[opus] [RFC PATCH v1] arm: kf_bfly4: Introduce ARM neon intrinsics

Viswanath Puttagunta viswanath.puttagunta at linaro.org
Fri Nov 21 15:43:20 PST 2014


On 14 November 2014 15:00, Timothy B. Terriberry <tterribe at xiph.org> wrote:
>
> Viswanath Puttagunta wrote:
>>
>> a. Will the approach of enabling this optimization using
>> --enable-armv7-neon-float or similar flag be acceptable to begin with?
>>
>> or
>>
>> b. Should the "AS_IF([test x"${enable_asm}" = x"yes"]".. section in
>> configure.ac be re-written?
>
>
> Automatic detection of compiler support in configure (in addition to run-time checks) would be preferable.
>
> We used a manual switches for, e.g., the SSE 4.2 intrinsics, just because the submitted auto-detection code was not very reliable, and no one has written anything better yet.

Please see my latest patch. It optimizes celt_pitch_xcorr for floating
point. As per your advise, I also added support for compile time and
run time detection.

You can abandon this patchset for now as I will work on integrating
NE10 FFT implementation later on..


More information about the opus mailing list