Merge branch 'master' into patch-2

ChocolArm64/AOpCodeTable.cs

@@ -374,16 +374,22 @@ namespace ChocolArm64
             SetA64("0x001110<<1xxxxx101011xxxxxxxxxx", AInstEmit.Sminp_V,       typeof(AOpCodeSimdReg));
             SetA64("0x001110<<1xxxxx100000xxxxxxxxxx", AInstEmit.Smlal_V,       typeof(AOpCodeSimdReg));
             SetA64("0x001110000xxxxx001011xxxxxxxxxx", AInstEmit.Smov_S,        typeof(AOpCodeSimdIns));
+            SetA64("0x001110<<1xxxxx101000xxxxxxxxxx", AInstEmit.Smlsl_V,       typeof(AOpCodeSimdReg));
             SetA64("0x001110<<1xxxxx110000xxxxxxxxxx", AInstEmit.Smull_V,       typeof(AOpCodeSimdReg));
+            SetA64("0x00111100>>>xxx100111xxxxxxxxxx", AInstEmit.Sqrshrn_V,     typeof(AOpCodeSimdShImm));
             SetA64("01011110<<100001010010xxxxxxxxxx", AInstEmit.Sqxtn_S,       typeof(AOpCodeSimd));
             SetA64("0x001110<<100001010010xxxxxxxxxx", AInstEmit.Sqxtn_V,       typeof(AOpCodeSimd));
             SetA64("01111110<<100001001010xxxxxxxxxx", AInstEmit.Sqxtun_S,      typeof(AOpCodeSimd));
             SetA64("0x101110<<100001001010xxxxxxxxxx", AInstEmit.Sqxtun_V,      typeof(AOpCodeSimd));
+            SetA64("0x00111100>>>xxx001001xxxxxxxxxx", AInstEmit.Srshr_V,       typeof(AOpCodeSimdShImm));
+            SetA64("0100111101xxxxxx001001xxxxxxxxxx", AInstEmit.Srshr_V,       typeof(AOpCodeSimdShImm));
             SetA64("0>001110<<1xxxxx010001xxxxxxxxxx", AInstEmit.Sshl_V,        typeof(AOpCodeSimdReg));
             SetA64("0x00111100>>>xxx101001xxxxxxxxxx", AInstEmit.Sshll_V,       typeof(AOpCodeSimdShImm));
-            SetA64("010111110>>>>xxx000001xxxxxxxxxx", AInstEmit.Sshr_S,        typeof(AOpCodeSimdShImm));
-            SetA64("0x0011110>>>>xxx000001xxxxxxxxxx", AInstEmit.Sshr_V,        typeof(AOpCodeSimdShImm));
-            SetA64("0x0011110>>>>xxx000101xxxxxxxxxx", AInstEmit.Ssra_V,        typeof(AOpCodeSimdShImm));
+            SetA64("0101111101xxxxxx000001xxxxxxxxxx", AInstEmit.Sshr_S,        typeof(AOpCodeSimdShImm));
+            SetA64("0x00111100>>>xxx000001xxxxxxxxxx", AInstEmit.Sshr_V,        typeof(AOpCodeSimdShImm));
+            SetA64("0100111101xxxxxx000001xxxxxxxxxx", AInstEmit.Sshr_V,        typeof(AOpCodeSimdShImm));
+            SetA64("0x00111100>>>xxx000101xxxxxxxxxx", AInstEmit.Ssra_V,        typeof(AOpCodeSimdShImm));
+            SetA64("0100111101xxxxxx000101xxxxxxxxxx", AInstEmit.Ssra_V,        typeof(AOpCodeSimdShImm));
             SetA64("0x00110000000000xxxxxxxxxxxxxxxx", AInstEmit.St__Vms,       typeof(AOpCodeSimdMemMs));
             SetA64("0x001100100xxxxxxxxxxxxxxxxxxxxx", AInstEmit.St__Vms,       typeof(AOpCodeSimdMemMs));
             SetA64("0x00110100x00000xxxxxxxxxxxxxxxx", AInstEmit.St__Vss,       typeof(AOpCodeSimdMemSs));
@@ -422,9 +428,11 @@ namespace ChocolArm64
             SetA64("0x101110<<100001010010xxxxxxxxxx", AInstEmit.Uqxtn_V,       typeof(AOpCodeSimd));
             SetA64("0>101110<<1xxxxx010001xxxxxxxxxx", AInstEmit.Ushl_V,        typeof(AOpCodeSimdReg));
             SetA64("0x10111100>>>xxx101001xxxxxxxxxx", AInstEmit.Ushll_V,       typeof(AOpCodeSimdShImm));
-            SetA64("011111110>>>>xxx000001xxxxxxxxxx", AInstEmit.Ushr_S,        typeof(AOpCodeSimdShImm));
-            SetA64("0x1011110>>>>xxx000001xxxxxxxxxx", AInstEmit.Ushr_V,        typeof(AOpCodeSimdShImm));
-            SetA64("0x1011110>>>>xxx000101xxxxxxxxxx", AInstEmit.Usra_V,        typeof(AOpCodeSimdShImm));
+            SetA64("0111111101xxxxxx000001xxxxxxxxxx", AInstEmit.Ushr_S,        typeof(AOpCodeSimdShImm));
+            SetA64("0x10111100>>>xxx000001xxxxxxxxxx", AInstEmit.Ushr_V,        typeof(AOpCodeSimdShImm));
+            SetA64("0110111101xxxxxx000001xxxxxxxxxx", AInstEmit.Ushr_V,        typeof(AOpCodeSimdShImm));
+            SetA64("0x10111100>>>xxx000101xxxxxxxxxx", AInstEmit.Usra_V,        typeof(AOpCodeSimdShImm));
+            SetA64("0110111101xxxxxx000101xxxxxxxxxx", AInstEmit.Usra_V,        typeof(AOpCodeSimdShImm));
             SetA64("0>001110<<0xxxxx000110xxxxxxxxxx", AInstEmit.Uzp1_V,        typeof(AOpCodeSimdReg));
             SetA64("0>001110<<0xxxxx010110xxxxxxxxxx", AInstEmit.Uzp2_V,        typeof(AOpCodeSimdReg));
             SetA64("0x001110<<100001001010xxxxxxxxxx", AInstEmit.Xtn_V,         typeof(AOpCodeSimd));

ChocolArm64/Instruction/AInstEmitSimdArithmetic.cs

@@ -65,11 +65,12 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimd Op = (AOpCodeSimd)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
+            int Elems = Bytes >> Op.Size;
 
             EmitVectorExtractZx(Context, Op.Rn, 0, Op.Size);
 
-            for (int Index = 1; Index < (Bytes >> Op.Size); Index++)
+            for (int Index = 1; Index < Elems; Index++)
             {
                 EmitVectorExtractZx(Context, Op.Rn, Index, Op.Size);
 
@@ -97,13 +98,16 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimd Op = (AOpCodeSimd)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
+            int Elems = Bytes >> Op.Size;
 
-            for (int Index = 0; Index < (Bytes >> Op.Size); Index++)
+            int ESize = 8 << Op.Size;
+
+            for (int Index = 0; Index < Elems; Index++)
             {
                 EmitVectorExtractZx(Context, Op.Rn, Index, Op.Size);
 
-                Context.EmitLdc_I4(8 << Op.Size);
+                Context.EmitLdc_I4(ESize);
 
                 Emit();
 
@@ -159,12 +163,19 @@ namespace ChocolArm64.Instruction
             AOpCodeSimdReg Op = (AOpCodeSimdReg)Context.CurrOp;
 
             int Elems = 8 >> Op.Size;
+
             int ESize = 8 << Op.Size;
 
             int Part = Op.RegisterSize == ARegisterSize.SIMD128 ? Elems : 0;
 
             long RoundConst = 1L << (ESize - 1);
 
+            if (Part != 0)
+            {
+                Context.EmitLdvec(Op.Rd);
+                Context.EmitStvectmp();
+            }
+
             for (int Index = 0; Index < Elems; Index++)
             {
                 EmitVectorExtractZx(Context, Op.Rn, Index, Op.Size + 1);
@@ -181,93 +192,18 @@ namespace ChocolArm64.Instruction
 
                 Context.EmitLsr(ESize);
 
-                EmitVectorInsert(Context, Op.Rd, Part + Index, Op.Size);
+                EmitVectorInsertTmp(Context, Part + Index, Op.Size);
             }
 
+            Context.EmitLdvectmp();
+            Context.EmitStvec(Op.Rd);
+
             if (Part == 0)
             {
                 EmitVectorZeroUpper(Context, Op.Rd);
             }
         }
 
-        private static void EmitSaturatingExtNarrow(AILEmitterCtx Context, bool SignedSrc, bool SignedDst, bool Scalar)
-        {
-            AOpCodeSimd Op = (AOpCodeSimd)Context.CurrOp;
-
-            int Elems = (!Scalar ? 8 >> Op.Size : 1);
-            int ESize = 8 << Op.Size;
-
-            int Part = (!Scalar & (Op.RegisterSize == ARegisterSize.SIMD128) ? Elems : 0);
-
-            int TMaxValue = (SignedDst ? (1 << (ESize - 1)) - 1 : (int)((1L << ESize) - 1L));
-            int TMinValue = (SignedDst ? -((1 << (ESize - 1))) : 0);
-
-            Context.EmitLdc_I8(0L);
-            Context.EmitSttmp();
-
-            for (int Index = 0; Index < Elems; Index++)
-            {
-                AILLabel LblLe    = new AILLabel();
-                AILLabel LblGeEnd = new AILLabel();
-
-                EmitVectorExtract(Context, Op.Rn, Index, Op.Size + 1, SignedSrc);
-
-                Context.Emit(OpCodes.Dup);
-
-                Context.EmitLdc_I4(TMaxValue);
-                Context.Emit(OpCodes.Conv_U8);
-
-                Context.Emit(SignedSrc ? OpCodes.Ble_S : OpCodes.Ble_Un_S, LblLe);
-
-                Context.Emit(OpCodes.Pop);
-
-                Context.EmitLdc_I4(TMaxValue);
-
-                Context.EmitLdc_I8(0x8000000L);
-                Context.EmitSttmp();
-
-                Context.Emit(OpCodes.Br_S, LblGeEnd);
-
-                Context.MarkLabel(LblLe);
-
-                Context.Emit(OpCodes.Dup);
-
-                Context.EmitLdc_I4(TMinValue);
-                Context.Emit(OpCodes.Conv_I8);
-
-                Context.Emit(SignedSrc ? OpCodes.Bge_S : OpCodes.Bge_Un_S, LblGeEnd);
-
-                Context.Emit(OpCodes.Pop);
-
-                Context.EmitLdc_I4(TMinValue);
-
-                Context.EmitLdc_I8(0x8000000L);
-                Context.EmitSttmp();
-
-                Context.MarkLabel(LblGeEnd);
-
-                if (Scalar)
-                {
-                    EmitVectorZeroLower(Context, Op.Rd);
-                }
-
-                EmitVectorInsert(Context, Op.Rd, Part + Index, Op.Size);
-            }
-
-            if (Part == 0)
-            {
-                EmitVectorZeroUpper(Context, Op.Rd);
-            }
-
-            Context.EmitLdarg(ATranslatedSub.StateArgIdx);
-            Context.EmitLdarg(ATranslatedSub.StateArgIdx);
-            Context.EmitCallPropGet(typeof(AThreadState), nameof(AThreadState.Fpsr));
-            Context.EmitLdtmp();
-            Context.Emit(OpCodes.Conv_I4);
-            Context.Emit(OpCodes.Or);
-            Context.EmitCallPropSet(typeof(AThreadState), nameof(AThreadState.Fpsr));
-        }
-
         public static void Fabd_S(AILEmitterCtx Context)
         {
             EmitScalarBinaryOpF(Context, () =>
@@ -338,7 +274,7 @@ namespace ChocolArm64.Instruction
 
             int SizeF = Op.Size & 1;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
 
             int Elems = Bytes >> SizeF + 2;
             int Half  = Elems >> 1;
@@ -655,106 +591,34 @@ namespace ChocolArm64.Instruction
 
         public static void Frecpe_S(AILEmitterCtx Context)
         {
-            EmitFrecpe(Context, 0, Scalar: true);
+            EmitScalarUnaryOpF(Context, () =>
+            {
+                EmitUnarySoftFloatCall(Context, nameof(ASoftFloat.RecipEstimate));
+            });
         }
 
         public static void Frecpe_V(AILEmitterCtx Context)
         {
-            AOpCodeSimd Op = (AOpCodeSimd)Context.CurrOp;
-
-            int SizeF = Op.Size & 1;
-
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
-
-            for (int Index = 0; Index < Bytes >> SizeF + 2; Index++)
+            EmitVectorUnaryOpF(Context, () =>
             {
-                EmitFrecpe(Context, Index, Scalar: false);
-            }
-
-            if (Op.RegisterSize == ARegisterSize.SIMD64)
-            {
-                EmitVectorZeroUpper(Context, Op.Rd);
-            }
-        }
-
-        private static void EmitFrecpe(AILEmitterCtx Context, int Index, bool Scalar)
-        {
-            AOpCodeSimd Op = (AOpCodeSimd)Context.CurrOp;
-
-            int SizeF = Op.Size & 1;
-
-            if (SizeF == 0)
-            {
-                Context.EmitLdc_R4(1);
-            }
-            else /* if (SizeF == 1) */
-            {
-                Context.EmitLdc_R8(1);
-            }
-
-            EmitVectorExtractF(Context, Op.Rn, Index, SizeF);
-
-            Context.Emit(OpCodes.Div);
-
-            if (Scalar)
-            {
-                EmitVectorZeroAll(Context, Op.Rd);
-            }
-
-            EmitVectorInsertF(Context, Op.Rd, Index, SizeF);
+                EmitUnarySoftFloatCall(Context, nameof(ASoftFloat.RecipEstimate));
+            });
         }
 
         public static void Frecps_S(AILEmitterCtx Context)
         {
-            EmitFrecps(Context, 0, Scalar: true);
+            EmitScalarBinaryOpF(Context, () =>
+            {
+                EmitBinarySoftFloatCall(Context, nameof(ASoftFloat.RecipStep));
+            });
         }
 
         public static void Frecps_V(AILEmitterCtx Context)
         {
-            AOpCodeSimd Op = (AOpCodeSimd)Context.CurrOp;
-
-            int SizeF = Op.Size & 1;
-
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
-
-            for (int Index = 0; Index < Bytes >> SizeF + 2; Index++)
+            EmitVectorBinaryOpF(Context, () =>
             {
-                EmitFrecps(Context, Index, Scalar: false);
-            }
-
-            if (Op.RegisterSize == ARegisterSize.SIMD64)
-            {
-                EmitVectorZeroUpper(Context, Op.Rd);
-            }
-        }
-
-        private static void EmitFrecps(AILEmitterCtx Context, int Index, bool Scalar)
-        {
-            AOpCodeSimdReg Op = (AOpCodeSimdReg)Context.CurrOp;
-
-            int SizeF = Op.Size & 1;
-
-            if (SizeF == 0)
-            {
-                Context.EmitLdc_R4(2);
-            }
-            else /* if (SizeF == 1) */
-            {
-                Context.EmitLdc_R8(2);
-            }
-
-            EmitVectorExtractF(Context, Op.Rn, Index, SizeF);
-            EmitVectorExtractF(Context, Op.Rm, Index, SizeF);
-
-            Context.Emit(OpCodes.Mul);
-            Context.Emit(OpCodes.Sub);
-
-            if (Scalar)
-            {
-                EmitVectorZeroAll(Context, Op.Rd);
-            }
-
-            EmitVectorInsertF(Context, Op.Rd, Index, SizeF);
+                EmitBinarySoftFloatCall(Context, nameof(ASoftFloat.RecipStep));
+            });
         }
 
         public static void Frinta_S(AILEmitterCtx Context)
@@ -956,7 +820,7 @@ namespace ChocolArm64.Instruction
 
             int SizeF = Op.Size & 1;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
 
             for (int Index = 0; Index < Bytes >> SizeF + 2; Index++)
             {
@@ -1188,6 +1052,15 @@ namespace ChocolArm64.Instruction
             });
         }
 
+        public static void Smlsl_V(AILEmitterCtx Context)
+        {
+            EmitVectorWidenRnRmTernaryOpSx(Context, () =>
+            {
+                Context.Emit(OpCodes.Mul);
+                Context.Emit(OpCodes.Sub);
+            });
+        }
+
         public static void Smull_V(AILEmitterCtx Context)
         {
             EmitVectorWidenRnRmBinaryOpSx(Context, () => Context.Emit(OpCodes.Mul));
@@ -1195,22 +1068,22 @@ namespace ChocolArm64.Instruction
 
         public static void Sqxtn_S(AILEmitterCtx Context)
         {
-            EmitSaturatingExtNarrow(Context, SignedSrc: true, SignedDst: true, Scalar: true);
+            EmitScalarSaturatingNarrowOpSxSx(Context, () => { });
         }
 
         public static void Sqxtn_V(AILEmitterCtx Context)
         {
-            EmitSaturatingExtNarrow(Context, SignedSrc: true, SignedDst: true, Scalar: false);
+            EmitVectorSaturatingNarrowOpSxSx(Context, () => { });
         }
 
         public static void Sqxtun_S(AILEmitterCtx Context)
         {
-            EmitSaturatingExtNarrow(Context, SignedSrc: true, SignedDst: false, Scalar: true);
+            EmitScalarSaturatingNarrowOpSxZx(Context, () => { });
         }
 
         public static void Sqxtun_V(AILEmitterCtx Context)
         {
-            EmitSaturatingExtNarrow(Context, SignedSrc: true, SignedDst: false, Scalar: false);
+            EmitVectorSaturatingNarrowOpSxZx(Context, () => { });
         }
 
         public static void Sub_S(AILEmitterCtx Context)
@@ -1284,11 +1157,12 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimd Op = (AOpCodeSimd)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
+            int Elems = Bytes >> Op.Size;
 
             EmitVectorExtractZx(Context, Op.Rn, 0, Op.Size);
 
-            for (int Index = 1; Index < (Bytes >> Op.Size); Index++)
+            for (int Index = 1; Index < Elems; Index++)
             {
                 EmitVectorExtractZx(Context, Op.Rn, Index, Op.Size);
 
@@ -1358,12 +1232,12 @@ namespace ChocolArm64.Instruction
 
         public static void Uqxtn_S(AILEmitterCtx Context)
         {
-            EmitSaturatingExtNarrow(Context, SignedSrc: false, SignedDst: false, Scalar: true);
+            EmitScalarSaturatingNarrowOpZxZx(Context, () => { });
         }
 
         public static void Uqxtn_V(AILEmitterCtx Context)
         {
-            EmitSaturatingExtNarrow(Context, SignedSrc: false, SignedDst: false, Scalar: false);
+            EmitVectorSaturatingNarrowOpZxZx(Context, () => { });
         }
     }
 }

ChocolArm64/Instruction/AInstEmitSimdCmp.cs

@@ -363,7 +363,7 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimd Op = (AOpCodeSimd)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
             int Elems = (!Scalar ? Bytes >> Op.Size : 1);
 
             ulong SzMask = ulong.MaxValue >> (64 - (8 << Op.Size));
@@ -407,7 +407,7 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdReg Op = (AOpCodeSimdReg)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
             int Elems = (!Scalar ? Bytes >> Op.Size : 1);
 
             ulong SzMask = ulong.MaxValue >> (64 - (8 << Op.Size));
@@ -454,7 +454,7 @@ namespace ChocolArm64.Instruction
 
             int SizeF = Op.Size & 1;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
 
             for (int Index = 0; Index < Bytes >> SizeF + 2; Index++)
             {

ChocolArm64/Instruction/AInstEmitSimdCvt.cs

@@ -45,10 +45,10 @@ namespace ChocolArm64.Instruction
             {
                 if (SizeF == 0)
                 {
-                    //TODO: This need the half precision floating point type,
-                    //that is not yet supported on .NET. We should probably
-                    //do our own implementation on the meantime.
-                    throw new NotImplementedException();
+                    EmitVectorExtractZx(Context, Op.Rn, Part + Index, 1);
+                    Context.Emit(OpCodes.Conv_U2);
+
+                    Context.EmitCall(typeof(ASoftFloat), nameof(ASoftFloat.ConvertHalfToSingle));
                 }
                 else /* if (SizeF == 1) */
                 {
@@ -337,7 +337,7 @@ namespace ChocolArm64.Instruction
 
             int FBits = GetFBits(Context);
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
 
             for (int Index = 0; Index < (Bytes >> SizeI); Index++)
             {
@@ -426,7 +426,7 @@ namespace ChocolArm64.Instruction
 
             int FBits = GetFBits(Context);
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
 
             for (int Index = 0; Index < (Bytes >> SizeI); Index++)
             {

ChocolArm64/Instruction/AInstEmitSimdHelper.cs

@@ -3,6 +3,7 @@ using ChocolArm64.State;
 using ChocolArm64.Translation;
 using System;
 using System.Reflection;
+using System.Reflection.Emit;
 using System.Runtime.CompilerServices;
 using System.Runtime.Intrinsics;
 using System.Runtime.Intrinsics.X86;
@@ -253,6 +254,26 @@ namespace ChocolArm64.Instruction
             Context.EmitCall(MthdInfo);
         }
 
+        public static void EmitBinarySoftFloatCall(AILEmitterCtx Context, string Name)
+        {
+            IAOpCodeSimd Op = (IAOpCodeSimd)Context.CurrOp;
+
+            int SizeF = Op.Size & 1;
+
+            MethodInfo MthdInfo;
+
+            if (SizeF == 0)
+            {
+                MthdInfo = typeof(ASoftFloat).GetMethod(Name, new Type[] { typeof(float), typeof(float) });
+            }
+            else /* if (SizeF == 1) */
+            {
+                MthdInfo = typeof(ASoftFloat).GetMethod(Name, new Type[] { typeof(double), typeof(double) });
+            }
+
+            Context.EmitCall(MthdInfo);
+        }
+
         public static void EmitScalarBinaryOpByElemF(AILEmitterCtx Context, Action Emit)
         {
             AOpCodeSimdRegElemF Op = (AOpCodeSimdRegElemF)Context.CurrOp;
@@ -397,7 +418,7 @@ namespace ChocolArm64.Instruction
 
             int SizeF = Op.Size & 1;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
 
             for (int Index = 0; Index < (Bytes >> SizeF + 2); Index++)
             {
@@ -447,7 +468,7 @@ namespace ChocolArm64.Instruction
 
             int SizeF = Op.Size & 1;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
 
             for (int Index = 0; Index < (Bytes >> SizeF + 2); Index++)
             {
@@ -507,9 +528,10 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimd Op = (AOpCodeSimd)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
+            int Elems = Bytes >> Op.Size;
 
-            for (int Index = 0; Index < (Bytes >> Op.Size); Index++)
+            for (int Index = 0; Index < Elems; Index++)
             {
                 if (Opers.HasFlag(OperFlags.Rd))
                 {
@@ -562,9 +584,10 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdReg Op = (AOpCodeSimdReg)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
+            int Elems = Bytes >> Op.Size;
 
-            for (int Index = 0; Index < (Bytes >> Op.Size); Index++)
+            for (int Index = 0; Index < Elems; Index++)
             {
                 if (Ternary)
                 {
@@ -602,9 +625,10 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdImm Op = (AOpCodeSimdImm)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
+            int Elems = Bytes >> Op.Size;
 
-            for (int Index = 0; Index < (Bytes >> Op.Size); Index++)
+            for (int Index = 0; Index < Elems; Index++)
             {
                 if (Binary)
                 {
@@ -719,11 +743,11 @@ namespace ChocolArm64.Instruction
             EmitVectorPairwiseOp(Context, Emit, false);
         }
 
-        private static void EmitVectorPairwiseOp(AILEmitterCtx Context, Action Emit, bool Signed)
+        public static void EmitVectorPairwiseOp(AILEmitterCtx Context, Action Emit, bool Signed)
         {
             AOpCodeSimdReg Op = (AOpCodeSimdReg)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
 
             int Elems = Bytes >> Op.Size;
             int Half  = Elems >> 1;
@@ -749,6 +773,127 @@ namespace ChocolArm64.Instruction
             }
         }
 
+        public static void EmitScalarSaturatingNarrowOpSxSx(AILEmitterCtx Context, Action Emit)
+        {
+            EmitSaturatingNarrowOp(Context, Emit, true, true, true);
+        }
+
+        public static void EmitScalarSaturatingNarrowOpSxZx(AILEmitterCtx Context, Action Emit)
+        {
+            EmitSaturatingNarrowOp(Context, Emit, true, false, true);
+        }
+
+        public static void EmitScalarSaturatingNarrowOpZxZx(AILEmitterCtx Context, Action Emit)
+        {
+            EmitSaturatingNarrowOp(Context, Emit, false, false, true);
+        }
+
+        public static void EmitVectorSaturatingNarrowOpSxSx(AILEmitterCtx Context, Action Emit)
+        {
+            EmitSaturatingNarrowOp(Context, Emit, true, true, false);
+        }
+
+        public static void EmitVectorSaturatingNarrowOpSxZx(AILEmitterCtx Context, Action Emit)
+        {
+            EmitSaturatingNarrowOp(Context, Emit, true, false, false);
+        }
+
+        public static void EmitVectorSaturatingNarrowOpZxZx(AILEmitterCtx Context, Action Emit)
+        {
+            EmitSaturatingNarrowOp(Context, Emit, false, false, false);
+        }
+
+        public static void EmitSaturatingNarrowOp(
+            AILEmitterCtx Context,
+            Action        Emit,
+            bool          SignedSrc,
+            bool          SignedDst,
+            bool          Scalar)
+        {
+            AOpCodeSimd Op = (AOpCodeSimd)Context.CurrOp;
+
+            int Elems = !Scalar ? 8 >> Op.Size : 1;
+
+            int ESize = 8 << Op.Size;
+
+            int Part = !Scalar && (Op.RegisterSize == ARegisterSize.SIMD128) ? Elems : 0;
+
+            long TMaxValue = SignedDst ? (1 << (ESize - 1)) - 1 : (1L << ESize) - 1L;
+            long TMinValue = SignedDst ? -((1 << (ESize - 1))) : 0;
+
+            Context.EmitLdc_I8(0L);
+            Context.EmitSttmp();
+
+            if (Part != 0)
+            {
+                Context.EmitLdvec(Op.Rd);
+                Context.EmitStvectmp();
+            }
+
+            for (int Index = 0; Index < Elems; Index++)
+            {
+                AILLabel LblLe    = new AILLabel();
+                AILLabel LblGeEnd = new AILLabel();
+
+                EmitVectorExtract(Context, Op.Rn, Index, Op.Size + 1, SignedSrc);
+
+                Emit();
+
+                Context.Emit(OpCodes.Dup);
+
+                Context.EmitLdc_I8(TMaxValue);
+
+                Context.Emit(SignedSrc ? OpCodes.Ble_S : OpCodes.Ble_Un_S, LblLe);
+
+                Context.Emit(OpCodes.Pop);
+
+                Context.EmitLdc_I8(TMaxValue);
+                Context.EmitLdc_I8(0x8000000L);
+                Context.EmitSttmp();
+
+                Context.Emit(OpCodes.Br_S, LblGeEnd);
+
+                Context.MarkLabel(LblLe);
+
+                Context.Emit(OpCodes.Dup);
+
+                Context.EmitLdc_I8(TMinValue);
+
+                Context.Emit(SignedSrc ? OpCodes.Bge_S : OpCodes.Bge_Un_S, LblGeEnd);
+
+                Context.Emit(OpCodes.Pop);
+
+                Context.EmitLdc_I8(TMinValue);
+                Context.EmitLdc_I8(0x8000000L);
+                Context.EmitSttmp();
+
+                Context.MarkLabel(LblGeEnd);
+
+                if (Scalar)
+                {
+                    EmitVectorZeroLower(Context, Op.Rd);
+                }
+
+                EmitVectorInsertTmp(Context, Part + Index, Op.Size);
+            }
+
+            Context.EmitLdvectmp();
+            Context.EmitStvec(Op.Rd);
+
+            if (Part == 0)
+            {
+                EmitVectorZeroUpper(Context, Op.Rd);
+            }
+
+            Context.EmitLdarg(ATranslatedSub.StateArgIdx);
+            Context.EmitLdarg(ATranslatedSub.StateArgIdx);
+            Context.EmitCallPropGet(typeof(AThreadState), nameof(AThreadState.Fpsr));
+            Context.EmitLdtmp();
+            Context.Emit(OpCodes.Conv_I4);
+            Context.Emit(OpCodes.Or);
+            Context.EmitCallPropSet(typeof(AThreadState), nameof(AThreadState.Fpsr));
+        }
+
         public static void EmitScalarSet(AILEmitterCtx Context, int Reg, int Size)
         {
             EmitVectorZeroAll(Context, Reg);

