AnsweredAssumed Answered

TCB block of fir accelerator from example in BSP of ADSP-21478

Question asked by jsksra1 on Nov 30, 2015
Latest reply on Dec 9, 2015 by Mitesh



In the Multi_Iteration_Mode example

TCB block for the fir accelerator is mentioned in one order and the TCB block defined in its reverse order.


And it will work only if we follow the code similarly, why we have to follow in this order??

And do we have to follow the reverse of TCB block mentioned for any TCBs like(FFT,DMA,SPORT etc )



Please find the below code snippet


/* TCB Structure

  CP[18:0]          ---------------      FIRCTL2

  CP[18:0] -0x1     ---------------      II ---- should point to x(n-N-1) ---- x(0)

  CP[18:0] -0x2     ----------------     IM

  CP[18:0] -0x3     -----------------    IL/IC

  CP[18:0] -0x4     -----------------    IB  ---- Base address for input circular buffer

  CP[18:0] -0x5     -----------------    OI

  CP[18:0] -0x6     -----------------    OM

  CP[18:0] -0x7     -----------------    OL/OC

  CP[18:0] -0x8     -----------------    OB  ---- Base address for output circular buffer

  CP[18:0] -0x9     -----------------    CI  --- Coefficient Index should point to c(N-1)

  CP[18:0] -0xA     -----------------    CM

  CP[18:0] -0xB     -----------------    CBL ---- Coefficient buffer Length

  CP[18:0] -0xC     -----------------    CP */



int TCB_Buf[13] = {0,TAPSIZE,-1,(int)Coeff_Buf+TAPS,(int)Out_Buf,WINDOWSIZE,1,(int)Out_Buf,(int)In_Buf,TAPSIZE+WINDOWSIZE-1,1,(int)In_Buf,0};