fs2open.github.com/code/graphics/2d.h

/*
 * Copyright (C) Volition, Inc. 1999.  All rights reserved.
 *
 * All source code herein is the property of Volition, Inc. You may not sell
 * or otherwise commercially exploit the source or things you created based on the
 * source.
 *
 */

#ifndef _GRAPHICS_H
#define _GRAPHICS_H

#include "globalincs/flagset.h"
#include "globalincs/pstypes.h"

#include "bmpman/bmpman.h"
#include "cfile/cfile.h"
#include "graphics/grinternal.h"
#include "graphics/tmapper.h"
#include "io/cursor.h"
#include "math/vecmat.h"
#include "osapi/osapi.h"
#include "utils/id.h"

// Forward definition
namespace graphics {
namespace util {
class UniformBuffer;
class GPUMemoryHeap;
} // namespace util
} // namespace graphics
namespace scripting {
template<typename condition_t>
class OverridableHook;
}

extern const float Default_min_draw_distance;
extern const float Default_max_draw_distance;
extern float Min_draw_distance_cockpit;
extern float Min_draw_distance;
extern float Max_draw_distance;
extern int Gr_inited;

// # Software Re-added by Kazan --- THIS HAS TO STAY -- It is used by standalone!
enum class GraphicsAPI : uint8_t { Default, Stub, OpenGL, Vulkan };

// z-buffering stuff
extern int gr_zbuffering, gr_zbuffering_mode;
extern int gr_global_zbuffering;

extern bool Gr_enable_soft_particles;

FLAG_LIST(FramebufferEffects){Thrusters = 0, Shockwaves, NUM_VALUES};
extern flagset<FramebufferEffects> Gr_framebuffer_effects;

enum class AntiAliasMode {
	None = 0,

	FXAA_Low = 1,
	FXAA_Medium = 2,
	FXAA_High = 3,

	SMAA_Low = 4,
	SMAA_Medium = 5,
	SMAA_High = 6,
	SMAA_Ultra = 7,
};
extern AntiAliasMode Gr_aa_mode;
extern AntiAliasMode Gr_aa_mode_last_frame;

bool gr_is_fxaa_mode(AntiAliasMode mode);
bool gr_is_smaa_mode(AntiAliasMode mode);

extern bool Gr_post_processing_enabled;

extern bool Gr_enable_vsync;

// HDR10 (PQ/ST.2084 + BT.2020) output. Currently only honored by the Vulkan renderer.
extern bool Gr_enable_hdr;
// True once the renderer has actually negotiated an HDR10 swap chain (read-only,
// set by the active renderer). Distinct from Gr_enable_hdr, which is the request.
extern bool Gr_hdr_output_active;
// Reference white luminance in nits (the brightness of SDR "paper white" / UI when HDR is active).
extern float Gr_hdr_paperwhite_nits;
// Display peak luminance in nits used for tone curve clamping and HDR10 metadata.
extern float Gr_hdr_peak_nits;

extern bool Deferred_lighting;
extern bool High_dynamic_range;

extern os::ViewportState Gr_configured_window_state;

extern const std::shared_ptr<scripting::OverridableHook<void>> OnFrameHook;

class material;
class model_material;
class particle_material;
class distortion_material;
class shield_material;
class movie_material;
class batched_bitmap_material;
class nanovg_material;
class decal_material;
class interface_material;

class transform_stack {

	matrix4 Current_transform;
	SCP_vector<matrix4> Stack;
public:
	transform_stack()
	{
		vm_matrix4_set_identity(&Current_transform);

		Stack.clear();
		Stack.push_back(Current_transform);
	}

	const matrix4 &get_transform() const
	{
		return Current_transform;
	}

	void clear()
	{
		vm_matrix4_set_identity(&Current_transform);

		Stack.clear();
		Stack.push_back(Current_transform);
	}

	void push_and_replace(const matrix4 &new_transform)
	{
		Current_transform = new_transform;
		Stack.push_back(Current_transform);
	}

	void push(const vec3d *pos, const matrix *orient, const vec3d *scale = NULL)
	{
		vec3d new_scale = SCALE_IDENTITY_VECTOR;
		matrix new_orient = IDENTITY_MATRIX;
		vec3d new_pos = ZERO_VECTOR;

		matrix4 current_transform_copy = Current_transform;
		matrix4 new_transform;

		if ( pos != NULL ) {
			new_pos = *pos;
		}

		if ( orient != NULL ) {
			new_orient = *orient;
		}

		if ( scale != NULL ) {
			new_scale = *scale;
		}

		vm_vec_scale(&new_orient.vec.rvec, new_scale.xyz.x);
		vm_vec_scale(&new_orient.vec.uvec, new_scale.xyz.y);
		vm_vec_scale(&new_orient.vec.fvec, new_scale.xyz.z);

		vm_matrix4_set_transform(&new_transform, &new_orient, &new_pos);

		vm_matrix4_x_matrix4(&Current_transform, &current_transform_copy, &new_transform);
		Stack.push_back(Current_transform);
	}

	void pop()
	{
		if ( Stack.size() > 1 ) {
			Stack.pop_back();
		}

		Current_transform = Stack.back();
	}

	size_t depth() {
		return Stack.size();
	}
};

enum primitive_type {
	PRIM_TYPE_POINTS,
	PRIM_TYPE_LINES,
	PRIM_TYPE_LINESTRIP,
	PRIM_TYPE_TRIS,
	PRIM_TYPE_TRISTRIP,
	PRIM_TYPE_TRIFAN,
};

enum shader_type {
	SDR_TYPE_NONE = -1,
	SDR_TYPE_MODEL,
	SDR_TYPE_EFFECT_PARTICLE,
	SDR_TYPE_EFFECT_DISTORTION,
	SDR_TYPE_POST_PROCESS_MAIN,
	SDR_TYPE_POST_PROCESS_BLUR,
	SDR_TYPE_POST_PROCESS_BLOOM_COMP,
	SDR_TYPE_POST_PROCESS_BRIGHTPASS,
	SDR_TYPE_POST_PROCESS_FXAA,
	SDR_TYPE_POST_PROCESS_FXAA_PREPASS,
	SDR_TYPE_POST_PROCESS_LIGHTSHAFTS,
	SDR_TYPE_POST_PROCESS_TONEMAPPING,
	SDR_TYPE_DEFERRED_LIGHTING,
	SDR_TYPE_DEFERRED_CLEAR,
	SDR_TYPE_VIDEO_PROCESS,
	SDR_TYPE_PASSTHROUGH_RENDER, //!< Shader for doing the old style fixed-function rendering. Only used internally, use
	                             //!< SDR_TYPE_DEFAULT_MATERIAL.
	SDR_TYPE_SHIELD_DECAL,
	SDR_TYPE_BATCHED_BITMAP,
	SDR_TYPE_DEFAULT_MATERIAL,
	SDR_TYPE_NANOVG,
	SDR_TYPE_DECAL,
	SDR_TYPE_SCENE_FOG,
	SDR_TYPE_VOLUMETRIC_FOG,
	SDR_TYPE_ROCKET_UI,
	SDR_TYPE_COPY,
	SDR_TYPE_COPY_WORLD,
	SDR_TYPE_MSAA_RESOLVE,

	SDR_TYPE_POST_PROCESS_SMAA_EDGE,
	SDR_TYPE_POST_PROCESS_SMAA_BLENDING_WEIGHT,
	SDR_TYPE_POST_PROCESS_SMAA_NEIGHBORHOOD_BLENDING,
	SDR_TYPE_POST_PROCESS_SMAA_RESOLVE,

