aboutsummaryrefslogtreecommitdiff
path: root/src/video_core
diff options
context:
space:
mode:
Diffstat (limited to 'src/video_core')
-rw-r--r--src/video_core/renderer_opengl/gl_rasterizer.cpp22
-rw-r--r--src/video_core/renderer_opengl/gl_shader_cache.cpp68
-rw-r--r--src/video_core/renderer_opengl/gl_shader_cache.h21
-rw-r--r--src/video_core/shader/decode/half_set_predicate.cpp19
4 files changed, 29 insertions, 101 deletions
diff --git a/src/video_core/renderer_opengl/gl_rasterizer.cpp b/src/video_core/renderer_opengl/gl_rasterizer.cpp
index bb09ecd52..01d89f47d 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer.cpp
+++ b/src/video_core/renderer_opengl/gl_rasterizer.cpp
@@ -537,8 +537,7 @@ std::pair<bool, bool> RasterizerOpenGL::ConfigureFramebuffers(
texture_cache.MarkDepthBufferInUse();
fbkey.zeta = depth_surface;
- fbkey.stencil_enable = regs.stencil_enable &&
- depth_surface->GetSurfaceParams().type == SurfaceType::DepthStencil;
+ fbkey.stencil_enable = depth_surface->GetSurfaceParams().type == SurfaceType::DepthStencil;
}
texture_cache.GuardRenderTargets(false);
@@ -577,16 +576,15 @@ void RasterizerOpenGL::ConfigureClearFramebuffer(OpenGLState& current_state, boo
if (depth_surface) {
const auto& params = depth_surface->GetSurfaceParams();
switch (params.type) {
- case VideoCore::Surface::SurfaceType::Depth: {
+ case VideoCore::Surface::SurfaceType::Depth:
depth_surface->Attach(GL_DEPTH_ATTACHMENT, GL_DRAW_FRAMEBUFFER);
glFramebufferTexture2D(GL_DRAW_FRAMEBUFFER, GL_STENCIL_ATTACHMENT, GL_TEXTURE_2D, 0, 0);
break;
- }
- case VideoCore::Surface::SurfaceType::DepthStencil: {
- depth_surface->Attach(GL_DEPTH_ATTACHMENT, GL_DRAW_FRAMEBUFFER);
+ case VideoCore::Surface::SurfaceType::DepthStencil:
+ depth_surface->Attach(GL_DEPTH_STENCIL_ATTACHMENT, GL_DRAW_FRAMEBUFFER);
break;
- }
- default: { UNIMPLEMENTED(); }
+ default:
+ UNIMPLEMENTED();
}
} else {
glFramebufferTexture2D(GL_DRAW_FRAMEBUFFER, GL_DEPTH_STENCIL_ATTACHMENT, GL_TEXTURE_2D, 0,
@@ -639,6 +637,7 @@ void RasterizerOpenGL::Clear() {
ASSERT_MSG(regs.zeta_enable != 0, "Tried to clear stencil but buffer is not enabled!");
use_stencil = true;
clear_state.stencil.test_enabled = true;
+
if (regs.clear_flags.stencil) {
// Stencil affects the clear so fill it with the used masks
clear_state.stencil.front.test_func = GL_ALWAYS;
@@ -1119,9 +1118,12 @@ void RasterizerOpenGL::SyncStencilTestState() {
if (!maxwell3d.dirty.stencil_test) {
return;
}
- const auto& regs = maxwell3d.regs;
+ maxwell3d.dirty.stencil_test = false;
+ const auto& regs = maxwell3d.regs;
state.stencil.test_enabled = regs.stencil_enable != 0;
+ state.MarkDirtyStencilState();
+
if (!regs.stencil_enable) {
return;
}
@@ -1150,8 +1152,6 @@ void RasterizerOpenGL::SyncStencilTestState() {
state.stencil.back.action_depth_fail = GL_KEEP;
state.stencil.back.action_depth_pass = GL_KEEP;
}
- state.MarkDirtyStencilState();
- maxwell3d.dirty.stencil_test = false;
}
void RasterizerOpenGL::SyncColorMask() {
diff --git a/src/video_core/renderer_opengl/gl_shader_cache.cpp b/src/video_core/renderer_opengl/gl_shader_cache.cpp
index cf6a5cddf..909ccb82c 100644
--- a/src/video_core/renderer_opengl/gl_shader_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_cache.cpp
@@ -348,23 +348,16 @@ Shader CachedShader::CreateKernelFromCache(const ShaderParameters& params,
}
std::tuple<GLuint, BaseBindings> CachedShader::GetProgramHandle(const ProgramVariant& variant) {
- GLuint handle{};