ChocolArm64/Instruction/AInstEmitSimdLogical.cs

@@ -55,7 +55,7 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdReg Op = (AOpCodeSimdReg)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
             int Elems = Bytes >> Op.Size;
 
             for (int Index = 0; Index < Elems; Index++)
@@ -195,7 +195,7 @@ namespace ChocolArm64.Instruction
                 throw new InvalidOperationException();
             }
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
             int Elems = Bytes >> Op.Size;
 
             int ContainerMask = (1 << (ContainerSize - Op.Size)) - 1;

ChocolArm64/Instruction/AInstEmitSimdMemory.cs

@@ -105,13 +105,14 @@ namespace ChocolArm64.Instruction
                     throw new InvalidOperationException();
                 }
 
-                int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+                int Bytes = Op.GetBitsCount() >> 3;
+                int Elems = Bytes >> Op.Size;
 
                 for (int SElem = 0; SElem < Op.SElems; SElem++)
                 {
                     int Rt = (Op.Rt + SElem) & 0x1f;
 
-                    for (int Index = 0; Index < (Bytes >> Op.Size); Index++)
+                    for (int Index = 0; Index < Elems; Index++)
                     {
                         EmitMemAddress();
 

ChocolArm64/Instruction/AInstEmitSimdMove.cs

@@ -14,9 +14,10 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdIns Op = (AOpCodeSimdIns)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
+            int Elems = Bytes >> Op.Size;
 
-            for (int Index = 0; Index < (Bytes >> Op.Size); Index++)
+            for (int Index = 0; Index < Elems; Index++)
             {
                 Context.EmitLdintzr(Op.Rn);
 
@@ -42,9 +43,10 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdIns Op = (AOpCodeSimdIns)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
+            int Elems = Bytes >> Op.Size;
 
-            for (int Index = 0; Index < (Bytes >> Op.Size); Index++)
+            for (int Index = 0; Index < Elems; Index++)
             {
                 EmitVectorExtractZx(Context, Op.Rn, Op.DstIndex, Op.Size);
 
@@ -64,7 +66,7 @@ namespace ChocolArm64.Instruction
             Context.EmitLdvec(Op.Rd);
             Context.EmitStvectmp();
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
 
             int Position = Op.Imm4;
 
@@ -338,19 +340,23 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdReg Op = (AOpCodeSimdReg)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Words = Op.GetBitsCount() >> 4;
+            int Pairs = Words >> Op.Size;
 
-            int Elems = Bytes >> Op.Size;
-
-            for (int Index = 0; Index < Elems; Index++)
+            for (int Index = 0; Index < Pairs; Index++)
             {
-                int Elem = (Index & ~1) + Part;
+                int Idx = Index << 1;
 
-                EmitVectorExtractZx(Context, (Index & 1) == 0 ? Op.Rn : Op.Rm, Elem, Op.Size);
+                EmitVectorExtractZx(Context, Op.Rn, Idx + Part, Op.Size);
+                EmitVectorExtractZx(Context, Op.Rm, Idx + Part, Op.Size);
 
-                EmitVectorInsert(Context, Op.Rd, Index, Op.Size);
+                EmitVectorInsertTmp(Context, Idx + 1, Op.Size);
+                EmitVectorInsertTmp(Context, Idx    , Op.Size);
             }
 
+            Context.EmitLdvectmp();
+            Context.EmitStvec(Op.Rd);
+
             if (Op.RegisterSize == ARegisterSize.SIMD64)
             {
                 EmitVectorZeroUpper(Context, Op.Rd);
@@ -361,20 +367,23 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdReg Op = (AOpCodeSimdReg)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Words = Op.GetBitsCount() >> 4;
+            int Pairs = Words >> Op.Size;
 
-            int Elems = Bytes >> Op.Size;
-            int Half  = Elems >> 1;
-
-            for (int Index = 0; Index < Elems; Index++)
+            for (int Index = 0; Index < Pairs; Index++)
             {
-                int Elem = Part + ((Index & (Half - 1)) << 1);
+                int Idx = Index << 1;
 
-                EmitVectorExtractZx(Context, Index < Half ? Op.Rn : Op.Rm, Elem, Op.Size);
+                EmitVectorExtractZx(Context, Op.Rn, Idx + Part, Op.Size);
+                EmitVectorExtractZx(Context, Op.Rm, Idx + Part, Op.Size);
 
-                EmitVectorInsert(Context, Op.Rd, Index, Op.Size);
+                EmitVectorInsertTmp(Context, Pairs + Index, Op.Size);
+                EmitVectorInsertTmp(Context,         Index, Op.Size);
             }
 
+            Context.EmitLdvectmp();
+            Context.EmitStvec(Op.Rd);
+
             if (Op.RegisterSize == ARegisterSize.SIMD64)
             {
                 EmitVectorZeroUpper(Context, Op.Rd);
@@ -385,20 +394,25 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdReg Op = (AOpCodeSimdReg)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Words = Op.GetBitsCount() >> 4;
+            int Pairs = Words >> Op.Size;
 
-            int Elems = Bytes >> Op.Size;
-            int Half  = Elems >> 1;
+            int Base = Part != 0 ? Pairs : 0;
 
-            for (int Index = 0; Index < Elems; Index++)
+            for (int Index = 0; Index < Pairs; Index++)
             {
-                int Elem = Part * Half + (Index >> 1);
+                int Idx = Index << 1;
 
-                EmitVectorExtractZx(Context, (Index & 1) == 0 ? Op.Rn : Op.Rm, Elem, Op.Size);
+                EmitVectorExtractZx(Context, Op.Rn, Base + Index, Op.Size);
+                EmitVectorExtractZx(Context, Op.Rm, Base + Index, Op.Size);
 
-                EmitVectorInsert(Context, Op.Rd, Index, Op.Size);
+                EmitVectorInsertTmp(Context, Idx + 1, Op.Size);
+                EmitVectorInsertTmp(Context, Idx    , Op.Size);
             }
 
+            Context.EmitLdvectmp();
+            Context.EmitStvec(Op.Rd);
+
             if (Op.RegisterSize == ARegisterSize.SIMD64)
             {
                 EmitVectorZeroUpper(Context, Op.Rd);

ChocolArm64/Instruction/AInstEmitSimdShift.cs

@@ -27,9 +27,7 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdShImm Op = (AOpCodeSimdShImm)Context.CurrOp;
 
-            int Shift = Op.Imm - (8 << Op.Size);
-
-            EmitVectorShImmBinaryZx(Context, () => Context.Emit(OpCodes.Shl), Shift);
+            EmitVectorShImmBinaryZx(Context, () => Context.Emit(OpCodes.Shl), GetImmShl(Op));
         }
 
         public static void Shll_V(AILEmitterCtx Context)
@@ -45,22 +43,21 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdShImm Op = (AOpCodeSimdShImm)Context.CurrOp;
 
-            int Shift = (8 << (Op.Size + 1)) - Op.Imm;
-
-            EmitVectorShImmNarrowBinaryZx(Context, () => Context.Emit(OpCodes.Shr_Un), Shift);
+            EmitVectorShImmNarrowBinaryZx(Context, () => Context.Emit(OpCodes.Shr_Un), GetImmShr(Op));
         }
 
         public static void Sli_V(AILEmitterCtx Context)
         {
             AOpCodeSimdShImm Op = (AOpCodeSimdShImm)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
+            int Elems = Bytes >> Op.Size;
 
-            int Shift = Op.Imm - (8 << Op.Size);
+            int Shift = GetImmShl(Op);
 
-            ulong Mask = Shift != 0 ? ulong.MaxValue >> (64 - Shift) : 0;            
+            ulong Mask = Shift != 0 ? ulong.MaxValue >> (64 - Shift) : 0;
 
-            for (int Index = 0; Index < (Bytes >> Op.Size); Index++)
+            for (int Index = 0; Index < Elems; Index++)
             {
                 EmitVectorExtractZx(Context, Op.Rn, Index, Op.Size);
 
@@ -84,6 +81,39 @@ namespace ChocolArm64.Instruction
             }
         }
 
+        public static void Sqrshrn_V(AILEmitterCtx Context)
+        {
+            AOpCodeSimdShImm Op = (AOpCodeSimdShImm)Context.CurrOp;
+
+            int Shift = GetImmShr(Op);
+
+            long RoundConst = 1L << (Shift - 1);
+
+            Action Emit = () =>
+            {
+                Context.EmitLdc_I8(RoundConst);
+
+                Context.Emit(OpCodes.Add);
+
+                Context.EmitLdc_I4(Shift);
+
+                Context.Emit(OpCodes.Shr);
+            };
+
+            EmitVectorSaturatingNarrowOpSxSx(Context, Emit);
+        }
+
+        public static void Srshr_V(AILEmitterCtx Context)
+        {
+            AOpCodeSimdShImm Op = (AOpCodeSimdShImm)Context.CurrOp;
+
+            int Shift = GetImmShr(Op);
+
+            long RoundConst = 1L << (Shift - 1);
+
+            EmitVectorRoundShImmBinarySx(Context, () => Context.Emit(OpCodes.Shr), Shift, RoundConst);
+        }
+
         public static void Sshl_V(AILEmitterCtx Context)
         {
             EmitVectorShl(Context, Signed: true);
@@ -93,9 +123,7 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdShImm Op = (AOpCodeSimdShImm)Context.CurrOp;
 
-            int Shift = Op.Imm - (8 << Op.Size);
-
-            EmitVectorShImmWidenBinarySx(Context, () => Context.Emit(OpCodes.Shl), Shift);
+            EmitVectorShImmWidenBinarySx(Context, () => Context.Emit(OpCodes.Shl), GetImmShl(Op));
         }
 
         public static void Sshr_S(AILEmitterCtx Context)
@@ -115,24 +143,20 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdShImm Op = (AOpCodeSimdShImm)Context.CurrOp;
 
-            int Shift = (8 << (Op.Size + 1)) - Op.Imm;
-
-            EmitVectorShImmBinarySx(Context, () => Context.Emit(OpCodes.Shr), Shift);
+            EmitVectorShImmBinarySx(Context, () => Context.Emit(OpCodes.Shr), GetImmShr(Op));
         }
 
         public static void Ssra_V(AILEmitterCtx Context)
         {
             AOpCodeSimdShImm Op = (AOpCodeSimdShImm)Context.CurrOp;
 
-            int Shift = (8 << (Op.Size + 1)) - Op.Imm;
-
             Action Emit = () =>
             {
                 Context.Emit(OpCodes.Shr);
                 Context.Emit(OpCodes.Add);
             };
 
-            EmitVectorShImmTernarySx(Context, Emit, Shift);
+            EmitVectorShImmTernarySx(Context, Emit, GetImmShr(Op));
         }
 
         public static void Ushl_V(AILEmitterCtx Context)
@@ -144,9 +168,7 @@ namespace ChocolArm64.Instruction
         {
             AOpCodeSimdShImm Op = (AOpCodeSimdShImm)Context.CurrOp;
 
-            int Shift = Op.Imm - (8 << Op.Size);
-
-            EmitVectorShImmWidenBinaryZx(Context, () => Context.Emit(OpCodes.Shl), Shift);
+            EmitVectorShImmWidenBinaryZx(Context, () => Context.Emit(OpCodes.Shl), GetImmShl(Op));
         }
 
         public static void Ushr_S(AILEmitterCtx Context)
@@ -251,28 +273,51 @@ namespace ChocolArm64.Instruction
             }
         }
 
+        [Flags]
+        private enum ShImmFlags
+        {
+            None = 0,
+
+            Signed  = 1 << 0,
+            Ternary = 1 << 1,
+            Rounded = 1 << 2,
+
+            SignedTernary = Signed | Ternary,
+            SignedRounded = Signed | Rounded
+        }
+
         private static void EmitVectorShImmBinarySx(AILEmitterCtx Context, Action Emit, int Imm)
         {
-            EmitVectorShImmOp(Context, Emit, Imm, false, true);
+            EmitVectorShImmOp(Context, Emit, Imm, ShImmFlags.Signed);
         }
 
         private static void EmitVectorShImmTernarySx(AILEmitterCtx Context, Action Emit, int Imm)
         {
-            EmitVectorShImmOp(Context, Emit, Imm, true, true);
+            EmitVectorShImmOp(Context, Emit, Imm, ShImmFlags.SignedTernary);
         }
 
         private static void EmitVectorShImmBinaryZx(AILEmitterCtx Context, Action Emit, int Imm)
         {
-            EmitVectorShImmOp(Context, Emit, Imm, false, false);
+            EmitVectorShImmOp(Context, Emit, Imm, ShImmFlags.None);
         }
 
-        private static void EmitVectorShImmOp(AILEmitterCtx Context, Action Emit, int Imm, bool Ternary, bool Signed)
+        private static void EmitVectorRoundShImmBinarySx(AILEmitterCtx Context, Action Emit, int Imm, long Rc)
+        {
+            EmitVectorShImmOp(Context, Emit, Imm, ShImmFlags.SignedRounded, Rc);
+        }
+
+        private static void EmitVectorShImmOp(AILEmitterCtx Context, Action Emit, int Imm, ShImmFlags Flags, long Rc = 0)
         {
             AOpCodeSimd Op = (AOpCodeSimd)Context.CurrOp;
 
-            int Bytes = Context.CurrOp.GetBitsCount() >> 3;
+            int Bytes = Op.GetBitsCount() >> 3;
+            int Elems = Bytes >> Op.Size;
 
-            for (int Index = 0; Index < (Bytes >> Op.Size); Index++)
+            bool Signed  = (Flags & ShImmFlags.Signed)  != 0;
+            bool Ternary = (Flags & ShImmFlags.Ternary) != 0;
+            bool Rounded = (Flags & ShImmFlags.Rounded) != 0;
+
+            for (int Index = 0; Index < Elems; Index++)
             {
                 if (Ternary)
                 {
@@ -281,6 +326,13 @@ namespace ChocolArm64.Instruction
 
                 EmitVectorExtract(Context, Op.Rn, Index, Op.Size, Signed);
 
+                if (Rounded)
+                {
+                    Context.EmitLdc_I8(Rc);
+
+                    Context.Emit(OpCodes.Add);
+                }
+
                 Context.EmitLdc_I4(Imm);
 
                 Emit();

ChocolArm64/Instruction/ASoftFallback.cs

@@ -12,22 +12,42 @@ namespace ChocolArm64.Instruction
 
         public static ulong CountLeadingSigns(ulong Value, int Size)
         {
-            return CountLeadingZeros((Value >> 1) ^ Value, Size - 1);
-        }
+            Value ^= Value >> 1;
 
-        public static ulong CountLeadingZeros(ulong Value, int Size)
-        {
-            int HighBit = Size - 1;
+            int HighBit = Size - 2;
 
             for (int Bit = HighBit; Bit >= 0; Bit--)
             {
-                if (((Value >> Bit) & 1) != 0)
+                if (((Value >> Bit) & 0b1) != 0)
                 {
                     return (ulong)(HighBit - Bit);
                 }
             }
 
-            return (ulong)Size;
+            return (ulong)(Size - 1);
+        }
+
+        private static readonly byte[] ClzNibbleTbl = { 4, 3, 2, 2, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0 };
+
+        public static ulong CountLeadingZeros(ulong Value, int Size)
+        {
+            if (Value == 0)
+            {
+                return (ulong)Size;
+            }
+
+            int NibbleIdx = Size;
+            int PreCount, Count = 0;
+
+            do
+            {
+                NibbleIdx -= 4;
+                PreCount = ClzNibbleTbl[(Value >> NibbleIdx) & 0b1111];
+                Count += PreCount;
+            }
+            while (PreCount == 4);
+
+            return (ulong)Count;
         }
 
         public static uint CountSetBits8(uint Value)
@@ -61,8 +81,8 @@ namespace ChocolArm64.Instruction
 
         private static uint Crc32w(uint Crc, uint Poly, uint Val)
         {
-            Crc = Crc32(Crc, Poly, (byte)(Val >> 0));
-            Crc = Crc32(Crc, Poly, (byte)(Val >> 8));
+            Crc = Crc32(Crc, Poly, (byte)(Val >> 0 ));
+            Crc = Crc32(Crc, Poly, (byte)(Val >> 8 ));
             Crc = Crc32(Crc, Poly, (byte)(Val >> 16));
             Crc = Crc32(Crc, Poly, (byte)(Val >> 24));
 
@@ -71,8 +91,8 @@ namespace ChocolArm64.Instruction
 
         private static uint Crc32x(uint Crc, uint Poly, ulong Val)
         {
-            Crc = Crc32(Crc, Poly, (byte)(Val >> 0));
-            Crc = Crc32(Crc, Poly, (byte)(Val >> 8));
+            Crc = Crc32(Crc, Poly, (byte)(Val >> 0 ));
+            Crc = Crc32(Crc, Poly, (byte)(Val >> 8 ));
             Crc = Crc32(Crc, Poly, (byte)(Val >> 16));
             Crc = Crc32(Crc, Poly, (byte)(Val >> 24));
             Crc = Crc32(Crc, Poly, (byte)(Val >> 32));
@@ -168,9 +188,10 @@ namespace ChocolArm64.Instruction
 
         public static long SMulHi128(long LHS, long RHS)
         {
-            long Result = (long)UMulHi128((ulong)(LHS), (ulong)(RHS));
+            long Result = (long)UMulHi128((ulong)LHS, (ulong)RHS);
             if (LHS < 0) Result -= RHS;
             if (RHS < 0) Result -= LHS;
+
             return Result;
         }
 
@@ -187,6 +208,7 @@ namespace ChocolArm64.Instruction
             ulong Z1 = T & 0xFFFFFFFF;
             ulong Z0 = T >> 32;
             Z1 += LLow * RHigh;
+
             return LHigh * RHigh + Z0 + (Z1 >> 32);
         }
     }

ChocolArm64/Instruction/ASoftFloat.cs

@@ -7,8 +7,10 @@ namespace ChocolArm64.Instruction
         static ASoftFloat()
         {
             InvSqrtEstimateTable = BuildInvSqrtEstimateTable();
+            RecipEstimateTable = BuildRecipEstimateTable();
         }
 
+        private static readonly byte[] RecipEstimateTable;
         private static readonly byte[] InvSqrtEstimateTable;
 
         private static byte[] BuildInvSqrtEstimateTable()
@@ -38,6 +40,22 @@ namespace ChocolArm64.Instruction
             return Table;
         }
 
+        private static byte[] BuildRecipEstimateTable()
+        {
+            byte[] Table = new byte[256];
+            for (ulong index = 0; index < 256; index++)
+            {
+                ulong a = index | 0x100;
+
+                a = (a << 1) + 1;
+                ulong b = 0x80000 / a;
+                b = (b + 1) >> 1;
+
+                Table[index] = (byte)(b & 0xFF);
+            }
+            return Table;
+        }
+
         public static float InvSqrtEstimate(float x)
         {
             return (float)InvSqrtEstimate((double)x);
@@ -105,5 +123,143 @@ namespace ChocolArm64.Instruction
             ulong result = x_sign | (result_exp << 52) | fraction;
             return BitConverter.Int64BitsToDouble((long)result);
         }
+
+        public static float RecipEstimate(float x)
+        {
+            return (float)RecipEstimate((double)x);
+        }
+
+        public static double RecipEstimate(double x)
+        {
+            ulong x_bits = (ulong)BitConverter.DoubleToInt64Bits(x);
+            ulong x_sign = x_bits & 0x8000000000000000;
+            ulong x_exp = (x_bits >> 52) & 0x7FF;
+            ulong scaled = x_bits & ((1ul << 52) - 1);
+
+            if (x_exp >= 2045)
+            {
+                if (x_exp == 0x7ff && scaled != 0)
+                {
+                    // NaN
+                    return BitConverter.Int64BitsToDouble((long)(x_bits | 0x0008000000000000));
+                }
+
+                // Infinity, or Out of range -> Zero
+                return BitConverter.Int64BitsToDouble((long)x_sign);
+            }
+
+            if (x_exp == 0)
+            {
+                if (scaled == 0)
+                {
+                    // Zero -> Infinity
+                    return BitConverter.Int64BitsToDouble((long)(x_sign | 0x7ff0000000000000));
+                }
+
+                // Denormal
+                if ((scaled & (1ul << 51)) == 0)
+                {
+                    x_exp = ~0ul;
+                    scaled <<= 2;
+                }
+                else
+                {
+                    scaled <<= 1;
+                }
+            }
+
+            scaled >>= 44;
+            scaled &= 0xFF;
+
+            ulong result_exp = (2045 - x_exp) & 0x7FF;
+            ulong estimate = (ulong)RecipEstimateTable[scaled];
+            ulong fraction = estimate << 44;
+
+            if (result_exp == 0)
+            {
+                fraction >>= 1;
+                fraction |= 1ul << 51;
+            }
+            else if (result_exp == 0x7FF)
+            {
+                result_exp = 0;
+                fraction >>= 2;
+                fraction |= 1ul << 50;
+            }
+
+            ulong result = x_sign | (result_exp << 52) | fraction;
+            return BitConverter.Int64BitsToDouble((long)result);
+        }
+
+        public static float RecipStep(float op1, float op2)
+        {
+            return (float)RecipStep((double)op1, (double)op2);
+        }
+
+        public static double RecipStep(double op1, double op2)
+        {
+            op1 = -op1;
+
+            ulong op1_bits = (ulong)BitConverter.DoubleToInt64Bits(op1);
+            ulong op2_bits = (ulong)BitConverter.DoubleToInt64Bits(op2);
+
+            ulong op1_sign = op1_bits & 0x8000000000000000;
+            ulong op2_sign = op2_bits & 0x8000000000000000;
+            ulong op1_other = op1_bits & 0x7FFFFFFFFFFFFFFF;
+            ulong op2_other = op2_bits & 0x7FFFFFFFFFFFFFFF;
+
+            bool inf1 = op1_other == 0x7ff0000000000000;
+            bool inf2 = op2_other == 0x7ff0000000000000;
+            bool zero1 = op1_other == 0;
+            bool zero2 = op2_other == 0;
+
+            if ((inf1 && zero2) || (zero1 && inf2))
+            {
+                return 2.0;
+            }
+            else if (inf1 || inf2)
+            {
+                // Infinity
+                return BitConverter.Int64BitsToDouble((long)(0x7ff0000000000000 | (op1_sign ^ op2_sign)));
+            }
+
+            return 2.0 + op1 * op2;
+        }
+
+        public static float ConvertHalfToSingle(ushort x)
+        {
+            uint x_sign = (uint)(x >> 15) & 0x0001;
+            uint x_exp = (uint)(x >> 10) & 0x001F;
+            uint x_mantissa = (uint)x & 0x03FF;
+
+            if (x_exp == 0 && x_mantissa == 0)
+            {
+                // Zero
+                return BitConverter.Int32BitsToSingle((int)(x_sign << 31));
+            }
+
+            if (x_exp == 0x1F)
+            {
+                // NaN or Infinity
+                return BitConverter.Int32BitsToSingle((int)((x_sign << 31) | 0x7F800000 | (x_mantissa << 13)));
+            }
+
+            int exponent = (int)x_exp - 15;
+
+            if (x_exp == 0)
+            {
+                // Denormal
+                x_mantissa <<= 1;
+                while ((x_mantissa & 0x0400) == 0)
+                {
+                    x_mantissa <<= 1;
+                    exponent--;
+                }
+                x_mantissa &= 0x03FF;
+            }
+
+            uint new_exp = (uint)((exponent + 127) & 0xFF) << 23;
+            return BitConverter.Int32BitsToSingle((int)((x_sign << 31) | new_exp | (x_mantissa << 13)));
+        }
     }
 }

ChocolArm64/Memory/AMemory.cs

@@ -33,19 +33,25 @@ namespace ChocolArm64.Memory
 
         private byte* RamPtr;
 
+        private int HostPageSize;
+
         public AMemory()
         {
             Manager = new AMemoryMgr();
 
             Monitors = new Dictionary<int, ArmMonitor>();
 
+            IntPtr Size = (IntPtr)AMemoryMgr.RamSize + AMemoryMgr.PageSize;
+
             if (RuntimeInformation.IsOSPlatform(OSPlatform.Windows))
             {
-                Ram = AMemoryWin32.Allocate((IntPtr)AMemoryMgr.RamSize + AMemoryMgr.PageSize);
+                Ram = AMemoryWin32.Allocate(Size);
+
+                HostPageSize = AMemoryWin32.GetPageSize(Ram, Size);
             }
             else
             {
-                Ram = Marshal.AllocHGlobal((IntPtr)AMemoryMgr.RamSize + AMemoryMgr.PageSize);
+                Ram = Marshal.AllocHGlobal(Size);
             }
 
             RamPtr = (byte*)Ram;
@@ -149,49 +155,53 @@ namespace ChocolArm64.Memory
             }
         }
 
-        public long GetHostPageSize()
+        public int GetHostPageSize()
         {
-            if (!RuntimeInformation.IsOSPlatform(OSPlatform.Windows))
-            {
-                return AMemoryMgr.PageSize;
-            }
-
-            IntPtr MemAddress = new IntPtr(RamPtr);
-            IntPtr MemSize    = new IntPtr(AMemoryMgr.RamSize);
-
-            long PageSize = AMemoryWin32.IsRegionModified(MemAddress, MemSize, Reset: false);
-
-            if (PageSize < 1)
-            {
-                throw new InvalidOperationException();
-            }
-
-            return PageSize;
+            return HostPageSize;
         }
 
-        public bool IsRegionModified(long Position, long Size)
+        public bool[] IsRegionModified(long Position, long Size)
         {
             if (!RuntimeInformation.IsOSPlatform(OSPlatform.Windows))
             {
-                return true;
+                return null;
             }
 
             long EndPos = Position + Size;
 
             if ((ulong)EndPos < (ulong)Position)
             {
-                return false;
+                return null;
             }
 
             if ((ulong)EndPos > AMemoryMgr.RamSize)
             {
-                return false;
+                return null;
             }
 
             IntPtr MemAddress = new IntPtr(RamPtr + Position);
             IntPtr MemSize    = new IntPtr(Size);
 
-            return AMemoryWin32.IsRegionModified(MemAddress, MemSize, Reset: true) != 0;
+            int HostPageMask = HostPageSize - 1;
+
+            Position &= ~HostPageMask;
+
+            Size = EndPos - Position;
+
+            IntPtr[] Addresses  = new IntPtr[(Size + HostPageMask) / HostPageSize];
+
+            AMemoryWin32.IsRegionModified(MemAddress, MemSize, Addresses, out int Count);
+
+            bool[] Modified = new bool[Addresses.Length];
+
+            for (int Index = 0; Index < Count; Index++)
+            {
+                long VA = Addresses[Index].ToInt64() - Ram.ToInt64();
+
+                Modified[(VA - Position) / HostPageSize] = true;
+            }
+
+            return Modified;
         }
 
         public sbyte ReadSByte(long Position)

ChocolArm64/Memory/AMemoryWin32.cs

@@ -49,7 +49,7 @@ namespace ChocolArm64.Memory
             VirtualFree(Address, IntPtr.Zero, MEM_RELEASE);
         }
 
-        public unsafe static long IsRegionModified(IntPtr Address, IntPtr Size, bool Reset)
+        public unsafe static int GetPageSize(IntPtr Address, IntPtr Size)
         {
             IntPtr[] Addresses = new IntPtr[1];
 
@@ -57,17 +57,36 @@ namespace ChocolArm64.Memory
 
             long Granularity;
 
-            int Flags = Reset ? WRITE_WATCH_FLAG_RESET : 0;
-
             GetWriteWatch(
-                Flags,
+                0,
                 Address,
                 Size,
                 Addresses,
                 &Count,
                 &Granularity);
 
-            return Count != 0 ? Granularity : 0;
+            return (int)Granularity;
+        }
+
+        public unsafe static void IsRegionModified(
+            IntPtr   Address,
+            IntPtr   Size,
+            IntPtr[] Addresses,
+            out int  AddrCount)
+        {
+            long Count = Addresses.Length;
+
+            long Granularity;
+
+            GetWriteWatch(
+                WRITE_WATCH_FLAG_RESET,
+                Address,
+                Size,
+                Addresses,
+                &Count,
+                &Granularity);
+
+            AddrCount = (int)Count;
         }
     }
 }

README.md

@@ -85,6 +85,8 @@ If you have some homebrew that currently doesn't work within the emulator, you c
 For help, support, suggestions, or if you just want to get in touch with the team; join our Discord server!  
 https://discord.gg/VkQYXAZ
 
+For donation support, please take a look at our Patreon: https://www.patreon.com/ryujinx
+
 **Running**
 
 To run this emulator, you need the .NET Core 2.1 (or higher) SDK *and* the OpenAL 11 Core SDK.  
@@ -92,6 +94,7 @@ Run `dotnet run -c Release -- path\to\homebrew.nro` inside the Ryujinx solution
 Run `dotnet run -c Release -- path\to\game_exefs_and_romfs_folder` to run official games (they need to be decrypted and extracted first!)
 
