Implement R16S & R16UI & R16I RenderTargetFormats & PixelFormats and more (R16_UNORM needed by Fate Extella) (#848)

* Implement R16S & R16UI & R16I RenderTargetFormats & PixelFormats


Do a separate function in order to get Bytes Per Pixel of DepthFormat


Apply the new function in gpu.h


delete unneeded white space

* correct merging error

4 files changed, 92 insertions, 19 deletions

diff --git a/src/video_core/gpu.cpp b/src/video_core/gpu.cpp
index 4ff4d71c5..b90937d17 100644
--- a/src/video_core/gpu.cpp
+++ b/src/video_core/gpu.cpp
@@ -34,19 +34,51 @@ u32 RenderTargetBytesPerPixel(RenderTargetFormat format) {
 
     switch (format) {
     case RenderTargetFormat::RGBA32_FLOAT:
+    case RenderTargetFormat::RGBA32_UINT:
         return 16;
     case RenderTargetFormat::RGBA16_FLOAT:
     case RenderTargetFormat::RG32_FLOAT:
         return 8;
     case RenderTargetFormat::RGBA8_UNORM:
+    case RenderTargetFormat::RGBA8_SRGB:
     case RenderTargetFormat::RGB10_A2_UNORM:
     case RenderTargetFormat::BGRA8_UNORM:
+    case RenderTargetFormat::RG16_UNORM:
+    case RenderTargetFormat::RG16_SNORM:
+    case RenderTargetFormat::RG16_UINT:
+    case RenderTargetFormat::RG16_SINT:
+    case RenderTargetFormat::RG16_FLOAT:
     case RenderTargetFormat::R32_FLOAT:
     case RenderTargetFormat::R11G11B10_FLOAT:
         return 4;
+    case RenderTargetFormat::R16_UNORM:
+    case RenderTargetFormat::R16_SNORM:
+    case RenderTargetFormat::R16_UINT:
+    case RenderTargetFormat::R16_SINT:
+    case RenderTargetFormat::R16_FLOAT:
+        return 2;
+    case RenderTargetFormat::R8_UNORM:
+        return 1;
     default:
         UNIMPLEMENTED_MSG("Unimplemented render target format {}", static_cast<u32>(format));
     }
 }
 
+u32 DepthFormatBytesPerPixel(DepthFormat format) {
+    switch (format) {
+    case DepthFormat::Z32_S8_X24_FLOAT:
+        return 8;
+    case DepthFormat::Z32_FLOAT:
+    case DepthFormat::S8_Z24_UNORM:
+    case DepthFormat::Z24_X8_UNORM:
+    case DepthFormat::Z24_S8_UNORM:
+    case DepthFormat::Z24_C8_UNORM:
+        return 4;
+    case DepthFormat::Z16_UNORM:
+        return 2;
+    default:
+        UNIMPLEMENTED_MSG("Unimplemented Depth format {}", static_cast<u32>(format));
+    }
+}
+
 } // namespace Tegra
diff --git a/src/video_core/gpu.h b/src/video_core/gpu.h
index 874eddd78..0164c747a 100644
--- a/src/video_core/gpu.h
+++ b/src/video_core/gpu.h
@@ -35,6 +35,10 @@ enum class RenderTargetFormat : u32 {
     R11G11B10_FLOAT = 0xE0,
     R32_FLOAT = 0xE5,
     B5G6R5_UNORM = 0xE8,
+    R16_UNORM = 0xEE,
+    R16_SNORM = 0xEF,
+    R16_SINT = 0xF0,
+    R16_UINT = 0xF1,
     R16_FLOAT = 0xF2,
     R8_UNORM = 0xF3,
 };
@@ -52,6 +56,9 @@ enum class DepthFormat : u32 {
 /// Returns the number of bytes per pixel of each rendertarget format.
 u32 RenderTargetBytesPerPixel(RenderTargetFormat format);
 
+/// Returns the number of bytes per pixel of each depth format.
+u32 DepthFormatBytesPerPixel(DepthFormat format);
+
 class DebugContext;
 