	SDR_TYPE_ENVMAP_SPHERE_WARP,

	SDR_TYPE_IRRADIANCE_MAP_GEN,

	SDR_TYPE_SHADOW_MAP_GEN,

	SDR_TYPE_GAMMA_BLIT,

	NUM_SHADER_TYPES
};


// Shader flags
#define SDR_FLAG_PARTICLE_POINT_GEN			(1<<0)

#define SDR_FLAG_BLUR_HORIZONTAL			(1<<0)
#define SDR_FLAG_BLUR_VERTICAL				(1<<1)

#define SDR_FLAG_NANOVG_EDGE_AA		(1<<0)

#define SDR_FLAG_DECAL_USE_NORMAL_MAP (1<<0)

#define SDR_FLAG_MSAA_SAMPLES_4 (1 << 0)
#define SDR_FLAG_MSAA_SAMPLES_8 (1 << 1)
#define SDR_FLAG_MSAA_SAMPLES_16 (1 << 2)

#define SDR_FLAG_VOLUMETRICS_DO_EDGE_SMOOTHING (1<<0)
#define SDR_FLAG_VOLUMETRICS_NOISE (1<<1)

#define SDR_FLAG_COPY_FROM_ARRAY (1 << 0)
#define SDR_FLAG_COPY_CLAMP01 (1 << 1)

#define SDR_FLAG_TONEMAPPING_LINEAR_OUT (1 << 0)

#define SDR_FLAG_GAMMA_HDR10 (1 << 0)

#define SDR_FLAG_ENV_MAP (1 << 0)
#define SDR_FLAG_DEFERRED_RT_SHADOWS (1 << 1)

#define SDR_FLAG_SHADOW_FALLBACK (1 << 0)

#define SDR_FLAG_SHADOW_FALLBACK (1 << 0)


enum class uniform_block_type {
	Lights = 0,
	ModelData = 1,
	NanoVGData = 2,
	DecalInfo = 3,
	DecalGlobals = 4,
	DeferredGlobals = 5,
	Matrices = 6,
	MovieData = 7,
	GenericData = 8,
	ShadowMapData = 9,
	ShadowCascadeParams = 10,

	NUM_BLOCK_TYPES
};

struct vertex_format_data
{
	enum vertex_format {
		POSITION4,
		POSITION3,
		POSITION2,
		COLOR3,
		COLOR4,
		COLOR4F,
		TEX_COORD2,
		TEX_COORD4,
		NORMAL,
		TANGENT,
		MODEL_ID,
		RADIUS,
		UVEC,
		MATRIX4,
	};

	vertex_format format_type;
	size_t stride;
	size_t offset;
	size_t divisor;
	size_t buffer_number;

	vertex_format_data(vertex_format i_format_type, size_t i_stride, size_t i_offset, size_t i_divisor, size_t i_buffer_number) :
	format_type(i_format_type), stride(i_stride), offset(i_offset), divisor(i_divisor), buffer_number(i_buffer_number) {}

	static inline uint mask(vertex_format v_format) { return 1 << v_format; }

	bool operator==(const vertex_format_data& other) const {
		return format_type == other.format_type && stride == other.stride && offset == other.offset;
	}
};
class vertex_layout
{
	SCP_vector<vertex_format_data> Vertex_components;

	uint Vertex_mask = 0;
	SCP_unordered_map<size_t, size_t> Vertex_stride;
public:
	vertex_layout() {}

	size_t get_num_vertex_components() const { return Vertex_components.size(); }

	const vertex_format_data* get_vertex_component(size_t index) const { return &Vertex_components[index]; }
	
	bool resident_vertex_format(vertex_format_data::vertex_format format_type) const;

	void add_vertex_component(vertex_format_data::vertex_format format_type, size_t stride, size_t offset, size_t divisor = 0, size_t buffer_number = 0);

	size_t get_vertex_stride(size_t buffer_number = 0) const { return Vertex_stride.at(buffer_number); }

	bool operator==(const vertex_layout& other) const;

	size_t hash() const;
};
namespace std {
template<> struct hash<vertex_format_data> {
	size_t operator()(const vertex_format_data& data) const;
};
template<> struct hash<vertex_layout> {
	size_t operator()(const vertex_layout& data) const;
};
}

enum class gr_capability {
	CAPABILITY_ENVIRONMENT_MAP,
	CAPABILITY_NORMAL_MAP,
	CAPABILITY_HEIGHT_MAP,
	CAPABILITY_SOFT_PARTICLES,
	CAPABILITY_DISTORTION,
	CAPABILITY_POST_PROCESSING,
	CAPABILITY_DEFERRED_LIGHTING,
	CAPABILITY_SHADOWS,
	CAPABILITY_THICK_OUTLINE,
	CAPABILITY_BATCHED_SUBMODELS,
	CAPABILITY_TIMESTAMP_QUERY,
	CAPABILITY_SEPARATE_BLEND_FUNCTIONS,
	CAPABILITY_PERSISTENT_BUFFER_MAPPING,
	CAPABILITY_BPTC,
	CAPABILITY_S3TC,
	CAPABILITY_LARGE_SHADER,
	CAPABILITY_INSTANCED_RENDERING,
	CAPABILITY_FAST_SHADOWS,
	CAPABILITY_QUERIES_REUSABLE,
	CAPABILITY_RAYTRACED_SHADOWS
};

struct gr_capability_def {
	gr_capability capability;
	const char* parse_name;
};

extern gr_capability_def gr_capabilities[];
extern const size_t gr_capabilities_num;

enum class gr_property
{
	UNIFORM_BUFFER_OFFSET_ALIGNMENT,
	UNIFORM_BUFFER_MAX_SIZE,
	MAX_ANISOTROPY
};

struct gr_buffer_handle_tag {
};
using gr_buffer_handle = ::util::ID<gr_buffer_handle_tag, int, -1>;

// stencil buffering stuff
extern int gr_stencil_mode;

/**
 * This is a structure used by the shader to keep track
 * of the values you want to use in the shade primitive.
 */
typedef struct shader {
	ubyte r, g, b, c; // factors and constant
	ubyte lookup[256];
} shader;

#define AC_TYPE_NONE		0		// Not an alphacolor
#define AC_TYPE_HUD		1		// Doesn't change hue depending on background.  Used for HUD stuff.
#define AC_TYPE_BLEND	2		// Changes hue depending on background.  Used for stars, etc.

// NEVER REFERENCE THESE VALUES OUTSIDE OF THE GRAPHICS LIBRARY!!!
// If you need to get the rgb values of a "color" struct call
// gr_get_colors after calling gr_set_colors_fast.
typedef struct color {
	int		is_alphacolor;
	int		alphacolor;
	int		magic;
	ubyte		red;
	ubyte		green;
	ubyte		blue;
	ubyte		alpha;
	ubyte		ac_type;							// The type of alphacolor.  See AC_TYPE_??? defines
	ubyte		raw8;
} color;

// Used by the team coloring code
typedef struct team_color {
	struct {
		float r, g, b;
	} base;
	struct {
		float r, g, b;
	} stripe;
} team_color;



typedef struct tsb_t {
	vec3d tangent;
	float scaler;
} tsb_t;

