singe/thirdparty/RmlUi/Backends/RmlUi_Renderer_DX11.cpp

#include "RmlUi_Renderer_DX11.h"
#include "RmlUi_Include_Windows.h"
#include <RmlUi/Core/Core.h>
#include <RmlUi/Core/DecorationTypes.h>
#include <RmlUi/Core/FileInterface.h>
#include <RmlUi/Core/Geometry.h>
#include <RmlUi/Core/Log.h>
#include <RmlUi/Core/MeshUtilities.h>
#include <RmlUi/Core/Platform.h>
#include <RmlUi/Core/SystemInterface.h>

// Helper for a common cleanup pattern in DX11
template <typename T>
static void Dx11CleanupResourceIfCreated(T*& resource)
{
	if (resource)
	{
		resource->Release();
		resource = nullptr;
	}
}

#ifndef RMLUI_PLATFORM_WIN32
	#error "Cannot build DX11 renderer on non-Windows platforms"
#endif // RMLUI_PLATFORM_WIN32

#include "d3dcompiler.h"

#ifdef RMLUI_DX_DEBUG
	#include <dxgidebug.h>
#endif

// Shader source code

#define MAX_NUM_STOPS 16
#define BLUR_SIZE 7
#define BLUR_NUM_WEIGHTS ((BLUR_SIZE + 1) / 2)

#define RMLUI_STRINGIFY_IMPL(x) #x
#define RMLUI_STRINGIFY(x) RMLUI_STRINGIFY_IMPL(x)

#define RMLUI_SHADER_HEADER "#define MAX_NUM_STOPS " RMLUI_STRINGIFY(MAX_NUM_STOPS) "\n#line " RMLUI_STRINGIFY(__LINE__) "\n"

constexpr const char shader_vert_main[] = RMLUI_SHADER_HEADER R"(
struct VS_Input
{
    float2 position : POSITION;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

struct PS_INPUT
{
    float4 position : SV_Position;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

cbuffer SharedConstantBuffer : register(b0)
{
    float4x4 m_transform;
    float2 m_translate;
};

PS_INPUT VSMain(const VS_Input IN)
{
    PS_INPUT result = (PS_INPUT)0;

    float2 translatedPos = IN.position + m_translate;
    float4 resPos = mul(m_transform, float4(translatedPos.x, translatedPos.y, 0.0, 1.0));

    result.position = resPos;
    result.color = IN.color;
    result.uv = IN.uv;

    return result;
};
)";

constexpr const char shader_frag_texture[] = RMLUI_SHADER_HEADER R"(
struct PS_Input
{
    float4 position : SV_Position;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

Texture2D g_InputTexture : register(t0);
SamplerState g_SamplerLinear : register(s0);

float4 PSMain(const PS_Input IN) : SV_TARGET
{
    return IN.color * g_InputTexture.Sample(g_SamplerLinear, IN.uv);
};
)";

constexpr const char shader_frag_color[] = RMLUI_SHADER_HEADER R"(
struct PS_Input
{
    float4 position : SV_Position;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

float4 PSMain(const PS_Input IN) : SV_TARGET
{
    return IN.color;
};
)";

enum class ShaderGradientFunction { Linear, Radial, Conic, RepeatingLinear, RepeatingRadial, RepeatingConic }; // Must match shader definitions below.

// We need to round up at compile-time so that we can address each element of the float4s
#define CEILING(x, y) (((x) + (y) - 1) / (y))
static const char shader_frag_gradient[] = RMLUI_SHADER_HEADER "#define MAX_NUM_STOPS_PACKED (uint)" RMLUI_STRINGIFY(CEILING(MAX_NUM_STOPS, 4)) R"(
#define LINEAR 0
#define RADIAL 1
#define CONIC 2
#define REPEATING_LINEAR 3
#define REPEATING_RADIAL 4
#define REPEATING_CONIC 5
#define PI 3.14159265

cbuffer SharedConstantBuffer : register(b0)
{
    float4x4 m_transform;
    float2 m_translate;

    // One to one translation of the OpenGL uniforms results in a LOT of wasted space due to CBuffer alignment rules.
    // Changes from GL3:
    // - Moved m_num_stops below m_func (saved 4 bytes of padding).
    // - Packed m_stop_positions into a float4[MAX_NUM_STOPS / 4] array, as each array element starts a new 16-byte row.
    // The below layout has 0 bytes of padding.

    int m_func;   // one of the above definitions
    int m_num_stops;
    float2 m_p;   // linear: starting point,         radial: center,                        conic: center
    float2 m_v;   // linear: vector to ending point, radial: 2d curvature (inverse radius), conic: angled unit vector
    float4 m_stop_colors[MAX_NUM_STOPS];
    float4 m_stop_positions[MAX_NUM_STOPS_PACKED]; // normalized, 0 -> starting point, 1 -> ending point
};

// Hide the way the data is packed in the cbuffer through a macro
// @NOTE: Hardcoded for MAX_NUM_STOPS 16.
//        i >> 2 => i >> sqrt(MAX_NUM_STOPS)
#define GET_STOP_POS(i) (m_stop_positions[i >> 2][i & 3])

struct PS_Input
{
    float4 position : SV_Position;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

#define glsl_mod(x,y) (((x)-(y)*floor((x)/(y))))

float4 lerp_stop_colors(float t) {
    float4 color = m_stop_colors[0];

    for (int i = 1; i < m_num_stops; i++)
        color = lerp(color, m_stop_colors[i], smoothstep(GET_STOP_POS(i-1), GET_STOP_POS(i), t));

    return color;
};

float4 PSMain(const PS_Input IN) : SV_TARGET
{
    float t = 0.0;

    if (m_func == LINEAR || m_func == REPEATING_LINEAR) {
        float dist_square = dot(m_v, m_v);
        float2 V = IN.uv.xy - m_p;
        t = dot(m_v, V) / dist_square;
    }
    else if (m_func == RADIAL || m_func == REPEATING_RADIAL) {
        float2 V = IN.uv.xy - m_p;
        t = length(m_v * V);
    }
    else if (m_func == CONIC || m_func == REPEATING_CONIC) {
        float2x2 R = float2x2(m_v.x, -m_v.y, m_v.y, m_v.x);
        float2 V = mul((IN.uv.xy - m_p), R);
        t = 0.5 + atan2(-V.x, V.y) / (2.0 * PI);
    }

    if (m_func == REPEATING_LINEAR || m_func == REPEATING_RADIAL || m_func == REPEATING_CONIC) {
        float t0 = GET_STOP_POS(0);
        float t1 = GET_STOP_POS(m_num_stops - 1);
        t = t0 + glsl_mod(t - t0, t1 - t0);
    }

    return IN.color * lerp_stop_colors(t);
};
)";

// "Creation" by Danilo Guanabara, based on: https://www.shadertoy.com/view/XsXXDn
static const char shader_frag_creation[] = RMLUI_SHADER_HEADER R"(
struct PS_Input
{
    float4 position : SV_Position;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

cbuffer SharedConstantBuffer : register(b0)
{
    float4x4 m_transform;
    float2 m_translate;
    float2 m_dimensions;
    float m_value;
};

#define glsl_mod(x,y) (((x)-(y)*floor((x)/(y))))

float4 PSMain(const PS_Input IN) : SV_TARGET
{
    float t = m_value;
    float3 c;
    float l;
    for (int i = 0; i < 3; i++) {
        float2 p = IN.uv;
        float2 uv = p;
        p -= .5;
        p.x *= m_dimensions.x / m_dimensions.y;
        float z = t + ((float)i) * .07;
        l = length(p);
        uv += p / l * (sin(z) + 1.) * abs(sin(l * 9. - z - z));
        c[i] = .01 / length(glsl_mod(uv, 1.) - .5);
    }
    return float4(c / l, IN.color.a);
};
)";

constexpr const char shader_vert_passthrough[] = RMLUI_SHADER_HEADER R"(
struct VS_Input
{
    float2 position : POSITION;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

struct PS_Input
{
    float4 position : SV_Position;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

cbuffer ConstantBuffer : register(b0)
{
    float4x4 m_transform;
    float2 m_translate;
};

PS_Input VSMain(const VS_Input IN)
{
    PS_Input result = (PS_Input)0;

    result.position = float4(IN.position.xy, 0.0f, 1.0f);
    result.uv = float2(IN.uv.x, 1.0f - IN.uv.y);

    return result;
};
)";

constexpr const char shader_frag_passthrough[] = RMLUI_SHADER_HEADER R"(
struct PS_Input
{
    float4 position : SV_Position;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

Texture2D g_InputTexture : register(t0);
SamplerState g_SamplerLinear : register(s0);

float4 PSMain(const PS_Input IN) : SV_TARGET
{
    return g_InputTexture.Sample(g_SamplerLinear, IN.uv);
};
)";

static const char shader_frag_color_matrix[] = RMLUI_SHADER_HEADER R"(

Texture2D g_InputTexture : register(t0);
SamplerState g_SamplerLinear : register(s0);

cbuffer ConstantBuffer : register(b1)
{
    float4x4 m_color_matrix;
};

struct PS_Input
{
    float4 position : SV_Position;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

float4 PSMain(const PS_Input IN) : SV_TARGET
{
    // The general case uses a 4x5 color matrix for full rgba transformation, plus a constant term with the last column.
    // However, we only consider the case of rgb transformations. Thus, we could in principle use a 3x4 matrix, but we
    // keep the alpha row for simplicity.
    // In the general case we should do the matrix transformation in non-premultiplied space. However, without alpha
    // transformations, we can do it directly in premultiplied space to avoid the extra division and multiplication
    // steps. In this space, the constant term needs to be multiplied by the alpha value, instead of unity.
    float4 texColor = g_InputTexture.Sample(g_SamplerLinear, IN.uv);
    float3 transformedColor = mul(m_color_matrix, texColor).rgb;
    return float4(transformedColor, texColor.a);
};
)";

static const char shader_frag_blend_mask[] = RMLUI_SHADER_HEADER R"(
Texture2D g_InputTexture : register(t0);
Texture2D g_MaskTexture : register(t1);
SamplerState g_SamplerLinear : register(s0);

struct PS_Input
{
    float4 position : SV_Position;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

float4 PSMain(const PS_Input IN) : SV_TARGET
{
    float4 texColor = g_InputTexture.Sample(g_SamplerLinear, IN.uv);
    float maskAlpha = g_MaskTexture.Sample(g_SamplerLinear, IN.uv).a;
    return texColor * maskAlpha;
};
)";

#define RMLUI_SHADER_BLUR_HEADER \
	RMLUI_SHADER_HEADER "\n#define BLUR_SIZE " RMLUI_STRINGIFY(BLUR_SIZE) "\n#define BLUR_NUM_WEIGHTS " RMLUI_STRINGIFY(BLUR_NUM_WEIGHTS)

static const char shader_vert_blur[] = RMLUI_SHADER_BLUR_HEADER R"(
struct VS_Input
{
    float2 position : POSITION;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

struct PS_INPUT
{
    float4 position : SV_Position;
    float2 uv[BLUR_SIZE] : TEXCOORD;
};

cbuffer SharedConstantBuffer : register(b0)
{
    float4x4 m_transform;
    float2 m_translate;
    float2 m_texelOffset;
    float4 m_weights;
    float2 m_texCoordMin;
    float2 m_texCoordMax;
};

PS_INPUT VSMain(const VS_Input IN)
{
    PS_INPUT result = (PS_INPUT)0;

    for (int i = 0; i < BLUR_SIZE; i++) {
        result.uv[i] = IN.uv - float(i - BLUR_NUM_WEIGHTS + 1) * m_texelOffset;
    }
    result.position = float4(IN.position.xy, 1.0, 1.0);

    return result;
};
)";

static const char shader_frag_blur[] = RMLUI_SHADER_BLUR_HEADER R"(
Texture2D g_InputTexture : register(t0);
SamplerState g_SamplerLinear : register(s0);

cbuffer SharedConstantBuffer : register(b0)
{
    float4x4 m_transform;
    float2 m_translate;
    float2 m_texelOffset;
    float4 m_weights;
    float2 m_texCoordMin;
    float2 m_texCoordMax;
};

struct PS_Input
{
    float4 position : SV_Position;
    float2 uv[BLUR_SIZE] : TEXCOORD;
};

float4 PSMain(const PS_Input IN) : SV_TARGET
{
    float4 color = float4(0.0, 0.0, 0.0, 0.0);
    for(int i = 0; i < BLUR_SIZE; i++)
    {
        float2 in_region = step(m_texCoordMin, IN.uv[i]) * step(IN.uv[i], m_texCoordMax);
        color += g_InputTexture.Sample(g_SamplerLinear, float2(IN.uv[i].x, 1.0f - IN.uv[i].y)) * in_region.x * in_region.y * m_weights[abs(i - BLUR_NUM_WEIGHTS + 1)];
    }
    return color;
};
)";

static const char shader_frag_drop_shadow[] = RMLUI_SHADER_HEADER R"(
Texture2D g_InputTexture : register(t0);
SamplerState g_SamplerLinear : register(s0);

cbuffer DropShadowBuffer : register(b1)
{
    float2 m_texCoordMin;
    float2 m_texCoordMax;
    float4 m_color;
};

struct PS_Input
{
    float4 position : SV_Position;
    float4 color : COLOR;
    float2 uv : TEXCOORD;
};

float4 PSMain(const PS_Input IN) : SV_TARGET
{
    float2 in_region = step(m_texCoordMin, IN.uv) * step(IN.uv, m_texCoordMax);
    return g_InputTexture.Sample(g_SamplerLinear, IN.uv).a * in_region.x * in_region.y * m_color;
};
)";

enum class ProgramId {
	None,
	Color,
	Texture,
	Gradient,
	Creation,
	Passthrough,
	ColorMatrix,
	BlendMask,
	Blur,
	DropShadow,
	Count,
};
enum class VertShaderId {
	Main,
	Passthrough,
	Blur,
	Count,
};
enum class FragShaderId {
	Color,
	Texture,
	Gradient,
	Creation,
	Passthrough,
	ColorMatrix,
	BlendMask,
	Blur,
	DropShadow,
	Count,
};
enum class ShaderType {
	Vertex,
	Fragment,
};

namespace Gfx {

struct VertShaderDefinition {
	VertShaderId id;
	const char* name_str;
	const char* code_str;
	const size_t code_size;
};
struct FragShaderDefinition {
	FragShaderId id;
	const char* name_str;
	const char* code_str;
	const size_t code_size;
};
struct ProgramDefinition {
	ProgramId id;
	const char* name_str;
	VertShaderId vert_shader;
	FragShaderId frag_shader;
};

