[Ffmpeg-devel-irc] ffmpeg-devel.log.20111222

[burek]

[00:02] <funman> michaelni: can you shout when you merge http://git.libav.org/?p=libav.git;a=commitdiff;h=2003a29284660255016b1ec4bd5dfb8d53e6a852 ? that will fix headaches in VLC
[00:03] <j-b> can you fix my headaches too?
[00:03] <j-b> just asking :)
[00:04] <nevcairiel> I still wonder why i never ran into the freeze or never had someone report one, maybe it really was something odd about vlc itself /me shrugs
[00:05] <j-b> on XP?
[00:05] <nevcairiel> any kind of windows really
[00:06] <j-b> it's almost unnoticeable on 7/VIsta but is on very-threaded context on XP, like VLC or multiple browser plugins like on Firefox.
[00:06] <CIA-49> ffmpeg: 03Tim Nicholson 07master * r7028c9f421 10ffmpeg/doc/ffmpeg.texi: Clearer explanation of audio channel mapping restrictions
[00:07] <nevcairiel> I just would've thought that some user of my DirectShow filter would have maybe experienced the freeze and reported it, but oh well, its fine this way for me too!
[00:07] <j-b> how much do you default your threads-number to?
[00:07] <nevcairiel> cpus * 1.5
[00:08] <nevcairiel> seemed to be the performance sweet spot
[00:08] <nevcairiel> but i'm not complaining, i just think its curious
[00:14] <Daemon404> only 2 more years
[00:14] <Daemon404> until xp extended support dies
[00:14] <Daemon404> :)
[00:15] <nevcairiel> The last hour i feel like i wanted to do something but forgot what .. i hate that..
[00:15] <j-b> oh, yeah \o/ 2014
[00:15] <Daemon404> i thought it was 2013
[00:15] <nevcairiel> 08/04/2014
[00:16] <Daemon404> damn
[00:16] <Daemon404> ms is supporting xp out-the-ass
[00:16] <nevcairiel> so roughly 2 years and 4 month
[00:16] Action: Daemon404 blames lolenterprise
[00:34] <pasteeater> my state government still uses xp, or at least the department i last visited
[00:34] <CIA-49> ffmpeg: 03Michael Niedermayer 07master * r8d960fbc70 10ffmpeg/libavformat/ipmovie.c: 
[00:34] <CIA-49> ffmpeg: ipmovie: fix FPE
[00:34] <CIA-49> ffmpeg: Fixed Ticket807
[00:34] <CIA-49> ffmpeg: Bug found by: Oana Stratulat
[00:34] <CIA-49> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:34] <CIA-49> ffmpeg: 03Michael Niedermayer 07master * r2b73cddd40 10ffmpeg/libavcodec/proresdec2.c: 
[00:34] <CIA-49> ffmpeg: proresdec: Check yuv slice data sizes.
[00:34] <CIA-49> ffmpeg: Fixes overread
[00:34] <CIA-49> ffmpeg: Fixes Ticket812
[00:34] <CIA-49> ffmpeg: Bug found by: Oana Stratulat
[00:34] <CIA-49> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:34] <CIA-49> ffmpeg: 03Michael Niedermayer 07master * ra40f43db64 10ffmpeg/libavcodec/error_resilience.c: 
[00:34] <CIA-49> ffmpeg: error_concealment: optimize guess_dc()
[00:34] <CIA-49> ffmpeg: Fixes Ticket811
[00:34] <CIA-49> ffmpeg: Bug found by: Oana Stratulat
[00:34] <CIA-49> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:44] <ubitux> "Buffering several frames is not supported. "
[00:44] <ubitux> huh.
[00:44] <ubitux> what's the point of the buffer filter?
[00:45] <caytchen> the buffer filter can not buffer several frames?
