Table lookup example

This example shows how the task context buffer is used as a large lookup table to convert the 16 bit input data to 8 bit output data.

The lookup table has 65536 entries defined:

For all -32768 <= in <32768
          / 0,        in < -4096
Table(n)=|  in/32     -4096 <=in<4096
          \ 255,      in >= 4096

The following is the stripped down code list. The routines of less interest have been removed to allow you to focus on the key features. Because the task context buffer (the lookup table) is already initialized by the host code and the table is used as read-only data, you do not need the context setup and context merge functions on the accelerator side.

Data structures shared by the host and accelerator

The following code segment shows the data structures shared by both the host and the accelerators. The my_task_context_t data structure contains the lookup table. The my_wb_parms_t data structure represents the parameter and context data for each work block.

/* ---------------------------------------------- */
/* data structures shared by host and accelerator */
/* ---------------------------------------------- */
typedef struct _my_task_context_t
{
    alf_data_byte_t table[65536];
} my_task_context_t;

typedef struct _my_wb_parms_t
{
    alf_data_uint32_t num_data;  /* number of data in this WB */
} my_wb_parms_t;

Task descriptor setup

The following code segment shows how the task descriptor is set up for this application. The task context-related information marked in bold in the code.

alf_task_desc_create(alf_handle, 0, &task_desc_handle;);
  /* set up the task descriptor ... ... */
  /* the computing kernel name */
  alf_task_desc_set_int64(task_desc_handle, 
     ALF_TASK_DESC_ACCEL_KERNEL_REF_L, "comp_kernel");

  /* the task context buffer size */
  alf_task_desc_set_int32(task_desc_handle,
     ALF_TASK_DESC_TSK_CTX_SIZE, sizeof(my_task_context_t));

  /* the work block parm buffer size */
  alf_task_desc_set_int32(task_desc_handle,
      ALF_TASK_DESC_WB_PARM_CTX_BUF_SIZE, sizeof(my_wb_parms_t));

  /* the input buffer size */
  alf_task_desc_set_int32(task_desc_handle,
     ALF_TASK_DESC_WB_IN_BUF_SIZE,
     PART_SIZE*sizeof(alf_data_int16_t));

  /* the output buffer size */
  alf_task_desc_set_int32(task_desc_handle,
     ALF_TASK_DESC_WB_OUT_BUF_SIZE, 
     PART_SIZE*sizeof(alf_data_byte_t));

  /* the task context entry */
  alf_task_desc_ctx_entry_add(task_desc_handle, ALF_DATA_BYTE,
    sizeof(my_task_context_t)/sizeof(alf_data_byte_t));

Work block setup

The following code segment shows the code for work block creation.

  /* creating wb and adding param & io buffer */
  for (i = 0; i < NUM_DATA; i += PART_SIZE) 
  {
       alf_wb_create(task_handle, ALF_WB_SINGLE, 0, &wb_handle);
       alf_wb_dtl_begin(wb_handle, ALF_BUF_IN, 0); /* input */
       alf_wb_dtl_entry_add(wb_handle, pcm16_in+i, PART_SIZE, ALF_DATA_INT16);
       alf_wb_dtl_end(wb_handle);
       alf_wb_dtl_begin(wb_handle, ALF_BUF_OUT, 0); /* output */
       alf_wb_dtl_entry_add(wb_handle, pcm8_out+i,PART_SIZE, ALF_DATA_BYTE);
       alf_wb_dtl_end(wb_handle);
       wb_parm.num_data = PART_SIZE;
       alf_wb_parm_add(wb_handle, (void *)&wb_parm, /* wb parm */
       sizeof(wb_parm)/sizeof(unsigned int), ALF_DATA_INT32, 0);
       alf_wb_enqueue(wb_handle);
  }

Accelerator code

The following code is the accelerator side code. The section of the code that modifies the task context is marked in bold.

/* ---------------------------------------------- */
/* the accelerator side code                      */
/* ---------------------------------------------- */
/* the computation kernel function */
int comp_kernel(void *p_task_context, void *p_parm_ctx_buffer,
                    void *p_input_buffer, void *p_output_buffer,
                    void *p_inout_buffer, unsigned int current_count,
                    unsigned int total_count)
{
    my_task_context_t *p_ctx = (my_task_context_t *) p_task_context;
    my_wb_parms_t *p_parm = (my_wb_parms_t *) p_parm_ctx_buffer;

    alf_data_int16_t *in = (alf_data_int16_t *)p_input_buffer;
    alf_data_byte_t *out = (alf_data_byte_t *)p_output_buffer;
    unsigned int size = p_parm->num_data;
    unsigned int i;

    // it is just a simple table lookup
    for(i=0;i<size;i++) 
    {
       out[i] = p_ctx->table[(unsigned short)in[i]];
    } 
    return 0;
}