Fix R4G4B4A4 format on Vulkan (#3696)

* Scale SamplesPassed counter by RT scale on report

Adds a scale factor for samples passed counter report based on the render target scale at the time. This ensures that when a game reads this counter, it appears similar to the result at 1x.

This doesn't cover cases where the the render target scale changes during the queried draws, though that might be better to handle along with other scope related issues in a future rework of counters. Games generally don't count for occlusion queries over render target changes anyways.

Fixes an issue in the Splatoon games where the special charge would scale too quickly at high res, points at the end of the game would be broken (but still provide a correct winner), and playing at a low res would make it impossible to swim in ink.

May also affect LOD scaling in The Witcher 3.

* Update Ryujinx.Graphics.Gpu/Engine/Threed/SemaphoreUpdater.cs

Co-authored-by: gdkchan <gab.dark.100@gmail.com>

Co-authored-by: gdkchan <gab.dark.100@gmail.com>

ARMeilleure/Decoders/OpCode32.cs

@@ -13,16 +13,13 @@ namespace ARMeilleure.Decoders
             Cond = (Condition)((uint)opCode >> 28);
         }
 
-        public bool IsThumb()
-        {
-            return this is OpCodeT16 || this is OpCodeT32;
-        }
+        public bool IsThumb { get; protected init; } = false;
 
         public uint GetPc()
         {
             // Due to backwards compatibility and legacy behavior of ARMv4 CPUs pipeline,
             // the PC actually points 2 instructions ahead.
-            if (IsThumb())
+            if (IsThumb)
             {
                 // PC is ahead by 4 in thumb mode whether or not the current instruction
                 // is 16 or 32 bit.

ARMeilleure/Decoders/OpCode32SimdCvtFI.cs

@@ -2,9 +2,10 @@
 {
     class OpCode32SimdCvtFI : OpCode32SimdS
     {
-        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdCvtFI(inst, address, opCode);
+        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdCvtFI(inst, address, opCode, false);
+        public new static OpCode CreateT32(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdCvtFI(inst, address, opCode, true);
 
-        public OpCode32SimdCvtFI(InstDescriptor inst, ulong address, int opCode) : base(inst, address, opCode)
+        public OpCode32SimdCvtFI(InstDescriptor inst, ulong address, int opCode, bool isThumb) : base(inst, address, opCode, isThumb)
         {
             Opc = (opCode >> 7) & 0x1;
 

ARMeilleure/Decoders/OpCode32SimdDupGP.cs

@@ -7,10 +7,13 @@
         public int Rt { get; }
         public bool Q { get; }
 
-        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdDupGP(inst, address, opCode);
+        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdDupGP(inst, address, opCode, false);
+        public static OpCode CreateT32(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdDupGP(inst, address, opCode, true);
 
-        public OpCode32SimdDupGP(InstDescriptor inst, ulong address, int opCode) : base(inst, address, opCode)
+        public OpCode32SimdDupGP(InstDescriptor inst, ulong address, int opCode, bool isThumb) : base(inst, address, opCode)
         {
+            IsThumb = isThumb;
+
             Size = 2 - (((opCode >> 21) & 0x2) | ((opCode >> 5) & 0x1)); // B:E - 0 for 32, 16 then 8.
             if (Size == -1)
             {

ARMeilleure/Decoders/OpCode32SimdImm44.cs

@@ -7,10 +7,13 @@
         public int Size { get; }
         public int Elems { get; }
 
-        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdImm44(inst, address, opCode);
+        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdImm44(inst, address, opCode, false);
+        public static OpCode CreateT32(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdImm44(inst, address, opCode, true);
 
-        public OpCode32SimdImm44(InstDescriptor inst, ulong address, int opCode) : base(inst, address, opCode)
+        public OpCode32SimdImm44(InstDescriptor inst, ulong address, int opCode, bool isThumb) : base(inst, address, opCode)
         {
+            IsThumb = isThumb;
+
             Size = (opCode >> 8) & 0x3;
 
             bool single = Size != 3;

ARMeilleure/Decoders/OpCode32SimdMemImm.cs

@@ -8,10 +8,13 @@
         public bool Add { get; }
         public int Immediate { get; }
 
-        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMemImm(inst, address, opCode);
+        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMemImm(inst, address, opCode, false);
+        public static OpCode CreateT32(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMemImm(inst, address, opCode, true);
 
-        public OpCode32SimdMemImm(InstDescriptor inst, ulong address, int opCode) : base(inst, address, opCode)
+        public OpCode32SimdMemImm(InstDescriptor inst, ulong address, int opCode, bool isThumb) : base(inst, address, opCode)
         {
+            IsThumb = isThumb;
+
             Immediate = opCode & 0xff;
 
             Rn = (opCode >> 16) & 0xf;

ARMeilleure/Decoders/OpCode32SimdMemMult.cs

@@ -12,10 +12,13 @@
         public bool DoubleWidth { get; }
         public bool Add { get; }
 
-        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMemMult(inst, address, opCode);
+        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMemMult(inst, address, opCode, false);
+        public static OpCode CreateT32(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMemMult(inst, address, opCode, true);
 
-        public OpCode32SimdMemMult(InstDescriptor inst, ulong address, int opCode) : base(inst, address, opCode)
+        public OpCode32SimdMemMult(InstDescriptor inst, ulong address, int opCode, bool isThumb) : base(inst, address, opCode)
         {
+            IsThumb = isThumb;
+
             Rn = (opCode >> 16) & 0xf;
 
             bool isLoad = (opCode & (1 << 20)) != 0;

ARMeilleure/Decoders/OpCode32SimdMovGp.cs

@@ -11,10 +11,13 @@
         public int Opc1 { get; }
         public int Opc2 { get; }
 
-        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMovGp(inst, address, opCode);
+        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMovGp(inst, address, opCode, false);
+        public static OpCode CreateT32(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMovGp(inst, address, opCode, true);
 
-        public OpCode32SimdMovGp(InstDescriptor inst, ulong address, int opCode) : base(inst, address, opCode)
+        public OpCode32SimdMovGp(InstDescriptor inst, ulong address, int opCode, bool isThumb) : base(inst, address, opCode)
         {
+            IsThumb = isThumb;
+
             // Which one is used is instruction dependant.
             Op = (opCode >> 20) & 0x1;
 

ARMeilleure/Decoders/OpCode32SimdMovGpDouble.cs

@@ -9,10 +9,13 @@
         public int Rt2 { get; }
         public int Op { get; }
 
-        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMovGpDouble(inst, address, opCode);
+        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMovGpDouble(inst, address, opCode, false);
+        public static OpCode CreateT32(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMovGpDouble(inst, address, opCode, true);
 
-        public OpCode32SimdMovGpDouble(InstDescriptor inst, ulong address, int opCode) : base(inst, address, opCode)
+        public OpCode32SimdMovGpDouble(InstDescriptor inst, ulong address, int opCode, bool isThumb) : base(inst, address, opCode)
         {
+            IsThumb = isThumb;
+
             // Which one is used is instruction dependant.
             Op = (opCode >> 20) & 0x1;
 

ARMeilleure/Decoders/OpCode32SimdMovGpElem.cs

@@ -11,10 +11,13 @@
 
         public int Index { get; }
 
-        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMovGpElem(inst, address, opCode);
+        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMovGpElem(inst, address, opCode, false);
+        public static OpCode CreateT32(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdMovGpElem(inst, address, opCode, true);
 
-        public OpCode32SimdMovGpElem(InstDescriptor inst, ulong address, int opCode) : base(inst, address, opCode)
+        public OpCode32SimdMovGpElem(InstDescriptor inst, ulong address, int opCode, bool isThumb) : base(inst, address, opCode)
         {
+            IsThumb = isThumb;
+
             Op = (opCode >> 20) & 0x1;
             U = ((opCode >> 23) & 1) != 0;
 

