summaryrefslogtreecommitdiffstats
path: root/src/video_core
diff options
context:
space:
mode:
Diffstat (limited to 'src/video_core')
-rw-r--r--src/video_core/buffer_cache/buffer_cache.h18
-rw-r--r--src/video_core/renderer_opengl/gl_buffer_cache.cpp31
-rw-r--r--src/video_core/renderer_opengl/gl_buffer_cache.h20
-rw-r--r--src/video_core/renderer_opengl/gl_device.cpp4
-rw-r--r--src/video_core/renderer_opengl/gl_device.h5
-rw-r--r--src/video_core/renderer_opengl/gl_rasterizer.cpp7
-rw-r--r--src/video_core/shader/decode/arithmetic.cpp11
-rw-r--r--src/video_core/shader/decode/arithmetic_half.cpp4
-rw-r--r--src/video_core/shader/decode/arithmetic_half_immediate.cpp8
-rw-r--r--src/video_core/shader/decode/ffma.cpp4
-rw-r--r--src/video_core/shader/decode/half_set.cpp4
-rw-r--r--src/video_core/shader/decode/half_set_predicate.cpp4
-rw-r--r--src/video_core/shader/decode/texture.cpp35
13 files changed, 93 insertions, 62 deletions
diff --git a/src/video_core/buffer_cache/buffer_cache.h b/src/video_core/buffer_cache/buffer_cache.h
index 2442ddfd6..4408b5001 100644
--- a/src/video_core/buffer_cache/buffer_cache.h
+++ b/src/video_core/buffer_cache/buffer_cache.h
@@ -12,6 +12,10 @@
#include <utility>
#include <vector>
+#include <boost/icl/interval_map.hpp>
+#include <boost/icl/interval_set.hpp>
+#include <boost/range/iterator_range.hpp>
+
#include "common/alignment.h"
#include "common/common_types.h"
#include "core/core.h"
@@ -30,7 +34,7 @@ public:
using BufferInfo = std::pair<const TBufferType*, u64>;
BufferInfo UploadMemory(GPUVAddr gpu_addr, std::size_t size, std::size_t alignment = 4,
- bool is_written = false) {
+ bool is_written = false, bool use_fast_cbuf = false) {
std::lock_guard lock{mutex};
auto& memory_manager = system.GPU().MemoryManager();
@@ -43,9 +47,13 @@ public:
// Cache management is a big overhead, so only cache entries with a given size.
// TODO: Figure out which size is the best for given games.
constexpr std::size_t max_stream_size = 0x800;
- if (size < max_stream_size) {
+ if (use_fast_cbuf || size < max_stream_size) {
if (!is_written && !IsRegionWritten(cache_addr, cache_addr + size - 1)) {
- return StreamBufferUpload(host_ptr, size, alignment);
+ if (use_fast_cbuf) {
+ return ConstBufferUpload(host_ptr, size);
+ } else {
+ return StreamBufferUpload(host_ptr, size, alignment);
+ }
}
}
@@ -152,6 +160,10 @@ protected:
virtual void CopyBlock(const TBuffer& src, const TBuffer& dst, std::size_t src_offset,
std::size_t dst_offset, std::size_t size) = 0;
+ virtual BufferInfo ConstBufferUpload(const void* raw_pointer, std::size_t size) {
+ return {};
+ }
+
/// Register an object into the cache
void Register(const MapInterval& new_map, bool inherit_written = false) {
const CacheAddr cache_ptr = new_map->GetStart();
diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.cpp b/src/video_core/renderer_opengl/gl_buffer_cache.cpp
index f8a807c84..0375fca17 100644
--- a/src/video_core/renderer_opengl/gl_buffer_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_buffer_cache.cpp
@@ -8,13 +8,17 @@
#include "common/assert.h"
#include "common/microprofile.h"
+#include "video_core/engines/maxwell_3d.h"
#include "video_core/rasterizer_interface.h"
#include "video_core/renderer_opengl/gl_buffer_cache.h"
+#include "video_core/renderer_opengl/gl_device.h"
#include "video_core/renderer_opengl/gl_rasterizer.h"
#include "video_core/renderer_opengl/gl_resource_manager.h"
namespace OpenGL {
+using Maxwell = Tegra::Engines::Maxwell3D::Regs;
+
MICROPROFILE_DEFINE(OpenGL_Buffer_Download, "OpenGL", "Buffer Download", MP_RGB(192, 192, 128));
CachedBufferBlock::CachedBufferBlock(CacheAddr cache_addr, const std::size_t size)
