[FFmpeg-devel] [RFC/PATCH] bitpacked_dec: Optimization for bitpacked_dec decoder performance
Devin Heitmueller
devin.heitmueller at ltnglobal.com
Sat May 6 15:13:44 EEST 2023
I added some instrumentation via the attached patch. You can see the
benefits here:
Before=1683378057.243350 After 1683378057.264239
Before=1683378083.335424 After 1683378083.356440
Before=1683378089.675400 After 1683378089.696512
Before=1683378151.792324 After 1683378151.813579
21 ms per run
After patch:
Before=1683378222.167796 After 1683378222.175760
Before=1683378233.131416 After 1683378233.139326
Before=1683378243.591895 After 1683378243.599840
8 ms per run
Note: this is a different platform than I did the original development
on, and apparently the improvement on this particular box is only 2.5x
rather than 4x.
Devin
On Sat, May 6, 2023 at 7:53 AM Paul B Mahol <onemda at gmail.com> wrote:
>
> On Sat, May 6, 2023 at 1:32 PM Lance Wang <lance.lmwang at gmail.com> wrote:
>
> > On Sat, May 6, 2023 at 4:58 AM Devin Heitmueller <
> > devin.heitmueller at ltnglobal.com> wrote:
> >
> > > Rework the code a bit to speed up the 10-bit bitpacked decoding
> > > routine. This is probably about as fast as I can get it without
> > > switching to assembly language.
> > >
> > > Demonstratable with:
> > >
> > > ./ffmpeg -f lavfi -i "smptehdbars=size=3840x2160" -c bitpacked -f image2
> > > -frames:v 1 source.yuv
> > > ./ffmpeg -f bitpacked -pix_fmt yuv422p10le -s 3840x2160 -c:v bitpacked -i
> > > source.yuv -pix_fmt yuv422p10le out.yuv
> > >
> > > On my development system, it went from 80ms for a 2160p frame
> > > down to 20ms (i.e. a 4X speedup). Good enough for now, I hope...
> > >
> > >
> > FYI, on my development system, I run two time for the original and modified
> > version and no obvious difference:
> > ./ffmpeg -f lavfi -i "smptehdbars=size=3840x2160" -c bitpacked -frames:v 25
> > source.yuv
> > time ./ffmpeg -f bitpacked -pix_fmt yuv422p10le -s 3840x2160 -c:v bitpacked
> > -i source.yuv -pix_fmt yuv422p10le out.yuv
> > frame= 25 fps=0.0 q=-0.0 Lsize= 810000kB time=00:00:00.96
> > bitrate=6912000.0kbits/s speed=1.13x
> >
> > real 0m0.961s
> > user 0m1.086s
> > sys 0m1.360s
> >
> > frame= 25 fps=0.0 q=-0.0 Lsize= 810000kB time=00:00:00.96
> > bitrate=6912000.0kbits/s speed=1.16x
> >
> > real 0m0.936s
> > user 0m1.358s
> > sys 0m1.350s
> >
> > after apply the patch:
> > frame= 25 fps=0.0 q=-0.0 Lsize= 810000kB time=00:00:00.96
> > bitrate=6912000.0kbits/s speed=1.14x
> >
> > real 0m0.953s
> > user 0m0.906s
> > sys 0m1.438s
> >
> > frame= 25 fps=0.0 q=-0.0 Lsize= 810000kB time=00:00:00.96
> > bitrate=6912000.0kbits/s speed=1.17x
> >
> > real 0m0.922s
> > user 0m0.926s
> > sys 0m1.066s
> >
>
> Only 25 frames?
> This is flawed.
>
>
> >
> >
> >
> > > Signed-off-by: Devin Heitmueller <dheitmueller at ltnglobal.com>
> > > ---
> > > libavcodec/bitpacked_dec.c | 17 +++++++----------
> > > 1 file changed, 7 insertions(+), 10 deletions(-)
> > >
> > > diff --git a/libavcodec/bitpacked_dec.c b/libavcodec/bitpacked_dec.c
> > > index a1ffef1..96aba27 100644
> > > --- a/libavcodec/bitpacked_dec.c
> > > +++ b/libavcodec/bitpacked_dec.c
> > > @@ -28,7 +28,6 @@
> > >
> > > #include "avcodec.h"
> > > #include "codec_internal.h"
> > > -#include "get_bits.h"
> > > #include "libavutil/imgutils.h"
> > > #include "thread.h"
> > >
> > > @@ -65,7 +64,7 @@ static int bitpacked_decode_yuv422p10(AVCodecContext
> > > *avctx, AVFrame *frame,
> > > {
> > > uint64_t frame_size = (uint64_t)avctx->width *
> > > (uint64_t)avctx->height * 20;
> > > uint64_t packet_size = (uint64_t)avpkt->size * 8;
> > > - GetBitContext bc;
> > > + uint8_t *src;
> > > uint16_t *y, *u, *v;
> > > int ret, i, j;
> > >
> > > @@ -79,20 +78,18 @@ static int bitpacked_decode_yuv422p10(AVCodecContext
> > > *avctx, AVFrame *frame,
> > > if (avctx->width % 2)
> > > return AVERROR_PATCHWELCOME;
> > >
> > > - ret = init_get_bits(&bc, avpkt->data, avctx->width * avctx->height *
> > > 20);
> > > - if (ret)
> > > - return ret;
> > > -
> > > + src = avpkt->data;
> > > for (i = 0; i < avctx->height; i++) {
> > > y = (uint16_t*)(frame->data[0] + i * frame->linesize[0]);
> > > u = (uint16_t*)(frame->data[1] + i * frame->linesize[1]);
> > > v = (uint16_t*)(frame->data[2] + i * frame->linesize[2]);
> > >
> > > for (j = 0; j < avctx->width; j += 2) {
> > > - *u++ = get_bits(&bc, 10);
> > > - *y++ = get_bits(&bc, 10);
> > > - *v++ = get_bits(&bc, 10);
> > > - *y++ = get_bits(&bc, 10);
> > > + *u++ = (src[0] << 2) | (src[1] >> 6);
> > > + *y++ = ((src[1] << 4) | (src[2] >> 4)) & 0x3ff;
> > > + *v++ = ((src[2] << 6) | (src[3] >> 2)) & 0x3ff;
> > > + *y++ = ((src[3] << 8) | (src[4])) & 0x3ff;
> > > + src += 5;
> > > }
> > > }
> > >
> > > --
> > > 1.8.3.1
> > >
> > > _______________________________________________
> > > 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".
> > >
> > _______________________________________________
> > 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".
> >
> _______________________________________________
> 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".
--
Devin Heitmueller, Senior Software Engineer
LTN Global Communications
o: +1 (301) 363-1001
w: https://ltnglobal.com e: devin.heitmueller at ltnglobal.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: timing.patch
Type: application/octet-stream
Size: 1066 bytes
Desc: not available
URL: <https://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20230506/e085e70f/attachment.obj>
More information about the ffmpeg-devel
mailing list