[Ffmpeg-devel-irc] ffmpeg-devel.log.20120909

burek burek021 at gmail.com
Mon Sep 10 02:05:02 CEST 2012


[00:02] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * rd40c0e4a67 10ffmpeg/ffserver.c: 
[00:02] <CIA-56> ffmpeg: ffserver: fix order of evaluation bugs
[00:02] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:02] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * rdd081f98dd 10ffmpeg/libavcodec/dct-test.c: 
[00:02] <CIA-56> ffmpeg: dct-test: fix order of evaluation bug
[00:02] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[02:41] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r31a797eb28 10ffmpeg/libswresample/ (rematrix.c swresample.c x86/swresample_x86.c): 
[02:41] <CIA-56> ffmpeg: swr: add av_cold to init/free functions
[02:41] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[02:41] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r59fdf543d5 10ffmpeg/libavformat/smush.c: 
[02:41] <CIA-56> ffmpeg: smush: read subversion for version==1 too.
[02:41] <CIA-56> ffmpeg: This avoids a gcc warning, otherwise it should have no effect.
[02:41] <CIA-56> ffmpeg: its a bit more consistent too
[02:41] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:37] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * rd48ebfbdea 10ffmpeg/libavcodec/mpeg4videodec.c: 
[03:37] <CIA-56> ffmpeg: mpeg4videodec: Fix "warning: dc_pred_dir may be used uninitialized in this function"
[03:37] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:05] <CIA-56> ffmpeg: 03Derek Buitenhuis 07master * re62e455f7b 10ffmpeg/libavcodec/ (10 files in 5 dirs): 
[04:05] <CIA-56> ffmpeg: dsputil/me/pixels: Actually use av_restrict
[04:05] <CIA-56> ffmpeg: Signed-off-by: Derek Buitenhuis <derek.buitenhuis at gmail.com>
[04:05] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[11:31] <CIA-56> ffmpeg: 03Stefano Sabatini 07master * ra4cb0af945 10ffmpeg/libavfilter/ (avfilter.h version.h vf_boxblur.c video.c): 
[11:31] <CIA-56> ffmpeg: lavfi: drop deprecated and pointless avfilter_default_end_frame() function
[11:31] <CIA-56> ffmpeg: Fix warning when compiling boxblur.
[11:31] <CIA-56> ffmpeg: While this is technically a major API break, practically there will be no
[11:31] <CIA-56> ffmpeg: one using that function since the filtering API is mostly private, so
[11:31] <CIA-56> ffmpeg: that function alone is not usable.
[11:31] <CIA-56> ffmpeg: 03Stefano Sabatini 07master * r3634441452 10ffmpeg/libavutil/samplefmt.h: lavu/samplefmt: add note regarding buf validity in av_samples_fill_arrays() doxy
[11:42] <CIA-56> ffmpeg: 03Stefano Sabatini 07master * reab438eabb 10ffmpeg/libavutil/samplefmt.h: 
[11:42] <CIA-56> ffmpeg: lavu/samplefmt: fix/clarify documentation for av_samples_fill_arrays()
[11:42] <CIA-56> ffmpeg: In particular, avoid improper use of the term "channels" in place of
[11:42] <CIA-56> ffmpeg: "planes", and fix erroneous documentation for the linesize parameter.
[12:40] <ubitux> nyuhu: we have AV_STRINGIFY
[12:41] <ubitux> instead of redefining it like you did
[12:41] <ubitux> (in the hue patch)
[13:07] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r363d302ea3 10ffmpeg/libavcodec/imc.c: 
[13:07] <CIA-56> ffmpeg: imc: use exp2 instead of pow(2,...)
[13:07] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[13:07] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * rb027156b19 10ffmpeg/ffprobe.c: ffprobe: use exp2 instead of pow(2, ...)
[13:07] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r4819d43d7f 10ffmpeg/libavcodec/aacpsy.c: 
[13:07] <CIA-56> ffmpeg: aacpsy: use exp2(f) instead of pow(f)(2,...)
