HEVC代码学习36：xTrMxN函数

之前在transformNxN的学习中提到，xT函数是前向变换的入口函数，实际上xT完成的是变换前的准备工作，真正的前向变换由xTrMxN函数实现。

先来看下xT函数。xT只是一个入口函数，首先定义了残差块block和系数块coeff，然后调用xTrMxN进行前向变换。

/** Wrapper function between HM interface and core NxN forward transform (2D)
 *  \param channelBitDepth bit depth of channel
 *  \param useDST
 *  \param piBlkResi input data (residual)
 *  \param uiStride stride of input residual data
 *  \param psCoeff output data (transform coefficients)
 *  \param iWidth transform width
 *  \param iHeight transform height
 *  \param maxLog2TrDynamicRange
 */
//前向变换的入口函数
Void TComTrQuant::xT( const Int channelBitDepth, Bool useDST, Pel* piBlkResi, UInt uiStride, TCoeff* psCoeff, Int iWidth, Int iHeight, const Int maxLog2TrDynamicRange )
{
#if MATRIX_MULT
  if( iWidth == iHeight)
  {
    xTr(channelBitDepth, piBlkResi, psCoeff, uiStride, (UInt)iWidth, useDST, maxLog2TrDynamicRange);
    return;
  }
#endif

  TCoeff block[ MAX_TU_SIZE * MAX_TU_SIZE ];        //同于存储残差块
  TCoeff coeff[ MAX_TU_SIZE * MAX_TU_SIZE ]