[00:45] <ubitux> according to that error message :p
[00:45] <ubitux> and i see only a single picref in the context struct
[00:51] <ubitux> i also don't get why the buffer filter has some scale heuristic since there is already a code to insert a scale filter in the filtergraph if needed
[01:11] <michaelni> ubitux, its for changing dimensions midstream
[02:05] <CIA-49> ffmpeg: 03Mans Rullgard 07master * rb1bcddfb6d 10ffmpeg/libpostproc/ (postprocess_altivec_template.c postprocess_template.c): 
[02:05] <CIA-49> ffmpeg: postproc: altivec: fix trivial cases of mixed declarations and code
[02:05] <CIA-49> ffmpeg: This moves declarations without initialisers or with constant
[02:05] <CIA-49> ffmpeg: initialisers to the start of a block, and adds do {} while(0)
[02:05] <CIA-49> ffmpeg: around some macros, thus allowing declarations within them.
[02:05] <CIA-49> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[02:05] <CIA-49> ffmpeg: 03Mans Rullgard 07master * r514294e8ca 10ffmpeg/libpostproc/postprocess_altivec_template.c: 
[02:05] <CIA-49> ffmpeg: postproc: altivec: fix remaining mixed declarations and code
[02:05] <CIA-49> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[02:05] <CIA-49> ffmpeg: 03Mans Rullgard 07master * rb49b836f3a 10ffmpeg/libpostproc/postprocess_altivec_template.c: 
[02:05] <CIA-49> ffmpeg: postproc: altivec: untemplate some variable names
[02:05] <CIA-49> ffmpeg: These variables are now in separate scopes and thus do not
[02:05] <CIA-49> ffmpeg: need unique names.
[02:05] <CIA-49> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[02:05] <CIA-49> ffmpeg: x86: Tighten register constraints for decode_significance*_x86.
[02:05] <CIA-49> ffmpeg: On 32-bit OS X with gcc 4.0/4.2 and shared libraries enabled, the ebx register
[02:05] <CIA-49> ffmpeg: is not available, but required to assemble the functions.
[02:05] <CIA-49> ffmpeg: This reverts commit 8742a4f to a simplified version of the original constraints.
[02:05] <CIA-49> ffmpeg: 03Tomas Härdin 07master * r6547fd9264 10ffmpeg/libavformat/mxfdec.c: 
[02:05] <CIA-49> ffmpeg: mxfdec: consider QuantizationBits between 17 and 24 to be pcm_s24*
[02:05] <CIA-49> ffmpeg: This fixes playback of BRD38772509.mxf.
[02:05] <CIA-49> ffmpeg: Signed-off-by: Janne Grunau <janne-libav at jannau.net>
[02:05] <CIA-49> ffmpeg: 03Martin Storsjö 07master * r9c984b1bd7 10ffmpeg/configure: 
[02:05] <CIA-49> ffmpeg: configure: Show whether the safe bitstream reader is enabled
[02:05] <CIA-49> (47 lines omitted)
[02:05] <CIA-49> ffmpeg: 03Tomas Härdin 07master * rc15da5941d 10ffmpeg/libavformat/mxf.c: 
[02:05] <CIA-49> ffmpeg: mxf: Add PictureEssenceCoding UL for V210.
[02:05] <CIA-49> ffmpeg: Signed-off-by: Janne Grunau <janne-libav at jannau.net>
[02:05] <CIA-49> ffmpeg: 03Michael Niedermayer 07master * r52c522c720 10ffmpeg/: (log message trimmed)
[02:05] <CIA-49> ffmpeg: Merge remote-tracking branch 'qatar/master'
[02:05] <CIA-49> ffmpeg: * qatar/master: (27 commits)
[02:05] <CIA-49> ffmpeg:  asfdec: add side data to ASFStream packet instead of output packet.
[02:05] <CIA-49> ffmpeg:  idroqdec: set AVFMTCTX_NOHEADER and create streams as they occur.
