Turn off vertex range culling in bezier/spline calls.

When we do lower res tess than the real PSP, we cant trust the game to not cause range culling to kick in. Fixes hrydgard#11692
konistehrad · Dec 17, 2020 · ba79de0 · ba79de0
1 parent 8fea940
commit ba79de0
Show file tree

Hide file tree

Showing 11 changed files with 43 additions and 19 deletions.
diff --git a/Common/GPU/ShaderWriter.cpp b/Common/GPU/ShaderWriter.cpp
@@ -83,6 +83,8 @@ void ShaderWriter::Preamble(const char **gl_extensions, size_t num_gl_extensions
 		case ShaderStage::Fragment:
 			W(vulkan_glsl_preamble_fs);
 			break;
+		default:
+			break;
 		}
 		break;
 	case HLSL_D3D11:
@@ -99,6 +101,8 @@ void ShaderWriter::Preamble(const char **gl_extensions, size_t num_gl_extensions
 				W(hlsl_d3d11_preamble_fs);
 			}
 			break;
+		default:
+			break;
 		}
 		break;
 	default:  // OpenGL
@@ -125,6 +129,8 @@ void ShaderWriter::Preamble(const char **gl_extensions, size_t num_gl_extensions
 			}
 			C("#define gl_VertexIndex gl_VertexID\n");
 			break;
+		default:
+			break;
 		}
 		if (!lang_.gles) {
 			C("#define lowp\n");
@@ -293,6 +299,8 @@ void ShaderWriter::DeclareSampler2D(const char *name, int binding) {
 	case HLSL_D3D11:
 		F("SamplerState %s : register(s%d);\n", name, binding);
 		break;
+	default:
+		break;
 	}
 }
 

diff --git a/GPU/Common/ShaderId.cpp b/GPU/Common/ShaderId.cpp
@@ -56,6 +56,7 @@ std::string VertexShaderDesc(const VShaderID &id) {
 	if (id.Bit(VS_BIT_HAS_TEXCOORD_TESS)) desc << "TessT ";
 	if (id.Bit(VS_BIT_HAS_NORMAL_TESS)) desc << "TessN ";
 	if (id.Bit(VS_BIT_NORM_REVERSE_TESS)) desc << "TessRevN ";
+	if (id.Bit(VS_BIT_VERTEX_RANGE_CULLING)) desc << "Cull ";
 
 	return desc.str();
 }
@@ -70,17 +71,20 @@ void ComputeVertexShaderID(VShaderID *id_out, u32 vertType, bool useHWTransform,
 	bool hasNormal = (vertType & GE_VTYPE_NRM_MASK) != 0;
 	bool hasTexcoord = (vertType & GE_VTYPE_TC_MASK) != 0;
 
-	bool doBezier = gstate_c.bezier;
-	bool doSpline = gstate_c.spline;
+	bool doBezier = gstate_c.submitType == SubmitType::HW_BEZIER;
+	bool doSpline = gstate_c.submitType == SubmitType::HW_SPLINE;
 
 	bool enableFog = gstate.isFogEnabled() && !isModeThrough && !gstate.isModeClear();
 	bool lmode = gstate.isUsingSecondaryColor() && gstate.isLightingEnabled() && !isModeThrough;
+	bool vertexRangeCulling = gstate_c.Supports(GPU_SUPPORTS_VS_RANGE_CULLING) &&
+		!isModeThrough && gstate_c.submitType == SubmitType::DRAW;  // neither hw nor sw spline/bezier. See #11692
 
 	VShaderID id;
 	id.SetBit(VS_BIT_LMODE, lmode);
 	id.SetBit(VS_BIT_IS_THROUGH, isModeThrough);
 	id.SetBit(VS_BIT_ENABLE_FOG, enableFog);
 	id.SetBit(VS_BIT_HAS_COLOR, hasColor);
+	id.SetBit(VS_BIT_VERTEX_RANGE_CULLING, vertexRangeCulling);
 
 	if (doTexture) {
 		id.SetBit(VS_BIT_DO_TEXTURE);

diff --git a/GPU/Common/ShaderId.h b/GPU/Common/ShaderId.h
@@ -15,7 +15,7 @@ enum VShaderBit : uint8_t {
 	VS_BIT_ENABLE_FOG = 2,
 	VS_BIT_HAS_COLOR = 3,
 	VS_BIT_DO_TEXTURE = 4,
-	// 5 is free.
+	VS_BIT_VERTEX_RANGE_CULLING = 5,
 	// 6 is free,
 	// 7 is free.
 	VS_BIT_USE_HW_TRANSFORM = 8,