ARMeilleure/Decoders/OpCode32SimdRegS.cs

@@ -4,9 +4,10 @@
     {
         public int Vn { get; }
 
-        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdRegS(inst, address, opCode);
+        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdRegS(inst, address, opCode, false);
+        public new static OpCode CreateT32(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdRegS(inst, address, opCode, true);
 
-        public OpCode32SimdRegS(InstDescriptor inst, ulong address, int opCode) : base(inst, address, opCode)
+        public OpCode32SimdRegS(InstDescriptor inst, ulong address, int opCode, bool isThumb) : base(inst, address, opCode, isThumb)
         {
             bool single = Size != 3;
             if (single)

ARMeilleure/Decoders/OpCode32SimdS.cs

@@ -8,10 +8,13 @@
         public int Opc2 { get; } // opc2 or RM (opc2<1:0>) [Vcvt, Vrint].
         public int Size { get; protected set; }
 
-        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdS(inst, address, opCode);
+        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdS(inst, address, opCode, false);
+        public static OpCode CreateT32(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdS(inst, address, opCode, true);
 
-        public OpCode32SimdS(InstDescriptor inst, ulong address, int opCode) : base(inst, address, opCode)
+        public OpCode32SimdS(InstDescriptor inst, ulong address, int opCode, bool isThumb) : base(inst, address, opCode)
         {
+            IsThumb = isThumb;
+
             Opc = (opCode >> 15) & 0x3;
             Opc2 = (opCode >> 16) & 0x7;
 

ARMeilleure/Decoders/OpCode32SimdSel.cs

@@ -4,9 +4,10 @@
     {
         public OpCode32SimdSelMode Cc { get; }
 
-        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdSel(inst, address, opCode);
+        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdSel(inst, address, opCode, false);
+        public new static OpCode CreateT32(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdSel(inst, address, opCode, true);
 
-        public OpCode32SimdSel(InstDescriptor inst, ulong address, int opCode) : base(inst, address, opCode)
+        public OpCode32SimdSel(InstDescriptor inst, ulong address, int opCode, bool isThumb) : base(inst, address, opCode, isThumb)
         {
             Cc = (OpCode32SimdSelMode)((opCode >> 20) & 3);
         }

ARMeilleure/Decoders/OpCode32SimdSpecial.cs

@@ -5,10 +5,13 @@
         public int Rt { get; }
         public int Sreg { get; }
 
-        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdSpecial(inst, address, opCode);
+        public new static OpCode Create(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdSpecial(inst, address, opCode, false);
+        public static OpCode CreateT32(InstDescriptor inst, ulong address, int opCode) => new OpCode32SimdSpecial(inst, address, opCode, true);
 
-        public OpCode32SimdSpecial(InstDescriptor inst, ulong address, int opCode) : base(inst, address, opCode)
+        public OpCode32SimdSpecial(InstDescriptor inst, ulong address, int opCode, bool isThumb) : base(inst, address, opCode)
         {
+            IsThumb = isThumb;
+
             Rt = (opCode >> 12) & 0xf;
             Sreg = (opCode >> 16) & 0xf;
         }

ARMeilleure/Decoders/OpCodeT16.cs

@@ -8,6 +8,7 @@ namespace ARMeilleure.Decoders
         {
             Cond = Condition.Al;
 
+            IsThumb = true;
             OpCodeSizeInBytes = 2;
         }
     }

ARMeilleure/Decoders/OpCodeT32.cs

@@ -8,6 +8,7 @@
         {
             Cond = Condition.Al;
 
+            IsThumb = true;
             OpCodeSizeInBytes = 4;
         }
     }

ARMeilleure/Decoders/OpCodeTable.cs

@@ -792,7 +792,61 @@ namespace ARMeilleure.Decoders
             SetA32("<<<<01101100xxxxxxxxxx000111xxxx", InstName.Uxtb16,  InstEmit32.Uxtb16,  OpCode32AluUx.Create);
             SetA32("<<<<01101111xxxxxxxxxx000111xxxx", InstName.Uxth,    InstEmit32.Uxth,    OpCode32AluUx.Create);
 
