2012/2/22 Kostya Shishkov <[email protected]>:
>> @@ -1022,25 +1022,10 @@ static void rv34_output_i16x16(RV34DecContext *r, 
>> int8_t *intra_types, int cbp)
>>                      q_ac = rv34_qscale_tab[s->qscale];
>>      uint8_t        *dst  = s->dest[0];
>>      DCTELEM        *ptr  = s->block[0];
>> -    int       avail[6*8] = {0};
>>      int i, j, itype, has_ac;
>>
>>      memset(block16, 0, 16 * sizeof(*block16));
>>
>> -    // Set neighbour information.
>> -    if(r->avail_cache[1])
>> -        avail[0] = 1;
>> -    if(r->avail_cache[2])
>> -        avail[1] = avail[2] = 1;
>> -    if(r->avail_cache[3])
>> -        avail[3] = avail[4] = 1;
>> -    if(r->avail_cache[4])
>> -        avail[5] = 1;
>> -    if(r->avail_cache[5])
>> -        avail[8] = avail[16] = 1;
>> -    if(r->avail_cache[9])
>> -        avail[24] = avail[32] = 1;
>> -
>>      has_ac = rv34_decode_block(block16, gb, r->cur_vlcs, 3, 0, q_dc, q_dc, 
>> q_ac);
>>      if(has_ac)
>>          r->rdsp.rv34_inv_transform(block16);
>> --

Yes the review at the time was really only related to this hunk.
Though I thought the other hunks might as well be included under that
'availability check' umbrella. Your call.

Christophe
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to