[Ffmpeg-devel-irc] ffmpeg-devel.log.20120713

burek burek021 at gmail.com
Sat Jul 14 02:05:04 CEST 2012


[00:57] <CIA-41> ffmpeg: 03Martin Storsjö 07master * r37eeb5e273 10ffmpeg/ (7 files in 3 dirs): 
[00:57] <CIA-41> ffmpeg: Support AAC encoding via the external library fdk-aac
[00:57] <CIA-41> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[00:57] <CIA-41> ffmpeg: 03Ronald S. Bultje 07master * r667fb97a65 10ffmpeg/ (4 files in 3 dirs): 
[00:57] <CIA-41> ffmpeg: dct/fft-test: use a replacement getopt() if the system has none present.
[00:57] <CIA-41> ffmpeg: This allows compiling and running these tests on systems lacking a built-
[00:57] <CIA-41> ffmpeg: in version of getopt(), such as MSVC.
[00:57] <CIA-41> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[00:57] <CIA-41> ffmpeg: 03Diego Biurrun 07master * rd648de61e6 10ffmpeg/ (libavcodec/Makefile libavformat/Makefile): build: Add missing build rules for the ISMV muxer
[00:57] <CIA-41> ffmpeg: 03Diego Biurrun 07master * rb7884ff7af 10ffmpeg/ (libavcodec/Makefile libavformat/Makefile): build: Fix MP2 muxer dependencies
[00:57] <CIA-41> ffmpeg: 03Diego Biurrun 07master * r8eea8fdceb 10ffmpeg/libavcodec/ (Makefile flac.c flacdec.c): 
[00:57] <CIA-41> ffmpeg: flac: Move flac functions shared between libraries to flac common code
[00:57] <CIA-41> ffmpeg: This fixes a number of flac-related build dependencies.
[00:57] <CIA-41> ffmpeg: 03Diego Biurrun 07master * r0da29727ea 10ffmpeg/libavcodec/Makefile: build: Fix Ogg demuxer dependencies
[00:57] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r620c6292b1 10ffmpeg/: (log message trimmed)
[00:57] <CIA-41> ffmpeg: Merge remote-tracking branch 'qatar/master'
[00:57] <CIA-41> ffmpeg: * qatar/master:
[00:57] <CIA-41> ffmpeg:  build: Fix Ogg demuxer dependencies
[00:57] <CIA-41> ffmpeg:  build: Fix FLAC demuxer dependencies
[00:57] <CIA-41> ffmpeg:  flac: Move flac functions shared between libraries to flac common code
[00:57] <CIA-41> ffmpeg:  build: Fix CAF demuxer dependencies
[00:57] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r81fe2e8b36 10ffmpeg/configure: 
[00:57] <CIA-41> ffmpeg: build_system: mark libfdk as non GPL compatible until the situation is clarified by someone qualified.
[00:57] <CIA-41> ffmpeg: 03Martin Storsjö 07master * r372597e538 10ffmpeg/libavcodec/ (avcodec.h options_table.h): 
[00:57] <CIA-41> ffmpeg: libavcodec: Add more AAC profiles
[00:57] <CIA-41> ffmpeg: The numerical values of the profiles are the MPEG4 Audio Object
[00:57] <CIA-41> ffmpeg: Type values, minus one.
[00:57] <CIA-41> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[00:57] <CIA-41> ffmpeg: 03Diego Biurrun 07master * ra519463366 10ffmpeg/libavcodec/Makefile: build: Fix CAF demuxer dependencies
[00:57] <CIA-41> ffmpeg: 03Diego Biurrun 07master * r7d605d5115 10ffmpeg/configure: 
[00:57] <CIA-41> ffmpeg: configure: Drop redundant mxf_d10 test dependency declaration
[00:57] <CIA-41> ffmpeg: The mxf_d10 test depends on avconv, which depends on avfilter, so there
[00:57] <CIA-41> ffmpeg: is no need to declare an avfilter dependency for the mxf_d10 test.
