On 07/07/14 17:07, Diego Biurrun wrote: > On Tue, Jun 03, 2014 at 08:07:30PM +0200, Diego Biurrun wrote: >> On Wed, Mar 26, 2014 at 06:28:10PM +0100, Luca Barbato wrote: >>> On 26/03/14 12:41, Diego Biurrun wrote: >>>> --- >>>> libavcodec/dsputil.c | 10 ++++++++-- >>>> libavcodec/dsputil.h | 1 - >>>> libavcodec/utils.c | 13 ------------- >>>> 3 files changed, 8 insertions(+), 16 deletions(-) >>> >>> The code seems more wrong than it was before. >>> >>> Would be better to keep a function and call it from the init_static_data >>> callback of the codecs that use it. >> >> It's used by quite a few codecs and subsystems and also by codecs that >> already declare such a callback, so I'm not sure how feasible this would >> be. Besides, we have many other codecs and subsystems that function in >> this way, so would this patch really make things considerably worse? > > ping > > I currently lack alternative suggestions and this patch is starting to > hold up the rest of the dsputil work.
Call the current ff_ function from the init_static_data callbacks when needed. lu _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
