Diamond75/Panda3DS
1
0
Fork 0
mirror of https://github.com/wheremyfoodat/Panda3DS.git synced 2025-06-03 12:27:21 +12:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge branch 'master' into io-file-cpp

Browse source
This commit is contained in:
wheremyfoodat 2023-06-28 02:53:26 +03:00
commit e09eaee6e7
9 changed files with 320 additions and 292 deletions
Download patch file Download diff file Expand all files Collapse all files

2
include/PICA/gpu.hpp
View file

@ -47,7 +47,7 @@ class GPU {
};
u64 getVertexShaderInputConfig() {
return u64(regs[PICAInternalRegs::VertexShaderInputCfgLow]) | (u64(regs[PICAInternalRegs::VertexShaderInputCfgHigh]) << 32);
return u64(regs[PICA::InternalRegs::VertexShaderInputCfgLow]) | (u64(regs[PICA::InternalRegs::VertexShaderInputCfgHigh]) << 32);
}
std::array<AttribInfo, maxAttribCount> attributeInfo; // Info for each of the 12 attributes

335
include/PICA/regs.hpp
View file

@ -1,130 +1,241 @@
#pragma once
#include "helpers.hpp"
namespace PICAInternalRegs {
enum : u32 {
// Rasterizer registers
ViewportWidth = 0x41,
ViewportInvw = 0x42,
ViewportHeight = 0x43,
ViewportInvh = 0x44,
namespace PICA {
namespace InternalRegs {
enum : u32 {
// Rasterizer registers
ViewportWidth = 0x41,
ViewportInvw = 0x42,
ViewportHeight = 0x43,
ViewportInvh = 0x44,
DepthScale = 0x4D,
DepthOffset = 0x4E,
ShaderOutputCount = 0x4F,
DepthScale = 0x4D,
DepthOffset = 0x4E,
ShaderOutputCount = 0x4F,
DepthmapEnable = 0x6D,
TexUnitCfg = 0x80,
DepthmapEnable = 0x6D,
TexUnitCfg = 0x80,
// Framebuffer registers
ColourOperation = 0x100,
BlendFunc = 0x101,
BlendColour = 0x103,
AlphaTestConfig = 0x104,
DepthAndColorMask = 0x107,
DepthBufferFormat = 0x116,
ColourBufferFormat = 0x117,
DepthBufferLoc = 0x11C,
ColourBufferLoc = 0x11D,
FramebufferSize = 0x11E,
// clang-format off
#define defineTexEnv(index, offset) \
TexEnv##index##Source = offset + 0, \
TexEnv##index##Operand = offset + 1, \
TexEnv##index##Combiner = offset + 2, \
TexEnv##index##Color = offset + 3, \
TexEnv##index##Scale = offset + 4,
// Geometry pipeline registers
VertexAttribLoc = 0x200,
AttribFormatLow = 0x201,
AttribFormatHigh = 0x202,
IndexBufferConfig = 0x227,
VertexCountReg = 0x228,
VertexOffsetReg = 0x22A,
SignalDrawArrays = 0x22E,
SignalDrawElements = 0x22F,
defineTexEnv(0, 0xC0)
defineTexEnv(1, 0xC8)
defineTexEnv(2, 0xD0)
defineTexEnv(3, 0xD8)
defineTexEnv(4, 0xF0)
defineTexEnv(5, 0xF8)
Attrib0Offset = 0x203,
Attrib1Offset = 0x206,
Attrib2Offset = 0x209,
Attrib3Offset = 0x20C,
Attrib4Offset = 0x20F,
Attrib5Offset = 0x212,
Attrib6Offset = 0x215,
Attrib7Offset = 0x218,
Attrib8Offset = 0x21B,
Attrib9Offset = 0x21E,
Attrib10Offset = 0x221,
Attrib11Offset = 0x224,
#undef defineTexEnv
// clang-format on
Attrib0Config2 = 0x205,
Attrib1Config2 = 0x208,
Attrib2Config2 = 0x20B,
Attrib3Config2 = 0x20E,
Attrib4Config2 = 0x211,
Attrib5Config2 = 0x214,
Attrib6Config2 = 0x217,
Attrib7Config2 = 0x21A,
Attrib8Config2 = 0x21D,
Attrib9Config2 = 0x220,
Attrib10Config2 = 0x223,
Attrib11Config2 = 0x226,
// Framebuffer registers
ColourOperation = 0x100,
BlendFunc = 0x101,
BlendColour = 0x103,
AlphaTestConfig = 0x104,
DepthAndColorMask = 0x107,
DepthBufferFormat = 0x116,
