Ryujinx/ARMeilleure/Translation/EmitterContext.cs

563 lines
16 KiB
C#
Raw Normal View History

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 20:56:22 +02:00
using ARMeilleure.IntermediateRepresentation;
using ARMeilleure.State;
using System;
using System.Collections.Generic;
using System.Runtime.InteropServices;
using static ARMeilleure.IntermediateRepresentation.OperandHelper;
namespace ARMeilleure.Translation
{
class EmitterContext
{
private Dictionary<Operand, BasicBlock> _irLabels;
private IntrusiveList<BasicBlock> _irBlocks;
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 20:56:22 +02:00
private BasicBlock _irBlock;
private bool _needsNewBlock;
public EmitterContext()
{
_irLabels = new Dictionary<Operand, BasicBlock>();
_irBlocks = new IntrusiveList<BasicBlock>();
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 20:56:22 +02:00
_needsNewBlock = true;
}
public Operand Add(Operand op1, Operand op2)
{
return Add(Instruction.Add, Local(op1.Type), op1, op2);
}
public Operand BitwiseAnd(Operand op1, Operand op2)
{
return Add(Instruction.BitwiseAnd, Local(op1.Type), op1, op2);
}
public Operand BitwiseExclusiveOr(Operand op1, Operand op2)
{
return Add(Instruction.BitwiseExclusiveOr, Local(op1.Type), op1, op2);
}
public Operand BitwiseNot(Operand op1)
{
return Add(Instruction.BitwiseNot, Local(op1.Type), op1);
}
public Operand BitwiseOr(Operand op1, Operand op2)
{
return Add(Instruction.BitwiseOr, Local(op1.Type), op1, op2);
}
public void Branch(Operand label)
{
Add(Instruction.Branch, null);
BranchToLabel(label);
}
public void BranchIfFalse(Operand label, Operand op1)
{
Add(Instruction.BranchIfFalse, null, op1);
BranchToLabel(label);
}
public void BranchIfTrue(Operand label, Operand op1)
{
Add(Instruction.BranchIfTrue, null, op1);
BranchToLabel(label);
}
public Operand ByteSwap(Operand op1)
{
return Add(Instruction.ByteSwap, Local(op1.Type), op1);
}
public Operand Call(Delegate func, params Operand[] callArgs)
{
// Add the delegate to the cache to ensure it will not be garbage collected.
func = DelegateCache.GetOrAdd(func);
IntPtr ptr = Marshal.GetFunctionPointerForDelegate<Delegate>(func);
OperandType returnType = GetOperandType(func.Method.ReturnType);
return Call(Const(ptr.ToInt64()), returnType, callArgs);
}
private static Dictionary<TypeCode, OperandType> _typeCodeToOperandTypeMap =
new Dictionary<TypeCode, OperandType>()
{
{ TypeCode.Boolean, OperandType.I32 },
{ TypeCode.Byte, OperandType.I32 },
{ TypeCode.Char, OperandType.I32 },
{ TypeCode.Double, OperandType.FP64 },
{ TypeCode.Int16, OperandType.I32 },
{ TypeCode.Int32, OperandType.I32 },
{ TypeCode.Int64, OperandType.I64 },
{ TypeCode.SByte, OperandType.I32 },
{ TypeCode.Single, OperandType.FP32 },
{ TypeCode.UInt16, OperandType.I32 },
{ TypeCode.UInt32, OperandType.I32 },
{ TypeCode.UInt64, OperandType.I64 }
};
private static OperandType GetOperandType(Type type)
{
if (_typeCodeToOperandTypeMap.TryGetValue(Type.GetTypeCode(type), out OperandType ot))
{
return ot;
}
else if (type == typeof(V128))
{
return OperandType.V128;
}
else if (type == typeof(void))
{
return OperandType.None;
}
throw new ArgumentException($"Invalid type \"{type.Name}\".");
}
public Operand Call(Operand address, OperandType returnType, params Operand[] callArgs)
{
Operand[] args = new Operand[callArgs.Length + 1];
args[0] = address;
Array.Copy(callArgs, 0, args, 1, callArgs.Length);
if (returnType != OperandType.None)
{
return Add(Instruction.Call, Local(returnType), args);
}
else
{
return Add(Instruction.Call, null, args);
}
}
public Operand CompareAndSwap128(Operand address, Operand expected, Operand desired)
{
return Add(Instruction.CompareAndSwap128, Local(OperandType.V128), address, expected, desired);
}
public Operand ConditionalSelect(Operand op1, Operand op2, Operand op3)
{
return Add(Instruction.ConditionalSelect, Local(op2.Type), op1, op2, op3);
}
public Operand ConvertI64ToI32(Operand op1)
{
if (op1.Type != OperandType.I64)
{
throw new ArgumentException($"Invalid operand type \"{op1.Type}\".");
}
return Add(Instruction.ConvertI64ToI32, Local(OperandType.I32), op1);
}
public Operand ConvertToFP(OperandType type, Operand op1)
{
return Add(Instruction.ConvertToFP, Local(type), op1);
}
public Operand ConvertToFPUI(OperandType type, Operand op1)
{
return Add(Instruction.ConvertToFPUI, Local(type), op1);
}