@@ -26,11 +30,22 @@ CachedBufferBlock::CachedBufferBlock(CacheAddr cache_addr, const std::size_t siz
CachedBufferBlock::~CachedBufferBlock() = default;
OGLBufferCache::OGLBufferCache(RasterizerOpenGL& rasterizer, Core::System& system,
- std::size_t stream_size)
- : VideoCommon::BufferCache<Buffer, GLuint, OGLStreamBuffer>{
- rasterizer, system, std::make_unique<OGLStreamBuffer>(stream_size, true)} {}
+ const Device& device, std::size_t stream_size)
+ : GenericBufferCache{rasterizer, system, std::make_unique<OGLStreamBuffer>(stream_size, true)} {
+ if (!device.HasFastBufferSubData()) {
+ return;
+ }
+
+ static constexpr auto size = static_cast<GLsizeiptr>(Maxwell::MaxConstBufferSize);
+ glCreateBuffers(static_cast<GLsizei>(std::size(cbufs)), std::data(cbufs));
+ for (const GLuint cbuf : cbufs) {
+ glNamedBufferData(cbuf, size, nullptr, GL_STREAM_DRAW);
+ }
+}
-OGLBufferCache::~OGLBufferCache() = default;
+OGLBufferCache::~OGLBufferCache() {
+ glDeleteBuffers(static_cast<GLsizei>(std::size(cbufs)), std::data(cbufs));
+}
Buffer OGLBufferCache::CreateBlock(CacheAddr cache_addr, std::size_t size) {
return std::make_shared<CachedBufferBlock>(cache_addr, size);
@@ -69,4 +84,12 @@ void OGLBufferCache::CopyBlock(const Buffer& src, const Buffer& dst, std::size_t
static_cast<GLsizeiptr>(size));
}
+OGLBufferCache::BufferInfo OGLBufferCache::ConstBufferUpload(const void* raw_pointer,
+ std::size_t size) {
+ DEBUG_ASSERT(cbuf_cursor < std::size(cbufs));
+ const GLuint& cbuf = cbufs[cbuf_cursor++];
+ glNamedBufferSubData(cbuf, 0, static_cast<GLsizeiptr>(size), raw_pointer);
+ return {&cbuf, 0};
+}
+
} // namespace OpenGL
diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.h b/src/video_core/renderer_opengl/gl_buffer_cache.h
index 022e7bfa9..8c7145443 100644
--- a/src/video_core/renderer_opengl/gl_buffer_cache.h
+++ b/src/video_core/renderer_opengl/gl_buffer_cache.h
@@ -4,10 +4,12 @@
#pragma once
+#include <array>
#include <memory>
#include "common/common_types.h"
#include "video_core/buffer_cache/buffer_cache.h"
+#include "video_core/engines/maxwell_3d.h"
#include "video_core/rasterizer_cache.h"
#include "video_core/renderer_opengl/gl_resource_manager.h"
#include "video_core/renderer_opengl/gl_stream_buffer.h"
@@ -18,12 +20,14 @@ class System;
namespace OpenGL {
+class Device;
class OGLStreamBuffer;
class RasterizerOpenGL;
class CachedBufferBlock;
using Buffer = std::shared_ptr<CachedBufferBlock>;
+using GenericBufferCache = VideoCommon::BufferCache<Buffer, GLuint, OGLStreamBuffer>;
class CachedBufferBlock : public VideoCommon::BufferBlock {
public:
@@ -38,14 +42,18 @@ private:
OGLBuffer gl_buffer{};
};
-class OGLBufferCache final : public VideoCommon::BufferCache<Buffer, GLuint, OGLStreamBuffer> {
+class OGLBufferCache final : public GenericBufferCache {
public:
explicit OGLBufferCache(RasterizerOpenGL& rasterizer, Core::System& system,
- std::size_t stream_size);
+ const Device& device, std::size_t stream_size);
~OGLBufferCache();
const GLuint* GetEmptyBuffer(std::size_t) override;
+ void Acquire() noexcept {
+ cbuf_cursor = 0;
+ }
+
protected:
Buffer CreateBlock(CacheAddr cache_addr, std::size_t size) override;
@@ -61,6 +69,14 @@ protected:
void CopyBlock(const Buffer& src, const Buffer& dst, std::size_t src_offset,
std::size_t dst_offset, std::size_t size) override;
+
+ BufferInfo ConstBufferUpload(const void* raw_pointer, std::size_t size) override;
+
+private:
+ std::size_t cbuf_cursor = 0;
+ std::array<GLuint, Tegra::Engines::Maxwell3D::Regs::MaxConstBuffers *
+ Tegra::Engines::Maxwell3D::Regs::MaxShaderProgram>
+ cbufs;
};
} // namespace OpenGL
diff --git a/src/video_core/renderer_opengl/gl_device.cpp b/src/video_core/renderer_opengl/gl_device.cpp
index 64de7e425..c65b24c69 100644