 **Compatibility**
+
 You can check out the compatibility list within the Wiki. Only a handful of games actually work.
 
 **Latest build**

Ryujinx.Audio/Adpcm/AdpcmDecoder.cs

@@ -0,0 +1,91 @@
+namespace Ryujinx.Audio.Adpcm
+{
+    public static class AdpcmDecoder
+    {
+        private const int SamplesPerFrame = 14;
+        private const int BytesPerFrame   = 8;
+
+        public static int[] Decode(byte[] Buffer, AdpcmDecoderContext Context)
+        {
+            int Samples = GetSamplesCountFromSize(Buffer.Length);
+
+            int[] Pcm = new int[Samples * 2];
+
+            short History0 = Context.History0;
+            short History1 = Context.History1;
+
+            int InputOffset  = 0;
+            int OutputOffset = 0;
+
+            while (InputOffset < Buffer.Length)
+            {
+                byte Header = Buffer[InputOffset++];
+
+                int Scale = 0x800 << (Header & 0xf);
+
+                int CoeffIndex = (Header >> 4) & 7;
+
+                short Coeff0 = Context.Coefficients[CoeffIndex * 2 + 0];
+                short Coeff1 = Context.Coefficients[CoeffIndex * 2 + 1];
+
+                int FrameSamples = SamplesPerFrame;
+
+                if (FrameSamples > Samples)
+                {
+                    FrameSamples = Samples;
+                }
+
+                int Value = 0;
+
+                for (int SampleIndex = 0; SampleIndex < FrameSamples; SampleIndex++)
+                {
+                    int Sample;
+
+                    if ((SampleIndex & 1) == 0)
+                    {
+                        Value = Buffer[InputOffset++];
+
+                        Sample = (Value << 24) >> 28;
+                    }
+                    else
+                    {
+                        Sample = (Value << 28) >> 28;
+                    }
+
+                    int Prediction = Coeff0 * History0 + Coeff1 * History1;
+
+                    Sample = (Sample * Scale + Prediction + 0x400) >> 11;
+
+                    short SaturatedSample = DspUtils.Saturate(Sample);
+
+                    History1 = History0;
+                    History0 = SaturatedSample;
+
+                    Pcm[OutputOffset++] = SaturatedSample;
+                    Pcm[OutputOffset++] = SaturatedSample;
+                }
+
+                Samples -= FrameSamples;
+            }
+
+            Context.History0 = History0;
+            Context.History1 = History1;
+
+            return Pcm;
+        }
+
+        public static long GetSizeFromSamplesCount(int SamplesCount)
+        {
+            int Frames = SamplesCount / SamplesPerFrame;
+
+            return Frames * BytesPerFrame;
+        }
+
+        public static int GetSamplesCountFromSize(long Size)
+        {
+            int Frames = (int)(Size / BytesPerFrame);
+
+            return Frames * SamplesPerFrame;
+        }
+    }
+}

Ryujinx.Audio/Adpcm/AdpcmDecoderContext.cs

@@ -0,0 +1,10 @@
+namespace Ryujinx.Audio.Adpcm
+{
+    public class AdpcmDecoderContext
+    {
+        public short[] Coefficients;
+
+        public short History0;
+        public short History1;
+    }
+}

Ryujinx.Audio/DspUtils.cs

@@ -0,0 +1,16 @@
+namespace Ryujinx.Audio.Adpcm
+{
+    public static class DspUtils
+    {
+        public static short Saturate(int Value)
+        {
+            if (Value > short.MaxValue)
+                Value = short.MaxValue;
+
+            if (Value < short.MinValue)
+                Value = short.MinValue;
+
+            return (short)Value;
+        }
+    }
+}

Ryujinx.Audio/IAalOutput.cs

@@ -2,11 +2,7 @@ namespace Ryujinx.Audio
 {
     public interface IAalOutput
     {
-        int OpenTrack(
-            int             SampleRate,
-            int             Channels,
-            ReleaseCallback Callback,
-            out AudioFormat Format);
+        int OpenTrack(int SampleRate, int Channels, ReleaseCallback Callback);
 
         void CloseTrack(int Track);
 
@@ -14,7 +10,7 @@ namespace Ryujinx.Audio
 
         long[] GetReleasedBuffers(int Track, int MaxCount);
 
-        void AppendBuffer(int Track, long Tag, byte[] Buffer);
+        void AppendBuffer<T>(int Track, long Tag, T[] Buffer)  where T : struct;
 
         void Start(int Track);
         void Stop(int Track);

Ryujinx.Audio/OpenAL/OpenALAudioOut.cs

@@ -3,6 +3,7 @@ using OpenTK.Audio.OpenAL;
 using System;
 using System.Collections.Concurrent;
 using System.Collections.Generic;
+using System.Runtime.InteropServices;
 using System.Threading;
 
 namespace Ryujinx.Audio.OpenAL
@@ -20,7 +21,7 @@ namespace Ryujinx.Audio.OpenAL
             public int SourceId { get; private set; }
 
             public int SampleRate { get; private set; }
-            
+
             public ALFormat Format { get; private set; }
 
             private ReleaseCallback Callback;
@@ -153,7 +154,7 @@ namespace Ryujinx.Audio.OpenAL
                     ShouldCallReleaseCallback = true;
                 }
             }
-            
+
             private void SyncQueuedTags()
             {
                 AL.GetSource(SourceId, ALGetSourcei.BuffersQueued,    out int QueuedCount);
@@ -226,15 +227,9 @@ namespace Ryujinx.Audio.OpenAL
             while (KeepPolling);
         }
 
-        public int OpenTrack(
-            int             SampleRate,
-            int             Channels,
-            ReleaseCallback Callback,
-            out AudioFormat Format)
+        public int OpenTrack(int SampleRate, int Channels, ReleaseCallback Callback)
         {
-            Format = AudioFormat.PcmInt16;
-
-            Track Td = new Track(SampleRate, GetALFormat(Channels, Format), Callback);
+            Track Td = new Track(SampleRate, GetALFormat(Channels), Callback);
 
             for (int Id = 0; Id < MaxTracks; Id++)
             {
@@ -247,31 +242,16 @@ namespace Ryujinx.Audio.OpenAL
             return -1;
         }
 
-        private ALFormat GetALFormat(int Channels, AudioFormat Format)
+        private ALFormat GetALFormat(int Channels)
         {
-            if (Channels < 1 || Channels > 2)
+            switch (Channels)
             {
-                throw new ArgumentOutOfRangeException(nameof(Channels));
+                case 1: return ALFormat.Mono16;
+                case 2: return ALFormat.Stereo16;
+                case 6: return ALFormat.Multi51Chn16Ext;
             }
 
-            if (Channels == 1)
-            {
-                switch (Format)
-                {
-                    case AudioFormat.PcmInt8:  return ALFormat.Mono8;
-                    case AudioFormat.PcmInt16: return ALFormat.Mono16;
-                }
-            }
-            else /* if (Channels == 2) */
-            {
-                switch (Format)
-                {
-                    case AudioFormat.PcmInt8:  return ALFormat.Stereo8;
-                    case AudioFormat.PcmInt16: return ALFormat.Stereo16;
-                }
-            }
-
-            throw new ArgumentException(nameof(Format));
+            throw new ArgumentOutOfRangeException(nameof(Channels));
         }
 
         public void CloseTrack(int Track)
@@ -288,7 +268,7 @@ namespace Ryujinx.Audio.OpenAL
             {
                 return Td.ContainsBuffer(Tag);
             }
-            
+
             return false;
         }
 
@@ -298,17 +278,19 @@ namespace Ryujinx.Audio.OpenAL
             {
                 return Td.GetReleasedBuffers(MaxCount);
             }
-            
+
             return null;
         }
 
-        public void AppendBuffer(int Track, long Tag, byte[] Buffer)
+        public void AppendBuffer<T>(int Track, long Tag, T[] Buffer) where T : struct
         {
             if (Tracks.TryGetValue(Track, out Track Td))
             {
                 int BufferId = Td.AppendBuffer(Tag);
 
-                AL.BufferData(BufferId, Td.Format, Buffer, Buffer.Length, Td.SampleRate);
+                int Size = Buffer.Length * Marshal.SizeOf<T>();
+
+                AL.BufferData<T>(BufferId, Td.Format, Buffer, Size, Td.SampleRate);
 
                 AL.SourceQueueBuffer(Td.SourceId, BufferId);
 
@@ -359,7 +341,5 @@ namespace Ryujinx.Audio.OpenAL
 
             return PlaybackState.Stopped;
         }
-
-
     }
 }

Ryujinx.Graphics/Gal/GalTextureFormat.cs

@@ -17,6 +17,7 @@ namespace Ryujinx.Graphics.Gal
         BC3          = 0x26,
         BC4          = 0x27,
         BC5          = 0x28,
+        ZF32         = 0x2f,
         Astc2D4x4    = 0x40,
         Astc2D5x5    = 0x41,
         Astc2D6x6    = 0x42,

Ryujinx.Graphics/Gal/IGalRasterizer.cs

@@ -2,6 +2,9 @@ namespace Ryujinx.Graphics.Gal
 {
     public interface IGalRasterizer
     {
+        void LockCaches();
+        void UnlockCaches();
+
         void ClearBuffers(GalClearBufferFlags Flags);
 
         bool IsVboCached(long Key, long DataSize);
@@ -46,9 +49,9 @@ namespace Ryujinx.Graphics.Gal
 
         void CreateIbo(long Key, byte[] Buffer);
 
-        void SetVertexArray(int VbIndex, int Stride, long VboKey, GalVertexAttrib[] Attribs);
+        void SetVertexArray(int Stride, long VboKey, GalVertexAttrib[] Attribs);
 
-        void SetIndexArray(long Key, int Size, GalIndexFormat Format);
+        void SetIndexArray(int Size, GalIndexFormat Format);
 
         void DrawArrays(int First, int PrimCount, GalPrimitiveType PrimType);
 

Ryujinx.Graphics/Gal/IGalShader.cs

@@ -18,6 +18,8 @@ namespace Ryujinx.Graphics.Gal
 
         void Bind(long Key);
 
+        void Unbind(GalShaderType Type);
+
         void BindProgram();
     }
 }

Ryujinx.Graphics/Gal/IGalTexture.cs

@@ -2,6 +2,9 @@ namespace Ryujinx.Graphics.Gal
 {
     public interface IGalTexture
     {
+        void LockCache();
+        void UnlockCache();
+
         void Create(long Key, byte[] Data, GalTexture Texture);
 
         bool TryGetCachedTexture(long Key, long DataSize, out GalTexture Texture);

Ryujinx.Graphics/Gal/OpenGL/OGLCachedResource.cs

@@ -36,6 +36,10 @@ namespace Ryujinx.Graphics.Gal.OpenGL
 
         private DeleteValue DeleteValueCallback;
 
+        private Queue<T> DeletePending;
+
+        private bool Locked;
+
         public OGLCachedResource(DeleteValue DeleteValueCallback)
         {
             if (DeleteValueCallback == null)
@@ -48,11 +52,33 @@ namespace Ryujinx.Graphics.Gal.OpenGL
             Cache = new Dictionary<long, CacheBucket>();
 
             SortedCache = new LinkedList<long>();
+
+            DeletePending = new Queue<T>();
+        }
+
+        public void Lock()
+        {
+            Locked = true;
+        }
+
+        public void Unlock()
+        {
+            Locked = false;
+
+            while (DeletePending.TryDequeue(out T Value))
+            {
+                DeleteValueCallback(Value);
+            }
+
+            ClearCacheIfNeeded();
         }
 
         public void AddOrUpdate(long Key, T Value, long Size)
         {
-            ClearCacheIfNeeded();
+            if (!Locked)
+            {
+                ClearCacheIfNeeded();
+            }
 
             LinkedListNode<long> Node = SortedCache.AddLast(Key);
 
@@ -60,7 +86,14 @@ namespace Ryujinx.Graphics.Gal.OpenGL
 
             if (Cache.TryGetValue(Key, out CacheBucket Bucket))
             {
-                DeleteValueCallback(Bucket.Value);
+                if (Locked)
+                {
+                    DeletePending.Enqueue(Bucket.Value);
+                }
+                else
+                {
+                    DeleteValueCallback(Bucket.Value);
+                }
 
                 SortedCache.Remove(Bucket.Node);
 
@@ -78,6 +111,12 @@ namespace Ryujinx.Graphics.Gal.OpenGL
             {
                 Value = Bucket.Value;
 
+                SortedCache.Remove(Bucket.Node);
+
+                LinkedListNode<long> Node = SortedCache.AddLast(Key);
+
+                Cache[Key] = new CacheBucket(Value, Bucket.DataSize, Node);
+
                 return true;
             }
 

Ryujinx.Graphics/Gal/OpenGL/OGLEnumConverter.cs

@@ -129,15 +129,16 @@ namespace Ryujinx.Graphics.Gal.OpenGL
         {
             switch (Format)
             {
-                case GalTextureFormat.R32G32B32A32: return (PixelFormat.Rgba, PixelType.Float);
-                case GalTextureFormat.R16G16B16A16: return (PixelFormat.Rgba, PixelType.HalfFloat);
-                case GalTextureFormat.A8B8G8R8:     return (PixelFormat.Rgba, PixelType.UnsignedByte);
-                case GalTextureFormat.R32:          return (PixelFormat.Red,  PixelType.Float);
-                case GalTextureFormat.A1B5G5R5:     return (PixelFormat.Rgba, PixelType.UnsignedShort5551);
-                case GalTextureFormat.B5G6R5:       return (PixelFormat.Rgb,  PixelType.UnsignedShort565);
-                case GalTextureFormat.G8R8:         return (PixelFormat.Rg,   PixelType.UnsignedByte);
-                case GalTextureFormat.R16:          return (PixelFormat.Red,  PixelType.HalfFloat);
-                case GalTextureFormat.R8:           return (PixelFormat.Red,  PixelType.UnsignedByte);
+                case GalTextureFormat.R32G32B32A32: return (PixelFormat.Rgba,           PixelType.Float);
+                case GalTextureFormat.R16G16B16A16: return (PixelFormat.Rgba,           PixelType.HalfFloat);
+                case GalTextureFormat.A8B8G8R8:     return (PixelFormat.Rgba,           PixelType.UnsignedByte);
+                case GalTextureFormat.R32:          return (PixelFormat.Red,            PixelType.Float);
+                case GalTextureFormat.A1B5G5R5:     return (PixelFormat.Rgba,           PixelType.UnsignedShort5551);
+                case GalTextureFormat.B5G6R5:       return (PixelFormat.Rgb,            PixelType.UnsignedShort565);
+                case GalTextureFormat.G8R8:         return (PixelFormat.Rg,             PixelType.UnsignedByte);
+                case GalTextureFormat.R16:          return (PixelFormat.Red,            PixelType.HalfFloat);
+                case GalTextureFormat.R8:           return (PixelFormat.Red,            PixelType.UnsignedByte);
+                case GalTextureFormat.ZF32:         return (PixelFormat.DepthComponent, PixelType.Float);
             }
 
             throw new NotImplementedException(Format.ToString());

Ryujinx.Graphics/Gal/OpenGL/OGLRasterizer.cs

@@ -71,6 +71,18 @@ namespace Ryujinx.Graphics.Gal.OpenGL
             IndexBuffer = new IbInfo();
         }
 
+        public void LockCaches()
+        {
+            VboCache.Lock();
+            IboCache.Lock();
+        }
+
+        public void UnlockCaches()
+        {
+            VboCache.Unlock();
+            IboCache.Unlock();
+        }
+
         public void ClearBuffers(GalClearBufferFlags Flags)
         {
             ClearBufferMask Mask = ClearBufferMask.ColorBufferBit;
@@ -223,7 +235,7 @@ namespace Ryujinx.Graphics.Gal.OpenGL
             GL.BufferData(BufferTarget.ElementArrayBuffer, Length, Buffer, BufferUsageHint.StreamDraw);
         }
 
-        public void SetVertexArray(int VbIndex, int Stride, long VboKey, GalVertexAttrib[] Attribs)
+        public void SetVertexArray(int Stride, long VboKey, GalVertexAttrib[] Attribs)
         {
             if (!VboCache.TryGetValue(VboKey, out int VboHandle))
             {
@@ -270,7 +282,7 @@ namespace Ryujinx.Graphics.Gal.OpenGL
             }
         }
 
-        public void SetIndexArray(long Key, int Size, GalIndexFormat Format)
+        public void SetIndexArray(int Size, GalIndexFormat Format)
         {
             IndexBuffer.Type = OGLEnumConverter.GetDrawElementsType(Format);
 

Ryujinx.Graphics/Gal/OpenGL/OGLShader.cs

@@ -118,6 +118,9 @@ namespace Ryujinx.Graphics.Gal.OpenGL
 
             if (IsDualVp)
             {
+                ShaderDumper.Dump(Memory, Position  + 0x50, Type, "a");
+                ShaderDumper.Dump(Memory, PositionB + 0x50, Type, "b");
+
                 Program = Decompiler.Decompile(
                     Memory,
                     Position  + 0x50,
@@ -126,6 +129,8 @@ namespace Ryujinx.Graphics.Gal.OpenGL
             }
             else
             {
+                ShaderDumper.Dump(Memory, Position + 0x50, Type);
+
                 Program = Decompiler.Decompile(Memory, Position + 0x50, Type);
             }
 
@@ -203,6 +208,18 @@ namespace Ryujinx.Graphics.Gal.OpenGL
             }
         }
 
+        public void Unbind(GalShaderType Type)
+        {
+            switch (Type)
+            {
+                case GalShaderType.Vertex:         Current.Vertex         = null; break;
+                case GalShaderType.TessControl:    Current.TessControl    = null; break;
+                case GalShaderType.TessEvaluation: Current.TessEvaluation = null; break;
+                case GalShaderType.Geometry:       Current.Geometry       = null; break;
+                case GalShaderType.Fragment:       Current.Fragment       = null; break;
+            }
+        }
+
         public void BindProgram()
         {
             if (Current.Vertex   == null ||

Ryujinx.Graphics/Gal/OpenGL/OGLTexture.cs

@@ -26,6 +26,16 @@ namespace Ryujinx.Graphics.Gal.OpenGL
             TextureCache = new OGLCachedResource<TCE>(DeleteTexture);
         }
 
+        public void LockCache()
+        {
+            TextureCache.Lock();
+        }
+
+        public void UnlockCache()
+        {
+            TextureCache.Unlock();
+        }
+
         private static void DeleteTexture(TCE CachedTexture)
         {
             GL.DeleteTexture(CachedTexture.Handle);

Ryujinx.Graphics/Gal/Shader/GlslDecompiler.cs

@@ -216,7 +216,7 @@ namespace Ryujinx.Graphics.Gal.Shader
 
         private void PrintDeclOutAttributes()
         {
-            if (Decl.ShaderType == GalShaderType.Vertex)
+            if (Decl.ShaderType != GalShaderType.Fragment)
             {
                 SB.AppendLine("layout (location = " + GlslDecl.PositionOutAttrLocation + ") out vec4 " + GlslDecl.PositionOutAttrName + ";");
             }
@@ -337,7 +337,10 @@ namespace Ryujinx.Graphics.Gal.Shader
             if (Decl.ShaderType == GalShaderType.Vertex)
             {
                 SB.AppendLine(IdentationStr + "gl_Position.xy *= " + GlslDecl.FlipUniformName + ";");
+            }
 
+            if (Decl.ShaderType != GalShaderType.Fragment)
+            {
                 SB.AppendLine(IdentationStr + GlslDecl.PositionOutAttrName + " = gl_Position;");
                 SB.AppendLine(IdentationStr + GlslDecl.PositionOutAttrName + ".w = 1;");
             }
@@ -598,9 +601,6 @@ namespace Ryujinx.Graphics.Gal.Shader
         {
             switch (Op.Inst)
             {
-                case ShaderIrInst.Frcp:
-                    return true;
-
                 case ShaderIrInst.Ipa:
                 case ShaderIrInst.Texq:
                 case ShaderIrInst.Texs:
@@ -608,8 +608,7 @@ namespace Ryujinx.Graphics.Gal.Shader
                     return false;
             }
 
-            return Op.OperandB != null ||
-                   Op.OperandC != null;
+            return true;
         }
 
         private string GetName(ShaderIrOperCbuf Cbuf)
@@ -711,13 +710,13 @@ namespace Ryujinx.Graphics.Gal.Shader
             }
             else
             {
-                return Imm.Value.ToString(CultureInfo.InvariantCulture);
+                return GetIntConst(Imm.Value);
             }
         }
 
         private string GetValue(ShaderIrOperImmf Immf)
         {
-            return Immf.Value.ToString(CultureInfo.InvariantCulture);
+            return GetFloatConst(Immf.Value);
         }
 
         private string GetName(ShaderIrOperPred Pred)
@@ -1047,7 +1046,7 @@ namespace Ryujinx.Graphics.Gal.Shader
 
                             if (!float.IsNaN(Value) && !float.IsInfinity(Value))
                             {
-                                return Value.ToString(CultureInfo.InvariantCulture);
+                                return GetFloatConst(Value);
                             }
                         }
                         break;
@@ -1064,6 +1063,20 @@ namespace Ryujinx.Graphics.Gal.Shader
             return Expr;
         }
 
+        private static string GetIntConst(int Value)
+        {
+            string Expr = Value.ToString(CultureInfo.InvariantCulture);
+
+            return Value < 0 ? "(" + Expr + ")" : Expr;
+        }
+
+        private static string GetFloatConst(float Value)
+        {
+            string Expr = Value.ToString(CultureInfo.InvariantCulture);
+
+            return Value < 0 ? "(" + Expr + ")" : Expr;
+        }
+
         private static OperType GetDstNodeType(ShaderIrNode Node)
         {
             //Special case instructions with the result type different

Ryujinx.Graphics/Gal/ShaderDumper.cs

@@ -0,0 +1,96 @@
+using System;
+using System.IO;
+
+namespace Ryujinx.Graphics.Gal
+{
+    static class ShaderDumper
+    {
+        private static string RuntimeDir;
+
+        private static int DumpIndex = 1;
+
+        public static void Dump(IGalMemory Memory, long Position, GalShaderType Type, string ExtSuffix = "")
+        {
+            if (string.IsNullOrWhiteSpace(GraphicsConfig.ShadersDumpPath))
+            {
+                return;
+            }
+
+            string FileName = "Shader" + DumpIndex.ToString("d4") + "." + ShaderExtension(Type) + ExtSuffix + ".bin";
+
+            string FilePath = Path.Combine(DumpDir(), FileName);
+
+            DumpIndex++;
+
+            using (FileStream Output = File.Create(FilePath))
+            using (BinaryWriter Writer = new BinaryWriter(Output))
+            {
+                long Offset = 0;
+
+                ulong Instruction = 0;
+
+                //Dump until a NOP instruction is found
+                while ((Instruction >> 52 & 0xfff8) != 0x50b0)
+                {
+                    uint Word0 = (uint)Memory.ReadInt32(Position + Offset + 0);
+                    uint Word1 = (uint)Memory.ReadInt32(Position + Offset + 4);
+
+                    Instruction = Word0 | (ulong)Word1 << 32;
+
+                    //Zero instructions (other kind of NOP) stop immediatly,
+                    //this is to avoid two rows of zeroes
+                    if (Instruction == 0)
+                    {
+                        break;
+                    }
+
+                    Writer.Write(Instruction);
+
+                    Offset += 8;
+                }
+
+                //Align to meet nvdisasm requeriments
+                while (Offset % 0x20 != 0)
+                {
+                    Writer.Write(0);
+
+                    Offset += 4;
+                }
+            }
+        }
+
+        private static string DumpDir()
+        {
+            if (string.IsNullOrEmpty(RuntimeDir))
+            {
+                int Index = 1;
+
+                do
+                {
+                    RuntimeDir = Path.Combine(GraphicsConfig.ShadersDumpPath, "Dumps" + Index.ToString("d2"));
+
+                    Index++;
+                }
+                while (Directory.Exists(RuntimeDir));
+
+                Directory.CreateDirectory(RuntimeDir);
+            }
+
+            return RuntimeDir;
+        }
+
+        private static string ShaderExtension(GalShaderType Type)
+        {
+            switch (Type)
+            {
+                case GalShaderType.Vertex:         return "vert";
+                case GalShaderType.TessControl:    return "tesc";
+                case GalShaderType.TessEvaluation: return "tese";
+                case GalShaderType.Geometry:       return "geom";
+                case GalShaderType.Fragment:       return "frag";
+
+                default: throw new ArgumentException(nameof(Type));
+            }
+        }
+    }
+}

Ryujinx.Graphics/GraphicsConfig.cs

@@ -0,0 +1,4 @@
+public static class GraphicsConfig
+{
+    public static string ShadersDumpPath;
+}

Ryujinx.HLE/Gpu/Engines/NvGpuEngine3d.cs

@@ -73,6 +73,8 @@ namespace Ryujinx.HLE.Gpu.Engines
 
         private void VertexEndGl(NvGpuVmm Vmm, NvGpuPBEntry PBEntry)
         {
+            LockCaches();
+
             SetFrameBuffer(Vmm, 0);
 
             long[] Keys = UploadShaders(Vmm);
@@ -90,6 +92,20 @@ namespace Ryujinx.HLE.Gpu.Engines
             UploadTextures(Vmm, Keys);
             UploadUniforms(Vmm);
             UploadVertexArrays(Vmm);
+
+            UnlockCaches();
+        }
+
+        private void LockCaches()
+        {
+            Gpu.Renderer.Rasterizer.LockCaches();
+            Gpu.Renderer.Texture.LockCache();
+        }
+
+        private void UnlockCaches()
+        {
+            Gpu.Renderer.Rasterizer.UnlockCaches();
+            Gpu.Renderer.Texture.UnlockCache();
         }
 
         private void ClearBuffers(NvGpuVmm Vmm, NvGpuPBEntry PBEntry)
@@ -156,6 +172,8 @@ namespace Ryujinx.HLE.Gpu.Engines
 
             for (; Index < 6; Index++)
             {
+                GalShaderType Type = GetTypeFromProgram(Index);
+
                 int Control = ReadRegister(NvGpuEngine3dReg.ShaderNControl + Index * 0x10);
                 int Offset  = ReadRegister(NvGpuEngine3dReg.ShaderNOffset  + Index * 0x10);
 
@@ -164,16 +182,16 @@ namespace Ryujinx.HLE.Gpu.Engines
 
                 if (!Enable)
                 {
+                    Gpu.Renderer.Shader.Unbind(Type);
+
                     continue;
                 }
 
                 long Key = BasePosition + (uint)Offset;
 
-                GalShaderType ShaderType = GetTypeFromProgram(Index);
+                Keys[(int)Type] = Key;
 
-                Keys[(int)ShaderType] = Key;
-
-                Gpu.Renderer.Shader.Create(Vmm, Key, ShaderType);
+                Gpu.Renderer.Shader.Create(Vmm, Key, Type);
                 Gpu.Renderer.Shader.Bind(Key);
             }
 
@@ -464,19 +482,17 @@ namespace Ryujinx.HLE.Gpu.Engines
 
             GalTextureSampler Sampler = TextureFactory.MakeSampler(Gpu, Vmm, TscPosition);
 
-            long TextureAddress = Vmm.ReadInt64(TicPosition + 4) & 0xffffffffffff;
+            long Key = Vmm.ReadInt64(TicPosition + 4) & 0xffffffffffff;
 
-            long Key = TextureAddress;
+            Key = Vmm.GetPhysicalAddress(Key);
 
-            TextureAddress = Vmm.GetPhysicalAddress(TextureAddress);
-
-            if (IsFrameBufferPosition(TextureAddress))
+            if (IsFrameBufferPosition(Key))
             {
                 //This texture is a frame buffer texture,
                 //we shouldn't read anything from memory and bind
                 //the frame buffer texture instead, since we're not
                 //really writing anything to memory.
-                Gpu.Renderer.FrameBuffer.BindTexture(TextureAddress, TexIndex);
+                Gpu.Renderer.FrameBuffer.BindTexture(Key, TexIndex);
             }
             else
             {
@@ -544,6 +560,8 @@ namespace Ryujinx.HLE.Gpu.Engines
         {
             long IndexPosition = MakeInt64From2xInt32(NvGpuEngine3dReg.IndexArrayAddress);
 
+            long IboKey = Vmm.GetPhysicalAddress(IndexPosition);
+
             int IndexEntryFmt = ReadRegister(NvGpuEngine3dReg.IndexArrayFormat);
             int IndexFirst    = ReadRegister(NvGpuEngine3dReg.IndexBatchFirst);
             int IndexCount    = ReadRegister(NvGpuEngine3dReg.IndexBatchCount);
@@ -561,16 +579,16 @@ namespace Ryujinx.HLE.Gpu.Engines
             {
                 int IbSize = IndexCount * IndexEntrySize;
 
-                bool IboCached = Gpu.Renderer.Rasterizer.IsIboCached(IndexPosition, (uint)IbSize);
+                bool IboCached = Gpu.Renderer.Rasterizer.IsIboCached(IboKey, (uint)IbSize);
 
-                if (!IboCached || Vmm.IsRegionModified(IndexPosition, (uint)IbSize, NvGpuBufferType.Index))
+                if (!IboCached || Vmm.IsRegionModified(IboKey, (uint)IbSize, NvGpuBufferType.Index))
                 {
                     byte[] Data = Vmm.ReadBytes(IndexPosition, (uint)IbSize);
 