- if (program_type == ProgramType::Geometry) {
- handle = GetGeometryShader(variant);
- } else {
- const auto [entry, is_cache_miss] = programs.try_emplace(variant);
- auto& program = entry->second;
- if (is_cache_miss) {
- program = TryLoadProgram(variant);
- if (!program) {
- program = SpecializeShader(code, entries, program_type, variant);
- disk_cache.SaveUsage(GetUsage(variant));
- }
-
- LabelGLObject(GL_PROGRAM, program->handle, cpu_addr);
+ const auto [entry, is_cache_miss] = programs.try_emplace(variant);
+ auto& program = entry->second;
+ if (is_cache_miss) {
+ program = TryLoadProgram(variant);
+ if (!program) {
+ program = SpecializeShader(code, entries, program_type, variant);
+ disk_cache.SaveUsage(GetUsage(variant));
}
- handle = program->handle;
+ LabelGLObject(GL_PROGRAM, program->handle, cpu_addr);
}
auto base_bindings = variant.base_bindings;
@@ -375,52 +368,9 @@ std::tuple<GLuint, BaseBindings> CachedShader::GetProgramHandle(const ProgramVar
base_bindings.gmem += static_cast<u32>(entries.global_memory_entries.size());
base_bindings.sampler += static_cast<u32>(entries.samplers.size());
- return {handle, base_bindings};
+ return {program->handle, base_bindings};
}
-GLuint CachedShader::GetGeometryShader(const ProgramVariant& variant) {
- const auto [entry, is_cache_miss] = geometry_programs.try_emplace(variant);
- auto& programs = entry->second;
-
- switch (variant.primitive_mode) {
- case GL_POINTS:
- return LazyGeometryProgram(programs.points, variant);
- case GL_LINES:
- case GL_LINE_STRIP:
- return LazyGeometryProgram(programs.lines, variant);
- case GL_LINES_ADJACENCY:
- case GL_LINE_STRIP_ADJACENCY:
- return LazyGeometryProgram(programs.lines_adjacency, variant);
- case GL_TRIANGLES:
- case GL_TRIANGLE_STRIP:
- case GL_TRIANGLE_FAN:
- return LazyGeometryProgram(programs.triangles, variant);
- case GL_TRIANGLES_ADJACENCY:
- case GL_TRIANGLE_STRIP_ADJACENCY:
- return LazyGeometryProgram(programs.triangles_adjacency, variant);
- default:
- UNREACHABLE_MSG("Unknown primitive mode.");
- return LazyGeometryProgram(programs.points, variant);
- }
-}
-
-GLuint CachedShader::LazyGeometryProgram(CachedProgram& target_program,
- const ProgramVariant& variant) {
- if (target_program) {
- return target_program->handle;
- }
- const auto [glsl_name, debug_name, vertices] = GetPrimitiveDescription(variant.primitive_mode);
- target_program = TryLoadProgram(variant);
- if (!target_program) {
- target_program = SpecializeShader(code, entries, program_type, variant);
- disk_cache.SaveUsage(GetUsage(variant));
- }
-
- LabelGLObject(GL_PROGRAM, target_program->handle, cpu_addr, debug_name);
-
- return target_program->handle;
-};
-
CachedProgram CachedShader::TryLoadProgram(const ProgramVariant& variant) const {
const auto found = precompiled_programs.find(GetUsage(variant));
if (found == precompiled_programs.end()) {
diff --git a/src/video_core/renderer_opengl/gl_shader_cache.h b/src/video_core/renderer_opengl/gl_shader_cache.h
index 2c8faf855..de195cc5d 100644
--- a/src/video_core/renderer_opengl/gl_shader_cache.h
+++ b/src/video_core/renderer_opengl/gl_shader_cache.h
@@ -86,22 +86,6 @@ private:
explicit CachedShader(const ShaderParameters& params, ProgramType program_type,
GLShader::ProgramResult result);
- // Geometry programs. These are needed because GLSL needs an input topology but it's not
- // declared by the hardware. Workaround this issue by generating a different shader per input
- // topology class.
