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

1 files changed, 390 insertions, 0 deletions

diff --git a/ARMeilleure/CodeGen/RegisterAllocators/LiveInterval.cs b/ARMeilleure/CodeGen/RegisterAllocators/LiveInterval.cs
new file mode 100644
index 00000000..18858a76
--- /dev/null
+++ b/ARMeilleure/CodeGen/RegisterAllocators/LiveInterval.cs
@@ -0,0 +1,390 @@
+using ARMeilleure.IntermediateRepresentation;
+using System;
+using System.Collections.Generic;
+using System.Diagnostics;
+using System.Linq;
+
+namespace ARMeilleure.CodeGen.RegisterAllocators
+{
+    class LiveInterval : IComparable<LiveInterval>
+    {
+        public const int NotFound = -1;
+
+        private LiveInterval _parent;
+
+        private SortedSet<int> _usePositions;
+
+        public int UsesCount => _usePositions.Count;
+
+        private List<LiveRange> _ranges;
+
+        private SortedList<int, LiveInterval> _childs;
+
+        public bool IsSplit => _childs.Count != 0;
+
+        public Operand Local { get; }
+
+        public Register Register { get; set; }
+
+        public int SpillOffset { get; private set; }
+
+        public bool IsSpilled => SpillOffset != -1;
+        public bool IsFixed { get; }
+
+        public bool IsEmpty => _ranges.Count == 0;
+
+        public LiveInterval(Operand local = null, LiveInterval parent = null)
+        {
+            Local   = local;
+            _parent = parent ?? this;
+
+            _usePositions = new SortedSet<int>();
+
+            _ranges = new List<LiveRange>();
+
+            _childs = new SortedList<int, LiveInterval>();
+
+            SpillOffset = -1;
+        }
+
+        public LiveInterval(Register register) : this()
+        {
+            IsFixed  = true;
+            Register = register;
+        }
+
+        public void SetStart(int position)
+        {
+            if (_ranges.Count != 0)
+            {
+                Debug.Assert(position != _ranges[0].End);
+
+                _ranges[0] = new LiveRange(position, _ranges[0].End);
+            }
+            else
+            {
+                _ranges.Add(new LiveRange(position, position + 1));
+            }
+        }
+
+        public int GetStart()
+        {
+            if (_ranges.Count == 0)
+            {
+                throw new InvalidOperationException("Empty interval.");
+            }
+
+            return _ranges[0].Start;
+        }
+
+        public void SetEnd(int position)
+        {
+            if (_ranges.Count != 0)
+            {
+                int lastIdx = _ranges.Count - 1;
+
+                Debug.Assert(position != _ranges[lastIdx].Start);
+
+                _ranges[lastIdx] = new LiveRange(_ranges[lastIdx].Start, position);
+            }
+            else
+            {
+                _ranges.Add(new LiveRange(position, position + 1));
+            }
+        }
+
+        public int GetEnd()
+        {
+            if (_ranges.Count == 0)
+            {
+                throw new InvalidOperationException("Empty interval.");
+            }
+
+            return _ranges[_ranges.Count - 1].End;
+        }
+
+        public void AddRange(int start, int end)
+        {
+            if (start >= end)
+            {
+                throw new ArgumentException("Invalid range start position " + start + ", " + end);
+            }
+
+            int index = _ranges.BinarySearch(new LiveRange(start, end));
+
+            if (index >= 0)
+            {
+                // New range insersects with an existing range, we need to remove
+                // all the intersecting ranges before adding the new one.
+                // We also extend the new range as needed, based on the values of
+                // the existing ranges being removed.
+                int lIndex = index;
+                int rIndex = index;
+
+                while (lIndex > 0 && _ranges[lIndex - 1].End >= start)
+                {
+                    lIndex--;
+                }
+
+                while (rIndex + 1 < _ranges.Count && _ranges[rIndex + 1].Start <= end)
+                {
+                    rIndex++;
+                }
+
+                if (start > _ranges[lIndex].Start)
+                {
+                    start = _ranges[lIndex].Start;
+                }
+
+                if (end < _ranges[rIndex].End)
+                {
+                    end = _ranges[rIndex].End;
+                }
+
+                _ranges.RemoveRange(lIndex, (rIndex - lIndex) + 1);
+
+                InsertRange(lIndex, start, end);
+            }
+            else
+            {
+                InsertRange(~index, start, end);
+            }
+        }
+
+        private void InsertRange(int index, int start, int end)
+        {
+            // Here we insert a new range on the ranges list.
+            // If possible, we extend an existing range rather than inserting a new one.
+            // We can extend an existing range if any of the following conditions are true:
+            // - The new range starts right after the end of the previous range on the list.
+            // - The new range ends right before the start of the next range on the list.
+            // If both cases are true, we can extend either one. We prefer to extend the
+            // previous range, and then remove the next one, but theres no specific reason
+            // for that, extending either one will do.