-            // FP & SIMD
+            // VFP
+            SetVfp("<<<<11101x110000xxxx101x11x0xxxx", InstName.Vabs,   InstEmit32.Vabs_S,   OpCode32SimdS.Create,           OpCode32SimdS.CreateT32);
+            SetVfp("<<<<11100x11xxxxxxxx101xx0x0xxxx", InstName.Vadd,   InstEmit32.Vadd_S,   OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("<<<<11101x11010xxxxx101x01x0xxxx", InstName.Vcmp,   InstEmit32.Vcmp,     OpCode32SimdS.Create,           OpCode32SimdS.CreateT32);
+            SetVfp("<<<<11101x11010xxxxx101x11x0xxxx", InstName.Vcmpe,  InstEmit32.Vcmpe,    OpCode32SimdS.Create,           OpCode32SimdS.CreateT32);
+            SetVfp("<<<<11101x110111xxxx101x11x0xxxx", InstName.Vcvt,   InstEmit32.Vcvt_FD,  OpCode32SimdS.Create,           OpCode32SimdS.CreateT32); // FP 32 and 64, scalar.
+            SetVfp("<<<<11101x11110xxxxx101x11x0xxxx", InstName.Vcvt,   InstEmit32.Vcvt_FI,  OpCode32SimdCvtFI.Create,       OpCode32SimdCvtFI.CreateT32); // FP32 to int.
+            SetVfp("<<<<11101x111000xxxx101xx1x0xxxx", InstName.Vcvt,   InstEmit32.Vcvt_FI,  OpCode32SimdCvtFI.Create,       OpCode32SimdCvtFI.CreateT32); // Int to FP32.
+            SetVfp("111111101x1111xxxxxx101xx1x0xxxx", InstName.Vcvt,   InstEmit32.Vcvt_RM,  OpCode32SimdCvtFI.Create,       OpCode32SimdCvtFI.CreateT32); // The many FP32 to int encodings (fp).
+            SetVfp("<<<<11101x00xxxxxxxx101xx0x0xxxx", InstName.Vdiv,   InstEmit32.Vdiv_S,   OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("<<<<11101xx0xxxxxxxx1011x0x10000", InstName.Vdup,   InstEmit32.Vdup,     OpCode32SimdDupGP.Create,       OpCode32SimdDupGP.CreateT32);
+            SetVfp("<<<<11101x10xxxxxxxx101xx0x0xxxx", InstName.Vfma,   InstEmit32.Vfma_S,   OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("<<<<11101x10xxxxxxxx101xx1x0xxxx", InstName.Vfms,   InstEmit32.Vfms_S,   OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("<<<<11101x01xxxxxxxx101xx1x0xxxx", InstName.Vfnma,  InstEmit32.Vfnma_S,  OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("<<<<11101x01xxxxxxxx101xx0x0xxxx", InstName.Vfnms,  InstEmit32.Vfnms_S,  OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("<<<<11001x01xxxxxxxx1011xxxxxxx0", InstName.Vldm,   InstEmit32.Vldm,     OpCode32SimdMemMult.Create,     OpCode32SimdMemMult.CreateT32);
+            SetVfp("<<<<11001x11xxxxxxxx1011xxxxxxx0", InstName.Vldm,   InstEmit32.Vldm,     OpCode32SimdMemMult.Create,     OpCode32SimdMemMult.CreateT32);
+            SetVfp("<<<<11010x11xxxxxxxx1011xxxxxxx0", InstName.Vldm,   InstEmit32.Vldm,     OpCode32SimdMemMult.Create,     OpCode32SimdMemMult.CreateT32);
+            SetVfp("<<<<11001x01xxxxxxxx1010xxxxxxxx", InstName.Vldm,   InstEmit32.Vldm,     OpCode32SimdMemMult.Create,     OpCode32SimdMemMult.CreateT32);
+            SetVfp("<<<<11001x11xxxxxxxx1010xxxxxxxx", InstName.Vldm,   InstEmit32.Vldm,     OpCode32SimdMemMult.Create,     OpCode32SimdMemMult.CreateT32);
+            SetVfp("<<<<11010x11xxxxxxxx1010xxxxxxxx", InstName.Vldm,   InstEmit32.Vldm,     OpCode32SimdMemMult.Create,     OpCode32SimdMemMult.CreateT32);
+            SetVfp("<<<<1101xx01xxxxxxxx101xxxxxxxxx", InstName.Vldr,   InstEmit32.Vldr,     OpCode32SimdMemImm.Create,      OpCode32SimdMemImm.CreateT32);
+            SetVfp("111111101x00xxxxxxxx10>>x0x0xxxx", InstName.Vmaxnm, InstEmit32.Vmaxnm_S, OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("111111101x00xxxxxxxx10>>x1x0xxxx", InstName.Vminnm, InstEmit32.Vminnm_S, OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("<<<<11100x00xxxxxxxx101xx0x0xxxx", InstName.Vmla,   InstEmit32.Vmla_S,   OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("<<<<11100x00xxxxxxxx101xx1x0xxxx", InstName.Vmls,   InstEmit32.Vmls_S,   OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("<<<<11100xx0xxxxxxxx1011xxx10000", InstName.Vmov,   InstEmit32.Vmov_G1,  OpCode32SimdMovGpElem.Create,   OpCode32SimdMovGpElem.CreateT32); // From gen purpose.
+            SetVfp("<<<<1110xxx1xxxxxxxx1011xxx10000", InstName.Vmov,   InstEmit32.Vmov_G1,  OpCode32SimdMovGpElem.Create,   OpCode32SimdMovGpElem.CreateT32); // To gen purpose.
+            SetVfp("<<<<1100010xxxxxxxxx101000x1xxxx", InstName.Vmov,   InstEmit32.Vmov_G2,  OpCode32SimdMovGpDouble.Create, OpCode32SimdMovGpDouble.CreateT32); // To/from gen purpose x2 and single precision x2.
+            SetVfp("<<<<1100010xxxxxxxxx101100x1xxxx", InstName.Vmov,   InstEmit32.Vmov_GD,  OpCode32SimdMovGpDouble.Create, OpCode32SimdMovGpDouble.CreateT32); // To/from gen purpose x2 and double precision.
+            SetVfp("<<<<1110000xxxxxxxxx1010x0010000", InstName.Vmov,   InstEmit32.Vmov_GS,  OpCode32SimdMovGp.Create,       OpCode32SimdMovGp.CreateT32); // To/from gen purpose and single precision.
+            SetVfp("<<<<11101x11xxxxxxxx101x0000xxxx", InstName.Vmov,   InstEmit32.Vmov_I,   OpCode32SimdImm44.Create,       OpCode32SimdImm44.CreateT32); // Scalar f16/32/64 based on size 01 10 11.
+            SetVfp("<<<<11101x110000xxxx101x01x0xxxx", InstName.Vmov,   InstEmit32.Vmov_S,   OpCode32SimdS.Create,           OpCode32SimdS.CreateT32);
+            SetVfp("<<<<11101111xxxxxxxx101000010000", InstName.Vmrs,   InstEmit32.Vmrs,     OpCode32SimdSpecial.Create,     OpCode32SimdSpecial.CreateT32);
+            SetVfp("<<<<11101110xxxxxxxx101000010000", InstName.Vmsr,   InstEmit32.Vmsr,     OpCode32SimdSpecial.Create,     OpCode32SimdSpecial.CreateT32);
+            SetVfp("<<<<11100x10xxxxxxxx101xx0x0xxxx", InstName.Vmul,   InstEmit32.Vmul_S,   OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("<<<<11101x110001xxxx101x01x0xxxx", InstName.Vneg,   InstEmit32.Vneg_S,   OpCode32SimdS.Create,           OpCode32SimdS.CreateT32);
+            SetVfp("<<<<11100x01xxxxxxxx101xx1x0xxxx", InstName.Vnmla,  InstEmit32.Vnmla_S,  OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("<<<<11100x01xxxxxxxx101xx0x0xxxx", InstName.Vnmls,  InstEmit32.Vnmls_S,  OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("<<<<11100x10xxxxxxxx101xx1x0xxxx", InstName.Vnmul,  InstEmit32.Vnmul_S,  OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+            SetVfp("111111101x1110xxxxxx101x01x0xxxx", InstName.Vrint,  InstEmit32.Vrint_RM, OpCode32SimdS.Create,           OpCode32SimdS.CreateT32);
+            SetVfp("<<<<11101x110110xxxx101x11x0xxxx", InstName.Vrint,  InstEmit32.Vrint_Z,  OpCode32SimdS.Create,           OpCode32SimdS.CreateT32);
+            SetVfp("<<<<11101x110111xxxx101x01x0xxxx", InstName.Vrintx, InstEmit32.Vrintx_S, OpCode32SimdS.Create,           OpCode32SimdS.CreateT32);
+            SetVfp("<<<<11101x110001xxxx101x11x0xxxx", InstName.Vsqrt,  InstEmit32.Vsqrt_S,  OpCode32SimdS.Create,           OpCode32SimdS.CreateT32);
+            SetVfp("111111100xxxxxxxxxxx101xx0x0xxxx", InstName.Vsel,   InstEmit32.Vsel,     OpCode32SimdSel.Create,         OpCode32SimdSel.CreateT32);
