opengl:将数组类型加载到统一缓冲区对象

j0pj023g  于 2021-07-03  发布在  Java
关注(0)|答案(1)|浏览(399)

这是片段着色器中的统一块

layout (std140)   uniform Material
{
 float array[2];
};

这是统一块对象。我正在使用LWJGL2.9.3

private UBO(int programID,String blockName)
  {
   blockID=GL31.glGetUniformBlockIndex(programID,blockName);  //Get index of uniform block

   IntBuffer indices=BufferUtils.createIntBuffer(16);
   GL31.glGetActiveUniformBlock(programID,blockID,GL31.GL_UNIFORM_BLOCK_ACTIVE_UNIFORM_INDICES,indices); //query all the indices of every variable in the uniform block
   indices.flip().limit(GL31.glGetActiveUniformBlocki(programID,blockID,GL31.GL_UNIFORM_BLOCK_ACTIVE_UNIFORMS)); //the above method requires an int buffer of 16 even if i dont have 16 variables, but after retriving indices limit the buffer to read only actual number of variables
   while(indices.hasRemaining())
   {
    int variableIndex=indices.get();

    String name=GL31.glGetActiveUniformName(programID
                                           ,variableIndex
                                           ,GL31.glGetActiveUniformsi(programID,variableIndex,GL31.GL_UNIFORM_NAME_LENGTH)); //get the name of the variable

    int offset=GL31.glGetActiveUniformsi(programID,variableIndex,GL31.GL_UNIFORM_OFFSET); //query the offset[will use it in buffer sub data]

    offsets.put(name,offset);
   }

   GL15.glBindBuffer(GL31.GL_UNIFORM_BUFFER,bufferID=GL15.glGenBuffers());

   GL15.glBufferData(GL31.GL_UNIFORM_BUFFER
                    ,GL31.glGetActiveUniformBlocki(programID,blockID,GL31.GL_UNIFORM_BLOCK_DATA_SIZE)
                    ,GL15.GL_DYNAMIC_DRAW); //allocate buffer to the actual size of the uniform block

   GL15.glBindBuffer(GL31.GL_UNIFORM_BUFFER,0);
  }

  private void bufferData(String name,Buffer value)
  { 
   int offset=offsets.get(name); //update float or int arrays
   if(value instanceof FloatBuffer){GL15.glBufferSubData(GL31.GL_UNIFORM_BUFFER,offset,(FloatBuffer)value);}
   else{GL15.glBufferSubData(GL31.GL_UNIFORM_BUFFER,offset,(IntBuffer)value);}
  }

对于常规的非数组变量,它可以正常工作,但是对于数组,浮点值和每个数组类型值的加载都不正确。我已经阅读了指定数组中每个元素都有16[vec4 size]对齐方式的文档,但我不知道如何在代码中指定
要更新浮点数组,我只需使用

bufferData("array[0]",BufferUtils.createFloatBuffer(2).put(new float[]{10.5f,3.0f}).flip());
//the array[0] dosen't actually refer to only the first element of the array it is the name of the array returned by GL31.glGetActiveUniformName()

但同样的,浮子都装错了。

nr9pn0ug

nr9pn0ug1#

我已经阅读了指定数组中的每个元素都有16[vec4 size]对齐的文档,但是我不知道如何在我的代码中指定这一点[…]
你说得对,你必须为每个元素添加3个浮点数:

BufferUtils.createFloatBuffer(8).put(new float[]{10.5f,0.0f,0.0f,0.0f,3.0f,0.0f,0.0f,0.0f}).flip()

但是,我建议使用单个 vec2 而不是 float array[2] :

layout (std140)   uniform Material
{
    vec2 array;
};

注意,在glsl中,可以使用swizzling或index操作符(例如。 array[0] 或者 array[1] ).

相关问题