/**
 * This should be basicly just like it is in the VB
 * a list of triangles and their associated normals
 */
class poly_list {
	// helper function struct that let's us sort the indices.
	// an instance is fed into std::sort and std::lower_bound.
	// overloaded operator() is used for the comparison function.
	struct finder {
		poly_list* search_list;
		bool compare_indices;
		vertex* vert_to_find;
		vec3d* norm_to_find;

		finder(poly_list* _search_list): search_list(_search_list), compare_indices(true), vert_to_find(NULL), norm_to_find(NULL) {}
		finder(poly_list* _search_list, vertex* _vert, vec3d* _norm): search_list(_search_list), compare_indices(false), vert_to_find(_vert), norm_to_find(_norm) {}

		bool operator()(const uint a, const uint b);
	};
public:
	poly_list(): n_verts(0), vert(NULL), norm(NULL), tsb(NULL), submodels(NULL), sorted_indices(NULL), currently_allocated(0) {}
	~poly_list();
	poly_list& operator=(const poly_list&);

	void allocate(int size);
	void make_index_buffer(SCP_vector<int> &vertex_list);
	void calculate_tangent();
	int n_verts;
	vertex *vert;
	vec3d *norm;
	tsb_t *tsb;
	int *submodels;

	uint *sorted_indices;

	int find_index(poly_list *plist, int idx);
	int find_index_fast(poly_list *plist, int idx);
private:
	int currently_allocated;
	int find_first_vertex(int idx);
	int find_first_vertex_fast(int idx);
	void generate_sorted_index_list();
};

class buffer_data
{
public:
	int flags;

	int texture;		// this is the texture the vertex buffer will use
	size_t n_verts;

	size_t index_offset;

	const uint *get_index() const
	{
		return index;
	}

        uint i_first, i_last;

	void release()
	{
		if (index) {
			delete [] index;
			index = NULL;
		}
	}

	void assign(size_t i, uint j)
	{
		const_cast<uint *>(index)[i] = j;
		if (i_first > i_last)
			i_first = i_last = j;
		else if (i_first > j)
			i_first = j;
		else if (i_last < j)
			i_last = j;
	}

	// Constructor

	buffer_data() :
	flags(0), texture(-1), n_verts(0), index_offset(0),
		i_first(1), i_last(0), index(NULL)
	{
	}

	explicit buffer_data(size_t n_vrts) :
		flags(0), texture(-1), n_verts(n_vrts), index_offset(0),
		i_first(1), i_last(0), index(NULL)
	{
		if ( n_verts > 0 ) {
			index = new(std::nothrow) uint[n_verts];
		} else {
			index = NULL;
		}
	}

	// Copy-constructor
	buffer_data(const buffer_data& other)
	{
		if ( other.index ) {
			index = new(std::nothrow) uint[other.n_verts];
			for (size_t i=0; i < other.n_verts; i++)
			{
				index[i] = other.index[i];
			}
		} else {
			index = NULL;
		}

		flags   = other.flags;
		texture = other.texture;
		n_verts = other.n_verts;

		i_first = other.i_first;
		i_last  = other.i_last;

		index_offset = other.index_offset;
	}

	// Copy-assignment operator
	buffer_data& operator=(const buffer_data& rhs)
	{
		if (this != &rhs)
		{
			if ( index ) {
				delete [] index;
			}

			if ( rhs.index && rhs.n_verts > 0 ) {
				index = new(std::nothrow) uint[rhs.n_verts];
				for (size_t i=0; i < rhs.n_verts; i++)
				{
					index[i] = rhs.index[i];
				}
			}

			flags   = rhs.flags;
			texture = rhs.texture;
			n_verts = rhs.n_verts;

			i_first = rhs.i_first;
			i_last  = rhs.i_last;

			index_offset = rhs.index_offset;
		}
		return *this;
	}

	// Destructor
	~buffer_data()
	{
		release();
	}

private:
	uint *index;
};

class vertex_buffer
{
public:
	int flags;

	size_t stride;
	size_t vertex_offset;
	size_t vertex_num_offset;

	poly_list *model_list;

	SCP_vector<buffer_data> tex_buf;

	vertex_layout layout;

	vertex_buffer() :
		flags(0), stride(0), vertex_offset(0), vertex_num_offset(0), model_list(NULL)
	{
	}

	~vertex_buffer()
	{
		clear();
	}

	void release()
	{
		if (model_list) {
			delete model_list;
			model_list = NULL;
		}

		for (SCP_vector<buffer_data>::iterator tbi = tex_buf.begin(); tbi != tex_buf.end(); ++tbi) {
			tbi->release();
		}
	}

	void clear()
	{
		release();
		tex_buf.clear();
	}
};

struct indexed_vertex_source {
	std::shared_ptr<uint8_t[]> Vertex_list = nullptr;
	std::shared_ptr<uint8_t[]> Index_list = nullptr;
	gr_buffer_handle Vbuffer_handle;
	size_t Vertex_offset = 0;
	size_t Base_vertex_offset = 0;

	gr_buffer_handle Ibuffer_handle;
	size_t Index_offset = 0;

	uint Vertex_list_size = 0;
	uint Index_list_size = 0;
};

struct light;

#define FIND_SCALED_NUM(x, x0, x1, y0, y1) ( ((((x) - (x0)) * ((y1) - (y0))) / ((x1) - (x0))) + (y0) )

#define GR_ALPHABLEND_NONE			0		// no blending
#define GR_ALPHABLEND_FILTER		1		// 50/50 mix of foreground, background, using intensity as alpha

#define GR_BITBLT_MODE_NORMAL		0		// Normal bitblting
#define GR_BITBLT_MODE_RLE			1		// RLE would be faster

// fog modes
#define GR_FOGMODE_NONE				0		// set this to turn off fog
#define GR_FOGMODE_FOG				1		// linear fog

enum class QueryType {
	Timestamp
};

enum class BufferType {
	Vertex,
	Index,
	Uniform
};

enum class BufferUsageHint { Static, Dynamic, Streaming, PersistentMapping };

/**
 * @brief Type of a graphics sync object
 */
typedef void* gr_sync;

typedef struct screen {
	int max_w = 0, max_h = 0; // Width and height
	int max_w_unscaled = 0, max_h_unscaled = 0;
	int max_w_unscaled_zoomed = 0, max_h_unscaled_zoomed = 0;
	int center_w = 0, center_h = 0; // Width and height of center monitor
	int center_offset_x = 0, center_offset_y = 0;
	int save_max_w = 0, save_max_h = 0; // Width and height
	int save_max_w_unscaled = 0, save_max_h_unscaled = 0;
	int save_max_w_unscaled_zoomed = 0, save_max_h_unscaled_zoomed = 0;
	int save_center_w = 0, save_center_h = 0; // Width and height of center monitor
	int save_center_offset_x = 0, save_center_offset_y = 0;
	int res = 0;                             // GR_640 or GR_1024
	GraphicsAPI mode = GraphicsAPI::Default;                            // What mode gr_init was called with.
	float aspect = 0.0f, clip_aspect = 0.0f; // Aspect ratio = 0, aspect of clip_width/clip_height
	int rowsize = 0;                         // What you need to add to go to next row (includes bytes_per_pixel)
	int bits_per_pixel = 0;                  // How many bits per pixel it is. (7,8,15,16,24,32)
	int bytes_per_pixel = 0;                 // How many bytes per pixel (1,2,3,4)
	int offset_x = 0, offset_y = 0;          // The offsets into the screen
	int offset_x_unscaled = 0, offset_y_unscaled = 0; // Offsets into the screen, in unscaled dimensions
	int clip_width = 0, clip_height = 0;
	int clip_width_unscaled = 0, clip_height_unscaled = 0; // Height and width of clip aread, in unscaled dimensions
	// center of clip area
	float clip_center_x = 0.0f, clip_center_y = 0.0f;

	float fog_near = 0.0f, fog_far = 0.0f;

