[FFmpeg-devel] [PATCH 10/21] avcodec/smacker: Use better nb_codes estimate when initializing VLC
Paul B Mahol
onemda at gmail.com
Sat Aug 1 16:58:44 EEST 2020
probably ok
On 8/1/20, Andreas Rheinhardt <andreas.rheinhardt at gmail.com> wrote:
> Using the real number of read codes allows to leave a loop in
> ff_init_vlc_sparse earlier; notice that all codes not explicitly
> set by reading data have been set to zero earlier (i.e. they are
> zero-length codes) and such codes are ignored by ff_init_vlc_sparse.
>
> This improves performance: When compiled with GCC 9, the time spent on
> one call to smka_decode_frame() for the sample from ticket #2425
> decreased from 2195367 decicycles to 2068389 decicycles. For Clang 9,
> it improved from 1602075 to 1534188 decicycles. These tests have been
> performed 20 times and each times the input file has been looped
> 32 times to get a sufficient number of frames.
>
> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt at gmail.com>
> ---
> libavcodec/smacker.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/libavcodec/smacker.c b/libavcodec/smacker.c
> index e6b163722a..27de30f0e1 100644
> --- a/libavcodec/smacker.c
> +++ b/libavcodec/smacker.c
> @@ -214,7 +214,7 @@ static int smacker_decode_header_tree(SmackVContext
> *smk, GetBitContext *gb, int
> goto error;
> skip_bits1(gb);
> if (h[i].current > 1) {
> - err = init_vlc(&vlc[i], SMKTREE_BITS, h[i].length,
> + err = init_vlc(&vlc[i], SMKTREE_BITS, h[i].current,
> INIT_VLC_DEFAULT_SIZES(h[i].lengths),
> INIT_VLC_DEFAULT_SIZES(h[i].bits),
> INIT_VLC_LE);
> @@ -657,7 +657,7 @@ static int smka_decode_frame(AVCodecContext *avctx, void
> *data,
> goto error;
> skip_bits1(&gb);
> if(h[i].current > 1) {
> - ret = init_vlc(&vlc[i], SMKTREE_BITS, h[i].length,
> + ret = init_vlc(&vlc[i], SMKTREE_BITS, h[i].current,
> h[i].lengths, sizeof(int), sizeof(int),
> h[i].bits, sizeof(uint32_t), sizeof(uint32_t),
> INIT_VLC_LE);
> if (ret < 0) {
> --
> 2.20.1
>
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request at ffmpeg.org with subject "unsubscribe".
More information about the ffmpeg-devel
mailing list