--- a/src/video_core/renderer_opengl/gl_device.cpp
+++ b/src/video_core/renderer_opengl/gl_device.cpp
@@ -51,8 +51,11 @@ bool HasExtension(const std::vector<std::string_view>& images, std::string_view
} // Anonymous namespace
Device::Device() {
+ const std::string_view vendor = reinterpret_cast<const char*>(glGetString(GL_VENDOR));
const std::vector extensions = GetExtensions();
+ const bool is_nvidia = vendor == "NVIDIA Corporation";
+
uniform_buffer_alignment = GetInteger<std::size_t>(GL_UNIFORM_BUFFER_OFFSET_ALIGNMENT);
shader_storage_alignment = GetInteger<std::size_t>(GL_SHADER_STORAGE_BUFFER_OFFSET_ALIGNMENT);
max_vertex_attributes = GetInteger<u32>(GL_MAX_VERTEX_ATTRIBS);
@@ -64,6 +67,7 @@ Device::Device() {
has_variable_aoffi = TestVariableAoffi();
has_component_indexing_bug = TestComponentIndexingBug();
has_precise_bug = TestPreciseBug();
+ has_fast_buffer_sub_data = is_nvidia;
LOG_INFO(Render_OpenGL, "Renderer_VariableAOFFI: {}", has_variable_aoffi);
LOG_INFO(Render_OpenGL, "Renderer_ComponentIndexingBug: {}", has_component_indexing_bug);
diff --git a/src/video_core/renderer_opengl/gl_device.h b/src/video_core/renderer_opengl/gl_device.h
index bb273c3d6..bf35bd0b6 100644
--- a/src/video_core/renderer_opengl/gl_device.h
+++ b/src/video_core/renderer_opengl/gl_device.h
@@ -54,6 +54,10 @@ public:
return has_precise_bug;
}
+ bool HasFastBufferSubData() const {
+ return has_fast_buffer_sub_data;
+ }
+
private:
static bool TestVariableAoffi();
static bool TestComponentIndexingBug();
@@ -69,6 +73,7 @@ private:
bool has_variable_aoffi{};
bool has_component_indexing_bug{};
bool has_precise_bug{};
+ bool has_fast_buffer_sub_data{};
};
} // namespace OpenGL
diff --git a/src/video_core/renderer_opengl/gl_rasterizer.cpp b/src/video_core/renderer_opengl/gl_rasterizer.cpp
index d1e147db8..e560d70d5 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer.cpp
+++ b/src/video_core/renderer_opengl/gl_rasterizer.cpp
@@ -67,7 +67,7 @@ static std::size_t GetConstBufferSize(const Tegra::Engines::ConstBufferInfo& buf
RasterizerOpenGL::RasterizerOpenGL(Core::System& system, Core::Frontend::EmuWindow& emu_window,
ScreenInfo& info)
: texture_cache{system, *this, device}, shader_cache{*this, system, emu_window, device},
- system{system}, screen_info{info}, buffer_cache{*this, system, STREAM_BUFFER_SIZE} {
+ system{system}, screen_info{info}, buffer_cache{*this, system, device, STREAM_BUFFER_SIZE} {
shader_program_manager = std::make_unique<GLShader::ProgramManager>();
state.draw.shader_program = 0;
state.Apply();
@@ -558,6 +558,8 @@ void RasterizerOpenGL::DrawPrelude() {
SyncPolygonOffset();
SyncAlphaTest();
+ buffer_cache.Acquire();
+
// Draw the vertex batch
const bool is_indexed = accelerate_draw == AccelDraw::Indexed;
@@ -879,7 +881,8 @@ void RasterizerOpenGL::SetupConstBuffer(const Tegra::Engines::ConstBufferInfo& b
const std::size_t size = Common::AlignUp(GetConstBufferSize(buffer, entry), sizeof(GLvec4));
const auto alignment = device.GetUniformBufferAlignment();
- const auto [cbuf, offset] = buffer_cache.UploadMemory(buffer.address, size, alignment);
+ const auto [cbuf, offset] = buffer_cache.UploadMemory(buffer.address, size, alignment, false,
+ device.HasFastBufferSubData());
bind_ubo_pushbuffer.Push(cbuf, offset, size);
}
diff --git a/src/video_core/shader/decode/arithmetic.cpp b/src/video_core/shader/decode/arithmetic.cpp
index 1473c282a..fcedd2af6 100644
--- a/src/video_core/shader/decode/arithmetic.cpp
+++ b/src/video_core/shader/decode/arithmetic.cpp
@@ -43,12 +43,12 @@ u32 ShaderIR::DecodeArithmetic(NodeBlock& bb, u32 pc) {
case OpCode::Id::FMUL_IMM: {
// FMUL does not have 'abs' bits and only the second operand has a 'neg' bit.
