Ogre:Hardwarebuffer

来源:互联网 发布:linux dd u盘启动 编辑:程序博客网 时间:2024/05/02 01:49

Ogre中的硬件缓存是指在显卡上的存储,这和在内存上的存储一样是可以访问的。有三种硬件缓存:HardwareVertexBuffer(顶点缓存,存储顶点的各种数据)、HardwareIndexBuffer(索引缓存,存储一个mesh的面片的顶点索引),HardwarePixelBuffer(纹理缓存,存储某个纹理贴图的数据)。这些数据在程序运行时都在显卡的存储上,然而你可以去读和写这些数据,来操控程序中物体的形状、纹理等。这个用处是非常大的。在Ogre中与访问这些硬件缓存有关的类及他们相互间的关系如下图: 
 

  1. hardwarebuffer 

根据这个图进行解释

1、最上面的hardwarevertexbuffer 
读写:如果mesh使用的所有子mesh共享buffer的形式,则用mesh的sharedvertexdata,否则用submesh的vertexdata来得到vertexdata结构,vertexdata封装了对该mesh的顶点缓存数据的访问方式,但是却不直接包含这些顶点缓存数据。vertexdata中的vettexbufferbinding可以知道当前的vertexdata对应了确切的硬件上的哪块buffer,可以通过vettexbufferbinding的getBuffer确切的得到该顶点缓存,而vertexdata中的vertexdeclaration则是一个对他对应的buffer进行各种访问的接口,里面有访问的格式等。如果要开始操纵这个buffer,需要将getbuffer得到的hardwarevertexbuffer调用lock,然后将这片缓存上锁,这个lock返回了一个void指针,指向的就是缓存数据。拿着这个指针就可以读取改写等

创建:使用hardwarebuffermanager的create来创建,创建后利用hardwarevertexbuffer的write写入数据

2.中间的hardwareindexbuffer 
读写:直接使用submesh的indexdata来得到一个indexdata结构,再调用它的hardwareindexbuffer的来得到这个顶点缓存,童年顶点缓存一样再调用lock来进行读写操作 
创建:同顶点缓存

3最下面的hardwarepixelbuffer

读写:从texture中可以直接得到这个hardwarepixelbuffer,然后对它lock后就可以得到一个pixelbox的数据,pixebox封装了所有纹理数据及其各种属性信息

创建:texture是由texturemanager创建的

下面是一些具体的使用硬件缓存的例子

读取顶点和索引缓存

    Ogre::MeshPtr meshPtr=mainEntity->getMesh(); 
   //假设这里使用的是share的形式 
    Ogre::VertexData* vertex_data=meshPtr->sharedVertexData;

   //得到位置数据的信息 
    const Ogre::VertexElement* posElem =vertex_data->vertexDeclaration->findElementBySemantic(Ogre::VES_POSITION);

//得到纹理坐标数据的信息 
    const Ogre::VertexElement* texcoElem=vertex_data->vertexDeclaration->findElementBySemantic(Ogre::VES_TEXTURE_COORDINATES ); 
    //得到位置和纹理的缓存 
    Ogre::HardwareVertexBufferSharedPtr posBuf =vertex_data->vertexBufferBinding->getBuffer(posElem->getSource()); 
    Ogre::HardwareVertexBufferSharedPtr texcoBuf =vertex_data->vertexBufferBinding->getBuffer(texcoElem->getSource()); 
     //顶点位置缓存的lock,读取 
    unsigned char* vertexPos =static_cast<unsigned char*>(posBuf->lock(Ogre::HardwareBuffer::HBL_READ_ONLY));

//将第一个点的位置读出 
    float* pReal;

//这个函数的作用是将当前vertexPos指向的数据用其他型(这里是float*)的指针指向,这样读出来的数据就是float型的了,或者用float型的数据进行写入 
    posElem->baseVertexPointerToElement(vertexPos, &pReal); 
   Ogre::Vector3 pt(pReal[0], pReal[1], pReal[2]); 
//访问之后要上锁 
    posBuf->unlock(); 
    