// clang-format off
static const VertShaderDefinition vert_shader_definitions[] = {
    {VertShaderId::Main,        "main",         shader_vert_main,           sizeof(shader_vert_main)},
    {VertShaderId::Passthrough, "passthrough",  shader_vert_passthrough,    sizeof(shader_vert_passthrough)},
    {VertShaderId::Blur,        "blur",         shader_vert_blur,           sizeof(shader_vert_blur)},
};
static const FragShaderDefinition frag_shader_definitions[] = {
    {FragShaderId::Color,       "color",        shader_frag_color,          sizeof(shader_frag_color)},
    {FragShaderId::Texture,     "texture",      shader_frag_texture,        sizeof(shader_frag_texture)},
    {FragShaderId::Gradient,    "gradient",     shader_frag_gradient,       sizeof(shader_frag_gradient)},
    {FragShaderId::Creation,    "creation",     shader_frag_creation,       sizeof(shader_frag_creation)},
    {FragShaderId::Passthrough, "passthrough",  shader_frag_passthrough,    sizeof(shader_frag_passthrough)},
    {FragShaderId::ColorMatrix, "color_matrix", shader_frag_color_matrix,   sizeof(shader_frag_color_matrix)},
    {FragShaderId::BlendMask,   "blend_mask",   shader_frag_blend_mask,     sizeof(shader_frag_blend_mask)},
    {FragShaderId::Blur,        "blur",         shader_frag_blur,           sizeof(shader_frag_blur)},
    {FragShaderId::DropShadow,  "drop_shadow",  shader_frag_drop_shadow,    sizeof(shader_frag_drop_shadow)},
};
static const ProgramDefinition program_definitions[] = {
    {ProgramId::Color,       "color",        VertShaderId::Main,        FragShaderId::Color},
    {ProgramId::Texture,     "texture",      VertShaderId::Main,        FragShaderId::Texture},
    {ProgramId::Gradient,    "gradient",     VertShaderId::Main,        FragShaderId::Gradient},
    {ProgramId::Creation,    "creation",     VertShaderId::Main,        FragShaderId::Creation},
    {ProgramId::Passthrough, "passthrough",  VertShaderId::Passthrough, FragShaderId::Passthrough},
    {ProgramId::ColorMatrix, "color_matrix", VertShaderId::Passthrough, FragShaderId::ColorMatrix},
    {ProgramId::BlendMask,   "blend_mask",   VertShaderId::Passthrough, FragShaderId::BlendMask},
    {ProgramId::Blur,        "blur",         VertShaderId::Blur,        FragShaderId::Blur},
    {ProgramId::DropShadow,  "drop_shadow",  VertShaderId::Passthrough, FragShaderId::DropShadow},
};
// clang-format on

template <typename T, typename Enum>
class EnumArray {
public:
	const T& operator[](Enum id) const
	{
		RMLUI_ASSERT((size_t)id < (size_t)Enum::Count);
		return ids[size_t(id)];
	}
	T& operator[](Enum id)
	{
		RMLUI_ASSERT((size_t)id < (size_t)Enum::Count);
		return ids[size_t(id)];
	}
	auto begin() const { return ids.begin(); }
	auto end() const { return ids.end(); }

private:
	Rml::Array<T, (size_t)Enum::Count> ids = {};
};

struct ShaderProgram {
public:
	ID3D11VertexShader* vertex_shader;
	ID3D11PixelShader* pixel_shader;
};

using Programs = EnumArray<ShaderProgram, ProgramId>;
using VertShaders = EnumArray<ID3DBlob*, VertShaderId>;
using FragShaders = EnumArray<ID3DBlob*, FragShaderId>;

struct ProgramData {
	Programs programs;
	VertShaders vert_shaders;
	FragShaders frag_shaders;
};

// Create the shader, 'shader_type' is either GL_VERTEX_SHADER or GL_FRAGMENT_SHADER.
static bool CreateShader(ID3DBlob*& out_shader_dxil, ShaderType shader_type, const char* code_string, const size_t code_length)
{
	RMLUI_ASSERT(shader_type == ShaderType::Vertex || shader_type == ShaderType::Fragment);

#ifdef RMLUI_DX_DEBUG
	static const UINT default_shader_flags = D3DCOMPILE_DEBUG | D3DCOMPILE_SKIP_OPTIMIZATION;
#else
	static const UINT default_shader_flags = 0;
#endif

	// Compile shader
	ID3DBlob* p_error_buff = nullptr;

	HRESULT result = D3DCompile(code_string, code_length, nullptr, nullptr, nullptr, shader_type == ShaderType::Vertex ? "VSMain" : "PSMain",
		shader_type == ShaderType::Vertex ? "vs_5_0" : "ps_5_0", default_shader_flags, 0, &out_shader_dxil, &p_error_buff);
	RMLUI_DX_ASSERTMSG(result, "failed to D3DCompile");
#ifdef RMLUI_DX_DEBUG
	if (FAILED(result))
	{
		Rml::Log::Message(Rml::Log::Type::LT_ERROR, "failed to compile shader: %s", (char*)p_error_buff->GetBufferPointer());
		return false;
	}
#endif

	Dx11CleanupResourceIfCreated(p_error_buff);

	return true;
}

static bool CreateProgram(ID3D11Device* p_device, ShaderProgram& out_program, ID3DBlob* vertex_shader, ID3DBlob* fragment_shader)
{
	RMLUI_ASSERT(vertex_shader);
	RMLUI_ASSERT(fragment_shader);

	HRESULT result =
		p_device->CreateVertexShader(vertex_shader->GetBufferPointer(), vertex_shader->GetBufferSize(), nullptr, &out_program.vertex_shader);
	RMLUI_DX_ASSERTMSG(result, "failed to CreateVertexShader");
	if (FAILED(result))
	{
#ifdef RMLUI_DX_DEBUG
		Rml::Log::Message(Rml::Log::Type::LT_ERROR, "failed to create vertex shader: %d", result);
#endif
		return false;
	}

	result = p_device->CreatePixelShader(fragment_shader->GetBufferPointer(), fragment_shader->GetBufferSize(), nullptr, &out_program.pixel_shader);
	RMLUI_DX_ASSERTMSG(result, "failed to CreatePixelShader");
	if (FAILED(result))
	{
#ifdef RMLUI_DX_DEBUG
		Rml::Log::Message(Rml::Log::Type::LT_ERROR, "failed to create pixel shader: %d", result);
#endif
		return false;
	}

	return true;
}

struct RenderTargetData {
	int width = 0, height = 0;
	ID3D11RenderTargetView* render_target_view = nullptr;                   // To write to colour attachment buffer
	ID3D11DepthStencilView* depth_stencil_view = nullptr;                   // To write to stencil buffer
	ID3D11Texture2D* render_target_texture = nullptr;                       // For MSAA resolve
	ID3D11ShaderResourceView* render_target_shader_resource_view = nullptr; // To blit
	bool owns_depth_stencil_buffer = false;
};

enum class RenderTargetAttachment { None, Depth, DepthStencil };

static bool CreateRenderTarget(ID3D11Device* p_device, RenderTargetData& out_rt, int width, int height, int samples,
	RenderTargetAttachment attachment, ID3D11DepthStencilView* shared_depth_stencil_buffer)
{
	// Generate render target
	D3D11_TEXTURE2D_DESC texture_desc = {};
	ZeroMemory(&texture_desc, sizeof(D3D11_TEXTURE2D_DESC));
	texture_desc.Width = width;
	texture_desc.Height = height;
	texture_desc.MipLevels = 1;
	texture_desc.ArraySize = 1;
	texture_desc.Format = DXGI_FORMAT_R8G8B8A8_UNORM;
	texture_desc.SampleDesc.Count = (samples > 1) ? samples : 1; // MSAA
	texture_desc.SampleDesc.Quality = (samples > 1) ? D3D11_STANDARD_MULTISAMPLE_PATTERN : 0;
	texture_desc.Usage = D3D11_USAGE_DEFAULT;
	texture_desc.BindFlags = D3D11_BIND_RENDER_TARGET | D3D11_BIND_SHADER_RESOURCE;
	texture_desc.CPUAccessFlags = 0;
	texture_desc.MiscFlags = 0;

	ID3D11Texture2D* rt_texture = nullptr;
	HRESULT result = p_device->CreateTexture2D(&texture_desc, nullptr, &rt_texture);
	RMLUI_DX_ASSERTMSG(result, "failed to CreateTexture2D");
	if (FAILED(result))
	{
		Rml::Log::Message(Rml::Log::LT_ERROR, "ID3D11Device::CreateTexture2D (%d)", result);
		return false;
	}

	D3D11_RENDER_TARGET_VIEW_DESC render_target_view_desc = {};
	render_target_view_desc.Format = texture_desc.Format;
	render_target_view_desc.ViewDimension = samples > 0 ? D3D11_RTV_DIMENSION_TEXTURE2DMS : D3D11_RTV_DIMENSION_TEXTURE2D;
	render_target_view_desc.Texture2D.MipSlice = 0;

	ID3D11RenderTargetView* render_target_view = nullptr;
	result = p_device->CreateRenderTargetView(rt_texture, &render_target_view_desc, &render_target_view);
	RMLUI_DX_ASSERTMSG(result, "failed to CreateRenderTargetView");
	if (FAILED(result))
	{
		Rml::Log::Message(Rml::Log::LT_ERROR, "ID3D11Device::CreateRenderTargetView (%d)", result);
		return false;
	}

	D3D11_SHADER_RESOURCE_VIEW_DESC render_target_shader_resource_view_desc = {};
	render_target_shader_resource_view_desc.Format = texture_desc.Format;
	render_target_shader_resource_view_desc.ViewDimension = samples > 0 ? D3D11_SRV_DIMENSION_TEXTURE2DMS : D3D11_SRV_DIMENSION_TEXTURE2D;
	render_target_shader_resource_view_desc.Texture2D.MostDetailedMip = 0;
	render_target_shader_resource_view_desc.Texture2D.MipLevels = 1;

	ID3D11ShaderResourceView* render_target_shader_resource_view = nullptr;
	result = p_device->CreateShaderResourceView(rt_texture, &render_target_shader_resource_view_desc, &render_target_shader_resource_view);
	RMLUI_DX_ASSERTMSG(result, "failed to CreateShaderResourceView");
	if (FAILED(result))
	{
		Rml::Log::Message(Rml::Log::LT_ERROR, "ID3D11Device::CreateShaderResourceView (%d)", result);
		return false;
	}

	// Generate stencil buffer if necessary
	ID3D11DepthStencilView* depth_stencil_view = nullptr;

	if (attachment != RenderTargetAttachment::None)
	{
		if (shared_depth_stencil_buffer)
		{
			// Share the depth/stencil buffer
			depth_stencil_view = shared_depth_stencil_buffer;
			depth_stencil_view->AddRef(); // Increase reference count since we're sharing it
		}
		else
		{
			// Create a new depth/stencil buffer
			D3D11_TEXTURE2D_DESC depth_desc = {};
			depth_desc.Width = width;
			depth_desc.Height = height;
			depth_desc.MipLevels = 1;
			depth_desc.ArraySize = 1;
			depth_desc.Format =
				(attachment == RenderTargetAttachment::DepthStencil) ? DXGI_FORMAT_D32_FLOAT_S8X24_UINT : DXGI_FORMAT_D24_UNORM_S8_UINT;
			depth_desc.SampleDesc = texture_desc.SampleDesc;
			depth_desc.Usage = D3D11_USAGE_DEFAULT;
			depth_desc.BindFlags = D3D11_BIND_DEPTH_STENCIL;
			depth_desc.CPUAccessFlags = 0;
			depth_desc.MiscFlags = 0;

			ID3D11Texture2D* depth_stencil_texture = nullptr;
			result = p_device->CreateTexture2D(&depth_desc, nullptr, &depth_stencil_texture);
			RMLUI_DX_ASSERTMSG(result, "failed to CreateTexture2D");
			if (FAILED(result))
			{
				Rml::Log::Message(Rml::Log::LT_ERROR, "ID3D11Device::CreateTexture2D (%d)", result);
				return false;
			}

			D3D11_DEPTH_STENCIL_VIEW_DESC dsvDesc = {};
			dsvDesc.Format = depth_desc.Format;
			dsvDesc.ViewDimension = (samples > 0) ? D3D11_DSV_DIMENSION_TEXTURE2DMS : D3D11_DSV_DIMENSION_TEXTURE2D;
			dsvDesc.Texture2D.MipSlice = 0;

			result = p_device->CreateDepthStencilView(depth_stencil_texture, &dsvDesc, &depth_stencil_view);
			RMLUI_DX_ASSERTMSG(result, "failed to CreateDepthStencilView");
			if (FAILED(result))
			{
				depth_stencil_texture->Release();
				Rml::Log::Message(Rml::Log::LT_ERROR, "ID3D11Device::CreateDepthStencilView (%d)", result);
				return false;
			}

			depth_stencil_texture->Release();
		}
	}

	out_rt = {};
	out_rt.width = width;
	out_rt.height = height;
	out_rt.render_target_view = render_target_view;
	out_rt.render_target_texture = rt_texture;
	out_rt.render_target_shader_resource_view = render_target_shader_resource_view;
	out_rt.depth_stencil_view = depth_stencil_view;
	out_rt.owns_depth_stencil_buffer = shared_depth_stencil_buffer != nullptr;

	return true;
}

static void DestroyRenderTarget(RenderTargetData& rt)
{
	Dx11CleanupResourceIfCreated(rt.render_target_view);
	Dx11CleanupResourceIfCreated(rt.render_target_texture);
	Dx11CleanupResourceIfCreated(rt.render_target_shader_resource_view);
	Dx11CleanupResourceIfCreated(rt.depth_stencil_view);
	rt = {};
}

static inline void BindTexture(ID3D11DeviceContext* context, const RenderTargetData& rt, uint32_t slot = 0)
{
	context->PSSetShaderResources(slot, 1, &rt.render_target_shader_resource_view);
}

static inline void BindRenderTarget(ID3D11DeviceContext* context, const RenderTargetData& rt)
{
	context->OMSetRenderTargets(1, &rt.render_target_view, rt.depth_stencil_view);
}

static bool CreateShaders(ID3D11Device* p_device, ProgramData& data)
{
	RMLUI_ASSERT(std::all_of(data.vert_shaders.begin(), data.vert_shaders.end(), [](auto&& value) { return value == nullptr; }));
	RMLUI_ASSERT(std::all_of(data.frag_shaders.begin(), data.frag_shaders.end(), [](auto&& value) { return value == nullptr; }));
	RMLUI_ASSERT(std::all_of(data.programs.begin(), data.programs.end(),
		[](const ShaderProgram& value) { return value.vertex_shader == nullptr || value.pixel_shader == nullptr; }));
	auto ReportError = [](const char* type, const char* name) {
		Rml::Log::Message(Rml::Log::LT_ERROR, "Could not create shader %s: '%s'.", type, name);
		return false;
	};

	for (const VertShaderDefinition& def : vert_shader_definitions)
	{
		if (!CreateShader(data.vert_shaders[def.id], ShaderType::Vertex, def.code_str, def.code_size))
			return ReportError("vertex shader", def.name_str);
	}

	for (const FragShaderDefinition& def : frag_shader_definitions)
	{
		if (!CreateShader(data.frag_shaders[def.id], ShaderType::Fragment, def.code_str, def.code_size))
			return ReportError("fragment shader", def.name_str);
	}

	for (const ProgramDefinition& def : program_definitions)
	{
		if (!CreateProgram(p_device, data.programs[def.id], data.vert_shaders[def.vert_shader], data.frag_shaders[def.frag_shader]))
			return ReportError("program", def.name_str);
	}

	return true;
}

static void DestroyShaders(const ProgramData& data)
{
	for (ShaderProgram programs : data.programs)
	{
		Dx11CleanupResourceIfCreated(programs.vertex_shader);
		Dx11CleanupResourceIfCreated(programs.pixel_shader);
	}

	for (ID3DBlob* shader_blob : data.vert_shaders)
	{
		Dx11CleanupResourceIfCreated(shader_blob);
	}

	for (ID3DBlob* shader_blob : data.frag_shaders)
	{
		Dx11CleanupResourceIfCreated(shader_blob);
	}
}

} // namespace Gfx

RenderInterface_DX11::RenderInterface_DX11(ID3D11Device* p_d3d_device)
{
	RMLUI_ASSERTMSG(p_d3d_device, "p_d3d_device cannot be nullptr!");

	// Assign D3D resources
	m_d3d_device = p_d3d_device;
	m_d3d_device->AddRef();
	m_d3d_device->QueryInterface(__uuidof(ID3D11Device1), (void**)&m_d3d_device_1);
	m_d3d_device_1->GetImmediateContext1(&m_d3d_context);
	m_render_layers.SetD3DResources(m_d3d_device);

	// Pre-cache quad for blitting
	Rml::Mesh mesh;
	Rml::MeshUtilities::GenerateQuad(mesh, Rml::Vector2f(-1), Rml::Vector2f(2), {});
	m_fullscreen_quad_geometry = RenderInterface_DX11::CompileGeometry(mesh.vertices, mesh.indices);

	if (!m_blend_state_enable)
	{
		// RmlUi serves vertex colors and textures with premultiplied alpha, set the blend mode accordingly.
		// Equivalent to glBlendFunc(GL_ONE, GL_ONE_MINUS_SRC_ALPHA).
		D3D11_BLEND_DESC blend_desc{};
		ZeroMemory(&blend_desc, sizeof(blend_desc));
		blend_desc.AlphaToCoverageEnable = false;
		blend_desc.IndependentBlendEnable = false;
		blend_desc.RenderTarget[0].BlendEnable = true;
		blend_desc.RenderTarget[0].SrcBlend = D3D11_BLEND_ONE;
		blend_desc.RenderTarget[0].DestBlend = D3D11_BLEND_INV_SRC_ALPHA;
		blend_desc.RenderTarget[0].BlendOp = D3D11_BLEND_OP_ADD;
		blend_desc.RenderTarget[0].SrcBlendAlpha = D3D11_BLEND_ONE;
		blend_desc.RenderTarget[0].DestBlendAlpha = D3D11_BLEND_INV_SRC_ALPHA;
		blend_desc.RenderTarget[0].BlendOpAlpha = D3D11_BLEND_OP_ADD;
		blend_desc.RenderTarget[0].RenderTargetWriteMask = D3D11_COLOR_WRITE_ENABLE_ALL;
		HRESULT result = m_d3d_device->CreateBlendState(&blend_desc, &m_blend_state_enable);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateBlendState");
#ifdef RMLUI_DX_DEBUG
		if (FAILED(result))
		{
			Rml::Log::Message(Rml::Log::LT_ERROR, "ID3D11Device::CreateBlendState (%d)", result);
			return;
		}
#endif