[00:58] <CIA-41> (1 lines omitted)
[01:40] Action: llogan tries fdk-aac
[01:56] <llogan> bah. fdk-aac gives me a segmentation fault.
[02:13] <Daemon404> was it merged properly?
[02:13] <Daemon404> i had backport(forward ported?) to ffmpeg last week
[02:13] <Daemon404> and mine was working fine 
[02:14] <kierank> sideported
[02:14] <Daemon404> yes
[02:14] <Daemon404> theres a term.
[02:14] <llogan> i'm too ignorant to know
[02:16] <Daemon404> yeah
[02:16] <Daemon404> it wasnt converted to ff_alloc_packet2
[02:16] <Daemon404> thats how i fixed it when i ported it.
[02:20] <Daemon404> llogan, 
[02:20] <Daemon404> -    if ((ret = ff_alloc_packet(avpkt, FFMAX(8192, 768 * avctx->channels)))) {
[02:20] <Daemon404> +    if ((ret = ff_alloc_packet2(avctx, avpkt, FFMAX(8192, 768 * avctx->channels)))) {
[02:23] <llogan> Daemon404: thanks
[02:24] <Daemon404> lemme know if it fixes it
[02:24] <Daemon404> and ill send to the ML
[02:26] <durandal_1707> i dont see why ff_alloc_packet could not work
[02:26] <llogan> Daemon404: i started working on something else so i might not be able to test soon
[02:27] <Daemon404> durandal_1707, probably got broken.
[02:27] <Daemon404> and i think everything else uses alloc_packet2
[02:27] <Daemon404> so noone notice
[02:27] <Daemon404> d
[02:30] <durandal_1707> Daemon404: right, 740b9f introduced crash
[02:35] <Daemon404> ...
[02:35] <Daemon404> $ ../ffmpeg/configure
[02:35] <Daemon404> ../ffmpeg/configure: internal error: too many vars
[02:35] <Daemon404> ../ffmpeg/configure[3616]: internal error: too many vars
[02:35] <Daemon404> install prefix            /usr/local
[02:35] <Daemon404> source path               /home/daemon404/dev/ffmpeg
[02:35] <Daemon404> C compiler                gcc
[02:35] <Daemon404> ARCH                      x86 (generic)
[02:35] <Daemon404> ../ffmpeg/configure[3624]: internal error: too many vars
[02:35] <Daemon404> really? openbsd sucks that much?
[02:36] <durandal_1707> bad shell?
[02:36] <llogan> burek: can you fill out the "Change information" box in the wiki for updates?
[02:36] <Daemon404> i recall michaelni talkign abotu openbsd before...
[02:37] <Daemon404> interesting
[02:37] <Daemon404> libav's configure works
[02:37] <Daemon404> ffmpeg's explodes
[02:39] <durandal_1707> Daemon404: libav is also smaller than ffmpeg
[02:39] <Daemon404> sholdnt be THAT much smaller
[02:39] <Daemon404> for the config script
[02:40] <ubitux> iirc openbsd fixed it; you can still use another shell than ksh
[02:41] <sandsmark> why would you use anything but ksh
[02:41] <Daemon404> ubitux, this is in 5.1
[02:41] <durandal_1707> Daemon404: upgrade your OpenBSD remote holes
[02:41] <Daemon404> http://www.openbsd.org/errata51.html
[02:41] <Daemon404> ^ already there
[02:42] <Daemon404> ubitux, and im using /bin.sh
[02:42] <Daemon404> not ksh
[02:42] <llogan> durandal_1707: are you Paul B Mahol?
[02:43] <durandal_1707> ubitux: i think that michaelni patched OpenBSD machines that are on fate
[02:43] <Daemon404> well it's nto an official errata patch
[02:44] <ubitux> the patch was sent and discussed on openbsd bugs
[02:44] <durandal_1707> llogan: I'm durandal_1707
[02:44] <Daemon404> ubitux, what came of it?
