Implement BFI, BRK, FLO, FSWZADD, PBK, SHFL and TXD shader instructions, misc. fixes

6 files changed, 53 insertions, 0 deletions

diff --git a/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/HelperFunctionNames.cs b/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/HelperFunctionNames.cs
new file mode 100644
index 00000000..f1540fbf
--- /dev/null
+++ b/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/HelperFunctionNames.cs
@@ -0,0 +1,11 @@
+namespace Ryujinx.Graphics.Shader.CodeGen.Glsl
+{
+    static class HelperFunctionNames
+    {
+        public static string Shuffle     = "Helper_Shuffle";
+        public static string ShuffleDown = "Helper_ShuffleDown";
+        public static string ShuffleUp   = "Helper_ShuffleUp";
+        public static string ShuffleXor  = "Helper_ShuffleXor";
+        public static string SwizzleAdd  = "Helper_SwizzleAdd";
+    }
+}
\ No newline at end of file
diff --git a/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/Shuffle.glsl b/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/Shuffle.glsl
new file mode 100644
index 00000000..380bc581
--- /dev/null
+++ b/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/Shuffle.glsl
@@ -0,0 +1,9 @@
+float Helper_Shuffle(float x, uint index, uint mask)
+{
+    uint clamp = mask & 0x1fu;
+    uint segMask = (mask >> 8) & 0x1fu;
+    uint minThreadId = gl_SubGroupInvocationARB & segMask;
+    uint maxThreadId = minThreadId | (clamp & ~segMask);
+    uint srcThreadId = (index & ~segMask) | minThreadId;
+    return (srcThreadId <= maxThreadId) ? readInvocationARB(x, srcThreadId) : x;
+}
\ No newline at end of file
diff --git a/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/ShuffleDown.glsl b/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/ShuffleDown.glsl
new file mode 100644
index 00000000..46750f20
--- /dev/null
+++ b/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/ShuffleDown.glsl
@@ -0,0 +1,9 @@
+float Helper_ShuffleDown(float x, uint index, uint mask)
+{
+    uint clamp = mask & 0x1fu;
+    uint segMask = (mask >> 8) & 0x1fu;
+    uint minThreadId = gl_SubGroupInvocationARB & segMask;
+    uint maxThreadId = minThreadId | (clamp & ~segMask);
+    uint srcThreadId = gl_SubGroupInvocationARB + index;
+    return (srcThreadId <= maxThreadId) ? readInvocationARB(x, srcThreadId) : x;
+}
\ No newline at end of file
diff --git a/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/ShuffleUp.glsl b/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/ShuffleUp.glsl
new file mode 100644
index 00000000..2bc83469
--- /dev/null
+++ b/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/ShuffleUp.glsl
@@ -0,0 +1,8 @@
+float Helper_ShuffleUp(float x, uint index, uint mask)
+{
+    uint clamp = mask & 0x1fu;
+    uint segMask = (mask >> 8) & 0x1fu;
+    uint minThreadId = gl_SubGroupInvocationARB & segMask;
+    uint srcThreadId = gl_SubGroupInvocationARB - index;
+    return (srcThreadId >= minThreadId) ? readInvocationARB(x, srcThreadId) : x;
+}
\ No newline at end of file
diff --git a/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/ShuffleXor.glsl b/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/ShuffleXor.glsl
new file mode 100644
index 00000000..1049e181
--- /dev/null
+++ b/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/ShuffleXor.glsl
@@ -0,0 +1,9 @@
+float Helper_ShuffleXor(float x, uint index, uint mask)
+{
+    uint clamp = mask & 0x1fu;
+    uint segMask = (mask >> 8) & 0x1fu;
+    uint minThreadId = gl_SubGroupInvocationARB & segMask;
+    uint maxThreadId = minThreadId | (clamp & ~segMask);
+    uint srcThreadId = gl_SubGroupInvocationARB ^ index;
+    return (srcThreadId <= maxThreadId) ? readInvocationARB(x, srcThreadId) : x;
+}
\ No newline at end of file
diff --git a/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/SwizzleAdd.glsl b/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/SwizzleAdd.glsl
new file mode 100644
index 00000000..7df3e57f
--- /dev/null
+++ b/Ryujinx.Graphics.Shader/CodeGen/Glsl/HelperFunctions/SwizzleAdd.glsl
@@ -0,0 +1,7 @@
+float Helper_SwizzleAdd(float x, float y, int mask)
+{
+    vec4 xLut = vec4(1.0, -1.0, 1.0, 0.0);
+    vec4 yLut = vec4(1.0, 1.0, -1.0, 1.0);
+    int lutIdx = mask >> int(gl_SubGroupInvocationARB & 3u) * 2;
+    return x * xLut[lutIdx] + y * yLut[lutIdx];
+}
\ No newline at end of file