 /**
diff --git a/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp b/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp
index 9fb734b77..15a33ed9b 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp
@@ -122,6 +122,9 @@ static constexpr std::array<FormatTuple, SurfaceParams::MaxPixelFormat> tex_form
     {GL_R32F, GL_RED, GL_FLOAT, ComponentType::Float, false},                  // R32F
     {GL_R16F, GL_RED, GL_HALF_FLOAT, ComponentType::Float, false},             // R16F
     {GL_R16, GL_RED, GL_UNSIGNED_SHORT, ComponentType::UNorm, false},          // R16UNORM
+    {GL_R16_SNORM, GL_RED, GL_SHORT, ComponentType::SNorm, false},             // R16S
+    {GL_R16UI, GL_RED_INTEGER, GL_UNSIGNED_SHORT, ComponentType::UInt, false}, // R16UI
+    {GL_R16I, GL_RED_INTEGER, GL_SHORT, ComponentType::SInt, false},           // R16I
     {GL_RG16, GL_RG, GL_UNSIGNED_SHORT, ComponentType::UNorm, false},          // RG16
     {GL_RG16F, GL_RG, GL_HALF_FLOAT, ComponentType::Float, false},             // RG16F
     {GL_RG16UI, GL_RG_INTEGER, GL_UNSIGNED_SHORT, ComponentType::UInt, false}, // RG16UI
@@ -239,13 +242,14 @@ static constexpr std::array<void (*)(u32, u32, u32, std::vector<u8>&, Tegra::GPU
         MortonCopy<true, PixelFormat::G8R8>,         MortonCopy<true, PixelFormat::BGRA8>,
         MortonCopy<true, PixelFormat::RGBA32F>,      MortonCopy<true, PixelFormat::RG32F>,
         MortonCopy<true, PixelFormat::R32F>,         MortonCopy<true, PixelFormat::R16F>,
-        MortonCopy<true, PixelFormat::R16UNORM>,     MortonCopy<true, PixelFormat::RG16>,
-        MortonCopy<true, PixelFormat::RG16F>,        MortonCopy<true, PixelFormat::RG16UI>,
-        MortonCopy<true, PixelFormat::RG16I>,        MortonCopy<true, PixelFormat::RG16S>,
-        MortonCopy<true, PixelFormat::RGB32F>,       MortonCopy<true, PixelFormat::SRGBA8>,
-        MortonCopy<true, PixelFormat::Z24S8>,        MortonCopy<true, PixelFormat::S8Z24>,
-        MortonCopy<true, PixelFormat::Z32F>,         MortonCopy<true, PixelFormat::Z16>,
-        MortonCopy<true, PixelFormat::Z32FS8>,
+        MortonCopy<true, PixelFormat::R16UNORM>,     MortonCopy<true, PixelFormat::R16S>,
+        MortonCopy<true, PixelFormat::R16UI>,        MortonCopy<true, PixelFormat::R16I>,
+        MortonCopy<true, PixelFormat::RG16>,         MortonCopy<true, PixelFormat::RG16F>,
+        MortonCopy<true, PixelFormat::RG16UI>,       MortonCopy<true, PixelFormat::RG16I>,
+        MortonCopy<true, PixelFormat::RG16S>,        MortonCopy<true, PixelFormat::RGB32F>,
+        MortonCopy<true, PixelFormat::SRGBA8>,       MortonCopy<true, PixelFormat::Z24S8>,
+        MortonCopy<true, PixelFormat::S8Z24>,        MortonCopy<true, PixelFormat::Z32F>,
+        MortonCopy<true, PixelFormat::Z16>,          MortonCopy<true, PixelFormat::Z32FS8>,
 };
 
 static constexpr std::array<void (*)(u32, u32, u32, std::vector<u8>&, Tegra::GPUVAddr),
@@ -276,6 +280,9 @@ static constexpr std::array<void (*)(u32, u32, u32, std::vector<u8>&, Tegra::GPU
         MortonCopy<false, PixelFormat::R32F>,
         MortonCopy<false, PixelFormat::R16F>,
         MortonCopy<false, PixelFormat::R16UNORM>,
+        MortonCopy<false, PixelFormat::R16S>,
+        MortonCopy<false, PixelFormat::R16UI>,
+        MortonCopy<false, PixelFormat::R16I>,
         MortonCopy<false, PixelFormat::RG16>,
         MortonCopy<false, PixelFormat::RG16F>,
         MortonCopy<false, PixelFormat::RG16UI>,
diff --git a/src/video_core/renderer_opengl/gl_rasterizer_cache.h b/src/video_core/renderer_opengl/gl_rasterizer_cache.h
index 829a76dfe..e24ba8cfe 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer_cache.h
+++ b/src/video_core/renderer_opengl/gl_rasterizer_cache.h
@@ -46,22 +46,25 @@ struct SurfaceParams {
         R32F = 20,
         R16F = 21,
         R16UNORM = 22,
-        RG16 = 23,
-        RG16F = 24,
-        RG16UI = 25,
-        RG16I = 26,
-        RG16S = 27,
-        RGB32F = 28,
-        SRGBA8 = 29,
+        R16S = 23,
+        R16UI = 24,
+        R16I = 25,
+        RG16 = 26,
+        RG16F = 27,
+        RG16UI = 28,
+        RG16I = 29,
+        RG16S = 30,
+        RGB32F = 31,
+        SRGBA8 = 32,
 
