[x264-devel] x86: Use SSE instead of SSE2 for copying data

[Henrik Gramner]

x264 | branch: master | Henrik Gramner <henrik at gramner.com> | Sun Jan 27 23:01:59 2013 +0100| [c63a518d43bb3822342513eb4af109551e86fbd2] | committer: Jason Garrett-Glaser

x86: Use SSE instead of SSE2 for copying data

Reduces code size because movaps/movups is one byte shorter than movdqa/movdqu.
Also merge MMX and SSE versions of memcpy_aligned into a single macro.

> http://git.videolan.org/gitweb.cgi/x264.git/?a=commit;h=c63a518d43bb3822342513eb4af109551e86fbd2
---

 common/cpu.c             |    2 +-
 common/pixel.c           |    4 ++
 common/x86/mc-a.asm      |    4 +-
 common/x86/mc-a2.asm     |   92 +++++++++++++++++++---------------------------
 common/x86/mc-c.c        |   44 +++++++++++-----------
 common/x86/predict-a.asm |   18 ++++-----
 common/x86/predict-c.c   |   20 +++++++---
 common/x86/predict.h     |    8 ++--
 common/x86/x86inc.asm    |    4 +-
 tools/checkasm.c         |    5 ++-
 10 files changed, 102 insertions(+), 99 deletions(-)

Diff:   http://git.videolan.org/gitweb.cgi/x264.git/?a=commitdiff;h=c63a518d43bb3822342513eb4af109551e86fbd2