public Operand Copy(Operand op1)
{
return Add(Instruction.Copy, Local(op1.Type), op1);
}
public Operand Copy(Operand dest, Operand op1)
{
if (dest.Kind != OperandKind.Register)
{
throw new ArgumentException($"Invalid dest operand kind \"{dest.Kind}\".");
}
return Add(Instruction.Copy, dest, op1);
}
public Operand CountLeadingZeros(Operand op1)
{
return Add(Instruction.CountLeadingZeros, Local(op1.Type), op1);
}
internal Operand CpuId()
{
return Add(Instruction.CpuId, Local(OperandType.I64));
}
public Operand Divide(Operand op1, Operand op2)
{
return Add(Instruction.Divide, Local(op1.Type), op1, op2);
}
public Operand DivideUI(Operand op1, Operand op2)
{
return Add(Instruction.DivideUI, Local(op1.Type), op1, op2);
}
public Operand ICompareEqual(Operand op1, Operand op2)
{
return Add(Instruction.CompareEqual, Local(OperandType.I32), op1, op2);
}
public Operand ICompareGreater(Operand op1, Operand op2)
{
return Add(Instruction.CompareGreater, Local(OperandType.I32), op1, op2);
}
public Operand ICompareGreaterOrEqual(Operand op1, Operand op2)
{
return Add(Instruction.CompareGreaterOrEqual, Local(OperandType.I32), op1, op2);
}
public Operand ICompareGreaterOrEqualUI(Operand op1, Operand op2)
{
return Add(Instruction.CompareGreaterOrEqualUI, Local(OperandType.I32), op1, op2);
}
public Operand ICompareGreaterUI(Operand op1, Operand op2)
{
return Add(Instruction.CompareGreaterUI, Local(OperandType.I32), op1, op2);
}
public Operand ICompareLess(Operand op1, Operand op2)
{
return Add(Instruction.CompareLess, Local(OperandType.I32), op1, op2);
}
public Operand ICompareLessOrEqual(Operand op1, Operand op2)
{
return Add(Instruction.CompareLessOrEqual, Local(OperandType.I32), op1, op2);
}
public Operand ICompareLessOrEqualUI(Operand op1, Operand op2)
{
return Add(Instruction.CompareLessOrEqualUI, Local(OperandType.I32), op1, op2);
}
public Operand ICompareLessUI(Operand op1, Operand op2)
{
return Add(Instruction.CompareLessUI, Local(OperandType.I32), op1, op2);
}
public Operand ICompareNotEqual(Operand op1, Operand op2)
{
return Add(Instruction.CompareNotEqual, Local(OperandType.I32), op1, op2);
}
public Operand Load(OperandType type, Operand address)
{
return Add(Instruction.Load, Local(type), address);
}
public Operand Load16(Operand address)
{
return Add(Instruction.Load16, Local(OperandType.I32), address);
}
public Operand Load8(Operand address)
{
return Add(Instruction.Load8, Local(OperandType.I32), address);
}
public Operand LoadArgument(OperandType type, int index)
{
return Add(Instruction.LoadArgument, Local(type), Const(index));
}
public void LoadFromContext()
{
_needsNewBlock = true;
Add(Instruction.LoadFromContext);
}
public Operand Multiply(Operand op1, Operand op2)
{
return Add(Instruction.Multiply, Local(op1.Type), op1, op2);
}
public Operand Multiply64HighSI(Operand op1, Operand op2)
{
return Add(Instruction.Multiply64HighSI, Local(OperandType.I64), op1, op2);
}
public Operand Multiply64HighUI(Operand op1, Operand op2)
{
return Add(Instruction.Multiply64HighUI, Local(OperandType.I64), op1, op2);
}
public Operand Negate(Operand op1)
{
return Add(Instruction.Negate, Local(op1.Type), op1);
}
public void Return()
{
Add(Instruction.Return);
_needsNewBlock = true;
}
public void Return(Operand op1)
{
Add(Instruction.Return, null, op1);
_needsNewBlock = true;
}
public Operand RotateRight(Operand op1, Operand op2)
{
return Add(Instruction.RotateRight, Local(op1.Type), op1, op2);
}
public Operand ShiftLeft(Operand op1, Operand op2)
{
return Add(Instruction.ShiftLeft, Local(op1.Type), op1, op2);
}
public Operand ShiftRightSI(Operand op1, Operand op2)
{
return Add(Instruction.ShiftRightSI, Local(op1.Type), op1, op2);
}
public Operand ShiftRightUI(Operand op1, Operand op2)
{
return Add(Instruction.ShiftRightUI, Local(op1.Type), op1, op2);
}
public Operand SignExtend16(OperandType type, Operand op1)
{
return Add(Instruction.SignExtend16, Local(type), op1);
}
public Operand SignExtend32(OperandType type, Operand op1)
{
return Add(Instruction.SignExtend32, Local(type), op1);
}
public Operand SignExtend8(OperandType type, Operand op1)
{
return Add(Instruction.SignExtend8, Local(type), op1);
}
public void Store(Operand address, Operand value)
{
Add(Instruction.Store, null, address, value);
}
public void Store16(Operand address, Operand value)
{
Add(Instruction.Store16, null, address, value);