	// the clip_l,r,t,b are used internally.  left and top are
	// actually always 0, but it's nice to have the code work with
	// arbitrary clipping regions.
	int clip_left = 0, clip_right = 0, clip_top = 0, clip_bottom = 0;
	// same as above except in unscaled dimensions
	int clip_left_unscaled = 0, clip_right_unscaled = 0, clip_top_unscaled = 0, clip_bottom_unscaled = 0;

	int current_alphablend_mode = 0; // See GR_ALPHABLEND defines above
	int current_bitblt_mode = 0;     // See GR_BITBLT_MODE defines above
	int current_bitmap = 0;
	color current_color{};
	color current_fog_color{};   // current fog color
	color current_clear_color{}; // current clear color
	shader current_shader{};
	float current_alpha = 0.0f;

	bool custom_size = false;
	int rendering_to_texture = 0; // wich texture we are rendering to, -1 if the back buffer
	int rendering_to_face = 0;    // wich face of the texture we are rendering to, -1 if the back buffer

	int envmap_render_target = 0;
	int irrmap_render_target = -1; // Irradiance map for diffuse env lighting.

	float line_width = 0.0f;

	// switch onscreen, offscreen
	std::function<void()> gf_flip;

	std::function<void()> gf_setup_frame;

	// sets the clipping region
	std::function<void(int x, int y, int w, int h, int resize_mode)> gf_set_clip;

	// resets the clipping region to entire screen
	std::function<void()> gf_reset_clip;

	// clears entire clipping region to current color
	std::function<void()> gf_clear;

	// dumps the current screen to a file
	std::function<void(const char* filename)> gf_print_screen;

	// dumps the current screen to a html blob string
	std::function<SCP_string()> gf_blob_screen;

	// transforms and dumps the current environment map to a file
	std::function<void(const char* filename)> gf_dump_envmap;

	// generate diffuse irradiance cubemap for IBL.
	std::function<void()> gf_calculate_irrmap;

	// Retrieves the zbuffer mode.
	std::function<int()> gf_zbuffer_get;

	// Sets mode.  Returns previous mode.
	std::function<int(int mode)> gf_zbuffer_set;

	// Clears the zbuffer.  If use_zbuffer is FALSE, then zbuffering mode is ignored and zbuffer is always off.
	std::function<void(int use_zbuffer)> gf_zbuffer_clear;

	// Set the stencil buffer mode. Returns previous mode
	std::function<int(int mode)> gf_stencil_set;

	// Clears the stencil buffer.
	std::function<void()> gf_stencil_clear;

	std::function<int(int mode, float alpha)> gf_alpha_mask_set;

	// Saves screen. Returns an id you pass to restore and free.
	std::function<int()> gf_save_screen;

	// Resets clip region and copies entire saved screen to the screen.
	std::function<void(int id)> gf_restore_screen;

	// Frees up a saved screen.
	std::function<void(int id)> gf_free_screen;

	// grab a region of the screen. assumes data is large enough
	std::function<void(int front, int w, int h, ubyte* data)> gf_get_region;

	// poly culling
	std::function<int(int cull)> gf_set_cull;

	// color buffer writes
	std::function<int(int mode)> gf_set_color_buffer;

	// preload a bitmap into texture memory
	std::function<int(int bitmap_num, int is_aabitmap)> gf_preload;

	// set the color to be used when clearing the background
	std::function<void(int r, int g, int b)> gf_set_clear_color;

	// Here be the bitmap functions
	std::function<void(bitmap_slot* slot, bool release)> gf_bm_free_data;
	std::function<void(bitmap_slot* slot)> gf_bm_create;
	std::function<void(bitmap_slot* slot)> gf_bm_init;
	std::function<void()> gf_bm_page_in_start;
	std::function<bool(int handle, bitmap* bm)> gf_bm_data;

	std::function<int(int handle, int* width, int* height, int* bpp, int* mm_lvl, int flags)> gf_bm_make_render_target;
	std::function<int(int handle, int face)> gf_bm_set_render_target;

	std::function<void(int)> gf_set_texture_addressing;

	std::function<gr_buffer_handle(BufferType type, BufferUsageHint usage, bool rt_capable)> gf_create_buffer;
	std::function<void(gr_buffer_handle handle)> gf_delete_buffer;

	// Optional lifecycle hooks for backends that maintain auxiliary per-model GPU
	// state (e.g. Vulkan's raytraced-shadow BLAS cache). Default to no-ops so
	// backends that don't need them (OpenGL, stub) don't have to assign anything.
	std::function<void(int pm_id)> gf_model_loaded = [](int) {};
	std::function<void(int pm_id)> gf_model_unloaded = [](int) {};

	// Rebuilds the raytraced-shadow top-level acceleration structure for the
	// current frame from the live shadow-casting object set. Called once per
	// frame from shadows_render_all(), alongside the cascaded shadow map pass --
	// the shading pass picks between the two per shadows_use_raytracing().
	// No-op default.
	std::function<void()> gf_build_shadow_tlas = []() {};

	std::function<void(gr_buffer_handle handle, size_t size, const void* data)> gf_update_buffer_data;
	std::function<void(gr_buffer_handle handle, size_t offset, size_t size, const void* data)>
		gf_update_buffer_data_offset;
	std::function<void*(gr_buffer_handle handle)> gf_map_buffer;
	std::function<void(gr_buffer_handle handle, size_t offset, size_t size)> gf_flush_mapped_buffer;
	std::function<void(void* data, size_t size)> gf_update_transform_buffer;

	// postprocessing effects
	std::function<void(const char*, int, const vec3d*)> gf_post_process_set_effect;
	std::function<void()> gf_post_process_set_defaults;

	std::function<void()> gf_post_process_begin;
	std::function<void()> gf_post_process_end;
	std::function<void()> gf_post_process_save_zbuffer;
	std::function<void()> gf_post_process_restore_zbuffer;

	std::function<void(bool clearNonColorBufs)> gf_deferred_lighting_begin;
	std::function<void()> gf_deferred_lighting_msaa;
	std::function<void()> gf_deferred_lighting_end;
	std::function<void()> gf_deferred_lighting_finish;

	std::function<void()> gf_scene_texture_begin;
	std::function<void()> gf_scene_texture_end;
	std::function<void()> gf_copy_effect_texture;

	std::function<void(int zbias)> gf_zbias;

	std::function<void(int)> gf_set_fill_mode;

	std::function<void(float width)> gf_set_line_width;

	std::function<void(material* material_def, float rad)> gf_sphere;

	std::function<int(shader_type type, unsigned int flags)> gf_maybe_create_shader;
	std::function<void(const std::function<void(size_t, size_t)>& progress_callback)> gf_recompile_all_shaders;

	std::function<void()> gf_clear_states;

	std::function<void(int bitmap_handle, int bpp, const ubyte* data, int width, int height)> gf_update_texture;
	// Reads the texture uploaded for a bitmap back from graphics memory.  Returns a vm_malloc'd buffer
	// (freed by the caller with vm_free) and reports the dimensions of the returned data, which may be
	// smaller than the bitmap if the texture was culled at upload; returns nullptr on failure.
	std::function<ubyte*(int bitmap_num, int* width_out, int* height_out)> gf_get_bitmap_from_texture;

	std::function<void(matrix4* shadow_view_matrix, const matrix* light_matrix, vec3d* eye_pos, bool first_pass)> gf_shadow_map_start;
	std::function<void()> gf_shadow_map_end;

	std::function<void()> gf_start_decal_pass;
	std::function<void()> gf_stop_decal_pass;

