse.LoadScalarVector128)); break;
case 3:
{
Type[] types = new Type[] { typeof(double*) };
context.EmitCall(typeof(Sse2).GetMethod(nameof(Sse2.LoadScalarVector128), types));
break;
}
case 4: context.EmitCall(typeof(Sse), nameof(Sse.LoadAlignedVector128)); break;
throw new InvalidOperationException($"Invalid vector load size of {1 << size} bytes.");
}
context.MarkLabel(lblEnd);
}
private static void EmitWriteInt(ILEmitterCtx context, int size)
{
EmitAddressCheck(context, size);
ILLabel lblFastPath = new ILLabel();
ILLabel lblSlowPath = new ILLabel();
ILLabel lblEnd = new ILLabel();
context.Emit(OpCodes.Brfalse_S, lblFastPath);
context.MarkLabel(lblSlowPath);
EmitWriteIntFallback(context, size);
context.Emit(OpCodes.Br, lblEnd);
context.MarkLabel(lblFastPath);
EmitPtPointerLoad(context, lblSlowPath);
context.EmitLdint(_tempIntValue);
if (size < 3)
{
context.Emit(OpCodes.Conv_U4);
}
switch (size)
{
case 0: context.Emit(OpCodes.Stind_I1); break;
case 1: context.Emit(OpCodes.Stind_I2); break;
case 2: context.Emit(OpCodes.Stind_I4); break;
case 3: context.Emit(OpCodes.Stind_I8); break;
}
context.MarkLabel(lblEnd);
}
private static void EmitWriteVector(ILEmitterCtx context, int size)
{
EmitAddressCheck(context, size);
ILLabel lblFastPath = new ILLabel();
ILLabel lblSlowPath = new ILLabel();
ILLabel lblEnd = new ILLabel();
context.Emit(OpCodes.Brfalse_S, lblFastPath);
context.MarkLabel(lblSlowPath);
EmitWriteVectorFallback(context, size);
context.Emit(OpCodes.Br, lblEnd);
context.MarkLabel(lblFastPath);
EmitPtPointerLoad(context, lblSlowPath);
context.EmitLdvec(_tempVecValue);
switch (size)
{
case 2: context.EmitCall(typeof(Sse), nameof(Sse.StoreScalar)); break;
case 3: context.EmitCall(typeof(Sse2), nameof(Sse2.StoreScalar)); break;
case 4: context.EmitCall(typeof(Sse), nameof(Sse.StoreAligned)); break;
default: throw new InvalidOperationException($"Invalid vector store size of {1 << size} bytes.");
}
context.MarkLabel(lblEnd);
}
private static void EmitAddressCheck(ILEmitterCtx context, int size)
{
long addressCheckMask = ~(context.Memory.AddressSpaceSize - 1);
addressCheckMask |= (1u << size) - 1;
context.EmitLdint(_tempIntAddress);
context.EmitLdc_I(addressCheckMask);
context.Emit(OpCodes.And);
}
private static void EmitPtPointerLoad(ILEmitterCtx context, ILLabel lblFallbackPath)
{
context.EmitLdc_I8(context.Memory.PageTable.ToInt64());
context.Emit(OpCodes.Conv_I);
int bit = MemoryManager.PageBits;
do
{
context.EmitLdint(_tempIntAddress);
if (context.CurrOp.RegisterSize == RegisterSize.Int32)
{
context.Emit(OpCodes.Conv_U8);
}
context.EmitLsr(bit);
bit += context.Memory.PtLevelBits;
if (bit < context.Memory.AddressSpaceBits)
{
context.EmitLdc_I8(context.Memory.PtLevelMask);
context.Emit(OpCodes.And);
}
context.EmitLdc_I8(IntPtr.Size);
context.Emit(OpCodes.Mul);
context.Emit(OpCodes.Conv_I);
context.Emit(OpCodes.Add);
context.Emit(OpCodes.Ldind_I);
}
while (bit < context.Memory.AddressSpaceBits);
if (!context.Memory.HasWriteWatchSupport)
{
context.Emit(OpCodes.Conv_U8);
context.EmitStint(_tempIntPtAddr);
context.EmitLdint(_tempIntPtAddr);
context.EmitLdc_I8(MemoryManager.PteFlagsMask);
context.Emit(OpCodes.And);