[02:05] <CIA-49> ffmpeg:  nellymoserdec: Indicate that the decoder can handle changed parameters
[02:05] <CIA-49> ffmpeg:  libavcodec: Apply parameter change side data when decoding audio
[02:06] <CIA-49> ffmpeg: libavformat: Add a utility function for adding parameter change side data
[02:06] <CIA-49> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[02:06] <CIA-49> ffmpeg: 03Martin Storsjö 07master * r251f320f7d 10ffmpeg/libavformat/flvdec.c: (log message trimmed)
[02:06] <CIA-49> ffmpeg: flvdec: Export new AAC/H.264 extradata as side data on the next packet
[02:06] <CIA-49> ffmpeg: Compared to just overwriting the old extradata, this has the
[02:06] <CIA-49> ffmpeg: advantage of letting the decoder know exactly when the
[02:06] <CIA-49> ffmpeg: extradata changed (otherwise it is changed immediately when the
[02:06] <CIA-49> ffmpeg: new extradata packet is demuxed, even if there's old queued packets
[02:06] <CIA-49> ffmpeg: awaiting to be decoded). This makes it easier for decoders to
[02:06] <CIA-49> (8 lines omitted)
[02:09] <michaelni> funman, j-b 2003a29284660255016b1ec4bd5dfb8d53e6a852 has been merged
[02:10] <funman> cheers :D
[02:10] <funman> millions of clueless windows users thank you
[02:11] <michaelni> :)
[02:12] <durandal_1707> someone with multimedia.cx account here?
[02:18] <Compn> yes
[02:18] <Compn> i got one durandal_1707
[02:20] <durandal_1707> there are couple of game formats listed as not supported in FFmpeg
[02:21] <caytchen> heresy
[02:22] <durandal_1707> want a list?
[02:23] <Compn> you can email mike at multimedia.cx , he does the game formats :)
[03:23] <CIA-49> ffmpeg: 03Michael Niedermayer 07master * r1674073b4d 10ffmpeg/libavutil/ (avutil.h utils.c): 
[03:23] <CIA-49> ffmpeg: libavutil: increase LIBAVUTIL_VERSION_MICRO to 100
[03:23] <CIA-49> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:23] <CIA-49> ffmpeg: 03Michael Niedermayer 07master * r2a85358d4f 10ffmpeg/libavformat/ (utils.c version.h): 
[03:23] <CIA-49> ffmpeg: libavformat: increase LIBAVFORMAT_VERSION_MICRO to 100
[03:23] <CIA-49> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:23] <CIA-49> ffmpeg: 03Michael Niedermayer 07master * rf0e90e0035 10ffmpeg/libavcodec/ (utils.c version.h): 
[03:23] <CIA-49> ffmpeg: lavc: make 100 the base for the micro version number.
[03:23] <CIA-49> ffmpeg: This way ffmpeg can be distinguished from the fork by a user
[03:23] <CIA-49> ffmpeg: application or a encoded file by a decoder.
[03:23] <CIA-49> ffmpeg: The highest value micro had, in the past, that i could find, was 6
[03:23] <CIA-49> ffmpeg: thus 100 should be safe.
[03:23] <CIA-49> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:23] <CIA-49> ffmpeg: 03Michael Niedermayer 07master * rd5874e9c2c 10ffmpeg/libavdevice/ (avdevice.c avdevice.h): 
[03:23] <CIA-49> ffmpeg: libavdevice: increase LIBAVDEVICE_VERSION_MICRO to 100
[03:23] <CIA-49> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:23] <CIA-49> ffmpeg: 03Michael Niedermayer 07master * r55a6d8d9fa 10ffmpeg/libpostproc/ (postprocess.c postprocess.h): 
[03:23] <CIA-49> ffmpeg: libpostproc: increase LIBPOSTPROC_VERSION_MICRO to 100
[03:23] <CIA-49> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:09] <CIA-49> ffmpeg: 03Michael Niedermayer 07master * r9eae8396f1 10ffmpeg/cmdutils.c: 
[04:09] <CIA-49> ffmpeg: cmdutils: restore vertical alignment of versions
[04:09] <CIA-49> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:31] <durandal_1707> ffmpeg fails to build for me
[04:42] <funman> durandal_1707: where does it fial?