    //读取索引信息 
          Ogre::SubMesh* submesh = meshPtr->getSubMesh( i );

     //得到这个submesh的indexdata 
        Ogre::IndexData* index_data = submesh->indexData; 
        int numTris = index_data->indexCount / 3;

//得到indexbuffer 
        Ogre::HardwareIndexBufferSharedPtr ibuf = index_data->indexBuffer; 
        bool use32bitindexes = (ibuf->getType() == Ogre::HardwareIndexBuffer::IT_32BIT);

//得到具体的索引缓存数据 
        unsigned long*  pLong = static_cast<unsigned long*>(ibuf->lock(Ogre::HardwareBuffer::HBL_READ_ONLY)); 
        unsigned short* pShort = reinterpret_cast<unsigned short*>(pLong);

          … … 
          ibuf->unlock();

访问纹理缓存

    Ogre::HardwarePixelBufferSharedPtr crossPixbufferPtr=texture.getPointer()->getBuffer(0,0);        
    crossPixbufferPtr->lock(Ogre::HardwareBuffer::HBL_NORMAL); 
    Ogre::PixelBox pb=crossPixbufferPtr->getCurrentLock(); 
    int height = pb.getHeight(); 
    int width = pb.getWidth(); 
    int pitch = pb.rowPitch; // Skip between rows of image    
    uint32* data=static_cast<uint32*>(pb.data);

     ……操纵data…… 
    crossPixbufferPtr->unlock();