+            SetVfp("<<<<11001x00xxxxxxxx1011xxxxxxx0", InstName.Vstm,   InstEmit32.Vstm,     OpCode32SimdMemMult.Create,     OpCode32SimdMemMult.CreateT32);
+            SetVfp("<<<<11001x10xxxxxxxx1011xxxxxxx0", InstName.Vstm,   InstEmit32.Vstm,     OpCode32SimdMemMult.Create,     OpCode32SimdMemMult.CreateT32);
+            SetVfp("<<<<11010x10xxxxxxxx1011xxxxxxx0", InstName.Vstm,   InstEmit32.Vstm,     OpCode32SimdMemMult.Create,     OpCode32SimdMemMult.CreateT32);
+            SetVfp("<<<<11001x00xxxxxxxx1010xxxxxxxx", InstName.Vstm,   InstEmit32.Vstm,     OpCode32SimdMemMult.Create,     OpCode32SimdMemMult.CreateT32);
+            SetVfp("<<<<11001x10xxxxxxxx1010xxxxxxxx", InstName.Vstm,   InstEmit32.Vstm,     OpCode32SimdMemMult.Create,     OpCode32SimdMemMult.CreateT32);
+            SetVfp("<<<<11010x10xxxxxxxx1010xxxxxxxx", InstName.Vstm,   InstEmit32.Vstm,     OpCode32SimdMemMult.Create,     OpCode32SimdMemMult.CreateT32);
+            SetVfp("<<<<1101xx00xxxxxxxx101xxxxxxxxx", InstName.Vstr,   InstEmit32.Vstr,     OpCode32SimdMemImm.Create,      OpCode32SimdMemImm.CreateT32);
+            SetVfp("<<<<11100x11xxxxxxxx101xx1x0xxxx", InstName.Vsub,   InstEmit32.Vsub_S,   OpCode32SimdRegS.Create,        OpCode32SimdRegS.CreateT32);
+
+            // ASIMD
             SetA32("111100111x110000xxx0001101x0xxx0", InstName.Aesd_V,      InstEmit32.Aesd_V,      OpCode32Simd.Create);
             SetA32("111100111x110000xxx0001100x0xxx0", InstName.Aese_V,      InstEmit32.Aese_V,      OpCode32Simd.Create);
             SetA32("111100111x110000xxx0001111x0xxx0", InstName.Aesimc_V,    InstEmit32.Aesimc_V,    OpCode32Simd.Create);
@@ -803,11 +857,9 @@ namespace ARMeilleure.Decoders
             SetA32("111100110x10xxx0xxx01100x1x0xxx0", InstName.Sha256su1_V, InstEmit32.Sha256su1_V, OpCode32SimdReg.Create);
             SetA32("1111001x0x<<xxxxxxxx0111xxx0xxxx", InstName.Vabd,        InstEmit32.Vabd_I,      OpCode32SimdReg.Create);
             SetA32("1111001x1x<<xxxxxxxx0111x0x0xxxx", InstName.Vabdl,       InstEmit32.Vabdl_I,     OpCode32SimdRegLong.Create);
-            SetA32("<<<<11101x110000xxxx101x11x0xxxx", InstName.Vabs,        InstEmit32.Vabs_S,      OpCode32SimdS.Create);
             SetA32("111100111x11<<01xxxx00110xx0xxxx", InstName.Vabs,        InstEmit32.Vabs_V,      OpCode32SimdCmpZ.Create);
             SetA32("111100111x111001xxxx01110xx0xxxx", InstName.Vabs,        InstEmit32.Vabs_V,      OpCode32SimdCmpZ.Create);
             SetA32("111100100xxxxxxxxxxx1000xxx0xxxx", InstName.Vadd,        InstEmit32.Vadd_I,      OpCode32SimdReg.Create);
-            SetA32("<<<<11100x11xxxxxxxx101xx0x0xxxx", InstName.Vadd,        InstEmit32.Vadd_S,      OpCode32SimdRegS.Create);
             SetA32("111100100x00xxxxxxxx1101xxx0xxxx", InstName.Vadd,        InstEmit32.Vadd_V,      OpCode32SimdReg.Create);
             SetA32("1111001x1x<<xxxxxxx00000x0x0xxxx", InstName.Vaddl,       InstEmit32.Vaddl_I,     OpCode32SimdRegLong.Create);
             SetA32("1111001x1x<<xxxxxxx00001x0x0xxxx", InstName.Vaddw,       InstEmit32.Vaddw_I,     OpCode32SimdRegWide.Create);
@@ -828,25 +880,13 @@ namespace ARMeilleure.Decoders
             SetA32("111100111x11xx01xxxx0x000xx0xxxx", InstName.Vcgt,        InstEmit32.Vcgt_Z,      OpCode32SimdCmpZ.Create);
             SetA32("111100111x11xx01xxxx0x011xx0xxxx", InstName.Vcle,        InstEmit32.Vcle_Z,      OpCode32SimdCmpZ.Create);
             SetA32("111100111x11xx01xxxx0x100xx0xxxx", InstName.Vclt,        InstEmit32.Vclt_Z,      OpCode32SimdCmpZ.Create);
-            SetA32("<<<<11101x11010xxxxx101x01x0xxxx", InstName.Vcmp,        InstEmit32.Vcmp,        OpCode32SimdS.Create);
-            SetA32("<<<<11101x11010xxxxx101x11x0xxxx", InstName.Vcmpe,       InstEmit32.Vcmpe,       OpCode32SimdS.Create);
             SetA32("111100111x110000xxxx01010xx0xxxx", InstName.Vcnt,        InstEmit32.Vcnt,        OpCode32SimdCmpZ.Create);
-            SetA32("<<<<11101x110111xxxx101x11x0xxxx", InstName.Vcvt,        InstEmit32.Vcvt_FD,     OpCode32SimdS.Create); // FP 32 and 64, scalar.
-            SetA32("<<<<11101x11110xxxxx101x11x0xxxx", InstName.Vcvt,        InstEmit32.Vcvt_FI,     OpCode32SimdCvtFI.Create); // FP32 to int.
-            SetA32("<<<<11101x111000xxxx101xx1x0xxxx", InstName.Vcvt,        InstEmit32.Vcvt_FI,     OpCode32SimdCvtFI.Create); // Int to FP32.
-            SetA32("111111101x1111xxxxxx101xx1x0xxxx", InstName.Vcvt,        InstEmit32.Vcvt_RM,     OpCode32SimdCvtFI.Create); // The many FP32 to int encodings (fp).
             SetA32("111100111x111011xxxx011xxxx0xxxx", InstName.Vcvt,        InstEmit32.Vcvt_V,      OpCode32SimdCmpZ.Create); // FP and integer, vector.
-            SetA32("<<<<11101x00xxxxxxxx101xx0x0xxxx", InstName.Vdiv,        InstEmit32.Vdiv_S,      OpCode32SimdRegS.Create);
-            SetA32("<<<<11101xx0xxxxxxxx1011x0x10000", InstName.Vdup,        InstEmit32.Vdup,        OpCode32SimdDupGP.Create);
             SetA32("111100111x11xxxxxxxx11000xx0xxxx", InstName.Vdup,        InstEmit32.Vdup_1,      OpCode32SimdDupElem.Create);
             SetA32("111100110x00xxxxxxxx0001xxx1xxxx", InstName.Veor,        InstEmit32.Veor_I,      OpCode32SimdBinary.Create);
             SetA32("111100101x11xxxxxxxxxxxxxxx0xxxx", InstName.Vext,        InstEmit32.Vext,        OpCode32SimdExt.Create);
-            SetA32("<<<<11101x10xxxxxxxx101xx0x0xxxx", InstName.Vfma,        InstEmit32.Vfma_S,      OpCode32SimdRegS.Create);
             SetA32("111100100x00xxxxxxxx1100xxx1xxxx", InstName.Vfma,        InstEmit32.Vfma_V,      OpCode32SimdReg.Create);
-            SetA32("<<<<11101x10xxxxxxxx101xx1x0xxxx", InstName.Vfms,        InstEmit32.Vfms_S,      OpCode32SimdRegS.Create);
             SetA32("111100100x10xxxxxxxx1100xxx1xxxx", InstName.Vfms,        InstEmit32.Vfms_V,      OpCode32SimdReg.Create);
-            SetA32("<<<<11101x01xxxxxxxx101xx1x0xxxx", InstName.Vfnma,       InstEmit32.Vfnma_S,     OpCode32SimdRegS.Create);
-            SetA32("<<<<11101x01xxxxxxxx101xx0x0xxxx", InstName.Vfnms,       InstEmit32.Vfnms_S,     OpCode32SimdRegS.Create);
             SetA32("1111001x0x<<xxxxxxxx0000xxx0xxxx", InstName.Vhadd,       InstEmit32.Vhadd,       OpCode32SimdReg.Create);
             SetA32("111101001x10xxxxxxxxxx00xxxxxxxx", InstName.Vld1,        InstEmit32.Vld1,        OpCode32SimdMemSingle.Create);
             SetA32("111101000x10xxxxxxxx0111xxxxxxxx", InstName.Vld1,        InstEmit32.Vld1,        OpCode32SimdMemPair.Create); // Regs = 1.
@@ -860,52 +900,31 @@ namespace ARMeilleure.Decoders
             SetA32("111101000x10xxxxxxxx010xxxxxxxxx", InstName.Vld3,        InstEmit32.Vld3,        OpCode32SimdMemPair.Create); // Inc = 1/2 (itype).
             SetA32("111101001x10xxxxxxxxxx11xxxxxxxx", InstName.Vld4,        InstEmit32.Vld4,        OpCode32SimdMemSingle.Create);
             SetA32("111101000x10xxxxxxxx000xxxxxxxxx", InstName.Vld4,        InstEmit32.Vld4,        OpCode32SimdMemPair.Create); // Inc = 1/2 (itype).