ColourBufferFormat = 0x117,
DepthBufferLoc = 0x11C,
ColourBufferLoc = 0x11D,
FramebufferSize = 0x11E,
AttribInfoStart = Attrib0Offset,
AttribInfoEnd = Attrib11Config2,
// Geometry pipeline registers
VertexAttribLoc = 0x200,
AttribFormatLow = 0x201,
AttribFormatHigh = 0x202,
IndexBufferConfig = 0x227,
VertexCountReg = 0x228,
VertexOffsetReg = 0x22A,
SignalDrawArrays = 0x22E,
SignalDrawElements = 0x22F,
// Fixed attribute registers
FixedAttribIndex = 0x232,
FixedAttribData0 = 0x233,
FixedAttribData1 = 0x234,
FixedAttribData2 = 0x235,
Attrib0Offset = 0x203,
Attrib1Offset = 0x206,
Attrib2Offset = 0x209,
Attrib3Offset = 0x20C,
Attrib4Offset = 0x20F,
Attrib5Offset = 0x212,
Attrib6Offset = 0x215,
Attrib7Offset = 0x218,
Attrib8Offset = 0x21B,
Attrib9Offset = 0x21E,
Attrib10Offset = 0x221,
Attrib11Offset = 0x224,
// Command processor registers
CmdBufSize0 = 0x238,
CmdBufSize1 = 0x239,
CmdBufAddr0 = 0x23A,
CmdBufAddr1 = 0x23B,
CmdBufTrigger0 = 0x23C,
CmdBufTrigger1 = 0x23D,
Attrib0Config2 = 0x205,
Attrib1Config2 = 0x208,
Attrib2Config2 = 0x20B,
Attrib3Config2 = 0x20E,
Attrib4Config2 = 0x211,
Attrib5Config2 = 0x214,
Attrib6Config2 = 0x217,
Attrib7Config2 = 0x21A,
Attrib8Config2 = 0x21D,
Attrib9Config2 = 0x220,
Attrib10Config2 = 0x223,
Attrib11Config2 = 0x226,
PrimitiveConfig = 0x25E,
PrimitiveRestart = 0x25F,
AttribInfoStart = Attrib0Offset,
AttribInfoEnd = Attrib11Config2,
// Vertex shader registers
VertexShaderAttrNum = 0x242,
VertexBoolUniform = 0x2B0,
VertexIntUniform0 = 0x2B1,
VertexIntUniform1 = 0x2B2,
VertexIntUniform2 = 0x2B3,
VertexIntUniform3 = 0x2B4,
// Fixed attribute registers
FixedAttribIndex = 0x232,
FixedAttribData0 = 0x233,
FixedAttribData1 = 0x234,
FixedAttribData2 = 0x235,
VertexShaderEntrypoint = 0x2BA,
VertexShaderTransferEnd = 0x2BF,
VertexFloatUniformIndex = 0x2C0,
VertexFloatUniformData0 = 0x2C1,
VertexFloatUniformData1 = 0x2C2,
VertexFloatUniformData2 = 0x2C3,
VertexFloatUniformData3 = 0x2C4,
VertexFloatUniformData4 = 0x2C5,
VertexFloatUniformData5 = 0x2C6,
VertexFloatUniformData6 = 0x2C7,
VertexFloatUniformData7 = 0x2C8,
// Command processor registers
CmdBufSize0 = 0x238,
CmdBufSize1 = 0x239,
CmdBufAddr0 = 0x23A,
CmdBufAddr1 = 0x23B,
CmdBufTrigger0 = 0x23C,
CmdBufTrigger1 = 0x23D,
VertexShaderInputBufferCfg = 0x2B9,
VertexShaderInputCfgLow = 0x2BB,
VertexShaderInputCfgHigh = 0x2BC,
PrimitiveConfig = 0x25E,
PrimitiveRestart = 0x25F,
VertexShaderTransferIndex = 0x2CB,
VertexShaderData0 = 0x2CC,
VertexShaderData1 = 0x2CD,
VertexShaderData2 = 0x2CE,
VertexShaderData3 = 0x2CF,
VertexShaderData4 = 0x2D0,
VertexShaderData5 = 0x2D1,
VertexShaderData6 = 0x2D2,
VertexShaderData7 = 0x2D3,
VertexShaderOpDescriptorIndex = 0x2D5,
VertexShaderOpDescriptorData0 = 0x2D6,
VertexShaderOpDescriptorData1 = 0x2D7,
VertexShaderOpDescriptorData2 = 0x2D8,
VertexShaderOpDescriptorData3 = 0x2D9,
VertexShaderOpDescriptorData4 = 0x2DA,
VertexShaderOpDescriptorData5 = 0x2DB,
VertexShaderOpDescriptorData6 = 0x2DC,
VertexShaderOpDescriptorData7 = 0x2DD,
// Vertex shader registers
VertexShaderAttrNum = 0x242,
VertexBoolUniform = 0x2B0,
VertexIntUniform0 = 0x2B1,
VertexIntUniform1 = 0x2B2,
VertexIntUniform2 = 0x2B3,
VertexIntUniform3 = 0x2B4,