context.Emit(OpCodes.Brtrue, lblFallbackPath);
context.EmitLdint(_tempIntPtAddr);
context.Emit(OpCodes.Conv_I);
}
context.EmitLdint(_tempIntAddress);
context.EmitLdc_I(MemoryManager.PageMask);
context.Emit(OpCodes.And);
context.Emit(OpCodes.Conv_I);
context.Emit(OpCodes.Add);
}
private static void EmitReadIntFallback(ILEmitterCtx context, int size)
{
context.EmitLdarg(TranslatedSub.MemoryArgIdx);
context.EmitLdint(_tempIntAddress);
if (context.CurrOp.RegisterSize == RegisterSize.Int32)
{
context.Emit(OpCodes.Conv_U8);
}
string fallbackMethodName = null;
switch (size)
{
case 0: fallbackMethodName = nameof(MemoryManager.ReadByte); break;
case 1: fallbackMethodName = nameof(MemoryManager.ReadUInt16); break;
case 2: fallbackMethodName = nameof(MemoryManager.ReadUInt32); break;
case 3: fallbackMethodName = nameof(MemoryManager.ReadUInt64); break;
}
context.EmitCall(typeof(MemoryManager), fallbackMethodName);
}
private static void EmitReadVectorFallback(ILEmitterCtx context, int size)
{
context.EmitLdarg(TranslatedSub.MemoryArgIdx);
context.EmitLdint(_tempIntAddress);
if (context.CurrOp.RegisterSize == RegisterSize.Int32)
{
context.Emit(OpCodes.Conv_U8);
}
string fallbackMethodName = null;
switch (size)
{
case 0: fallbackMethodName = nameof(MemoryManager.ReadVector8); break;
case 1: fallbackMethodName = nameof(MemoryManager.ReadVector16); break;
case 2: fallbackMethodName = nameof(MemoryManager.ReadVector32); break;
case 3: fallbackMethodName = nameof(MemoryManager.ReadVector64); break;
case 4: fallbackMethodName = nameof(MemoryManager.ReadVector128); break;
}
context.EmitCall(typeof(MemoryManager), fallbackMethodName);
}
private static void EmitWriteIntFallback(ILEmitterCtx context, int size)
{
context.EmitLdarg(TranslatedSub.MemoryArgIdx);
context.EmitLdint(_tempIntAddress);
if (context.CurrOp.RegisterSize == RegisterSize.Int32)
{
context.Emit(OpCodes.Conv_U8);
}
context.EmitLdint(_tempIntValue);
if (size < 3)
{
context.Emit(OpCodes.Conv_U4);
}
string fallbackMethodName = null;
switch (size)
{
case 0: fallbackMethodName = nameof(MemoryManager.WriteByte); break;
case 1: fallbackMethodName = nameof(MemoryManager.WriteUInt16); break;
case 2: fallbackMethodName = nameof(MemoryManager.WriteUInt32); break;
case 3: fallbackMethodName = nameof(MemoryManager.WriteUInt64); break;
}
context.EmitCall(typeof(MemoryManager), fallbackMethodName);
}
private static void EmitWriteVectorFallback(ILEmitterCtx context, int size)
{
context.EmitLdarg(TranslatedSub.MemoryArgIdx);
context.EmitLdint(_tempIntAddress);
if (context.CurrOp.RegisterSize == RegisterSize.Int32)
{
context.Emit(OpCodes.Conv_U8);
}
context.EmitLdvec(_tempVecValue);
string fallbackMethodName = null;
switch (size)
{
case 0: fallbackMethodName = nameof(MemoryManager.WriteVector8); break;
case 1: fallbackMethodName = nameof(MemoryManager.WriteVector16); break;
case 2: fallbackMethodName = nameof(MemoryManager.WriteVector32); break;
case 3: fallbackMethodName = nameof(MemoryManager.WriteVector64); break;
case 4: fallbackMethodName = nameof(MemoryManager.WriteVector128Internal); break;
}
context.EmitCall(typeof(MemoryManager), fallbackMethodName);
}
}
}