[02:44] <ubitux> i thought it get applied
[02:44] <Daemon404> (also lol @ gcc 4.2.1)
[02:44] <Daemon404> ubitux, maybe in cvs, but it didnt make it to any errata
[02:45] <durandal_1707> what is put in errata is political decision
[02:46] <ubitux> http://marc.info/?l=openbsd-bugs&m=132793345717890&w=2
[02:46] <ubitux> http://marc.info/?l=openbsd-bugs&m=132817614319025&w=2
[02:47] <Daemon404> ubitux, im not even using ksh though
[02:47] <ubitux> sh isn't linked to ksh? isn't the configure script trying to use it instead?
[02:47] <Daemon404> $ md5 /bin/ksh /bin/sh
[02:47] <Daemon404> MD5 (/bin/ksh) = abc475c8a2d32bbd434de546b1171df8
[02:47] <Daemon404> MD5 (/bin/sh) = abc475c8a2d32bbd434de546b1171df8
[02:48] <Daemon404> yeah. nvm.
[02:48] <ubitux> :)
[02:50] <Daemon404> "FWIW, the problem doesn't apply to the ffmpeg we currently have in
[02:50] <Daemon404> ports."
[02:50] Action: Daemon404 lol
[02:51] <Daemon404> inb4 0.5
[02:52] <Daemon404> how odd... it was committed to openbsd's cvs in feb, before the 5.1 lock, but 5.1 doesnt have the fix.
[02:55] <durandal_1707> llogan: do you have more questions for me?
[02:57] <durandal_1707> Daemon404: this new encoded is obviously not distributable
[02:59] <Daemon404> its nto obvious whatsoever/
[02:59] <llogan> durandal_1707: no
[03:00] <Daemon404> (my personal interpretation is that it IS gpl-compatible)
[03:00] <Daemon404> but someone is emailing sflc iirc
[03:02] <durandal_1707> Daemon404: so you gonna post/push patch or what?
[03:03] <Daemon404> durandal_1707, nobody confirmed it worked
[03:04] <Daemon404> i will send it  ML once it is confirmed, alogn with a discussion about the regression
[03:04] <durandal_1707> Daemon404: i can confirm that replacing ff_alloc_packet2 with ff_alloc_packet in pamenc.c causes segv
[03:05] <durandal_1707> i just lost half of minute for nothing
[03:05] <Daemon404> ...?
[03:07] <durandal_1707> Daemon404: i just confirmed your earlier statement
[03:07] <Daemon404> yes i know
[03:08] <Daemon404> im going to run a build locally and test fdk-aac
[03:08] <Daemon404> and then send it.
[03:11] <durandal_1707> Daemon404: send patch for what?
[03:13] <Daemon404> moving fdk-aac over to alloc_packet2, as well as start a discussion on the regression in 740b9f
[03:14] <durandal_1707> Daemon404: i have fix for that regression already
[03:14] <Daemon404> fdk should use alloc_packet2 anyway
[03:22] <Daemon404> maybe a second bug in fdk...
[03:22] Action: Daemon404 tests more
[03:29] <durandal_1707> Daemon404: reviewed
[03:42] <CIA-41> ffmpeg: 03Paul B Mahol 07master * rd3abbb1d13 10ffmpeg/libavcodec/utils.c: 
[03:42] <CIA-41> ffmpeg: lavc: fix ff_alloc_packet()
[03:42] <CIA-41> ffmpeg: Regression introduced in 740b9ff44.
[03:42] <CIA-41> ffmpeg: Signed-off-by: Paul B Mahol <onemda at gmail.com>
[03:43] <Daemon404> heh, nice extremely undescriptive, unreviewed, and poorly formatted (code) commit :|
[03:45] <durandal_1707> i dont have double standards
[03:45] <Daemon404> hmm?