[04:43] <CIA-49> ffmpeg: 03Paul B Mahol 07master * ra4209ad631 10ffmpeg/ (6 files in 3 dirs): 
[04:43] <CIA-49> ffmpeg: lavf: Add SMJPEG demuxer.
[04:43] <CIA-49> ffmpeg: Signed-off-by: Paul B Mahol <onemda at gmail.com>
[04:43] <CIA-49> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:44] <durandal_1707> /usr/home/r/git/ffmpeg/cmdutils.c:582: undefined reference to swresample_configuration
[04:45] <durandal_1707> done distclean, will try again
[04:49] <CIA-49> ffmpeg: 03Michael Niedermayer 07master * rab6603b1d8 10ffmpeg/libswresample/libswresample.v: 
[04:49] <CIA-49> ffmpeg: libswresample.v: add swresample* for cmdutils.c
[04:49] <CIA-49> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:50] <michaelni> durandal_1707, retry now
[05:02] <durandal_1707> builds fine now
[05:04] <durandal_1707> would this break abi?
[05:11] <michaelni> not sure i understand you
[05:12] <michaelni> no function was removed or renamed
[05:26] <durandal_1707> heh resurected vivo demuxer just segfaulted on first sample
[05:46] <Compn> vivo demuxer? :)
[05:46] Action: Compn isnt paying attention
[05:47] <Compn> drv said hed post it when he gets time to review it again
[05:48] <durandal_1707> that was in 2009 no?
[05:48] <Compn> that was his reply to my mail from last month 
[05:48] <Compn> er not last month
[05:49] <Compn> august 2011
[05:49] <Compn> guess drv has been busy ;)
[06:02] <durandal_1707> wtf, mplayer now crash every time it finish playing
[10:13] <CIA-49> ffmpeg: 03Stefano Sabatini 07master * r386aee6864 10ffmpeg/libavfilter/ (formats.c internal.h sink_buffer.c): 
[10:13] <CIA-49> ffmpeg: sink_buffer: copy list of provided formats in the context
[10:13] <CIA-49> ffmpeg: A list of formats may have been dynamically created by the calling code,
[10:13] <CIA-49> ffmpeg: and thus should not be referenced by the sink buffer context.
[10:13] <CIA-49> ffmpeg: Avoid possible invalid data reference.
[10:13] <CIA-49> ffmpeg: 03Nicolas George 07master * racb67c5bbc 10ffmpeg/libavdevice/lavfi.c: 
[10:13] <CIA-49> ffmpeg: lavd/lavfi: fix two memleaks.
[10:13] <CIA-49> ffmpeg: Signed-off-by: Nicolas George <nicolas.george at normalesup.org>
[10:13] <CIA-49> ffmpeg: Signed-off-by: Stefano Sabatini <stefasab at gmail.com>
[10:27] <Tjoppen> michaelni: forgot to mention I also pushed the fuzz fixes to github: https://github.com/Tjoppen/FFmpeg/tree/fuzz_fixes
[11:09] <CIA-49> ffmpeg: 03Stefano Sabatini 07master * r834f80db22 10ffmpeg/doc/platform.texi: platform.texi: close @enumerate section in the "Using shared libraries" subsection
[11:55] <Tjoppen> ffmpeg dups one frame on pal dv25 input. that doesn't seem right
[14:32] <CIA-49> ffmpeg: 03Tomas Härdin 07master * re7839602f4 10ffmpeg/libavformat/mxfdec.c: 
[14:32] <CIA-49> ffmpeg: mxfdec: Never seek back in local sets and KLVs
[14:32] <CIA-49> ffmpeg: Specially crafted files can lead the parsing code to take too long.