-            SetA32("<<<<11001x01xxxxxxxx1011xxxxxxx0", InstName.Vldm,        InstEmit32.Vldm,        OpCode32SimdMemMult.Create);
-            SetA32("<<<<11001x11xxxxxxxx1011xxxxxxx0", InstName.Vldm,        InstEmit32.Vldm,        OpCode32SimdMemMult.Create);
-            SetA32("<<<<11010x11xxxxxxxx1011xxxxxxx0", InstName.Vldm,        InstEmit32.Vldm,        OpCode32SimdMemMult.Create);
-            SetA32("<<<<11001x01xxxxxxxx1010xxxxxxxx", InstName.Vldm,        InstEmit32.Vldm,        OpCode32SimdMemMult.Create);
-            SetA32("<<<<11001x11xxxxxxxx1010xxxxxxxx", InstName.Vldm,        InstEmit32.Vldm,        OpCode32SimdMemMult.Create);
-            SetA32("<<<<11010x11xxxxxxxx1010xxxxxxxx", InstName.Vldm,        InstEmit32.Vldm,        OpCode32SimdMemMult.Create);
-            SetA32("<<<<1101xx01xxxxxxxx101xxxxxxxxx", InstName.Vldr,        InstEmit32.Vldr,        OpCode32SimdMemImm.Create);
             SetA32("1111001x0x<<xxxxxxxx0110xxx0xxxx", InstName.Vmax,        InstEmit32.Vmax_I,      OpCode32SimdReg.Create);
             SetA32("111100100x00xxxxxxxx1111xxx0xxxx", InstName.Vmax,        InstEmit32.Vmax_V,      OpCode32SimdReg.Create);
             SetA32("1111001x0x<<xxxxxxxx0110xxx1xxxx", InstName.Vmin,        InstEmit32.Vmin_I,      OpCode32SimdReg.Create);
             SetA32("111100100x10xxxxxxxx1111xxx0xxxx", InstName.Vmin,        InstEmit32.Vmin_V,      OpCode32SimdReg.Create);
-            SetA32("111111101x00xxxxxxxx10>>x0x0xxxx", InstName.Vmaxnm,      InstEmit32.Vmaxnm_S,    OpCode32SimdRegS.Create);
             SetA32("111100110x0xxxxxxxxx1111xxx1xxxx", InstName.Vmaxnm,      InstEmit32.Vmaxnm_V,    OpCode32SimdReg.Create);
-            SetA32("111111101x00xxxxxxxx10>>x1x0xxxx", InstName.Vminnm,      InstEmit32.Vminnm_S,    OpCode32SimdRegS.Create);
             SetA32("111100110x1xxxxxxxxx1111xxx1xxxx", InstName.Vminnm,      InstEmit32.Vminnm_V,    OpCode32SimdReg.Create);
             SetA32("1111001x1x<<xxxxxxxx000xx1x0xxxx", InstName.Vmla,        InstEmit32.Vmla_1,      OpCode32SimdRegElem.Create);
             SetA32("111100100xxxxxxxxxxx1001xxx0xxxx", InstName.Vmla,        InstEmit32.Vmla_I,      OpCode32SimdReg.Create);
-            SetA32("<<<<11100x00xxxxxxxx101xx0x0xxxx", InstName.Vmla,        InstEmit32.Vmla_S,      OpCode32SimdRegS.Create);
             SetA32("111100100x00xxxxxxxx1101xxx1xxxx", InstName.Vmla,        InstEmit32.Vmla_V,      OpCode32SimdReg.Create);
             SetA32("1111001x1x<<xxxxxxx01000x0x0xxxx", InstName.Vmlal,       InstEmit32.Vmlal_I,     OpCode32SimdRegLong.Create);
             SetA32("1111001x1x<<xxxxxxxx010xx1x0xxxx", InstName.Vmls,        InstEmit32.Vmls_1,      OpCode32SimdRegElem.Create);
-            SetA32("<<<<11100x00xxxxxxxx101xx1x0xxxx", InstName.Vmls,        InstEmit32.Vmls_S,      OpCode32SimdRegS.Create);
             SetA32("111100100x10xxxxxxxx1101xxx1xxxx", InstName.Vmls,        InstEmit32.Vmls_V,      OpCode32SimdReg.Create);
             SetA32("111100110xxxxxxxxxxx1001xxx0xxxx", InstName.Vmls,        InstEmit32.Vmls_I,      OpCode32SimdReg.Create);
             SetA32("1111001x1x<<xxxxxxx01010x0x0xxxx", InstName.Vmlsl,       InstEmit32.Vmlsl_I,     OpCode32SimdRegLong.Create);
-            SetA32("<<<<11100xx0xxxxxxxx1011xxx10000", InstName.Vmov,        InstEmit32.Vmov_G1,     OpCode32SimdMovGpElem.Create); // From gen purpose.
-            SetA32("<<<<1110xxx1xxxxxxxx1011xxx10000", InstName.Vmov,        InstEmit32.Vmov_G1,     OpCode32SimdMovGpElem.Create); // To gen purpose.
-            SetA32("<<<<1100010xxxxxxxxx101000x1xxxx", InstName.Vmov,        InstEmit32.Vmov_G2,     OpCode32SimdMovGpDouble.Create); // To/from gen purpose x2 and single precision x2.
-            SetA32("<<<<1100010xxxxxxxxx101100x1xxxx", InstName.Vmov,        InstEmit32.Vmov_GD,     OpCode32SimdMovGpDouble.Create); // To/from gen purpose x2 and double precision.
-            SetA32("<<<<1110000xxxxxxxxx1010x0010000", InstName.Vmov,        InstEmit32.Vmov_GS,     OpCode32SimdMovGp.Create); // To/from gen purpose and single precision.
             SetA32("1111001x1x000xxxxxxx0xx00x01xxxx", InstName.Vmov,        InstEmit32.Vmov_I,      OpCode32SimdImm.Create); // D/Q vector I32.
-            SetA32("<<<<11101x11xxxxxxxx101x0000xxxx", InstName.Vmov,        InstEmit32.Vmov_I,      OpCode32SimdImm44.Create); // Scalar f16/32/64 based on size 01 10 11.
             SetA32("1111001x1x000xxxxxxx10x00x01xxxx", InstName.Vmov,        InstEmit32.Vmov_I,      OpCode32SimdImm.Create); // D/Q I16.
             SetA32("1111001x1x000xxxxxxx11xx0x01xxxx", InstName.Vmov,        InstEmit32.Vmov_I,      OpCode32SimdImm.Create); // D/Q (dt - from cmode).
             SetA32("1111001x1x000xxxxxxx11100x11xxxx", InstName.Vmov,        InstEmit32.Vmov_I,      OpCode32SimdImm.Create); // D/Q I64.
-            SetA32("<<<<11101x110000xxxx101x01x0xxxx", InstName.Vmov,        InstEmit32.Vmov_S,      OpCode32SimdS.Create);
             SetA32("1111001x1x001000xxx0101000x1xxxx", InstName.Vmovl,       InstEmit32.Vmovl,       OpCode32SimdLong.Create);
             SetA32("1111001x1x010000xxx0101000x1xxxx", InstName.Vmovl,       InstEmit32.Vmovl,       OpCode32SimdLong.Create);
             SetA32("1111001x1x100000xxx0101000x1xxxx", InstName.Vmovl,       InstEmit32.Vmovl,       OpCode32SimdLong.Create);
             SetA32("111100111x11<<10xxxx001000x0xxx0", InstName.Vmovn,       InstEmit32.Vmovn,       OpCode32SimdMovn.Create);
-            SetA32("<<<<11101111xxxxxxxx101000010000", InstName.Vmrs,        InstEmit32.Vmrs,        OpCode32SimdSpecial.Create);
-            SetA32("<<<<11101110xxxxxxxx101000010000", InstName.Vmsr,        InstEmit32.Vmsr,        OpCode32SimdSpecial.Create);
             SetA32("1111001x1x<<xxxxxxxx100xx1x0xxxx", InstName.Vmul,        InstEmit32.Vmul_1,      OpCode32SimdRegElem.Create);
             SetA32("111100100x<<xxxxxxxx1001xxx1xxxx", InstName.Vmul,        InstEmit32.Vmul_I,      OpCode32SimdReg.Create);
             SetA32("111100110x00xxxxxxxx1001xxx1xxxx", InstName.Vmul,        InstEmit32.Vmul_I,      OpCode32SimdReg.Create);
-            SetA32("<<<<11100x10xxxxxxxx101xx0x0xxxx", InstName.Vmul,        InstEmit32.Vmul_S,      OpCode32SimdRegS.Create);
             SetA32("111100110x00xxxxxxxx1101xxx1xxxx", InstName.Vmul,        InstEmit32.Vmul_V,      OpCode32SimdReg.Create);
             SetA32("1111001x1x<<xxxxxxx01010x1x0xxxx", InstName.Vmull,       InstEmit32.Vmull_1,     OpCode32SimdRegElemLong.Create);
             SetA32("1111001x1x<<xxxxxxx01100x0x0xxxx", InstName.Vmull,       InstEmit32.Vmull_I,     OpCode32SimdRegLong.Create);
@@ -914,12 +933,8 @@ namespace ARMeilleure.Decoders
             SetA32("1111001x1x000xxxxxxx0xx00x11xxxx", InstName.Vmvn,        InstEmit32.Vmvn_II,     OpCode32SimdImm.Create); // D/Q vector I32.
