Archived
1
0
Fork 0
forked from Mirror/Ryujinx
This repository has been archived on 2024-10-11. You can view files and clone it, but cannot push or open issues or pull requests.
jinx/ARMeilleure/Instructions/InstEmitFlowHelper.cs
riperiperi d904706fc0
Use a Jump Table for direct and indirect calls/jumps, removing transitions to managed (#975)
* Implement Jump Table for Native Calls

NOTE: this slows down rejit considerably! Not recommended to be used
without codegen optimisation or AOT.

- Does not work on Linux
- A32 needs an additional commit.

* A32 Support

(WIP)

* Actually write Direct Call pointers to the table

That would help.

* Direct Calls: Rather than returning to the translator, attempt to keep within the native stack frame.

A return to the translator can still happen, but only by exceptionally
bubbling up to it.

Also:
- Always translate lowCq as a function. Faster interop with the direct
jumps, and this will be useful in future if we want to do speculative
translation.
- Tail Call Detection: after the decoding stage, detect if we do a tail
call, and avoid translating into it. Detected if a jump is made to an
address outwith the contiguous sequence of blocks surrounding the entry
point. The goal is to reduce code touched by jit and rejit.

* A32 Support

* Use smaller max function size for lowCq, fix exceptional returns

When a return has an unexpected value and there is no code block
following this one, we now return the value rather than continuing.

* CompareAndSwap (buggy)

* Ensure CompareAndSwap does not get optimized away.

* Use CompareAndSwap to make the dynamic table thread safe.

* Tail call for linux, throw on too many arguments.

* Combine CompareAndSwap 128 and 32/64.

They emit different IR instructions since their PreAllocator behaviour
is different, but now they just have one function on EmitterContext.

* Fix issues separating from optimisations.

* Use a stub to find and execute missing functions.

This allows us to skip doing many runtime comparisons and branches, and reduces the amount of code we need to emit significantly.

For the indirect call table, this stub also does the work of moving in the highCq address to the table when one is found.

* Make Jump Tables and Jit Cache dynmically resize

Reserve virtual memory, commit as needed.

* Move TailCallRemover to its own class.

* Multithreaded Translation (based on heuristic)

A poor one, at that. Need to get core count for a better one, which
means a lot of OS specific garbage.

* Better priority management for background threads.

* Bound core limit a bit more

Past a certain point the load is not paralellizable and starts stealing from the main thread. Likely due to GC, memory, heap allocation thread contention. Reduce by one core til optimisations come to improve the situation.

* Fix memory management on linux.

* Temporary solution to some sync problems.

This will make sure threads exit correctly, most of the time. There is a potential race where setting the sync counter to 0 does nothing (counter stays at what it was before, thread could take too long to exit), but we need to find a better way to do this anyways. Synchronization frequency has been tightened as we never enter blockwise segments of code. Essentially this means, check every x functions or loop iterations, before lowcq blocks existed and were worth just as much. Ideally it should be done in a better way, since functions can be anywhere from 1 to 5000 instructions. (maybe based on host timer, or an interrupt flag from a scheduler thread)

* Address feedback minus CompareAndSwap change.

* Use default ReservedRegion granularity.

* Merge CompareAndSwap with its V128 variant.

* We already got the source, no need to do it again.

* Make sure all background translation threads exit.

* Fix CompareAndSwap128

Detection criteria was a bit scuffed.