if (instr.fmul.tab5cb8_2 != 0) {
- LOG_WARNING(HW_GPU, "FMUL tab5cb8_2({}) is not implemented",
- instr.fmul.tab5cb8_2.Value());
+ LOG_DEBUG(HW_GPU, "FMUL tab5cb8_2({}) is not implemented",
+ instr.fmul.tab5cb8_2.Value());
}
if (instr.fmul.tab5c68_0 != 1) {
- LOG_WARNING(HW_GPU, "FMUL tab5cb8_0({}) is not implemented",
- instr.fmul.tab5c68_0.Value());
+ LOG_DEBUG(HW_GPU, "FMUL tab5cb8_0({}) is not implemented",
+ instr.fmul.tab5c68_0.Value());
}
op_b = GetOperandAbsNegFloat(op_b, false, instr.fmul.negate_b);
@@ -144,10 +144,11 @@ u32 ShaderIR::DecodeArithmetic(NodeBlock& bb, u32 pc) {
case OpCode::Id::RRO_C:
case OpCode::Id::RRO_R:
case OpCode::Id::RRO_IMM: {
+ LOG_DEBUG(HW_GPU, "(STUBBED) RRO used");
+
// Currently RRO is only implemented as a register move.
op_b = GetOperandAbsNegFloat(op_b, instr.alu.abs_b, instr.alu.negate_b);
SetRegister(bb, instr.gpr0, op_b);
- LOG_WARNING(HW_GPU, "RRO instruction is incomplete");
break;
}
default:
diff --git a/src/video_core/shader/decode/arithmetic_half.cpp b/src/video_core/shader/decode/arithmetic_half.cpp
index b06cbe441..ee7d9a29d 100644
--- a/src/video_core/shader/decode/arithmetic_half.cpp
+++ b/src/video_core/shader/decode/arithmetic_half.cpp
@@ -21,8 +21,8 @@ u32 ShaderIR::DecodeArithmeticHalf(NodeBlock& bb, u32 pc) {
if (opcode->get().GetId() == OpCode::Id::HADD2_C ||
opcode->get().GetId() == OpCode::Id::HADD2_R) {
- if (instr.alu_half.ftz != 0) {
- LOG_WARNING(HW_GPU, "{} FTZ not implemented", opcode->get().GetName());
+ if (instr.alu_half.ftz == 0) {
+ LOG_DEBUG(HW_GPU, "{} without FTZ is not implemented", opcode->get().GetName());
}
}
diff --git a/src/video_core/shader/decode/arithmetic_half_immediate.cpp b/src/video_core/shader/decode/arithmetic_half_immediate.cpp
index 6466fc011..d179b9873 100644
--- a/src/video_core/shader/decode/arithmetic_half_immediate.cpp
+++ b/src/video_core/shader/decode/arithmetic_half_immediate.cpp
@@ -19,12 +19,12 @@ u32 ShaderIR::DecodeArithmeticHalfImmediate(NodeBlock& bb, u32 pc) {
const auto opcode = OpCode::Decode(instr);
if (opcode->get().GetId() == OpCode::Id::HADD2_IMM) {
- if (instr.alu_half_imm.ftz != 0) {
- LOG_WARNING(HW_GPU, "{} FTZ not implemented", opcode->get().GetName());
+ if (instr.alu_half_imm.ftz == 0) {
+ LOG_DEBUG(HW_GPU, "{} without FTZ is not implemented", opcode->get().GetName());
}
} else {
- if (instr.alu_half_imm.precision != Tegra::Shader::HalfPrecision::None) {
- LOG_WARNING(HW_GPU, "{} FTZ not implemented", opcode->get().GetName());
+ if (instr.alu_half_imm.precision != Tegra::Shader::HalfPrecision::FTZ) {
+ LOG_DEBUG(HW_GPU, "{} without FTZ is not implemented", opcode->get().GetName());
}
}
diff --git a/src/video_core/shader/decode/ffma.cpp b/src/video_core/shader/decode/ffma.cpp
index ca2f39e8d..5973588d6 100644
--- a/src/video_core/shader/decode/ffma.cpp
+++ b/src/video_core/shader/decode/ffma.cpp
@@ -19,10 +19,10 @@ u32 ShaderIR::DecodeFfma(NodeBlock& bb, u32 pc) {
UNIMPLEMENTED_IF_MSG(instr.ffma.cc != 0, "FFMA cc not implemented");
if (instr.ffma.tab5980_0 != 1) {