创建顶点缓存和索引缓存,进而根据其创建一个自定义的mesh

void createColourCube() 
    { 
    /// Create the mesh via the MeshManager 
    Ogre::MeshPtr msh = MeshManager::getSingleton().createManual("ColourCube", "General"); 
    /// Create one submesh 
    SubMesh* sub = msh->createSubMesh(); 
    const float sqrt13 = 0.577350269f; /* sqrt(1/3) */ 
    /// Define the vertices (8 vertices, each consisting of 2 groups of 3 floats 
    const size_t nVertices = 8; 
    const size_t vbufCount = 3*2*nVertices; 
    float vertices[vbufCount] = { 
            -100.0,100.0,-100.0,        //0 position 
            -sqrt13,sqrt13,-sqrt13,     //0 normal 
            100.0,100.0,-100.0,         //1 position 
            sqrt13,sqrt13,-sqrt13,      //1 normal 
            100.0,-100.0,-100.0,        //2 position 
            sqrt13,-sqrt13,-sqrt13,     //2 normal 
            -100.0,-100.0,-100.0,       //3 position 
            -sqrt13,-sqrt13,-sqrt13,    //3 normal 
            -100.0,100.0,100.0,         //4 position 
            -sqrt13,sqrt13,sqrt13,      //4 normal 
            100.0,100.0,100.0,          //5 position 
            sqrt13,sqrt13,sqrt13,       //5 normal 
            100.0,-100.0,100.0,         //6 position 
            sqrt13,-sqrt13,sqrt13,      //6 normal 
            -100.0,-100.0,100.0,        //7 position 
            -sqrt13,-sqrt13,sqrt13,     //7 normal 
    }; 
    RenderSystem* rs = Root::getSingleton().getRenderSystem(); 
    RGBA colours[nVertices]; 
    RGBA *pColour = colours; 
    // Use render system to convert colour value since colour packing varies 
    rs->convertColourValue(ColourValue(1.0,0.0,0.0), pColour++); //0 colour 
    rs->convertColourValue(ColourValue(1.0,1.0,0.0), pColour++); //1 colour 
    rs->convertColourValue(ColourValue(0.0,1.0,0.0), pColour++); //2 colour 
    rs->convertColourValue(ColourValue(0.0,0.0,0.0), pColour++); //3 colour 
    rs->convertColourValue(ColourValue(1.0,0.0,1.0), pColour++); //4 colour 
    rs->convertColourValue(ColourValue(1.0,1.0,1.0), pColour++); //5 colour 
    rs->convertColourValue(ColourValue(0.0,1.0,1.0), pColour++); //6 colour 
    rs->convertColourValue(ColourValue(0.0,0.0,1.0), pColour++); //7 colour 
    /// Define 12 triangles (two triangles per cube face) 
    /// The values in this table refer to vertices in the above table 
    const size_t ibufCount = 36; 
    unsigned short faces[ibufCount] = { 
            0,2,3, 
            0,1,2, 
            1,6,2, 
            1,5,6, 
            4,6,5, 
            4,7,6, 
            0,7,4, 
            0,3,7, 
            0,5,1, 
            0,4,5, 
            2,7,3, 
            2,6,7 
    }; 
    /// Create vertex data structure for 8 vertices shared between submeshes 
    msh->sharedVertexData = new VertexData(); 
    msh->sharedVertexData->vertexCount = nVertices; 
    /// Create declaration (memory format) of vertex data 
    VertexDeclaration* decl = msh->sharedVertexData->vertexDeclaration; 
    size_t offset = 0; 
    // 1st buffer 
    decl->addElement(0, offset, VET_FLOAT3, VES_POSITION); 
    offset += VertexElement::getTypeSize(VET_FLOAT3); 
    decl->addElement(0, offset, VET_FLOAT3, VES_NORMAL); 
    offset += VertexElement::getTypeSize(VET_FLOAT3); 
    /// Allocate vertex buffer of the requested number of vertices (vertexCount) 
    /// and bytes per vertex (offset) 
    HardwareVertexBufferSharedPtr vbuf = 
        HardwareBufferManager::getSingleton().createVertexBuffer( 
        offset, msh->sharedVertexData->vertexCount, HardwareBuffer::HBU_STATIC_WRITE_ONLY); 
    /// Upload the vertex data to the card 
    vbuf->writeData(0, vbuf->getSizeInBytes(), vertices, true); 
    /// Set vertex buffer binding so buffer 0 is bound to our vertex buffer 
    VertexBufferBinding* bind = msh->sharedVertexData->vertexBufferBinding; 
    bind->setBinding(0, vbuf); 
    // 2nd buffer 
    offset = 0; 
    decl->addElement(1, offset, VET_COLOUR, VES_DIFFUSE); 
    offset += VertexElement::getTypeSize(VET_COLOUR); 
    /// Allocate vertex buffer of the requested number of vertices (vertexCount) 
    /// and bytes per vertex (offset) 
    vbuf = HardwareBufferManager::getSingleton().createVertexBuffer( 
        offset, msh->sharedVertexData->vertexCount, HardwareBuffer::HBU_STATIC_WRITE_ONLY); 
    /// Upload the vertex data to the card 
    vbuf->writeData(0, vbuf->getSizeInBytes(), colours, true); 
    /// Set vertex buffer binding so buffer 1 is bound to our colour buffer 
    bind->setBinding(1, vbuf); 
    /// Allocate index buffer of the requested number of vertices (ibufCount) 
    HardwareIndexBufferSharedPtr ibuf = HardwareBufferManager::getSingleton(). 
        createIndexBuffer( 
        HardwareIndexBuffer::IT_16BIT, 
        ibufCount, 
        HardwareBuffer::HBU_STATIC_WRITE_ONLY); 
    /// Upload the index data to the card 
    ibuf->writeData(0, ibuf->getSizeInBytes(), faces, true); 
    /// Set parameters of the submesh 
    sub->useSharedVertices = true; 
    sub->indexData->indexBuffer = ibuf; 
    sub->indexData->indexCount = ibufCount; 
    sub->indexData->indexStart = 0; 
    /// Set bounding information (for culling) 
    msh->_setBounds(AxisAlignedBox(-100,-100,-100,100,100,100)); 
    msh->_setBoundingSphereRadius(Math::Sqrt(3*100*100)); 
    /// Notify -Mesh object that it has been loaded 
    msh->load(); 
    }

然后可以从mesh直接创建entity放在场景中

Entity*thisEntity = sceneManager->createEntity("cc", "ColourCube");

本文来自CSDN博客,转载请标明出处:http://blog.csdn.net/leonwei/archive/2010/08/10/5799770.aspx

原创粉丝点击