[FFmpeg-devel] [PATCH 2/2] libavutil: Make changes in softfloat needed for fixed point aac decoder.

[Nedeljko Babic]

>> 
>> As I said in the comment of this patch and before in comments to review,
>> I modified the code in the softfloat to be more usable in the implementation
>> of fixed point aac decoder.
>> Fixed point aac decoder was developed by using our float emulation and it was
>> more convenient to change ffmpeg softfloat than to make changes in aac since
>> softfloat is not used anywhere in ffmpeg currently.
>> 
>> It is clear to me that maybe for the number that is power of two it is more
>> convenient to use original way, but for other numbers it is basically the same.
>> 
>> On the other hand there is one subtraction less in the implementation of this
>> function if precision is used as argument.
>> 
>> And at the end, basically, I do not have a problem with changing this function
>> back to original if you want me to and comments above were just for the sake of
>> discussion :)
>
>you can add a 2nd function if you feel the subtraction matters
>but there should be one with simple and a human understandable
>interface

Ok. Than I'll leave the original function. 
I think it would bloat the code if I add basically the same function just because
of a subtraction.

Thanks,
Nedeljko