irrlicht/include/SMaterialLayer.h
cutealien 9ce63bc7d3 Fix SMaterialLayer operator!= and optimize operator=
SMaterialLayers are now identical when both have identity matrices.
Before it didn't consider them identical when one layer had set the identity matrix explicitly and the other didn't.
Generally didn't matter, just caused very rarely some extra state switches in the drivers. And just as rarely had a cheaper comparison. Just seems more correct this way.

operator= no longer releases texture memory which was allocated at one point. 
Unless explicitly requested such memory is now always released later in the destructor.
This can avoid quite a few memory allocations/released in the driver. Usually not a noticeable performance difference on most platforms. But it can help avoid memory fragmentation.
We instead use an extra bool now to tell if the texture memory is used. So slight increase in SMaterialLayer and SMaterial size. But I did a quick performance test and this had no negative influence here, while it did improve speed in the case where it switched between material layers using/not using texture matrices a bit.

git-svn-id: svn://svn.code.sf.net/p/irrlicht/code/trunk@6488 dfc29bdd-3216-0410-991c-e03cc46cb475
2023-05-07 14:51:09 +00:00

259 lines
7.6 KiB
C++

// Copyright (C) 2002-2012 Nikolaus Gebhardt
// This file is part of the "Irrlicht Engine".
// For conditions of distribution and use, see copyright notice in irrlicht.h
#ifndef S_MATERIAL_LAYER_H_INCLUDED
#define S_MATERIAL_LAYER_H_INCLUDED
#include "matrix4.h"
#include "irrAllocator.h"
namespace irr
{
namespace video
{
class ITexture;
//! Texture coord clamp mode outside [0.0, 1.0]
enum E_TEXTURE_CLAMP
{
//! Texture repeats
ETC_REPEAT = 0,
//! Texture is clamped to the last pixel
ETC_CLAMP,
//! Texture is clamped to the edge pixel
ETC_CLAMP_TO_EDGE,
//! Texture is clamped to the border pixel (if exists)
ETC_CLAMP_TO_BORDER,
//! Texture is alternatingly mirrored (0..1..0..1..0..)
ETC_MIRROR,
//! Texture is mirrored once and then clamped (0..1..0)
ETC_MIRROR_CLAMP,
//! Texture is mirrored once and then clamped to edge
ETC_MIRROR_CLAMP_TO_EDGE,
//! Texture is mirrored once and then clamped to border
ETC_MIRROR_CLAMP_TO_BORDER
};
static const char* const aTextureClampNames[] = {
"texture_clamp_repeat",
"texture_clamp_clamp",
"texture_clamp_clamp_to_edge",
"texture_clamp_clamp_to_border",
"texture_clamp_mirror",
"texture_clamp_mirror_clamp",
"texture_clamp_mirror_clamp_to_edge",
"texture_clamp_mirror_clamp_to_border", 0};
//! Struct for holding material parameters which exist per texture layer
// Note for implementers: Serialization is in CNullDriver
class SMaterialLayer
{
public:
//! Default constructor
SMaterialLayer() : Texture(0), TextureWrapU(ETC_REPEAT), TextureWrapV(ETC_REPEAT), TextureWrapW(ETC_REPEAT),
BilinearFilter(true), TrilinearFilter(false), AnisotropicFilter(0), LODBias(0),
TextureMatrix(0), TextureMatrixUsed(false)
{
}
//! Copy constructor
/** \param other Material layer to copy from. */
SMaterialLayer(const SMaterialLayer& other)
{
// This pointer is checked during assignment
TextureMatrix = 0;
*this = other;
}
//! Destructor
~SMaterialLayer()
{
if ( TextureMatrix )
{
MatrixAllocator.destruct(TextureMatrix);
MatrixAllocator.deallocate(TextureMatrix);
}
}
//! Assignment operator
/** \param other Material layer to copy from.
\return This material layer, updated. */
SMaterialLayer& operator=(const SMaterialLayer& other)
{
// Check for self-assignment!
if (this == &other)
return *this;
Texture = other.Texture;
if (other.TextureMatrixUsed)
{
if (TextureMatrix)
{
*TextureMatrix = *other.TextureMatrix;
}
else
{
TextureMatrix = MatrixAllocator.allocate(1);
MatrixAllocator.construct(TextureMatrix,*other.TextureMatrix);
}
TextureMatrixUsed = true;
}
else
{
TextureMatrixUsed = false;
}
TextureWrapU = other.TextureWrapU;
TextureWrapV = other.TextureWrapV;
TextureWrapW = other.TextureWrapW;
BilinearFilter = other.BilinearFilter;
TrilinearFilter = other.TrilinearFilter;
AnisotropicFilter = other.AnisotropicFilter;
LODBias = other.LODBias;
return *this;
}