-                    Gpu.Renderer.Rasterizer.CreateIbo(IndexPosition, Data);
+                    Gpu.Renderer.Rasterizer.CreateIbo(IboKey, Data);
                 }
 
-                Gpu.Renderer.Rasterizer.SetIndexArray(IndexPosition, IbSize, IndexFormat);
+                Gpu.Renderer.Rasterizer.SetIndexArray(IbSize, IndexFormat);
             }
 
             List<GalVertexAttrib>[] Attribs = new List<GalVertexAttrib>[32];
@@ -619,20 +637,22 @@ namespace Ryujinx.HLE.Gpu.Engines
                     continue;
                 }
 
+                long VboKey = Vmm.GetPhysicalAddress(VertexPosition);
+
                 int Stride = Control & 0xfff;
 
                 long VbSize = (VertexEndPos - VertexPosition) + 1;
 
-                bool VboCached = Gpu.Renderer.Rasterizer.IsVboCached(VertexPosition, VbSize);
+                bool VboCached = Gpu.Renderer.Rasterizer.IsVboCached(VboKey, VbSize);
 
-                if (!VboCached || Vmm.IsRegionModified(VertexPosition, VbSize, NvGpuBufferType.Vertex))
+                if (!VboCached || Vmm.IsRegionModified(VboKey, VbSize, NvGpuBufferType.Vertex))
                 {
                     byte[] Data = Vmm.ReadBytes(VertexPosition, VbSize);
 
-                    Gpu.Renderer.Rasterizer.CreateVbo(VertexPosition, Data);
+                    Gpu.Renderer.Rasterizer.CreateVbo(VboKey, Data);
                 }
 
-                Gpu.Renderer.Rasterizer.SetVertexArray(Index, Stride, VertexPosition, Attribs[Index].ToArray());
+                Gpu.Renderer.Rasterizer.SetVertexArray(Stride, VboKey, Attribs[Index].ToArray());
             }
 
             GalPrimitiveType PrimType = (GalPrimitiveType)(PrimCtrl & 0xffff);
@@ -641,7 +661,7 @@ namespace Ryujinx.HLE.Gpu.Engines
             {
                 int VertexBase = ReadRegister(NvGpuEngine3dReg.VertexArrayElemBase);
 
-                Gpu.Renderer.Rasterizer.DrawElements(IndexPosition, IndexFirst, VertexBase, PrimType);
+                Gpu.Renderer.Rasterizer.DrawElements(IboKey, IndexFirst, VertexBase, PrimType);
             }
             else
             {

Ryujinx.HLE/Gpu/Engines/NvGpuFifo.cs

@@ -1,5 +1,6 @@
 using Ryujinx.HLE.Gpu.Memory;
 using System.Collections.Concurrent;
+using System.Threading;
 
 namespace Ryujinx.HLE.Gpu.Engines
 {
@@ -18,6 +19,8 @@ namespace Ryujinx.HLE.Gpu.Engines
 
         private NvGpuEngine[] SubChannels;
 
+        public AutoResetEvent Event { get; private set; }
+
         private struct CachedMacro
         {
             public int Position { get; private set; }
@@ -60,6 +63,8 @@ namespace Ryujinx.HLE.Gpu.Engines
             Macros = new CachedMacro[MacrosCount];
 
             Mme = new int[MmeWords];
+
+            Event = new AutoResetEvent(false);
         }
 
         public void PushBuffer(NvGpuVmm Vmm, NvGpuPBEntry[] Buffer)
@@ -68,6 +73,8 @@ namespace Ryujinx.HLE.Gpu.Engines
             {
                 BufferQueue.Enqueue((Vmm, PBEntry));
             }
+
+            Event.Set();
         }
 
         public void DispatchCalls()

Ryujinx.HLE/Gpu/Memory/NvGpuBufferType.cs

@@ -4,6 +4,7 @@ namespace Ryujinx.HLE.Gpu.Memory
     {
         Index,
         Vertex,
-        Texture
+        Texture,
+        Count
     }
 }

Ryujinx.HLE/Gpu/Memory/NvGpuVmm.cs

@@ -274,11 +274,9 @@ namespace Ryujinx.HLE.Gpu.Memory
             PageTable[L0][L1] = TgtAddr;
         }
 
-        public bool IsRegionModified(long Position, long Size, NvGpuBufferType BufferType)
+        public bool IsRegionModified(long PA, long Size, NvGpuBufferType BufferType)
         {
-            long PA = GetPhysicalAddress(Position);
-
-            return Cache.IsRegionModified(Memory, BufferType, Position, PA, Size);
+            return Cache.IsRegionModified(Memory, BufferType, PA, Size);
         }
 
         public byte ReadByte(long Position)

Ryujinx.HLE/Gpu/Memory/NvGpuVmmCache.cs

@@ -11,43 +11,53 @@ namespace Ryujinx.HLE.Gpu.Memory
 
         private class CachedPage
         {
-            private List<(long Start, long End)> Regions;
+            private struct Range
+            {
+                public long Start;
+                public long End;
+
+                public Range(long Start, long End)
+                {
+                    this.Start = Start;
+                    this.End   = End;
+                }
+            }
+
+            private List<Range>[] Regions;
 
             public LinkedListNode<long> Node { get; set; }
 
-            public int Count => Regions.Count;
-
             public int Timestamp { get; private set; }
 
-            public long PABase { get; private set; }
-
-            public NvGpuBufferType BufferType { get; private set; }
-
-            public CachedPage(long PABase, NvGpuBufferType BufferType)
+            public CachedPage()
             {
-                this.PABase     = PABase;
-                this.BufferType = BufferType;
+                Regions = new List<Range>[(int)NvGpuBufferType.Count];
 
-                Regions = new List<(long, long)>();
+                for (int Index = 0; Index < Regions.Length; Index++)
+                {
+                    Regions[Index] = new List<Range>();
+                }
             }
 
-            public bool AddRange(long Start, long End)
+            public bool AddRange(long Start, long End, NvGpuBufferType BufferType)
             {
-                for (int Index = 0; Index < Regions.Count; Index++)
-                {
-                    (long RgStart, long RgEnd) = Regions[Index];
+                List<Range> BtRegions = Regions[(int)BufferType];
 
-                    if (Start >= RgStart && End <= RgEnd)
+                for (int Index = 0; Index < BtRegions.Count; Index++)
+                {
+                    Range Rg = BtRegions[Index];
+
+                    if (Start >= Rg.Start && End <= Rg.End)
                     {
                         return false;
                     }
 
-                    if (Start <= RgEnd && RgStart <= End)
+                    if (Start <= Rg.End && Rg.Start <= End)
                     {
-                        long MinStart = Math.Min(RgStart, Start);
-                        long MaxEnd   = Math.Max(RgEnd,   End);
+                        long MinStart = Math.Min(Rg.Start, Start);
+                        long MaxEnd   = Math.Max(Rg.End,   End);
 
-                        Regions[Index] = (MinStart, MaxEnd);
+                        BtRegions[Index] = new Range(MinStart, MaxEnd);
 
                         Timestamp = Environment.TickCount;
 
@@ -55,12 +65,24 @@ namespace Ryujinx.HLE.Gpu.Memory
                     }
                 }
 
-                Regions.Add((Start, End));
+                BtRegions.Add(new Range(Start, End));
 
                 Timestamp = Environment.TickCount;
 
                 return true;
             }
+
+            public int GetTotalCount()
+            {
+                int Count = 0;
+
+                for (int Index = 0; Index < Regions.Length; Index++)
+                {
+                    Count += Regions[Index].Count;
+                }
+
+                return Count;
+            }
         }
 
         private Dictionary<long, CachedPage> Cache;
@@ -76,71 +98,61 @@ namespace Ryujinx.HLE.Gpu.Memory
             SortedCache = new LinkedList<long>();
         }
 
-        public bool IsRegionModified(
-            AMemory         Memory,
-            NvGpuBufferType BufferType,
-            long            VA,
-            long            PA,
-            long            Size)
+        public bool IsRegionModified(AMemory Memory, NvGpuBufferType BufferType, long PA, long Size)
         {
+            bool[] Modified = Memory.IsRegionModified(PA, Size);
+
+            if (Modified == null)
+            {
+                return true;
+            }
+
             ClearCachedPagesIfNeeded();
 
             long PageSize = Memory.GetHostPageSize();
 
             long Mask = PageSize - 1;
 
-            long VAEnd = VA + Size;
             long PAEnd = PA + Size;
 
             bool RegMod = false;
 
-            while (VA < VAEnd)
-            {
-                long Key    = VA & ~Mask;
-                long PABase = PA & ~Mask;
+            int Index = 0;
+
+            while (PA < PAEnd)
+            {
+                long Key = PA & ~Mask;
 
-                long VAPgEnd = Math.Min((VA + PageSize) & ~Mask, VAEnd);
                 long PAPgEnd = Math.Min((PA + PageSize) & ~Mask, PAEnd);
 
                 bool IsCached = Cache.TryGetValue(Key, out CachedPage Cp);
 
-                bool PgReset = false;
-
-                if (!IsCached)
+                if (IsCached)
                 {
-                    Cp = new CachedPage(PABase, BufferType);
+                    CpCount -= Cp.GetTotalCount();
 
-                    Cache.Add(Key, Cp);
+                    SortedCache.Remove(Cp.Node);
                 }
                 else
                 {
-                    CpCount -= Cp.Count;
+                    Cp = new CachedPage();
 
-                    SortedCache.Remove(Cp.Node);
-
-                    if (Cp.PABase     != PABase ||
-                        Cp.BufferType != BufferType)
-                    {
-                        PgReset = true;
-                    }
+                    Cache.Add(Key, Cp);
                 }
 
-                PgReset |= Memory.IsRegionModified(PA, PAPgEnd - PA) && IsCached;
-
-                if (PgReset)
+                if (Modified[Index++] && IsCached)
                 {
-                    Cp = new CachedPage(PABase, BufferType);
+                    Cp = new CachedPage();
 
                     Cache[Key] = Cp;
                 }
 
                 Cp.Node = SortedCache.AddLast(Key);
 
-                RegMod |= Cp.AddRange(VA, VAPgEnd);
+                RegMod |= Cp.AddRange(PA, PAPgEnd, BufferType);
 
-                CpCount += Cp.Count;
+                CpCount += Cp.GetTotalCount();
 
-                VA = VAPgEnd;
                 PA = PAPgEnd;
             }
 
@@ -169,7 +181,7 @@ namespace Ryujinx.HLE.Gpu.Memory
 
                 Cache.Remove(Key);
 
-                CpCount -= Cp.Count;
+                CpCount -= Cp.GetTotalCount();
 
                 TimeDelta = RingDelta(Cp.Timestamp, Timestamp);
             }

Ryujinx.HLE/Gpu/Texture/TextureHelper.cs

@@ -28,23 +28,30 @@ namespace Ryujinx.HLE.Gpu.Texture
         {
             switch (Texture.Format)
             {
-                case GalTextureFormat.R32G32B32A32: return Texture.Width * Texture.Height * 16;
-                case GalTextureFormat.R16G16B16A16: return Texture.Width * Texture.Height * 8;
-                case GalTextureFormat.A8B8G8R8:     return Texture.Width * Texture.Height * 4;
-                case GalTextureFormat.R32:          return Texture.Width * Texture.Height * 4;
-                case GalTextureFormat.A1B5G5R5:     return Texture.Width * Texture.Height * 2;
-                case GalTextureFormat.B5G6R5:       return Texture.Width * Texture.Height * 2;
-                case GalTextureFormat.G8R8:         return Texture.Width * Texture.Height * 2;
-                case GalTextureFormat.R16:          return Texture.Width * Texture.Height * 2;
-                case GalTextureFormat.R8:           return Texture.Width * Texture.Height;
+                case GalTextureFormat.R32G32B32A32:
+                    return Texture.Width * Texture.Height * 16;
+
+                case GalTextureFormat.R16G16B16A16:
+                    return Texture.Width * Texture.Height * 8;
+
+                case GalTextureFormat.A8B8G8R8:
+                case GalTextureFormat.R32:
+                case GalTextureFormat.ZF32:
+                    return Texture.Width * Texture.Height * 4;
+
+                case GalTextureFormat.A1B5G5R5:
+                case GalTextureFormat.B5G6R5:
+                case GalTextureFormat.G8R8:
+                case GalTextureFormat.R16:
+                    return Texture.Width * Texture.Height * 2;
+
+                case GalTextureFormat.R8:
+                    return Texture.Width * Texture.Height;
 
                 case GalTextureFormat.BC1:
                 case GalTextureFormat.BC4:
                 {
-                    int W = (Texture.Width  + 3) / 4;
-                    int H = (Texture.Height + 3) / 4;
-
-                    return W * H * 8;
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 4, 4, 8);
                 }
 
                 case GalTextureFormat.BC7U:
@@ -53,16 +60,86 @@ namespace Ryujinx.HLE.Gpu.Texture
                 case GalTextureFormat.BC5:
                 case GalTextureFormat.Astc2D4x4:
                 {
-                    int W = (Texture.Width  + 3) / 4;
-                    int H = (Texture.Height + 3) / 4;
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 4, 4, 16);
+                }
 
-                    return W * H * 16;
+                case GalTextureFormat.Astc2D5x5:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 5, 5, 16);
+                }
+
+                case GalTextureFormat.Astc2D6x6:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 6, 6, 16);
+                }
+
+                case GalTextureFormat.Astc2D8x8:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 8, 8, 16);
+                }
+
+                case GalTextureFormat.Astc2D10x10:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 10, 10, 16);
+                }
+
+                case GalTextureFormat.Astc2D12x12:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 12, 12, 16);
+                }
+
+                case GalTextureFormat.Astc2D5x4:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 5, 4, 16);
+                }
+
+                case GalTextureFormat.Astc2D6x5:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 6, 5, 16);
+                }
+
+                case GalTextureFormat.Astc2D8x6:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 8, 6, 16);
+                }
+
+                case GalTextureFormat.Astc2D10x8:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 10, 8, 16);
+                }
+
+                case GalTextureFormat.Astc2D12x10:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 12, 10, 16);
+                }
+
+                case GalTextureFormat.Astc2D8x5:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 8, 5, 16);
+                }
+
+                case GalTextureFormat.Astc2D10x5:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 10, 5, 16);
+                }
+
+                case GalTextureFormat.Astc2D10x6:
+                {
+                    return CompressedTextureSize(Texture.Width, Texture.Height, 10, 6, 16);
                 }
             }
 
             throw new NotImplementedException(Texture.Format.ToString());
         }
 
+        public static int CompressedTextureSize(int TextureWidth, int TextureHeight, int BlockWidth, int BlockHeight, int Bpb)
+        {
+            int W = (TextureWidth  + (BlockWidth - 1)) / BlockWidth;
+            int H = (TextureHeight + (BlockHeight - 1)) / BlockHeight;
+
+            return W * H * Bpb;
+        }
+
         public static (AMemory Memory, long Position) GetMemoryAndPosition(
             IAMemory Memory,
             long     Position)

Ryujinx.HLE/Gpu/Texture/TextureReader.cs

@@ -10,22 +10,36 @@ namespace Ryujinx.HLE.Gpu.Texture
         {
             switch (Texture.Format)
             {
-                case GalTextureFormat.R32G32B32A32: return Read16Bpp   (Memory, Texture);
-                case GalTextureFormat.R16G16B16A16: return Read8Bpp    (Memory, Texture);
-                case GalTextureFormat.A8B8G8R8:     return Read4Bpp    (Memory, Texture);
-                case GalTextureFormat.R32:          return Read4Bpp    (Memory, Texture);
-                case GalTextureFormat.A1B5G5R5:     return Read5551    (Memory, Texture);
-                case GalTextureFormat.B5G6R5:       return Read565     (Memory, Texture);
-                case GalTextureFormat.G8R8:         return Read2Bpp    (Memory, Texture);
-                case GalTextureFormat.R16:          return Read2Bpp    (Memory, Texture);
-                case GalTextureFormat.R8:           return Read1Bpp    (Memory, Texture);
-                case GalTextureFormat.BC7U:         return Read16Bpt4x4(Memory, Texture);
-                case GalTextureFormat.BC1:          return Read8Bpt4x4 (Memory, Texture);
-                case GalTextureFormat.BC2:          return Read16Bpt4x4(Memory, Texture);
-                case GalTextureFormat.BC3:          return Read16Bpt4x4(Memory, Texture);
-                case GalTextureFormat.BC4:          return Read8Bpt4x4 (Memory, Texture);
-                case GalTextureFormat.BC5:          return Read16Bpt4x4(Memory, Texture);
-                case GalTextureFormat.Astc2D4x4:    return Read16Bpt4x4(Memory, Texture);
+                case GalTextureFormat.R32G32B32A32: return Read16Bpp                 (Memory, Texture);
+                case GalTextureFormat.R16G16B16A16: return Read8Bpp                  (Memory, Texture);
+                case GalTextureFormat.A8B8G8R8:     return Read4Bpp                  (Memory, Texture);
+                case GalTextureFormat.R32:          return Read4Bpp                  (Memory, Texture);
+                case GalTextureFormat.A1B5G5R5:     return Read5551                  (Memory, Texture);
+                case GalTextureFormat.B5G6R5:       return Read565                   (Memory, Texture);
+                case GalTextureFormat.G8R8:         return Read2Bpp                  (Memory, Texture);
+                case GalTextureFormat.R16:          return Read2Bpp                  (Memory, Texture);
+                case GalTextureFormat.R8:           return Read1Bpp                  (Memory, Texture);
+                case GalTextureFormat.BC7U:         return Read16BptCompressedTexture(Memory, Texture, 4, 4);
+                case GalTextureFormat.BC1:          return Read8Bpt4x4               (Memory, Texture);
+                case GalTextureFormat.BC2:          return Read16BptCompressedTexture(Memory, Texture, 4, 4);
+                case GalTextureFormat.BC3:          return Read16BptCompressedTexture(Memory, Texture, 4, 4);
+                case GalTextureFormat.BC4:          return Read8Bpt4x4               (Memory, Texture);
+                case GalTextureFormat.BC5:          return Read16BptCompressedTexture(Memory, Texture, 4, 4);
+                case GalTextureFormat.ZF32:         return Read4Bpp                  (Memory, Texture);
+                case GalTextureFormat.Astc2D4x4:    return Read16BptCompressedTexture(Memory, Texture, 4, 4);
+                case GalTextureFormat.Astc2D5x5:    return Read16BptCompressedTexture(Memory, Texture, 5, 5);
+                case GalTextureFormat.Astc2D6x6:    return Read16BptCompressedTexture(Memory, Texture, 6, 6);
+                case GalTextureFormat.Astc2D8x8:    return Read16BptCompressedTexture(Memory, Texture, 8, 8);
+                case GalTextureFormat.Astc2D10x10:  return Read16BptCompressedTexture(Memory, Texture, 10, 10);
+                case GalTextureFormat.Astc2D12x12:  return Read16BptCompressedTexture(Memory, Texture, 12, 12);
+                case GalTextureFormat.Astc2D5x4:    return Read16BptCompressedTexture(Memory, Texture, 5, 4);
+                case GalTextureFormat.Astc2D6x5:    return Read16BptCompressedTexture(Memory, Texture, 6, 5);
+                case GalTextureFormat.Astc2D8x6:    return Read16BptCompressedTexture(Memory, Texture, 8, 6);
+                case GalTextureFormat.Astc2D10x8:   return Read16BptCompressedTexture(Memory, Texture, 10, 8);
+                case GalTextureFormat.Astc2D12x10:  return Read16BptCompressedTexture(Memory, Texture, 12, 10);
+                case GalTextureFormat.Astc2D8x5:    return Read16BptCompressedTexture(Memory, Texture, 8, 5);
+                case GalTextureFormat.Astc2D10x5:   return Read16BptCompressedTexture(Memory, Texture, 10, 5);
+                case GalTextureFormat.Astc2D10x6:   return Read16BptCompressedTexture(Memory, Texture, 10, 6);
             }
 
             throw new NotImplementedException(Texture.Format.ToString());
@@ -306,10 +320,10 @@ namespace Ryujinx.HLE.Gpu.Texture
             return Output;
         }
 
-        private unsafe static byte[] Read16Bpt4x4(IAMemory Memory, TextureInfo Texture)
+        private unsafe static byte[] Read16BptCompressedTexture(IAMemory Memory, TextureInfo Texture, int BlockWidth, int BlockHeight)
         {
-            int Width  = (Texture.Width  + 3) / 4;
-            int Height = (Texture.Height + 3) / 4;
+            int Width  = (Texture.Width  + (BlockWidth - 1)) / BlockWidth;
+            int Height = (Texture.Height + (BlockHeight - 1)) / BlockHeight;
 
             byte[] Output = new byte[Width * Height * 16];
 

Ryujinx.HLE/OsHle/Kernel/SvcThread.cs

@@ -87,7 +87,7 @@ namespace Ryujinx.HLE.OsHle.Kernel
 
             KThread CurrThread = Process.GetThread(ThreadState.Tpidr);
 
-            if (TimeoutNs == 0)
+            if (TimeoutNs == 0 || TimeoutNs == ulong.MaxValue)
             {
                 Process.Scheduler.Yield(CurrThread);
             }

Ryujinx.HLE/OsHle/Services/Aud/AudErr.cs

@@ -3,6 +3,7 @@ namespace Ryujinx.HLE.OsHle.Services.Aud
     static class AudErr
     {
         public const int DeviceNotFound        = 1;
+        public const int UnsupportedRevision   = 2;
         public const int UnsupportedSampleRate = 3;
     }
 }

Ryujinx.HLE/OsHle/Services/Aud/AudioOut/AudioOutData.cs

@@ -1,6 +1,6 @@
 using System.Runtime.InteropServices;
 
-namespace Ryujinx.HLE.OsHle.Services.Aud
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioOut
 {
     [StructLayout(LayoutKind.Sequential)]
     struct AudioOutData

Ryujinx.HLE/OsHle/Services/Aud/AudioOut/IAudioOut.cs

@@ -1,12 +1,11 @@
 using ChocolArm64.Memory;
 using Ryujinx.Audio;
-using Ryujinx.HLE.Logging;
 using Ryujinx.HLE.OsHle.Handles;
 using Ryujinx.HLE.OsHle.Ipc;
 using System;
 using System.Collections.Generic;
 
-namespace Ryujinx.HLE.OsHle.Services.Aud
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioOut
 {
     class IAudioOut : IpcService, IDisposable
     {

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/AudioConsts.cs

@@ -0,0 +1,8 @@
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    static class AudioConsts
+    {
+        public const int HostSampleRate    = 48000;
+        public const int HostChannelsCount = 2;
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/BehaviorIn.cs

@@ -0,0 +1,11 @@
+using System.Runtime.InteropServices;
+
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    [StructLayout(LayoutKind.Sequential, Size = 0x10, Pack = 4)]
+    struct BehaviorIn
+    {
+        public long Unknown0;
+        public long Unknown8;
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/BiquadFilter.cs

@@ -0,0 +1,16 @@
+using System.Runtime.InteropServices;
+
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    [StructLayout(LayoutKind.Sequential, Size = 0xc, Pack = 1)]
+    struct BiquadFilter
+    {
+        public byte  Enable;
+        public byte  Padding;
+        public short B0;
+        public short B1;
+        public short B2;
+        public short A1;
+        public short A2;
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/IAudioRenderer.cs

@@ -0,0 +1,316 @@
+using ChocolArm64.Memory;
+using Ryujinx.Audio;
+using Ryujinx.Audio.Adpcm;
+using Ryujinx.HLE.Logging;
+using Ryujinx.HLE.OsHle.Handles;
+using Ryujinx.HLE.OsHle.Ipc;
+using Ryujinx.HLE.OsHle.Utilities;
+using System;
+using System.Collections.Generic;
+using System.Runtime.InteropServices;
+
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    class IAudioRenderer : IpcService, IDisposable
+    {
+        //This is the amount of samples that are going to be appended
+        //each time that RequestUpdateAudioRenderer is called. Ideally,
+        //this value shouldn't be neither too small (to avoid the player
+        //starving due to running out of samples) or too large (to avoid
+        //high latency).
+        private const int MixBufferSamplesCount = 960;
+
+        private Dictionary<int, ServiceProcessRequest> m_Commands;
+
+        public override IReadOnlyDictionary<int, ServiceProcessRequest> Commands => m_Commands;
+
+        private KEvent UpdateEvent;
+
+        private AMemory Memory;
+
+        private IAalOutput AudioOut;
+
+        private AudioRendererParameter Params;
+
+        private MemoryPoolContext[] MemoryPools;
+
+        private VoiceContext[] Voices;
+
+        private int Track;
+
+        public IAudioRenderer(AMemory Memory, IAalOutput AudioOut, AudioRendererParameter Params)
+        {
+            m_Commands = new Dictionary<int, ServiceProcessRequest>()
+            {
+                { 4, RequestUpdateAudioRenderer },
+                { 5, StartAudioRenderer         },
+                { 6, StopAudioRenderer          },
+                { 7, QuerySystemEvent           }
+            };
+
+            UpdateEvent = new KEvent();
+
+            this.Memory   = Memory;
+            this.AudioOut = AudioOut;
+            this.Params   = Params;
+
+            Track = AudioOut.OpenTrack(
+                AudioConsts.HostSampleRate,
+                AudioConsts.HostChannelsCount,
+                AudioCallback);
+
+            MemoryPools = CreateArray<MemoryPoolContext>(Params.EffectCount + Params.VoiceCount * 4);
+
+            Voices = CreateArray<VoiceContext>(Params.VoiceCount);
+
+            InitializeAudioOut();
+        }
+
+        private void AudioCallback()
+        {
+            UpdateEvent.WaitEvent.Set();
+        }
+
+        private static T[] CreateArray<T>(int Size) where T : new()
+        {
+            T[] Output = new T[Size];
+
+            for (int Index = 0; Index < Size; Index++)
+            {
+                Output[Index] = new T();
+            }
+
+            return Output;
+        }
+
+        private void InitializeAudioOut()
+        {
+            AppendMixedBuffer(0);
+            AppendMixedBuffer(1);
+            AppendMixedBuffer(2);
+
+            AudioOut.Start(Track);
+        }
+
+        public long RequestUpdateAudioRenderer(ServiceCtx Context)
+        {
+            long OutputPosition = Context.Request.ReceiveBuff[0].Position;
+            long OutputSize     = Context.Request.ReceiveBuff[0].Size;
+
+            AMemoryHelper.FillWithZeros(Context.Memory, OutputPosition, (int)OutputSize);
+
+            long InputPosition = Context.Request.SendBuff[0].Position;
+
+            StructReader Reader = new StructReader(Context.Memory, InputPosition);
+            StructWriter Writer = new StructWriter(Context.Memory, OutputPosition);
+
+            UpdateDataHeader InputHeader = Reader.Read<UpdateDataHeader>();
+
+            Reader.Read<BehaviorIn>(InputHeader.BehaviorSize);
+
+            MemoryPoolIn[] MemoryPoolsIn = Reader.Read<MemoryPoolIn>(InputHeader.MemoryPoolSize);
+
+            for (int Index = 0; Index < MemoryPoolsIn.Length; Index++)
+            {
+                MemoryPoolIn MemoryPool = MemoryPoolsIn[Index];
+
+                if (MemoryPool.State == MemoryPoolState.RequestAttach)
+                {
+                    MemoryPools[Index].OutStatus.State = MemoryPoolState.Attached;
+                }
+                else if (MemoryPool.State == MemoryPoolState.RequestDetach)
+                {
+                    MemoryPools[Index].OutStatus.State = MemoryPoolState.Detached;
+                }
+            }
+
+            Reader.Read<VoiceChannelResourceIn>(InputHeader.VoiceResourceSize);
+
+            VoiceIn[] VoicesIn = Reader.Read<VoiceIn>(InputHeader.VoiceSize);
+
+            for (int Index = 0; Index < VoicesIn.Length; Index++)
+            {
+                VoiceIn Voice = VoicesIn[Index];
+
+                VoiceContext VoiceCtx = Voices[Index];
+
+                VoiceCtx.SetAcquireState(Voice.Acquired != 0);
+
+                if (Voice.Acquired == 0)
+                {
+                    continue;
+                }
+
+                if (Voice.FirstUpdate != 0)
+                {
+                    VoiceCtx.AdpcmCtx = GetAdpcmDecoderContext(
+                        Voice.AdpcmCoeffsPosition,
+                        Voice.AdpcmCoeffsSize);
+
+                    VoiceCtx.SampleFormat  = Voice.SampleFormat;
+                    VoiceCtx.SampleRate    = Voice.SampleRate;
+                    VoiceCtx.ChannelsCount = Voice.ChannelsCount;
+
+                    VoiceCtx.SetBufferIndex(Voice.BaseWaveBufferIndex);
+                }
+
+                VoiceCtx.WaveBuffers[0] = Voice.WaveBuffer0;
+                VoiceCtx.WaveBuffers[1] = Voice.WaveBuffer1;
+                VoiceCtx.WaveBuffers[2] = Voice.WaveBuffer2;
+                VoiceCtx.WaveBuffers[3] = Voice.WaveBuffer3;
+                VoiceCtx.Volume         = Voice.Volume;
+                VoiceCtx.PlayState      = Voice.PlayState;
+            }
+
+            UpdateAudio();
+
+            UpdateDataHeader OutputHeader = new UpdateDataHeader();
+
+            int UpdateHeaderSize = Marshal.SizeOf<UpdateDataHeader>();
+
+            OutputHeader.Revision               = IAudioRendererManager.RevMagic;
+            OutputHeader.BehaviorSize           = 0xb0;
+            OutputHeader.MemoryPoolSize         = (Params.EffectCount + Params.VoiceCount * 4) * 0x10;
+            OutputHeader.VoiceSize              = Params.VoiceCount  * 0x10;
+            OutputHeader.EffectSize             = Params.EffectCount * 0x10;
+            OutputHeader.SinkSize               = Params.SinkCount   * 0x20;
+            OutputHeader.PerformanceManagerSize = 0x10;
+            OutputHeader.TotalSize              = UpdateHeaderSize             +
+                                                  OutputHeader.BehaviorSize    +
+                                                  OutputHeader.MemoryPoolSize +
+                                                  OutputHeader.VoiceSize      +
+                                                  OutputHeader.EffectSize     +
+                                                  OutputHeader.SinkSize       +
+                                                  OutputHeader.PerformanceManagerSize;
+
+            Writer.Write(OutputHeader);
+
+            foreach (MemoryPoolContext MemoryPool in MemoryPools)
+            {
+                Writer.Write(MemoryPool.OutStatus);
+            }
+
+            foreach (VoiceContext Voice in Voices)
+            {
+                Writer.Write(Voice.OutStatus);
+            }
+
+            return 0;
+        }
+
+        public long StartAudioRenderer(ServiceCtx Context)
+        {
+            Context.Ns.Log.PrintStub(LogClass.ServiceAudio, "Stubbed.");
+
+            return 0;
+        }
+
+        public long StopAudioRenderer(ServiceCtx Context)
+        {
+            Context.Ns.Log.PrintStub(LogClass.ServiceAudio, "Stubbed.");
+
+            return 0;
+        }
+
+        public long QuerySystemEvent(ServiceCtx Context)
+        {
+            int Handle = Context.Process.HandleTable.OpenHandle(UpdateEvent);
+
+            Context.Response.HandleDesc = IpcHandleDesc.MakeCopy(Handle);
+
+            return 0;
+        }
+
+        private AdpcmDecoderContext GetAdpcmDecoderContext(long Position, long Size)
+        {
+            if (Size == 0)
+            {
+                return null;
+            }
+
+            AdpcmDecoderContext Context = new AdpcmDecoderContext();
+
+            Context.Coefficients = new short[Size >> 1];
+
+            for (int Offset = 0; Offset < Size; Offset += 2)
+            {
+                Context.Coefficients[Offset >> 1] = Memory.ReadInt16(Position + Offset);
+            }
+
+            return Context;
+        }
+
+        private void UpdateAudio()
+        {
+            long[] Released = AudioOut.GetReleasedBuffers(Track, 2);
+
+            for (int Index = 0; Index < Released.Length; Index++)
+            {
+                AppendMixedBuffer(Released[Index]);
+            }
+        }
+
+        private void AppendMixedBuffer(long Tag)
+        {
+            int[] MixBuffer = new int[MixBufferSamplesCount * AudioConsts.HostChannelsCount];
+
+            foreach (VoiceContext Voice in Voices)
+            {
+                if (!Voice.Playing)
+                {
+                    continue;
+                }
+
+                int OutOffset = 0;
+
+                int PendingSamples = MixBufferSamplesCount;
+
+                while (PendingSamples > 0)
+                {
+                    int[] Samples = Voice.GetBufferData(Memory, PendingSamples, out int ReturnedSamples);
+
+                    if (ReturnedSamples == 0)
+                    {
+                        break;
+                    }
+
+                    PendingSamples -= ReturnedSamples;
+
+                    for (int Offset = 0; Offset < Samples.Length; Offset++)
+                    {
+                        int Sample = (int)(Samples[Offset] * Voice.Volume);
+
+                        MixBuffer[OutOffset++] += Sample;
+                    }
+                }
+            }
+
+            AudioOut.AppendBuffer(Track, Tag, GetFinalBuffer(MixBuffer));
+        }
+
+        private static short[] GetFinalBuffer(int[] Buffer)
+        {
+            short[] Output = new short[Buffer.Length];
+
+            for (int Offset = 0; Offset < Buffer.Length; Offset++)
+            {
+                Output[Offset] = DspUtils.Saturate(Buffer[Offset]);
+            }
+
+            return Output;
+        }
+
+        public void Dispose()
+        {
+            Dispose(true);
+        }
+
+        protected virtual void Dispose(bool Disposing)
+        {
+            if (Disposing)
+            {
+                UpdateEvent.Dispose();
+            }
+        }
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/MemoryPoolContext.cs

