On 07/19/2012 03:41 AM, Luca Barbato wrote:
> From: Michael Bradshaw <[email protected]>

>  static av_cold int libopenjpeg_decode_init(AVCodecContext *avctx)
> @@ -54,6 +225,7 @@ static av_cold int libopenjpeg_decode_init(AVCodecContext 
> *avctx)
>      LibOpenJPEGContext *ctx = avctx->priv_data;
>  
>      opj_set_default_decoder_parameters(&ctx->dec_params);
> +    avcodec_get_frame_defaults(&ctx->image);
>      avctx->coded_frame = &ctx->image;
>      return 0;
>  }
> @@ -77,10 +249,10 @@ static int libopenjpeg_decode_frame(AVCodecContext 
> *avctx,
>  

This function is huge and is pretty much resetting the decoder all the
time, I wonder if we couldn't keep some of it out the way or reset it
only when dimensions or pixel format change.

I might try later. dropping the overinlining might be good though.

lu

-- 

Luca Barbato
Gentoo/linux
http://dev.gentoo.org/~lu_zero

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

Reply via email to