		ZeroMemory(&blend_desc, sizeof(blend_desc));
		blend_desc.RenderTarget[0].RenderTargetWriteMask = 0;
		result = m_d3d_device->CreateBlendState(&blend_desc, &m_blend_state_disable_color);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateBlendState");
#ifdef RMLUI_DX_DEBUG
		if (FAILED(result))
		{
			Rml::Log::Message(Rml::Log::LT_ERROR, "ID3D11Device::CreateBlendState (%d)", result);
			return;
		}
#endif

		blend_desc.RenderTarget[0].BlendEnable = true;
		blend_desc.RenderTarget[0].SrcBlend = D3D11_BLEND_BLEND_FACTOR;
		blend_desc.RenderTarget[0].DestBlend = D3D11_BLEND_ZERO;
		blend_desc.RenderTarget[0].BlendOp = D3D11_BLEND_OP_ADD;
		blend_desc.RenderTarget[0].SrcBlendAlpha = D3D11_BLEND_BLEND_FACTOR;
		blend_desc.RenderTarget[0].DestBlendAlpha = D3D11_BLEND_ZERO;
		blend_desc.RenderTarget[0].BlendOpAlpha = D3D11_BLEND_OP_ADD;
		blend_desc.RenderTarget[0].RenderTargetWriteMask = D3D11_COLOR_WRITE_ENABLE_ALL;
		result = m_d3d_device->CreateBlendState(&blend_desc, &m_blend_state_color_filter);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateBlendState");
#ifdef RMLUI_DX_DEBUG
		if (FAILED(result))
		{
			Rml::Log::Message(Rml::Log::LT_ERROR, "ID3D11Device::CreateBlendState (%d)", result);
			return;
		}
#endif
	}

	// Scissor regions require a rasterizer state. Cache one for scissor on and off
	{
		D3D11_RASTERIZER_DESC rasterizerDesc{};
		rasterizerDesc.FillMode = D3D11_FILL_SOLID;
		rasterizerDesc.CullMode = D3D11_CULL_NONE;
		rasterizerDesc.FrontCounterClockwise = false;
		rasterizerDesc.DepthBias = D3D11_DEFAULT_DEPTH_BIAS;
		rasterizerDesc.SlopeScaledDepthBias = D3D11_DEFAULT_SLOPE_SCALED_DEPTH_BIAS;
		rasterizerDesc.DepthBiasClamp = D3D11_DEFAULT_DEPTH_BIAS_CLAMP;
		rasterizerDesc.DepthClipEnable = false;
		rasterizerDesc.ScissorEnable = true;
		rasterizerDesc.MultisampleEnable = NUM_MSAA_SAMPLES > 1;
		rasterizerDesc.AntialiasedLineEnable = NUM_MSAA_SAMPLES > 1;

		HRESULT result = m_d3d_device->CreateRasterizerState(&rasterizerDesc, &m_rasterizer_state_scissor_enabled);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateRasterizerState");
#ifdef RMLUI_DX_DEBUG
		if (FAILED(result))
		{
			Rml::Log::Message(Rml::Log::LT_ERROR, "ID3D11Device::CreateRasterizerState (scissor: enabled) (%d)", result);
			return;
		}
#endif

		rasterizerDesc.ScissorEnable = false;

		result = m_d3d_device->CreateRasterizerState(&rasterizerDesc, &m_rasterizer_state_scissor_disabled);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateRasterizerState");
#ifdef RMLUI_DX_DEBUG
		if (FAILED(result))
		{
			Rml::Log::Message(Rml::Log::LT_ERROR, "ID3D11Device::CreateRasterizerState (scissor: disabled) (%d)", result);
			return;
		}
#endif
	}

	// Compile and load shaders
	auto mut_program_data = Rml::MakeUnique<Gfx::ProgramData>();
	if (Gfx::CreateShaders(m_d3d_device, *mut_program_data))
	{
		program_data = std::move(mut_program_data);
	}

	// Create vertex layout. This will be constant to avoid copying to an intermediate struct.
	{
		D3D11_INPUT_ELEMENT_DESC polygonLayout[] = {
			{"POSITION", 0, DXGI_FORMAT_R32G32_FLOAT, 0, D3D11_APPEND_ALIGNED_ELEMENT, D3D11_INPUT_PER_VERTEX_DATA, 0},
			{"COLOR", 0, DXGI_FORMAT_R8G8B8A8_UNORM, 0, D3D11_APPEND_ALIGNED_ELEMENT, D3D11_INPUT_PER_VERTEX_DATA, 0},
			{"TEXCOORD", 0, DXGI_FORMAT_R32G32_FLOAT, 0, D3D11_APPEND_ALIGNED_ELEMENT, D3D11_INPUT_PER_VERTEX_DATA, 0},
		};
		uint32_t numElements = sizeof(polygonLayout) / sizeof(polygonLayout[0]);

		HRESULT result =
			m_d3d_device->CreateInputLayout(polygonLayout, numElements, program_data->vert_shaders[VertShaderId::Main]->GetBufferPointer(),
				program_data->vert_shaders[VertShaderId::Main]->GetBufferSize(), &m_vertex_layout);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateInputLayout");
		if (FAILED(result))
		{
			return;
		}
	}

	// Create constant buffers. This is so that we can bind uniforms such as translation and color to the shaders.
	{
		D3D11_BUFFER_DESC cbufferDesc{};
		cbufferDesc.Usage = D3D11_USAGE_DYNAMIC;
		cbufferDesc.ByteWidth = sizeof(ShaderCbuffer);
		cbufferDesc.BindFlags = D3D11_BIND_CONSTANT_BUFFER;
		cbufferDesc.CPUAccessFlags = D3D11_CPU_ACCESS_WRITE;
		cbufferDesc.MiscFlags = 0;
		cbufferDesc.StructureByteStride = 0;

		HRESULT result = m_d3d_device->CreateBuffer(&cbufferDesc, nullptr, &m_shader_buffer);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateBuffer");
		if (FAILED(result))
		{
			return;
		}

		cbufferDesc.Usage = D3D11_USAGE_DYNAMIC;
		cbufferDesc.ByteWidth = sizeof(ColorMatrixCbuffer);
		cbufferDesc.BindFlags = D3D11_BIND_CONSTANT_BUFFER;
		cbufferDesc.CPUAccessFlags = D3D11_CPU_ACCESS_WRITE;
		cbufferDesc.MiscFlags = 0;
		cbufferDesc.StructureByteStride = 0;

		result = m_d3d_device->CreateBuffer(&cbufferDesc, nullptr, &m_color_matrix_cbuffer);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateBuffer");
		if (FAILED(result))
		{
			return;
		}

		cbufferDesc.Usage = D3D11_USAGE_DYNAMIC;
		cbufferDesc.ByteWidth = sizeof(ShaderCbuffer::Blur);
		cbufferDesc.BindFlags = D3D11_BIND_CONSTANT_BUFFER;
		cbufferDesc.CPUAccessFlags = D3D11_CPU_ACCESS_WRITE;
		cbufferDesc.MiscFlags = 0;
		cbufferDesc.StructureByteStride = 0;

		result = m_d3d_device->CreateBuffer(&cbufferDesc, nullptr, &m_blur_cbuffer);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateBuffer");
		if (FAILED(result))
		{
			return;
		}

		cbufferDesc.Usage = D3D11_USAGE_DYNAMIC;
		cbufferDesc.ByteWidth = sizeof(ShaderCbuffer::DropShadow);
		cbufferDesc.BindFlags = D3D11_BIND_CONSTANT_BUFFER;
		cbufferDesc.CPUAccessFlags = D3D11_CPU_ACCESS_WRITE;
		cbufferDesc.MiscFlags = 0;
		cbufferDesc.StructureByteStride = 0;

		result = m_d3d_device->CreateBuffer(&cbufferDesc, nullptr, &m_drop_shadow_cbuffer);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateBuffer");
		if (FAILED(result))
		{
			return;
		}
	}

	// Create sampler state for textures
	{
		D3D11_SAMPLER_DESC samplerDesc{};
		samplerDesc.Filter = D3D11_FILTER_MIN_MAG_MIP_LINEAR;
		samplerDesc.AddressU = D3D11_TEXTURE_ADDRESS_WRAP;
		samplerDesc.AddressV = D3D11_TEXTURE_ADDRESS_WRAP;
		samplerDesc.AddressW = D3D11_TEXTURE_ADDRESS_WRAP;
		samplerDesc.MipLODBias = 0.0f;
		samplerDesc.MaxAnisotropy = 1;
		samplerDesc.ComparisonFunc = D3D11_COMPARISON_ALWAYS;
		samplerDesc.BorderColor[0] = 0;
		samplerDesc.BorderColor[1] = 0;
		samplerDesc.BorderColor[2] = 0;
		samplerDesc.BorderColor[3] = 0;
		samplerDesc.MinLOD = 0;
		samplerDesc.MaxLOD = D3D11_FLOAT32_MAX;

		HRESULT result = m_d3d_device->CreateSamplerState(&samplerDesc, &m_sampler_state);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateSamplerState");
		if (FAILED(result))
		{
			return;
		}
	}

	// Create depth stencil states
	{
		D3D11_DEPTH_STENCIL_DESC desc;
		ZeroMemory(&desc, sizeof(desc));
		desc.DepthEnable = false;
		desc.StencilReadMask = D3D11_DEFAULT_STENCIL_READ_MASK;
		desc.StencilWriteMask = 0;
		desc.FrontFace.StencilFunc = D3D11_COMPARISON_ALWAYS;
		desc.FrontFace.StencilPassOp = D3D11_STENCIL_OP_KEEP;
		desc.FrontFace.StencilFailOp = D3D11_STENCIL_OP_KEEP;
		desc.FrontFace.StencilDepthFailOp = D3D11_STENCIL_OP_KEEP;
		desc.BackFace = desc.FrontFace;
		desc.StencilEnable = true;
		// Disabled
		m_d3d_device->CreateDepthStencilState(&desc, &m_depth_stencil_state_disable);

		desc.StencilEnable = true;
		desc.StencilReadMask = D3D11_DEFAULT_STENCIL_READ_MASK;
		desc.StencilWriteMask = D3D11_DEFAULT_STENCIL_WRITE_MASK;
		desc.FrontFace.StencilFunc = D3D11_COMPARISON_ALWAYS;
		desc.FrontFace.StencilPassOp = D3D11_STENCIL_OP_REPLACE;
		desc.FrontFace.StencilFailOp = D3D11_STENCIL_OP_KEEP;
		desc.FrontFace.StencilDepthFailOp = D3D11_STENCIL_OP_KEEP;
		desc.BackFace = desc.FrontFace;
		// Set and SetInverse
		m_d3d_device->CreateDepthStencilState(&desc, &m_depth_stencil_state_stencil_set);

		desc.FrontFace.StencilFunc = D3D11_COMPARISON_ALWAYS;
		desc.FrontFace.StencilPassOp = D3D11_STENCIL_OP_INCR;
		desc.FrontFace.StencilFailOp = D3D11_STENCIL_OP_KEEP;
		desc.FrontFace.StencilDepthFailOp = D3D11_STENCIL_OP_KEEP;
		desc.BackFace = desc.FrontFace;
		// Intersect
		m_d3d_device->CreateDepthStencilState(&desc, &m_depth_stencil_state_stencil_intersect);

		desc.StencilWriteMask = 0;
		desc.FrontFace.StencilFunc = D3D11_COMPARISON_EQUAL;
		desc.FrontFace.StencilPassOp = D3D11_STENCIL_OP_KEEP;
		desc.FrontFace.StencilFailOp = D3D11_STENCIL_OP_KEEP;
		desc.FrontFace.StencilDepthFailOp = D3D11_STENCIL_OP_KEEP;
		desc.BackFace = desc.FrontFace;
		// Stencil test
		m_d3d_device->CreateDepthStencilState(&desc, &m_depth_stencil_state_stencil_test);
	}
}

RenderInterface_DX11::~RenderInterface_DX11()
{
	if (m_fullscreen_quad_geometry)
	{
		RenderInterface_DX11::ReleaseGeometry(m_fullscreen_quad_geometry);
		m_fullscreen_quad_geometry = {};
	}

	if (program_data)
	{
		Gfx::DestroyShaders(*program_data);
		program_data.reset();
	}

	// Cleans up all general resources
	Dx11CleanupResourceIfCreated(m_sampler_state);
	Dx11CleanupResourceIfCreated(m_blend_state_enable);
	Dx11CleanupResourceIfCreated(m_blend_state_color_filter);
	Dx11CleanupResourceIfCreated(m_blend_state_disable_color);
	Dx11CleanupResourceIfCreated(m_depth_stencil_state_disable);
	Dx11CleanupResourceIfCreated(m_depth_stencil_state_stencil_intersect);
	Dx11CleanupResourceIfCreated(m_depth_stencil_state_stencil_set);
	Dx11CleanupResourceIfCreated(m_depth_stencil_state_stencil_test);
	Dx11CleanupResourceIfCreated(m_rasterizer_state_scissor_disabled);
	Dx11CleanupResourceIfCreated(m_rasterizer_state_scissor_enabled);
	Dx11CleanupResourceIfCreated(m_shader_buffer);
	Dx11CleanupResourceIfCreated(m_color_matrix_cbuffer);
	Dx11CleanupResourceIfCreated(m_blur_cbuffer);
	Dx11CleanupResourceIfCreated(m_drop_shadow_cbuffer);
	Dx11CleanupResourceIfCreated(m_vertex_layout);
	Dx11CleanupResourceIfCreated(m_d3d_context);
	Dx11CleanupResourceIfCreated(m_d3d_device_1);
	Dx11CleanupResourceIfCreated(m_d3d_device);
}

void RenderInterface_DX11::BeginFrame()
{
	RMLUI_ASSERT(m_viewport_width >= 1 && m_viewport_height >= 1);
	RMLUI_ASSERTMSG(m_d3d_context, "d3d_context cannot be nullptr!");
	RMLUI_ASSERTMSG(m_d3d_device, "d3d_device cannot be nullptr!");