- LOG_WARNING(HW_GPU, "FFMA tab5980_0({}) not implemented", instr.ffma.tab5980_0.Value());
+ LOG_DEBUG(HW_GPU, "FFMA tab5980_0({}) not implemented", instr.ffma.tab5980_0.Value());
}
if (instr.ffma.tab5980_1 != 0) {
- LOG_WARNING(HW_GPU, "FFMA tab5980_1({}) not implemented", instr.ffma.tab5980_1.Value());
+ LOG_DEBUG(HW_GPU, "FFMA tab5980_1({}) not implemented", instr.ffma.tab5980_1.Value());
}
const Node op_a = GetRegister(instr.gpr8);
diff --git a/src/video_core/shader/decode/half_set.cpp b/src/video_core/shader/decode/half_set.cpp
index 48ca7a4af..848e46874 100644
--- a/src/video_core/shader/decode/half_set.cpp
+++ b/src/video_core/shader/decode/half_set.cpp
@@ -20,8 +20,8 @@ u32 ShaderIR::DecodeHalfSet(NodeBlock& bb, u32 pc) {
const Instruction instr = {program_code[pc]};
const auto opcode = OpCode::Decode(instr);
- if (instr.hset2.ftz != 0) {
- LOG_WARNING(HW_GPU, "{} FTZ not implemented", opcode->get().GetName());
+ if (instr.hset2.ftz == 0) {
+ LOG_DEBUG(HW_GPU, "{} without FTZ is not implemented", opcode->get().GetName());
}
Node op_a = UnpackHalfFloat(GetRegister(instr.gpr8), instr.hset2.type_a);
diff --git a/src/video_core/shader/decode/half_set_predicate.cpp b/src/video_core/shader/decode/half_set_predicate.cpp
index fec8f2dbe..310655619 100644
--- a/src/video_core/shader/decode/half_set_predicate.cpp
+++ b/src/video_core/shader/decode/half_set_predicate.cpp
@@ -19,7 +19,9 @@ u32 ShaderIR::DecodeHalfSetPredicate(NodeBlock& bb, u32 pc) {
const Instruction instr = {program_code[pc]};
const auto opcode = OpCode::Decode(instr);
- LOG_DEBUG(HW_GPU, "ftz={}", static_cast<u32>(instr.hsetp2.ftz));
+ if (instr.hsetp2.ftz != 0) {
+ LOG_DEBUG(HW_GPU, "{} without FTZ is not implemented", opcode->get().GetName());
+ }
Node op_a = UnpackHalfFloat(GetRegister(instr.gpr8), instr.hsetp2.type_a);
op_a = GetOperandAbsNegHalf(op_a, instr.hsetp2.abs_a, instr.hsetp2.negate_a);
diff --git a/src/video_core/shader/decode/texture.cpp b/src/video_core/shader/decode/texture.cpp
index ca690b58b..bb926a132 100644
--- a/src/video_core/shader/decode/texture.cpp
+++ b/src/video_core/shader/decode/texture.cpp
@@ -44,10 +44,6 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) {
bool is_bindless = false;
switch (opcode->get().GetId()) {
case OpCode::Id::TEX: {
- if (instr.tex.UsesMiscMode(TextureMiscMode::NODEP)) {
- LOG_WARNING(HW_GPU, "TEX.NODEP implementation is incomplete");
- }
-
const TextureType texture_type{instr.tex.texture_type};
const bool is_array = instr.tex.array != 0;
const bool is_aoffi = instr.tex.UsesMiscMode(TextureMiscMode::AOFFI);
@@ -62,10 +58,6 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) {
UNIMPLEMENTED_IF_MSG(instr.tex.UsesMiscMode(TextureMiscMode::AOFFI),
"AOFFI is not implemented");
- if (instr.tex.UsesMiscMode(TextureMiscMode::NODEP)) {
- LOG_WARNING(HW_GPU, "TEX.NODEP implementation is incomplete");
- }
-
const TextureType texture_type{instr.tex_b.texture_type};
const bool is_array = instr.tex_b.array != 0;
const bool is_aoffi = instr.tex.UsesMiscMode(TextureMiscMode::AOFFI);
@@ -82,10 +74,6 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) {