             SetA32("1111001x1x000xxxxxxx10x00x11xxxx", InstName.Vmvn,        InstEmit32.Vmvn_II,     OpCode32SimdImm.Create);
             SetA32("1111001x1x000xxxxxxx110x0x11xxxx", InstName.Vmvn,        InstEmit32.Vmvn_II,     OpCode32SimdImm.Create);
-            SetA32("<<<<11101x110001xxxx101x01x0xxxx", InstName.Vneg,        InstEmit32.Vneg_S,      OpCode32SimdS.Create);
             SetA32("111100111x11<<01xxxx00111xx0xxxx", InstName.Vneg,        InstEmit32.Vneg_V,      OpCode32SimdCmpZ.Create);
             SetA32("111100111x111001xxxx01111xx0xxxx", InstName.Vneg,        InstEmit32.Vneg_V,      OpCode32SimdCmpZ.Create);
-            SetA32("<<<<11100x01xxxxxxxx101xx1x0xxxx", InstName.Vnmla,       InstEmit32.Vnmla_S,     OpCode32SimdRegS.Create);
-            SetA32("<<<<11100x01xxxxxxxx101xx0x0xxxx", InstName.Vnmls,       InstEmit32.Vnmls_S,     OpCode32SimdRegS.Create);
-            SetA32("<<<<11100x10xxxxxxxx101xx1x0xxxx", InstName.Vnmul,       InstEmit32.Vnmul_S,     OpCode32SimdRegS.Create);
             SetA32("111100100x11xxxxxxxx0001xxx1xxxx", InstName.Vorn,        InstEmit32.Vorn_I,      OpCode32SimdBinary.Create);
             SetA32("111100100x10xxxxxxxx0001xxx1xxxx", InstName.Vorr,        InstEmit32.Vorr_I,      OpCode32SimdBinary.Create);
             SetA32("1111001x1x000xxxxxxx<<x10x01xxxx", InstName.Vorr,        InstEmit32.Vorr_II,     OpCode32SimdImm.Create);
@@ -944,21 +959,20 @@ namespace ARMeilleure.Decoders
             SetA32("111100100x00xxxxxxxx1111xxx1xxxx", InstName.Vrecps,      InstEmit32.Vrecps,      OpCode32SimdReg.Create);
             SetA32("111100111x11xx00xxxx000<<xx0xxxx", InstName.Vrev,        InstEmit32.Vrev,        OpCode32SimdRev.Create);
             SetA32("1111001x0x<<xxxxxxxx0001xxx0xxxx", InstName.Vrhadd,      InstEmit32.Vrhadd,      OpCode32SimdReg.Create);
-            SetA32("111111101x1110xxxxxx101x01x0xxxx", InstName.Vrint,       InstEmit32.Vrint_RM,    OpCode32SimdS.Create);
-            SetA32("<<<<11101x110110xxxx101x11x0xxxx", InstName.Vrint,       InstEmit32.Vrint_Z,     OpCode32SimdS.Create);
-            SetA32("<<<<11101x110111xxxx101x01x0xxxx", InstName.Vrintx,      InstEmit32.Vrintx_S,    OpCode32SimdS.Create);
+            SetA32("111100111x111010xxxx01010xx0xxxx", InstName.Vrinta,      InstEmit32.Vrinta_V,    OpCode32SimdCmpZ.Create);
+            SetA32("111100111x111010xxxx01101xx0xxxx", InstName.Vrintm,      InstEmit32.Vrintm_V,    OpCode32SimdCmpZ.Create);
+            SetA32("111100111x111010xxxx01000xx0xxxx", InstName.Vrintn,      InstEmit32.Vrintn_V,    OpCode32SimdCmpZ.Create);
+            SetA32("111100111x111010xxxx01111xx0xxxx", InstName.Vrintp,      InstEmit32.Vrintp_V,    OpCode32SimdCmpZ.Create);
             SetA32("1111001x1x>>>xxxxxxx0010>xx1xxxx", InstName.Vrshr,       InstEmit32.Vrshr,       OpCode32SimdShImm.Create);
             SetA32("111100101x>>>xxxxxxx100001x1xxx0", InstName.Vrshrn,      InstEmit32.Vrshrn,      OpCode32SimdShImmNarrow.Create);
             SetA32("111100111x111011xxxx010x1xx0xxxx", InstName.Vrsqrte,     InstEmit32.Vrsqrte,     OpCode32SimdSqrte.Create);
             SetA32("111100100x10xxxxxxxx1111xxx1xxxx", InstName.Vrsqrts,     InstEmit32.Vrsqrts,     OpCode32SimdReg.Create);
             SetA32("1111001x1x>>>xxxxxxx0011>xx1xxxx", InstName.Vrsra,       InstEmit32.Vrsra,       OpCode32SimdShImm.Create);
-            SetA32("111111100xxxxxxxxxxx101xx0x0xxxx", InstName.Vsel,        InstEmit32.Vsel,        OpCode32SimdSel.Create);
             SetA32("111100101x>>>xxxxxxx0101>xx1xxxx", InstName.Vshl,        InstEmit32.Vshl,        OpCode32SimdShImm.Create);
             SetA32("1111001x0xxxxxxxxxxx0100xxx0xxxx", InstName.Vshl,        InstEmit32.Vshl_I,      OpCode32SimdReg.Create);
             SetA32("1111001x1x>>>xxxxxxx101000x1xxxx", InstName.Vshll,       InstEmit32.Vshll,       OpCode32SimdShImmLong.Create); // A1 encoding.
             SetA32("1111001x1x>>>xxxxxxx0000>xx1xxxx", InstName.Vshr,        InstEmit32.Vshr,        OpCode32SimdShImm.Create);
             SetA32("111100101x>>>xxxxxxx100000x1xxx0", InstName.Vshrn,       InstEmit32.Vshrn,       OpCode32SimdShImmNarrow.Create);
-            SetA32("<<<<11101x110001xxxx101x11x0xxxx", InstName.Vsqrt,       InstEmit32.Vsqrt_S,     OpCode32SimdS.Create);
             SetA32("1111001x1x>>>xxxxxxx0001>xx1xxxx", InstName.Vsra,        InstEmit32.Vsra,        OpCode32SimdShImm.Create);
             SetA32("111101001x00xxxxxxxx<<00xxxxxxxx", InstName.Vst1,        InstEmit32.Vst1,        OpCode32SimdMemSingle.Create);
             SetA32("111101000x00xxxxxxxx0111xxxxxxxx", InstName.Vst1,        InstEmit32.Vst1,        OpCode32SimdMemPair.Create); // Regs = 1.
@@ -972,15 +986,7 @@ namespace ARMeilleure.Decoders
             SetA32("111101000x00xxxxxxxx010xxxxxxxxx", InstName.Vst3,        InstEmit32.Vst3,        OpCode32SimdMemPair.Create); // Inc = 1/2 (itype).
             SetA32("111101001x00xxxxxxxx<<11xxxxxxxx", InstName.Vst4,        InstEmit32.Vst4,        OpCode32SimdMemSingle.Create);
             SetA32("111101000x00xxxxxxxx000xxxxxxxxx", InstName.Vst4,        InstEmit32.Vst4,        OpCode32SimdMemPair.Create); // Inc = 1/2 (itype).
-            SetA32("<<<<11001x00xxxxxxxx1011xxxxxxx0", InstName.Vstm,        InstEmit32.Vstm,        OpCode32SimdMemMult.Create);
-            SetA32("<<<<11001x10xxxxxxxx1011xxxxxxx0", InstName.Vstm,        InstEmit32.Vstm,        OpCode32SimdMemMult.Create);
-            SetA32("<<<<11010x10xxxxxxxx1011xxxxxxx0", InstName.Vstm,        InstEmit32.Vstm,        OpCode32SimdMemMult.Create);
-            SetA32("<<<<11001x00xxxxxxxx1010xxxxxxxx", InstName.Vstm,        InstEmit32.Vstm,        OpCode32SimdMemMult.Create);
-            SetA32("<<<<11001x10xxxxxxxx1010xxxxxxxx", InstName.Vstm,        InstEmit32.Vstm,        OpCode32SimdMemMult.Create);
-            SetA32("<<<<11010x10xxxxxxxx1010xxxxxxxx", InstName.Vstm,        InstEmit32.Vstm,        OpCode32SimdMemMult.Create);
-            SetA32("<<<<1101xx00xxxxxxxx101xxxxxxxxx", InstName.Vstr,        InstEmit32.Vstr,        OpCode32SimdMemImm.Create);
             SetA32("111100110xxxxxxxxxxx1000xxx0xxxx", InstName.Vsub,        InstEmit32.Vsub_I,      OpCode32SimdReg.Create);
-            SetA32("<<<<11100x11xxxxxxxx101xx1x0xxxx", InstName.Vsub,        InstEmit32.Vsub_S,      OpCode32SimdRegS.Create);
             SetA32("111100100x10xxxxxxxx1101xxx0xxxx", InstName.Vsub,        InstEmit32.Vsub_V,      OpCode32SimdReg.Create);
             SetA32("1111001x1x<<xxxxxxx00010x0x0xxxx", InstName.Vsubl,       InstEmit32.Vsubl_I,     OpCode32SimdRegLong.Create);
             SetA32("1111001x1x<<xxxxxxx00011x0x0xxxx", InstName.Vsubw,       InstEmit32.Vsubw_I,     OpCode32SimdRegWide.Create);
@@ -1248,6 +1254,18 @@ namespace ARMeilleure.Decoders
             Set(reversedEncoding, AllInstT32, new InstDescriptor(name, emitter), reversedMakeOp);
         }
 
