avcodec: split generic YV12/NV12 copy out of SSE2 functions
This is required by the next commit. This also avoids making two copies on non-x86 platforms, where USWC optimizations are irrelevant. (cherry picked from commit 07d8e576c7a0dcc7863d6853f620d5a99ebba55f) Conflicts: modules/codec/avcodec/copy.c
Showing
This diff is collapsed.
Please register or sign in to comment