//! Gets the texture transformation matrix
/** \return Texture matrix of this layer. */
core::matrix4& getTextureMatrix()
{
if (!TextureMatrix)
{
TextureMatrix = MatrixAllocator.allocate(1);
MatrixAllocator.construct(TextureMatrix,core::IdentityMatrix);
TextureMatrixUsed = true;
}
else if ( !TextureMatrixUsed )
{
*TextureMatrix = core::IdentityMatrix;
TextureMatrixUsed = true;
}
return *TextureMatrix;
}
//! Gets the immutable texture transformation matrix
/** \return Texture matrix of this layer. */
const core::matrix4& getTextureMatrix() const
{
if (TextureMatrixUsed)
return *TextureMatrix;
else
return core::IdentityMatrix;
}
//! Sets the texture transformation matrix to mat
/** NOTE: Pipelines can ignore this matrix when the
texture is 0.
\param mat New texture matrix for this layer. */
void setTextureMatrix(const core::matrix4& mat)
{
if (!TextureMatrix)
{
TextureMatrix = MatrixAllocator.allocate(1);
MatrixAllocator.construct(TextureMatrix,mat);
}
else
*TextureMatrix = mat;
TextureMatrixUsed = true;
}
//! Check if we have set a custom texture matrix
//! Note that otherwise we get an IdentityMatrix as default
inline bool hasSetTextureMatrix() const
{
return TextureMatrixUsed;
}
//! Reset texture matrix to identity matrix
/** \param releaseMemory Releases also texture memory. Otherwise done in destructor */
void resetTextureMatrix(bool releaseMemory=true)
{
if ( TextureMatrix && releaseMemory)
{
MatrixAllocator.destruct(TextureMatrix);
MatrixAllocator.deallocate(TextureMatrix);
TextureMatrix = 0;
}
TextureMatrixUsed = false;
}
//! Inequality operator
/** \param b Layer to compare to.
\return True if layers are different, else false. */
inline bool operator!=(const SMaterialLayer& b) const
{
bool different =
Texture != b.Texture ||
TextureWrapU != b.TextureWrapU ||
TextureWrapV != b.TextureWrapV ||
TextureWrapW != b.TextureWrapW ||
BilinearFilter != b.BilinearFilter ||
TrilinearFilter != b.TrilinearFilter ||
AnisotropicFilter != b.AnisotropicFilter ||
LODBias != b.LODBias;
if (different)
return true;
else
{
different = (TextureMatrixUsed && b.TextureMatrixUsed && (*TextureMatrix != *b.TextureMatrix))
|| (TextureMatrixUsed && !b.TextureMatrixUsed && (*TextureMatrix != core::IdentityMatrix))
|| (!TextureMatrixUsed && b.TextureMatrixUsed && (core::IdentityMatrix != *b.TextureMatrix));
}
return different;
}
//! Equality operator
/** \param b Layer to compare to.
\return True if layers are equal, else false. */
inline bool operator==(const SMaterialLayer& b) const
{ return !(b!=*this); }
//! Texture
ITexture* Texture;
//! Texture Clamp Mode
/** Values are taken from E_TEXTURE_CLAMP. */
u8 TextureWrapU:4;
u8 TextureWrapV:4;
u8 TextureWrapW:4;
//! Is bilinear filtering enabled? Default: true
bool BilinearFilter:1;
//! Is trilinear filtering enabled? Default: false
/** If the trilinear filter flag is enabled,
the bilinear filtering flag is ignored. */
bool TrilinearFilter:1;
//! Is anisotropic filtering enabled? Default: 0, disabled
/** In Irrlicht you can use anisotropic texture filtering
in conjunction with bilinear or trilinear texture
filtering to improve rendering results. Primitives
will look less blurry with this flag switched on. The number gives
the maximal anisotropy degree, and is often in the range 2-16.
Value 1 is equivalent to 0, but should be avoided. */
u8 AnisotropicFilter;
//! Bias for the mipmap choosing decision.
/** This value can make the textures more or less blurry than with the
default value of 0. The value (divided by 8.f) is added to the mipmap level
chosen initially, and thus takes a smaller mipmap for a region
if the value is positive. */
s8 LODBias;
private:
friend class SMaterial;
irr::core::irrAllocator<irr::core::matrix4> MatrixAllocator;
//! Texture Matrix
/** Do not access this element directly as the internal
resource management has to cope with Null pointers etc. */
core::matrix4* TextureMatrix;
bool TextureMatrixUsed; // TextureMatrix memory stays until destructor even when unused to avoid unnecessary allocation/de-allocations
};
} // end namespace video
} // end namespace irr
#endif // S_MATERIAL_LAYER_H_INCLUDED