ffmpeg | branch: master | Pierre-Anthony Lemieux | Mon Sep
19 08:26:18 2022 -0700| [0e402ebf487075727187c78f85908a492eed3e59] | committer:
Zane van Iperen
format/imfdec: improve error handling when selecting tracks for playback
Reviewed-by: Zane van Iperen
>
ffmpeg | branch: master | Lynne | Wed Sep 28 00:07:53 2022
+0200| [0bc7ba448868f74866800f956ad7e4835b976f47] | committer: Lynne
configure: remove mdct15 from the encoder/decoder's list of requirements
>
ffmpeg | branch: master | Andreas Rheinhardt |
Tue Sep 27 01:40:24 2022 +0200| [2664b39d54aa44ba185ea39c3933b80d9dfcb542] |
committer: Andreas Rheinhardt
avocdec/snowenc: Fix left shift of negative number
Fixes the vsynth(1|2|_lena)-snow-ll FATE-tests.
Reviewed-by: Michael Niedermayer
ffmpeg | branch: master | Timo Rothenpieler | Tue Sep
27 19:35:37 2022 +0200| [59cb0bd23d61f6ea3bfd86558346e2720aba7f06] | committer:
Timo Rothenpieler
avfilter/vf_extractplanes: add missing break; statement
>
ffmpeg | branch: master | Derek Buitenhuis | Thu
Sep 22 14:11:31 2022 +0100| [f5cd00bf5265cb1d10de925c4d0b3ebb191f0d74] |
committer: Derek Buitenhuis
ffprobe: Check for invalid matrix error when printing rotation
av_display_rotation_get will return NAN when the display matrix is invalid,
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:49 2022 +0300| [09f907999f6ff4204d5848e5fd01e1143cb76d9c] | committer:
Lynne
lavc/aacpsdsp: RISC-V V hybrid_analysis_ileave
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=09f907999f6ff4204d5848e5fd01e1143cb76d9c
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:51 2022 +0300| [c03f9654c997b33b8028eb71c9e7ba61fd53a813] | committer:
Lynne
lavc/aacpsdsp: RISC-V V stereo_interpolate[0]
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=c03f9654c997b33b8028eb71c9e7ba61fd53a813
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:50 2022 +0300| [a15edb0bc0108362fa3c71de3bf763072341b8b0] | committer:
Lynne
lavc/aacpsdsp: RISC-V V hybrid_synthesis_deint
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=a15edb0bc0108362fa3c71de3bf763072341b8b0
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:48 2022 +0300| [15c3a0bd6ec1383e26ca6a41ea8daa95dc0e1736] | committer:
Lynne
lavc/aacpsdsp: RISC-V V hybrid_analysis
This starts with one-time initialisation of the 26 constant factors
like
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:47 2022 +0300| [e180326a0b72bbbdee51810592b16178a48797f3] | committer:
Lynne
lavc/aacpsdsp: RISC-V V mul_pair_single
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=e180326a0b72bbbdee51810592b16178a48797f3
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:42 2022 +0300| [f41ae62f39ad1f91a3817325fdbba7304aba4641] | committer:
Lynne
lavc/audiodsp: RISC-V V scalarproduct_int16
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=f41ae62f39ad1f91a3817325fdbba7304aba4641
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:46 2022 +0300| [b0cacf4c3f8fe451e22a43462ea68ffa1e2e09ec] | committer:
Lynne
lavc/aacpsdsp: RISC-V V add_squares
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=b0cacf4c3f8fe451e22a43462ea68ffa1e2e09ec
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:45 2022 +0300| [453aba71e6db37d367269cc080dae4c1548db33a] | committer:
Lynne
lavc/vorbisdsp: RISC-V V inverse_coupling
This uses the following vectorisation:
for (i = 0; i < blocksize; i++) {
ang[i] = mag[i] -
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:44 2022 +0300| [220dfd0945ee8481d8cdbf713f515a94ceee9992] | committer:
Lynne
lavc/fmtconvert: RISC-V V int32_to_float_fmul_array8
>
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:41 2022 +0300| [f127a5d29d7aee99cb4dd4d576d375c678b3c388] | committer:
Lynne
lavc/audiodsp: RISC-V V vector_clipf
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=f127a5d29d7aee99cb4dd4d576d375c678b3c388
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:43 2022 +0300| [47a10b9a99130457c27b220afba7d7de4a69bb18] | committer:
Lynne
lavc/fmtconvert: RISC-V V int32_to_float_fmul_scalar
>
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:40 2022 +0300| [27da9514c3a7b506e3c3e0ada5dee67566c5aadf] | committer:
Lynne
lavc/audiodsp: RISC-V V vector_clip_int32
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=27da9514c3a7b506e3c3e0ada5dee67566c5aadf
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:39 2022 +0300| [c1bb19e263e14887ad286c16352edbaa39be4f66] | committer:
Lynne
lavu/fixeddsp: RISC-V V butterflies_fixed
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=c1bb19e263e14887ad286c16352edbaa39be4f66
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:38 2022 +0300| [cd77662953942cf166f0115b21cb8619e2e8f630] | committer:
Lynne