[14:32] <CIA-49> ffmpeg: We fix a lot of these problems by not allowing local tags to extend past the
[14:32] <CIA-49> ffmpeg: end of the set and not allowing other KLVs to be read past the end of
[14:32] <CIA-49> ffmpeg: themselves.
[14:32] <CIA-49> ffmpeg: 03Tomas Härdin 07master * r2116e4ba91 10ffmpeg/libavformat/mxfdec.c: 
[14:32] <CIA-49> ffmpeg: mxfdec: Check url_feof() in mxf_read_local_tags()
[14:32] <CIA-49> ffmpeg: This fixes the infinite loop with zzuf2.mxf
[14:33] <CIA-49> ffmpeg: 03Tomas Härdin 07master * r5e5e69d078 10ffmpeg/libavformat/mxfdec.c: 
[14:33] <CIA-49> ffmpeg: mxfdec: Check for NULL component
[14:33] <CIA-49> ffmpeg: This fixes SIGSEGV with zzuf1.mxf
[14:33] <CIA-49> ffmpeg: 03Tomas Härdin 07master * r4ae7d45170 10ffmpeg/libavformat/mxfdec.c: 
[14:33] <CIA-49> ffmpeg: mxfdec: Fix infinite loop in mxf_packet_timestamps()
[14:33] <CIA-49> ffmpeg: This can happen if an index table segment has a very large IndexStartPosition.
[14:33] <CIA-49> ffmpeg: zzuf3.mxf is an example of such a file.
[14:33] <CIA-49> ffmpeg: 03Tomas Härdin 07master * r46d65fb8a5 10ffmpeg/libavformat/mxfdec.c: 
[14:33] <CIA-49> ffmpeg: mxfdec: Sanity check PreviousPartition
[14:33] <CIA-49> ffmpeg: Without this certain files could get the demuxer stuck in a loop
[14:33] <CIA-49> ffmpeg: 03Michael Niedermayer 07master * rdd1fb65287 10ffmpeg/: (log message trimmed)
[14:33] <CIA-49> ffmpeg: Merge remote-tracking branch 'tjoppen/fuzz_fixes'
[14:33] <CIA-49> ffmpeg: * tjoppen/fuzz_fixes:
[14:33] <CIA-49> ffmpeg:  mxfdec: Don't crash in mxf_packet_timestamps() if current_edit_unit overflows
[14:33] <CIA-49> ffmpeg:  mxfdec: Zero nb_ptses in mxf_compute_ptses_fake_index()
[14:33] <CIA-49> ffmpeg:  mxfdec: Sanity check PreviousPartition
[14:33] <CIA-49> ffmpeg:  mxfdec: Never seek back in local sets and KLVs
[14:33] <CIA-49> ffmpeg: 03Tomas Härdin 07master * r184f479096 10ffmpeg/libavformat/mxfdec.c: 
[14:33] <CIA-49> ffmpeg: mxfdec: Move the current_partition check inside mxf_read_header()
[14:33] <CIA-49> ffmpeg: This fixes SIGSEGV on files where this is actually the case, such as zzuf4.mxf
[14:33] <CIA-49> ffmpeg: 03Tomas Härdin 07master * rcc1dda851b 10ffmpeg/libavformat/mxfdec.c: mxfdec: Don't crash in mxf_packet_timestamps() if current_edit_unit overflows
[14:33] <CIA-49> ffmpeg: 03Tomas Härdin 07master * r0cd21ddaed 10ffmpeg/libavformat/mxfdec.c: 
[14:33] <CIA-49> ffmpeg: mxfdec: Zero nb_ptses in mxf_compute_ptses_fake_index()
[14:33] <CIA-49> ffmpeg: This fixes SIGSEGV on files with both CBR and VBR index segments (zzuf6.mxf).
[14:40] <OanaStra_> michaelni: are you here ?
[22:29] <cbsrobot> damn I hate these gotos in libopenjepgdec
[00:00] --- Fri Dec 23 2011