ryujinx/ARMeilleure/Instructions/InstEmitMemoryEx.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

261 lines
10 KiB
C#

using ARMeilleure.Decoders;
using ARMeilleure.IntermediateRepresentation;
using ARMeilleure.Translation;
using System;
using System.Diagnostics;
using static ARMeilleure.Instructions.InstEmitHelper;
using static ARMeilleure.IntermediateRepresentation.OperandHelper;
namespace ARMeilleure.Instructions
{
static partial class InstEmit
{
[Flags]
private enum AccessType
{
None = 0,
Ordered = 1,
Exclusive = 2,
OrderedEx = Ordered | Exclusive
}
public static void Clrex(ArmEmitterContext context)
{
context.Call(new _Void(NativeInterface.ClearExclusive));
}
public static void Dmb(ArmEmitterContext context) => EmitBarrier(context);
public static void Dsb(ArmEmitterContext context) => EmitBarrier(context);
public static void Ldar(ArmEmitterContext context) => EmitLdr(context, AccessType.Ordered);
public static void Ldaxr(ArmEmitterContext context) => EmitLdr(context, AccessType.OrderedEx);
public static void Ldxr(ArmEmitterContext context) => EmitLdr(context, AccessType.Exclusive);
public static void Ldxp(ArmEmitterContext context) => EmitLdp(context, AccessType.Exclusive);
public static void Ldaxp(ArmEmitterContext context) => EmitLdp(context, AccessType.OrderedEx);
private static void EmitLdr(ArmEmitterContext context, AccessType accType)
{
EmitLoadEx(context, accType, pair: false);
}
private static void EmitLdp(ArmEmitterContext context, AccessType accType)
{
EmitLoadEx(context, accType, pair: true);
}
private static void EmitLoadEx(ArmEmitterContext context, AccessType accType, bool pair)
{
OpCodeMemEx op = (OpCodeMemEx)context.CurrOp;
bool ordered = (accType & AccessType.Ordered) != 0;
bool exclusive = (accType & AccessType.Exclusive) != 0;
if (ordered)
{
EmitBarrier(context);
}
Operand address = context.Copy(GetIntOrSP(context, op.Rn));
if (pair)
{
// Exclusive loads should be atomic. For pairwise loads, we need to
// read all the data at once. For a 32-bits pairwise load, we do a
// simple 64-bits load, for a 128-bits load, we need to call a special
// method to read 128-bits atomically.
if (op.Size == 2)
{
Operand value = EmitLoad(context, address, exclusive, 3);
Operand valueLow = context.ConvertI64ToI32(value);
valueLow = context.ZeroExtend32(OperandType.I64, valueLow);
Operand valueHigh = context.ShiftRightUI(value, Const(32));
SetIntOrZR(context, op.Rt, valueLow);
SetIntOrZR(context, op.Rt2, valueHigh);
}
else if (op.Size == 3)
{
Operand value = EmitLoad(context, address, exclusive, 4);
Operand valueLow = context.VectorExtract(OperandType.I64, value, 0);
Operand valueHigh = context.VectorExtract(OperandType.I64, value, 1);
SetIntOrZR(context, op.Rt, valueLow);
SetIntOrZR(context, op.Rt2, valueHigh);
}
else
{
throw new InvalidOperationException($"Invalid load size of {1 << op.Size} bytes.");
}
}
else
{
// 8, 16, 32 or 64-bits (non-pairwise) load.
Operand value = EmitLoad(context, address, exclusive, op.Size);
SetIntOrZR(context, op.Rt, value);
}
}
private static Operand EmitLoad(
ArmEmitterContext context,
Operand address,
bool exclusive,
int size)
{
Delegate fallbackMethodDlg = null;
if (exclusive)
{
switch (size)
{
case 0: fallbackMethodDlg = new _U8_U64 (NativeInterface.ReadByteExclusive); break;
case 1: fallbackMethodDlg = new _U16_U64 (NativeInterface.ReadUInt16Exclusive); break;
case 2: fallbackMethodDlg = new _U32_U64 (NativeInterface.ReadUInt32Exclusive); break;
case 3: fallbackMethodDlg = new _U64_U64 (NativeInterface.ReadUInt64Exclusive); break;
case 4: fallbackMethodDlg = new _V128_U64(NativeInterface.ReadVector128Exclusive); break;
}
}
else
{
switch (size)
{
case 0: fallbackMethodDlg = new _U8_U64 (NativeInterface.ReadByte); break;
case 1: fallbackMethodDlg = new _U16_U64 (NativeInterface.ReadUInt16); break;
case 2: fallbackMethodDlg = new _U32_U64 (NativeInterface.ReadUInt32); break;
case 3: fallbackMethodDlg = new _U64_U64 (NativeInterface.ReadUInt64); break;
case 4: fallbackMethodDlg = new _V128_U64(NativeInterface.ReadVector128); break;
}
}
return context.Call(fallbackMethodDlg, address);
}
public static void Pfrm(ArmEmitterContext context)
{
// Memory Prefetch, execute as no-op.
}
public static void Stlr(ArmEmitterContext context) => EmitStr(context, AccessType.Ordered);
public static void Stlxr(ArmEmitterContext context) => EmitStr(context, AccessType.OrderedEx);
public static void Stxr(ArmEmitterContext context) => EmitStr(context, AccessType.Exclusive);
public static void Stxp(ArmEmitterContext context) => EmitStp(context, AccessType.Exclusive);
public static void Stlxp(ArmEmitterContext context) => EmitStp(context, AccessType.OrderedEx);
private static void EmitStr(ArmEmitterContext context, AccessType accType)
{
EmitStoreEx(context, accType, pair: false);
}
private static void EmitStp(ArmEmitterContext context, AccessType accType)
{
EmitStoreEx(context, accType, pair: true);
}
private static void EmitStoreEx(ArmEmitterContext context, AccessType accType, bool pair)
{
OpCodeMemEx op = (OpCodeMemEx)context.CurrOp;
bool ordered = (accType & AccessType.Ordered) != 0;
bool exclusive = (accType & AccessType.Exclusive) != 0;
if (ordered)
{
EmitBarrier(context);
}
Operand address = context.Copy(GetIntOrSP(context, op.Rn));
Operand t = GetIntOrZR(context, op.Rt);
Operand s = null;
if (pair)
{
Debug.Assert(op.Size == 2 || op.Size == 3, "Invalid size for pairwise store.");
Operand t2 = GetIntOrZR(context, op.Rt2);
Operand value;
if (op.Size == 2)
{
value = context.BitwiseOr(t, context.ShiftLeft(t2, Const(32)));
}
else /* if (op.Size == 3) */
{
value = context.VectorInsert(context.VectorZero(), t, 0);
value = context.VectorInsert(value, t2, 1);
}
s = EmitStore(context, address, value, exclusive, op.Size + 1);
}
else
{
s = EmitStore(context, address, t, exclusive, op.Size);
}
if (s != null)
{
// This is only needed for exclusive stores. The function returns 0
// when the store is successful, and 1 otherwise.
SetIntOrZR(context, op.Rs, s);
}
}
private static Operand EmitStore(
ArmEmitterContext context,
Operand address,
Operand value,
bool exclusive,
int size)
{
if (size < 3)
{
value = context.ConvertI64ToI32(value);
}
Delegate fallbackMethodDlg = null;
if (exclusive)
{
switch (size)
{
case 0: fallbackMethodDlg = new _S32_U64_U8 (NativeInterface.WriteByteExclusive); break;
case 1: fallbackMethodDlg = new _S32_U64_U16 (NativeInterface.WriteUInt16Exclusive); break;
case 2: fallbackMethodDlg = new _S32_U64_U32 (NativeInterface.WriteUInt32Exclusive); break;
case 3: fallbackMethodDlg = new _S32_U64_U64 (NativeInterface.WriteUInt64Exclusive); break;
case 4: fallbackMethodDlg = new _S32_U64_V128(NativeInterface.WriteVector128Exclusive); break;
}
return context.Call(fallbackMethodDlg, address, value);
}
else
{
switch (size)
{
case 0: fallbackMethodDlg = new _Void_U64_U8 (NativeInterface.WriteByte); break;
case 1: fallbackMethodDlg = new _Void_U64_U16 (NativeInterface.WriteUInt16); break;
case 2: fallbackMethodDlg = new _Void_U64_U32 (NativeInterface.WriteUInt32); break;
case 3: fallbackMethodDlg = new _Void_U64_U64 (NativeInterface.WriteUInt64); break;
case 4: fallbackMethodDlg = new _Void_U64_V128(NativeInterface.WriteVector128); break;
}
context.Call(fallbackMethodDlg, address, value);
return null;
}
}
private static void EmitBarrier(ArmEmitterContext context)
{
// Note: This barrier is most likely not necessary, and probably
// doesn't make any difference since we need to do a ton of stuff
// (software MMU emulation) to read or write anything anyway.
}
}
}