LGTM. Diego Elio Pettenò — Flameeyes [email protected] — http://blog.flameeyes.eu/
On Mon, Sep 2, 2013 at 5:08 PM, Vittorio Giovara <[email protected] > wrote: > --- > I did a simple performance test for this, as suggested on IRC and got > > WITHOUT > real 2m57.078s > user 4m16.308s > sys 0m31.410s > > WITH > real 2m56.643s > user 4m16.052s > sys 0m31.714s > > Plus I think code is more readable. I think there might be quite a few > places that could receive the same treatment. > > Cheers, > Vittorio > > libavcodec/mpeg12enc.c | 14 ++------------ > 1 file changed, 2 insertions(+), 12 deletions(-) > > diff --git a/libavcodec/mpeg12enc.c b/libavcodec/mpeg12enc.c > index 0ede7c4..a69d9b6 100644 > --- a/libavcodec/mpeg12enc.c > +++ b/libavcodec/mpeg12enc.c > @@ -439,15 +439,13 @@ static inline void put_mb_modes(MpegEncContext *s, > int n, int bits, int has_mv, > } > } > > -static av_always_inline void mpeg1_encode_mb_internal(MpegEncContext *s, > - int16_t > block[6][64], > - int motion_x, int > motion_y, > - int mb_block_count) > +void ff_mpeg1_encode_mb(MpegEncContext *s, int16_t block[6][64], int > motion_x, int motion_y) > { > int i, cbp; > const int mb_x = s->mb_x; > const int mb_y = s->mb_y; > const int first_mb = mb_x == s->resync_mb_x && mb_y == s->resync_mb_y; > + const int mb_block_count = s->chroma_format == CHROMA_420 ? 6 : 8; > > /* compute cbp */ > cbp = 0; > @@ -671,14 +669,6 @@ static av_always_inline void > mpeg1_encode_mb_internal(MpegEncContext *s, > } > } > > -void ff_mpeg1_encode_mb(MpegEncContext *s, int16_t block[6][64], int > motion_x, int motion_y) > -{ > - if (s->chroma_format == CHROMA_420) > - mpeg1_encode_mb_internal(s, block, motion_x, motion_y, 6); > - else > - mpeg1_encode_mb_internal(s, block, motion_x, motion_y, 8); > -} > - > // RAL: Parameter added: f_or_b_code > static void mpeg1_encode_motion(MpegEncContext *s, int val, int > f_or_b_code) > { > -- > 1.7.9.5 > > _______________________________________________ > libav-devel mailing list > [email protected] > https://lists.libav.org/mailman/listinfo/libav-devel > _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
