Re: [FFmpeg-devel] [PATCH] avfilter/vf_quirc: add missing header to unbreak build

2024-08-20 Thread Gyan Doshi
Patch withdrawn, superfluous since 1afe42852b. On 2024-08-19 08:33 pm, Gyan Doshi wrote: filters.h has been needed since 42cbf66fff for FilterLink. --- libavfilter/vf_quirc.c | 1 + 1 file changed, 1 insertion(+) diff --git a/libavfilter/vf_quirc.c b/libavfilter/vf_quirc.c index 760f5d97de..

[FFmpeg-devel] [PATCH 1/6] avcodec/magicyuvenc: better slice height

2024-08-20 Thread Michael Niedermayer
Fixes: Use of uninitialized value Fixes: 71072/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_MAGICYUV_fuzzer-4835252046987264 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer --- libavcodec/magicyuvenc.c

[FFmpeg-devel] [PATCH 2/6] avformat/apetag: Check APETAGEX

2024-08-20 Thread Michael Niedermayer
Fixes: Use of uninitialized value Fixes: 71074/clusterfuzz-testcase-minimized-ffmpeg_IO_DEMUXER_fuzzer-5697034877730816 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer --- libavformat/apetag.c | 3 ++- 1 fil

[FFmpeg-devel] [PATCH 3/6] avcodec/vc1_block: propagate error codes

2024-08-20 Thread Michael Niedermayer
Fixes: use of uninitialized value Fixes: 71228/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_VC1IMAGE_fuzzer-6188476880453632 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer --- libavcodec/vc1_block.c |

[FFmpeg-devel] [PATCH 4/6] avcodec/notchlc: Check bytes left before reading

2024-08-20 Thread Michael Niedermayer
Fixes: Use of uninitialized value Fixes: 71230/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_NOTCHLC_fuzzer-4624502095413248 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer --- libavcodec/notchlc.c | 3 +

[FFmpeg-devel] [PATCH 5/6] tools/target_swr_fuzzer: Check av_samples_fill_arrays() for failure

2024-08-20 Thread Michael Niedermayer
Fixes: use of uninitialized value Fixes: 71242/clusterfuzz-testcase-minimized-ffmpeg_SWR_fuzzer-4905557943713792 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer --- tools/target_swr_fuzzer.c | 9 +++-- 1

[FFmpeg-devel] [PATCH 6/6] avformat/argo_brp: Check that ASF chunk header is completely read

2024-08-20 Thread Michael Niedermayer
Fixes: Use of uninitialized value Fixes: 71280/clusterfuzz-testcase-minimized-ffmpeg_dem_ARGO_BRP_fuzzer-4692991866896384 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer --- libavformat/argo_brp.c | 4 ++--

Re: [FFmpeg-devel] Does rtspenc actually support AVFMT_GLOBALHEADER?

2024-08-20 Thread John Cox
On Mon, 19 Aug 2024 at 19:32, Martin Storsjö wrote: > > On Mon, 19 Aug 2024, John Cox wrote: > > > Does rtspenc actually support AVFMT_GLOBALHEADER? It is specified in the > > FFOutputFormat flags but I can't see anywhere in the code where > > extradata is referenced like it is in other output for

[FFmpeg-devel] [FEATURE] Cut a video (-ss) with timings non-aligned on keyframes, with minimal re-encoding

2024-08-20 Thread basj
>> More generally, which is the recommanded way to cut a video with a specific >> starting point and specific length, with minimal re-encoding? >> Millions of hours of CPU-time are probably wasted to reencode >> already-perfectly-encoded content, just for cutting ;) > >Only do remux without trans

Re: [FFmpeg-devel] [FEATURE] Cut a video (-ss) with timings non-aligned on keyframes, with minimal re-encoding

2024-08-20 Thread Lynne via ffmpeg-devel
On 20/08/2024 14:13, b...@gget.it wrote: More generally, which is the recommanded way to cut a video with a specific starting point and specific length, with minimal re-encoding? Millions of hours of CPU-time are probably wasted to reencode already-perfectly-encoded content, just for cutting ;)

