On Fri, Jul 08, 2011 at 11:51:47AM -0700, Ronald S. Bultje wrote: > From: "Ronald S. Bultje" <[email protected]> > > Fix an overflow in 16-bit vertical scaling (e.g. using -flags lanczos), > fix a crash in bilinear vertical scaling when using C code if output > format was 8-bpc RGB but without alpha (e.g. RGB24), fix an underflow > in 16-bit horizontal scaling when using e.g. lanczos scaling, and fix > a bug when converting from 16-bit to 8-bit and using MMX optimizations, > where we would wrongly use the second (unused) hald of the U plane, > rather than the V plane, to write out the V pixels.
Sounds like multiple commits squashed into one. s/hald/half/ Diego _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
