Add NEON versions of the SATD functions.
This gives CIF encoding a 37.5% speed up on a Cortex A8. svn path=/trunk/theora/; revision=17874
Showing
- lib/Makefile.am 4 additions, 0 deletionslib/Makefile.am
- lib/arm/armenc.c 3 additions, 0 deletionslib/arm/armenc.c
- lib/arm/armenc.h 7 additions, 0 deletionslib/arm/armenc.h
- lib/arm/armencfrag.s 220 additions, 0 deletionslib/arm/armencfrag.s
- lib/arm/armfrag.s 1 addition, 2 deletionslib/arm/armfrag.s
Loading
Please register or sign in to comment