}
public void Store8(Operand address, Operand value)
{
Add(Instruction.Store8, null, address, value);
}
public void StoreToContext()
{
Add(Instruction.StoreToContext);
_needsNewBlock = true;
}
public Operand Subtract(Operand op1, Operand op2)
{
return Add(Instruction.Subtract, Local(op1.Type), op1, op2);
}
public Operand VectorCreateScalar(Operand value)
{
return Add(Instruction.VectorCreateScalar, Local(OperandType.V128), value);
}
public Operand VectorExtract(OperandType type, Operand vector, int index)
{
return Add(Instruction.VectorExtract, Local(type), vector, Const(index));
}
public Operand VectorExtract16(Operand vector, int index)
{
return Add(Instruction.VectorExtract16, Local(OperandType.I32), vector, Const(index));
}
public Operand VectorExtract8(Operand vector, int index)
{
return Add(Instruction.VectorExtract8, Local(OperandType.I32), vector, Const(index));
}
public Operand VectorInsert(Operand vector, Operand value, int index)
{
return Add(Instruction.VectorInsert, Local(OperandType.V128), vector, value, Const(index));
}
public Operand VectorInsert16(Operand vector, Operand value, int index)
{
return Add(Instruction.VectorInsert16, Local(OperandType.V128), vector, value, Const(index));
}
public Operand VectorInsert8(Operand vector, Operand value, int index)
{
return Add(Instruction.VectorInsert8, Local(OperandType.V128), vector, value, Const(index));
}
public Operand VectorZero()
{
return Add(Instruction.VectorZero, Local(OperandType.V128));
}
public Operand VectorZeroUpper64(Operand vector)
{
return Add(Instruction.VectorZeroUpper64, Local(OperandType.V128), vector);
}
public Operand VectorZeroUpper96(Operand vector)
{
return Add(Instruction.VectorZeroUpper96, Local(OperandType.V128), vector);
}
public Operand ZeroExtend16(OperandType type, Operand op1)
{
return Add(Instruction.ZeroExtend16, Local(type), op1);
}
public Operand ZeroExtend32(OperandType type, Operand op1)
{
return Add(Instruction.ZeroExtend32, Local(type), op1);
}
public Operand ZeroExtend8(OperandType type, Operand op1)
{
return Add(Instruction.ZeroExtend8, Local(type), op1);
}
private Operand Add(Instruction inst, Operand dest = null, params Operand[] sources)
{
if (_needsNewBlock)
{
NewNextBlock();
}
Operation operation = new Operation(inst, dest, sources);
_irBlock.Operations.AddLast(operation);
return dest;
}
public Operand AddIntrinsic(Intrinsic intrin, params Operand[] args)
{
return Add(intrin, Local(OperandType.V128), args);
}
public Operand AddIntrinsicInt(Intrinsic intrin, params Operand[] args)
{
return Add(intrin, Local(OperandType.I32), args);
}
public Operand AddIntrinsicLong(Intrinsic intrin, params Operand[] args)
{
return Add(intrin, Local(OperandType.I64), args);
}
private Operand Add(Intrinsic intrin, Operand dest, params Operand[] sources)
{
if (_needsNewBlock)
{
NewNextBlock();
}
IntrinsicOperation operation = new IntrinsicOperation(intrin, dest, sources);
_irBlock.Operations.AddLast(operation);
return dest;
}
private void BranchToLabel(Operand label)
{
if (!_irLabels.TryGetValue(label, out BasicBlock branchBlock))
{
branchBlock = new BasicBlock();
_irLabels.Add(label, branchBlock);
}
_irBlock.Branch = branchBlock;
_needsNewBlock = true;
}
public void MarkLabel(Operand label)
{
if (_irLabels.TryGetValue(label, out BasicBlock nextBlock))
{
nextBlock.Index = _irBlocks.Count;
_irBlocks.AddLast(nextBlock);
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 20:56:22 +02:00
NextBlock(nextBlock);
}
else
{
NewNextBlock();
_irLabels.Add(label, _irBlock);
}
}
private void NewNextBlock()
{
BasicBlock block = new BasicBlock(_irBlocks.Count);
_irBlocks.AddLast(block);
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 20:56:22 +02:00
NextBlock(block);
}
private void NextBlock(BasicBlock nextBlock)
{
if (_irBlock != null && !EndsWithUnconditional(_irBlock))
{
_irBlock.Next = nextBlock;
}
_irBlock = nextBlock;
_needsNewBlock = false;
}
private static bool EndsWithUnconditional(BasicBlock block)
{
Operation lastOp = block.GetLastOp() as Operation;
if (lastOp == null)
{
return false;
}
return lastOp.Instruction == Instruction.Branch ||
lastOp.Instruction == Instruction.Return;
}
public ControlFlowGraph GetControlFlowGraph()
{
return new ControlFlowGraph(_irBlocks.First, _irBlocks);
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 20:56:22 +02:00
}
}
}