	// new drawing functions
	std::function<
		void(model_material* material_info, indexed_vertex_source* vert_source, vertex_buffer* bufferp, size_t texi)>
		gf_render_model;
	std::function<void(gr_buffer_handle ubo_handle, size_t ubo_offset, size_t ubo_size,
		vertex_buffer* buffer, indexed_vertex_source* vert_src, size_t texi)>
		gf_render_shadow_draw;
	std::function<void(shield_material* material_info,
		primitive_type prim_type,
		vertex_layout* layout,
		gr_buffer_handle buffer_handle,
		int n_verts)>
		gf_render_shield_impact;
	std::function<void(material* material_info,
		primitive_type prim_type,
		vertex_layout* layout,
		int offset,
		int n_verts,
		gr_buffer_handle buffer_handle,
		size_t buffer_offset)>
		gf_render_primitives;
	std::function<void(particle_material* material_info,
		primitive_type prim_type,
		vertex_layout* layout,
		int offset,
		int n_verts,
		gr_buffer_handle buffer_handle)>
		gf_render_primitives_particle;
	std::function<void(distortion_material* material_info,
		primitive_type prim_type,
		vertex_layout* layout,
		int offset,
		int n_verts,
		gr_buffer_handle buffer_handle)>
		gf_render_primitives_distortion;
	std::function<void(movie_material* material_info,
		primitive_type prim_type,
		vertex_layout* layout,
		int n_verts,
		gr_buffer_handle buffer,
		size_t buffer_offset)>
		gf_render_movie;
	std::function<void(batched_bitmap_material* material_info,
		primitive_type prim_type,
		vertex_layout* layout,
		int offset,
		int n_verts,
		gr_buffer_handle buffer_handle)>
		gf_render_primitives_batched;
	std::function<void(nanovg_material* material_info,
		primitive_type prim_type,
		vertex_layout* layout,
		int offset,
		int n_verts,
		gr_buffer_handle buffer_handle)>
		gf_render_nanovg;
	std::function<void(decal_material* material_info,
		primitive_type prim_type,
		vertex_layout* layout,
		int num_elements,
		const indexed_vertex_source& buffers,
		const gr_buffer_handle& instance_buffer,
		int num_instances)>
		gf_render_decals;
	void (*gf_render_rocket_primitives)(interface_material* material_info,
		primitive_type prim_type,
		vertex_layout* layout,
		int n_indices,
		gr_buffer_handle vertex_buffer,
		gr_buffer_handle index_buffer);

	std::function<bool(gr_capability capability)> gf_is_capable;
	std::function<bool(gr_property property, void* destination)> gf_get_property;

	std::function<void(const char* name)> gf_push_debug_group;
	std::function<void()> gf_pop_debug_group;

	std::function<int()> gf_create_query_object;
	std::function<void(int obj, QueryType type)> gf_query_value;
	std::function<bool(int obj)> gf_query_value_available;
	std::function<uint64_t(int obj)> gf_get_query_value;
	std::function<void(int obj)> gf_delete_query_object;

	std::unique_ptr<os::Viewport> (*gf_create_viewport)(const os::ViewPortProperties& props);
	std::function<void(os::Viewport* view)> gf_use_viewport;

	std::function<void(uniform_block_type bind_point, size_t offset, size_t size, gr_buffer_handle buffer)>
		gf_bind_uniform_buffer;

	std::function<gr_sync()> gf_sync_fence;
	std::function<bool(gr_sync sync, uint64_t timeoutns)> gf_sync_wait;
	std::function<void(gr_sync sync)> gf_sync_delete;

	std::function<void(int x, int y, int width, int height)> gf_set_viewport;

	std::function<void(bool set_override)> gf_override_fog;

	// ImGui backend integration
	std::function<void()> gf_imgui_new_frame;
	std::function<void()> gf_imgui_render_draw_data;

	//OpenXR functions
	std::function<SCP_vector<const char*>()> gf_openxr_get_extensions;
	std::function<bool()> gf_openxr_test_capabilities;
	std::function<bool()> gf_openxr_create_session;
	std::function<int64_t(const SCP_vector<int64_t>&)> gf_openxr_get_swapchain_format;
	std::function<bool()> gf_openxr_acquire_swapchain_buffers;
	std::function<bool()> gf_openxr_flip;
} screen;

/**
 * @brief Scripting context render values
 */
typedef struct lua_screen {
	bool active = false;
	bool force_fso_context = false;
	color current_color = {
		0,   // is_alphacolor
		0,   // alphacolor
		0,   // magic
		255, // red
		255, // green
		255, // blue
		255, // alpha
		0,   // ac_type
		0    // raw8
	};
	float line_width = 1.0f;
	int current_font_index = 0;
} lua_screen;

extern lua_screen gr_lua_screen;

bool gr_lua_context_active();

// Macros to easily choose been which context to use for color and line width
// Note that font is handled slightly differently for the normal game context by using it's own global in the FontManager namespace
// So FontManager::getCurrentFontIndex() is effectively its own macro

// Gets the current color between the game context and the lua context if active
#define GR_CURRENT_COLOR (gr_lua_context_active() ? gr_lua_screen.current_color : gr_screen.current_color)

// Gets the current line width between the game context and the lua contet if active
#define GR_CURRENT_LINE_WIDTH (gr_lua_context_active() ? gr_lua_screen.line_width : gr_screen.line_width)


// handy macro
#define GR_MAYBE_CLEAR_RES(bmap)		do  { int bmw = -1; int bmh = -1; if(bmap != -1){ bm_get_info( bmap, &bmw, &bmh, NULL, NULL, NULL); if((bmw != gr_screen.max_w) || (bmh != gr_screen.max_h)){gr_clear();} } else {gr_clear();} } while(false);

//Window's interface to set up graphics:
//--------------------------------------
// Call this at application startup

#define GR_DEFAULT (-1)

// resolution constants   - always keep resolutions in ascending order and starting from 0  
#define GR_NUM_RESOLUTIONS			2
#define GR_640							0		// 640 x 480
#define GR_1024						1		// 1024 x 768

#define GR_1024_THRESHOLD_WIDTH		1024
#define GR_1024_THRESHOLD_HEIGHT	600

extern const char *Resolution_prefixes[GR_NUM_RESOLUTIONS];

extern bool gr_init(std::unique_ptr<os::GraphicsOperations>&& graphicsOps, GraphicsAPI d_mode = GraphicsAPI::Default,
					int d_width = GR_DEFAULT, int d_height = GR_DEFAULT, int d_depth = GR_DEFAULT);

// The render API (GraphicsAPI::OpenGL/GraphicsAPI::Vulkan) selected via the "Graphics.RenderAPI" in-game option (or a mod's
// default settings table). The -vulkan/-opengl command line flags still win over this; see Cmdline_graphics_api.
extern GraphicsAPI gr_get_configured_render_api();

// The display name of a render API, e.g. "OpenGL". Returns "Unknown???" for GraphicsAPI values that are not
// a real backend, such as GraphicsAPI::Stub.
extern const char* gr_render_api_name(GraphicsAPI api);

extern void gr_screen_resize(int width, int height);
extern int gr_get_resolution_class(int width, int height);

/**
 * @brief Converts a position or delta from OS window pixels into internal render-target pixels
 *
 * @details When -window_res is in effect the frame is rendered into an offscreen buffer sized
 * gr_screen.max_w/max_h and only stretched to the window when we flip, so everything the engine
 * draws lives in render space while SDL reports input in window space. These are pure scale
 * factors with no offset, so they are equally valid for absolute positions and for deltas.
 *
 * Note that this is unrelated to the gr_resize_screen_pos() family, which handles the retail
 * 640/1024 menu coordinate system.
 */
void gr_window_to_render_pos(float& x, float& y);

/**
 * @brief Converts a position or delta from internal render-target pixels into OS window pixels
 *
 * @see gr_window_to_render_pos
 */
void gr_render_to_window_pos(float& x, float& y);

// Call this when your app ends.
extern void gr_close();

extern screen gr_screen;

#define GR_FILL_MODE_WIRE 1
#define GR_FILL_MODE_SOLID 2

#define GR_ZBUFF_NONE	0
#define GR_ZBUFF_WRITE	(1<<0)
#define GR_ZBUFF_READ	(1<<1)
#define GR_ZBUFF_FULL	(GR_ZBUFF_WRITE|GR_ZBUFF_READ)