[07:40] <ubitux> burek: sorry for the delay with -re, i'll commit tonight
[11:41] <burek> ubitux, ok :) thanks again :)
[16:04] <ubitux> http://git.videolan.org/?p=ffmpeg.git;a=blob;f=libavfilter/libmpcodecs/vf_geq.c;hb=HEAD#l55
[16:05] <ubitux> xi = x = [...];
[16:05] <ubitux> x -= xi;
[16:05] <ubitux> ’ x = 0, right?
[16:05] <ubitux> or i'm missing something?
[16:06] <ubitux> oh the types.
[16:06] Action: ubitux feels stupid again
[16:43] <burek> ubitux, can I bother you for one more docs edit :)
[16:43] <burek> (btw I see it will be easier that I learn git usage) :)
[16:44] <burek> The synopsis for FFmpeg cmd line usage does not cover multiple inputs/outputs, so it might me a good idea either to expand the generic syntax line or to provide additional line, describing cmd usage for multiple inputs/outputs
[16:44] <burek> something like: The generic syntax is: ffmpeg [global options] [[infile options][‘-i’ infile1]] [[infile options][‘-i’ infile2]] [[infile options][‘-i’ infile3]]... {[outfile options] outfile1}... {[outfile options] outfile2}... {[outfile options] outfile3}...
[16:49] <burek> btw, is this possible in ffmpeg, using only 1 line: ffmpeg -i input -vcodec ... -acodec ... -f (some_cool_piping_format) | ffmpeg -i - -c copy out1 -c copy out2 -c copy out3 ...
[17:00] <yetifoot> burek, you could do that with image2pipe, but with audio i don't know
[17:00] <yetifoot> oh sorry nevermind
[17:04] <Daemon404> burek, eh?
[17:04] <Daemon404> why would you need to do that
[17:04] <Daemon404> ffmpeg supports multiple output files
[17:04] <burek> Daemon404, because this way you have only 1 encoding process
[17:05] <Daemon404> humm i see what you mean
[17:05] <Daemon404> in that case, it is entirely possible (i think)
[18:05] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * rc5cc94982d 10ffmpeg/libavformat/matroskaenc.c: 
[18:06] <CIA-41> ffmpeg: matroskaenc: switch to av_assert
[18:06] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:06] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r1cef5a0196 10ffmpeg/ffmpeg.c: 
[18:06] <CIA-41> ffmpeg: ffmpeg: fix type of av_log() argument
[18:06] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:06] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r790e25ef61 10ffmpeg/doc/general.texi: 
[18:06] <CIA-41> ffmpeg: general.texi: fix project reference
[18:06] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:06] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * rff2fda57df 10ffmpeg/doc/protocols.texi: 
[18:06] <CIA-41> ffmpeg: doc/protocol.texi: fix missed application reference
[18:06] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:43] <CIA-41> ffmpeg: 03burek 07master * rcb59e62abb 10ffmpeg/doc/ffmpeg.texi: 
[18:43] <CIA-41> ffmpeg: doc/ffmpeg: document more deeply -re option.
[18:43] <CIA-41> ffmpeg: Signed-off-by: Clément BSsch <ubitux at gmail.com>
[18:43] <CIA-41> ffmpeg: Signed-off-by: Lou Logan <lou at lrcd.com>
[18:45] <ubitux> burek: the double [[ ]] around infile stuff most likely mean you can have it multiple times
[18:45] <ubitux> burek: i'de better add more example a bit beloew
[18:45] <ubitux> -e
[18:45] <ubitux> -e two times.
[19:41] <burek> ubitux, is there (anywhere) in those docs, explained what does [] [[]] `` mean?