	// Backup DX11 state
	{
		m_previous_d3d_state.scissor_rects_count = m_previous_d3d_state.viewports_count = D3D11_VIEWPORT_AND_SCISSORRECT_OBJECT_COUNT_PER_PIPELINE;
		m_d3d_context->RSGetScissorRects(&m_previous_d3d_state.scissor_rects_count, m_previous_d3d_state.scissor_rects);
		m_d3d_context->RSGetViewports(&m_previous_d3d_state.viewports_count, m_previous_d3d_state.viewports);
		m_d3d_context->RSGetState(&m_previous_d3d_state.rasterizer_state);
		m_d3d_context->OMGetBlendState(&m_previous_d3d_state.blend_state, m_previous_d3d_state.blend_factor, &m_previous_d3d_state.sample_mask);
		m_d3d_context->OMGetDepthStencilState(&m_previous_d3d_state.depth_stencil_state, &m_previous_d3d_state.stencil_ref);
		m_d3d_context->PSGetShaderResources(0, 1, &m_previous_d3d_state.pixel_shader_shader_resource);
		m_d3d_context->PSGetSamplers(0, 1, &m_previous_d3d_state.pixel_shader_sampler);
		m_previous_d3d_state.pixel_shader_instances_count = m_previous_d3d_state.vertex_shader_instances_count =
			m_previous_d3d_state.geometry_shader_instances_count = 256;
		m_d3d_context->PSGetShader(&m_previous_d3d_state.pixel_shader, m_previous_d3d_state.pixel_shader_instances,
			&m_previous_d3d_state.pixel_shader_instances_count);
		m_d3d_context->VSGetShader(&m_previous_d3d_state.vertex_shader, m_previous_d3d_state.vertex_shader_instances,
			&m_previous_d3d_state.vertex_shader_instances_count);
		m_d3d_context->VSGetConstantBuffers(0, 1, &m_previous_d3d_state.vertex_shader_constant_buffer);
		m_d3d_context->GSGetShader(&m_previous_d3d_state.geometry_shader, m_previous_d3d_state.geometry_shader_instances,
			&m_previous_d3d_state.geometry_shader_instances_count);

		m_d3d_context->IAGetPrimitiveTopology(&m_previous_d3d_state.primitive_topology);
		m_d3d_context->IAGetIndexBuffer(&m_previous_d3d_state.index_buffer, &m_previous_d3d_state.index_buffer_format,
			&m_previous_d3d_state.index_buffer_offset);
		m_d3d_context->IAGetVertexBuffers(0, 1, &m_previous_d3d_state.vertex_buffer, &m_previous_d3d_state.vertex_buffer_stride,
			&m_previous_d3d_state.vertex_buffer_offset);
		m_d3d_context->IAGetInputLayout(&m_previous_d3d_state.input_layout);
	}

	// Initialise DX11 state for RmlUi
	D3D11_VIEWPORT d3dviewport;
	d3dviewport.TopLeftX = 0;
	d3dviewport.TopLeftY = 0;
	d3dviewport.Width = (float)m_viewport_width;
	d3dviewport.Height = (float)m_viewport_height;
	d3dviewport.MinDepth = 0.0f;
	d3dviewport.MaxDepth = 1.0f;
	m_d3d_context->RSSetViewports(1, &d3dviewport);

	SetBlendState(m_blend_state_enable);
	m_d3d_context->RSSetState(m_rasterizer_state_scissor_disabled); // Disable scissor
	m_stencil_ref_value = 0;                                        // Reset stencil
	m_d3d_context->OMSetDepthStencilState(m_depth_stencil_state_disable, 0);

	SetTransform(nullptr);

	m_render_layers.BeginFrame(m_viewport_width, m_viewport_height);
	Gfx::BindRenderTarget(m_d3d_context, m_render_layers.GetTopLayer());
	float clearColor[4] = {0.0f, 0.0f, 0.0f, 0.0f};
	m_d3d_context->ClearRenderTargetView(m_render_layers.GetTopLayer().render_target_view, clearColor);

	UseProgram(ProgramId::None);
	m_scissor_state = Rml::Rectanglei::MakeInvalid();
}

void RenderInterface_DX11::EndFrame(ID3D11RenderTargetView* p_render_target_view)
{
	RMLUI_ASSERTMSG(p_render_target_view, "p_render_target_view cannot be nullptr!");

	const Gfx::RenderTargetData& rt_active = m_render_layers.GetTopLayer();
	const Gfx::RenderTargetData& rt_postprocess = m_render_layers.GetPostprocessPrimary();

	// Resolve MSAA to the post-process framebuffer.
	m_d3d_context->ResolveSubresource(rt_postprocess.render_target_texture, 0, rt_active.render_target_texture, 0, DXGI_FORMAT_R8G8B8A8_UNORM);

	// Draw to bound_render_target (usually the swapchain RTV)
	m_d3d_context->OMSetRenderTargets(1, &p_render_target_view, nullptr);

	// Assuming we have an opaque background, we can just write to it with the premultiplied alpha blend mode and we'll get the correct result.
	// Instead, if we had a transparent destination that didn't use premultiplied alpha, we would need to perform a manual un-premultiplication step.
	Gfx::BindTexture(m_d3d_context, rt_postprocess);

	UseProgram(ProgramId::Passthrough);

	DrawFullscreenQuad();

	m_render_layers.EndFrame();

	// Reset internal state
	m_current_blend_state = nullptr;

	// Restore modified DX state
	{
		m_d3d_context->RSSetScissorRects(m_previous_d3d_state.scissor_rects_count, m_previous_d3d_state.scissor_rects);
		m_d3d_context->RSSetViewports(m_previous_d3d_state.viewports_count, m_previous_d3d_state.viewports);
		m_d3d_context->RSSetState(m_previous_d3d_state.rasterizer_state);
		if (m_previous_d3d_state.rasterizer_state)
		{
			m_previous_d3d_state.rasterizer_state->Release();
		}
		m_d3d_context->OMSetBlendState(m_previous_d3d_state.blend_state, m_previous_d3d_state.blend_factor, m_previous_d3d_state.sample_mask);
		if (m_previous_d3d_state.blend_state)
		{
			m_previous_d3d_state.blend_state->Release();
		}
		m_d3d_context->OMSetDepthStencilState(m_previous_d3d_state.depth_stencil_state, m_previous_d3d_state.stencil_ref);
		if (m_previous_d3d_state.depth_stencil_state)
		{
			m_previous_d3d_state.depth_stencil_state->Release();
		}
		m_d3d_context->PSSetShaderResources(0, 1, &m_previous_d3d_state.pixel_shader_shader_resource);
		if (m_previous_d3d_state.pixel_shader_shader_resource)
		{
			m_previous_d3d_state.pixel_shader_shader_resource->Release();
		}
		m_d3d_context->PSSetSamplers(0, 1, &m_previous_d3d_state.pixel_shader_sampler);
		if (m_previous_d3d_state.pixel_shader_sampler)
		{
			m_previous_d3d_state.pixel_shader_sampler->Release();
		}
		m_d3d_context->PSSetShader(m_previous_d3d_state.pixel_shader, m_previous_d3d_state.pixel_shader_instances,
			m_previous_d3d_state.pixel_shader_instances_count);
		if (m_previous_d3d_state.pixel_shader)
		{
			m_previous_d3d_state.pixel_shader->Release();
		}
		for (UINT i = 0; i < m_previous_d3d_state.pixel_shader_instances_count; i++)
		{
			if (m_previous_d3d_state.pixel_shader_instances[i])
			{
				m_previous_d3d_state.pixel_shader_instances[i]->Release();
			}
		}
		m_d3d_context->VSSetShader(m_previous_d3d_state.vertex_shader, m_previous_d3d_state.vertex_shader_instances,
			m_previous_d3d_state.vertex_shader_instances_count);
		if (m_previous_d3d_state.vertex_shader)
		{
			m_previous_d3d_state.vertex_shader->Release();
		}
		m_d3d_context->VSSetConstantBuffers(0, 1, &m_previous_d3d_state.vertex_shader_constant_buffer);
		if (m_previous_d3d_state.vertex_shader_constant_buffer)
		{
			m_previous_d3d_state.vertex_shader_constant_buffer->Release();
		}
		m_d3d_context->GSSetShader(m_previous_d3d_state.geometry_shader, m_previous_d3d_state.geometry_shader_instances,
			m_previous_d3d_state.geometry_shader_instances_count);
		if (m_previous_d3d_state.geometry_shader)
		{
			m_previous_d3d_state.geometry_shader->Release();
		}
		for (UINT i = 0; i < m_previous_d3d_state.vertex_shader_instances_count; i++)
		{
			if (m_previous_d3d_state.vertex_shader_instances[i])
			{
				m_previous_d3d_state.vertex_shader_instances[i]->Release();
			}
		}
		m_d3d_context->IASetPrimitiveTopology(m_previous_d3d_state.primitive_topology);
		m_d3d_context->IASetIndexBuffer(m_previous_d3d_state.index_buffer, m_previous_d3d_state.index_buffer_format,
			m_previous_d3d_state.index_buffer_offset);
		if (m_previous_d3d_state.index_buffer)
		{
			m_previous_d3d_state.index_buffer->Release();
		}
		m_d3d_context->IASetVertexBuffers(0, 1, &m_previous_d3d_state.vertex_buffer, &m_previous_d3d_state.vertex_buffer_stride,
			&m_previous_d3d_state.vertex_buffer_offset);
		if (m_previous_d3d_state.vertex_buffer)
		{
			m_previous_d3d_state.vertex_buffer->Release();
		}
		m_d3d_context->IASetInputLayout(m_previous_d3d_state.input_layout);
		if (m_previous_d3d_state.input_layout)
		{
			m_previous_d3d_state.input_layout->Release();
		}
	}
}

Rml::CompiledGeometryHandle RenderInterface_DX11::CompileGeometry(Rml::Span<const Rml::Vertex> vertices, Rml::Span<const int> indices)
{
	ID3D11Buffer* vertex_buffer = nullptr;
	ID3D11Buffer* index_buffer = nullptr;

	// Bind vertex buffer
	{
		D3D11_BUFFER_DESC vertexBufferDesc{};
		vertexBufferDesc.Usage = D3D11_USAGE_DEFAULT;
		vertexBufferDesc.ByteWidth = static_cast<UINT>(sizeof(Rml::Vertex) * vertices.size());
		vertexBufferDesc.BindFlags = D3D11_BIND_VERTEX_BUFFER;
		vertexBufferDesc.CPUAccessFlags = 0;
		vertexBufferDesc.MiscFlags = 0;
		vertexBufferDesc.StructureByteStride = 0;

		D3D11_SUBRESOURCE_DATA vertexData{};
		vertexData.pSysMem = vertices.data();
		vertexData.SysMemPitch = 0;
		vertexData.SysMemSlicePitch = 0;

		HRESULT result = m_d3d_device->CreateBuffer(&vertexBufferDesc, &vertexData, &vertex_buffer);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateBuffer");
		if (FAILED(result))
		{
			return Rml::CompiledGeometryHandle(0);
		}
	}

	// Bind index buffer
	{
		D3D11_BUFFER_DESC indexBufferDesc{};
		indexBufferDesc.Usage = D3D11_USAGE_DEFAULT;
		indexBufferDesc.ByteWidth = static_cast<UINT>(sizeof(int) * indices.size());
		indexBufferDesc.BindFlags = D3D11_BIND_INDEX_BUFFER;
		indexBufferDesc.CPUAccessFlags = 0;
		indexBufferDesc.MiscFlags = 0;
		indexBufferDesc.StructureByteStride = 0;

		D3D11_SUBRESOURCE_DATA indexData{};
		indexData.pSysMem = indices.data();
		indexData.SysMemPitch = 0;
		indexData.SysMemSlicePitch = 0;

		HRESULT result = m_d3d_device->CreateBuffer(&indexBufferDesc, &indexData, &index_buffer);
		RMLUI_DX_ASSERTMSG(result, "failed to CreateBuffer");
		if (FAILED(result))
		{
			Dx11CleanupResourceIfCreated(vertex_buffer);
			return Rml::CompiledGeometryHandle(0);
		}
	}

	DX11_GeometryData* geometryData = new DX11_GeometryData;

	geometryData->vertex_buffer = vertex_buffer;
	geometryData->index_buffer = index_buffer;
	geometryData->index_count = indices.size();

	return (Rml::CompiledGeometryHandle)geometryData;
}

void RenderInterface_DX11::ReleaseGeometry(Rml::CompiledGeometryHandle handle)
{
	DX11_GeometryData* geometry = (DX11_GeometryData*)handle;

	Dx11CleanupResourceIfCreated(geometry->vertex_buffer);
	Dx11CleanupResourceIfCreated(geometry->index_buffer);

	delete geometry;
}

void RenderInterface_DX11::RenderGeometry(Rml::CompiledGeometryHandle handle, Rml::Vector2f translation, Rml::TextureHandle texture)
{
	DX11_GeometryData* geometry = (DX11_GeometryData*)handle;

	if (m_translation != translation)
	{
		m_translation = translation;
		m_cbuffer_dirty = true;
	}

	if (texture == TexturePostprocess || texture == TexturePostprocessNoBinding)
	{
		// Do nothing.
	}
	else if (texture)
	{
		// Texture available
		UseProgram(ProgramId::Texture);
		if (texture != TextureEnableWithoutBinding)
		{
			ID3D11ShaderResourceView* texture_view = (ID3D11ShaderResourceView*)texture;
			m_d3d_context->PSSetShaderResources(0, 1, &texture_view);
		}
		m_d3d_context->PSSetSamplers(0, 1, &m_sampler_state);
	}
	else
	{
		// No texture, use color
		UseProgram(ProgramId::Color);
	}

	if (texture != TexturePostprocessNoBinding)
	{
		UpdateConstantBuffer();

		m_d3d_context->VSSetConstantBuffers(0, 1, &m_shader_buffer);
		m_d3d_context->PSSetConstantBuffers(0, 1, &m_shader_buffer);
	}

	// Bind vertex and index buffers, issue draw call
	uint32_t stride = sizeof(Rml::Vertex);
	uint32_t offset = 0;
	m_d3d_context->IASetInputLayout(m_vertex_layout);
	m_d3d_context->IASetVertexBuffers(0, 1, &geometry->vertex_buffer, &stride, &offset);
	m_d3d_context->IASetIndexBuffer(geometry->index_buffer, DXGI_FORMAT_R32_UINT, 0);
	m_d3d_context->IASetPrimitiveTopology(D3D11_PRIMITIVE_TOPOLOGY_TRIANGLELIST);
	m_d3d_context->DrawIndexed(static_cast<UINT>(geometry->index_count), 0, 0);

	// Unbind resources
	ID3D11ShaderResourceView* const nullSRV = nullptr;
	m_d3d_context->PSSetShaderResources(0, 1, &nullSRV);
}

// Set to byte packing, or the compiler will expand our struct, which means it won't read correctly from file
#pragma pack(1)
struct TGAHeader {
	char idLength;
	char colourMapType;
	char dataType;
	short int colourMapOrigin;
	short int colourMapLength;
	char colourMapDepth;
	short int xOrigin;
	short int yOrigin;
	short int width;
	short int height;
	char bitsPerPixel;
	char imageDescriptor;
};
// Restore packing
#pragma pack()

Rml::TextureHandle RenderInterface_DX11::LoadTexture(Rml::Vector2i& texture_dimensions, const Rml::String& source)
{
	Rml::FileInterface* file_interface = Rml::GetFileInterface();
	Rml::FileHandle file_handle = file_interface->Open(source);
	if (!file_handle)
	{
		return false;
	}

	file_interface->Seek(file_handle, 0, SEEK_END);
	size_t buffer_size = file_interface->Tell(file_handle);
	file_interface->Seek(file_handle, 0, SEEK_SET);

	if (buffer_size <= sizeof(TGAHeader))
	{
		Rml::Log::Message(Rml::Log::LT_ERROR, "Texture file size is smaller than TGAHeader, file is not a valid TGA image.");
		file_interface->Close(file_handle);
		return false;
	}

	using Rml::byte;
	Rml::UniquePtr<byte[]> buffer(new byte[buffer_size]);
	file_interface->Read(buffer.get(), buffer_size, file_handle);
	file_interface->Close(file_handle);

	TGAHeader header;
	memcpy(&header, buffer.get(), sizeof(TGAHeader));

	int color_mode = header.bitsPerPixel / 8;
	const size_t image_size = header.width * header.height * 4; // We always make 32bit textures

	if (header.dataType != 2)
	{
		Rml::Log::Message(Rml::Log::LT_ERROR, "Only 24/32bit uncompressed TGAs are supported.");
		return false;
	}

	// Ensure we have at least 3 colors
	if (color_mode < 3)
	{
		Rml::Log::Message(Rml::Log::LT_ERROR, "Only 24 and 32bit textures are supported.");
		return false;
	}

	const byte* image_src = buffer.get() + sizeof(TGAHeader);
	Rml::UniquePtr<byte[]> image_dest_buffer(new byte[image_size]);
	byte* image_dest = image_dest_buffer.get();