#define GR_STENCIL_NONE		0
#define GR_STENCIL_READ		1
#define GR_STENCIL_WRITE	2

#define GR_RESIZE_NONE				0
#define GR_RESIZE_FULL				1
#define GR_RESIZE_FULL_CENTER		2
#define GR_RESIZE_MENU				3
#define GR_RESIZE_MENU_ZOOMED		4
#define GR_RESIZE_MENU_NO_OFFSET	5
#define GR_RESIZE_REPLACE			6

void gr_set_screen_scale(int x, int y, int zoom_x = -1, int zoom_y = -1, int max_x = gr_screen.max_w, int max_y = gr_screen.max_h, int center_x = gr_screen.center_w, int center_y = gr_screen.center_h, bool force_stretch = false);
void gr_reset_screen_scale();
bool gr_unsize_screen_pos(int *x, int *y, int *w = NULL, int *h = NULL, int resize_mode = GR_RESIZE_FULL);
bool gr_resize_screen_pos(int *x, int *y, int *w = NULL, int *h = NULL, int resize_mode = GR_RESIZE_FULL);
bool gr_unsize_screen_posf(float *x, float *y, float *w = NULL, float *h = NULL, int resize_mode = GR_RESIZE_FULL);
bool gr_resize_screen_posf(float *x, float *y, float *w = NULL, float *h = NULL, int resize_mode = GR_RESIZE_FULL);

// Does formatted printing.  This calls gr_string after formatting,
// so if you don't need to format the string, then call gr_string
// directly.
extern void gr_printf(int x, int y, SCP_FORMAT_STRING const char *format, ...) SCP_FORMAT_STRING_ARGS(3, 4);
extern void gr_printf(int x, int y, size_t len, SCP_FORMAT_STRING const char *format, ...) SCP_FORMAT_STRING_ARGS(4, 5);
// same as gr_printf but positions text correctly in menus
extern void gr_printf_menu(int x, int y, SCP_FORMAT_STRING const char *format, ...) SCP_FORMAT_STRING_ARGS(3, 4);
extern void gr_printf_menu(int x, int y, size_t len, SCP_FORMAT_STRING const char *format, ...) SCP_FORMAT_STRING_ARGS(4, 5);
// same as gr_printf_menu but accounts for menu zooming
extern void gr_printf_menu_zoomed(int x, int y, SCP_FORMAT_STRING const char *format, ...) SCP_FORMAT_STRING_ARGS(3, 4);
extern void gr_printf_menu_zoomed(int x, int y, size_t len, SCP_FORMAT_STRING const char *format, ...) SCP_FORMAT_STRING_ARGS(4, 5);
// same as gr_printf but doesn't resize for non-standard resolutions
extern void gr_printf_no_resize(int x, int y, SCP_FORMAT_STRING const char *format, ...) SCP_FORMAT_STRING_ARGS(3, 4);
extern void gr_printf_no_resize(int x, int y, size_t len, SCP_FORMAT_STRING const char *format, ...) SCP_FORMAT_STRING_ARGS(4, 5);

// Returns the size of the string in pixels in w and h
extern void gr_get_string_size( int *w, int *h, const char * text, float scaleMultiplier = 1.0f, size_t len = std::string::npos);

// Returns the height of the current font
extern int gr_get_font_height();

// Returns a scaled amount of lines per the current font --wookieejedi
extern int gr_get_dynamic_font_lines(int number_default_lines);

extern io::mouse::Cursor* Web_cursor;

// Called by OS when application gets/looses focus
extern void gr_activate(bool active);

#define GR_CALL(x) (x)

// These macros make the function indirection look like the
// old Descent-style gr_xxx calls.

#define gr_print_screen		GR_CALL(gr_screen.gf_print_screen)
#define gr_blob_screen		GR_CALL(gr_screen.gf_blob_screen)
#define gr_dump_envmap		GR_CALL(gr_screen.gf_dump_envmap)

void gr_request_screenshot(const char* filename);
bool gr_is_screenshot_requested();

//#define gr_flip				GR_CALL(gr_screen.gf_flip)
void gr_flip(bool execute_scripting = true);

inline void gr_setup_frame() {
	gr_screen.gf_setup_frame();
}

//#define gr_set_clip			GR_CALL(gr_screen.gf_set_clip)
inline void gr_set_clip(int x, int y, int w, int h, int resize_mode=GR_RESIZE_FULL)
{
	gr_screen.gf_set_clip(x, y, w, h, resize_mode);
}
#define gr_reset_clip		GR_CALL(gr_screen.gf_reset_clip)

void gr_set_bitmap(int bitmap_num, int alphablend = GR_ALPHABLEND_NONE, int bitbltmode = GR_BITBLT_MODE_NORMAL, float alpha = 1.0f);

#define gr_clear				GR_CALL(gr_screen.gf_clear)

#define gr_zbuffer_get		GR_CALL(gr_screen.gf_zbuffer_get)
#define gr_zbuffer_set		GR_CALL(gr_screen.gf_zbuffer_set)
#define gr_zbuffer_clear	GR_CALL(gr_screen.gf_zbuffer_clear)

#define gr_stencil_set		GR_CALL(gr_screen.gf_stencil_set)
#define gr_stencil_clear	GR_CALL(gr_screen.gf_stencil_clear)

#define gr_alpha_mask_set	GR_CALL(gr_screen.gf_alpha_mask_set)

#define gr_save_screen		GR_CALL(gr_screen.gf_save_screen)
#define gr_restore_screen	GR_CALL(gr_screen.gf_restore_screen)
#define gr_free_screen		GR_CALL(gr_screen.gf_free_screen)

#define gr_get_region		GR_CALL(gr_screen.gf_get_region)

#define gr_set_cull			GR_CALL(gr_screen.gf_set_cull)
#define gr_set_color_buffer	GR_CALL(gr_screen.gf_set_color_buffer)


#define gr_preload			GR_CALL(gr_screen.gf_preload)

#define gr_set_clear_color	GR_CALL(gr_screen.gf_set_clear_color)

// Here be the bitmap functions
#define gr_bm_free_data				GR_CALL(gr_screen.gf_bm_free_data)
#define gr_bm_create				GR_CALL(gr_screen.gf_bm_create)
#define gr_bm_init					GR_CALL(gr_screen.gf_bm_init)
#define gr_bm_page_in_start			GR_CALL(gr_screen.gf_bm_page_in_start)
#define gr_bm_data					GR_CALL(gr_screen.gf_bm_data)

#define gr_bm_make_render_target					GR_CALL(gr_screen.gf_bm_make_render_target)

inline int gr_bm_set_render_target(int n, int face = -1)
{
	return gr_screen.gf_bm_set_render_target(n, face);
}

#define gr_set_texture_addressing GR_CALL(gr_screen.gf_set_texture_addressing)

/**
 * @brief Create a GPU buffer
 * @param type The buffer type (Vertex, Index, Uniform)
 * @param usage Usage hint for optimization
 * @param rt_capable If true (Vulkan only), the buffer is created with the extra
 *        usage flags (VK_BUFFER_USAGE_SHADER_DEVICE_ADDRESS_BIT and
 *        VK_BUFFER_USAGE_ACCELERATION_STRUCTURE_BUILD_INPUT_READ_ONLY_BIT_KHR)
 *        needed to use it as acceleration structure geometry input. Ignored by
 *        backends without raytraced shadow support.
 */
inline gr_buffer_handle gr_create_buffer(BufferType type, BufferUsageHint usage, bool rt_capable = false)
{
	return gr_screen.gf_create_buffer(type, usage, rt_capable);
}