[19:59] <ubitux> i guess it's a convention
[20:00] <ubitux> most ppl don't try to decipher that anyway
[20:00] <ubitux> they just look at the examples
[20:01] <ubitux> and try to match their needs using various examples
[20:04] <burek> exactly my point too :)
[20:36] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r0eb0b310c4 10ffmpeg/libavcodec/ (avcodec.h options_table.h utils.c version.h): 
[20:36] <CIA-41> ffmpeg: lavc: add pkt_timebase to allow avcodec to adjust pkt_dts/pts when needed
[20:36] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:11] <CIA-41> ffmpeg: 03Compn 07master * rc3a4c1bd5a 10ffmpeg/libavformat/riff.c: 
[21:11] <CIA-41> ffmpeg: add PLV1 fourcc to mpeg4
[21:11] <CIA-41> ffmpeg: works on http://samples.ffmpeg.org/V-codecs/PLV1/york-water-meteor.avi
[21:13] <Compn> interesting
[21:13] <Compn> new version of pelco codec uses ffmpeg
[21:13] <Daemon404> inb4 illegally
[21:13] <Compn> :\
[21:41] <sandsmark> http://tdistler.com/2011/01/06/ffmpeg-and-invalid-buffer-reads seems like they have been using it for a while
[21:42] <kierank> they should use av_malloc
[21:53] <durandal_1707> Daemon404: have you ever heard about not commiting own work on master branch?
[21:59] <Daemon404> durandal_1707, um what?
[22:00] <Daemon404> if youre referring to my github, that is where michael merges from.
[22:13] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r051dd2eea6 10ffmpeg/libavcodec/ffv1.c: 
[22:13] <CIA-41> ffmpeg: ffv1dec: print timestamp at which a CRC mismatch happened
[22:13] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[22:13] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * raf285072b5 10ffmpeg/libavcodec/ffv1.c: 
[22:13] <CIA-41> ffmpeg: ffv1enc: use default number of slices if unspecified
[22:13] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[22:54] <CIA-41> ffmpeg: 03Martin Storsjö 07master * r7548831c5e 10ffmpeg/libavcodec/libfdk-aacenc.c: 
[22:54] <CIA-41> ffmpeg: libfdk-aac: Allow setting the encoder bandwidth/cutoff frequency
[22:54] <CIA-41> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[22:54] <CIA-41> ffmpeg: 03Kostya Shishkov 07master * r5b7e3a195d 10ffmpeg/ (8 files in 4 dirs): MS Expression Encoder Screen decoder
[22:54] <CIA-41> ffmpeg: 03Diego Biurrun 07master * r47aed43990 10ffmpeg/libavformat/Makefile: build: Fix APE tag dependencies
[22:54] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r674216991d 10ffmpeg/: (log message trimmed)
[22:54] <CIA-41> ffmpeg: Merge remote-tracking branch 'qatar/master'
[22:54] <CIA-41> ffmpeg: * qatar/master:
[22:54] <CIA-41> ffmpeg:  build: Fix APE tag dependencies
[22:54] <CIA-41> ffmpeg:  MS Expression Encoder Screen decoder
[22:54] <CIA-41> ffmpeg:  libfdk-aac: Allow setting the encoder bandwidth/cutoff frequency
[22:54] <CIA-41> ffmpeg: Conflicts:
[23:09] <CIA-41> ffmpeg: 03Paul B Mahol 07master * r1033807313 10ffmpeg/libavcodec/exr.c: 
[23:09] <CIA-41> ffmpeg: exr: frame multithreading support
[23:09] <CIA-41> ffmpeg: Signed-off-by: Paul B Mahol <onemda at gmail.com>
[23:15] <CIA-41> ffmpeg: 03Paul B Mahol 07master * rfedefe4a87 10ffmpeg/libavcodec/exr.c: 
[23:15] <CIA-41> ffmpeg: exr: remove double ;;
[23:15] <CIA-41> ffmpeg: Signed-off-by: Paul B Mahol <onemda at gmail.com>
[00:00] --- Sat Jul 14 2012


More information about the Ffmpeg-devel-irc mailing list