@@ -0,0 +1,12 @@
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    class MemoryPoolContext
+    {
+        public MemoryPoolOut OutStatus;
+
+        public MemoryPoolContext()
+        {
+            OutStatus.State = MemoryPoolState.Detached;
+        }
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/MemoryPoolIn.cs

@@ -0,0 +1,14 @@
+using System.Runtime.InteropServices;
+
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    [StructLayout(LayoutKind.Sequential, Size = 0x20, Pack = 4)]
+    struct MemoryPoolIn
+    {
+        public long            Address;
+        public long            Size;
+        public MemoryPoolState State;
+        public int             Unknown14;
+        public long            Unknown18;
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/MemoryPoolOut.cs

@@ -0,0 +1,12 @@
+using System.Runtime.InteropServices;
+
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    [StructLayout(LayoutKind.Sequential, Size = 0x10, Pack = 4)]
+    struct MemoryPoolOut
+    {
+        public MemoryPoolState State;
+        public int             Unknown14;
+        public long            Unknown18;
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/MemoryPoolState.cs

@@ -1,4 +1,4 @@
-namespace Ryujinx.HLE.OsHle.Services.Aud
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
 {
     enum MemoryPoolState : int
     {

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/PlayState.cs

@@ -0,0 +1,9 @@
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    enum PlayState : byte
+    {
+        Playing = 0,
+        Stopped = 1,
+        Paused  = 2
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/Resampler.cs

@@ -0,0 +1,191 @@
+using System;
+
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    static class Resampler
+    {
+#region "LookUp Tables"
+        private static short[] CurveLut0 = new short[]
+        {
+            6600,  19426, 6722,  3,     6479,  19424, 6845,  9,     6359,  19419, 6968,  15,    6239,  19412, 7093,  22,
+            6121,  19403, 7219,  28,    6004,  19391, 7345,  34,    5888,  19377, 7472,  41,    5773,  19361, 7600,  48,
+            5659,  19342, 7728,  55,    5546,  19321, 7857,  62,    5434,  19298, 7987,  69,    5323,  19273, 8118,  77,
+            5213,  19245, 8249,  84,    5104,  19215, 8381,  92,    4997,  19183, 8513,  101,   4890,  19148, 8646,  109,
+            4785,  19112, 8780,  118,   4681,  19073, 8914,  127,   4579,  19031, 9048,  137,   4477,  18988, 9183,  147,
+            4377,  18942, 9318,  157,   4277,  18895, 9454,  168,   4179,  18845, 9590,  179,   4083,  18793, 9726,  190,
+            3987,  18738, 9863,  202,   3893,  18682, 10000, 215,   3800,  18624, 10137, 228,   3709,  18563, 10274, 241,
+            3618,  18500, 10411, 255,   3529,  18436, 10549, 270,   3441,  18369, 10687, 285,   3355,  18300, 10824, 300,
+            3269,  18230, 10962, 317,   3186,  18157, 11100, 334,   3103,  18082, 11238, 351,   3022,  18006, 11375, 369,
+            2942,  17927, 11513, 388,   2863,  17847, 11650, 408,   2785,  17765, 11788, 428,   2709,  17681, 11925, 449,
+            2635,  17595, 12062, 471,   2561,  17507, 12198, 494,   2489,  17418, 12334, 517,   2418,  17327, 12470, 541,
+            2348,  17234, 12606, 566,   2280,  17140, 12741, 592,   2213,  17044, 12876, 619,   2147,  16946, 13010, 647,
+            2083,  16846, 13144, 675,   2020,  16745, 13277, 704,   1958,  16643, 13409, 735,   1897,  16539, 13541, 766,
+            1838,  16434, 13673, 798,   1780,  16327, 13803, 832,   1723,  16218, 13933, 866,   1667,  16109, 14062, 901,
+            1613,  15998, 14191, 937,   1560,  15885, 14318, 975,   1508,  15772, 14445, 1013,  1457,  15657, 14571, 1052,
+            1407,  15540, 14695, 1093,  1359,  15423, 14819, 1134,  1312,  15304, 14942, 1177,  1266,  15185, 15064, 1221,
+            1221,  15064, 15185, 1266,  1177,  14942, 15304, 1312,  1134,  14819, 15423, 1359,  1093,  14695, 15540, 1407,
+            1052,  14571, 15657, 1457,  1013,  14445, 15772, 1508,  975,   14318, 15885, 1560,  937,   14191, 15998, 1613,
+            901,   14062, 16109, 1667,  866,   13933, 16218, 1723,  832,   13803, 16327, 1780,  798,   13673, 16434, 1838,
+            766,   13541, 16539, 1897,  735,   13409, 16643, 1958,  704,   13277, 16745, 2020,  675,   13144, 16846, 2083,
+            647,   13010, 16946, 2147,  619,   12876, 17044, 2213,  592,   12741, 17140, 2280,  566,   12606, 17234, 2348,
+            541,   12470, 17327, 2418,  517,   12334, 17418, 2489,  494,   12198, 17507, 2561,  471,   12062, 17595, 2635,
+            449,   11925, 17681, 2709,  428,   11788, 17765, 2785,  408,   11650, 17847, 2863,  388,   11513, 17927, 2942,
+            369,   11375, 18006, 3022,  351,   11238, 18082, 3103,  334,   11100, 18157, 3186,  317,   10962, 18230, 3269,
+            300,   10824, 18300, 3355,  285,   10687, 18369, 3441,  270,   10549, 18436, 3529,  255,   10411, 18500, 3618,
+            241,   10274, 18563, 3709,  228,   10137, 18624, 3800,  215,   10000, 18682, 3893,  202,   9863,  18738, 3987,
+            190,   9726,  18793, 4083,  179,   9590,  18845, 4179,  168,   9454,  18895, 4277,  157,   9318,  18942, 4377,
+            147,   9183,  18988, 4477,  137,   9048,  19031, 4579,  127,   8914,  19073, 4681,  118,   8780,  19112, 4785,
+            109,   8646,  19148, 4890,  101,   8513,  19183, 4997,  92,    8381,  19215, 5104,  84,    8249,  19245, 5213,
+            77,    8118,  19273, 5323,  69,    7987,  19298, 5434,  62,    7857,  19321, 5546,  55,    7728,  19342, 5659,
+            48,    7600,  19361, 5773,  41,    7472,  19377, 5888,  34,    7345,  19391, 6004,  28,    7219,  19403, 6121,
+            22,    7093,  19412, 6239,  15,    6968,  19419, 6359,  9,     6845,  19424, 6479,  3,     6722,  19426, 6600
+        };
+
+        private static short[] CurveLut1 = new short[]
+        {
+            -68,   32639, 69,    -5,    -200,  32630, 212,   -15,   -328,  32613, 359,   -26,   -450,  32586, 512,   -36,
+            -568,  32551, 669,   -47,   -680,  32507, 832,   -58,   -788,  32454, 1000,  -69,   -891,  32393, 1174,  -80,
+            -990,  32323, 1352,  -92,   -1084, 32244, 1536,  -103,  -1173, 32157, 1724,  -115,  -1258, 32061, 1919,  -128,
+            -1338, 31956, 2118,  -140,  -1414, 31844, 2322,  -153,  -1486, 31723, 2532,  -167,  -1554, 31593, 2747,  -180,
+            -1617, 31456, 2967,  -194,  -1676, 31310, 3192,  -209,  -1732, 31157, 3422,  -224,  -1783, 30995, 3657,  -240,
+            -1830, 30826, 3897,  -256,  -1874, 30649, 4143,  -272,  -1914, 30464, 4393,  -289,  -1951, 30272, 4648,  -307,
+            -1984, 30072, 4908,  -325,  -2014, 29866, 5172,  -343,  -2040, 29652, 5442,  -362,  -2063, 29431, 5716,  -382,
+            -2083, 29203, 5994,  -403,  -2100, 28968, 6277,  -424,  -2114, 28727, 6565,  -445,  -2125, 28480, 6857,  -468,
+            -2133, 28226, 7153,  -490,  -2139, 27966, 7453,  -514,  -2142, 27700, 7758,  -538,  -2142, 27428, 8066,  -563,
+            -2141, 27151, 8378,  -588,  -2136, 26867, 8694,  -614,  -2130, 26579, 9013,  -641,  -2121, 26285, 9336,  -668,
+            -2111, 25987, 9663,  -696,  -2098, 25683, 9993,  -724,  -2084, 25375, 10326, -753,  -2067, 25063, 10662, -783,
+            -2049, 24746, 11000, -813,  -2030, 24425, 11342, -844,  -2009, 24100, 11686, -875,  -1986, 23771, 12033, -907,
+            -1962, 23438, 12382, -939,  -1937, 23103, 12733, -972,  -1911, 22764, 13086, -1005, -1883, 22422, 13441, -1039,
+            -1855, 22077, 13798, -1072, -1825, 21729, 14156, -1107, -1795, 21380, 14516, -1141, -1764, 21027, 14877, -1176,
+            -1732, 20673, 15239, -1211, -1700, 20317, 15602, -1246, -1667, 19959, 15965, -1282, -1633, 19600, 16329, -1317,
+            -1599, 19239, 16694, -1353, -1564, 18878, 17058, -1388, -1530, 18515, 17423, -1424, -1495, 18151, 17787, -1459,
+            -1459, 17787, 18151, -1495, -1424, 17423, 18515, -1530, -1388, 17058, 18878, -1564, -1353, 16694, 19239, -1599,
+            -1317, 16329, 19600, -1633, -1282, 15965, 19959, -1667, -1246, 15602, 20317, -1700, -1211, 15239, 20673, -1732,
+            -1176, 14877, 21027, -1764, -1141, 14516, 21380, -1795, -1107, 14156, 21729, -1825, -1072, 13798, 22077, -1855,
+            -1039, 13441, 22422, -1883, -1005, 13086, 22764, -1911, -972,  12733, 23103, -1937, -939,  12382, 23438, -1962,
+            -907,  12033, 23771, -1986, -875,  11686, 24100, -2009, -844,  11342, 24425, -2030, -813,  11000, 24746, -2049,
+            -783,  10662, 25063, -2067, -753,  10326, 25375, -2084, -724,  9993,  25683, -2098, -696,  9663,  25987, -2111,
+            -668,  9336,  26285, -2121, -641,  9013,  26579, -2130, -614,  8694,  26867, -2136, -588,  8378,  27151, -2141,
+            -563,  8066,  27428, -2142, -538,  7758,  27700, -2142, -514,  7453,  27966, -2139, -490,  7153,  28226, -2133,
+            -468,  6857,  28480, -2125, -445,  6565,  28727, -2114, -424,  6277,  28968, -2100, -403,  5994,  29203, -2083,
+            -382,  5716,  29431, -2063, -362,  5442,  29652, -2040, -343,  5172,  29866, -2014, -325,  4908,  30072, -1984,
+            -307,  4648,  30272, -1951, -289,  4393,  30464, -1914, -272,  4143,  30649, -1874, -256,  3897,  30826, -1830,
+            -240,  3657,  30995, -1783, -224,  3422,  31157, -1732, -209,  3192,  31310, -1676, -194,  2967,  31456, -1617,
+            -180,  2747,  31593, -1554, -167,  2532,  31723, -1486, -153,  2322,  31844, -1414, -140,  2118,  31956, -1338,
+            -128,  1919,  32061, -1258, -115,  1724,  32157, -1173, -103,  1536,  32244, -1084, -92,   1352,  32323, -990,
+            -80,   1174,  32393, -891,  -69,   1000,  32454, -788,  -58,   832,   32507, -680,  -47,   669,   32551, -568,
+            -36,   512,   32586, -450,  -26,   359,   32613, -328,  -15,   212,   32630, -200,  -5,    69,    32639, -68
+        };
+
+        private static short[] CurveLut2 = new short[]
+        {
+            3195,  26287, 3329,  -32,   3064,  26281, 3467,  -34,   2936,  26270, 3608,  -38,   2811,  26253, 3751,  -42,
+            2688,  26230, 3897,  -46,   2568,  26202, 4046,  -50,   2451,  26169, 4199,  -54,   2338,  26130, 4354,  -58,
+            2227,  26085, 4512,  -63,   2120,  26035, 4673,  -67,   2015,  25980, 4837,  -72,   1912,  25919, 5004,  -76,
+            1813,  25852, 5174,  -81,   1716,  25780, 5347,  -87,   1622,  25704, 5522,  -92,   1531,  25621, 5701,  -98,
+            1442,  25533, 5882,  -103,  1357,  25440, 6066,  -109,  1274,  25342, 6253,  -115,  1193,  25239, 6442,  -121,
+            1115,  25131, 6635,  -127,  1040,  25018, 6830,  -133,  967,   24899, 7027,  -140,  897,   24776, 7227,  -146,
+            829,   24648, 7430,  -153,  764,   24516, 7635,  -159,  701,   24379, 7842,  -166,  641,   24237, 8052,  -174,
+            583,   24091, 8264,  -181,  526,   23940, 8478,  -187,  472,   23785, 8695,  -194,  420,   23626, 8914,  -202,
+            371,   23462, 9135,  -209,  324,   23295, 9358,  -215,  279,   23123, 9583,  -222,  236,   22948, 9809,  -230,
+            194,   22769, 10038, -237,  154,   22586, 10269, -243,  117,   22399, 10501, -250,  81,    22208, 10735, -258,
+            47,    22015, 10970, -265,  15,    21818, 11206, -271,  -16,   21618, 11444, -277,  -44,   21415, 11684, -283,
+            -71,   21208, 11924, -290,  -97,   20999, 12166, -296,  -121,  20786, 12409, -302,  -143,  20571, 12653, -306,
+            -163,  20354, 12898, -311,  -183,  20134, 13143, -316,  -201,  19911, 13389, -321,  -218,  19686, 13635, -325,
+            -234,  19459, 13882, -328,  -248,  19230, 14130, -332,  -261,  18998, 14377, -335,  -273,  18765, 14625, -337,
+            -284,  18531, 14873, -339,  -294,  18295, 15121, -341,  -302,  18057, 15369, -341,  -310,  17817, 15617, -341,
+            -317,  17577, 15864, -340,  -323,  17335, 16111, -340,  -328,  17092, 16357, -338,  -332,  16848, 16603, -336,
+            -336,  16603, 16848, -332,  -338,  16357, 17092, -328,  -340,  16111, 17335, -323,  -340,  15864, 17577, -317,
+            -341,  15617, 17817, -310,  -341,  15369, 18057, -302,  -341,  15121, 18295, -294,  -339,  14873, 18531, -284,
+            -337,  14625, 18765, -273,  -335,  14377, 18998, -261,  -332,  14130, 19230, -248,  -328,  13882, 19459, -234,
+            -325,  13635, 19686, -218,  -321,  13389, 19911, -201,  -316,  13143, 20134, -183,  -311,  12898, 20354, -163,
+            -306,  12653, 20571, -143,  -302,  12409, 20786, -121,  -296,  12166, 20999, -97,   -290,  11924, 21208, -71,
+            -283,  11684, 21415, -44,   -277,  11444, 21618, -16,   -271,  11206, 21818, 15,    -265,  10970, 22015, 47,
+            -258,  10735, 22208, 81,    -250,  10501, 22399, 117,   -243,  10269, 22586, 154,   -237,  10038, 22769, 194,
+            -230,  9809,  22948, 236,   -222,  9583,  23123, 279,   -215,  9358,  23295, 324,   -209,  9135,  23462, 371,
+            -202,  8914,  23626, 420,   -194,  8695,  23785, 472,   -187,  8478,  23940, 526,   -181,  8264,  24091, 583,
+            -174,  8052,  24237, 641,   -166,  7842,  24379, 701,   -159,  7635,  24516, 764,   -153,  7430,  24648, 829,
+            -146,  7227,  24776, 897,   -140,  7027,  24899, 967,   -133,  6830,  25018, 1040,  -127,  6635,  25131, 1115,
+            -121,  6442,  25239, 1193,  -115,  6253,  25342, 1274,  -109,  6066,  25440, 1357,  -103,  5882,  25533, 1442,
+            -98,   5701,  25621, 1531,  -92,   5522,  25704, 1622,  -87,   5347,  25780, 1716,  -81,   5174,  25852, 1813,
+            -76,   5004,  25919, 1912,  -72,   4837,  25980, 2015,  -67,   4673,  26035, 2120,  -63,   4512,  26085, 2227,
+            -58,   4354,  26130, 2338,  -54,   4199,  26169, 2451,  -50,   4046,  26202, 2568,  -46,   3897,  26230, 2688,
+            -42,   3751,  26253, 2811,  -38,   3608,  26270, 2936,  -34,   3467,  26281, 3064,  -32,   3329,  26287, 3195
+        };
+#endregion
+
+        public static int[] Resample2Ch(
+            int[]   Buffer,
+            int     SrcSampleRate,
+            int     DstSampleRate,
+            int     SamplesCount,
+            ref int FracPart)
+        {
+            if (Buffer == null)
+            {
+                throw new ArgumentNullException(nameof(Buffer));
+            }
+
+            if (SrcSampleRate <= 0)
+            {
+                throw new ArgumentOutOfRangeException(nameof(SrcSampleRate));
+            }
+
+            if (DstSampleRate <= 0)
+            {
+                throw new ArgumentOutOfRangeException(nameof(DstSampleRate));
+            }
+
+            double Ratio = (double)SrcSampleRate / DstSampleRate;
+
+            int NewSamplesCount = (int)(SamplesCount / Ratio);
+
+            int Step = (int)(Ratio * 0x8000);
+
+            int[] Output = new int[NewSamplesCount * 2];
+
+            short[] Lut;
+
+            if (Step > 0xaaaa)
+            {
+                Lut = CurveLut0;
+            }
+            else if (Step <= 0x8000)
+            {
+                Lut = CurveLut1;
+            }
+            else
+            {
+                Lut = CurveLut2;
+            }
+
+            int InOffs = 0;
+
+            for (int OutOffs = 0; OutOffs < Output.Length; OutOffs += 2)
+            {
+                int LutIndex = (FracPart >> 8) * 4;
+
+                int Sample0 = Buffer[(InOffs + 0) * 2 + 0] * Lut[LutIndex + 0] +
+                              Buffer[(InOffs + 1) * 2 + 0] * Lut[LutIndex + 1] +
+                              Buffer[(InOffs + 2) * 2 + 0] * Lut[LutIndex + 2] +
+                              Buffer[(InOffs + 3) * 2 + 0] * Lut[LutIndex + 3];
+
+                int Sample1 = Buffer[(InOffs + 0) * 2 + 1] * Lut[LutIndex + 0] +
+                              Buffer[(InOffs + 1) * 2 + 1] * Lut[LutIndex + 1] +
+                              Buffer[(InOffs + 2) * 2 + 1] * Lut[LutIndex + 2] +
+                              Buffer[(InOffs + 3) * 2 + 1] * Lut[LutIndex + 3];
+
+                int NewOffset = FracPart + Step;
+
+                InOffs += NewOffset >> 15;
+
+                FracPart = NewOffset & 0x7fff;
+
+                Output[OutOffs + 0] = Sample0 >> 15;
+                Output[OutOffs + 1] = Sample1 >> 15;
+            }
+
+            return Output;
+        }
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/UpdateDataHeader.cs

@@ -1,15 +1,15 @@
-namespace Ryujinx.HLE.OsHle.Services.Aud
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
 {
     struct UpdateDataHeader
     {
         public int Revision;
         public int BehaviorSize;
-        public int MemoryPoolsSize;
-        public int VoicesSize;
+        public int MemoryPoolSize;
+        public int VoiceSize;
         public int VoiceResourceSize;
-        public int EffectsSize;
-        public int MixesSize;
-        public int SinksSize;
+        public int EffectSize;
+        public int MixeSize;
+        public int SinkSize;
         public int PerformanceManagerSize;
         public int Unknown24;
         public int Unknown28;

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/VoiceChannelResourceIn.cs

@@ -0,0 +1,10 @@
+using System.Runtime.InteropServices;
+
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    [StructLayout(LayoutKind.Sequential, Size = 0x70, Pack = 1)]
+    struct VoiceChannelResourceIn
+    {
+        //???
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/VoiceContext.cs

@@ -0,0 +1,188 @@
+using ChocolArm64.Memory;
+using Ryujinx.Audio.Adpcm;
+using System;
+
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    class VoiceContext
+    {
+        private bool Acquired;
+        private bool BufferReload;
+
+        private int ResamplerFracPart;
+
+        private int BufferIndex;
+        private int Offset;
+
+        public int SampleRate;
+        public int ChannelsCount;
+
+        public float Volume;
+
+        public PlayState PlayState;
+
+        public SampleFormat SampleFormat;
+
+        public AdpcmDecoderContext AdpcmCtx;
+
+        public WaveBuffer[] WaveBuffers;
+
+        public VoiceOut OutStatus;
+
+        private int[] Samples;
+
+        public bool Playing => Acquired && PlayState == PlayState.Playing;
+
+        public VoiceContext()
+        {
+            WaveBuffers = new WaveBuffer[4];
+        }
+
+        public void SetAcquireState(bool NewState)
+        {
+            if (Acquired && !NewState)
+            {
+                //Release.
+                Reset();
+            }
+
+            Acquired = NewState;
+        }
+
+        private void Reset()
+        {
+            BufferReload = true;
+
+            BufferIndex = 0;
+            Offset      = 0;
+
+            OutStatus.PlayedSamplesCount     = 0;
+            OutStatus.PlayedWaveBuffersCount = 0;
+            OutStatus.VoiceDropsCount        = 0;
+        }
+
+        public int[] GetBufferData(AMemory Memory, int MaxSamples, out int SamplesCount)
+        {
+            if (!Playing)
+            {
+                SamplesCount = 0;
+
+                return null;
+            }
+
+            if (BufferReload)
+            {
+                BufferReload = false;
+
+                UpdateBuffer(Memory);
+            }
+
+            WaveBuffer Wb = WaveBuffers[BufferIndex];
+
+            int MaxSize = Samples.Length - Offset;
+
+            int Size = MaxSamples * AudioConsts.HostChannelsCount;
+
+            if (Size > MaxSize)
+            {
+                Size = MaxSize;
+            }
+
+            int[] Output = new int[Size];
+
+            Array.Copy(Samples, Offset, Output, 0, Size);
+
+            SamplesCount = Size / AudioConsts.HostChannelsCount;
+
+            OutStatus.PlayedSamplesCount += SamplesCount;
+
+            Offset += Size;
+
+            if (Offset == Samples.Length)
+            {
+                Offset = 0;
+
+                if (Wb.Looping == 0)
+                {
+                    SetBufferIndex((BufferIndex + 1) & 3);
+                }
+
+                OutStatus.PlayedWaveBuffersCount++;
+
+                if (Wb.LastBuffer != 0)
+                {
+                    PlayState = PlayState.Paused;
+                }
+            }
+
+            return Output;
+        }
+
+        private void UpdateBuffer(AMemory Memory)
+        {
+            //TODO: Implement conversion for formats other
+            //than interleaved stereo (2 channels).
+            //As of now, it assumes that HostChannelsCount == 2.
+            WaveBuffer Wb = WaveBuffers[BufferIndex];
+
+            if (SampleFormat == SampleFormat.PcmInt16)
+            {
+                int SamplesCount = (int)(Wb.Size / (sizeof(short) * ChannelsCount));
+
+                Samples = new int[SamplesCount * AudioConsts.HostChannelsCount];
+
+                if (ChannelsCount == 1)
+                {
+                    for (int Index = 0; Index < SamplesCount; Index++)
+                    {
+                        short Sample = Memory.ReadInt16(Wb.Position + Index * 2);
+
+                        Samples[Index * 2 + 0] = Sample;
+                        Samples[Index * 2 + 1] = Sample;
+                    }
+                }
+                else
+                {
+                    for (int Index = 0; Index < SamplesCount * 2; Index++)
+                    {
+                        Samples[Index] = Memory.ReadInt16(Wb.Position + Index * 2);
+                    }
+                }
+            }
+            else if (SampleFormat == SampleFormat.Adpcm)
+            {
+                byte[] Buffer = Memory.ReadBytes(Wb.Position, Wb.Size);
+
+                Samples = AdpcmDecoder.Decode(Buffer, AdpcmCtx);
+            }
+            else
+            {
+                throw new InvalidOperationException();
+            }
+
+            if (SampleRate != AudioConsts.HostSampleRate)
+            {
+                //TODO: We should keep the frames being discarded (see the 4 below)
+                //on a buffer and include it on the next samples buffer, to allow
+                //the resampler to do seamless interpolation between wave buffers.
+                int SamplesCount = Samples.Length / AudioConsts.HostChannelsCount;
+
+                SamplesCount = Math.Max(SamplesCount - 4, 0);
+
+                Samples = Resampler.Resample2Ch(
+                    Samples,
+                    SampleRate,
+                    AudioConsts.HostSampleRate,
+                    SamplesCount,
+                    ref ResamplerFracPart);
+            }
+        }
+
+        public void SetBufferIndex(int Index)
+        {
+            BufferIndex = Index & 3;
+
+            BufferReload = true;
+        }
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/VoiceIn.cs

