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.Decoders;
|
|
|
|
using ARMeilleure.IntermediateRepresentation;
|
|
|
|
using ARMeilleure.State;
|
|
|
|
using ARMeilleure.Translation;
|
|
|
|
using System;
|
|
|
|
|
|
|
|
using static ARMeilleure.Instructions.InstEmitHelper;
|
|
|
|
using static ARMeilleure.IntermediateRepresentation.OperandHelper;
|
|
|
|
|
|
|
|
namespace ARMeilleure.Instructions
|
|
|
|
{
|
|
|
|
static class InstEmitAluHelper
|
|
|
|
{
|
|
|
|
public static void EmitNZFlagsCheck(ArmEmitterContext context, Operand d)
|
|
|
|
{
|
|
|
|
SetFlag(context, PState.NFlag, context.ICompareLess (d, Const(d.Type, 0)));
|
|
|
|
SetFlag(context, PState.ZFlag, context.ICompareEqual(d, Const(d.Type, 0)));
|
|
|
|
}
|
|
|
|
|
|
|
|
public static void EmitAdcsCCheck(ArmEmitterContext context, Operand n, Operand d)
|
|
|
|
{
|
|
|
|
// C = (Rd == Rn && CIn) || Rd < Rn
|
|
|
|
Operand cIn = GetFlag(PState.CFlag);
|
|
|
|
|
|
|
|
Operand cOut = context.BitwiseAnd(context.ICompareEqual(d, n), cIn);
|
|
|
|
|
|
|
|
cOut = context.BitwiseOr(cOut, context.ICompareLessUI(d, n));
|
|
|
|
|
|
|
|
SetFlag(context, PState.CFlag, cOut);
|
|
|
|
}
|
|
|
|
|
|
|
|
public static void EmitAddsCCheck(ArmEmitterContext context, Operand n, Operand d)
|
|
|
|
{
|
|
|
|
// C = Rd < Rn
|
|
|
|
SetFlag(context, PState.CFlag, context.ICompareLessUI(d, n));
|
|
|
|
}
|
|
|
|
|
|
|
|
public static void EmitAddsVCheck(ArmEmitterContext context, Operand n, Operand m, Operand d)
|
|
|
|
{
|
|
|
|
// V = (Rd ^ Rn) & ~(Rn ^ Rm) < 0
|
|
|
|
Operand vOut = context.BitwiseExclusiveOr(d, n);
|
|
|
|
|
|
|
|
vOut = context.BitwiseAnd(vOut, context.BitwiseNot(context.BitwiseExclusiveOr(n, m)));
|
|
|
|
|
|
|
|
vOut = context.ICompareLess(vOut, Const(vOut.Type, 0));
|
|
|
|
|
|
|
|
SetFlag(context, PState.VFlag, vOut);
|
|
|
|
}
|
|
|
|
|
|
|
|
public static void EmitSbcsCCheck(ArmEmitterContext context, Operand n, Operand m)
|
|
|
|
{
|
|
|
|
// C = (Rn == Rm && CIn) || Rn > Rm
|
|
|
|
Operand cIn = GetFlag(PState.CFlag);
|
|
|
|
|
|
|
|
Operand cOut = context.BitwiseAnd(context.ICompareEqual(n, m), cIn);
|
|
|
|
|
|
|
|
cOut = context.BitwiseOr(cOut, context.ICompareGreaterUI(n, m));
|
|
|
|
|
|
|
|
SetFlag(context, PState.CFlag, cOut);
|
|
|
|
}
|
|
|
|
|
|
|
|
public static void EmitSubsCCheck(ArmEmitterContext context, Operand n, Operand m)
|
|
|
|
{
|
|
|
|
// C = Rn >= Rm
|
|
|
|
SetFlag(context, PState.CFlag, context.ICompareGreaterOrEqualUI(n, m));
|
|
|
|
}
|
|
|
|
|
|
|
|
public static void EmitSubsVCheck(ArmEmitterContext context, Operand n, Operand m, Operand d)
|
|
|
|
{
|
|
|
|
// V = (Rd ^ Rn) & (Rn ^ Rm) < 0
|
|
|
|
Operand vOut = context.BitwiseExclusiveOr(d, n);
|
|
|
|
|
|
|
|
vOut = context.BitwiseAnd(vOut, context.BitwiseExclusiveOr(n, m));
|
|
|
|
|
|
|
|
vOut = context.ICompareLess(vOut, Const(vOut.Type, 0));
|
|
|
|
|
|
|
|
SetFlag(context, PState.VFlag, vOut);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
public static Operand GetAluN(ArmEmitterContext context)
|
|
|
|
{
|
|
|
|
if (context.CurrOp is IOpCodeAlu op)
|
|
|
|
{
|
|
|
|
if (op.DataOp == DataOp.Logical || op is IOpCodeAluRs)
|
|
|
|
{
|
|
|
|
return GetIntOrZR(context, op.Rn);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
return GetIntOrSP(context, op.Rn);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else if (context.CurrOp is IOpCode32Alu op32)
|
|
|
|
{
|
|
|
|
return GetIntA32(context, op32.Rn);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
throw InvalidOpCodeType(context.CurrOp);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
public static Operand GetAluM(ArmEmitterContext context, bool setCarry = true)
|
|
|
|
{
|
|
|
|
switch (context.CurrOp)
|
|
|
|
{
|
|
|
|
// ARM32.
