shader/other: Implement thread comparisons (NV_shader_thread_group)

Hardware S2R special registers match gl_Thread*MaskNV. We can trivially
implement these using Nvidia's extension on OpenGL or naively stubbing
them with the ARB instructions to match. This might cause issues if the
host device warp size doesn't match Nvidia's. That said, this is
unlikely on proper shaders.

Refer to the attached url for more documentation about these flags.
https://www.khronos.org/registry/OpenGL/extensions/NV/NV_shader_thread_group.txt

2 files changed, 26 insertions, 0 deletions

diff --git a/src/video_core/shader/decode/other.cpp b/src/video_core/shader/decode/other.cpp
index d4f95b18c..399a455c4 100644
--- a/src/video_core/shader/decode/other.cpp
+++ b/src/video_core/shader/decode/other.cpp
@@ -109,6 +109,27 @@ u32 ShaderIR::DecodeOther(NodeBlock& bb, u32 pc) {
                 return Operation(OperationCode::WorkGroupIdY);
             case SystemVariable::CtaIdZ:
                 return Operation(OperationCode::WorkGroupIdZ);
+            case SystemVariable::EqMask:
+            case SystemVariable::LtMask:
+            case SystemVariable::LeMask:
+            case SystemVariable::GtMask:
+            case SystemVariable::GeMask:
+                uses_warps = true;
+                switch (instr.sys20) {
+                case SystemVariable::EqMask:
+                    return Operation(OperationCode::ThreadEqMask);
+                case SystemVariable::LtMask:
+                    return Operation(OperationCode::ThreadLtMask);
+                case SystemVariable::LeMask:
+                    return Operation(OperationCode::ThreadLeMask);
+                case SystemVariable::GtMask:
+                    return Operation(OperationCode::ThreadGtMask);
+                case SystemVariable::GeMask:
+                    return Operation(OperationCode::ThreadGeMask);
+                default:
+                    UNREACHABLE();
+                    return Immediate(0u);
+                }
             default:
                 UNIMPLEMENTED_MSG("Unhandled system move: {}",
                                   static_cast<u32>(instr.sys20.Value()));
diff --git a/src/video_core/shader/node.h b/src/video_core/shader/node.h
index f75b62240..cce8aeebe 100644
--- a/src/video_core/shader/node.h
+++ b/src/video_core/shader/node.h
@@ -226,6 +226,11 @@ enum class OperationCode {
     VoteEqual,    /// (bool) -> bool
 
     ThreadId,       /// () -> uint
+    ThreadEqMask,   /// () -> uint
+    ThreadGeMask,   /// () -> uint
+    ThreadGtMask,   /// () -> uint
+    ThreadLeMask,   /// () -> uint
+    ThreadLtMask,   /// () -> uint
     ShuffleIndexed, /// (uint value, uint index) -> uint
 
     MemoryBarrierGL, /// () -> void