@@ -0,0 +1,49 @@
+using System.Runtime.InteropServices;
+
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    [StructLayout(LayoutKind.Sequential, Size = 0x170, Pack = 1)]
+    struct VoiceIn
+    {
+        public int VoiceSlot;
+        public int NodeId;
+
+        public byte FirstUpdate;
+        public byte Acquired;
+
+        public PlayState PlayState;
+
+        public SampleFormat SampleFormat;
+
+        public int SampleRate;
+
+        public int Priority;
+
+        public int Unknown14;
+
+        public int ChannelsCount;
+
+        public float Pitch;
+        public float Volume;
+
+        public BiquadFilter BiquadFilter0;
+        public BiquadFilter BiquadFilter1;
+
+        public int AppendedWaveBuffersCount;
+
+        public int BaseWaveBufferIndex;
+
+        public int Unknown44;
+
+        public long AdpcmCoeffsPosition;
+        public long AdpcmCoeffsSize;
+
+        public int VoiceDestination;
+        public int Padding;
+
+        public WaveBuffer WaveBuffer0;
+        public WaveBuffer WaveBuffer1;
+        public WaveBuffer WaveBuffer2;
+        public WaveBuffer WaveBuffer3;
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/VoiceOut.cs

@@ -0,0 +1,12 @@
+using System.Runtime.InteropServices;
+
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    [StructLayout(LayoutKind.Sequential, Size = 0x10, Pack = 4)]
+    struct VoiceOut
+    {
+        public long PlayedSamplesCount;
+        public int  PlayedWaveBuffersCount;
+        public int  VoiceDropsCount; //?
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRenderer/WaveBuffer.cs

@@ -0,0 +1,20 @@
+using System.Runtime.InteropServices;
+
+namespace Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer
+{
+    [StructLayout(LayoutKind.Sequential, Size = 0x38, Pack = 1)]
+    struct WaveBuffer
+    {
+        public long  Position;
+        public long  Size;
+        public int   FirstSampleOffset;
+        public int   LastSampleOffset;
+        public byte  Looping;
+        public byte  LastBuffer;
+        public short Unknown1A;
+        public int   Unknown1C;
+        public long  AdpcmLoopContextPosition;
+        public long  AdpcmLoopContextSize;
+        public long  Unknown30;
+    }
+}

Ryujinx.HLE/OsHle/Services/Aud/AudioRendererParameter.cs

@@ -8,14 +8,14 @@ namespace Ryujinx.HLE.OsHle.Services.Aud
         public int SampleRate;
         public int SampleCount;
         public int Unknown8;
-        public int UnknownC;
+        public int MixCount;
         public int VoiceCount;
         public int SinkCount;
         public int EffectCount;
-        public int Unknown1C;
-        public int Unknown20;
+        public int PerformanceManagerCount;
+        public int VoiceDropEnable;
         public int SplitterCount;
-        public int Unknown28;
+        public int SplitterDestinationDataCount;
         public int Unknown2C;
         public int Revision;
     }

Ryujinx.HLE/OsHle/Services/Aud/IAudioOutManager.cs

@@ -3,6 +3,7 @@ using Ryujinx.Audio;
 using Ryujinx.HLE.Logging;
 using Ryujinx.HLE.OsHle.Handles;
 using Ryujinx.HLE.OsHle.Ipc;
+using Ryujinx.HLE.OsHle.Services.Aud.AudioOut;
 using System.Collections.Generic;
 using System.Text;
 
@@ -14,6 +15,10 @@ namespace Ryujinx.HLE.OsHle.Services.Aud
     {
         private const string DefaultAudioOutput = "DeviceOut";
 
+        private const int DefaultSampleRate = 48000;
+
+        private const int DefaultChannelsCount = 2;
+
         private Dictionary<int, ServiceProcessRequest> m_Commands;
 
         public override IReadOnlyDictionary<int, ServiceProcessRequest> Commands => m_Commands;
@@ -122,7 +127,12 @@ namespace Ryujinx.HLE.OsHle.Services.Aud
             int SampleRate = Context.RequestData.ReadInt32();
             int Channels   = Context.RequestData.ReadInt32();
 
-            if (SampleRate != 48000)
+            if (SampleRate == 0)
+            {
+                SampleRate = DefaultSampleRate;
+            }
+
+            if (SampleRate != DefaultSampleRate)
             {
                 Context.Ns.Log.PrintWarning(LogClass.Audio, "Invalid sample rate!");
 
@@ -133,7 +143,7 @@ namespace Ryujinx.HLE.OsHle.Services.Aud
 
             if (Channels == 0)
             {
-                Channels = 2;
+                Channels = DefaultChannelsCount;
             }
 
             KEvent ReleaseEvent = new KEvent();
@@ -145,13 +155,13 @@ namespace Ryujinx.HLE.OsHle.Services.Aud
 
             IAalOutput AudioOut = Context.Ns.AudioOut;
 
-            int Track = AudioOut.OpenTrack(SampleRate, 2, Callback, out AudioFormat Format);
+            int Track = AudioOut.OpenTrack(SampleRate, Channels, Callback);
 
             MakeObject(Context, new IAudioOut(AudioOut, ReleaseEvent, Track));
 
             Context.ResponseData.Write(SampleRate);
             Context.ResponseData.Write(Channels);
-            Context.ResponseData.Write((int)Format);
+            Context.ResponseData.Write((int)SampleFormat.PcmInt16);
             Context.ResponseData.Write((int)PlaybackState.Stopped);
 
             return 0;

Ryujinx.HLE/OsHle/Services/Aud/IAudioRenderer.cs

@@ -1,136 +0,0 @@
-using ChocolArm64.Memory;
-using Ryujinx.HLE.Logging;
-using Ryujinx.HLE.OsHle.Handles;
-using Ryujinx.HLE.OsHle.Ipc;
-using System;
-using System.Collections.Generic;
-using System.Runtime.InteropServices;
-
-namespace Ryujinx.HLE.OsHle.Services.Aud
-{
-    class IAudioRenderer : IpcService, IDisposable
-    {
-        private Dictionary<int, ServiceProcessRequest> m_Commands;
-
-        public override IReadOnlyDictionary<int, ServiceProcessRequest> Commands => m_Commands;
-
-        private KEvent UpdateEvent;
-
-        private AudioRendererParameter Params;
-
-        public IAudioRenderer(AudioRendererParameter Params)
-        {
-            m_Commands = new Dictionary<int, ServiceProcessRequest>()
-            {
-                { 4, RequestUpdateAudioRenderer },
-                { 5, StartAudioRenderer         },
-                { 6, StopAudioRenderer          },
-                { 7, QuerySystemEvent           }
-            };
-
-            UpdateEvent = new KEvent();
-
-            this.Params = Params;
-        }
-
-        public long RequestUpdateAudioRenderer(ServiceCtx Context)
-        {
-            long OutputPosition = Context.Request.ReceiveBuff[0].Position;
-            long OutputSize     = Context.Request.ReceiveBuff[0].Size;
-
-            AMemoryHelper.FillWithZeros(Context.Memory, OutputPosition, (int)OutputSize);
-
-            long InputPosition = Context.Request.SendBuff[0].Position;
-
-            UpdateDataHeader InputDataHeader = AMemoryHelper.Read<UpdateDataHeader>(Context.Memory, InputPosition);
-
-            UpdateDataHeader OutputDataHeader = new UpdateDataHeader();
-
-            int UpdateHeaderSize = Marshal.SizeOf<UpdateDataHeader>();
-
-            OutputDataHeader.Revision               = Params.Revision;
-            OutputDataHeader.BehaviorSize           = 0xb0;
-            OutputDataHeader.MemoryPoolsSize        = (Params.EffectCount + Params.VoiceCount * 4) * 0x10;
-            OutputDataHeader.VoicesSize             = Params.VoiceCount  * 0x10;
-            OutputDataHeader.EffectsSize            = Params.EffectCount * 0x10;
-            OutputDataHeader.SinksSize              = Params.SinkCount   * 0x20;
-            OutputDataHeader.PerformanceManagerSize = 0x10;
-            OutputDataHeader.TotalSize              = UpdateHeaderSize                 +
-                                                      OutputDataHeader.BehaviorSize    +
-                                                      OutputDataHeader.MemoryPoolsSize +
-                                                      OutputDataHeader.VoicesSize      +
-                                                      OutputDataHeader.EffectsSize     +
-                                                      OutputDataHeader.SinksSize       +
-                                                      OutputDataHeader.PerformanceManagerSize;
-
-            AMemoryHelper.Write(Context.Memory, OutputPosition, OutputDataHeader);
-
-            int InMemoryPoolOffset = UpdateHeaderSize + InputDataHeader.BehaviorSize;
-
-            int OutMemoryPoolOffset = UpdateHeaderSize;
-
-            for (int Offset = 0; Offset < OutputDataHeader.MemoryPoolsSize; Offset += 0x10, InMemoryPoolOffset += 0x20)
-            {
-                MemoryPoolState PoolState = (MemoryPoolState)Context.Memory.ReadInt32(InputPosition + InMemoryPoolOffset + 0x10);
-
-                //TODO: Figure out what the other values does.
-                if (PoolState == MemoryPoolState.RequestAttach)
-                {
-                    Context.Memory.WriteInt32(OutputPosition + OutMemoryPoolOffset + Offset, (int)MemoryPoolState.Attached);
-                }
-                else if (PoolState == MemoryPoolState.RequestDetach)
-                {
-                    Context.Memory.WriteInt32(OutputPosition + OutMemoryPoolOffset + Offset, (int)MemoryPoolState.Detached);
-                }
-            }
-
-            int OutVoicesOffset = OutMemoryPoolOffset + OutputDataHeader.MemoryPoolsSize;
-
-            for (int Offset = 0; Offset < OutputDataHeader.VoicesSize; Offset += 0x10)
-            {
-                Context.Memory.WriteInt32(OutputPosition + OutVoicesOffset + Offset + 8, (int)VoicePlaybackState.Finished);
-            }
-
-            //TODO: We shouldn't be signaling this here.
-            UpdateEvent.WaitEvent.Set();
-
-            return 0;
-        }
-
-        public long StartAudioRenderer(ServiceCtx Context)
-        {
-            Context.Ns.Log.PrintStub(LogClass.ServiceAudio, "Stubbed.");
-
-            return 0;
-        }
-
-        public long StopAudioRenderer(ServiceCtx Context)
-        {
-            Context.Ns.Log.PrintStub(LogClass.ServiceAudio, "Stubbed.");
-
-            return 0;
-        }
-
-        public long QuerySystemEvent(ServiceCtx Context)
-        {
-            int Handle = Context.Process.HandleTable.OpenHandle(UpdateEvent);
-
-            Context.Response.HandleDesc = IpcHandleDesc.MakeCopy(Handle);
-
-            return 0;
-        }
-
-        public void Dispose()
-        {
-            Dispose(true);
-        }
-
-        protected virtual void Dispose(bool Disposing)
-        {
-            if (Disposing)
-            {
-                UpdateEvent.Dispose();
-            }
-        }
-    }
-}

Ryujinx.HLE/OsHle/Services/Aud/IAudioRendererManager.cs

@@ -1,7 +1,11 @@
+using Ryujinx.Audio;
 using Ryujinx.HLE.Logging;
 using Ryujinx.HLE.OsHle.Ipc;
+using Ryujinx.HLE.OsHle.Services.Aud.AudioRenderer;
+using Ryujinx.HLE.OsHle.Utilities;
 using System.Collections.Generic;
-using System.Runtime.InteropServices;
+
+using static Ryujinx.HLE.OsHle.ErrorCode;
 
 namespace Ryujinx.HLE.OsHle.Services.Aud
 {
@@ -12,6 +16,10 @@ namespace Ryujinx.HLE.OsHle.Services.Aud
                                       ('V' << 16) |
                                       ('0' << 24);
 
+        private const int Rev = 4;
+
+        public const int RevMagic = Rev0Magic + (Rev << 24);
+
         private Dictionary<int, ServiceProcessRequest> m_Commands;
 
         public override IReadOnlyDictionary<int, ServiceProcessRequest> Commands => m_Commands;
@@ -28,76 +36,69 @@ namespace Ryujinx.HLE.OsHle.Services.Aud
 
         public long OpenAudioRenderer(ServiceCtx Context)
         {
-            //Same buffer as GetAudioRendererWorkBufferSize is receive here.
+            IAalOutput AudioOut = Context.Ns.AudioOut;
 
-            AudioRendererParameter Params = new AudioRendererParameter();
+            AudioRendererParameter Params = GetAudioRendererParameter(Context);
 
-            Params.SampleRate    = Context.RequestData.ReadInt32();
-            Params.SampleCount   = Context.RequestData.ReadInt32();
-            Params.Unknown8      = Context.RequestData.ReadInt32();
-            Params.UnknownC      = Context.RequestData.ReadInt32();
-            Params.VoiceCount    = Context.RequestData.ReadInt32();
-            Params.SinkCount     = Context.RequestData.ReadInt32();
-            Params.EffectCount   = Context.RequestData.ReadInt32();
-            Params.Unknown1C     = Context.RequestData.ReadInt32();
-            Params.Unknown20     = Context.RequestData.ReadInt32();
-            Params.SplitterCount = Context.RequestData.ReadInt32();
-            Params.Unknown28     = Context.RequestData.ReadInt32();
-            Params.Unknown2C     = Context.RequestData.ReadInt32();
-            Params.Revision      = Context.RequestData.ReadInt32();
-
-            MakeObject(Context, new IAudioRenderer(Params));
+            MakeObject(Context, new IAudioRenderer(Context.Memory, AudioOut, Params));
 
             return 0;
         }
 
         public long GetAudioRendererWorkBufferSize(ServiceCtx Context)
         {
-            long SampleRate = Context.RequestData.ReadUInt32();
-            long Unknown4   = Context.RequestData.ReadUInt32();
-            long Unknown8   = Context.RequestData.ReadUInt32();
-            long UnknownC   = Context.RequestData.ReadUInt32();
-            long Unknown10  = Context.RequestData.ReadUInt32(); //VoiceCount
-            long Unknown14  = Context.RequestData.ReadUInt32(); //SinkCount
-            long Unknown18  = Context.RequestData.ReadUInt32(); //EffectCount
-            long Unknown1c  = Context.RequestData.ReadUInt32(); //Boolean
-            long Unknown20  = Context.RequestData.ReadUInt32(); //Not used here in FW3.0.1 - Boolean
-            long Unknown24  = Context.RequestData.ReadUInt32();
-            long Unknown28  = Context.RequestData.ReadUInt32(); //SplitterCount
-            long Unknown2c  = Context.RequestData.ReadUInt32(); //Not used here in FW3.0.1
-            int  RevMagic   = Context.RequestData.ReadInt32();
+            AudioRendererParameter Params = GetAudioRendererParameter(Context);
 
-            int Version = (RevMagic - Rev0Magic) >> 24;
+            int Revision = (Params.Revision - Rev0Magic) >> 24;
 
-            if (Version <= 3) //REV3 Max is supported
+            if (Revision <= Rev)
             {
-                long Size  = RoundUp(Unknown8 * 4, 64);
-                     Size += (UnknownC << 10);
-                     Size += (UnknownC + 1) * 0x940;
-                     Size += Unknown10 * 0x3F0;
-                     Size += RoundUp((UnknownC + 1) * 8, 16);
-                     Size += RoundUp(Unknown10 * 8, 16);
-                     Size += RoundUp((0x3C0 * (Unknown14 + UnknownC) + 4 * Unknown4) * (Unknown8 + 6), 64);
-                     Size += 0x2C0 * (Unknown14 + UnknownC) + 0x30 * (Unknown18 + (4 * Unknown10)) + 0x50;
+                bool IsSplitterSupported = Revision >= 3;
 
-                if (Version >= 3) //IsSplitterSupported
+                long Size;
+
+                Size  = IntUtils.AlignUp(Params.Unknown8 * 4, 64);
+                Size += Params.MixCount * 0x400;
+                Size += (Params.MixCount + 1) * 0x940;
+                Size += Params.VoiceCount * 0x3F0;
+                Size += IntUtils.AlignUp((Params.MixCount + 1) * 8, 16);
+                Size += IntUtils.AlignUp(Params.VoiceCount * 8, 16);
+                Size += IntUtils.AlignUp(
+                    ((Params.SinkCount + Params.MixCount) * 0x3C0 + Params.SampleCount * 4) *
+                    (Params.Unknown8 + 6), 64);
+                Size += (Params.SinkCount + Params.MixCount) * 0x2C0;
+                Size += (Params.EffectCount + Params.VoiceCount * 4) * 0x30 + 0x50;
+
+                if (IsSplitterSupported)
                 {
-                    Size += RoundUp((NodeStatesGetWorkBufferSize((int)UnknownC + 1) + EdgeMatrixGetWorkBufferSize((int)UnknownC + 1)), 16);
-                    Size += 0xE0 * Unknown28 + 0x20 * Unknown24 + RoundUp(Unknown28 * 4, 16);
+                    Size += IntUtils.AlignUp((
+                        NodeStatesGetWorkBufferSize(Params.MixCount + 1) +
+                        EdgeMatrixGetWorkBufferSize(Params.MixCount + 1)), 16);
+
+                    Size += Params.SplitterDestinationDataCount * 0xE0;
+                    Size += Params.SplitterCount * 0x20;
+                    Size += IntUtils.AlignUp(Params.SplitterDestinationDataCount * 4, 16);
                 }
 
-                Size = 0x4C0 * Unknown18 + RoundUp(Size, 64) + 0x170 * Unknown14 + ((Unknown10 << 8) | 0x40);
+                Size = Params.EffectCount * 0x4C0 +
+                       Params.SinkCount * 0x170 +
+                       Params.VoiceCount * 0x100 +
+                       IntUtils.AlignUp(Size, 64) + 0x40;
 
-                if (Unknown1c >= 1)
+                if (Params.PerformanceManagerCount >= 1)
                 {
-                    Size += ((((Unknown18 + Unknown14 + Unknown10 + UnknownC + 1) * 16) + 0x658) * (Unknown1c + 1) + 0x13F) & ~0x3FL;
+                    Size += (((Params.EffectCount +
+                               Params.SinkCount +
+                               Params.VoiceCount +
+                               Params.MixCount + 1) * 16 + 0x658) *
+                               (Params.PerformanceManagerCount + 1) + 0x13F) & ~0x3FL;
                 }
 
-                long WorkBufferSize = (Size + 0x1907D) & ~0xFFFL;
+                Size = (Size + 0x1907D) & ~0xFFFL;
 
-                Context.ResponseData.Write(WorkBufferSize);
+                Context.ResponseData.Write(Size);
 
-                Context.Ns.Log.PrintDebug(LogClass.ServiceAudio, $"WorkBufferSize is 0x{WorkBufferSize:x16}.");
+                Context.Ns.Log.PrintDebug(LogClass.ServiceAudio, $"WorkBufferSize is 0x{Size:x16}.");
 
                 return 0;
             }
@@ -105,20 +106,36 @@ namespace Ryujinx.HLE.OsHle.Services.Aud
             {
                 Context.ResponseData.Write(0L);
 
-                Context.Ns.Log.PrintError(LogClass.ServiceAudio, $"Library Revision 0x{RevMagic:x8} is not supported!");
+                Context.Ns.Log.PrintWarning(LogClass.ServiceAudio, $"Library Revision 0x{Params.Revision:x8} is not supported!");
 
-                return 0x499;
+                return MakeError(ErrorModule.Audio, AudErr.UnsupportedRevision);
             }
         }
 
-        private static long RoundUp(long Value, int Size)
+        private AudioRendererParameter GetAudioRendererParameter(ServiceCtx Context)
         {
-            return (Value + (Size - 1)) & ~((long)Size - 1);
+            AudioRendererParameter Params = new AudioRendererParameter();
+
+            Params.SampleRate                   = Context.RequestData.ReadInt32();
+            Params.SampleCount                  = Context.RequestData.ReadInt32();
+            Params.Unknown8                     = Context.RequestData.ReadInt32();
+            Params.MixCount                     = Context.RequestData.ReadInt32();
+            Params.VoiceCount                   = Context.RequestData.ReadInt32();
+            Params.SinkCount                    = Context.RequestData.ReadInt32();
+            Params.EffectCount                  = Context.RequestData.ReadInt32();
+            Params.PerformanceManagerCount      = Context.RequestData.ReadInt32();
+            Params.VoiceDropEnable              = Context.RequestData.ReadInt32();
+            Params.SplitterCount                = Context.RequestData.ReadInt32();
+            Params.SplitterDestinationDataCount = Context.RequestData.ReadInt32();
+            Params.Unknown2C                    = Context.RequestData.ReadInt32();
+            Params.Revision                     = Context.RequestData.ReadInt32();
+
+            return Params;
         }
 
         private static int NodeStatesGetWorkBufferSize(int Value)
         {
-            int Result = (int)RoundUp(Value, 64);
+            int Result = IntUtils.AlignUp(Value, 64);
 
             if (Result < 0)
             {
@@ -130,7 +147,7 @@ namespace Ryujinx.HLE.OsHle.Services.Aud
 
         private static int EdgeMatrixGetWorkBufferSize(int Value)
         {
-            int Result = (int)RoundUp(Value * Value, 64);
+            int Result = IntUtils.AlignUp(Value * Value, 64);
 
             if (Result < 0)
             {

Ryujinx.HLE/OsHle/Services/Aud/SampleFormat.cs

@@ -1,12 +1,12 @@
-namespace Ryujinx.Audio
+namespace Ryujinx.HLE.OsHle.Services.Aud
 {
-    public enum AudioFormat
+    enum SampleFormat : byte
     {
         Invalid  = 0,
         PcmInt8  = 1,
         PcmInt16 = 2,
-        PcmImt24 = 3,
-        PcmImt32 = 4,
+        PcmInt24 = 3,
+        PcmInt32 = 4,
         PcmFloat = 5,
         Adpcm    = 6
     }

Ryujinx.HLE/OsHle/Services/Aud/VoiceState.cs

@@ -1,9 +0,0 @@
-namespace Ryujinx.HLE.OsHle.Services.Aud
-{
-    enum VoicePlaybackState : int
-    {
-        Playing  = 0,
-        Finished = 1,
-        Paused   = 2
-    }
-}

Ryujinx.HLE/OsHle/Services/Nv/INvDrvServices.cs

@@ -23,11 +23,11 @@ namespace Ryujinx.HLE.OsHle.Services.Nv
         private static Dictionary<string, IoctlProcessor> IoctlProcessors =
                    new Dictionary<string, IoctlProcessor>()
         {
-            { "/dev/nvhost-as-gpu",   ProcessIoctlNvGpuAS       },
-            { "/dev/nvhost-ctrl",     ProcessIoctlNvHostCtrl    },
-            { "/dev/nvhost-ctrl-gpu", ProcessIoctlNvGpuGpu      },
-            { "/dev/nvhost-gpu",      ProcessIoctlNvHostChannel },
-            { "/dev/nvmap",           ProcessIoctlNvMap         }
+            { "/dev/nvhost-as-gpu",   ProcessIoctlNvGpuAS    },
+            { "/dev/nvhost-ctrl",     ProcessIoctlNvHostCtrl },
+            { "/dev/nvhost-ctrl-gpu", ProcessIoctlNvGpuGpu   },
+            { "/dev/nvhost-gpu",      ProcessIoctlNvHostGpu  },
+            { "/dev/nvmap",           ProcessIoctlNvMap      }
         };
 
         public static GlobalStateTable Fds { get; private set; }
@@ -44,6 +44,7 @@ namespace Ryujinx.HLE.OsHle.Services.Nv
                 { 3,  Initialize       },
                 { 4,  QueryEvent       },
                 { 8,  SetClientPid     },
+                { 11, Ioctl            },
                 { 13, FinishInitialize }
             };
 
@@ -162,9 +163,9 @@ namespace Ryujinx.HLE.OsHle.Services.Nv
             return ProcessIoctl(Context, Cmd, NvGpuGpuIoctl.ProcessIoctl);
         }
 
-        private static int ProcessIoctlNvHostChannel(ServiceCtx Context, int Cmd)
+        private static int ProcessIoctlNvHostGpu(ServiceCtx Context, int Cmd)
         {
-            return ProcessIoctl(Context, Cmd, NvHostChannelIoctl.ProcessIoctl);
+            return ProcessIoctl(Context, Cmd, NvHostChannelIoctl.ProcessIoctlGpu);
         }
 
         private static int ProcessIoctlNvMap(ServiceCtx Context, int Cmd)
@@ -207,6 +208,8 @@ namespace Ryujinx.HLE.OsHle.Services.Nv
 
             NvGpuASIoctl.UnloadProcess(Process);
 
+            NvHostChannelIoctl.UnloadProcess(Process);
+
             NvHostCtrlIoctl.UnloadProcess(Process);
 
             NvMapIoctl.UnloadProcess(Process);

Ryujinx.HLE/OsHle/Services/Nv/NvHostChannel/NvChannel.cs

@@ -0,0 +1,7 @@
+namespace Ryujinx.HLE.OsHle.Services.Nv.NvHostChannel
+{
+    class NvChannel
+    {
+        public int Timeout;
+    }
+}

Ryujinx.HLE/OsHle/Services/Nv/NvHostChannel/NvChannelName.cs

@@ -0,0 +1,7 @@
+namespace Ryujinx.HLE.OsHle.Services.Nv.NvHostChannel
+{
+    enum NvChannelName
+    {
+        Gpu
+    }
+}

Ryujinx.HLE/OsHle/Services/Nv/NvHostChannel/NvHostChannelIoctl.cs

@@ -3,23 +3,50 @@ using Ryujinx.HLE.Gpu.Memory;
 using Ryujinx.HLE.Logging;
 using Ryujinx.HLE.OsHle.Services.Nv.NvGpuAS;
 using System;
+using System.Collections.Concurrent;
 
 namespace Ryujinx.HLE.OsHle.Services.Nv.NvHostChannel
 {
     class NvHostChannelIoctl
     {
-        public static int ProcessIoctl(ServiceCtx Context, int Cmd)
+        private class ChannelsPerProcess
+        {
+            public ConcurrentDictionary<NvChannelName, NvChannel> Channels { get; private set; }
+
+            public ChannelsPerProcess()
+            {
+                Channels = new ConcurrentDictionary<NvChannelName, NvChannel>();
+
+                Channels.TryAdd(NvChannelName.Gpu, new NvChannel());
+            }
+        }
+
+        private static ConcurrentDictionary<Process, ChannelsPerProcess> Channels;
+
+        static NvHostChannelIoctl()
+        {
+            Channels = new ConcurrentDictionary<Process, ChannelsPerProcess>();
+        }
+
+        public static int ProcessIoctlGpu(ServiceCtx Context, int Cmd)
+        {
+            return ProcessIoctl(Context, NvChannelName.Gpu, Cmd);
+        }
+
+        public static int ProcessIoctl(ServiceCtx Context, NvChannelName Channel, int Cmd)
         {
             switch (Cmd & 0xffff)
             {
-                case 0x4714: return SetUserData     (Context);
-                case 0x4801: return SetNvMap        (Context);
-                case 0x4808: return SubmitGpfifo    (Context);
-                case 0x4809: return AllocObjCtx     (Context);
-                case 0x480b: return ZcullBind       (Context);
-                case 0x480c: return SetErrorNotifier(Context);
-                case 0x480d: return SetPriority     (Context);
-                case 0x481a: return AllocGpfifoEx2  (Context);
+                case 0x4714: return SetUserData      (Context);
+                case 0x4801: return SetNvMap         (Context);
+                case 0x4803: return SetTimeout       (Context, Channel);
+                case 0x4808: return SubmitGpfifo     (Context);
+                case 0x4809: return AllocObjCtx      (Context);
+                case 0x480b: return ZcullBind        (Context);
+                case 0x480c: return SetErrorNotifier (Context);
+                case 0x480d: return SetPriority      (Context);
+                case 0x481a: return AllocGpfifoEx2   (Context);
+                case 0x481b: return KickoffPbWithAttr(Context);
             }
 
             throw new NotImplementedException(Cmd.ToString("x8"));
@@ -45,6 +72,15 @@ namespace Ryujinx.HLE.OsHle.Services.Nv.NvHostChannel
             return NvResult.Success;
         }
 
