Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
 help / color / mirror / Atom feed
From: Arnie Chang <arnie.chang@sifive.com>
To: ffmpeg-devel@ffmpeg.org
Cc: Arnie Chang <arnie.chang@sifive.com>
Subject: [FFmpeg-devel] [PATCH 0/5] RISC-V: Improve H264 decoding performance using RVV intrinsic
Date: Tue,  9 May 2023 17:50:25 +0800
Message-ID: <20230509095030.25506-1-arnie.chang@sifive.com> (raw)

We are submitting a set of patches that significantly improve H.264 decoding performance
by utilizing RVV intrinsic code. The average speedup(FPS) achieved by these patches is more than 2x,
as experimented on 720P videos running on an internal FPGA board.

Patch1: add support for RVV intrinsic code in the configure file
Patch2: optimize chroma motion compensation
Patch3: optimize luma motion compensation
Patch4: optimize dsp functions, such as IDCT, in-loop filtering, and weighed filtering
Patch5: optimize intra prediction

Arnie Chang (5):
  configure: Add detection of RISC-V vector intrinsic support
  lavc/h264chroma: Add vectorized implementation of chroma MC for RISC-V
  lavc/h264qpel: Add vectorized implementation of luma MC for RISC-V
  lavc/h264dsp: Add vectorized implementation of DSP functions for
    RISC-V
  lavc/h264pred: Add vectorized implementation of intra prediction for
    RISC-V

 configure                                 |    2 +
 libavcodec/h264chroma.c                   |    2 +
 libavcodec/h264chroma.h                   |    1 +
 libavcodec/h264dsp.c                      |    2 +
 libavcodec/h264dsp.h                      |    3 +-
 libavcodec/h264pred.c                     |    2 +
 libavcodec/h264pred.h                     |    3 +-
 libavcodec/h264qpel.c                     |    2 +
 libavcodec/h264qpel.h                     |    1 +
 libavcodec/riscv/Makefile                 |   11 +
 libavcodec/riscv/h264_chroma_init_riscv.c |   45 +
 libavcodec/riscv/h264_dsp_init_riscv.c    |   68 ++
 libavcodec/riscv/h264_idct.c              |  482 +++++++++
 libavcodec/riscv/h264_idct.h              |   46 +
 libavcodec/riscv/h264_inloop.c            |  669 ++++++++++++
 libavcodec/riscv/h264_inloop.h            |   47 +
 libavcodec/riscv/h264_lowpass.h           |  249 +++++
 libavcodec/riscv/h264_mc_chroma.c         |  821 ++++++++++++++
 libavcodec/riscv/h264_mc_chroma.h         |   40 +
 libavcodec/riscv/h264_mc_luma.c           |  412 +++++++
 libavcodec/riscv/h264_mc_luma.h           |  101 ++
 libavcodec/riscv/h264_mc_luma_avg16.h     | 1183 +++++++++++++++++++++
 libavcodec/riscv/h264_mc_luma_avg8.h      |  773 ++++++++++++++
 libavcodec/riscv/h264_mc_luma_put16.h     |  963 +++++++++++++++++
 libavcodec/riscv/h264_mc_luma_put8.h      |  648 +++++++++++
 libavcodec/riscv/h264_pred.c              |  884 +++++++++++++++
 libavcodec/riscv/h264_pred.h              |   53 +
 libavcodec/riscv/h264_pred_init_riscv.c   |   67 ++
 libavcodec/riscv/h264_qpel_init_riscv.c   |  107 ++
 libavcodec/riscv/h264_utility.h           |   75 ++
 libavcodec/riscv/h264_weighted_sum.c      |  273 +++++
 libavcodec/riscv/h264_weighted_sum.h      |   47 +
 32 files changed, 8080 insertions(+), 2 deletions(-)
 create mode 100644 libavcodec/riscv/h264_chroma_init_riscv.c
 create mode 100644 libavcodec/riscv/h264_dsp_init_riscv.c
 create mode 100644 libavcodec/riscv/h264_idct.c
 create mode 100644 libavcodec/riscv/h264_idct.h
 create mode 100644 libavcodec/riscv/h264_inloop.c
 create mode 100644 libavcodec/riscv/h264_inloop.h
 create mode 100644 libavcodec/riscv/h264_lowpass.h
 create mode 100644 libavcodec/riscv/h264_mc_chroma.c
 create mode 100644 libavcodec/riscv/h264_mc_chroma.h
 create mode 100644 libavcodec/riscv/h264_mc_luma.c
 create mode 100644 libavcodec/riscv/h264_mc_luma.h
 create mode 100644 libavcodec/riscv/h264_mc_luma_avg16.h
 create mode 100644 libavcodec/riscv/h264_mc_luma_avg8.h
 create mode 100644 libavcodec/riscv/h264_mc_luma_put16.h
 create mode 100644 libavcodec/riscv/h264_mc_luma_put8.h
 create mode 100644 libavcodec/riscv/h264_pred.c
 create mode 100644 libavcodec/riscv/h264_pred.h
 create mode 100644 libavcodec/riscv/h264_pred_init_riscv.c
 create mode 100644 libavcodec/riscv/h264_qpel_init_riscv.c
 create mode 100644 libavcodec/riscv/h264_utility.h
 create mode 100644 libavcodec/riscv/h264_weighted_sum.c
 create mode 100644 libavcodec/riscv/h264_weighted_sum.h

-- 
2.17.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

             reply	other threads:[~2023-05-09  9:50 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-05-09  9:50 Arnie Chang [this message]
2023-05-09  9:50 ` [FFmpeg-devel] [PATCH 1/5] configure: Add detection of RISC-V vector intrinsic support Arnie Chang
2023-05-09  9:50 ` [FFmpeg-devel] [PATCH 2/5] lavc/h264chroma: Add vectorized implementation of chroma MC for RISC-V Arnie Chang
2023-05-09  9:50 ` [FFmpeg-devel] [PATCH 3/5] lavc/h264qpel: Add vectorized implementation of luma " Arnie Chang
2023-05-09  9:50 ` [FFmpeg-devel] [PATCH 4/5] lavc/h264dsp: Add vectorized implementation of DSP functions " Arnie Chang
2023-05-09  9:50 ` [FFmpeg-devel] [PATCH 5/5] lavc/h264pred: Add vectorized implementation of intra prediction " Arnie Chang
2023-05-09 15:47 ` [FFmpeg-devel] [PATCH 0/5] RISC-V: Improve H264 decoding performance using RVV intrinsic Lynne
2023-05-09 16:51 ` Rémi Denis-Courmont
2023-05-10  8:46   ` Arnie Chang
2023-05-10 11:41     ` Lynne
2023-05-10 12:14     ` Rémi Denis-Courmont

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20230509095030.25506-1-arnie.chang@sifive.com \
    --to=arnie.chang@sifive.com \
    --cc=ffmpeg-devel@ffmpeg.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

This inbox may be cloned and mirrored by anyone:

	git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \
		ffmpegdev@gitmailbox.com
	public-inbox-index ffmpegdev

Example config snippet for mirrors.


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git