|
|
|
|
case OpCode32AluImm op:
|
|
|
|
{
|
|
|
|
if (op.SetFlags && op.IsRotated)
|
|
|
|
{
|
|
|
|
SetFlag(context, PState.CFlag, Const((uint)op.Immediate >> 31));
|
|
|
|
}
|
|
|
|
|
|
|
|
return Const(op.Immediate);
|
|
|
|
}
|
|
|
|
|
|
|
|
case OpCode32AluRsImm op: return GetMShiftedByImmediate(context, op, setCarry);
|
|
|
|
|
|
|
|
case OpCodeT16AluImm8 op: return Const(op.Immediate);
|
|
|
|
|
|
|
|
// ARM64.
|
|
|
|
case IOpCodeAluImm op:
|
|
|
|
{
|
|
|
|
if (op.GetOperandType() == OperandType.I32)
|
|
|
|
{
|
|
|
|
return Const((int)op.Immediate);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
return Const(op.Immediate);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
case IOpCodeAluRs op:
|
|
|
|
{
|
|
|
|
Operand value = GetIntOrZR(context, op.Rm);
|
|
|
|
|
|
|
|
switch (op.ShiftType)
|
|
|
|
{
|
|
|
|
case ShiftType.Lsl: value = context.ShiftLeft (value, Const(op.Shift)); break;
|
|
|
|
case ShiftType.Lsr: value = context.ShiftRightUI(value, Const(op.Shift)); break;
|
|
|
|
case ShiftType.Asr: value = context.ShiftRightSI(value, Const(op.Shift)); break;
|
|
|
|
case ShiftType.Ror: value = context.RotateRight (value, Const(op.Shift)); break;
|
|
|
|
}
|
|
|
|
|
|
|
|
return value;
|
|
|
|
}
|
|
|
|
|
|
|
|
case IOpCodeAluRx op:
|
|
|
|
{
|
|
|
|
Operand value = GetExtendedM(context, op.Rm, op.IntType);
|
|
|
|
|
|
|
|
value = context.ShiftLeft(value, Const(op.Shift));
|
|
|
|
|
|
|
|
return value;
|
|
|
|
}
|
|
|
|
|
|
|
|
default: throw InvalidOpCodeType(context.CurrOp);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
private static Exception InvalidOpCodeType(OpCode opCode)
|
|
|
|
{
|
|
|
|
return new InvalidOperationException($"Invalid OpCode type \"{opCode?.GetType().Name ?? "null"}\".");
|
|
|
|
}
|
|
|
|
|
|
|
|
// ARM32 helpers.
|
|
|
|
private static Operand GetMShiftedByImmediate(ArmEmitterContext context, OpCode32AluRsImm op, bool setCarry)
|
|
|
|
{
|
|
|
|
Operand m = GetIntA32(context, op.Rm);
|
|
|
|
|
|
|
|
int shift = op.Imm;
|
|
|
|
|
|
|
|
if (shift == 0)
|
|
|
|
{
|
|
|
|
switch (op.ShiftType)
|
|
|
|
{
|
|
|
|
case ShiftType.Lsr: shift = 32; break;
|
|
|
|
case ShiftType.Asr: shift = 32; break;
|
|
|
|
case ShiftType.Ror: shift = 1; break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (shift != 0)
|
|
|
|
{
|
|
|
|
setCarry &= op.SetFlags;
|
|
|
|
|
|
|
|
switch (op.ShiftType)
|
|
|
|
{
|
|
|
|
case ShiftType.Lsl: m = GetLslC(context, m, setCarry, shift); break;
|
|
|
|
case ShiftType.Lsr: m = GetLsrC(context, m, setCarry, shift); break;
|
|
|
|
case ShiftType.Asr: m = GetAsrC(context, m, setCarry, shift); break;
|
|
|
|
case ShiftType.Ror:
|
|
|
|
if (op.Imm != 0)
|
|
|
|
{
|
|
|
|
m = GetRorC(context, m, setCarry, shift);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
m = GetRrxC(context, m, setCarry);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return m;
|
|
|
|
}
|
|
|
|
|
|
|
|
private static Operand GetLslC(ArmEmitterContext context, Operand m, bool setCarry, int shift)
|
|
|
|
{
|
|
|
|
if ((uint)shift > 32)
|
|
|
|
{
|
|
|
|
return GetShiftByMoreThan32(context, setCarry);
|
|
|
|
}
|
|
|
|
else if (shift == 32)
|
|
|
|
{
|
|
|
|
if (setCarry)
|
|
|
|
{
|
|
|
|
SetCarryMLsb(context, m);
|
|
|
|
}
|
|
|
|
|
|
|
|
return Const(0);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