VertexShaderEntrypoint = 0x2BA,
VertexShaderTransferEnd = 0x2BF,
VertexFloatUniformIndex = 0x2C0,
VertexFloatUniformData0 = 0x2C1,
VertexFloatUniformData1 = 0x2C2,
VertexFloatUniformData2 = 0x2C3,
VertexFloatUniformData3 = 0x2C4,
VertexFloatUniformData4 = 0x2C5,
VertexFloatUniformData5 = 0x2C6,
VertexFloatUniformData6 = 0x2C7,
VertexFloatUniformData7 = 0x2C8,
VertexShaderInputBufferCfg = 0x2B9,
VertexShaderInputCfgLow = 0x2BB,
VertexShaderInputCfgHigh = 0x2BC,
VertexShaderTransferIndex = 0x2CB,
VertexShaderData0 = 0x2CC,
VertexShaderData1 = 0x2CD,
VertexShaderData2 = 0x2CE,
VertexShaderData3 = 0x2CF,
VertexShaderData4 = 0x2D0,
VertexShaderData5 = 0x2D1,
VertexShaderData6 = 0x2D2,
VertexShaderData7 = 0x2D3,
VertexShaderOpDescriptorIndex = 0x2D5,
VertexShaderOpDescriptorData0 = 0x2D6,
VertexShaderOpDescriptorData1 = 0x2D7,
VertexShaderOpDescriptorData2 = 0x2D8,
VertexShaderOpDescriptorData3 = 0x2D9,
VertexShaderOpDescriptorData4 = 0x2DA,
VertexShaderOpDescriptorData5 = 0x2DB,
VertexShaderOpDescriptorData6 = 0x2DC,
VertexShaderOpDescriptorData7 = 0x2DD,
};
}
enum class TextureFmt : u32 {
RGBA8 = 0x0,
RGB8 = 0x1,
RGBA5551 = 0x2,
RGB565 = 0x3,
RGBA4 = 0x4,
IA8 = 0x5,
RG8 = 0x6,
I8 = 0x7,
A8 = 0x8,
IA4 = 0x9,
I4 = 0xA,
A4 = 0xB,
ETC1 = 0xC,
ETC1A4 = 0xD,
};
}
enum class ColorFmt : u32 {
RGBA8 = 0x0,
RGB8 = 0x1,
RGBA5551 = 0x2,
RGB565 = 0x3,
RGBA4 = 0x4,
};
enum class DepthFmt : u32 {
Depth16 = 0,
Unknown1 = 1, // Technically selectable, but function is unknown
Depth24 = 2,
Depth24Stencil8 = 3,
};
// Returns the string representation of a texture format
inline constexpr const char* textureFormatToString(TextureFmt fmt) {
switch (fmt) {
case TextureFmt::RGBA8: return "RGBA8";
case TextureFmt::RGB8: return "RGB8";
case TextureFmt::RGBA5551: return "RGBA5551";
case TextureFmt::RGB565: return "RGB565";
case TextureFmt::RGBA4: return "RGBA4";
case TextureFmt::IA8: return "IA8";
case TextureFmt::RG8: return "RG8";
case TextureFmt::I8: return "I8";
case TextureFmt::A8: return "A8";
case TextureFmt::IA4: return "IA4";
case TextureFmt::I4: return "I4";
case TextureFmt::A4: return "A4";
case TextureFmt::ETC1: return "ETC1";
case TextureFmt::ETC1A4: return "ETC1A4";
default: return "Unknown";
}
}
inline constexpr const char* textureFormatToString(ColorFmt fmt) {
return textureFormatToString(static_cast<TextureFmt>(fmt));
}
inline constexpr bool hasStencil(DepthFmt format) { return format == PICA::DepthFmt::Depth24Stencil8; }
// Size occupied by each pixel in bytes
// All formats are 16BPP except for RGBA8 (32BPP) and BGR8 (24BPP)
inline constexpr usize sizePerPixel(TextureFmt format) {
switch (format) {
case TextureFmt::RGB8: return 3;
case TextureFmt::RGBA8: return 4;
default: return 2;
}
}
inline constexpr usize sizePerPixel(ColorFmt format) {
return sizePerPixel(static_cast<TextureFmt>(format));
}
inline constexpr usize sizePerPixel(DepthFmt format) {
switch (format) {
case DepthFmt::Depth16: return 2;
case DepthFmt::Depth24: return 3;
case DepthFmt::Depth24Stencil8: return 4;
default: return 1; // Invalid format
}
}
enum class PrimType : u32 {
TriangleList = 0,
TriangleStrip = 1,
TriangleFan = 2,
GeometryPrimitive = 3,
};
} // namespace PICA