[FFmpeg-devel] [PATCH v2 0/4] add bdof to the vvc decoder

2024-08-20 Thread Nuo Mi
Major changes since v1: remove hadd and pmulld as James suggested reduce vextracti128 Nuo Mi (4): avcodec/vvcdec: misc, rename BDOF_BLOCK_SIZE to BDOF_MIN_BLOCK_SIZE avcodec/vvcdec: bdof, do not pad sources and gradients to simplify the code x86/vvcdec: inter, add optical flow avx2 code

[FFmpeg-devel] [PATCH v2 1/4] avcodec/vvcdec: misc, rename BDOF_BLOCK_SIZE to BDOF_MIN_BLOCK_SIZE

2024-08-20 Thread Nuo Mi
--- libavcodec/vvc/dsp.c| 4 ++-- libavcodec/vvc/inter_template.c | 10 +- 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/libavcodec/vvc/dsp.c b/libavcodec/vvc/dsp.c index 648d54ebb2..7463d8c9de 100644 --- a/libavcodec/vvc/dsp.c +++ b/libavcodec/vvc/dsp.c @@ -8

[FFmpeg-devel] [PATCH v2 2/4] avcodec/vvcdec: bdof, do not pad sources and gradients to simplify the code

2024-08-20 Thread Nuo Mi
--- libavcodec/vvc/dsp.c| 25 + libavcodec/vvc/dsp.h| 4 +- libavcodec/vvc/inter_template.c | 65 ++--- 3 files changed, 31 insertions(+), 63 deletions(-) diff --git a/libavcodec/vvc/dsp.c b/libavcodec/vvc/dsp.c index 7463d8c9de..43

[FFmpeg-devel] [PATCH v2 3/4] x86/vvcdec: inter, add optical flow avx2 code

2024-08-20 Thread Nuo Mi
BDoF used about 10%–25% of the CPU for some clips. Here are the FPS for one run; please ignore the negative values, as they may be due to round-to-round variation clips | before | after | delta ||---|---

[FFmpeg-devel] [PATCH v2 4/4] checkasm: add vvc_bdof test

2024-08-20 Thread Nuo Mi
apply_bdof_8_8x16_c: 5776.5 apply_bdof_8_8x16_avx2: 396.2 apply_bdof_8_16x8_c: 5722.0 apply_bdof_8_16x8_avx2: 216.0 apply_bdof_8_16x16_c: 11213.2 apply_bdof_8_16x16_avx2: 434.5 apply_bdof_10_8x16_c: 5657.7 apply_bdof_10_8x16_avx2: 1096.0 apply_bdof_10_16x8_c: 5531.7 apply_bdof_10_16x8_avx2: 212.5 a

Re: [FFmpeg-devel] [PATCH 3/4] x86/vvcdec: inter, add optical flow avx2 code

2024-08-20 Thread Nuo Mi
On Sun, Aug 18, 2024 at 11:18 AM James Almer wrote: > On 8/17/2024 10:48 PM, Nuo Mi wrote: > > +pxorm6, m6 > > +phaddw m%2, m6 > > +phaddw m%2, m6 > > Horizonal adds are slow. Can't you do this with normal adds, shifts and > blend? >

Re: [FFmpeg-devel] [FEATURE] Cut a video (-ss) with timings non-aligned on keyframes, with minimal re-encoding

2024-08-20 Thread Michael Niedermayer
On Tue, Aug 20, 2024 at 02:36:53PM +0200, Lynne via ffmpeg-devel wrote: > On 20/08/2024 14:13, b...@gget.it wrote: > > > > More generally, which is the recommanded way to cut a video with a > > > > specific starting point and specific length, with minimal re-encoding? > > > > Millions of hours of

Re: [FFmpeg-devel] [PATCH 1/3] lavu/opt: document underlying C types for enum AVOptionType