	// Targa is BGR, swap to RGB, flip Y axis, and convert to premultiplied alpha.
	for (long y = 0; y < header.height; y++)
	{
		long read_index = y * header.width * color_mode;
		long write_index = ((header.imageDescriptor & 32) != 0) ? read_index : (header.height - y - 1) * header.width * 4;
		for (long x = 0; x < header.width; x++)
		{
			image_dest[write_index] = image_src[read_index + 2];
			image_dest[write_index + 1] = image_src[read_index + 1];
			image_dest[write_index + 2] = image_src[read_index];
			if (color_mode == 4)
			{
				const byte alpha = image_src[read_index + 3];
				for (size_t j = 0; j < 3; j++)
					image_dest[write_index + j] = byte((image_dest[write_index + j] * alpha) / 255);
				image_dest[write_index + 3] = alpha;
			}
			else
				image_dest[write_index + 3] = 255;

			write_index += 4;
			read_index += color_mode;
		}
	}

	texture_dimensions.x = header.width;
	texture_dimensions.y = header.height;

	return GenerateTexture({image_dest, image_size}, texture_dimensions);
}

Rml::TextureHandle RenderInterface_DX11::GenerateTexture(Rml::Span<const Rml::byte> source, Rml::Vector2i source_dimensions)
{
	ID3D11Texture2D* gpu_texture = nullptr;
	ID3D11ShaderResourceView* gpu_texture_view = nullptr;

	D3D11_TEXTURE2D_DESC textureDesc{};
	textureDesc.Width = source_dimensions.x;
	textureDesc.Height = source_dimensions.y;
	textureDesc.MipLevels = 0;
	textureDesc.ArraySize = 1;
	textureDesc.Format = DXGI_FORMAT_R8G8B8A8_UNORM;
	textureDesc.SampleDesc.Count = 1;
	textureDesc.SampleDesc.Quality = 0;
	textureDesc.Usage = D3D11_USAGE_DEFAULT;
	textureDesc.BindFlags = D3D11_BIND_SHADER_RESOURCE | D3D11_BIND_RENDER_TARGET;
	textureDesc.CPUAccessFlags = 0;
	textureDesc.MiscFlags = D3D11_RESOURCE_MISC_GENERATE_MIPS;

	HRESULT result = m_d3d_device->CreateTexture2D(&textureDesc, nullptr, &gpu_texture);
	RMLUI_DX_ASSERTMSG(result, "failed to CreateTexture2D");
	if (FAILED(result))
	{
		return Rml::TextureHandle(0);
	}

	// Set the row pitch of the raw image data
	uint32_t rowPitch = (source_dimensions.x * 4) * sizeof(unsigned char);

	// Copy the raw image data into the texture
	if (source.size() > 0)
	{
		m_d3d_context->UpdateSubresource(gpu_texture, 0, nullptr, source.data(), rowPitch, 0);
	}

	// Setup the shader resource view description.
	D3D11_SHADER_RESOURCE_VIEW_DESC srvDesc{};
	srvDesc.Format = textureDesc.Format;
	srvDesc.ViewDimension = D3D11_SRV_DIMENSION_TEXTURE2D;
	srvDesc.Texture2D.MostDetailedMip = 0;
	srvDesc.Texture2D.MipLevels = static_cast<UINT>(-1);

	// Create the shader resource view for the texture.
	result = m_d3d_device->CreateShaderResourceView(gpu_texture, &srvDesc, &gpu_texture_view);
	RMLUI_DX_ASSERTMSG(result, "failed to CreateShaderResourceView");
	if (FAILED(result))
	{
		Dx11CleanupResourceIfCreated(gpu_texture);
		return Rml::TextureHandle(0);
	}
	Dx11CleanupResourceIfCreated(gpu_texture);

	// Generate mipmaps for this texture.
	if (source.size() > 0)
	{
		m_d3d_context->GenerateMips(gpu_texture_view);
	}

	return Rml::TextureHandle(gpu_texture_view);
}

void RenderInterface_DX11::ReleaseTexture(Rml::TextureHandle texture_handle)
{
	ID3D11ShaderResourceView* texture_view = (ID3D11ShaderResourceView*)texture_handle;
	Dx11CleanupResourceIfCreated(texture_view);
}

static Rml::Colourf ConvertToColorf(Rml::ColourbPremultiplied c0)
{
	Rml::Colourf result;
	for (int i = 0; i < 4; i++)
		result[i] = (1.f / 255.f) * float(c0[i]);
	return result;
}

void RenderInterface_DX11::DrawFullscreenQuad()
{
	RenderGeometry(m_fullscreen_quad_geometry, {}, RenderInterface_DX11::TexturePostprocess);
}

void RenderInterface_DX11::DrawFullscreenQuad(Rml::Vector2f uv_offset, Rml::Vector2f uv_scaling)
{
	Rml::Mesh mesh;
	Rml::MeshUtilities::GenerateQuad(mesh, Rml::Vector2f(-1), Rml::Vector2f(2), {});
	if (uv_offset != Rml::Vector2f() || uv_scaling != Rml::Vector2f(1.f))
	{
		for (Rml::Vertex& vertex : mesh.vertices)
			vertex.tex_coord = (vertex.tex_coord * uv_scaling) + uv_offset;
	}
	const Rml::CompiledGeometryHandle geometry = CompileGeometry(mesh.vertices, mesh.indices);
	RenderGeometry(geometry, {}, RenderInterface_DX11::TexturePostprocess);
	ReleaseGeometry(geometry);
}

/// Flip vertical axis of the rectangle, and move its origin to the vertically opposite side of the viewport.
/// @note Changes coordinate system from RmlUi to OpenGL, or equivalently in reverse.
/// @note The Rectangle::Top and Rectangle::Bottom members will have reverse meaning in the returned rectangle.
static Rml::Rectanglei VerticallyFlipped(Rml::Rectanglei rect, int viewport_height)
{
	RMLUI_ASSERT(rect.Valid());
	Rml::Rectanglei flipped_rect = rect;
	flipped_rect.p0.y = viewport_height - rect.p1.y;
	flipped_rect.p1.y = viewport_height - rect.p0.y;
	return flipped_rect;
}

void RenderInterface_DX11::SetScissor(Rml::Rectanglei region, bool vertically_flip)
{
	if (region.Valid() != m_scissor_state.Valid())
	{
		if (region.Valid())
		{
			m_d3d_context->RSSetState(m_rasterizer_state_scissor_enabled);
			m_scissor_enabled = true;
		}
		else
		{
			m_d3d_context->RSSetState(m_rasterizer_state_scissor_disabled);
			m_scissor_enabled = false;
		}
	}

	if (region.Valid() && vertically_flip)
		region = VerticallyFlipped(region, m_viewport_height);

	if (region.Valid() && region != m_scissor_state)
	{
		// Some render APIs don't like offscreen positions (WebGL in particular), so clamp them to the viewport.
		const int x_min = Rml::Math::Clamp(region.Left(), 0, m_viewport_width);
		const int y_min = Rml::Math::Clamp(region.Top(), 0, m_viewport_height);
		const int x_max = Rml::Math::Clamp(region.Right(), 0, m_viewport_width);
		const int y_max = Rml::Math::Clamp(region.Bottom(), 0, m_viewport_height);

		D3D11_RECT rect_scissor = {};
		rect_scissor.left = x_min;
		rect_scissor.top = y_min;
		rect_scissor.right = x_max;
		rect_scissor.bottom = y_max;

		m_d3d_context->RSSetScissorRects(1, &rect_scissor);
	}

	m_scissor_state = region;
}

void RenderInterface_DX11::EnableScissorRegion(bool enable)
{
	// Assume enable is immediately followed by a SetScissorRegion() call, and ignore it here.
	if (!enable)
		SetScissor(Rml::Rectanglei::MakeInvalid(), false);
}

void RenderInterface_DX11::SetScissorRegion(Rml::Rectanglei region)
{
	SetScissor(region);
}

void RenderInterface_DX11::SetViewport(const int width, const int height)
{
	m_viewport_width = width;
	m_viewport_height = height;
	m_projection = Rml::Matrix4f::ProjectOrtho(0, (float)width, (float)height, 0, -10000, 10000);
}

void RenderInterface_DX11::Clear(ID3D11RenderTargetView* p_render_target_view)
{
	float clearColor[4] = {0.0f, 0.0f, 0.0f, 1.0f};
	m_d3d_context->ClearRenderTargetView(p_render_target_view, clearColor);
}

void RenderInterface_DX11::RestorePartialRenderState()
{
	UseProgram(ProgramId::None);

	ID3D11BlendState* restore_blend_state = m_current_blend_state;
	m_current_blend_state = nullptr;
	SetBlendState(restore_blend_state);
}

void RenderInterface_DX11::SetBlendState(ID3D11BlendState* blend_state)
{
	if (!m_current_blend_state_enabled || blend_state != m_current_blend_state)
	{
		const float blend_factor[4] = {1.0f, 1.0f, 1.0f, 1.0f};
		m_d3d_context->OMSetBlendState(blend_state, blend_factor, 0xFFFFFFFF);
		m_current_blend_state = blend_state;
		m_current_blend_state_enabled = true;
	}
}

void RenderInterface_DX11::DisableBlend()
{
	if (m_current_blend_state_enabled)
	{
		m_d3d_context->OMSetBlendState(nullptr, nullptr, 0xFFFFFFFF);
		m_current_blend_state_enabled = false;
	}
}

void RenderInterface_DX11::SetTransform(const Rml::Matrix4f* new_transform)
{
	m_transform = (new_transform ? (m_projection * (*new_transform)) : m_projection);
	m_cbuffer_dirty = true;
}

void RenderInterface_DX11::BlitRenderTarget(const Gfx::RenderTargetData& source, const Gfx::RenderTargetData& dest, int srcX0, int srcY0, int srcX1,
	int srcY1, int dstX0, int dstY0, int dstX1, int dstY1)
{
	int src_width = srcX1 - srcX0;
	int src_height = srcY1 - srcY0;
	int dest_width = dstX1 - dstX0;
	int dest_height = dstY1 - dstY0;

	bool is_flipped = src_width < 0 || src_height < 0 || dest_width < 0 || dest_height < 0;
	bool is_stretched = src_width != dest_width || src_height != dest_height;
	bool is_full_copy = src_width == dest_width && src_height == dest_height && srcX0 == 0 && srcY0 == 0 && dstX0 == 0 && dstY0 == 0;

	if (is_flipped || is_stretched || !is_full_copy)
	{
		// Full draw call. Slow path because no equivalent in DX11

		// Cache current state
		ID3D11RenderTargetView* original_render_target_view = nullptr;
		ID3D11DepthStencilView* original_depth_stencil_view = nullptr;
		m_d3d_context->OMGetRenderTargets(1, &original_render_target_view, &original_depth_stencil_view);

		// Bind temporary as the render target
		const Gfx::RenderTargetData& temporary_rt = m_render_layers.GetTemporary();

		// Unbind existing textures to prevent warning spam
		ID3D11ShaderResourceView* null_shader_resource_views[2] = {nullptr, nullptr};
		m_d3d_context->PSSetShaderResources(0, 2, null_shader_resource_views);

		// Disable blending
		DisableBlend();

		// Resolve from source to the temporary first
		m_d3d_context->ResolveSubresource(temporary_rt.render_target_texture, 0, source.render_target_texture, 0, DXGI_FORMAT_R8G8B8A8_UNORM);

		// Bind destination as our final render texture
		Gfx::BindRenderTarget(m_d3d_context, dest);

		D3D11_VIEWPORT d3dviewport{};
		d3dviewport.TopLeftX = 0;
		d3dviewport.TopLeftY = 0;
		d3dviewport.Width = (float)dest.width;
		d3dviewport.Height = (float)dest.height;
		d3dviewport.MinDepth = 0.0f;
		d3dviewport.MaxDepth = 1.0f;
		m_d3d_context->RSSetViewports(1, &d3dviewport);

		// Bind texture
		m_d3d_context->PSSetSamplers(0, 1, &m_sampler_state);
		Gfx::BindTexture(m_d3d_context, temporary_rt);
		UseProgram(ProgramId::Passthrough);

		// Draw a quad into temporary with source bound as the texture, and the UVs lerping to match the new dimensions
		// We want to map UV   0 -  1 to src min-max
		// We want to map pos -1 - +1 to dst min max

		float uv_x_min = float(srcX0) / float(source.width);  // Map to 0
		float uv_y_max = float(srcY0) / float(source.height); // Map to 0
		float uv_x_max = float(srcX1) / float(source.width);  // Map to +1
		float uv_y_min = float(srcY1) / float(source.height); // Map to +1

		float pos_x_min = (dstX0 / float(dest.width)) * 2.0f - 1.0f;
		float pos_y_min = ((dest.height - dstY0 - dest_height) / float(dest.height)) * 2.0f - 1.0f;
		float pos_x_max = ((dstX0 + dest_width) / float(dest.width)) * 2.0f - 1.0f;
		float pos_y_max = ((dest.height - dstY0) / float(dest.height)) * 2.0f - 1.0f;

		Rml::Mesh mesh;
		const int v0 = (int)mesh.vertices.size();
		const int i0 = (int)mesh.indices.size();

		mesh.vertices.resize(mesh.vertices.size() + 4);
		mesh.indices.resize(mesh.indices.size() + 6);
		Rml::Vertex* vertices = mesh.vertices.data();
		int* indices = mesh.indices.data();

		vertices[v0 + 0].position.x = pos_x_min;
		vertices[v0 + 0].position.y = pos_y_min;
		vertices[v0 + 0].tex_coord.x = uv_x_min;
		vertices[v0 + 0].tex_coord.y = uv_y_min;

		vertices[v0 + 1].position.x = pos_x_max;
		vertices[v0 + 1].position.y = pos_y_min;
		vertices[v0 + 1].tex_coord.x = uv_x_max;
		vertices[v0 + 1].tex_coord.y = uv_y_min;

		vertices[v0 + 2].position.x = pos_x_max;
		vertices[v0 + 2].position.y = pos_y_max;
		vertices[v0 + 2].tex_coord.x = uv_x_max;
		vertices[v0 + 2].tex_coord.y = uv_y_max;

		vertices[v0 + 3].position.x = pos_x_min;
		vertices[v0 + 3].position.y = pos_y_max;
		vertices[v0 + 3].tex_coord.x = uv_x_min;
		vertices[v0 + 3].tex_coord.y = uv_y_max;

		indices[i0 + 0] = v0 + 0;
		indices[i0 + 1] = v0 + 3;
		indices[i0 + 2] = v0 + 1;

		indices[i0 + 3] = v0 + 1;
		indices[i0 + 4] = v0 + 3;
		indices[i0 + 5] = v0 + 2;

		const Rml::CompiledGeometryHandle geometry = CompileGeometry(mesh.vertices, mesh.indices);
		RenderGeometry(geometry, {}, RenderInterface_DX11::TexturePostprocess);
		ReleaseGeometry(geometry);

		// Restore state
		m_d3d_context->OMSetRenderTargets(1, &original_render_target_view, original_depth_stencil_view);
		SetBlendState(m_current_blend_state);

		d3dviewport.Width = static_cast<FLOAT>(m_viewport_width);
		d3dviewport.Height = static_cast<FLOAT>(m_viewport_height);
		m_d3d_context->RSSetViewports(1, &d3dviewport);
		m_d3d_context->PSSetShaderResources(0, 2, null_shader_resource_views);
		Dx11CleanupResourceIfCreated(original_render_target_view);
		Dx11CleanupResourceIfCreated(original_depth_stencil_view);
	}
	else
	{
		// Resolve and move on
		m_d3d_context->ResolveSubresource(dest.render_target_texture, 0, source.render_target_texture, 0, DXGI_FORMAT_R8G8B8A8_UNORM);
	}
}

void RenderInterface_DX11::BlitLayerToPostprocessPrimary(Rml::LayerHandle layer_handle)
{
	const Gfx::RenderTargetData& source = m_render_layers.GetLayer(layer_handle);
	const Gfx::RenderTargetData& destination = m_render_layers.GetPostprocessPrimary();

	// Blit and resolve MSAA. Any active scissor state will restrict the size of the blit region.
	BlitRenderTarget(source, destination, 0, 0, source.width, source.height, 0, 0, destination.width, destination.height);
}

Rml::LayerHandle RenderInterface_DX11::PushLayer()
{
	const Rml::LayerHandle layer_handle = m_render_layers.PushLayer();

	Gfx::BindRenderTarget(m_d3d_context, m_render_layers.GetTopLayer());
	float colors[4] = {0, 0, 0, 0};
	m_d3d_context->ClearRenderTargetView(m_render_layers.GetTopLayer().render_target_view, colors);

	return layer_handle;
}

void RenderInterface_DX11::CompositeLayers(Rml::LayerHandle source_handle, Rml::LayerHandle destination_handle, Rml::BlendMode blend_mode,
	Rml::Span<const Rml::CompiledFilterHandle> filters)
{
	using Rml::BlendMode;

