On Mon, Jan 21, 2013 at 06:57:43AM +0100, Luca Barbato wrote:
> From: "Ronald S. Bultje" <[email protected]>
> 
> Slate them to be removed on the next major bump.
> 
> They are functionally irrelevant, and most of them are unused,
> except the vp3 one, which is used wrongly in the bfin arch
> optimizations.

See the patch I just sent, which will make the VP3 one unused as well.
This patch ought to be rebased on top of that one.

> --- a/libavcodec/avcodec.h
> +++ b/libavcodec/avcodec.h
> @@ -2680,20 +2680,30 @@ typedef struct AVCodecContext {
>  #define FF_IDCT_ALTIVEC       8
>  #define FF_IDCT_SH4           9
>  #define FF_IDCT_SIMPLEARM     10
> +#if FF_API_IDCT
>  #define FF_IDCT_H264          11
>  #define FF_IDCT_VP3           12
> +#endif
>  #define FF_IDCT_IPP           13
>  #define FF_IDCT_XVIDMMX       14
> +#if FF_API_IDCT
>  #define FF_IDCT_CAVS          15
> +#endif
>  #define FF_IDCT_SIMPLEARMV5TE 16
>  #define FF_IDCT_SIMPLEARMV6   17
>  #define FF_IDCT_SIMPLEVIS     18
> +#if FF_API_IDCT
>  #define FF_IDCT_WMV2          19
> +#endif
>  #define FF_IDCT_FAAN          20
> +#if FF_API_IDCT
>  #define FF_IDCT_EA            21
> +#endif
>  #define FF_IDCT_SIMPLENEON    22
>  #define FF_IDCT_SIMPLEALPHA   23
> +#if FF_API_IDCT
>  #define FF_IDCT_BINK          24
> +#endif

It would be nice if these could be grouped together under a single
ifdef instead.

> --- a/libavcodec/bfin/dsputil_bfin.c
> +++ b/libavcodec/bfin/dsputil_bfin.c
> @@ -257,7 +257,8 @@ void ff_dsputil_init_bfin( DSPContext* c, AVCodecContext 
> *avctx )
>          if (avctx->dct_algo == FF_DCT_AUTO)
>              c->fdct                  = ff_bfin_fdct;
>  
> -        if (avctx->idct_algo == FF_IDCT_VP3) {
> +        // FIXME convert to VP3DSPContext
> +        if (0) { // avctx->idct_algo == FF_IDCT_VP3) {
>              c->idct_permutation_type = FF_NO_IDCT_PERM;
>              c->idct                  = ff_bfin_vp3_idct;
>              c->idct_add              = ff_bfin_vp3_idct_add;

see above

Diego
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to