#define gr_delete_buffer GR_CALL(gr_screen.gf_delete_buffer)

/**
 * @brief Notify the graphics backend that a model finished loading (its GPU
 * vertex/index buffers are already uploaded at this point). No-op on backends
 * that don't need per-model GPU state.
 */
inline void gr_model_loaded(int pm_id)
{
	gr_screen.gf_model_loaded(pm_id);
}

/**
 * @brief Notify the graphics backend that a model is about to be freed, so it
 * can release any auxiliary per-model GPU state before the model's own GPU
 * buffers are destroyed.
 */
inline void gr_model_unloaded(int pm_id)
{
	gr_screen.gf_model_unloaded(pm_id);
}

#define gr_build_shadow_tlas GR_CALL(gr_screen.gf_build_shadow_tlas)

#define gr_update_buffer_data GR_CALL(gr_screen.gf_update_buffer_data)
#define gr_update_buffer_data_offset GR_CALL(gr_screen.gf_update_buffer_data_offset)
inline void* gr_map_buffer(gr_buffer_handle handle)
{
	return gr_screen.gf_map_buffer(handle);
}
inline void gr_flush_mapped_buffer(gr_buffer_handle handle, size_t offset, size_t size)
{
	gr_screen.gf_flush_mapped_buffer(handle, offset, size);
}
#define gr_update_transform_buffer GR_CALL(gr_screen.gf_update_transform_buffer)

#define gr_scene_texture_begin GR_CALL(gr_screen.gf_scene_texture_begin)
#define gr_scene_texture_end GR_CALL(gr_screen.gf_scene_texture_end)
#define gr_copy_effect_texture GR_CALL(gr_screen.gf_copy_effect_texture)

#define gr_post_process_set_effect GR_CALL(gr_screen.gf_post_process_set_effect)
#define gr_post_process_set_defaults GR_CALL(gr_screen.gf_post_process_set_defaults)
#define gr_post_process_begin GR_CALL(gr_screen.gf_post_process_begin)
#define gr_post_process_end GR_CALL(gr_screen.gf_post_process_end)
#define gr_post_process_save_zbuffer GR_CALL(gr_screen.gf_post_process_save_zbuffer)
inline void gr_post_process_restore_zbuffer()
{
	gr_screen.gf_post_process_restore_zbuffer();
}

#define gr_deferred_lighting_begin		GR_CALL(gr_screen.gf_deferred_lighting_begin)
#define gr_deferred_lighting_msaa		GR_CALL(gr_screen.gf_deferred_lighting_msaa)
#define gr_deferred_lighting_end		GR_CALL(gr_screen.gf_deferred_lighting_end)
#define gr_deferred_lighting_finish		GR_CALL(gr_screen.gf_deferred_lighting_finish)

#define	gr_zbias						GR_CALL(gr_screen.gf_zbias)
#define	gr_set_fill_mode				GR_CALL(gr_screen.gf_set_fill_mode)

#define gr_set_line_width				GR_CALL(gr_screen.gf_set_line_width)

#define gr_sphere						GR_CALL(gr_screen.gf_sphere)

#define gr_maybe_create_shader			GR_CALL(gr_screen.gf_maybe_create_shader)
#define gr_recompile_all_shaders		GR_CALL(gr_screen.gf_recompile_all_shaders)

#define gr_clear_states					GR_CALL(gr_screen.gf_clear_states)

#define gr_update_texture				GR_CALL(gr_screen.gf_update_texture)
#define gr_get_bitmap_from_texture		GR_CALL(gr_screen.gf_get_bitmap_from_texture)

#define gr_shadow_map_start				GR_CALL(gr_screen.gf_shadow_map_start)
#define gr_shadow_map_end				GR_CALL(gr_screen.gf_shadow_map_end)
#define gr_render_shield_impact			GR_CALL(gr_screen.gf_render_shield_impact)

#define gr_override_fog					GR_CALL(gr_screen.gf_override_fog)

#define gr_imgui_new_frame				GR_CALL(gr_screen.gf_imgui_new_frame)
#define gr_imgui_render_draw_data		GR_CALL(gr_screen.gf_imgui_render_draw_data)

/**
 * @brief Starts an ImGui frame, sized to the render target rather than to the OS window
 *
 * @details Drives the renderer backend, the SDL platform backend and ImGui itself, then corrects
 * the display size the platform backend derived from the window. Must be paired with
 * ImGui::Render() and gr_imgui_render_draw_data().
 */
void gr_imgui_begin_frame();

inline void gr_render_primitives(material* material_info,
	primitive_type prim_type,
	vertex_layout* layout,
	int vert_offset,
	int n_verts,
	gr_buffer_handle buffer_handle = gr_buffer_handle(),
	size_t buffer_offset = 0)
{
	gr_screen
		.gf_render_primitives(material_info, prim_type, layout, vert_offset, n_verts, buffer_handle, buffer_offset);
}

inline void gr_render_primitives_particle(particle_material* material_info,
	primitive_type prim_type,
	vertex_layout* layout,
	int offset,
	int n_verts,
	gr_buffer_handle buffer_handle = gr_buffer_handle())
{
	gr_screen.gf_render_primitives_particle(material_info, prim_type, layout, offset, n_verts, buffer_handle);
}

inline void gr_render_primitives_batched(batched_bitmap_material* material_info,
	primitive_type prim_type,
	vertex_layout* layout,
	int offset,
	int n_verts,
	gr_buffer_handle buffer_handle = gr_buffer_handle())
{
	gr_screen.gf_render_primitives_batched(material_info, prim_type, layout, offset, n_verts, buffer_handle);
}

inline void gr_render_nanovg(nanovg_material* material_info,
	primitive_type prim_type,
	vertex_layout* layout,
	int offset,
	int n_verts,
	gr_buffer_handle buffer_handle)
{
	gr_screen.gf_render_nanovg(material_info, prim_type, layout, offset, n_verts, buffer_handle);
}

inline void gr_render_primitives_distortion(distortion_material* material_info,
	primitive_type prim_type,
	vertex_layout* layout,
	int offset,
	int n_verts,
	gr_buffer_handle buffer_handle = gr_buffer_handle())
{
	gr_screen.gf_render_primitives_distortion(material_info, prim_type, layout, offset, n_verts, buffer_handle);
}

inline void gr_render_movie(movie_material* material_info,
	primitive_type prim_type,
	vertex_layout* layout,
	int n_verts,
	gr_buffer_handle buffer,
	size_t buffer_offset = 0)
{
	gr_screen.gf_render_movie(material_info, prim_type, layout, n_verts, buffer, buffer_offset);
}

inline void gr_render_model(model_material* material_info, indexed_vertex_source *vert_source, vertex_buffer* bufferp, size_t texi)
{
	gr_screen.gf_render_model(material_info, vert_source, bufferp, texi);
}

inline void gr_render_shadow_draw(gr_buffer_handle ubo_handle, size_t ubo_offset, size_t ubo_size,
                                   vertex_buffer* buffer, indexed_vertex_source* vert_src, size_t texi)
{
	gr_screen.gf_render_shadow_draw(ubo_handle, ubo_offset, ubo_size, buffer, vert_src, texi);
}

inline void gr_render_rocket_primitives(interface_material* material_info,
	primitive_type prim_type,
	vertex_layout* layout,
	int n_indices,
	gr_buffer_handle vertex_buffer,
	gr_buffer_handle index_buffer)
{
	gr_screen.gf_render_rocket_primitives(material_info, prim_type, layout, n_indices, vertex_buffer, index_buffer);
}

inline bool gr_is_capable(gr_capability capability)
{
	return gr_screen.gf_is_capable(capability);
}

inline bool gr_get_property(gr_property property, void* destination)
{
	return gr_screen.gf_get_property(property, destination);
}