diff --git a/GPU/Common/ShaderUniforms.cpp b/GPU/Common/ShaderUniforms.cpp
@@ -215,7 +215,7 @@ void BaseUpdateUniforms(UB_VS_FS_Base *ub, uint64_t dirtyUniforms, bool flipView
 		const int h = gstate.getTextureHeight(0);
 		const float widthFactor = (float)w * invW;
 		const float heightFactor = (float)h * invH;
-		if (gstate_c.bezier || gstate_c.spline) {
+		if (gstate_c.submitType == SubmitType::HW_BEZIER || gstate_c.submitType == SubmitType::HW_SPLINE) {
 			// When we are generating UV coordinates through the bezier/spline, we need to apply the scaling.
 			// However, this is missing a check that we're not getting our UV:s supplied for us in the vertices.
 			ub->uvScaleOffset[0] = gstate_c.uv.uScale * widthFactor;

diff --git a/GPU/Common/VertexShaderGenerator.cpp b/GPU/Common/VertexShaderGenerator.cpp
@@ -207,6 +207,8 @@ bool GenerateVertexShader(const VShaderID &id, char *buffer, const ShaderLanguag
 	}
 	bool texCoordInVec3 = false;
 
+	bool vertexRangeCulling = id.Bit(VS_BIT_VERTEX_RANGE_CULLING);
+
 	if (compat.shaderLanguage == GLSL_VULKAN) {
 		WRITE(p, "\n");
 		WRITE(p, "layout (std140, set = 0, binding = 3) uniform baseVars {\n%s};\n", ub_baseStr);
@@ -1086,7 +1088,7 @@ bool GenerateVertexShader(const VShaderID &id, char *buffer, const ShaderLanguag
 			WRITE(p, "  %sv_fogdepth = (viewPos.z + u_fogcoef.x) * u_fogcoef.y;\n", compat.vsOutPrefix);
 	}
 
-	if (!isModeThrough && gstate_c.Supports(GPU_SUPPORTS_VS_RANGE_CULLING)) {
+	if (vertexRangeCulling) {
 		WRITE(p, "  vec3 projPos = outPos.xyz / outPos.w;\n");
 		// Vertex range culling doesn't happen when depth is clamped, so only do this if in range.
 		WRITE(p, "  if (u_cullRangeMin.w <= 0.0 || (projPos.z >= u_cullRangeMin.z && projPos.z <= u_cullRangeMax.z)) {\n");

diff --git a/GPU/D3D11/DrawEngineD3D11.cpp b/GPU/D3D11/DrawEngineD3D11.cpp
@@ -341,7 +341,7 @@ void DrawEngineD3D11::DoFlush() {
 	ApplyDrawState(prim);
 
 	// Always use software for flat shading to fix the provoking index.
-	bool tess = gstate_c.bezier || gstate_c.spline;
+	bool tess = gstate_c.submitType == SubmitType::HW_BEZIER || gstate_c.submitType == SubmitType::HW_SPLINE;
 	bool useHWTransform = CanUseHardwareTransform(prim) && (tess || gstate.getShadeMode() != GE_SHADE_FLAT);
 
 	if (useHWTransform) {

diff --git a/GPU/Directx9/ShaderManagerDX9.cpp b/GPU/Directx9/ShaderManagerDX9.cpp
@@ -543,7 +543,7 @@ void ShaderManagerDX9::DirtyLastShader() { // disables vertex arrays
 
 VSShader *ShaderManagerDX9::ApplyShader(bool useHWTransform, bool useHWTessellation, u32 vertType) {
 	// Always use software for flat shading to fix the provoking index.
-	bool tess = gstate_c.bezier || gstate_c.spline;
+	bool tess = gstate_c.submitType == SubmitType::HW_BEZIER || gstate_c.submitType == SubmitType::HW_SPLINE;
 	useHWTransform = useHWTransform && (tess || gstate.getShadeMode() != GE_SHADE_FLAT);
 
 	VShaderID VSID;

diff --git a/GPU/GLES/ShaderManagerGLES.cpp b/GPU/GLES/ShaderManagerGLES.cpp
@@ -409,7 +409,7 @@ void LinkedShader::UpdateUniforms(u32 vertType, const ShaderID &vsid, bool useBu
 		const float widthFactor = (float)w * invW;
 		const float heightFactor = (float)h * invH;
 		float uvscaleoff[4];
-		if (gstate_c.bezier || gstate_c.spline) {
+		if (gstate_c.submitType == SubmitType::HW_BEZIER || gstate_c.submitType == SubmitType::HW_SPLINE) {
 			// When we are generating UV coordinates through the bezier/spline, we need to apply the scaling.
 			// However, this is missing a check that we're not getting our UV:s supplied for us in the vertices.
 			uvscaleoff[0] = gstate_c.uv.uScale * widthFactor;