+        private static void SetVfp(string encoding, InstName name, InstEmitter emitter, MakeOp makeOpA32, MakeOp makeOpT32)
+        {
+            SetA32(encoding, name, emitter, makeOpA32);
+
+            string thumbEncoding = encoding;
+            if (thumbEncoding.StartsWith("<<<<"))
+            {
+                thumbEncoding = "1110" + thumbEncoding.Substring(4);
+            }
+            SetT32(thumbEncoding, name, emitter, makeOpT32);
+        }
+
         private static void SetA64(string encoding, InstName name, InstEmitter emitter, MakeOp makeOp)
         {
             Set(encoding, AllInstA64, new InstDescriptor(name, emitter), makeOp);

ARMeilleure/Instructions/InstEmitAluHelper.cs

@@ -128,7 +128,7 @@ namespace ARMeilleure.Instructions
         {
             Debug.Assert(value.Type == OperandType.I32);
 
-            if (((OpCode32)context.CurrOp).IsThumb())
+            if (((OpCode32)context.CurrOp).IsThumb)
             {
                 bool isReturn = IsA32Return(context);
                 if (!isReturn)

ARMeilleure/Instructions/InstEmitFlow32.cs

@@ -34,7 +34,7 @@ namespace ARMeilleure.Instructions
 
             uint pc = op.GetPc();
 
-            bool isThumb = ((OpCode32)context.CurrOp).IsThumb();
+            bool isThumb = ((OpCode32)context.CurrOp).IsThumb;
 
             uint currentPc = isThumb
                 ? pc | 1
@@ -61,7 +61,7 @@ namespace ARMeilleure.Instructions
             Operand addr = context.Copy(GetIntA32(context, op.Rm));
             Operand bitOne = context.BitwiseAnd(addr, Const(1));
 
-            bool isThumb = ((OpCode32)context.CurrOp).IsThumb();
+            bool isThumb = ((OpCode32)context.CurrOp).IsThumb;
 
             uint currentPc = isThumb
                 ? (pc - 2) | 1

ARMeilleure/Instructions/InstEmitSimdCvt32.cs

@@ -323,6 +323,60 @@ namespace ARMeilleure.Instructions
             }
         }
 
+        // VRINTA (vector).
+        public static void Vrinta_V(ArmEmitterContext context)
+        {
+            EmitVectorUnaryOpF32(context, (m) => EmitRoundMathCall(context, MidpointRounding.AwayFromZero, m));
+        }
+
+        // VRINTM (vector).
+        public static void Vrintm_V(ArmEmitterContext context)
+        {
+            if (Optimizations.UseSse2)
+            {
+                EmitVectorUnaryOpSimd32(context, (m) =>
+                {
+                    return context.AddIntrinsic(Intrinsic.X86Roundps, m, Const(X86GetRoundControl(FPRoundingMode.TowardsMinusInfinity)));
+                });
+            }
+            else
+            {
+                EmitVectorUnaryOpF32(context, (m) => EmitUnaryMathCall(context, nameof(Math.Floor), m));
+            }
+        }
+
+        // VRINTN (vector).
+        public static void Vrintn_V(ArmEmitterContext context)
+        {
+            if (Optimizations.UseSse2)
+            {
+                EmitVectorUnaryOpSimd32(context, (m) =>
+                {
+                    return context.AddIntrinsic(Intrinsic.X86Roundps, m, Const(X86GetRoundControl(FPRoundingMode.ToNearest)));
+                });
+            }
+            else
+            {
+                EmitVectorUnaryOpF32(context, (m) => EmitRoundMathCall(context, MidpointRounding.ToEven, m));
+            }
+        }
+
+        // VRINTP (vector).
+        public static void Vrintp_V(ArmEmitterContext context)
+        {
+            if (Optimizations.UseSse2)
+            {
+                EmitVectorUnaryOpSimd32(context, (m) =>
+                {
+                    return context.AddIntrinsic(Intrinsic.X86Roundps, m, Const(X86GetRoundControl(FPRoundingMode.TowardsPlusInfinity)));
+                });
+            }
+            else
+            {
+                EmitVectorUnaryOpF32(context, (m) => EmitUnaryMathCall(context, nameof(Math.Ceiling), m));
+            }
+        }
+
         // VRINTZ (floating-point).
         public static void Vrint_Z(ArmEmitterContext context)
         {

ARMeilleure/Instructions/InstName.cs

@@ -636,6 +636,10 @@ namespace ARMeilleure.Instructions
         Vrev,
         Vrhadd,
         Vrint,
+        Vrinta,
+        Vrintm,
+        Vrintn,
+        Vrintp,
         Vrintx,
         Vrshr,
         Vrshrn,

Ryujinx.Graphics.Gpu/Engine/Threed/SemaphoreUpdater.cs

@@ -1,4 +1,5 @@
 using Ryujinx.Graphics.GAL;
+using System;
 
 namespace Ryujinx.Graphics.Gpu.Engine.Threed
 {
@@ -151,10 +152,21 @@ namespace Ryujinx.Graphics.Gpu.Engine.Threed
 
             ulong ticks = _context.GetTimestamp();
 
+            float divisor = type switch
+            {
+                ReportCounterType.SamplesPassed => _channel.TextureManager.RenderTargetScale * _channel.TextureManager.RenderTargetScale,
+                _ => 1f
+            };
+
             ICounterEvent counter = null;
 
             void resultHandler(object evt, ulong result)
             {
+                if (divisor != 1f)
+                {
+                    result = (ulong)MathF.Ceiling(result / divisor);
+                }
+
                 CounterData counterData = new CounterData
                 {
                     Counter = result,