inline void gr_push_debug_group(const char* name)
{
	gr_screen.gf_push_debug_group(name);
}

inline void gr_pop_debug_group()
{
	gr_screen.gf_pop_debug_group();
}

inline int gr_create_query_object()
{
	return gr_screen.gf_create_query_object();
}

inline void gr_query_value(int obj, QueryType type)
{
	gr_screen.gf_query_value(obj, type);
}

inline bool gr_query_value_available(int obj)
{
	return gr_screen.gf_query_value_available(obj);
}

inline std::uint64_t gr_get_query_value(int obj)
{
	return gr_screen.gf_get_query_value(obj);
}

inline void gr_delete_query_object(int obj)
{
	gr_screen.gf_delete_query_object(obj);
}

inline std::unique_ptr<os::Viewport> gr_create_viewport(const os::ViewPortProperties& props)
{
	return gr_screen.gf_create_viewport(props);
}
inline void gr_use_viewport(os::Viewport* view)
{
	gr_screen.gf_use_viewport(view);
}
inline void gr_set_viewport(int x, int y, int width, int height)
{
	gr_screen.gf_set_viewport(x, y, width, height);
}

inline void gr_bind_uniform_buffer(uniform_block_type bind_point, size_t offset, size_t size, gr_buffer_handle buffer)
{
	gr_screen.gf_bind_uniform_buffer(bind_point, offset, size, buffer);
}

inline gr_sync gr_sync_fence()
{
	return gr_screen.gf_sync_fence();
}
inline bool gr_sync_wait(gr_sync sync, uint64_t timeoutns)
{
	return gr_screen.gf_sync_wait(sync, timeoutns);
}
inline void gr_sync_delete(gr_sync sync)
{
	gr_screen.gf_sync_delete(sync);
}

//OpenXR
#define gr_openxr_get_extensions GR_CALL(gr_screen.gf_openxr_get_extensions)
#define gr_openxr_test_capabilities GR_CALL(gr_screen.gf_openxr_test_capabilities)
#define gr_openxr_create_session GR_CALL(gr_screen.gf_openxr_create_session)
#define gr_openxr_get_swapchain_format GR_CALL(gr_screen.gf_openxr_get_swapchain_format)
#define gr_openxr_acquire_swapchain_buffers GR_CALL(gr_screen.gf_openxr_acquire_swapchain_buffers)
#define gr_openxr_flip GR_CALL(gr_screen.gf_openxr_flip)

// color functions
void gr_init_color(color *c, int r, int g, int b);
void gr_init_alphacolor( color *clr, int r, int g, int b, int alpha, int type = AC_TYPE_HUD );
void gr_set_color( int r, int g, int b );
void gr_set_color_fast(const color *dst);
bool gr_compare_color_values(const color& clr1, const color& clr2);

// shader functions
void gr_create_shader(shader *shade, ubyte r, ubyte g, ubyte b, ubyte c);
void gr_set_shader(shader *shade);

// new bitmap functions
void gr_bitmap(int x, int y, int resize_mode = GR_RESIZE_FULL, bool mirror = false, float scale_factor = 1.0f);
void gr_bitmap_uv(int _x, int _y, int _w, int _h, float _u0, float _v0, float _u1, float _v1, int resize_mode = GR_RESIZE_FULL);

// special function for drawing polylines. this function is specifically intended for
// polylines where each section is no more than 90 degrees away from a previous section.
// Moreover, it is _really_ intended for use with 45 degree angles. 
void gr_pline_special(SCP_vector<vec3d> *pts, int thickness,int resize_mode=GR_RESIZE_FULL);

#define VB_FLAG_POSITION	(1<<0)
#define VB_FLAG_RHW			(1<<1)	//incompatable with the next normal
#define VB_FLAG_NORMAL		(1<<2)
#define VB_FLAG_DIFUSE		(1<<3)
#define VB_FLAG_SPECULAR	(1<<4)
#define VB_FLAG_UV1			(1<<5)	//how many UV coords, only use one of these
#define VB_FLAG_UV2			(1<<6)
#define VB_FLAG_UV3			(1<<7)
#define VB_FLAG_UV4			(1<<8)
#define VB_FLAG_TANGENT		(1<<9)
#define VB_FLAG_LARGE_INDEX	(1<<10)
#define VB_FLAG_MODEL_ID	(1<<11)
#define VB_FLAG_TRANS		(1<<12)

/**
* @brief Prints the current time
*
* Draws the timestamp of the current #Missiontime in the format @c h:mm:ss at the specified coordinates
*
* @param x The x position where the timestamp should be draw
* @param y The y position where the timestamp should be draw
* @param timestamp The timestamp (in 65536ths of a second) to be printed
* @param resize_mode The resize mode to use
*/
void gr_print_timestamp(int x, int y, fix timestamp, int resize_mode);

namespace graphics {
class DebugScope {
 public:
	explicit DebugScope(const char* name) {
		gr_push_debug_group(name);
	}
	~DebugScope() {
		gr_pop_debug_group();
	}
};
}

#if !defined(NDEBUG) || defined(FS_OPENGL_DEBUG)
#define GR_DEBUG_SCOPE(name) ::graphics::DebugScope SCP_TOKEN_CONCAT(gr_scope, __LINE__)(name)
#else
#define GR_DEBUG_SCOPE(name) do {} while(false)
#endif

enum AnimatedShader {
	ANIMATED_SHADER_LOADOUTSELECT_FS1= 0,
	ANIMATED_SHADER_LOADOUTSELECT_FS2= 1,
	ANIMATED_SHADER_CLOAK = 2,
};

/**
 * @brief Retreives a uniform buffer for storing uniform block data
 * @param type The type of uniform data that will be stored in the buffer
 * @param num_elements The number of elements that will be used in the buffer
 * @param element_size_override Override what the size of the element should be instead of relying on the preconfigured
 * size for that block type
 * @return A structure which gives access to a memory buffer where the uniform data can be stored
 */
graphics::util::UniformBuffer gr_get_uniform_buffer(uniform_block_type type, size_t num_elements,
                                                    size_t element_size_override = 0);

struct VideoModeData {
	uint32_t width = 0;
	uint32_t height = 0;
	uint32_t bit_depth = 0;
};

struct DisplayData {
	uint32_t index;
	SCP_string name;

	int32_t x = 0;
	int32_t y = 0;
	int32_t width = 0;
	int32_t height = 0;

	SCP_vector<VideoModeData> video_modes;
};

SCP_vector<DisplayData> gr_enumerate_displays();

enum class GpuHeap {
	ModelVertex = 0,
	ModelIndex,

	NUM_VALUES
};

/**
 * @brief Allocates storage on the specified GPU heap and stores data in that storage
 * @param heap_type The heap type to store this memory on
 * @param size The size of the data
 * @param data A pointer to the data
 * @param[out] offset_out The offset at which the data is stored in the buffer
 * @param[out] handle_out The handle of the buffer object this data is stored in
 */
void gr_heap_allocate(GpuHeap heap_type, size_t size, void* data, size_t& offset_out, gr_buffer_handle& handle_out);

/**
 * @brief Deallocates memory previously allocated with gr_heap_allocate.
 * @param heap_type The heap type to deallocate this memory from
 * @param data_offset The offset at which the data is stored.
 */
void gr_heap_deallocate(GpuHeap heap_type, size_t data_offset);

void gr_set_gamma(float gamma);

void gr_get_post_process_effect_names(SCP_vector<SCP_string> &names);

bool gr_is_viewport_window();

void gr_uniform_buffer_managers_init();

SDL_DisplayID gr_get_preferred_display();

// Include this last to make the 2D rendering function available everywhere
#include "graphics/render.h"

#endif