On 12/07/2012 07:20 AM, Måns Rullgård wrote: > Christophe Gisquet <christophe.gisq...@gmail.com> writes: > >> 2012/11/27 Justin Ruggles <justin.rugg...@gmail.com>: >>> The patch itself looks good. Only the neon version would need to be >>> modified. The altivec version already handles multiples of 8. >>> >>> The (yet to be ported) tak decoder could also use this. >> >> This tak decoder seems to require down to multiples of 4, not sure if >> that's worth it. > > Can the buffers be padded to allow some overread?
Yes that would probably work. I'll give it a try. FWIW, the only 2 sizes that require a multiple of 4 are 4 and 12. The only size that requires a multiple of 8 is 24. The rest are multiples of 16. -Justin _______________________________________________ libav-devel mailing list libav-devel@libav.org https://lists.libav.org/mailman/listinfo/libav-devel