lavu/floatdsp: RISC-V V scalarproduct_float
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=cd77662953942cf166f0115b21cb8619e2e8f630
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:37 2022 +0300| [b493370662eaa87f84450e876b8b61b34d91f23d] | committer:
Lynne
lavu/floatdsp: RISC-V V vector_fmul_window
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=b493370662eaa87f84450e876b8b61b34d91f23d
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:36 2022 +0300| [9aeb6aca3a9fa8caf8db263de4fa9f8077ba6634] | committer:
Lynne
lavu/floatdsp: RISC-V V vector_fmul_reverse
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=9aeb6aca3a9fa8caf8db263de4fa9f8077ba6634
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:33 2022 +0300| [d120ab5b915fb422247f9e4dd5b823d1b47d6adf] | committer:
Lynne
lavu/floatdsp: RISC-V V vector_dmac_scalar
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=d120ab5b915fb422247f9e4dd5b823d1b47d6adf
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:32 2022 +0300| [c3db27ba9552fc3cb764bf9eb717840d981c5779] | committer:
Lynne
lavu/floatdsp: RISC-V V vector_fmac_scalar
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=c3db27ba9552fc3cb764bf9eb717840d981c5779
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:35 2022 +0300| [47ce9735cc8362ea6b3a508dd660233af28d77ba] | committer:
Lynne
lavu/floatdsp: RISC-V V butterflies_float
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=47ce9735cc8362ea6b3a508dd660233af28d77ba
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:34 2022 +0300| [f4ea45040f4c0d896fce66a6f52deb53249ed77f] | committer:
Lynne
lavu/floatdsp: RISC-V V vector_fmul_add
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=f4ea45040f4c0d896fce66a6f52deb53249ed77f
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:29 2022 +0300| [89b7ec65a8c50a88ca730c666410f5743f600634] | committer:
Lynne
lavu/floatdsp: RISC-V V vector_dmul_scalar
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=89b7ec65a8c50a88ca730c666410f5743f600634
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:31 2022 +0300| [da169a210dbd33d5d34baaf3d188ca5388b1b267] | committer:
Lynne
lavu/floatdsp: RISC-V V vector_dmul
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=da169a210dbd33d5d34baaf3d188ca5388b1b267
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:25 2022 +0300| [0c0a3deb1826638915775daa7cefb891a300060b] | committer:
Lynne
lavu/cpu: CPU flags for the RISC-V Vector extension
RVV defines a total of 12 different extensions, including:
- 5 different instruction subsets:
-
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:30 2022 +0300| [7058af9969b737adbb1cd302cf8fa5feb7bc9e2b] | committer:
Lynne
lavu/floatdsp: RISC-V V vector_fmul
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=7058af9969b737adbb1cd302cf8fa5feb7bc9e2b
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:24 2022 +0300| [1edac8eb468cb9a769ded6197dbce515b71b137e] | committer:
Lynne
lavc/pixblockdsp: RISC-V I get_pixels
Benchmarks on SiFive U74-MC (courtesy of Shanghai StarFive Tech):
get_pixels_c: 180.0
get_pixels_rvi: 136.7
>
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:28 2022 +0300| [a6c10d05fe26dcfb5920fa4ce8ea8f74bf5f82dc] | committer:
Lynne
lavu/floatdsp: RISC-V V vector_fmul_scalar
This is based on existing code from the VLC git tree with two minor
changes to account for the different
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:26 2022 +0300| [1b6aee52a5634976c6fd212a28175512eb7865f8] | committer:
Lynne
configure: probe RISC-V Vector extension
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=1b6aee52a5634976c6fd212a28175512eb7865f8
---
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:27 2022 +0300| [39357cad37e865b723c7a97adc27d4a6b4388008] | committer:
Lynne
lavu/riscv: fallback macros for SH{1, 2, 3}ADD
Those mnemonics require the very latest binutils release at the time of
writing. These macros provide
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:23 2022 +0300| [04d092e7d5204f1aebb7e61f92bb263873e0f735] | committer:
Lynne
lavc/audiodsp: RISC-V F vector_clipf
RV64G supports MIN & MAX instructions natively only on floating point
registers, not general purpose ones. The
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:22 2022 +0300| [746f1ff36ac0d232687820fbde4e4efc79093af7] | committer:
Lynne
lavu/riscv: initial common header for assembler macros
>
ffmpeg | branch: master | Rémi Denis-Courmont | Mon Sep 26
17:52:21 2022 +0300| [b95e2fbd85e93b58d019ad0e3da64a72ee7cc3d5] | committer:
Lynne
lavu/cpu: detect RISC-V base extensions
This introduces compile-time and run-time CPU detection on RISC-V. In
practice, I doubt that FFmpeg will ever
36 matches
Mail list logo