Ryujinx.Graphics.Vulkan/FormatCapabilities.cs

@@ -108,6 +108,10 @@ namespace Ryujinx.Graphics.Vulkan
                 {
                     format = VkFormat.D32SfloatS8Uint;
                 }
+                else if (srcFormat == GAL.Format.R4G4B4A4Unorm)
+                {
+                    format = VkFormat.R4G4B4A4UnormPack16;
+                }
                 else
                 {
                     Logger.Error?.Print(LogClass.Gpu, $"Format {srcFormat} is not supported by the host.");

Ryujinx.Graphics.Vulkan/FormatTable.cs

@@ -6,11 +6,11 @@ namespace Ryujinx.Graphics.Vulkan
 {
     static class FormatTable
     {
-        private static readonly VkFormat[] Table;
+        private static readonly VkFormat[] _table;
 
         static FormatTable()
         {
-            Table = new VkFormat[Enum.GetNames(typeof(Format)).Length];
+            _table = new VkFormat[Enum.GetNames(typeof(Format)).Length];
 
             Add(Format.R8Unorm,             VkFormat.R8Unorm);
             Add(Format.R8Snorm,             VkFormat.R8SNorm);
@@ -68,7 +68,7 @@ namespace Ryujinx.Graphics.Vulkan
             Add(Format.D32FloatS8Uint,      VkFormat.D32SfloatS8Uint);
             Add(Format.R8G8B8A8Srgb,        VkFormat.R8G8B8A8Srgb);
             Add(Format.R4G4Unorm,           VkFormat.R4G4UnormPack8);
-            Add(Format.R4G4B4A4Unorm,       VkFormat.R4G4B4A4UnormPack16);
+            Add(Format.R4G4B4A4Unorm,       VkFormat.A4B4G4R4UnormPack16Ext);
             Add(Format.R5G5B5X1Unorm,       VkFormat.A1R5G5B5UnormPack16);
             Add(Format.R5G5B5A1Unorm,       VkFormat.A1R5G5B5UnormPack16);
             Add(Format.R5G6B5Unorm,         VkFormat.R5G6B5UnormPack16);
@@ -161,12 +161,12 @@ namespace Ryujinx.Graphics.Vulkan
 
         private static void Add(Format format, VkFormat vkFormat)
         {
-            Table[(int)format] = vkFormat;
+            _table[(int)format] = vkFormat;
         }
 
         public static VkFormat GetFormat(Format format)
         {
-            return Table[(int)format];
+            return _table[(int)format];
         }
     }
 }

Ryujinx.Graphics.Vulkan/TextureView.cs

@@ -74,17 +74,7 @@ namespace Ryujinx.Graphics.Vulkan
                 swizzleR = swizzleB;
                 swizzleB = temp;
             }
-            else if (info.Format == GAL.Format.R4G4B4A4Unorm)
-            {
-                var tempG = swizzleG;
-                var tempB = swizzleB;
-
-                swizzleB = swizzleA;
-                swizzleG = swizzleR;
-                swizzleR = tempG;
-                swizzleA = tempB;
-            }
-            else if (info.Format == GAL.Format.A1B5G5R5Unorm)
+            else if (VkFormat == VkFormat.R4G4B4A4UnormPack16 || info.Format == GAL.Format.A1B5G5R5Unorm)
             {
                 var tempB = swizzleB;
                 var tempA = swizzleA;

Ryujinx.Tests/Cpu/CpuTestSimdCvt32.cs

@@ -13,6 +13,16 @@ namespace Ryujinx.Tests.Cpu
 #if SimdCvt32
 
 #region "ValueSource (Opcodes)"
+        private static uint[] _Vrint_AMNP_V_F32_()
+        {
+            return new uint[]
+            {
+                0xf3ba0500u, // VRINTA.F32 Q0, Q0
+                0xf3ba0680u, // VRINTM.F32 Q0, Q0
+                0xf3ba0400u, // VRINTN.F32 Q0, Q0
+                0xf3ba0780u  // VRINTP.F32 Q0, Q0
+            };
+        }
 #endregion
 
 #region "ValueSource (Types)"
@@ -64,6 +74,47 @@ namespace Ryujinx.Tests.Cpu
             }
         }
 
+        private static IEnumerable<ulong> _2S_F_()
+        {
+            yield return 0xFF7FFFFFFF7FFFFFul; // -Max Normal    (float.MinValue)
+            yield return 0x8080000080800000ul; // -Min Normal
+            yield return 0x807FFFFF807FFFFFul; // -Max Subnormal
+            yield return 0x8000000180000001ul; // -Min Subnormal (-float.Epsilon)
+            yield return 0x7F7FFFFF7F7FFFFFul; // +Max Normal    (float.MaxValue)
+            yield return 0x0080000000800000ul; // +Min Normal
+            yield return 0x007FFFFF007FFFFFul; // +Max Subnormal
+            yield return 0x0000000100000001ul; // +Min Subnormal (float.Epsilon)
+
+            if (!NoZeros)
+            {
+                yield return 0x8000000080000000ul; // -Zero
+                yield return 0x0000000000000000ul; // +Zero
+            }
+
+            if (!NoInfs)
+            {
+                yield return 0xFF800000FF800000ul; // -Infinity
+                yield return 0x7F8000007F800000ul; // +Infinity
+            }
+
+            if (!NoNaNs)
+            {
+                yield return 0xFFC00000FFC00000ul; // -QNaN (all zeros payload) (float.NaN)
+                yield return 0xFFBFFFFFFFBFFFFFul; // -SNaN (all ones  payload)
+                yield return 0x7FC000007FC00000ul; // +QNaN (all zeros payload) (-float.NaN) (DefaultNaN)
+                yield return 0x7FBFFFFF7FBFFFFFul; // +SNaN (all ones  payload)
+            }
+
+            for (int cnt = 1; cnt <= RndCnt; cnt++)
+            {
+                ulong rnd1 = GenNormalS();
+                ulong rnd2 = GenSubnormalS();
+
+                yield return (rnd1 << 32) | rnd1;
+                yield return (rnd2 << 32) | rnd2;
+            }
+        }
+
         private static IEnumerable<ulong> _1D_F_()
         {
             yield return 0xFFEFFFFFFFFFFFFFul; // -Max Normal    (double.MinValue)
@@ -224,6 +275,35 @@ namespace Ryujinx.Tests.Cpu
             CompareAgainstUnicorn();
         }
 
+        [Test, Pairwise] [Explicit]
+        public void Vrint_AMNP_V_F32([ValueSource(nameof(_Vrint_AMNP_V_F32_))] uint opcode,
+                                     [Values(0u, 1u, 2u, 3u)] uint rd,
+                                     [Values(0u, 1u, 2u, 3u)] uint rm,
+                                     [ValueSource(nameof(_2S_F_))] ulong d0,
+                                     [ValueSource(nameof(_2S_F_))] ulong d1,
+                                     [ValueSource(nameof(_2S_F_))] ulong d2,
+                                     [ValueSource(nameof(_2S_F_))] ulong d3,
+                                     [Values] bool q)
+        {
+            if (q)
+            {
+                opcode |= 1 << 6;
+
+                rd >>= 1; rd <<= 1;
+                rm >>= 1; rm <<= 1;
+            }
+
+            opcode |= ((rd & 0xf) << 12) | ((rd & 0x10) << 18);
+            opcode |= ((rm & 0xf) << 0)  | ((rm & 0x10) << 1);
+
+            V128 v0 = MakeVectorE0E1(d0, d1);
+            V128 v1 = MakeVectorE0E1(d2, d3);
+
+            SingleOpcode(opcode, v0: v0, v1: v1);
+
+            CompareAgainstUnicorn();
+        }
+
         [Test, Pairwise, Description("VRINTX.F<size> <Sd>, <Sm>")]
         public void Vrintx_S([Values(0u, 1u)] uint rd,
                              [Values(0u, 1u)] uint rm,
@@ -253,7 +333,7 @@ namespace Ryujinx.Tests.Cpu
             }
 
             opcode |= ((size & 3) << 8);
-            
+
             int fpscr = (int)rMode << (int)Fpcr.RMode;
             SingleOpcode(opcode, v0: v0, v1: v1, v2: v2, fpscr: fpscr);