         MaxColorFormat,
 
         // DepthStencil formats
-        Z24S8 = 30,
-        S8Z24 = 31,
-        Z32F = 32,
-        Z16 = 33,
-        Z32FS8 = 34,
+        Z24S8 = 33,
+        S8Z24 = 34,
+        Z32F = 35,
+        Z16 = 36,
+        Z32FS8 = 37,
 
         MaxDepthStencilFormat,
 
@@ -122,6 +125,9 @@ struct SurfaceParams {
             1, // R32F
             1, // R16F
             1, // R16UNORM
+            1, // R16S
+            1, // R16UI
+            1, // R16I
             1, // RG16
             1, // RG16F
             1, // RG16UI
@@ -168,6 +174,9 @@ struct SurfaceParams {
             32,  // R32F
             16,  // R16F
             16,  // R16UNORM
+            16,  // R16S
+            16,  // R16UI
+            16,  // R16I
             32,  // RG16
             32,  // RG16F
             32,  // RG16UI
@@ -245,6 +254,14 @@ struct SurfaceParams {
             return PixelFormat::RG16S;
         case Tegra::RenderTargetFormat::R16_FLOAT:
             return PixelFormat::R16F;
+        case Tegra::RenderTargetFormat::R16_UNORM:
+            return PixelFormat::R16UNORM;
+        case Tegra::RenderTargetFormat::R16_SNORM:
+            return PixelFormat::R16S;
+        case Tegra::RenderTargetFormat::R16_UINT:
+            return PixelFormat::R16UI;
+        case Tegra::RenderTargetFormat::R16_SINT:
+            return PixelFormat::R16I;
         case Tegra::RenderTargetFormat::R32_FLOAT:
             return PixelFormat::R32F;
         default:
@@ -293,6 +310,12 @@ struct SurfaceParams {
                 return PixelFormat::R16F;
             case Tegra::Texture::ComponentType::UNORM:
                 return PixelFormat::R16UNORM;
+            case Tegra::Texture::ComponentType::SNORM:
+                return PixelFormat::R16S;
+            case Tegra::Texture::ComponentType::UINT:
+                return PixelFormat::R16UI;
+            case Tegra::Texture::ComponentType::SINT:
+                return PixelFormat::R16I;
             }
             LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
                          static_cast<u32>(component_type));
@@ -376,9 +399,11 @@ struct SurfaceParams {
         case Tegra::RenderTargetFormat::RGB10_A2_UNORM:
         case Tegra::RenderTargetFormat::R8_UNORM:
         case Tegra::RenderTargetFormat::RG16_UNORM:
+        case Tegra::RenderTargetFormat::R16_UNORM:
         case Tegra::RenderTargetFormat::B5G6R5_UNORM:
             return ComponentType::UNorm;
         case Tegra::RenderTargetFormat::RG16_SNORM:
+        case Tegra::RenderTargetFormat::R16_SNORM:
             return ComponentType::SNorm;
         case Tegra::RenderTargetFormat::RGBA16_FLOAT:
         case Tegra::RenderTargetFormat::R11G11B10_FLOAT:
@@ -390,8 +415,10 @@ struct SurfaceParams {
             return ComponentType::Float;
         case Tegra::RenderTargetFormat::RGBA32_UINT:
         case Tegra::RenderTargetFormat::RG16_UINT:
+        case Tegra::RenderTargetFormat::R16_UINT:
             return ComponentType::UInt;
         case Tegra::RenderTargetFormat::RG16_SINT:
+        case Tegra::RenderTargetFormat::R16_SINT:
             return ComponentType::SInt;
         default:
             LOG_CRITICAL(HW_GPU, "Unimplemented format={}", static_cast<u32>(format));