On 8/11/20, [email protected] <[email protected]> wrote: > On Tue, Aug 11, 2020 at 05:24:06PM +0200, Paul B Mahol wrote: >> NAK >> >> On 8/11/20, [email protected] <[email protected]> wrote: >> > From: Limin Wang <[email protected]> >> > >> > max is initialized to ((1 << depth) - 1) * 256 * 128 before. >> > >> > Signed-off-by: Limin Wang <[email protected]> >> > --- >> > libavfilter/vf_w3fdif.c | 2 +- >> > 1 file changed, 1 insertion(+), 1 deletion(-) >> > >> > diff --git a/libavfilter/vf_w3fdif.c b/libavfilter/vf_w3fdif.c >> > index 5d64dbd..0f9efbf 100644 >> > --- a/libavfilter/vf_w3fdif.c >> > +++ b/libavfilter/vf_w3fdif.c >> > @@ -163,7 +163,7 @@ static void filter_scale(uint8_t *out_pixel, const >> > int32_t *work_pixel, int line >> > int j; >> > >> > for (j = 0; j < linesize; j++, out_pixel++, work_pixel++) >> > - *out_pixel = av_clip(*work_pixel, 0, 255 * 256 * 128) >> 15; >> > + *out_pixel = av_clip(*work_pixel, 0, max) >> 15; >> > } >> > >> > static void filter16_simple_low(int32_t *work_line, >> > -- >> > 1.8.3.1 >> > >> > _______________________________________________ >> > ffmpeg-devel mailing list >> > [email protected] >> > https://ffmpeg.org/mailman/listinfo/ffmpeg-devel >> > >> > To unsubscribe, visit link above, or email >> > [email protected] with subject "unsubscribe". >> >> >> Change is pointless and give no real benefits. > > 1. max is input as parameter and calculated already, why use it is > pointless? > 2. keep 8 bit and 16bit with same code so that we can remove the duplicate > code > later.
Code is not duplicated at all. Different types are used. You can off course use macros if you are really bored but regressions in performance are unacceptable. Your patch also ruin performance. > > -- > Thanks, > Limin Wang > _______________________________________________ > ffmpeg-devel mailing list > [email protected] > https://ffmpeg.org/mailman/listinfo/ffmpeg-devel > > To unsubscribe, visit link above, or email > [email protected] with subject "unsubscribe". _______________________________________________ ffmpeg-devel mailing list [email protected] https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email [email protected] with subject "unsubscribe".
