On Mon, May 13, 2024 at 6:49 PM James Almer <jamr...@gmail.com> wrote: > On 5/6/2024 2:49 PM, Rémi Denis-Courmont wrote: > > Le maanantaina 6. toukokuuta 2024, 20.18.11 EEST Ramiro Polla a écrit : > >> I'll send a similar patch to fix checkasm/idctdsp after this is merged. > > > > The idctdsp test does not actually test the iDCT, but only the trivial-ish > > add/put helpers, so it does not care about the context. You're welcome to > > fix > > it anyway of course. > > I personally find it ugly how we're storing a whole AVCodecContext on > stack in these tests just to pass two ints to an init function. > Maybe we can make said values be input parameters for these instead of a > pointer to avctx.
It could make sense for fdct, but for idct we need a few more parameters (bits_per_raw_sample, codec_id, flags, idct_algo, lowres). That would make the function calls much longer, and in that case I'd prefer just keeping AVCodecContext. Or having an input parameter structure for each *dsp context, but that seems a bit overkill. _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