	// Blit source layer to postprocessing buffer. Do this regardless of whether we actually have any filters to be
	// applied, because we need to resolve the multi-sampled framebuffer in any case.
	// @performance If we have BlendMode::Replace and no filters or mask then we can just blit directly to the destination.
	BlitLayerToPostprocessPrimary(source_handle);

	// Render the filters, the PostprocessPrimary framebuffer is used for both input and output.
	RenderFilters(filters);

	// Render to the destination layer.
	Gfx::BindRenderTarget(m_d3d_context, m_render_layers.GetLayer(destination_handle));
	Gfx::BindTexture(m_d3d_context, m_render_layers.GetPostprocessPrimary());

	UseProgram(ProgramId::Passthrough);

	if (blend_mode == BlendMode::Replace)
		DisableBlend();

	DrawFullscreenQuad();

	if (blend_mode == BlendMode::Replace)
		SetBlendState(m_current_blend_state);

	if (destination_handle != m_render_layers.GetTopLayerHandle())
		Gfx::BindRenderTarget(m_d3d_context, m_render_layers.GetTopLayer());
}

void RenderInterface_DX11::PopLayer()
{
	m_render_layers.PopLayer();
	Gfx::BindRenderTarget(m_d3d_context, m_render_layers.GetTopLayer());
}

Rml::TextureHandle RenderInterface_DX11::SaveLayerAsTexture()
{
	RMLUI_ASSERT(m_scissor_state.Valid());
	const Rml::Rectanglei bounds = m_scissor_state;

	Rml::TextureHandle render_texture = GenerateTexture({}, bounds.Size());
	if (!render_texture)
	{
		return {};
	}

	BlitLayerToPostprocessPrimary(m_render_layers.GetTopLayerHandle());

	EnableScissorRegion(false);

	const Gfx::RenderTargetData& source = m_render_layers.GetPostprocessPrimary();
	const Gfx::RenderTargetData& destination = m_render_layers.GetPostprocessSecondary();

	// Flip the image vertically, as that convention is used for textures, and move to origin.
	BlitRenderTarget(source, destination, bounds.Left(), source.height - bounds.Bottom(), // src0
		bounds.Right(), source.height - bounds.Top(),                                     // src1
		0, bounds.Height(),                                                               // dst0
		bounds.Width(), 0                                                                 // dst1
	);

	// Now we need to copy the destination texture to the final texture for rendering
	// Bind the destination texture as the source for copying to the final texture (render_texture)
	// after extracting the associated resource with it
	ID3D11ShaderResourceView* texture_view = (ID3D11ShaderResourceView*)render_texture;
	ID3D11Resource* texture_resource = nullptr;
	texture_view->GetResource(&texture_resource);

	// Copy the destination texture to the final texture resource
	D3D11_BOX copy_box{};
	copy_box.left = bounds.Left();
	copy_box.right = bounds.Right();
	copy_box.top = bounds.Top();
	copy_box.bottom = bounds.Bottom();
	copy_box.front = 0;
	copy_box.back = 1;

	// Copy the blitted content from the destination to the final texture
	m_d3d_context->CopySubresourceRegion(texture_resource, 0, 0, 0, 0, destination.render_target_texture, 0, &copy_box);

	// Restore state and free memory
	SetScissor(bounds);
	Gfx::BindRenderTarget(m_d3d_context, m_render_layers.GetTopLayer());
	texture_resource->Release();

	return render_texture;
}

enum class FilterType { Invalid = 0, Passthrough, Blur, DropShadow, ColorMatrix, MaskImage };
struct CompiledFilter {
	FilterType type;

	// Passthrough
	float blend_factor;

	// Blur
	float sigma;

	// Drop shadow
	Rml::Vector2f offset;
	Rml::ColourbPremultiplied color;

	// ColorMatrix
	Rml::Matrix4f color_matrix;
};

Rml::CompiledFilterHandle RenderInterface_DX11::SaveLayerAsMaskImage()
{
	BlitLayerToPostprocessPrimary(m_render_layers.GetTopLayerHandle());

	const Gfx::RenderTargetData& source = m_render_layers.GetPostprocessPrimary();
	const Gfx::RenderTargetData& destination = m_render_layers.GetBlendMask();

	// Unbind resources
	ID3D11ShaderResourceView* const nullSRV = nullptr;
	m_d3d_context->PSSetShaderResources(0, 1, &nullSRV);

	Gfx::BindRenderTarget(m_d3d_context, destination);
	Gfx::BindTexture(m_d3d_context, source);
	UseProgram(ProgramId::Passthrough);
	ID3D11BlendState* blend_state_backup = m_current_blend_state;
	DisableBlend();

	DrawFullscreenQuad();

	// Unbind resources
	m_d3d_context->PSSetShaderResources(0, 1, &nullSRV);

	SetBlendState(blend_state_backup);
	Gfx::BindRenderTarget(m_d3d_context, m_render_layers.GetTopLayer());

	CompiledFilter filter = {};
	filter.type = FilterType::MaskImage;
	return reinterpret_cast<Rml::CompiledFilterHandle>(new CompiledFilter(std::move(filter)));
}

Rml::CompiledFilterHandle RenderInterface_DX11::CompileFilter(const Rml::String& name, const Rml::Dictionary& parameters)
{
	CompiledFilter filter = {};

	if (name == "opacity")
	{
		filter.type = FilterType::Passthrough;
		filter.blend_factor = Rml::Get(parameters, "value", 1.0f);
	}
	else if (name == "blur")
	{
		filter.type = FilterType::Blur;
		filter.sigma = Rml::Get(parameters, "sigma", 1.0f);
	}
	else if (name == "drop-shadow")
	{
		filter.type = FilterType::DropShadow;
		filter.sigma = Rml::Get(parameters, "sigma", 0.f);
		filter.color = Rml::Get(parameters, "color", Rml::Colourb()).ToPremultiplied();
		filter.offset = Rml::Get(parameters, "offset", Rml::Vector2f(0.f));
	}
	else if (name == "brightness")
	{
		filter.type = FilterType::ColorMatrix;
		const float value = Rml::Get(parameters, "value", 1.0f);
		filter.color_matrix = Rml::Matrix4f::Diag(value, value, value, 1.f);
	}
	else if (name == "contrast")
	{
		filter.type = FilterType::ColorMatrix;
		const float value = Rml::Get(parameters, "value", 1.0f);
		const float grayness = 0.5f - 0.5f * value;
		filter.color_matrix = Rml::Matrix4f::Diag(value, value, value, 1.f);
		filter.color_matrix.SetColumn(3, Rml::Vector4f(grayness, grayness, grayness, 1.f));
	}
	else if (name == "invert")
	{
		filter.type = FilterType::ColorMatrix;
		const float value = Rml::Math::Clamp(Rml::Get(parameters, "value", 1.0f), 0.f, 1.f);
		const float inverted = 1.f - 2.f * value;
		filter.color_matrix = Rml::Matrix4f::Diag(inverted, inverted, inverted, 1.f);
		filter.color_matrix.SetColumn(3, Rml::Vector4f(value, value, value, 1.f));
	}
	else if (name == "grayscale")
	{
		filter.type = FilterType::ColorMatrix;
		const float value = Rml::Get(parameters, "value", 1.0f);
		const float rev_value = 1.f - value;
		const Rml::Vector3f gray = value * Rml::Vector3f(0.2126f, 0.7152f, 0.0722f);
		// clang-format off
        filter.color_matrix = Rml::Matrix4f::FromRows(
            {gray.x + rev_value, gray.y,             gray.z,             0.f},
            {gray.x,             gray.y + rev_value, gray.z,             0.f},
            {gray.x,             gray.y,             gray.z + rev_value, 0.f},
            {0.f,                0.f,                0.f,                1.f}
        );
		// clang-format on
	}
	else if (name == "sepia")
	{
		filter.type = FilterType::ColorMatrix;
		const float value = Rml::Get(parameters, "value", 1.0f);
		const float rev_value = 1.f - value;
		const Rml::Vector3f r_mix = value * Rml::Vector3f(0.393f, 0.769f, 0.189f);
		const Rml::Vector3f g_mix = value * Rml::Vector3f(0.349f, 0.686f, 0.168f);
		const Rml::Vector3f b_mix = value * Rml::Vector3f(0.272f, 0.534f, 0.131f);
		// clang-format off
        filter.color_matrix = Rml::Matrix4f::FromRows(
            {r_mix.x + rev_value, r_mix.y,             r_mix.z,             0.f},
            {g_mix.x,             g_mix.y + rev_value, g_mix.z,             0.f},
            {b_mix.x,             b_mix.y,             b_mix.z + rev_value, 0.f},
            {0.f,                 0.f,                 0.f,                 1.f}
        );
		// clang-format on
	}
	else if (name == "hue-rotate")
	{
		// Hue-rotation and saturation values based on: https://www.w3.org/TR/filter-effects-1/#attr-valuedef-type-huerotate
		filter.type = FilterType::ColorMatrix;
		const float value = Rml::Get(parameters, "value", 1.0f);
		const float s = Rml::Math::Sin(value);
		const float c = Rml::Math::Cos(value);
		// clang-format off
        filter.color_matrix = Rml::Matrix4f::FromRows(
            {0.213f + 0.787f * c - 0.213f * s,  0.715f - 0.715f * c - 0.715f * s,  0.072f - 0.072f * c + 0.928f * s,  0.f},
            {0.213f - 0.213f * c + 0.143f * s,  0.715f + 0.285f * c + 0.140f * s,  0.072f - 0.072f * c - 0.283f * s,  0.f},
            {0.213f - 0.213f * c - 0.787f * s,  0.715f - 0.715f * c + 0.715f * s,  0.072f + 0.928f * c + 0.072f * s,  0.f},
            {0.f,                               0.f,                               0.f,                               1.f}
        );
		// clang-format on
	}
	else if (name == "saturate")
	{
		filter.type = FilterType::ColorMatrix;
		const float value = Rml::Get(parameters, "value", 1.0f);
		// clang-format off
        filter.color_matrix = Rml::Matrix4f::FromRows(
            {0.213f + 0.787f * value,  0.715f - 0.715f * value,  0.072f - 0.072f * value,  0.f},
            {0.213f - 0.213f * value,  0.715f + 0.285f * value,  0.072f - 0.072f * value,  0.f},
            {0.213f - 0.213f * value,  0.715f - 0.715f * value,  0.072f + 0.928f * value,  0.f},
            {0.f,                      0.f,                      0.f,                      1.f}
        );
		// clang-format on
	}

	if (filter.type != FilterType::Invalid)
		return reinterpret_cast<Rml::CompiledFilterHandle>(new CompiledFilter(std::move(filter)));

	Rml::Log::Message(Rml::Log::LT_WARNING, "Unsupported filter type '%s'.", name.c_str());
	return {};
}

void RenderInterface_DX11::ReleaseFilter(Rml::CompiledFilterHandle filter)
{
	delete reinterpret_cast<CompiledFilter*>(filter);
}

enum class CompiledShaderType { Invalid = 0, Gradient, Creation };
struct CompiledShader {
	CompiledShaderType type;

	// Gradient
	ShaderGradientFunction gradient_function;
	Rml::Vector2f p;
	Rml::Vector2f v;
	Rml::Vector<float> stop_positions;
	Rml::Vector<Rml::Colourf> stop_colors;

	// Shader
	Rml::Vector2f dimensions;
};

Rml::CompiledShaderHandle RenderInterface_DX11::CompileShader(const Rml::String& name, const Rml::Dictionary& parameters)
{
	auto ApplyColorStopList = [](CompiledShader& shader, const Rml::Dictionary& shader_parameters) {
		auto it = shader_parameters.find("color_stop_list");
		RMLUI_ASSERT(it != shader_parameters.end() && it->second.GetType() == Rml::Variant::COLORSTOPLIST);
		const Rml::ColorStopList& color_stop_list = it->second.GetReference<Rml::ColorStopList>();
		const int num_stops = Rml::Math::Min((int)color_stop_list.size(), MAX_NUM_STOPS);

		shader.stop_positions.resize(num_stops);
		shader.stop_colors.resize(num_stops);
		for (int i = 0; i < num_stops; i++)
		{
			const Rml::ColorStop& stop = color_stop_list[i];
			RMLUI_ASSERT(stop.position.unit == Rml::Unit::NUMBER);
			shader.stop_positions[i] = stop.position.number;
			shader.stop_colors[i] = ConvertToColorf(stop.color);
		}
	};

	CompiledShader shader = {};

	if (name == "linear-gradient")
	{
		shader.type = CompiledShaderType::Gradient;
		const bool repeating = Rml::Get(parameters, "repeating", false);
		shader.gradient_function = (repeating ? ShaderGradientFunction::RepeatingLinear : ShaderGradientFunction::Linear);
		shader.p = Rml::Get(parameters, "p0", Rml::Vector2f(0.f));
		shader.v = Rml::Get(parameters, "p1", Rml::Vector2f(0.f)) - shader.p;
		ApplyColorStopList(shader, parameters);
	}
	else if (name == "radial-gradient")
	{
		shader.type = CompiledShaderType::Gradient;
		const bool repeating = Rml::Get(parameters, "repeating", false);
		shader.gradient_function = (repeating ? ShaderGradientFunction::RepeatingRadial : ShaderGradientFunction::Radial);
		shader.p = Rml::Get(parameters, "center", Rml::Vector2f(0.f));
		shader.v = Rml::Vector2f(1.f) / Rml::Get(parameters, "radius", Rml::Vector2f(1.f));
		ApplyColorStopList(shader, parameters);
	}
	else if (name == "conic-gradient")
	{
		shader.type = CompiledShaderType::Gradient;
		const bool repeating = Rml::Get(parameters, "repeating", false);
		shader.gradient_function = (repeating ? ShaderGradientFunction::RepeatingConic : ShaderGradientFunction::Conic);
		shader.p = Rml::Get(parameters, "center", Rml::Vector2f(0.f));
		const float angle = Rml::Get(parameters, "angle", 0.f);
		shader.v = {Rml::Math::Cos(angle), Rml::Math::Sin(angle)};
		ApplyColorStopList(shader, parameters);
	}
	else if (name == "shader")
	{
		const Rml::String value = Rml::Get(parameters, "value", Rml::String());
		if (value == "creation")
		{
			shader.type = CompiledShaderType::Creation;
			shader.dimensions = Rml::Get(parameters, "dimensions", Rml::Vector2f(0.f));
		}
	}

	if (shader.type != CompiledShaderType::Invalid)
		return reinterpret_cast<Rml::CompiledShaderHandle>(new CompiledShader(std::move(shader)));

	Rml::Log::Message(Rml::Log::LT_WARNING, "Unsupported shader type '%s'.", name.c_str());
	return {};
}

void RenderInterface_DX11::RenderShader(Rml::CompiledShaderHandle shader_handle, Rml::CompiledGeometryHandle geometry_handle,
	Rml::Vector2f translation, Rml::TextureHandle /*texture*/)
{
	RMLUI_ASSERT(shader_handle && geometry_handle);
	const CompiledShader& shader = *reinterpret_cast<CompiledShader*>(shader_handle);
	const CompiledShaderType type = shader.type;

	DX11_GeometryData* geometry = (DX11_GeometryData*)geometry_handle;

	switch (type)
	{
	case CompiledShaderType::Gradient:
	{
		RMLUI_ASSERT(shader.stop_positions.size() == shader.stop_colors.size());
		const int num_stops = (int)shader.stop_positions.size();

		UseProgram(ProgramId::Gradient);

		D3D11_MAPPED_SUBRESOURCE mappedResource{};
		// Lock the constant buffer so it can be written to.
		HRESULT result = m_d3d_context->Map(m_shader_buffer, 0, D3D11_MAP_WRITE_DISCARD, 0, &mappedResource);
		if (FAILED(result))
		{
			return;
		}

		// Get a pointer to the data in the constant buffer.
		ShaderCbuffer* dataPtr = (ShaderCbuffer*)mappedResource.pData;