+            int? extIndex = null;
+
+            if (index > 0 && _ranges[index - 1].End == start)
+            {
+                start = _ranges[index - 1].Start;
+
+                extIndex = index - 1;
+            }
+
+            if (index < _ranges.Count && _ranges[index].Start == end)
+            {
+                end = _ranges[index].End;
+
+                if (extIndex.HasValue)
+                {
+                    _ranges.RemoveAt(index);
+                }
+                else
+                {
+                    extIndex = index;
+                }
+            }
+
+            if (extIndex.HasValue)
+            {
+                _ranges[extIndex.Value] = new LiveRange(start, end);
+            }
+            else
+            {
+                _ranges.Insert(index, new LiveRange(start, end));
+            }
+        }
+
+        public void AddUsePosition(int position)
+        {
+            _usePositions.Add(position);
+        }
+
+        public bool Overlaps(int position)
+        {
+            return _ranges.BinarySearch(new LiveRange(position, position + 1)) >= 0;
+        }
+
+        public bool Overlaps(LiveInterval other)
+        {
+            foreach (LiveRange range in other._ranges)
+            {
+                if (_ranges.BinarySearch(range) >= 0)
+                {
+                    return true;
+                }
+            }
+
+            return false;
+        }
+
+        public int GetOverlapPosition(LiveInterval other)
+        {
+            foreach (LiveRange range in other._ranges)
+            {
+                int overlapIndex = _ranges.BinarySearch(range);
+
+                if (overlapIndex >= 0)
+                {
+                    // It's possible that we have multiple overlaps within a single interval,
+                    // in this case, we pick the one with the lowest start position, since
+                    // we return the first overlap position.
+                    while (overlapIndex > 0 && _ranges[overlapIndex - 1].End > range.Start)
+                    {
+                        overlapIndex--;
+                    }
+
+                    LiveRange overlappingRange = _ranges[overlapIndex];
+
+                    return overlappingRange.Start;
+                }
+            }
+
+            return NotFound;
+        }
+
+        public IEnumerable<LiveInterval> SplitChilds()
+        {
+            return _childs.Values;
+        }
+
+        public IEnumerable<int> UsePositions()
+        {
+            return _usePositions;
+        }
+
+        public int FirstUse()
+        {
+            if (_usePositions.Count == 0)
+            {
+                return NotFound;
+            }
+
+            return _usePositions.First();
+        }
+
+        public int NextUseAfter(int position)
+        {
+            foreach (int usePosition in _usePositions)
+            {
+                if (usePosition >= position)
+                {
+                    return usePosition;
+                }
+            }
+
+            return NotFound;
+        }
+
+        public LiveInterval Split(int position)
+        {
+            LiveInterval right = new LiveInterval(Local, _parent);
+
+            int splitIndex = 0;
+
+            for (; splitIndex < _ranges.Count; splitIndex++)
+            {
+                LiveRange range = _ranges[splitIndex];
+
+                if (position > range.Start && position <= range.End)
+                {
+                    right._ranges.Add(new LiveRange(position, range.End));
+
+                    range = new LiveRange(range.Start, position);
+
+                    _ranges[splitIndex++] = range;
+
+                    break;
+                }
+
+                if (range.Start >= position)
+                {
+                    break;
+                }
+            }
+
+            if (splitIndex < _ranges.Count)
+            {
+                int count = _ranges.Count - splitIndex;
+
+                right._ranges.AddRange(_ranges.GetRange(splitIndex, count));
+
+                _ranges.RemoveRange(splitIndex, count);
+            }
+
+            foreach (int usePosition in _usePositions.Where(x => x >= position))
+            {
+                right._usePositions.Add(usePosition);
+            }
+
+            _usePositions.RemoveWhere(x => x >= position);
+
+            Debug.Assert(_ranges.Count != 0, "Left interval is empty after split.");
+
+            Debug.Assert(right._ranges.Count != 0, "Right interval is empty after split.");
+
+            AddSplitChild(right);
+
+            return right;
+        }
+
+        private void AddSplitChild(LiveInterval child)
+        {
+            Debug.Assert(!child.IsEmpty, "Trying to insert a empty interval.");
+
+            _parent._childs.Add(child.GetStart(), child);
+        }
+
+        public LiveInterval GetSplitChild(int position)
+        {
+            if (Overlaps(position))
+            {
+                return this;
+            }
+
+            foreach (LiveInterval splitChild in _childs.Values)
+            {
+                if (splitChild.Overlaps(position))
+                {
+                    return splitChild;
+                }
+            }
+
+            return null;
+        }
+
+        public bool TrySpillWithSiblingOffset()
+        {
+            foreach (LiveInterval splitChild in _parent._childs.Values)
+            {
+                if (splitChild.IsSpilled)
+                {
+                    Spill(splitChild.SpillOffset);
+
+                    return true;
+                }
+            }
+
+            return false;
+        }
+
+        public void Spill(int offset)
+        {
+            SpillOffset = offset;
+        }
+
+        public int CompareTo(LiveInterval other)
+        {
+            if (_ranges.Count == 0 || other._ranges.Count == 0)
+            {
+                return _ranges.Count.CompareTo(other._ranges.Count);
+            }
+
+            return _ranges[0].Start.CompareTo(other._ranges[0].Start);
+        }
+
+        public override string ToString()
+        {
+            return string.Join("; ", _ranges);
+        }
+    }
+}
\ No newline at end of file