+        private static int SetTimeout(ServiceCtx Context, NvChannelName Channel)
+        {
+            long InputPosition = Context.Request.GetBufferType0x21().Position;
+
+            GetChannel(Context, Channel).Timeout = Context.Memory.ReadInt32(InputPosition);
+
+            return NvResult.Success;
+        }
+
         private static int SubmitGpfifo(ServiceCtx Context)
         {
             long InputPosition  = Context.Request.GetBufferType0x21().Position;
@@ -58,15 +94,7 @@ namespace Ryujinx.HLE.OsHle.Services.Nv.NvHostChannel
             {
                 long Gpfifo = Context.Memory.ReadInt64(InputPosition + 0x18 + Index * 8);
 
-                long VA = Gpfifo & 0xff_ffff_ffff;
-
-                int Size = (int)(Gpfifo >> 40) & 0x7ffffc;
-
-                byte[] Data = Vmm.ReadBytes(VA, Size);
-
-                NvGpuPBEntry[] PushBuffer = NvGpuPushBuffer.Decode(Data);
-
-                Context.Ns.Gpu.Fifo.PushBuffer(Vmm, PushBuffer);
+                PushGpfifo(Context, Vmm, Gpfifo);
             }
 
             Args.SyncptId    = 0;
@@ -126,5 +154,57 @@ namespace Ryujinx.HLE.OsHle.Services.Nv.NvHostChannel
 
             return NvResult.Success;
         }
+
+        private static int KickoffPbWithAttr(ServiceCtx Context)
+        {
+            long InputPosition  = Context.Request.GetBufferType0x21().Position;
+            long OutputPosition = Context.Request.GetBufferType0x22().Position;
+
+            NvHostChannelSubmitGpfifo Args = AMemoryHelper.Read<NvHostChannelSubmitGpfifo>(Context.Memory, InputPosition);
+
+            NvGpuVmm Vmm = NvGpuASIoctl.GetVmm(Context);
+
+            for (int Index = 0; Index < Args.NumEntries; Index++)
+            {
+                long Gpfifo = Context.Memory.ReadInt64(Args.Address + Index * 8);
+
+                PushGpfifo(Context, Vmm, Gpfifo);
+            }
+
+            Args.SyncptId    = 0;
+            Args.SyncptValue = 0;
+
+            AMemoryHelper.Write(Context.Memory, OutputPosition, Args);
+
+            return NvResult.Success;
+        }
+
+        private static void PushGpfifo(ServiceCtx Context, NvGpuVmm Vmm, long Gpfifo)
+        {
+            long VA = Gpfifo & 0xff_ffff_ffff;
+
+            int Size = (int)(Gpfifo >> 40) & 0x7ffffc;
+
+            byte[] Data = Vmm.ReadBytes(VA, Size);
+
+            NvGpuPBEntry[] PushBuffer = NvGpuPushBuffer.Decode(Data);
+
+            Context.Ns.Gpu.Fifo.PushBuffer(Vmm, PushBuffer);
+        }
+
+        public static NvChannel GetChannel(ServiceCtx Context, NvChannelName Channel)
+        {
+            ChannelsPerProcess Cpp = Channels.GetOrAdd(Context.Process, (Key) =>
+            {
+                return new ChannelsPerProcess();
+            });
+
+            return Cpp.Channels[Channel];
+        }
+
+        public static void UnloadProcess(Process Process)
+        {
+            Channels.TryRemove(Process, out _);
+        }
     }
 }

Ryujinx.HLE/OsHle/Services/Nv/NvHostChannel/NvHostChannelSubmitGpfifo.cs

@@ -2,7 +2,7 @@ namespace Ryujinx.HLE.OsHle.Services.Nv.NvHostChannel
 {
     struct NvHostChannelSubmitGpfifo
     {
-        public long Gpfifo;
+        public long Address;
         public int  NumEntries;
         public int  Flags;
         public int  SyncptId;

Ryujinx.HLE/OsHle/Services/Nv/NvHostCtrl/NvHostCtrlIoctl.cs

@@ -2,6 +2,7 @@ using ChocolArm64.Memory;
 using Ryujinx.HLE.Logging;
 using System;
 using System.Collections.Concurrent;
+using System.Text;
 using System.Threading;
 
 namespace Ryujinx.HLE.OsHle.Services.Nv.NvHostCtrl
@@ -10,9 +11,16 @@ namespace Ryujinx.HLE.OsHle.Services.Nv.NvHostCtrl
     {
         private static ConcurrentDictionary<Process, NvHostCtrlUserCtx> UserCtxs;
 
+        private static bool IsProductionMode = true;
+
         static NvHostCtrlIoctl()
         {
             UserCtxs = new ConcurrentDictionary<Process, NvHostCtrlUserCtx>();
+
+            if (Set.NxSettings.Settings.TryGetValue("nv!rmos_set_production_mode", out object ProductionModeSetting))
+            {
+                IsProductionMode = ((string)ProductionModeSetting) != "0"; // Default value is ""
+            }
         }
 
         public static int ProcessIoctl(ServiceCtx Context, int Cmd)
@@ -71,17 +79,52 @@ namespace Ryujinx.HLE.OsHle.Services.Nv.NvHostCtrl
 
         private static int GetConfig(ServiceCtx Context)
         {
-            long InputPosition  = Context.Request.GetBufferType0x21().Position;
-            long OutputPosition = Context.Request.GetBufferType0x22().Position;
+            if (!IsProductionMode)
+            {
+                long InputPosition  = Context.Request.GetBufferType0x21().Position;
+                long OutputPosition = Context.Request.GetBufferType0x22().Position;
 
-            string Nv   = AMemoryHelper.ReadAsciiString(Context.Memory, InputPosition + 0,    0x41);
-            string Name = AMemoryHelper.ReadAsciiString(Context.Memory, InputPosition + 0x41, 0x41);
+                string Domain = AMemoryHelper.ReadAsciiString(Context.Memory, InputPosition + 0, 0x41);
+                string Name   = AMemoryHelper.ReadAsciiString(Context.Memory, InputPosition + 0x41, 0x41);
 
-            Context.Memory.WriteByte(OutputPosition + 0x82, 0);
+                if (Set.NxSettings.Settings.TryGetValue($"{Domain}!{Name}", out object NvSetting))
+                {
+                    byte[] SettingBuffer = new byte[0x101];
 
-            Context.Ns.Log.PrintStub(LogClass.ServiceNv, "Stubbed.");
+                    if (NvSetting is string StringValue)
+                    {
+                        if (StringValue.Length > 0x100)
+                        {
+                            Context.Ns.Log.PrintError(Logging.LogClass.ServiceNv, $"{Domain}!{Name} String value size is too big!");
+                        }
+                        else
+                        {
+                            SettingBuffer = Encoding.ASCII.GetBytes(StringValue + "\0");
+                        }
+                    }
 
-            return NvResult.Success;
+                    if (NvSetting is int IntValue)
+                    {
+                        SettingBuffer = BitConverter.GetBytes(IntValue);
+                    }
+                    else if (NvSetting is bool BoolValue)
+                    {
+                        SettingBuffer[0] = BoolValue ? (byte)1 : (byte)0;
+                    }
+                    else
+                    {
+                        throw new NotImplementedException(NvSetting.GetType().Name);
+                    }
+
+                    Context.Memory.WriteBytes(OutputPosition + 0x82, SettingBuffer);
+
+                    Context.Ns.Log.PrintDebug(Logging.LogClass.ServiceNv, $"Got setting {Domain}!{Name}");
+                }
+                
+                return NvResult.Success;
+            }
+
+            return NvResult.NotAvailableInProduction;
         }
 
         private static int EventWait(ServiceCtx Context)
@@ -352,4 +395,4 @@ namespace Ryujinx.HLE.OsHle.Services.Nv.NvHostCtrl
             UserCtxs.TryRemove(Process, out _);
         }
     }
-}
+}

Ryujinx.HLE/OsHle/Services/Nv/NvMap/NvMapIoctl.cs

@@ -48,7 +48,7 @@ namespace Ryujinx.HLE.OsHle.Services.Nv.NvMap
                 return NvResult.InvalidInput;
             }
 
-            int Size = IntUtils.RoundUp(Args.Size, NvGpuVmm.PageSize);
+            int Size = IntUtils.AlignUp(Args.Size, NvGpuVmm.PageSize);
 
             Args.Handle = AddNvMap(Context, new NvMapHandle(Size));
 
@@ -121,7 +121,7 @@ namespace Ryujinx.HLE.OsHle.Services.Nv.NvMap
                 Map.Align =       Args.Align;
                 Map.Kind  = (byte)Args.Kind;
 
-                int Size = IntUtils.RoundUp(Map.Size, NvGpuVmm.PageSize);
+                int Size = IntUtils.AlignUp(Map.Size, NvGpuVmm.PageSize);
 
                 long Address = Args.Address;
 

Ryujinx.HLE/OsHle/Services/Nv/NvResult.cs

@@ -2,12 +2,13 @@ namespace Ryujinx.HLE.OsHle.Services.Nv
 {
     static class NvResult
     {
-        public const int Success      = 0;
-        public const int TryAgain     = -11;
-        public const int OutOfMemory  = -12;
-        public const int InvalidInput = -22;
-        public const int NotSupported = -25;
-        public const int Restart      = -85;
-        public const int TimedOut     = -110;
+        public const int NotAvailableInProduction = 196614;
+        public const int Success                  = 0;
+        public const int TryAgain                 = -11;
+        public const int OutOfMemory              = -12;
+        public const int InvalidInput             = -22;
+        public const int NotSupported             = -25;
+        public const int Restart                  = -85;
+        public const int TimedOut                 = -110;
     }
-}
+}

Ryujinx.HLE/OsHle/Services/ServiceFactory.cs

@@ -18,6 +18,7 @@ using Ryujinx.HLE.OsHle.Services.Prepo;
 using Ryujinx.HLE.OsHle.Services.Set;
 using Ryujinx.HLE.OsHle.Services.Sfdnsres;
 using Ryujinx.HLE.OsHle.Services.Sm;
+using Ryujinx.HLE.OsHle.Services.Spl;
 using Ryujinx.HLE.OsHle.Services.Ssl;
 using Ryujinx.HLE.OsHle.Services.Vi;
 using System;
@@ -66,6 +67,9 @@ namespace Ryujinx.HLE.OsHle.Services
                 case "caps:ss":
                     return new IScreenshotService();
 
+                case "csrng":
+                    return new IRandomInterface();
+
                 case "friend:a":
                     return new IServiceCreator();
 

Ryujinx.HLE/OsHle/Services/Spl/IRandomInterface.cs

@@ -0,0 +1,50 @@
+using Ryujinx.HLE.OsHle.Ipc;
+using System;
+using System.Collections.Generic;
+using System.Security.Cryptography;
+
+namespace Ryujinx.HLE.OsHle.Services.Spl
+{
+    class IRandomInterface : IpcService, IDisposable
+    {
+        private Dictionary<int, ServiceProcessRequest> m_Commands;
+
+        public override IReadOnlyDictionary<int, ServiceProcessRequest> Commands => m_Commands;
+
+        private RNGCryptoServiceProvider Rng;
+
+        public IRandomInterface()
+        {
+            m_Commands = new Dictionary<int, ServiceProcessRequest>()
+            {
+                { 0, GetRandomBytes }
+            };
+
+            Rng = new RNGCryptoServiceProvider();
+        }
+
+        public long GetRandomBytes(ServiceCtx Context)
+        {
+            byte[] RandomBytes = new byte[Context.Request.ReceiveBuff[0].Size];
+
+            Rng.GetBytes(RandomBytes);
+
+            Context.Memory.WriteBytes(Context.Request.ReceiveBuff[0].Position, RandomBytes);
+
+            return 0;
+        }
+
+        public void Dispose()
+        {
+            Dispose(true);
+        }
+
+        protected virtual void Dispose(bool Disposing)
+        {
+            if (Disposing)
+            {
+                Rng.Dispose();
+            }
+        }
+    }
+}

Ryujinx.HLE/OsHle/Services/Time/ITimeZoneService.cs

@@ -2,6 +2,7 @@ using Ryujinx.HLE.Logging;
 using Ryujinx.HLE.OsHle.Ipc;
 using System;
 using System.Collections.Generic;
+using System.Text;
 
 namespace Ryujinx.HLE.OsHle.Services.Time
 {
@@ -13,20 +14,31 @@ namespace Ryujinx.HLE.OsHle.Services.Time
 
         private static readonly DateTime Epoch = new DateTime(1970, 1, 1, 0, 0, 0, DateTimeKind.Utc);
 
+        private TimeZoneInfo TimeZone = TimeZoneInfo.Local;
+
         public ITimeZoneService()
         {
             m_Commands = new Dictionary<int, ServiceProcessRequest>()
             {
-                { 0,   GetDeviceLocationName    },
-                { 101, ToCalendarTimeWithMyRule }
+                { 0,   GetDeviceLocationName     },
+                { 1,   SetDeviceLocationName     },
+                { 2,   GetTotalLocationNameCount },
+                { 3,   LoadLocationNameList      },
+                { 4,   LoadTimeZoneRule          },
+                { 100, ToCalendarTime            },
+                { 101, ToCalendarTimeWithMyRule  }
             };
         }
 
         public long GetDeviceLocationName(ServiceCtx Context)
         {
-            Context.Ns.Log.PrintStub(LogClass.ServiceTime, "Stubbed.");
+            char[] TzName = TimeZone.Id.ToCharArray();
 
-            for (int Index = 0; Index < 0x24; Index++)
+            Context.ResponseData.Write(TzName);
+
+            int Padding = 0x24 - TzName.Length;
+
+            for (int Index = 0; Index < Padding; Index++)
             {
                 Context.ResponseData.Write((byte)0);
             }
@@ -34,11 +46,94 @@ namespace Ryujinx.HLE.OsHle.Services.Time
             return 0;
         }
 
-        public long ToCalendarTimeWithMyRule(ServiceCtx Context)
+        public long SetDeviceLocationName(ServiceCtx Context)
         {
-            long PosixTime = Context.RequestData.ReadInt64();
+            byte[] LocationName = Context.RequestData.ReadBytes(0x24);
+            string TzID         = Encoding.ASCII.GetString(LocationName).TrimEnd('\0');
 
-            DateTime CurrentTime = Epoch.AddSeconds(PosixTime).ToLocalTime();
+            long ResultCode = 0;
+
+            try
+            {
+                TimeZone = TimeZoneInfo.FindSystemTimeZoneById(TzID);
+            }
+            catch (TimeZoneNotFoundException e)
+            {
+                ResultCode = 0x7BA74;
+            }
+
+            return ResultCode;
+        }
+
+        public long GetTotalLocationNameCount(ServiceCtx Context)
+        {
+            Context.ResponseData.Write(TimeZoneInfo.GetSystemTimeZones().Count);
+
+            return 0;
+        }
+
+        public long LoadLocationNameList(ServiceCtx Context)
+        {
+            long BufferPosition = Context.Response.SendBuff[0].Position;
+            long BufferSize     = Context.Response.SendBuff[0].Size;
+
+            int i = 0;
+            foreach (TimeZoneInfo info in TimeZoneInfo.GetSystemTimeZones())
+            {
+                byte[] TzData = Encoding.ASCII.GetBytes(info.Id);
+
+                Context.Memory.WriteBytes(BufferPosition + i, TzData);
+
+                int Padding = 0x24 - TzData.Length;
+
+                for (int Index = 0; Index < Padding; Index++)
+                {
+                    Context.ResponseData.Write((byte)0);
+                }
+
+                i += 0x24;
+            }
+            return 0;
+        }
+
+        public long LoadTimeZoneRule(ServiceCtx Context)
+        {
+            long BufferPosition = Context.Request.ReceiveBuff[0].Position;
+            long BufferSize     = Context.Request.ReceiveBuff[0].Size;
+
+            if (BufferSize != 0x4000)
+            {
+                Context.Ns.Log.PrintWarning(LogClass.ServiceTime, $"TimeZoneRule buffer size is 0x{BufferSize:x} (expected 0x4000)");
+            }
+
+            long ResultCode = 0;
+
+            byte[] LocationName = Context.RequestData.ReadBytes(0x24);
+            string TzID         = Encoding.ASCII.GetString(LocationName).TrimEnd('\0');
+
+            // Check if the Time Zone exists, otherwise error out.
+            try
+            {
+                TimeZoneInfo Info = TimeZoneInfo.FindSystemTimeZoneById(TzID);
+                byte[] TzData     = Encoding.ASCII.GetBytes(Info.Id);
+
+                // FIXME: This is not in ANY cases accurate, but the games don't about the content of the buffer, they only pass it.
+                // TODO: Reverse the TZif2 conversion in PCV to make this match with real hardware.
+                Context.Memory.WriteBytes(BufferPosition, TzData);
+            }
+            catch (TimeZoneNotFoundException e)
+            {
+                Context.Ns.Log.PrintWarning(LogClass.ServiceTime, $"Timezone not found for string: {TzID} (len: {TzID.Length})");
+                ResultCode = 0x7BA74;
+            }
+
+            return ResultCode;
+        }
+
+        private long ToCalendarTimeWithTz(ServiceCtx Context, long PosixTime, TimeZoneInfo Info)
+        {
+            DateTime CurrentTime = Epoch.AddSeconds(PosixTime);
+            CurrentTime          = TimeZoneInfo.ConvertTimeFromUtc(CurrentTime, Info);
 
             Context.ResponseData.Write((ushort)CurrentTime.Year);
             Context.ResponseData.Write((byte)CurrentTime.Month);
@@ -46,31 +141,54 @@ namespace Ryujinx.HLE.OsHle.Services.Time
             Context.ResponseData.Write((byte)CurrentTime.Hour);
             Context.ResponseData.Write((byte)CurrentTime.Minute);
             Context.ResponseData.Write((byte)CurrentTime.Second);
-            Context.ResponseData.Write((byte)0);
-
-            /* Thanks to TuxSH
-                struct CalendarAdditionalInfo {
-	                u32 tm_wday; //day of week [0,6] (Sunday = 0)
-	                s32 tm_yday; //day of year [0,365]
-	                struct timezone {
-		                char[8] tz_name;
-		                bool isDaylightSavingTime;
-		                s32 utcOffsetSeconds;
-	                };
-                };
-            */
+            Context.ResponseData.Write((byte)0); //MilliSecond ?
             Context.ResponseData.Write((int)CurrentTime.DayOfWeek);
-
             Context.ResponseData.Write(CurrentTime.DayOfYear - 1);
-
-            //TODO: Find out the names used.
-            Context.ResponseData.Write(new byte[8]);
-
+            Context.ResponseData.Write(new byte[8]); //TODO: Find out the names used.
             Context.ResponseData.Write((byte)(CurrentTime.IsDaylightSavingTime() ? 1 : 0));
-
-            Context.ResponseData.Write((int)TimeZoneInfo.Local.GetUtcOffset(CurrentTime).TotalSeconds);
+            Context.ResponseData.Write((int)Info.GetUtcOffset(CurrentTime).TotalSeconds);
 
             return 0;
         }
+
+        public long ToCalendarTime(ServiceCtx Context)
+        {
+            long PosixTime      = Context.RequestData.ReadInt64();
+            long BufferPosition = Context.Request.SendBuff[0].Position;
+            long BufferSize     = Context.Request.SendBuff[0].Size;
+
+            if (BufferSize != 0x4000)
+            {
+                Context.Ns.Log.PrintWarning(LogClass.ServiceTime, $"TimeZoneRule buffer size is 0x{BufferSize:x} (expected 0x4000)");
+            }
+
+            // TODO: Reverse the TZif2 conversion in PCV to make this match with real hardware.
+            byte[] TzData = Context.Memory.ReadBytes(BufferPosition, 0x24);
+            string TzID   = Encoding.ASCII.GetString(TzData).TrimEnd('\0');
+
+            long ResultCode = 0;
+
+            // Check if the Time Zone exists, otherwise error out.
+            try
+            {
+                TimeZoneInfo Info = TimeZoneInfo.FindSystemTimeZoneById(TzID);
+
+                ResultCode = ToCalendarTimeWithTz(Context, PosixTime, Info);
+            }
+            catch (TimeZoneNotFoundException e)
+            {
+                Context.Ns.Log.PrintWarning(LogClass.ServiceTime, $"Timezone not found for string: {TzID} (len: {TzID.Length})");
+                ResultCode = 0x7BA74;
+            }
+
+            return ResultCode;
+        }
+
+        public long ToCalendarTimeWithMyRule(ServiceCtx Context)
+        {
+            long PosixTime = Context.RequestData.ReadInt64();
+
+            return ToCalendarTimeWithTz(Context, PosixTime, TimeZone);
+        }
     }
-}
+}

Ryujinx.HLE/OsHle/Utilities/IntUtils.cs

@@ -2,12 +2,12 @@ namespace Ryujinx.HLE.OsHle.Utilities
 {
     static class IntUtils
     {
-        public static int RoundUp(int Value, int Size)
+        public static int AlignUp(int Value, int Size)
         {
             return (Value + (Size - 1)) & ~(Size - 1);
         }
 
-        public static long RoundUp(long Value, int Size)
+        public static long AlignUp(long Value, int Size)
         {
             return (Value + (Size - 1)) & ~((long)Size - 1);
         }

Ryujinx.HLE/OsHle/Utilities/StructReader.cs

@@ -0,0 +1,45 @@
+using ChocolArm64.Memory;
+using System.Runtime.InteropServices;
+
+namespace Ryujinx.HLE.OsHle.Utilities
+{
+    class StructReader
+    {
+        private AMemory Memory;
+
+        public long Position { get; private set; }
+
+        public StructReader(AMemory Memory, long Position)
+        {
+            this.Memory   = Memory;
+            this.Position = Position;
+        }
+
+        public T Read<T>() where T : struct
+        {
+            T Value = AMemoryHelper.Read<T>(Memory, Position);
+
+            Position += Marshal.SizeOf<T>();
+
+            return Value;
+        }
+
+        public T[] Read<T>(int Size) where T : struct
+        {
+            int StructSize = Marshal.SizeOf<T>();
+
+            int Count = Size / StructSize;
+
+            T[] Output = new T[Count];
+
+            for (int Index = 0; Index < Count; Index++)
+            {
+                Output[Index] = AMemoryHelper.Read<T>(Memory, Position);
+
+                Position += StructSize;
+            }
+
+            return Output;
+        }
+    }
+}

Ryujinx.HLE/OsHle/Utilities/StructWriter.cs

@@ -0,0 +1,25 @@
+using ChocolArm64.Memory;
+using System.Runtime.InteropServices;
+
+namespace Ryujinx.HLE.OsHle.Utilities
+{
+    class StructWriter
+    {
+        private AMemory Memory;
+
+        public long Position { get; private set; }
+
+        public StructWriter(AMemory Memory, long Position)
+        {
+            this.Memory   = Memory;
+            this.Position = Position;
+        }
+
+        public void Write<T>(T Value) where T : struct
+        {
+            AMemoryHelper.Write(Memory, Position, Value);
+
+            Position += Marshal.SizeOf<T>();
+        }
+    }
+}

Ryujinx.HLE/Switch.cs

@@ -71,6 +71,11 @@ namespace Ryujinx.HLE
             Os.LoadProgram(FileName);
         }
 
+        public bool WaitFifo()
+        {
+            return Gpu.Fifo.Event.WaitOne(8);
+        }
+
         public void ProcessFrame()
         {
             Gpu.Fifo.DispatchCalls();

Ryujinx.Tests/Cpu/CpuTestSimdArithmetic.cs

@@ -9,46 +9,6 @@ namespace Ryujinx.Tests.Cpu
 {
     public class CpuTestSimdArithmetic : CpuTest
     {
-        [TestCase(0xE228420u,   0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul)]
-        [TestCase(0xE228420u,   0x00000000FFFFFFFFul, 0x00000000FFFFFFFFul, 0x0000000000000001ul, 0x0000000000000001ul, 0x00000000FFFFFF00ul, 0x0000000000000000ul)]
-        [TestCase(0xE228420u,   0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFEFEFEFEFEFEFEFEul, 0x0000000000000000ul)]
-        [TestCase(0xE228420u,   0x0102030405060708ul, 0xAAAAAAAAAAAAAAAAul, 0x0807060504030201ul, 0x2222222222222222ul, 0x0909090909090909ul, 0x0000000000000000ul)]
-        [TestCase(0x4E228420u,  0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul)]
-        [TestCase(0x4E228420u,  0x00000000FFFFFFFFul, 0x00000000FFFFFFFFul, 0x0000000000000001ul, 0x0000000000000001ul, 0x00000000FFFFFF00ul, 0x00000000FFFFFF00ul)]
-        [TestCase(0x4E228420u,  0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFEFEFEFEFEFEFEFEul, 0xFEFEFEFEFEFEFEFEul)]
-        [TestCase(0x4E228420u,  0x0102030405060708ul, 0xAAAAAAAAAAAAAAAAul, 0x0807060504030201ul, 0x2222222222222222ul, 0x0909090909090909ul, 0xCCCCCCCCCCCCCCCCul)]
-        [TestCase(0xE628420u,   0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul)]
-        [TestCase(0xE628420u,   0x00000000FFFFFFFFul, 0x00000000FFFFFFFFul, 0x0000000000000001ul, 0x0000000000000001ul, 0x00000000FFFF0000ul, 0x0000000000000000ul)]
-        [TestCase(0xE628420u,   0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFEFFFEFFFEFFFEul, 0x0000000000000000ul)]
-        [TestCase(0xE628420u,   0x0102030405060708ul, 0xAAAAAAAAAAAAAAAAul, 0x0807060504030201ul, 0x2222222222222222ul, 0x0909090909090909ul, 0x0000000000000000ul)]
-        [TestCase(0x4E628420u,  0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul)]
-        [TestCase(0x4E628420u,  0x00000000FFFFFFFFul, 0x00000000FFFFFFFFul, 0x0000000000000001ul, 0x0000000000000001ul, 0x00000000FFFF0000ul, 0x00000000FFFF0000ul)]
-        [TestCase(0x4E628420u,  0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFEFFFEFFFEFFFEul, 0xFFFEFFFEFFFEFFFEul)]
-        [TestCase(0x4E628420u,  0x0102030405060708ul, 0xAAAAAAAAAAAAAAAAul, 0x0807060504030201ul, 0x2222222222222222ul, 0x0909090909090909ul, 0xCCCCCCCCCCCCCCCCul)]
-        [TestCase(0xEA28420u,   0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul)]
-        [TestCase(0xEA28420u,   0x00000000FFFFFFFFul, 0x00000000FFFFFFFFul, 0x0000000000000001ul, 0x0000000000000001ul, 0x0000000000000000ul, 0x0000000000000000ul)]
-        [TestCase(0xEA28420u,   0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFEFFFFFFFEul, 0x0000000000000000ul)]
-        [TestCase(0xEA28420u,   0x0102030405060708ul, 0xAAAAAAAAAAAAAAAAul, 0x0807060504030201ul, 0x2222222222222222ul, 0x0909090909090909ul, 0x0000000000000000ul)]
-        [TestCase(0x4EA28420u,  0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul)]
-        [TestCase(0x4EA28420u,  0x00000000FFFFFFFFul, 0x00000000FFFFFFFFul, 0x0000000000000001ul, 0x0000000000000001ul, 0x0000000000000000ul, 0x0000000000000000ul)]
-        [TestCase(0x4EA28420u,  0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFEFFFFFFFEul, 0xFFFFFFFEFFFFFFFEul)]
-        [TestCase(0x4EA28420u,  0x0102030405060708ul, 0xAAAAAAAAAAAAAAAAul, 0x0807060504030201ul, 0x2222222222222222ul, 0x0909090909090909ul, 0xCCCCCCCCCCCCCCCCul)]
-        [TestCase(0x4EE28420u,  0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul, 0x0000000000000000ul)]
-        [TestCase(0x4EE28420u,  0x00000000FFFFFFFFul, 0x00000000FFFFFFFFul, 0x0000000000000001ul, 0x0000000000000001ul, 0x0000000100000000ul, 0x0000000100000000ul)]
-        [TestCase(0x4EE28420u,  0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFFul, 0xFFFFFFFFFFFFFFFEul, 0xFFFFFFFFFFFFFFFEul)]
-        [TestCase(0x4EE28420u,  0x0102030405060708ul, 0xAAAAAAAAAAAAAAAAul, 0x0807060504030201ul, 0x2222222222222222ul, 0x0909090909090909ul, 0xCCCCCCCCCCCCCCCCul)]
-        public void Add_V(uint Opcode, ulong A0, ulong A1, ulong B0, ulong B1, ulong Result0, ulong Result1)
-        {
-            Vector128<float> V1 = MakeVectorE0E1(A0, A1);
-            Vector128<float> V2 = MakeVectorE0E1(B0, B1);
-            AThreadState ThreadState = SingleOpcode(Opcode, V1: V1, V2: V2);
-            Assert.Multiple(() =>
-            {
-                Assert.AreEqual(Result0, GetVectorE0(ThreadState.V0));
-                Assert.AreEqual(Result1, GetVectorE1(ThreadState.V0));
-            });
-        }
-
         [TestCase(0x1E224820u, 0x0000000000000000ul, 0x0000000080000000ul, 0x0000000000000000ul)]
         [TestCase(0x1E224820u, 0x0000000080000000ul, 0x0000000000000000ul, 0x0000000000000000ul)]
         [TestCase(0x1E224820u, 0x0000000080000000ul, 0x0000000080000000ul, 0x0000000080000000ul)]
@@ -163,26 +123,18 @@ namespace Ryujinx.Tests.Cpu
             Assert.That(Sse41.Extract(ThreadState.V6, (byte)0), Is.EqualTo(A * B));
         }
 
