fbarchard@google.com 78a3a6b345 Change Any functions that convert 1 to 1 formats, memcpy style, so use C for remainder to allow a minimum width of 1. This has some advantages - allows function to be used even with SIMD that only allows aligned memory. Fewer macros, used by more functions. SIMD is not used unaligned avoiding page/cache split. No overlap so it can be used in place. Disadvantage is it will be slower if close to the maximum number of non-SIMD pixels. ... BUG=373 TESTED=libyuv_unittest still passes R=brucedawson@google.com, tpsiaki@google.com Review URL: https://webrtc-codereview.appspot.com/23209004 git-svn-id: http://libyuv.googlecode.com/svn/trunk@1141 16f28f9a-4ce2-e073-06de-1de4eb20be90		2014-10-24 22:17:59 +00:00
..
libyuv	Change Any functions that convert 1 to 1 formats, memcpy style, so use C for remainder to allow a minimum width of 1. This has some advantages - allows function to be used even with SIMD that only allows aligned memory. Fewer macros, used by more functions. SIMD is not used unaligned avoiding page/cache split. No overlap so it can be used in place. Disadvantage is it will be slower if close to the maximum number of non-SIMD pixels.	2014-10-24 22:17:59 +00:00
libyuv.h	FixedDiv1 using a single 64/32 divide. Removes size restriction from slope.	2014-01-02 22:32:09 +00:00