		// Copy the data to the GPU
		dataPtr->common.transform = m_transform;
		dataPtr->common.translation = translation;
		dataPtr->gradient.func = static_cast<int>(shader.gradient_function);
		dataPtr->gradient.p = shader.p;
		dataPtr->gradient.v = shader.v;
		dataPtr->gradient.num_stops = num_stops;
		// Reset stop positions and colours to 0
		memset(dataPtr->gradient.stop_positions, 0, sizeof(dataPtr->gradient.stop_positions));
		memset(dataPtr->gradient.stop_colors, 0, sizeof(dataPtr->gradient.stop_colors));
		// Copy to stop position and colours
		memcpy_s(&dataPtr->gradient.stop_positions, num_stops * sizeof(float), shader.stop_positions.data(), num_stops * sizeof(float));
		memcpy_s(&dataPtr->gradient.stop_colors, num_stops * sizeof(Rml::Vector4f), shader.stop_colors.data(), num_stops * sizeof(Rml::Vector4f));

		// Upload to the GPU.
		m_d3d_context->Unmap(m_shader_buffer, 0);

		// Issue draw call
		m_d3d_context->IASetInputLayout(m_vertex_layout);
		m_d3d_context->VSSetConstantBuffers(0, 1, &m_shader_buffer);
		m_d3d_context->PSSetConstantBuffers(0, 1, &m_shader_buffer);
		uint32_t stride = sizeof(Rml::Vertex);
		uint32_t offset = 0;
		m_d3d_context->IASetVertexBuffers(0, 1, &geometry->vertex_buffer, &stride, &offset);
		m_d3d_context->IASetIndexBuffer(geometry->index_buffer, DXGI_FORMAT_R32_UINT, 0);
		m_d3d_context->IASetPrimitiveTopology(D3D11_PRIMITIVE_TOPOLOGY_TRIANGLELIST);
		m_d3d_context->DrawIndexed(static_cast<UINT>(geometry->index_count), 0, 0);
	}
	break;
	case CompiledShaderType::Creation:
	{
		const double time = Rml::GetSystemInterface()->GetElapsedTime();

		UseProgram(ProgramId::Creation);

		D3D11_MAPPED_SUBRESOURCE mappedResource{};
		// Lock the constant buffer so it can be written to.
		HRESULT result = m_d3d_context->Map(m_shader_buffer, 0, D3D11_MAP_WRITE_DISCARD, 0, &mappedResource);
		if (FAILED(result))
		{
			return;
		}

		// Get a pointer to the data in the constant buffer.
		ShaderCbuffer* dataPtr = (ShaderCbuffer*)mappedResource.pData;

		// Copy the data to the GPU
		dataPtr->common.transform = m_transform;
		dataPtr->common.translation = translation;
		dataPtr->creation.value = (float)time;
		dataPtr->creation.dimensions = shader.dimensions;

		// Upload to the GPU.
		m_d3d_context->Unmap(m_shader_buffer, 0);

		// Issue draw call
		m_d3d_context->IASetInputLayout(m_vertex_layout);
		m_d3d_context->VSSetConstantBuffers(0, 1, &m_shader_buffer);
		m_d3d_context->PSSetConstantBuffers(0, 1, &m_shader_buffer);
		uint32_t stride = sizeof(Rml::Vertex);
		uint32_t offset = 0;
		m_d3d_context->IASetVertexBuffers(0, 1, &geometry->vertex_buffer, &stride, &offset);
		m_d3d_context->IASetIndexBuffer(geometry->index_buffer, DXGI_FORMAT_R32_UINT, 0);
		m_d3d_context->IASetPrimitiveTopology(D3D11_PRIMITIVE_TOPOLOGY_TRIANGLELIST);
		m_d3d_context->DrawIndexed(static_cast<UINT>(geometry->index_count), 0, 0);
	}
	break;
	case CompiledShaderType::Invalid:
	{
		Rml::Log::Message(Rml::Log::LT_WARNING, "Unhandled render shader %d.", (int)type);
	}
	break;
	}
}

void RenderInterface_DX11::ReleaseShader(Rml::CompiledShaderHandle shader_handle)
{
	delete reinterpret_cast<CompiledShader*>(shader_handle);
}

static void SigmaToParameters(const float desired_sigma, int& out_pass_level, float& out_sigma)
{
	constexpr int max_num_passes = 10;
	static_assert(max_num_passes < 31, "");
	constexpr float max_single_pass_sigma = 3.0f;
	out_pass_level = Rml::Math::Clamp(Rml::Math::Log2(int(desired_sigma * (2.f / max_single_pass_sigma))), 0, max_num_passes);
	out_sigma = Rml::Math::Clamp(desired_sigma / float(1 << out_pass_level), 0.0f, max_single_pass_sigma);
}

static void SetTexCoordLimits(Rml::Vector2f& p_tex_coord_min, Rml::Vector2f& p_tex_coord_max, Rml::Rectanglei rectangle_flipped,
	Rml::Vector2i framebuffer_size)
{
	// Offset by half-texel values so that texture lookups are clamped to fragment centers, thereby avoiding color
	// bleeding from neighboring texels due to bilinear interpolation.
	const Rml::Vector2f min = (Rml::Vector2f(rectangle_flipped.p0) + Rml::Vector2f(0.5f)) / Rml::Vector2f(framebuffer_size);
	const Rml::Vector2f max = (Rml::Vector2f(rectangle_flipped.p1) - Rml::Vector2f(0.5f)) / Rml::Vector2f(framebuffer_size);

	p_tex_coord_min.x = min.x;
	p_tex_coord_min.y = min.y;
	p_tex_coord_max.x = max.x;
	p_tex_coord_max.y = max.y;
}

static void SetBlurWeights(Rml::Vector4f& p_weights, float sigma)
{
	constexpr int num_weights = BLUR_NUM_WEIGHTS;
	float weights[num_weights];
	float normalization = 0.0f;
	for (int i = 0; i < num_weights; i++)
	{
		if (Rml::Math::Absolute(sigma) < 0.1f)
			weights[i] = float(i == 0);
		else
			weights[i] = Rml::Math::Exp(-float(i * i) / (2.0f * sigma * sigma)) / (Rml::Math::SquareRoot(2.f * Rml::Math::RMLUI_PI) * sigma);

		normalization += (i == 0 ? 1.f : 2.0f) * weights[i];
	}
	for (float& weight : weights)
		weight /= normalization;

	p_weights.x = weights[0];
	p_weights.y = weights[1];
	p_weights.z = weights[2];
	p_weights.w = weights[3];
}

void RenderInterface_DX11::RenderBlur(float sigma, const Gfx::RenderTargetData& source_destination, const Gfx::RenderTargetData& temp,
	const Rml::Rectanglei window_flipped)
{
	RMLUI_ASSERT(&source_destination != &temp && source_destination.width == temp.width && source_destination.height == temp.height);
	RMLUI_ASSERT(window_flipped.Valid());

	if (window_flipped.Top() == window_flipped.Bottom() || window_flipped.Left() == window_flipped.Right())
	{
		// In some cases the bounds are 0 width, so early exit since there's nothing to blur
		return;
	}

	int pass_level = 0;
	SigmaToParameters(sigma, pass_level, sigma);
	if (sigma == 0)
	{
		// If sigma is zero we don't have anything to blur, so early exit since blurring will have no effect on the final image
		return;
	}

	const Rml::Rectanglei original_scissor = m_scissor_state;

	// Begin by downscaling so that the blur pass can be done at a reduced resolution for large sigma.
	Rml::Rectanglei scissor = window_flipped;

	UseProgram(ProgramId::Passthrough);
	SetScissor(scissor, true);

	// Downscale by iterative half-scaling with bilinear filtering, to reduce aliasing.
	D3D11_VIEWPORT d3dviewport{};
	d3dviewport.TopLeftX = 0;
	d3dviewport.TopLeftY = static_cast<FLOAT>(source_destination.height / 2);
	d3dviewport.Width = static_cast<FLOAT>(source_destination.width / 2);
	d3dviewport.Height = static_cast<FLOAT>(source_destination.height / 2);
	d3dviewport.MinDepth = 0.0f;
	d3dviewport.MaxDepth = 1.0f;
	m_d3d_context->RSSetViewports(1, &d3dviewport);

	// Scale UVs if we have even dimensions, such that texture fetches align perfectly between texels, thereby producing a 50% blend of
	// neighboring texels.
	const Rml::Vector2f uv_scaling = {(source_destination.width % 2 == 1) ? (1.f - 1.f / float(source_destination.width)) : 1.f,
		(source_destination.height % 2 == 1) ? (1.f - 1.f / float(source_destination.height)) : 1.f};

	for (int i = 0; i < pass_level; i++)
	{
		scissor.p0 = (scissor.p0 + Rml::Vector2i(1)) / 2;
		scissor.p1 = Rml::Math::Max(scissor.p1 / 2, scissor.p0);
		const bool from_source = (i % 2 == 0);
		Gfx::BindRenderTarget(m_d3d_context, (from_source ? temp : source_destination));
		Gfx::BindTexture(m_d3d_context, from_source ? source_destination : temp);
		SetScissor(scissor, true);

		DrawFullscreenQuad({}, uv_scaling);
	}

	d3dviewport.TopLeftX = 0;
	d3dviewport.TopLeftY = 0;
	d3dviewport.Width = static_cast<FLOAT>(source_destination.width);
	d3dviewport.Height = static_cast<FLOAT>(source_destination.height);
	d3dviewport.MinDepth = 0.0f;
	d3dviewport.MaxDepth = 1.0f;
	m_d3d_context->RSSetViewports(1, &d3dviewport);

	// Ensure texture data end up in the temp buffer. Depending on the last downscaling, we might need to move it from the source_destination buffer.
	const bool transfer_to_temp_buffer = (pass_level % 2 == 0);
	if (transfer_to_temp_buffer)
	{
		Gfx::BindRenderTarget(m_d3d_context, temp);
		Gfx::BindTexture(m_d3d_context, source_destination);
		DrawFullscreenQuad();
	}

	// Set up uniforms.
	UseProgram(ProgramId::Blur);
	D3D11_MAPPED_SUBRESOURCE mappedResource{};
	// Lock the constant buffer so it can be written to.
	HRESULT result = m_d3d_context->Map(m_shader_buffer, 0, D3D11_MAP_WRITE_DISCARD, 0, &mappedResource);
	if (FAILED(result))
	{
		return;
	}

	// Get a pointer to the data in the constant buffer.
	ShaderCbuffer* dataPtr = (ShaderCbuffer*)mappedResource.pData;

	SetBlurWeights(dataPtr->blur.weights, sigma);
	SetTexCoordLimits(dataPtr->blur.texcoord_min, dataPtr->blur.texcoord_max, scissor, {source_destination.width, source_destination.height});

	Rml::Vector2f texel_offset = Rml::Vector2f(0.f, 1.f) * (1.0f / float(temp.height));
	dataPtr->blur.texel_offset = texel_offset;

	// Upload to the GPU.
	m_d3d_context->Unmap(m_shader_buffer, 0);
	// Unbind second pass
	ID3D11Buffer* null_buffer = nullptr;
	m_d3d_context->VSSetConstantBuffers(1, 1, &null_buffer);

	// Blur render pass - vertical.
	Gfx::BindRenderTarget(m_d3d_context, source_destination);
	Gfx::BindTexture(m_d3d_context, temp);
	DrawFullscreenQuad();

	// Blur render pass - horizontal.
	Gfx::BindRenderTarget(m_d3d_context, temp);
	Gfx::BindTexture(m_d3d_context, source_destination);

	// Add a 1px transparent border around the blur region by first clearing with a padded scissor. This helps prevent
	// artifacts when upscaling the blur result in the later step. On Intel and AMD, we have observed that during
	// blitting with linear filtering, pixels outside the 'src' region can be blended into the output. On the other
	// hand, it looks like Nvidia clamps the pixels to the source edge, which is what we really want. Regardless, we
	// work around the issue with this extra step.
	SetScissor(scissor.Extend(1), true);
	// D3D11 does not support clearing based on scissor rects. So we use a D3D11Context1 to use ClearView, and define the D3D11_RECt here
	Rml::Rectanglei scissor_ext = scissor.Extend(1);
	scissor_ext = VerticallyFlipped(scissor_ext, m_viewport_height);

	// Some render APIs don't like offscreen positions (WebGL in particular), so clamp them to the viewport.
	const int x_min = Rml::Math::Clamp(scissor_ext.Left(), 0, m_viewport_width);
	const int y_min = Rml::Math::Clamp(scissor_ext.Top(), 0, m_viewport_height);
	const int x_max = Rml::Math::Clamp(scissor_ext.Right(), 0, m_viewport_width);
	const int y_max = Rml::Math::Clamp(scissor_ext.Bottom(), 0, m_viewport_height);

	D3D11_RECT rect_scissor = {};
	rect_scissor.left = x_min;
	rect_scissor.top = y_min;
	rect_scissor.right = x_max;
	rect_scissor.bottom = y_max;

	FLOAT clear_color[4] = {0.0f, 0.0f, 0.0f, 0.0f};
	m_d3d_context->ClearView(temp.render_target_view, clear_color, &rect_scissor, 1);
	SetScissor(scissor, true);

	// Update texel offset for upscale pass
	result = m_d3d_context->Map(m_blur_cbuffer, 0, D3D11_MAP_WRITE_DISCARD, 0, &mappedResource);
	if (FAILED(result))
	{
		ID3D11ShaderResourceView* const nullSRV = nullptr;
		m_d3d_context->PSSetShaderResources(0, 1, &nullSRV);
		return;
	}

	ShaderCbuffer::Blur* dataPtrBlur = (ShaderCbuffer::Blur*)mappedResource.pData;

	texel_offset = Rml::Vector2f(1.f, 0.f) * (1.0f / float(source_destination.width));
	dataPtrBlur->texel_offset = texel_offset;

	// Upload to the GPU.
	m_d3d_context->Unmap(m_blur_cbuffer, 0);
	m_d3d_context->VSSetConstantBuffers(0, 1, &m_blur_cbuffer);

	RenderGeometry(m_fullscreen_quad_geometry, {}, RenderInterface_DX11::TexturePostprocessNoBinding);

	// Blit the blurred image to the scissor region with upscaling.
	Rml::Rectanglei window_flipped_twice = VerticallyFlipped(window_flipped, m_viewport_height);
	SetScissor(window_flipped_twice, false);

	const Rml::Vector2i src_min = scissor.p0;
	const Rml::Vector2i src_max = scissor.p1;
	const Rml::Vector2i dst_min = window_flipped_twice.p0;
	const Rml::Vector2i dst_max = window_flipped_twice.p1;
	BlitRenderTarget(temp, source_destination, src_min.x, src_min.y, src_max.x, src_max.y, dst_min.x, dst_max.y, dst_max.x, dst_min.y);

	// The above upscale blit might be jittery at low resolutions (large pass levels). This is especially noticeable when moving an element with
	// backdrop blur around or when trying to click/hover an element within a blurred region since it may be rendered at an offset. For more stable
	// and accurate rendering we next upscale the blur image by an exact power-of-two. However, this may not fill the edges completely so we need to
	// do the above first. Note that this strategy may sometimes result in visible seams. Alternatively, we could try to enlarge the window to the
	// next power-of-two size and then downsample and blur that.