- struct GeometryPrograms {
- CachedProgram points;
- CachedProgram lines;
- CachedProgram lines_adjacency;
- CachedProgram triangles;
- CachedProgram triangles_adjacency;
- };
-
- GLuint GetGeometryShader(const ProgramVariant& variant);
-
- /// Generates a geometry shader or returns one that already exists.
- GLuint LazyGeometryProgram(CachedProgram& target_program, const ProgramVariant& variant);
-
CachedProgram TryLoadProgram(const ProgramVariant& variant) const;
ShaderDiskCacheUsage GetUsage(const ProgramVariant& variant) const;
@@ -117,11 +101,6 @@ private:
std::size_t shader_length{};
std::unordered_map<ProgramVariant, CachedProgram> programs;
- std::unordered_map<ProgramVariant, GeometryPrograms> geometry_programs;
-
- std::unordered_map<u32, GLuint> cbuf_resource_cache;
- std::unordered_map<u32, GLuint> gmem_resource_cache;
- std::unordered_map<u32, GLint> uniform_cache;
};
class ShaderCacheOpenGL final : public RasterizerCache<Shader> {
diff --git a/src/video_core/shader/decode/half_set_predicate.cpp b/src/video_core/shader/decode/half_set_predicate.cpp
index afea33e5f..840694527 100644
--- a/src/video_core/shader/decode/half_set_predicate.cpp
+++ b/src/video_core/shader/decode/half_set_predicate.cpp
@@ -42,9 +42,8 @@ u32 ShaderIR::DecodeHalfSetPredicate(NodeBlock& bb, u32 pc) {
cond = instr.hsetp2.reg.cond;
h_and = instr.hsetp2.reg.h_and;
op_b =
- UnpackHalfFloat(GetOperandAbsNegHalf(GetRegister(instr.gpr20), instr.hsetp2.reg.abs_b,
- instr.hsetp2.reg.negate_b),
- instr.hsetp2.reg.type_b);
+ GetOperandAbsNegHalf(UnpackHalfFloat(GetRegister(instr.gpr20), instr.hsetp2.reg.type_b),
+ instr.hsetp2.reg.abs_b, instr.hsetp2.reg.negate_b);
break;
default:
UNREACHABLE();
@@ -52,22 +51,22 @@ u32 ShaderIR::DecodeHalfSetPredicate(NodeBlock& bb, u32 pc) {
}
const OperationCode combiner = GetPredicateCombiner(instr.hsetp2.op);
- const Node combined_pred = GetPredicate(instr.hsetp2.pred3, instr.hsetp2.neg_pred);
+ const Node combined_pred = GetPredicate(instr.hsetp2.pred39, instr.hsetp2.neg_pred);
const auto Write = [&](u64 dest, Node src) {
SetPredicate(bb, dest, Operation(combiner, std::move(src), combined_pred));
};
const Node comparison = GetPredicateComparisonHalf(cond, op_a, op_b);
- const u64 first = instr.hsetp2.pred0;
- const u64 second = instr.hsetp2.pred39;
+ const u64 first = instr.hsetp2.pred3;
+ const u64 second = instr.hsetp2.pred0;
if (h_and) {
- const Node joined = Operation(OperationCode::LogicalAnd2, comparison);
+ Node joined = Operation(OperationCode::LogicalAnd2, comparison);
Write(first, joined);
- Write(second, Operation(OperationCode::LogicalNegate, joined));
+ Write(second, Operation(OperationCode::LogicalNegate, std::move(joined)));
} else {
- Write(first, Operation(OperationCode::LogicalPick2, comparison, Immediate(0u)));
- Write(second, Operation(OperationCode::LogicalPick2, comparison, Immediate(1u)));
+ Write(first, Operation(OperationCode::LogicalPick2, comparison, Immediate(0U)));
+ Write(second, Operation(OperationCode::LogicalPick2, comparison, Immediate(1U)));
}
return pc;
generated by cgit v1.2.3 (git 2.25.1) at 2026-01-01 04:27:51 +0000