* Address Comments.
2020-03-12 14:20:55 +11:00

352 lines
13 KiB
C#

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 InstEmitFlowHelper
{
public const ulong CallFlag = 1;
public static void EmitCondBranch(ArmEmitterContext context, Operand target, Condition cond)
{
if (cond != Condition.Al)
{
context.BranchIfTrue(target, GetCondTrue(context, cond));
}
else
{
context.Branch(target);
}
}
public static Operand GetCondTrue(ArmEmitterContext context, Condition condition)
{
Operand cmpResult = context.TryGetComparisonResult(condition);
if (cmpResult != null)
{
return cmpResult;
}
Operand value = Const(1);
Operand Inverse(Operand val)
{
return context.BitwiseExclusiveOr(val, Const(1));
}
switch (condition)
{
case Condition.Eq:
value = GetFlag(PState.ZFlag);
break;
case Condition.Ne:
value = Inverse(GetFlag(PState.ZFlag));
break;
case Condition.GeUn:
value = GetFlag(PState.CFlag);
break;
case Condition.LtUn:
value = Inverse(GetFlag(PState.CFlag));
break;
case Condition.Mi:
value = GetFlag(PState.NFlag);
break;
case Condition.Pl:
value = Inverse(GetFlag(PState.NFlag));
break;
case Condition.Vs:
value = GetFlag(PState.VFlag);
break;
case Condition.Vc:
value = Inverse(GetFlag(PState.VFlag));
break;
case Condition.GtUn:
{
Operand c = GetFlag(PState.CFlag);
Operand z = GetFlag(PState.ZFlag);
value = context.BitwiseAnd(c, Inverse(z));
break;
}
case Condition.LeUn:
{
Operand c = GetFlag(PState.CFlag);
Operand z = GetFlag(PState.ZFlag);
value = context.BitwiseOr(Inverse(c), z);
break;
}
case Condition.Ge:
{
Operand n = GetFlag(PState.NFlag);
Operand v = GetFlag(PState.VFlag);
value = context.ICompareEqual(n, v);
break;
}
case Condition.Lt:
{
Operand n = GetFlag(PState.NFlag);
Operand v = GetFlag(PState.VFlag);
value = context.ICompareNotEqual(n, v);
break;
}
case Condition.Gt:
{
Operand n = GetFlag(PState.NFlag);
Operand z = GetFlag(PState.ZFlag);
Operand v = GetFlag(PState.VFlag);
value = context.BitwiseAnd(Inverse(z), context.ICompareEqual(n, v));
break;
}
case Condition.Le:
{
Operand n = GetFlag(PState.NFlag);
Operand z = GetFlag(PState.ZFlag);
Operand v = GetFlag(PState.VFlag);
value = context.BitwiseOr(z, context.ICompareNotEqual(n, v));
break;
}
}
return value;
}
public static void EmitCall(ArmEmitterContext context, ulong immediate)
{
EmitJumpTableBranch(context, Const(immediate));
}
private static void EmitNativeCall(ArmEmitterContext context, Operand nativeContextPtr, Operand funcAddr, bool isJump = false)
{
context.StoreToContext();
Operand returnAddress;
if (isJump)
{
context.Tailcall(funcAddr, nativeContextPtr);
}
else
{
returnAddress = context.Call(funcAddr, OperandType.I64, nativeContextPtr);
context.LoadFromContext();
EmitContinueOrReturnCheck(context, returnAddress);
}
}
private static void EmitNativeCall(ArmEmitterContext context, Operand funcAddr, bool isJump = false)
{
EmitNativeCall(context, context.LoadArgument(OperandType.I64, 0), funcAddr, isJump);
}
public static void EmitVirtualCall(ArmEmitterContext context, Operand target)
{
EmitVirtualCallOrJump(context, target, isJump: false);
}
public static void EmitVirtualJump(ArmEmitterContext context, Operand target, bool isReturn)
{
EmitVirtualCallOrJump(context, target, isJump: true, isReturn: isReturn);
}
private static void EmitVirtualCallOrJump(ArmEmitterContext context, Operand target, bool isJump, bool isReturn = false)
{
if (isReturn)
{
context.Return(target);
}
else
{
EmitJumpTableBranch(context, target, isJump);
}
}
private static void EmitContinueOrReturnCheck(ArmEmitterContext context, Operand returnAddress)
{
// Note: The return value of a translated function is always an Int64 with the
// address execution has returned to. We expect this address to be immediately after the
// current instruction, if it isn't we keep returning until we reach the dispatcher.
Operand nextAddr = Const(GetNextOpAddress(context.CurrOp));
// Try to continue within this block.
// If the return address isn't to our next instruction, we need to return so the JIT can figure out what to do.
Operand lblContinue = Label();
// We need to clear out the call flag for the return address before comparing it.
context.BranchIfTrue(lblContinue, context.ICompareEqual(context.BitwiseAnd(returnAddress, Const(~CallFlag)), nextAddr));
context.Return(returnAddress);
context.MarkLabel(lblContinue);
if (context.CurrBlock.Next == null)
{
// No code following this instruction, try and find the next block and jump to it.
EmitTailContinue(context, nextAddr);
}
}
private static ulong GetNextOpAddress(OpCode op)
{
return op.Address + (ulong)op.OpCodeSizeInBytes;
}
public static void EmitTailContinue(ArmEmitterContext context, Operand address, bool allowRejit = false)