	// We need to flip the scissor rect to have it match DX11 coords
	const Rml::Vector2i target_min = src_min * (1 << pass_level);
	const Rml::Vector2i target_max = src_max * (1 << pass_level);
	if (target_min != window_flipped.p0 || target_max != window_flipped.p1)
	{
		window_flipped_twice = VerticallyFlipped(scissor, m_viewport_height);
		const Rml::Vector2i target_min_blit = window_flipped_twice.p0 * (1 << pass_level);
		const Rml::Vector2i target_max_blit = window_flipped_twice.p1 * (1 << pass_level);
		BlitRenderTarget(temp, source_destination, src_min.x, src_min.y, src_max.x, src_max.y, target_min_blit.x, target_max_blit.y,
			target_max_blit.x, target_min_blit.y);
	}

	// Restore render state.
	SetScissor(original_scissor);

	ID3D11ShaderResourceView* const nullSRV = nullptr;
	m_d3d_context->PSSetShaderResources(0, 1, &nullSRV);
}

void RenderInterface_DX11::RenderFilters(Rml::Span<const Rml::CompiledFilterHandle> filter_handles)
{
	m_d3d_context->PSSetSamplers(0, 1, &m_sampler_state);

	// To unbind slots when necessary
	ID3D11ShaderResourceView* null_shader_resource_views[2] = {nullptr, nullptr};

	for (const Rml::CompiledFilterHandle filter_handle : filter_handles)
	{
		const CompiledFilter& filter = *reinterpret_cast<const CompiledFilter*>(filter_handle);
		const FilterType type = filter.type;

		switch (type)
		{
		case FilterType::Passthrough:
		{
			UseProgram(ProgramId::Passthrough);
			const float blend_factor[4] = {filter.blend_factor, filter.blend_factor, filter.blend_factor, filter.blend_factor};
			m_d3d_context->OMSetBlendState(m_blend_state_color_filter, blend_factor, 0xFFFFFFFF);
			m_current_blend_state = m_blend_state_color_filter;

			const Gfx::RenderTargetData& source = m_render_layers.GetPostprocessPrimary();
			const Gfx::RenderTargetData& destination = m_render_layers.GetPostprocessSecondary();
			Gfx::BindRenderTarget(m_d3d_context, destination);
			Gfx::BindTexture(m_d3d_context, source);
			m_d3d_context->PSSetSamplers(0, 1, &m_sampler_state);

			DrawFullscreenQuad();

			m_render_layers.SwapPostprocessPrimarySecondary();
			SetBlendState(m_blend_state_enable);
			m_d3d_context->PSSetShaderResources(0, 2, null_shader_resource_views);
		}
		break;
		case FilterType::Blur:
		{
			ID3D11BlendState* blend_state_backup = m_current_blend_state;
			DisableBlend();

			const Gfx::RenderTargetData& source_destination = m_render_layers.GetPostprocessPrimary();
			const Gfx::RenderTargetData& temp = m_render_layers.GetPostprocessSecondary();

			const Rml::Rectanglei window_flipped = VerticallyFlipped(m_scissor_state, m_viewport_height);
			RenderBlur(filter.sigma, source_destination, temp, window_flipped);

			SetBlendState(blend_state_backup);
		}
		break;
		case FilterType::DropShadow:
		{
			UseProgram(ProgramId::DropShadow);
			ID3D11BlendState* blend_state_backup = m_current_blend_state;
			DisableBlend();

			const Gfx::RenderTargetData& primary = m_render_layers.GetPostprocessPrimary();
			const Gfx::RenderTargetData& secondary = m_render_layers.GetPostprocessSecondary();
			Gfx::BindRenderTarget(m_d3d_context, secondary);
			Gfx::BindTexture(m_d3d_context, primary);

			D3D11_MAPPED_SUBRESOURCE mappedResource{};
			// Lock the constant buffer so it can be written to.
			HRESULT result = m_d3d_context->Map(m_drop_shadow_cbuffer, 0, D3D11_MAP_WRITE_DISCARD, 0, &mappedResource);
			if (FAILED(result))
			{
				return;
			}

			// Get a pointer to the data in the constant buffer.
			ShaderCbuffer::DropShadow* dataPtr = (ShaderCbuffer::DropShadow*)mappedResource.pData;

			Rml::Colourf color = ConvertToColorf(filter.color);

			// Copy the data to the GPU
			dataPtr->color.x = color.red;
			dataPtr->color.y = color.green;
			dataPtr->color.z = color.blue;
			dataPtr->color.w = color.alpha;

			const Rml::Rectanglei window_flipped = VerticallyFlipped(m_scissor_state, m_viewport_height);
			SetTexCoordLimits(dataPtr->texcoord_min, dataPtr->texcoord_max, m_scissor_state, {primary.width, primary.height});

			// Upload to the GPU.
			m_d3d_context->Unmap(m_drop_shadow_cbuffer, 0);
			m_d3d_context->PSSetConstantBuffers(1, 1, &m_drop_shadow_cbuffer);

			m_cbuffer_dirty = true;

			const Rml::Vector2f uv_offset = filter.offset / Rml::Vector2f(-(float)m_viewport_width, (float)m_viewport_height);
			DrawFullscreenQuad(uv_offset);

			if (filter.sigma >= 0.5f)
			{
				const Gfx::RenderTargetData& tertiary = m_render_layers.GetPostprocessTertiary();
				RenderBlur(filter.sigma, secondary, tertiary, window_flipped);
			}

			UseProgram(ProgramId::Passthrough);
			Gfx::BindRenderTarget(m_d3d_context, secondary);
			Gfx::BindTexture(m_d3d_context, primary);
			SetBlendState(blend_state_backup);
			DrawFullscreenQuad();

			m_render_layers.SwapPostprocessPrimarySecondary();

			ID3D11ShaderResourceView* const nullSRV = nullptr;
			m_d3d_context->PSSetShaderResources(0, 1, &nullSRV);
			m_d3d_context->PSSetShaderResources(1, 1, &nullSRV);
		}
		break;
		case FilterType::ColorMatrix:
		{
			UseProgram(ProgramId::ColorMatrix);

			ID3D11BlendState* blend_state_backup = m_current_blend_state;
			DisableBlend();

			const Gfx::RenderTargetData& source = m_render_layers.GetPostprocessPrimary();
			const Gfx::RenderTargetData& destination = m_render_layers.GetPostprocessSecondary();
			Gfx::BindRenderTarget(m_d3d_context, destination);
			Gfx::BindTexture(m_d3d_context, source);

			D3D11_MAPPED_SUBRESOURCE mappedResource{};
			// Lock the constant buffer so it can be written to.
			HRESULT result = m_d3d_context->Map(m_color_matrix_cbuffer, 0, D3D11_MAP_WRITE_DISCARD, 0, &mappedResource);
			if (FAILED(result))
			{
				// Restore blend state
				SetBlendState(blend_state_backup);

				ID3D11ShaderResourceView* const nullSRV = nullptr;
				m_d3d_context->PSSetShaderResources(0, 1, &nullSRV);
				m_cbuffer_dirty = true;
				return;
			}

			// Get a pointer to the data in the constant buffer.
			ColorMatrixCbuffer* dataPtr = (ColorMatrixCbuffer*)mappedResource.pData;

			constexpr bool transpose = std::is_same<decltype(filter.color_matrix), Rml::RowMajorMatrix4f>::value;

			// Copy the data to the GPU
			memcpy_s(&dataPtr->color_matrix, sizeof(dataPtr->color_matrix),
				transpose ? filter.color_matrix.Transpose().data() : filter.color_matrix.data(), sizeof(dataPtr->color_matrix));

			// Upload to the GPU.
			m_d3d_context->Unmap(m_color_matrix_cbuffer, 0);

			m_d3d_context->PSSetConstantBuffers(1, 1, &m_color_matrix_cbuffer);

			DrawFullscreenQuad();

			// Unbind cbuffer
			ID3D11Buffer* null_cbuffer = nullptr;
			m_d3d_context->PSSetConstantBuffers(1, 1, &null_cbuffer);

			m_render_layers.SwapPostprocessPrimarySecondary();

			// Restore blend state
			SetBlendState(blend_state_backup);

			ID3D11ShaderResourceView* const nullSRV = nullptr;
			m_d3d_context->PSSetShaderResources(0, 1, &nullSRV);
			m_cbuffer_dirty = true;
		}
		break;
		case FilterType::MaskImage:
		{
			UseProgram(ProgramId::BlendMask);
			ID3D11BlendState* blend_state_backup = m_current_blend_state;
			DisableBlend();

			const Gfx::RenderTargetData& source = m_render_layers.GetPostprocessPrimary();
			const Gfx::RenderTargetData& blend_mask = m_render_layers.GetBlendMask();
			const Gfx::RenderTargetData& destination = m_render_layers.GetPostprocessSecondary();

			Gfx::BindRenderTarget(m_d3d_context, destination);
			Gfx::BindTexture(m_d3d_context, source, 0);
			Gfx::BindTexture(m_d3d_context, blend_mask, 1);

			DrawFullscreenQuad();

			m_render_layers.SwapPostprocessPrimarySecondary();
			SetBlendState(blend_state_backup);
			m_d3d_context->PSSetShaderResources(0, 2, null_shader_resource_views);
		}
		break;
		case FilterType::Invalid:
		{
			Rml::Log::Message(Rml::Log::LT_WARNING, "Unhandled render filter %d.", (int)type);
		}
		break;
		}

		// Unbind resources
		ID3D11ShaderResourceView* const nullSRV = nullptr;
		m_d3d_context->PSSetShaderResources(0, 1, &nullSRV);
	}
}

void RenderInterface_DX11::UpdateConstantBuffer()
{
	if (m_cbuffer_dirty)
	{
		D3D11_MAPPED_SUBRESOURCE mappedResource{};
		// Lock the constant buffer so it can be written to.
		HRESULT result = m_d3d_context->Map(m_shader_buffer, 0, D3D11_MAP_WRITE_DISCARD, 0, &mappedResource);
		if (FAILED(result))
		{
			return;
		}

		// Get a pointer to the data in the constant buffer.
		ShaderCbuffer* dataPtr = (ShaderCbuffer*)mappedResource.pData;

		// Copy the data to the GPU
		dataPtr->common.transform = m_transform;
		dataPtr->common.translation = m_translation;

		// Upload to the GPU.
		m_d3d_context->Unmap(m_shader_buffer, 0);
		m_cbuffer_dirty = false;
	}
}

void RenderInterface_DX11::UseProgram(ProgramId program_id)
{
	RMLUI_ASSERT(program_data);
	if (active_program != program_id)
	{
		if (program_id != ProgramId::None)
		{
			m_d3d_context->VSSetShader(program_data->programs[program_id].vertex_shader, nullptr, 0);
			m_d3d_context->PSSetShader(program_data->programs[program_id].pixel_shader, nullptr, 0);
		}
		active_program = program_id;
	}
}

void RenderInterface_DX11::EnableClipMask(bool enable)
{
	m_is_stencil_enabled = enable;
	if (enable)
	{
		m_d3d_context->OMSetDepthStencilState(m_depth_stencil_state_stencil_test, m_stencil_ref_value);
	}
	else
	{
		m_d3d_context->OMSetDepthStencilState(m_depth_stencil_state_disable, 0);
	}
}

void RenderInterface_DX11::RenderToClipMask(Rml::ClipMaskOperation operation, Rml::CompiledGeometryHandle geometry, Rml::Vector2f translation)
{
	RMLUI_ASSERT(m_is_stencil_enabled);
	using Rml::ClipMaskOperation;

	ID3D11DepthStencilState* stencil_state = m_depth_stencil_state_disable;

	UINT stencil_test_value = m_stencil_ref_value;
	switch (operation)
	{
	case ClipMaskOperation::Set:
	{
		stencil_test_value = 1;
		stencil_state = m_depth_stencil_state_stencil_set;
	}
	break;
	case ClipMaskOperation::SetInverse:
	{
		stencil_test_value = 0;
		stencil_state = m_depth_stencil_state_stencil_set;
	}
	break;
	case ClipMaskOperation::Intersect:
	{
		stencil_test_value += 1;
		stencil_state = m_depth_stencil_state_stencil_intersect;
	}
	break;
	}

	// Disable writing to the color of the render target
	float blendFactor[4] = {1.0f, 1.0f, 1.0f, 1.0f};
	m_d3d_context->OMSetBlendState(m_blend_state_disable_color, blendFactor, 0xffffffff);

	Gfx::BindRenderTarget(m_d3d_context, m_render_layers.GetTopLayer());
	m_d3d_context->OMSetDepthStencilState(stencil_state, 1);

	bool clear_stencil = (operation == ClipMaskOperation::Set || operation == ClipMaskOperation::SetInverse);
	if (clear_stencil)
	{
		// Clear stencil buffer
		Rml::LayerHandle layer_handle = m_render_layers.GetTopLayerHandle();
		const Gfx::RenderTargetData& rtData = m_render_layers.GetLayer(layer_handle);
		m_d3d_context->ClearDepthStencilView(rtData.depth_stencil_view, D3D11_CLEAR_STENCIL, 1.0f, 0);
	}

	RenderGeometry(geometry, translation, {});

	// Restore blend state
	m_d3d_context->OMSetBlendState(m_current_blend_state, blendFactor, 0xffffffff);
	// Restore state
	m_stencil_ref_value = stencil_test_value;
	m_d3d_context->OMSetDepthStencilState(m_depth_stencil_state_stencil_test, m_stencil_ref_value);
}

RenderInterface_DX11::RenderLayerStack::RenderLayerStack()
{
	m_rt_postprocess.resize(5);
}

RenderInterface_DX11::RenderLayerStack::~RenderLayerStack()
{
	DestroyRenderTargets();
}

void RenderInterface_DX11::RenderLayerStack::SetD3DResources(ID3D11Device* device)
{
	RMLUI_ASSERTMSG(!m_d3d_device, "D3D11Device has already been set!");
	m_d3d_device = device;
}

Rml::LayerHandle RenderInterface_DX11::RenderLayerStack::PushLayer()
{
	RMLUI_ASSERT(m_layers_size <= (int)m_rt_layers.size());

	if (m_layers_size == (int)m_rt_layers.size())
	{
		// All framebuffers should share a single stencil buffer.
		ID3D11DepthStencilView* shared_depth_stencil = (m_rt_layers.empty() ? nullptr : m_rt_layers.front().depth_stencil_view);

		m_rt_layers.push_back(Gfx::RenderTargetData{});
		Gfx::CreateRenderTarget(m_d3d_device, m_rt_layers.back(), m_width, m_height, NUM_MSAA_SAMPLES, Gfx::RenderTargetAttachment::DepthStencil,
			shared_depth_stencil);
	}

	m_layers_size += 1;
	return GetTopLayerHandle();
}

void RenderInterface_DX11::RenderLayerStack::PopLayer()
{
	RMLUI_ASSERT(m_layers_size > 0);
	m_layers_size -= 1;
}

const Gfx::RenderTargetData& RenderInterface_DX11::RenderLayerStack::GetLayer(Rml::LayerHandle layer) const
{
	RMLUI_ASSERT((size_t)layer < (size_t)m_layers_size);
	return m_rt_layers[layer];
}

const Gfx::RenderTargetData& RenderInterface_DX11::RenderLayerStack::GetTopLayer() const
{
	return GetLayer(GetTopLayerHandle());
}

Rml::LayerHandle RenderInterface_DX11::RenderLayerStack::GetTopLayerHandle() const
{
	RMLUI_ASSERT(m_layers_size > 0);
	return static_cast<Rml::LayerHandle>(m_layers_size - 1);
}

void RenderInterface_DX11::RenderLayerStack::SwapPostprocessPrimarySecondary()
{
	std::swap(m_rt_postprocess[0], m_rt_postprocess[1]);
}

void RenderInterface_DX11::RenderLayerStack::BeginFrame(int new_width, int new_height)
{
	RMLUI_ASSERT(m_layers_size == 0);

	if (new_width != m_width || new_height != m_height)
	{
		m_width = new_width;
		m_height = new_height;

		DestroyRenderTargets();
	}

	PushLayer();
}

void RenderInterface_DX11::RenderLayerStack::EndFrame()
{
	RMLUI_ASSERT(m_layers_size == 1);
	PopLayer();
}

void RenderInterface_DX11::RenderLayerStack::DestroyRenderTargets()
{
	RMLUI_ASSERTMSG(m_layers_size == 0, "Do not call this during frame rendering, that is, between BeginFrame() and EndFrame().");

	for (Gfx::RenderTargetData& fb : m_rt_layers)
		Gfx::DestroyRenderTarget(fb);

	m_rt_layers.clear();

	for (Gfx::RenderTargetData& fb : m_rt_postprocess)
		Gfx::DestroyRenderTarget(fb);
}

const Gfx::RenderTargetData& RenderInterface_DX11::RenderLayerStack::EnsureRenderTargetPostprocess(int index)
{
	RMLUI_ASSERT(index < (int)m_rt_postprocess.size())
	Gfx::RenderTargetData& rt = m_rt_postprocess[index];
	if (!rt.render_target_view)
		Gfx::CreateRenderTarget(m_d3d_device, rt, m_width, m_height, 0, Gfx::RenderTargetAttachment::None, nullptr);
	return rt;
}