21
include/renderer_gl/renderer_gl.hpp
View file

@ -7,6 +7,7 @@
#include "opengl.hpp"
#include "surface_cache.hpp"
#include "textures.hpp"
#include "PICA/regs.hpp"
// More circular dependencies!
class GPU;
@ -45,12 +46,12 @@ class Renderer {
OpenGL::uvec2 fbSize; // The size of the framebuffer (ie both the colour and depth buffer)'
u32 colourBufferLoc; // Location in 3DS VRAM for the colour buffer
ColourBuffer::Formats colourBufferFormat; // Format of the colours stored in the colour buffer
u32 colourBufferLoc; // Location in 3DS VRAM for the colour buffer
PICA::ColorFmt colourBufferFormat; // Format of the colours stored in the colour buffer
// Same for the depth/stencil buffer
u32 depthBufferLoc;
DepthBuffer::Formats depthBufferFormat;
PICA::DepthFmt depthBufferFormat;
// Dummy VAO/VBO for blitting the final output
OpenGL::VertexArray dummyVAO;
@ -75,23 +76,19 @@ class Renderer {
void getGraphicsContext(); // Set up graphics context for rendering
void clearBuffer(u32 startAddress, u32 endAddress, u32 value, u32 control); // Clear a GPU buffer in VRAM
void displayTransfer(u32 inputAddr, u32 outputAddr, u32 inputSize, u32 outputSize, u32 flags); // Perform display transfer
void drawVertices(OpenGL::Primitives primType, std::span<const Vertex> vertices); // Draw the given vertices
void drawVertices(PICA::PrimType primType, std::span<const Vertex> vertices); // Draw the given vertices
void setFBSize(u32 width, u32 height) {
fbSize.x() = width;
fbSize.y() = height;
}
void setColourFormat(ColourBuffer::Formats format) { colourBufferFormat = format; }
void setColourFormat(u32 format) { colourBufferFormat = static_cast<ColourBuffer::Formats>(format); }
void setDepthFormat(DepthBuffer::Formats format) { depthBufferFormat = format; }
void setDepthFormat(u32 format) {
if (format == 1) {
void setColourFormat(PICA::ColorFmt format) { colourBufferFormat = format; }
void setDepthFormat(PICA::DepthFmt format) {
if (format == PICA::DepthFmt::Unknown1) {
Helpers::panic("[PICA] Undocumented depth-stencil mode!");
}
depthBufferFormat = static_cast<DepthBuffer::Formats>(format);
depthBufferFormat = format;
}
void setColourBufferLoc(u32 loc) { colourBufferLoc = loc; }