if (setCarry)
|
|
|
|
{
|
|
|
|
Operand cOut = context.ShiftRightUI(m, Const(32 - shift));
|
|
|
|
|
|
|
|
cOut = context.BitwiseAnd(cOut, Const(1));
|
|
|
|
|
|
|
|
SetFlag(context, PState.CFlag, cOut);
|
|
|
|
}
|
|
|
|
|
|
|
|
return context.ShiftLeft(m, Const(shift));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
private static Operand GetLsrC(ArmEmitterContext context, Operand m, bool setCarry, int shift)
|
|
|
|
{
|
|
|
|
if ((uint)shift > 32)
|
|
|
|
{
|
|
|
|
return GetShiftByMoreThan32(context, setCarry);
|
|
|
|
}
|
|
|
|
else if (shift == 32)
|
|
|
|
{
|
|
|
|
if (setCarry)
|
|
|
|
{
|
|
|
|
SetCarryMMsb(context, m);
|
|
|
|
}
|
|
|
|
|
|
|
|
return Const(0);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
if (setCarry)
|
|
|
|
{
|
|
|
|
SetCarryMShrOut(context, m, shift);
|
|
|
|
}
|
|
|
|
|
|
|
|
return context.ShiftRightUI(m, Const(shift));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
private static Operand GetShiftByMoreThan32(ArmEmitterContext context, bool setCarry)
|
|
|
|
{
|
|
|
|
if (setCarry)
|
|
|
|
{
|
2019-12-07 13:45:32 +01:00
|
|
|
SetFlag(context, PState.CFlag, Const(0));
|
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
|
|
|
}
|
|
|
|
|
|
|
|
return Const(0);
|
|
|
|
}
|
|
|
|
|
|
|
|
private static Operand GetAsrC(ArmEmitterContext context, Operand m, bool setCarry, int shift)
|
|
|
|
{
|
|
|
|
if ((uint)shift >= 32)
|
|
|
|
{
|
|
|
|
m = context.ShiftRightSI(m, Const(31));
|
|
|
|
|
|
|
|
if (setCarry)
|
|
|
|
{
|
|
|
|
SetCarryMLsb(context, m);
|
|
|
|
}
|
|
|
|
|
|
|
|
return m;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
if (setCarry)
|
|
|
|
{
|
|
|
|
SetCarryMShrOut(context, m, shift);
|
|
|
|
}
|
|
|
|
|
|
|
|
return context.ShiftRightSI(m, Const(shift));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
private static Operand GetRorC(ArmEmitterContext context, Operand m, bool setCarry, int shift)
|
|
|
|
{
|
|
|
|
shift &= 0x1f;
|
|
|
|
|
|
|
|
m = context.RotateRight(m, Const(shift));
|
|
|
|
|
|
|
|
if (setCarry)
|
|
|
|
{
|
|
|
|
SetCarryMMsb(context, m);
|
|
|
|
}
|
|
|
|
|
|
|
|
return m;
|
|
|
|
}
|
|
|
|
|
|
|
|
private static Operand GetRrxC(ArmEmitterContext context, Operand m, bool setCarry)
|
|
|
|
{
|
|
|
|
// Rotate right by 1 with carry.
|
|
|
|
Operand cIn = context.Copy(GetFlag(PState.CFlag));
|
|
|
|
|
|
|
|
if (setCarry)
|
|
|
|
{
|
|
|
|
SetCarryMLsb(context, m);
|
|
|
|
}
|
|
|
|
|
|
|
|
m = context.ShiftRightUI(m, Const(1));
|
|
|
|
|
|
|
|
m = context.BitwiseOr(m, context.ShiftLeft(cIn, Const(31)));
|
|
|
|
|
|
|
|
return m;
|
|
|
|
}
|
|
|
|
|
|
|
|
private static void SetCarryMLsb(ArmEmitterContext context, Operand m)
|
|
|
|
{
|
|
|
|
SetFlag(context, PState.CFlag, context.BitwiseAnd(m, Const(1)));
|
|
|
|
}
|
|
|
|
|
|
|
|
private static void SetCarryMMsb(ArmEmitterContext context, Operand m)
|
|
|
|
{
|
|
|
|
SetFlag(context, PState.CFlag, context.ShiftRightUI(m, Const(31)));
|
|
|
|
}
|
|
|
|
|
|
|
|
private static void SetCarryMShrOut(ArmEmitterContext context, Operand m, int shift)
|
|
|
|
{
|
|
|
|
Operand cOut = context.ShiftRightUI(m, Const(shift - 1));
|
|
|
|
|
|
|
|
cOut = context.BitwiseAnd(cOut, Const(1));
|
|
|
|
|
|
|
|
SetFlag(context, PState.CFlag, cOut);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|