-        [Test, Description("FRECPE D0, D1")]
-        public void Frecpe_S([Random(100)] double A)
+        [TestCase(0x00000000u, 0x7F800000u)]
+        [TestCase(0x80000000u, 0xFF800000u)]
+        [TestCase(0x00FFF000u, 0x7E000000u)]
+        [TestCase(0x41200000u, 0x3DCC8000u)]
+        [TestCase(0xC1200000u, 0xBDCC8000u)]
+        [TestCase(0x001FFFFFu, 0x7F800000u)]
+        [TestCase(0x007FF000u, 0x7E800000u)]
+        public void Frecpe_S(uint A, uint Result)
         {
-            AThreadState ThreadState = SingleOpcode(0x5EE1D820, V1: MakeVectorE0(A));
-
-            Assert.That(VectorExtractDouble(ThreadState.V0, 0), Is.EqualTo(1 / A));
-        }
-
-        [Test, Description("FRECPE V2.4S, V0.4S")]
-        public void Frecpe_V([Random(100)] float A)
-        {
-            AThreadState ThreadState = SingleOpcode(0x4EA1D802, V0: Sse.SetAllVector128(A));
-
-            Assert.Multiple(() =>
-            {
-                Assert.That(Sse41.Extract(ThreadState.V2, (byte)0), Is.EqualTo(1 / A));
-                Assert.That(Sse41.Extract(ThreadState.V2, (byte)1), Is.EqualTo(1 / A));
-                Assert.That(Sse41.Extract(ThreadState.V2, (byte)2), Is.EqualTo(1 / A));
-                Assert.That(Sse41.Extract(ThreadState.V2, (byte)3), Is.EqualTo(1 / A));
-            });
+            Vector128<float> V1 = MakeVectorE0(A);
+            AThreadState ThreadState = SingleOpcode(0x5EA1D820, V1: V1);
+            Assert.AreEqual(Result, GetVectorE0(ThreadState.V0));
         }
 
         [Test, Description("FRECPS D0, D1, D2")]
@@ -202,12 +154,14 @@ namespace Ryujinx.Tests.Cpu
                 V2: Sse.SetAllVector128(A),
                 V0: Sse.SetAllVector128(B));
 
+            float Result = (float)(2 - ((double)A * (double)B));
+
             Assert.Multiple(() =>
             {
-                Assert.That(Sse41.Extract(ThreadState.V4, (byte)0), Is.EqualTo(2 - (A * B)));
-                Assert.That(Sse41.Extract(ThreadState.V4, (byte)1), Is.EqualTo(2 - (A * B)));
-                Assert.That(Sse41.Extract(ThreadState.V4, (byte)2), Is.EqualTo(2 - (A * B)));
-                Assert.That(Sse41.Extract(ThreadState.V4, (byte)3), Is.EqualTo(2 - (A * B)));
+                Assert.That(Sse41.Extract(ThreadState.V4, (byte)0), Is.EqualTo(Result));
+                Assert.That(Sse41.Extract(ThreadState.V4, (byte)1), Is.EqualTo(Result));
+                Assert.That(Sse41.Extract(ThreadState.V4, (byte)2), Is.EqualTo(Result));
+                Assert.That(Sse41.Extract(ThreadState.V4, (byte)3), Is.EqualTo(Result));
             });
         }
 

Ryujinx.Tests/Cpu/CpuTestSimdCvt.cs

@@ -0,0 +1,40 @@
+using ChocolArm64.State;
+
+using NUnit.Framework;
+
+using System.Runtime.Intrinsics;
+using System.Runtime.Intrinsics.X86;
+
+namespace Ryujinx.Tests.Cpu
+{
+    public class CpuTestSimdCvt : CpuTest
+    {
+        [TestCase((ushort)0x0000, 0x00000000u)] // Positive Zero
+        [TestCase((ushort)0x8000, 0x80000000u)] // Negative Zero
+        [TestCase((ushort)0x3E00, 0x3FC00000u)] // +1.5
+        [TestCase((ushort)0xBE00, 0xBFC00000u)] // -1.5
+        [TestCase((ushort)0xFFFF, 0xFFFFE000u)] // -QNaN
+        [TestCase((ushort)0x7C00, 0x7F800000u)] // +Inf
+        [TestCase((ushort)0x3C00, 0x3F800000u)] // 1.0
+        [TestCase((ushort)0x3C01, 0x3F802000u)] // 1.0009765625
+        [TestCase((ushort)0xC000, 0xC0000000u)] // -2.0
+        [TestCase((ushort)0x7BFF, 0x477FE000u)] // 65504.0 (Largest Normal)
+        [TestCase((ushort)0x03FF, 0x387FC000u)] // 0.00006097555 (Largest Subnormal)
+        [TestCase((ushort)0x0001, 0x33800000u)] // 5.96046448e-8 (Smallest Subnormal)
+        public void Fcvtl_V_f16(ushort Value, uint Result)
+        {
+            uint Opcode = 0x0E217801;
+            Vector128<float> V0 = Sse.StaticCast<ushort, float>(Sse2.SetAllVector128(Value));
+
+            AThreadState ThreadState = SingleOpcode(Opcode, V0: V0);
+
+            Assert.Multiple(() =>
+            {
+                Assert.That(Sse41.Extract(Sse.StaticCast<float, uint>(ThreadState.V1), (byte)0), Is.EqualTo(Result));
+                Assert.That(Sse41.Extract(Sse.StaticCast<float, uint>(ThreadState.V1), (byte)1), Is.EqualTo(Result));
+                Assert.That(Sse41.Extract(Sse.StaticCast<float, uint>(ThreadState.V1), (byte)2), Is.EqualTo(Result));
+                Assert.That(Sse41.Extract(Sse.StaticCast<float, uint>(ThreadState.V1), (byte)3), Is.EqualTo(Result));
+            });
+        }
+    }
+}

Ryujinx.Tests/Cpu/CpuTestSimdMove.cs

@@ -1,136 +0,0 @@
-using ChocolArm64.State;
-
-using NUnit.Framework;
-
-using System.Runtime.Intrinsics;
-using System.Runtime.Intrinsics.X86;
-
-namespace Ryujinx.Tests.Cpu
-{
-    public class CpuTestSimdMove : CpuTest
-    {
-        [Test, Description("TRN1 V0.4S, V1.4S, V2.4S")]
-        public void Trn1_V_4S([Random(2)] uint A0, [Random(2)] uint A1, [Random(2)] uint A2, [Random(2)] uint A3,
-                              [Random(2)] uint B0, [Random(2)] uint B1, [Random(2)] uint B2, [Random(2)] uint B3)
-        {
-            uint Opcode = 0x4E822820;
-            Vector128<float> V1 = Sse.StaticCast<uint, float>(Sse2.SetVector128(A3, A2, A1, A0));
-            Vector128<float> V2 = Sse.StaticCast<uint, float>(Sse2.SetVector128(B3, B2, B1, B0));
-
-            AThreadState ThreadState = SingleOpcode(Opcode, V1: V1, V2: V2);
-
-            Assert.Multiple(() =>
-            {
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, uint>(ThreadState.V0), (byte)0), Is.EqualTo(A0));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, uint>(ThreadState.V0), (byte)1), Is.EqualTo(B0));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, uint>(ThreadState.V0), (byte)2), Is.EqualTo(A2));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, uint>(ThreadState.V0), (byte)3), Is.EqualTo(B2));
-            });
-        }
-
-        [Test, Description("TRN1 V0.8B, V1.8B, V2.8B")]
-        public void Trn1_V_8B([Random(2)] byte A0, [Random(1)] byte A1, [Random(2)] byte A2, [Random(1)] byte A3,
-                              [Random(2)] byte A4, [Random(1)] byte A5, [Random(2)] byte A6, [Random(1)] byte A7,
-                              [Random(2)] byte B0, [Random(1)] byte B1, [Random(2)] byte B2, [Random(1)] byte B3,
-                              [Random(2)] byte B4, [Random(1)] byte B5, [Random(2)] byte B6, [Random(1)] byte B7)
-        {
-            uint Opcode = 0x0E022820;
-            Vector128<float> V1 = Sse.StaticCast<byte, float>(Sse2.SetVector128(0, 0, 0, 0, 0, 0, 0, 0, A7, A6, A5, A4, A3, A2, A1, A0));
-            Vector128<float> V2 = Sse.StaticCast<byte, float>(Sse2.SetVector128(0, 0, 0, 0, 0, 0, 0, 0, B7, B6, B5, B4, B3, B2, B1, B0));
-
-            AThreadState ThreadState = SingleOpcode(Opcode, V1: V1, V2: V2);
-
-            Assert.Multiple(() =>
-            {
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)0), Is.EqualTo(A0));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)1), Is.EqualTo(B0));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)2), Is.EqualTo(A2));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)3), Is.EqualTo(B2));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)4), Is.EqualTo(A4));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)5), Is.EqualTo(B4));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)6), Is.EqualTo(A6));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)7), Is.EqualTo(B6));
-            });
-        }
-
-        [Test, Description("TRN2 V0.4S, V1.4S, V2.4S")]
-        public void Trn2_V_4S([Random(2)] uint A0, [Random(2)] uint A1, [Random(2)] uint A2, [Random(2)] uint A3,
-                              [Random(2)] uint B0, [Random(2)] uint B1, [Random(2)] uint B2, [Random(2)] uint B3)
-        {
-            uint Opcode = 0x4E826820;
-            Vector128<float> V1 = Sse.StaticCast<uint, float>(Sse2.SetVector128(A3, A2, A1, A0));
-            Vector128<float> V2 = Sse.StaticCast<uint, float>(Sse2.SetVector128(B3, B2, B1, B0));
-
-            AThreadState ThreadState = SingleOpcode(Opcode, V1: V1, V2: V2);
-
-            Assert.Multiple(() =>
-            {
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, uint>(ThreadState.V0), (byte)0), Is.EqualTo(A1));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, uint>(ThreadState.V0), (byte)1), Is.EqualTo(B1));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, uint>(ThreadState.V0), (byte)2), Is.EqualTo(A3));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, uint>(ThreadState.V0), (byte)3), Is.EqualTo(B3));
-            });
-        }
-
-        [Test, Description("TRN2 V0.8B, V1.8B, V2.8B")]
-        public void Trn2_V_8B([Random(1)] byte A0, [Random(2)] byte A1, [Random(1)] byte A2, [Random(2)] byte A3,
-                              [Random(1)] byte A4, [Random(2)] byte A5, [Random(1)] byte A6, [Random(2)] byte A7,
-                              [Random(1)] byte B0, [Random(2)] byte B1, [Random(1)] byte B2, [Random(2)] byte B3,
-                              [Random(1)] byte B4, [Random(2)] byte B5, [Random(1)] byte B6, [Random(2)] byte B7)
-        {
-            uint Opcode = 0x0E026820;
-            Vector128<float> V1 = Sse.StaticCast<byte, float>(Sse2.SetVector128(0, 0, 0, 0, 0, 0, 0, 0, A7, A6, A5, A4, A3, A2, A1, A0));
-            Vector128<float> V2 = Sse.StaticCast<byte, float>(Sse2.SetVector128(0, 0, 0, 0, 0, 0, 0, 0, B7, B6, B5, B4, B3, B2, B1, B0));
-
-            AThreadState ThreadState = SingleOpcode(Opcode, V1: V1, V2: V2);
-
-            Assert.Multiple(() =>
-            {
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)0), Is.EqualTo(A1));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)1), Is.EqualTo(B1));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)2), Is.EqualTo(A3));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)3), Is.EqualTo(B3));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)4), Is.EqualTo(A5));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)5), Is.EqualTo(B5));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)6), Is.EqualTo(A7));
-                Assert.That(Sse41.Extract(Sse.StaticCast<float, byte>(ThreadState.V0), (byte)7), Is.EqualTo(B7));
-            });
-        }
-
-        [TestCase(0u, 0u, 0x2313221221112010ul, 0x0000000000000000ul)]
-        [TestCase(1u, 0u, 0x2313221221112010ul, 0x2717261625152414ul)]
-        [TestCase(0u, 1u, 0x2322131221201110ul, 0x0000000000000000ul)]
-        [TestCase(1u, 1u, 0x2322131221201110ul, 0x2726171625241514ul)]
-        [TestCase(0u, 2u, 0x2322212013121110ul, 0x0000000000000000ul)]
-        [TestCase(1u, 2u, 0x2322212013121110ul, 0x2726252417161514ul)]
-        [TestCase(1u, 3u, 0x1716151413121110ul, 0x2726252423222120ul)]
-        public void Zip1_V(uint Q, uint size, ulong Result_0, ulong Result_1)
-        {
-            // ZIP1 V0.<T>, V1.<T>, V2.<T>
-            uint Opcode = 0x0E023820 | (Q << 30) | (size << 22);
-            Vector128<float> V1 = MakeVectorE0E1(0x1716151413121110, 0x1F1E1D1C1B1A1918);
-            Vector128<float> V2 = MakeVectorE0E1(0x2726252423222120, 0x2F2E2D2C2B2A2928);
-            AThreadState ThreadState = SingleOpcode(Opcode, V1: V1, V2: V2);
-            Assert.AreEqual(Result_0, GetVectorE0(ThreadState.V0));
-            Assert.AreEqual(Result_1, GetVectorE1(ThreadState.V0));
-        }
-
-        [TestCase(0u, 0u, 0x2717261625152414ul, 0x0000000000000000ul)]
-        [TestCase(1u, 0u, 0x2B1B2A1A29192818ul, 0x2F1F2E1E2D1D2C1Cul)]
-        [TestCase(0u, 1u, 0x2726171625241514ul, 0x0000000000000000ul)]
-        [TestCase(1u, 1u, 0x2B2A1B1A29281918ul, 0x2F2E1F1E2D2C1D1Cul)]
-        [TestCase(0u, 2u, 0x2726252417161514ul, 0x0000000000000000ul)]
-        [TestCase(1u, 2u, 0x2B2A29281B1A1918ul, 0x2F2E2D2C1F1E1D1Cul)]
-        [TestCase(1u, 3u, 0x1F1E1D1C1B1A1918ul, 0x2F2E2D2C2B2A2928ul)]
-        public void Zip2_V(uint Q, uint size, ulong Result_0, ulong Result_1)
-        {
-            // ZIP2 V0.<T>, V1.<T>, V2.<T>
-            uint Opcode = 0x0E027820 | (Q << 30) | (size << 22);
-            Vector128<float> V1 = MakeVectorE0E1(0x1716151413121110, 0x1F1E1D1C1B1A1918);
-            Vector128<float> V2 = MakeVectorE0E1(0x2726252423222120, 0x2F2E2D2C2B2A2928);
-            AThreadState ThreadState = SingleOpcode(Opcode, V1: V1, V2: V2);
-            Assert.AreEqual(Result_0, GetVectorE0(ThreadState.V0));
-            Assert.AreEqual(Result_1, GetVectorE1(ThreadState.V0));
-        }
-    }
-}

Ryujinx.Tests/Cpu/Tester/Instructions.cs

@@ -4655,6 +4655,74 @@ namespace Ryujinx.Tests.Cpu.Tester
             Vpart(d, part, result);
         }
 
+        // trn1_advsimd.html
+        public static void Trn1_V(bool Q, Bits size, Bits Rm, Bits Rn, Bits Rd)
+        {
+            const bool op = false;
+
+            /* Decode */
+            int d = (int)UInt(Rd);
+            int n = (int)UInt(Rn);
+            int m = (int)UInt(Rm);
+
+            /* if size:Q == '110' then ReservedValue(); */
+
+            int esize = 8 << (int)UInt(size);
+            int datasize = (Q ? 128 : 64);
+            int elements = datasize / esize;
+            int part = (int)UInt(op);
+            int pairs = elements / 2;
+
+            /* Operation */
+            /* CheckFPAdvSIMDEnabled64(); */
+
+            Bits result = new Bits(datasize);
+            Bits operand1 = V(datasize, n);
+            Bits operand2 = V(datasize, m);
+
+            for (int p = 0; p <= pairs - 1; p++)
+            {
+                Elem(result, 2 * p + 0, esize, Elem(operand1, 2 * p + part, esize));
+                Elem(result, 2 * p + 1, esize, Elem(operand2, 2 * p + part, esize));
+            }
+
+            V(d, result);
+        }
+
+        // trn2_advsimd.html
+        public static void Trn2_V(bool Q, Bits size, Bits Rm, Bits Rn, Bits Rd)
+        {
+            const bool op = true;
+
+            /* Decode */
+            int d = (int)UInt(Rd);
+            int n = (int)UInt(Rn);
+            int m = (int)UInt(Rm);
+
+            /* if size:Q == '110' then ReservedValue(); */
+
+            int esize = 8 << (int)UInt(size);
+            int datasize = (Q ? 128 : 64);
+            int elements = datasize / esize;
+            int part = (int)UInt(op);
+            int pairs = elements / 2;
+
+            /* Operation */
+            /* CheckFPAdvSIMDEnabled64(); */
+
+            Bits result = new Bits(datasize);
+            Bits operand1 = V(datasize, n);
+            Bits operand2 = V(datasize, m);
+
+            for (int p = 0; p <= pairs - 1; p++)
+            {
+                Elem(result, 2 * p + 0, esize, Elem(operand1, 2 * p + part, esize));
+                Elem(result, 2 * p + 1, esize, Elem(operand2, 2 * p + part, esize));
+            }
+
+            V(d, result);
+        }
+
         // uaba_advsimd.html
         public static void Uaba_V(bool Q, Bits size, Bits Rm, Bits Rn, Bits Rd)
         {
@@ -4832,6 +4900,146 @@ namespace Ryujinx.Tests.Cpu.Tester
 
             V(d, result);
         }
+
+        // uzp1_advsimd.html
+        public static void Uzp1_V(bool Q, Bits size, Bits Rm, Bits Rn, Bits Rd)
+        {
+            const bool op = false;
+
+            /* Decode */
+            int d = (int)UInt(Rd);
+            int n = (int)UInt(Rn);
+            int m = (int)UInt(Rm);
+
+            /* if size:Q == '110' then ReservedValue(); */
+
+            int esize = 8 << (int)UInt(size);
+            int datasize = (Q ? 128 : 64);
+            int elements = datasize / esize;
+            int part = (int)UInt(op);
+
+            /* Operation */
+            /* CheckFPAdvSIMDEnabled64(); */
+
+            Bits result = new Bits(datasize);
+            Bits operandl = V(datasize, n);
+            Bits operandh = V(datasize, m);
+
+            Bits zipped = Bits.Concat(operandh, operandl);
+
+            for (int e = 0; e <= elements - 1; e++)
+            {
+                Elem(result, e, esize, Elem(zipped, 2 * e + part, esize));
+            }
+
+            V(d, result);
+        }
+
+        // uzp2_advsimd.html
+        public static void Uzp2_V(bool Q, Bits size, Bits Rm, Bits Rn, Bits Rd)
+        {
+            const bool op = true;
+
+            /* Decode */
+            int d = (int)UInt(Rd);
+            int n = (int)UInt(Rn);
+            int m = (int)UInt(Rm);
+
+            /* if size:Q == '110' then ReservedValue(); */
+
+            int esize = 8 << (int)UInt(size);
+            int datasize = (Q ? 128 : 64);
+            int elements = datasize / esize;
+            int part = (int)UInt(op);
+
+            /* Operation */
+            /* CheckFPAdvSIMDEnabled64(); */
+
+            Bits result = new Bits(datasize);
+            Bits operandl = V(datasize, n);
+            Bits operandh = V(datasize, m);
+
+            Bits zipped = Bits.Concat(operandh, operandl);
+
+            for (int e = 0; e <= elements - 1; e++)
+            {
+                Elem(result, e, esize, Elem(zipped, 2 * e + part, esize));
+            }
+
+            V(d, result);
+        }
+
+        // zip1_advsimd.html
+        public static void Zip1_V(bool Q, Bits size, Bits Rm, Bits Rn, Bits Rd)
+        {
+            const bool op = false;
+
+            /* Decode */
+            int d = (int)UInt(Rd);
+            int n = (int)UInt(Rn);
+            int m = (int)UInt(Rm);
+
+            /* if size:Q == '110' then ReservedValue(); */
+
+            int esize = 8 << (int)UInt(size);
+            int datasize = (Q ? 128 : 64);
+            int elements = datasize / esize;
+            int part = (int)UInt(op);
+            int pairs = elements / 2;
+
+            /* Operation */
+            /* CheckFPAdvSIMDEnabled64(); */
+
+            Bits result = new Bits(datasize);
+            Bits operand1 = V(datasize, n);
+            Bits operand2 = V(datasize, m);
+
+            int @base = part * pairs;
+
+            for (int p = 0; p <= pairs - 1; p++)
+            {
+                Elem(result, 2 * p + 0, esize, Elem(operand1, @base + p, esize));
+                Elem(result, 2 * p + 1, esize, Elem(operand2, @base + p, esize));
+            }
+
+            V(d, result);
+        }
+
+        // zip2_advsimd.html
+        public static void Zip2_V(bool Q, Bits size, Bits Rm, Bits Rn, Bits Rd)
+        {
+            const bool op = true;
+
+            /* Decode */
+            int d = (int)UInt(Rd);
+            int n = (int)UInt(Rn);
+            int m = (int)UInt(Rm);
+
+            /* if size:Q == '110' then ReservedValue(); */
+
+            int esize = 8 << (int)UInt(size);
+            int datasize = (Q ? 128 : 64);
+            int elements = datasize / esize;
+            int part = (int)UInt(op);
+            int pairs = elements / 2;
+
+            /* Operation */
+            /* CheckFPAdvSIMDEnabled64(); */
+
+            Bits result = new Bits(datasize);
+            Bits operand1 = V(datasize, n);
+            Bits operand2 = V(datasize, m);
+
+            int @base = part * pairs;
+
+            for (int p = 0; p <= pairs - 1; p++)
+            {
+                Elem(result, 2 * p + 0, esize, Elem(operand1, @base + p, esize));
+                Elem(result, 2 * p + 1, esize, Elem(operand2, @base + p, esize));
+            }
+
+            V(d, result);
+        }
 #endregion
     }
 }

Ryujinx/Config.cs

@@ -29,6 +29,8 @@ namespace Ryujinx
 
             AOptimizations.DisableMemoryChecks = !Convert.ToBoolean(Parser.Value("Enable_Memory_Checks"));
 
+            GraphicsConfig.ShadersDumpPath = Parser.Value("Graphics_Shaders_Dump_Path");
+
             Log.SetEnable(LogLevel.Debug,   Convert.ToBoolean(Parser.Value("Logging_Enable_Debug")));
             Log.SetEnable(LogLevel.Stub,    Convert.ToBoolean(Parser.Value("Logging_Enable_Stub")));
             Log.SetEnable(LogLevel.Info,    Convert.ToBoolean(Parser.Value("Logging_Enable_Info")));

Ryujinx/Ryujinx.conf

@@ -1,6 +1,9 @@
 #Enable cpu memory checks (slow)
 Enable_Memory_Checks = false
 
+#Dump shaders in local directory (e.g. `C:\ShaderDumps`)
+Graphics_Shaders_Dump_Path =
+
 #Enable print debug logs
 Logging_Enable_Debug = false
 

Ryujinx/Ui/GLScreen.cs

@@ -5,6 +5,9 @@ using Ryujinx.Graphics.Gal;
 using Ryujinx.HLE;
 using Ryujinx.HLE.Input;
 using System;
+using System.Threading;
+
+using Stopwatch = System.Diagnostics.Stopwatch;
 
 namespace Ryujinx
 {
@@ -16,6 +19,8 @@ namespace Ryujinx
         private const float TouchScreenRatioX = (float)TouchScreenWidth  / TouchScreenHeight;
         private const float TouchScreenRatioY = (float)TouchScreenHeight / TouchScreenWidth;
 
+        private const int TargetFPS = 60;
+
         private Switch Ns;
 
         private IGalRenderer Renderer;
@@ -24,6 +29,14 @@ namespace Ryujinx
 
         private MouseState? Mouse = null;
 
+        private Thread RenderThread;
+
+        private bool ResizeEvent;
+
+        private bool TitleEvent;
+
+        private string NewTitle;
+
         public GLScreen(Switch Ns, IGalRenderer Renderer)
             : base(1280, 720,
             new GraphicsMode(), "Ryujinx", 0,
@@ -36,13 +49,85 @@ namespace Ryujinx
             Location = new Point(
                 (DisplayDevice.Default.Width  / 2) - (Width  / 2),
                 (DisplayDevice.Default.Height / 2) - (Height / 2));
+
+            ResizeEvent = false;
+
+            TitleEvent = false;
         }
 
-        protected override void OnLoad(EventArgs e)
+        private void RenderLoop()
         {
-            VSync = VSyncMode.On;
+            MakeCurrent();
+
+            Stopwatch Chrono = new Stopwatch();
+
+            Chrono.Start();
+
+            long TicksPerFrame = Stopwatch.Frequency / TargetFPS;
+
+            long Ticks = 0;
+
+            while (Exists && !IsExiting)
+            {
+                if (Ns.WaitFifo())
+                {
+                    Ns.ProcessFrame();
+                }
+
+                Renderer.RunActions();
+
+                if (ResizeEvent)
+                {
+                    ResizeEvent = false;
+
+                    Renderer.FrameBuffer.SetWindowSize(Width, Height);
+                }
+
+                Ticks += Chrono.ElapsedTicks;
+
+                Chrono.Restart();
+
+                if (Ticks >= TicksPerFrame)
+                {
+                    RenderFrame();
+
+                    //Queue max. 1 vsync
+                    Ticks = Math.Min(Ticks - TicksPerFrame, TicksPerFrame);
+                }
+            }
+        }
+
+        public void MainLoop()
+        {
+            VSync = VSyncMode.Off;
+
+            Visible = true;
 
             Renderer.FrameBuffer.SetWindowSize(Width, Height);
+
+            Context.MakeCurrent(null);
+
+            //OpenTK doesn't like sleeps in its thread, to avoid this a renderer thread is created
+            RenderThread = new Thread(RenderLoop);
+
+            RenderThread.Start();
+
+            while (Exists && !IsExiting)
+            {
+                ProcessEvents();
+
+                if (!IsExiting)
+                {
+                    UpdateFrame();
+
+                    if (TitleEvent)
+                    {
+                        TitleEvent = false;
+
+                        Title = NewTitle;
+                    }
+                }
+            }
         }
         
         private bool IsGamePadButtonPressedFromString(GamePadState GamePad, string Button)
@@ -99,7 +184,7 @@ namespace Ryujinx
             }
         }
 
-        protected override void OnUpdateFrame(FrameEventArgs e)
+        private new void UpdateFrame()
         {
             HidControllerButtons CurrentButton = 0;
             HidJoystickPosition  LeftJoystick;
@@ -278,13 +363,9 @@ namespace Ryujinx
                 CurrentButton,
                 LeftJoystick,
                 RightJoystick);
-
-            Ns.ProcessFrame();
-
-            Renderer.RunActions();
         }
 
-        protected override void OnRenderFrame(FrameEventArgs e)
+        private new void RenderFrame()
         {
             Renderer.FrameBuffer.Render();
 
@@ -293,16 +374,25 @@ namespace Ryujinx
             double HostFps = Ns.Statistics.GetSystemFrameRate();
             double GameFps = Ns.Statistics.GetGameFrameRate();
 
-            Title = $"Ryujinx | Host FPS: {HostFps:0.0} | Game FPS: {GameFps:0.0}";
+            NewTitle = $"Ryujinx | Host FPS: {HostFps:0.0} | Game FPS: {GameFps:0.0}";
+
+            TitleEvent = true;
 
             SwapBuffers();
 
             Ns.Os.SignalVsync();
         }
 
+        protected override void OnUnload(EventArgs e)
+        {
+            RenderThread.Join();
+
+            base.OnUnload(e);
+        }
+
         protected override void OnResize(EventArgs e)
         {
-            Renderer.FrameBuffer.SetWindowSize(Width, Height);
+            ResizeEvent = true;
         }
 
         protected override void OnKeyDown(KeyboardKeyEventArgs e)

Ryujinx/Ui/Program.cs

@@ -67,7 +67,7 @@ namespace Ryujinx
                     Screen.Exit();
                 };
 
-                Screen.Run(0.0, 60.0);
+                Screen.MainLoop();
             }
 
             Environment.Exit(0);