[FFmpeg-devel] [PATCH][RFC] swresample/dither: use integer arithmetic

Ganesh Ajjanagadde gajjanag at mit.edu
Sun Aug 23 22:26:45 CEST 2015


On Sun, Aug 23, 2015 at 4:17 PM, James Almer <jamrial at gmail.com> wrote:
> On 23/08/15 5:09 PM, Ganesh Ajjanagadde wrote:
>> On Sun, Aug 23, 2015 at 3:42 PM, James Almer <jamrial at gmail.com> wrote:
>>> On 23/08/15 4:10 PM, Ganesh Ajjanagadde wrote:
>>>> I notice one use in entire codebase (line 515 of avformat/dump.c), so
>>>> it should be ok in my view.
>>>
>>> Nevermind, missed this.
>>>
>>> Not sure how it works with msvc 2012 though, assuming the msdn link
>>> documentation is right.
>>
>> You might be onto something, note that this was introduce in June 2015,
>> commit ID: df22e30172b09cda4d6f7d4f43508284be65848a.
>>
>> Maybe build on such MSVC broke since then?
>> That would perhaps explain why it was not caught yet,
>> assuming we don't have many MSVC 2012 testers.
>
> We have a msvc 2012 client and it compiles that file just fine without any
> related warning.
>
> http://fate.ffmpeg.org/log.cgi?time=20150822170001&log=compile&slot=x86_32-msvc11-windows-native

Their documentation gives:
https://msdn.microsoft.com/en-us/library/a206stx2%28v=vs.110%29.aspx
-> 2012 has only labs()
https://msdn.microsoft.com/en-us/library/a206stx2%28v=vs.120%29.aspx
-> 2013 onwards has both.

wiki says:
FFmpeg can be built with MSVC 2012 or earlier using a C99-to-C89
conversion utility and wrapper, or with MSVC 2013 natively.

Maybe conversion utility was being run for 2012 client, allowing safe
use of llabs.
Assuming I am right here, I can use llabs safely without any issues.

Thanks for checking the compat stuff.

> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel


More information about the ffmpeg-devel mailing list