const bool depth_compare = instr.texs.UsesMiscMode(TextureMiscMode::DC);
const auto process_mode = instr.texs.GetTextureProcessMode();
- if (instr.texs.UsesMiscMode(TextureMiscMode::NODEP)) {
- LOG_WARNING(HW_GPU, "TEXS.NODEP implementation is incomplete");
- }
-
const Node4 components =
GetTexsCode(instr, texture_type, process_mode, depth_compare, is_array);
@@ -107,10 +95,6 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) {
UNIMPLEMENTED_IF_MSG(instr.tld4.UsesMiscMode(TextureMiscMode::PTP),
"PTP is not implemented");
- if (instr.tld4.UsesMiscMode(TextureMiscMode::NODEP)) {
- LOG_WARNING(HW_GPU, "TLD4.NODEP implementation is incomplete");
- }
-
const auto texture_type = instr.tld4.texture_type.Value();
const bool depth_compare = is_bindless ? instr.tld4_b.UsesMiscMode(TextureMiscMode::DC)
: instr.tld4.UsesMiscMode(TextureMiscMode::DC);
@@ -125,9 +109,6 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) {
case OpCode::Id::TLD4S: {
UNIMPLEMENTED_IF_MSG(instr.tld4s.UsesMiscMode(TextureMiscMode::AOFFI),
"AOFFI is not implemented");
- if (instr.tld4s.UsesMiscMode(TextureMiscMode::NODEP)) {
- LOG_WARNING(HW_GPU, "TLD4S.NODEP implementation is incomplete");
- }
const bool depth_compare = instr.tld4s.UsesMiscMode(TextureMiscMode::DC);
const Node op_a = GetRegister(instr.gpr8);
@@ -164,10 +145,6 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) {
is_bindless = true;
[[fallthrough]];
case OpCode::Id::TXQ: {
- if (instr.txq.UsesMiscMode(TextureMiscMode::NODEP)) {
- LOG_WARNING(HW_GPU, "TXQ.NODEP implementation is incomplete");
- }
-
// TODO: The new commits on the texture refactor, change the way samplers work.
// Sadly, not all texture instructions specify the type of texture their sampler
// uses. This must be fixed at a later instance.
@@ -205,10 +182,6 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) {
UNIMPLEMENTED_IF_MSG(instr.tmml.UsesMiscMode(Tegra::Shader::TextureMiscMode::NDV),
"NDV is not implemented");
- if (instr.tmml.UsesMiscMode(TextureMiscMode::NODEP)) {
- LOG_WARNING(HW_GPU, "TMML.NODEP implementation is incomplete");
- }
-
auto texture_type = instr.tmml.texture_type.Value();
const bool is_array = instr.tmml.array != 0;
const auto& sampler =
@@ -254,10 +227,6 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) {
UNIMPLEMENTED_IF_MSG(instr.tld.ms, "MS is not implemented");
UNIMPLEMENTED_IF_MSG(instr.tld.cl, "CL is not implemented");
- if (instr.tld.nodep_flag) {
- LOG_WARNING(HW_GPU, "TLD.NODEP implementation is incomplete");
- }
-
WriteTexInstructionFloat(bb, instr, GetTldCode(instr));
break;
}
@@ -269,10 +238,6 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) {
"AOFFI is not implemented");
UNIMPLEMENTED_IF_MSG(instr.tlds.UsesMiscMode(TextureMiscMode::MZ), "MZ is not implemented");
- if (instr.tlds.UsesMiscMode(TextureMiscMode::NODEP)) {
- LOG_WARNING(HW_GPU, "TLDS.NODEP implementation is incomplete");
- }
-
const Node4 components = GetTldsCode(instr, texture_type, is_array);
if (instr.tlds.fp32_flag) {
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-25 06:58:31 +0100