diff --git a/GPU/GPUCommon.cpp b/GPU/GPUCommon.cpp
@@ -1830,20 +1830,21 @@ void GPUCommon::Execute_Bezier(u32 op, u32 diff) {
 
 	if (drawEngineCommon_->CanUseHardwareTessellation(surface.primType)) {
 		gstate_c.Dirty(DIRTY_VERTEXSHADER_STATE);
-		gstate_c.bezier = true;
+		gstate_c.submitType = SubmitType::HW_BEZIER;
 		if (gstate_c.spline_num_points_u != surface.num_points_u) {
 			gstate_c.Dirty(DIRTY_BEZIERSPLINE);
 			gstate_c.spline_num_points_u = surface.num_points_u;
 		}
+	} else {
+		gstate_c.submitType = SubmitType::BEZIER;
 	}
 
 	int bytesRead = 0;
 	UpdateUVScaleOffset();
 	drawEngineCommon_->SubmitCurve(control_points, indices, surface, gstate.vertType, &bytesRead, "bezier");
 
-	if (gstate_c.bezier)
-		gstate_c.Dirty(DIRTY_VERTEXSHADER_STATE);
-	gstate_c.bezier = false;
+	gstate_c.Dirty(DIRTY_VERTEXSHADER_STATE);
+	gstate_c.submitType = SubmitType::DRAW;
 
 	// After drawing, we advance pointers - see SubmitPrim which does the same.
 	int count = surface.num_points_u * surface.num_points_v;
@@ -1896,20 +1897,21 @@ void GPUCommon::Execute_Spline(u32 op, u32 diff) {
 
 	if (drawEngineCommon_->CanUseHardwareTessellation(surface.primType)) {
 		gstate_c.Dirty(DIRTY_VERTEXSHADER_STATE);
-		gstate_c.spline = true;
+		gstate_c.submitType = SubmitType::HW_SPLINE;
 		if (gstate_c.spline_num_points_u != surface.num_points_u) {
 			gstate_c.Dirty(DIRTY_BEZIERSPLINE);
 			gstate_c.spline_num_points_u = surface.num_points_u;
 		}
+	} else {
+		gstate_c.submitType = SubmitType::SPLINE;
 	}
 
 	int bytesRead = 0;
 	UpdateUVScaleOffset();
 	drawEngineCommon_->SubmitCurve(control_points, indices, surface, gstate.vertType, &bytesRead, "spline");
 
-	if (gstate_c.spline)
-		gstate_c.Dirty(DIRTY_VERTEXSHADER_STATE);
-	gstate_c.spline = false;
+	gstate_c.Dirty(DIRTY_VERTEXSHADER_STATE);
+	gstate_c.submitType = SubmitType::DRAW;
 
 	// After drawing, we advance pointers - see SubmitPrim which does the same.
 	int count = surface.num_points_u * surface.num_points_v;

diff --git a/GPU/GPUState.h b/GPU/GPUState.h
@@ -504,6 +504,14 @@ struct KnownVertexBounds {
 	u16 maxV;
 };
 
+enum class SubmitType {
+	DRAW,
+	BEZIER,
+	SPLINE,
+	HW_BEZIER,
+	HW_SPLINE,
+};
+
 struct GPUStateCache {
 	bool Supports(u32 flags) { return (featureFlags & flags) != 0; } // Return true if ANY of flags are true.
 	bool SupportsAll(u32 flags) { return (featureFlags & flags) == flags; } // Return true if ALL flags are true.
@@ -602,8 +610,8 @@ struct GPUStateCache {
 	}
 	u32 curRTOffsetX;
 
-	bool bezier;
-	bool spline;
+	// Set if we are doing hardware bezier/spline.
+	SubmitType submitType;
 	int spline_num_points_u;
 
 	bool useShaderDepal;

diff --git a/GPU/Vulkan/DrawEngineVulkan.cpp b/GPU/Vulkan/DrawEngineVulkan.cpp
@@ -598,7 +598,7 @@ void DrawEngineVulkan::DoFlush() {
 
 	FrameData *frame = &frame_[vulkan_->GetCurFrame()];
 
-	bool tess = gstate_c.bezier || gstate_c.spline;
+	bool tess = gstate_c.submitType == SubmitType::HW_BEZIER || gstate_c.submitType == SubmitType::HW_SPLINE;
 
 	bool textureNeedsApply = false;
 	if (gstate_c.IsDirty(DIRTY_TEXTURE_IMAGE | DIRTY_TEXTURE_PARAMS) && !gstate.isModeClear() && gstate.isTextureMapEnabled()) {