[FFmpeg-devel] [PATCH] configure: add host libm capability detection

[Ganesh Ajjanagadde]

On Thu, Nov 26, 2015 at 6:52 AM, Derek Buitenhuis
<derek.buitenhuis at gmail.com> wrote:
> On 11/26/2015 7:49 AM, Hendrik Leppkes wrote:
>> I'm really not a fan of this change. configure is slow and complex
>> enough as it is, and host tools don't need to be micro-optimized,
>> since they are never run by a user.
>
> I agree. It's opening an endless can complexity for essentially no gain.

How does it open an "endless can of complexity"? There are not many
host funcs needed. Gain of a couple of 0.1 ms startup across various
codecs is not minimal. In that case, why even bother keeping the
"hardcoded tables" option? The complexity of that and associated
ifdefry (note: it is being done at every tablegen file, not just
localized to a avutil/host_libm) is definitely at least as great as
that of doing this change.

>
> We went down this path 2+ years ago (in Libav), and in the end it was not done.
>
> - Derek
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel