mpegaudio: replace memcpy by SIMD code
authorChristophe GISQUET <christophe.gisquet@gmail.com>
Wed, 18 Jan 2012 21:34:29 +0000 (22:34 +0100)
committerRonald S. Bultje <rsbultje@gmail.com>
Thu, 16 Feb 2012 04:11:54 +0000 (20:11 -0800)
commitf3e084909bff422f0c853507a82f92ff2efc0d28
treef40d51982942666861616d941aebcf910f1b6341
parentae591aeea58d64399b8281be31dacec0de85ae04
mpegaudio: replace memcpy by SIMD code

By replacing memcpy with an unrolled loop using the alignment knowledge
it has, some speedup can be obtained.

Before (gcc 4.6.1): ~400 cycles
After: ~370 cycles

Overall, around 2% speed increase when decoding a 2400s mp3 to f32le.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
libavcodec/x86/mpegaudiodec_mmx.c