2024-08-20 Thread Michael Niedermayer
Hi On Sun, Aug 18, 2024 at 01:26:53PM +0200, Anton Khirnov wrote: > --- > libavutil/opt.h | 78 +++-- > 1 file changed, 75 insertions(+), 3 deletions(-) > > diff --git a/libavutil/opt.h b/libavutil/opt.h > index 07e27a9208..23bc495158 100644 > --- a/li

Re: [FFmpeg-devel] [FEATURE] Cut a video (-ss) with timings non-aligned on keyframes, with minimal re-encoding

2024-08-20 Thread Lynne via ffmpeg-devel
On 20/08/2024 22:33, Michael Niedermayer wrote: On Tue, Aug 20, 2024 at 02:36:53PM +0200, Lynne via ffmpeg-devel wrote: On 20/08/2024 14:13, b...@gget.it wrote: More generally, which is the recommanded way to cut a video with a specific starting point and specific length, with minimal re-encod

Re: [FFmpeg-devel] [PATCH 2/3] lavu/opt: forward av_opt_get_video_rate() to av_opt_get_q()

2024-08-20 Thread Michael Niedermayer
On Sun, Aug 18, 2024 at 01:26:54PM +0200, Anton Khirnov wrote: > The two functions are exactly the same. > --- > libavutil/opt.c | 13 + > 1 file changed, 1 insertion(+), 12 deletions(-) LGTM thx [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB No g

[FFmpeg-devel] [PATCH] libavdevice: fix compilation for Mac OS X 10.7-10.12, iOS < 11

2024-08-20 Thread Erik Bråthen Solem
avfoundation.m uses constants prefixed with AVMediaType on Mac OS X > 10.6. In 10.7 through 10.12 their type was NSString*, but starting with 10.13 a new AVMediaType struct type was introduced. In avfoundation.m, the function getDevicesWithMediaType takes this struct as parameter, which breaks supp

Re: [FFmpeg-devel] [PATCH] libavdevice: fix compilation for Mac OS X 10.7-10.12, iOS < 11

2024-08-20 Thread epirat07
On 21 Aug 2024, at 0:43, Erik Bråthen Solem wrote: > avfoundation.m uses constants prefixed with AVMediaType on Mac OS X > 10.6. > In 10.7 through 10.12 their type was NSString*, but starting with 10.13 a > new AVMediaType struct type was introduced. In avfoundation.m, the function > getDevicesW

Re: [FFmpeg-devel] [PATCH] libavdevice: fix compilation for Mac OS X 10.7-10.12, iOS < 11

2024-08-20 Thread Gnattu OC via ffmpeg-devel
> On Aug 21, 2024, at 07:17, epira...@gmail.com wrote: > > > > On 21 Aug 2024, at 0:43, Erik Bråthen Solem wrote: > >> avfoundation.m uses constants prefixed with AVMediaType on Mac OS X > 10.6. >> In 10.7 through 10.12 their type was NSString*, but starting with 10.13 a >> new AVMediaType st

Re: [FFmpeg-devel] [PATCH] libavdevice: fix compilation for Mac OS X 10.7-10.12, iOS < 11

2024-08-20 Thread epirat07
On 21 Aug 2024, at 3:09, Gnattu OC via ffmpeg-devel wrote: >> On Aug 21, 2024, at 07:17, epira...@gmail.com wrote: >> >> >> >> On 21 Aug 2024, at 0:43, Erik Bråthen Solem wrote: >> >>> avfoundation.m uses constants prefixed with AVMediaType on Mac OS X > 10.6. >>> In 10.7 through 10.12 their typ

[FFmpeg-devel] [PATCH] fate/jpeg2000dec: add selected conformance codestreams from ISO/IEC 15444-4

2024-08-20 Thread pal
From: Pierre-Anthony Lemieux --- tests/fate/jpeg2000.mak | 129 +-- tests/ref/fate/jpeg2000dec-ds0_hm_15_b8 | 6 ++ tests/ref/fate/jpeg2000dec-ds0_ht_02_b11 | 6 ++ tests/ref/fate/jpeg2000dec-ds0_ht_02_b12 | 6 ++ tests/ref/fate/jpeg2000dec-ds0_ht_03_b