55
include/renderer_gl/surfaces.hpp
View file

@ -1,4 +1,5 @@
#pragma once
#include "PICA/regs.hpp"
#include "boost/icl/interval.hpp"
#include "helpers.hpp"
#include "opengl.hpp"
@ -7,18 +8,8 @@ template <typename T>
using Interval = boost::icl::right_open_interval<T>;
struct ColourBuffer {
enum class Formats : u32 {
RGBA8 = 0,
BGR8 = 1,
RGB5A1 = 2,
RGB565 = 3,
RGBA4 = 4,
Trash1 = 5, Trash2 = 6, Trash3 = 7 // Technically selectable, but their function is unknown
};
u32 location;
Formats format;
PICA::ColorFmt format;
OpenGL::uvec2 size;
bool valid;
@ -30,7 +21,7 @@ struct ColourBuffer {
ColourBuffer() : valid(false) {}
ColourBuffer(u32 loc, Formats format, u32 x, u32 y, bool valid = true)
ColourBuffer(u32 loc, PICA::ColorFmt format, u32 x, u32 y, bool valid = true)
: location(loc), format(format), size({x, y}), valid(valid) {
u64 endLoc = (u64)loc + sizeInBytes();
@ -78,31 +69,14 @@ struct ColourBuffer {
size.x() == other.size.x() && size.y() == other.size.y();
}
// Size occupied by each pixel in bytes
// All formats are 16BPP except for RGBA8 (32BPP) and BGR8 (24BPP)
size_t sizePerPixel() {
switch (format) {
case Formats::BGR8: return 3;
case Formats::RGBA8: return 4;
default: return 2;
}
}
size_t sizeInBytes() {
return (size_t)size.x() * (size_t)size.y() * sizePerPixel();
return (size_t)size.x() * (size_t)size.y() * PICA::sizePerPixel(format);
}
};
struct DepthBuffer {
enum class Formats : u32 {
Depth16 = 0,
Garbage = 1,
Depth24 = 2,
Depth24Stencil8 = 3
};
u32 location;
Formats format;
PICA::DepthFmt format;
OpenGL::uvec2 size; // Implicitly set to the size of the framebuffer
bool valid;
@ -113,7 +87,7 @@ struct DepthBuffer {
DepthBuffer() : valid(false) {}
DepthBuffer(u32 loc, Formats format, u32 x, u32 y, bool valid = true) :
DepthBuffer(u32 loc, PICA::DepthFmt format, u32 x, u32 y, bool valid = true) :
location(loc), format(format), size({x, y}), valid(valid) {
u64 endLoc = (u64)loc + sizeInBytes();
@ -121,10 +95,6 @@ struct DepthBuffer {
range = Interval<u32>(loc, (u32)endLoc);
}
bool hasStencil() {
return format == Formats::Depth24Stencil8;
}
void allocate() {
// Create texture for the FBO, setting up filters and the like
// Reading back the current texture is slow, but allocate calls should be few and far between.
@ -167,18 +137,7 @@ struct DepthBuffer {
size.x() == other.size.x() && size.y() == other.size.y();
}
// Size occupied by each pixel in bytes
size_t sizePerPixel() {
switch (format) {
case Formats::Depth16: return 2;
case Formats::Depth24: return 3;
case Formats::Depth24Stencil8: return 4;
default: return 1; // Invalid format
}
}
size_t sizeInBytes() {
return (size_t)size.x() * (size_t)size.y() * sizePerPixel();
return (size_t)size.x() * (size_t)size.y() * PICA::sizePerPixel(format);
}
};

31
include/renderer_gl/textures.hpp
View file

@ -1,6 +1,7 @@
#pragma once
#include <array>
#include <string>
#include "PICA/regs.hpp"
#include "boost/icl/interval.hpp"
#include "helpers.hpp"
#include "opengl.hpp"
@ -9,28 +10,9 @@ template <typename T>
using Interval = boost::icl::right_open_interval<T>;
struct Texture {
enum class Formats : u32 {
RGBA8 = 0,
RGB8 = 1,
RGBA5551 = 2,
RGB565 = 3,
RGBA4 = 4,
IA8 = 5,
RG8 = 6,
I8 = 7,
A8 = 8,
IA4 = 9,
I4 = 10,
A4 = 11,
ETC1 = 12,
ETC1A4 = 13,
Trash1 = 14, Trash2 = 15 // TODO: What are these?
};
u32 location;
u32 config; // Magnification/minification filter, wrapping configs, etc
Formats format;
PICA::TextureFmt format;
OpenGL::uvec2 size;
bool valid;
@ -41,7 +23,7 @@ struct Texture {
Texture() : valid(false) {}
Texture(u32 loc, Formats format, u32 x, u32 y, u32 config, bool valid = true)
Texture(u32 loc, PICA::TextureFmt format, u32 x, u32 y, u32 config, bool valid = true)
: location(loc), format(format), size({x, y}), config(config), valid(valid) {
u64 endLoc = (u64)loc + sizeInBytes();
@ -62,7 +44,7 @@ struct Texture {
void free();
u64 sizeInBytes();
u32 decodeTexel(u32 u, u32 v, Formats fmt, const void* data);
u32 decodeTexel(u32 u, u32 v, PICA::TextureFmt fmt, const void* data);
// Get the morton interleave offset of a texel based on its U and V values
static u32 mortonInterleave(u32 u, u32 v);
@ -70,12 +52,9 @@ struct Texture {
static u32 getSwizzledOffset(u32 u, u32 v, u32 width, u32 bytesPerPixel);
static u32 getSwizzledOffset_4bpp(u32 u, u32 v, u32 width);
// Returns the string representation of a texture format
static std::string textureFormatToString(Formats fmt);
// Returns the format of this texture as a string
std::string formatToString() {
return textureFormatToString(format);
return PICA::textureFormatToString(format);
}
// Returns the texel at coordinates (u, v) of an ETC1(A4) texture