{
bool useTailContinue = true; // Left option here as it may be useful if we need to return to managed rather than tail call in future. (eg. for debug)
if (useTailContinue)
{
if (allowRejit)
{
address = context.BitwiseOr(address, Const(1L));
}
Operand fallbackAddr = context.Call(new _U64_U64(NativeInterface.GetFunctionAddress), address);
EmitNativeCall(context, fallbackAddr, true);
}
else
{
context.Return(address);
}
}
private static void EmitNativeCallWithGuestAddress(ArmEmitterContext context, Operand funcAddr, Operand guestAddress, bool isJump)
{
Operand nativeContextPtr = context.LoadArgument(OperandType.I64, 0);
context.Store(context.Add(nativeContextPtr, Const(NativeContext.GetCallAddressOffset())), guestAddress);
EmitNativeCall(context, nativeContextPtr, funcAddr, isJump);
}
private static void EmitBranchFallback(ArmEmitterContext context, Operand address, bool isJump)
{
address = context.BitwiseOr(address, Const(address.Type, (long)CallFlag)); // Set call flag.
Operand fallbackAddr = context.Call(new _U64_U64(NativeInterface.GetFunctionAddress), address);
EmitNativeCall(context, fallbackAddr, isJump);
}
public static void EmitDynamicTableCall(ArmEmitterContext context, Operand tableAddress, Operand address, bool isJump)
{
// Loop over elements of the dynamic table. Unrolled loop.
Operand endLabel = Label();
Operand fallbackLabel = Label();
Action<Operand> emitTableEntry = (Operand entrySkipLabel) =>
{
// Try to take this entry in the table if its guest address equals 0.
Operand gotResult = context.CompareAndSwap(tableAddress, Const(0L), address);
// Is the address ours? (either taken via CompareAndSwap (0), or what was already here)
context.BranchIfFalse(entrySkipLabel, context.BitwiseOr(context.ICompareEqual(gotResult, address), context.ICompareEqual(gotResult, Const(0L))));
// It's ours, so what function is it pointing to?
Operand targetFunctionPtr = context.Add(tableAddress, Const(8L));
Operand targetFunction = context.Load(OperandType.I64, targetFunctionPtr);
// Call the function.
// We pass in the entry address as the guest address, as the entry may need to be updated by the indirect call stub.
EmitNativeCallWithGuestAddress(context, targetFunction, tableAddress, isJump);
context.Branch(endLabel);
};
// Currently this uses a size of 1, as higher values inflate code size for no real benefit.
for (int i = 0; i < JumpTable.DynamicTableElems; i++)
{
if (i == JumpTable.DynamicTableElems - 1)
{
emitTableEntry(fallbackLabel); // If this is the last entry, avoid emitting the additional label and add.
}
else
{
Operand nextLabel = Label();
emitTableEntry(nextLabel);
context.MarkLabel(nextLabel);
tableAddress = context.Add(tableAddress, Const((long)JumpTable.JumpTableStride)); // Move to the next table entry.
}
}
context.MarkLabel(fallbackLabel);
EmitBranchFallback(context, address, isJump);
context.MarkLabel(endLabel);
}
public static void EmitJumpTableBranch(ArmEmitterContext context, Operand address, bool isJump = false)
{
if (address.Type == OperandType.I32)
{
address = context.ZeroExtend32(OperandType.I64, address);
}
// TODO: Constant folding. Indirect calls are slower in the best case and emit more code so we want to avoid them when possible.
bool isConst = address.Kind == OperandKind.Constant;
long constAddr = (long)address.Value;
if (!context.HighCq)
{
// Don't emit indirect calls or jumps if we're compiling in lowCq mode.
// This avoids wasting space on the jump and indirect tables.
// Just ask the translator for the function address.
EmitBranchFallback(context, address, isJump);
}
else if (!isConst)
{
// Virtual branch/call - store first used addresses on a small table for fast lookup.
int entry = context.JumpTable.ReserveDynamicEntry(isJump);
int jumpOffset = entry * JumpTable.JumpTableStride * JumpTable.DynamicTableElems;
Operand dynTablePtr = Const(context.JumpTable.DynamicPointer.ToInt64() + jumpOffset);
EmitDynamicTableCall(context, dynTablePtr, address, isJump);
}
else
{
int entry = context.JumpTable.ReserveTableEntry(context.BaseAddress & (~3L), constAddr, isJump);
int jumpOffset = entry * JumpTable.JumpTableStride + 8; // Offset directly to the host address.
// TODO: Relocatable jump table ptr for AOT. Would prefer a solution to patch this constant into functions as they are loaded rather than calculate at runtime.
Operand tableEntryPtr = Const(context.JumpTable.JumpPointer.ToInt64() + jumpOffset);
Operand funcAddr = context.Load(OperandType.I64, tableEntryPtr);
EmitNativeCallWithGuestAddress(context, funcAddr, address, isJump); // Call the function directly. If it's not present yet, this will call the direct call stub.
}
}
}
}