[x265] [PATCH] removed loop, using part variable to test all partition for chroma_copy_pp

[Steve Borho]

On Tue, Nov 5, 2013 at 3:24 AM, Deepthi Nandakumar <
deepthi at multicorewareinc.com> wrote:

>
>
>
> On Mon, Nov 4, 2013 at 7:36 PM, <praveen at multicorewareinc.com> wrote:
>
>> # HG changeset patch
>> # User Praveen Tiwari
>> # Date 1383573863 -19800
>> # Node ID 20b712dce6c237f64ea3ca2a06fc4cb9370ad91f
>> # Parent  690639d8d50283feeb6d208ba70d350aa2b7b196
>> removed loop, using part variable to test all partition for chroma_copy_pp
>>
>> diff -r 690639d8d502 -r 20b712dce6c2 source/test/pixelharness.cpp
>> --- a/source/test/pixelharness.cpp      Mon Nov 04 18:57:06 2013 +0530
>> +++ b/source/test/pixelharness.cpp      Mon Nov 04 19:34:23 2013 +0530
>> @@ -645,6 +645,14 @@
>>          }
>>      }
>>
>> What does this mean? You are calling check_block_copy_pp recursively??
>


not sure what you're asking, but "if (opt.chroma_copy_pp[part])" checks
that an optimized function for this partition exists.


>
>> +    if (opt.chroma_copy_pp[part])
>> +    {
>> +        if (!check_block_copy_pp(ref.chroma_copy_pp[part],
>> opt.chroma_copy_pp[part]))
>> +        {
>> +            printf("chroma_copy_pp[%s] failed\n", lumaPartStr[part]);
>> +            return false;
>> +        }
>> +    }
>>      return true;
>>  }
>>
>> @@ -803,18 +811,6 @@
>>              return false;
>>          }
>>      }
>> -
>>
> Why should check chroma be a part of check luma?
>

the chroma and luma partitions are deliberately aligned, so an iteration
over one can be used as an iteration over both;  you just have to be
careful to index into the right string table in the printf.


>
>
>> -    for (int i = 0; i < NUM_CHROMA_PARTITIONS; i++)
>> -    {
>> -      if (opt.chroma_copy_pp[i])
>> -      {
>> -          if (!check_block_copy_pp(ref.chroma_copy_pp[i],
>> opt.chroma_copy_pp[i]))
>> -          {
>> -              printf("chroma_copy_pp[%s] failed\n", chromaPartStr[i]);
>> -              return false;
>> -          }
>> -      }
>> -    }
>>      return true;
>>  }
>>
>> _______________________________________________
>> x265-devel mailing list
>> x265-devel at videolan.org
>> https://mailman.videolan.org/listinfo/x265-devel
>>
>
>
> _______________________________________________
> x265-devel mailing list
> x265-devel at videolan.org
> https://mailman.videolan.org/listinfo/x265-devel
>
>


-- 
Steve Borho
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20131105/09049734/attachment.html>