[13:07] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[13:30] <CIA-56> ffmpeg: 03Hendrik Leppkes 07master * rfb4e983e0c 10ffmpeg/libavcodec/x86/mlpdsp.c: 
[13:30] <CIA-56> ffmpeg: x86: mlpdsp: mlp_filter_channel_x86 requires inline asm
[13:30] <CIA-56> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[13:31] <CIA-56> ffmpeg: 03Diego Biurrun 07master * re0c6cce447 10ffmpeg/ (21 files in 5 dirs): 
[13:31] <CIA-56> ffmpeg: x86: Replace checks for CPU extensions and flags by convenience macros
[13:31] <CIA-56> ffmpeg: This separates code relying on inline from that relying on external
[13:31] <CIA-56> ffmpeg: assembly and fixes instances where the coalesced check was incorrect.
[13:31] <CIA-56> ffmpeg: 03Mans Rullgard 07master * r6a0200f24d 10ffmpeg/configure: 
[13:31] <CIA-56> ffmpeg: configure: msvc: fix/simplify setting of flags for hostcc
[13:31] <CIA-56> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[13:31] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r77aedc77ab 10ffmpeg/: (log message trimmed)
[13:31] <CIA-56> ffmpeg: Merge remote-tracking branch 'qatar/master'
[13:31] <CIA-56> ffmpeg: * qatar/master:
[13:31] <CIA-56> ffmpeg:  swscale: Provide the right alignment for external mmx asm
[13:31] <CIA-56> ffmpeg:  x86: Replace checks for CPU extensions and flags by convenience macros
[13:31] <CIA-56> ffmpeg:  configure: msvc: fix/simplify setting of flags for hostcc
[13:31] <CIA-56> ffmpeg:  x86: mlpdsp: mlp_filter_channel_x86 requires inline asm
[13:31] <CIA-56> ffmpeg: 03Martin Storsjö 07master * r75c37c5ace 10ffmpeg/libswscale/utils.c: 
[13:31] <CIA-56> ffmpeg: swscale: Provide the right alignment for external mmx asm
[13:31] <CIA-56> ffmpeg: This reverts parts of e0c6cce4472. There is external mmx asm that
[13:31] <CIA-56> ffmpeg: requires this alignment.
[13:31] <CIA-56> ffmpeg: This fixes crashes when using swscale in builds with external mmx,
[13:31] <CIA-56> ffmpeg: without inline assembly.
[13:31] <CIA-56> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[14:23] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r9c6af3a3db 10ffmpeg/ffserver.c: 
[14:23] <CIA-56> ffmpeg: ffserver: increase buffer size in http_send_too_busy_reply()
[14:23] <CIA-56> ffmpeg: also add a assert0 to make sure the data is not truncated (which
[14:23] <CIA-56> ffmpeg: should be impossible now)
[14:23] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[14:23] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * rd7cb5a8daf 10ffmpeg/ffmpeg_filter.c: 
[14:23] <CIA-56> ffmpeg: ffmpeg_filter: fix unsafe snprintf() usage
[14:23] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[14:23] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r9fea619f19 10ffmpeg/libavcodec/ass.c: 
[14:23] <CIA-56> ffmpeg: ass: assert that the timsstamps fitted in the buffers
[14:23] <CIA-56> ffmpeg: Failure of the assert would cause various problems later
[14:23] <CIA-56> ffmpeg: if we continue.
[14:23] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[14:23] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * rb9a07e787b 10ffmpeg/libavcodec/srtdec.c: 
[14:48] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r7992814920 10ffmpeg/ffprobe.c: 
[14:48] <CIA-56> ffmpeg: ffprobe: avoid unsafe usage of the snprintf() return code
[14:48] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[14:48] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r07108cafc6 10ffmpeg/libavformat/gxfenc.c: 
[14:48] <CIA-56> ffmpeg: gxfenc: assert that the to be written data has not been truncated.
[14:48] <CIA-56> ffmpeg: Otherwise out of array reads would happen later
[14:48] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[14:48] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r17f9626b5d 10ffmpeg/libavcodec/tiff.c: 
[14:48] <CIA-56> ffmpeg: tiff/doubles2str: check for truncation
[14:48] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[14:48] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r0196bc6f14 10ffmpeg/libavcodec/tiff.c: 
[14:48] <CIA-56> ffmpeg: tiff/doubles2str: factor out component_len
[14:48] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[14:48] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * reb336cb881 10ffmpeg/libavformat/gxfenc.c: 
[14:48] <CIA-56> ffmpeg: gxfenc: dont duplicate buffer size
[14:48] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[15:03] <Daemon404> that fate bot is kinda useful
[15:03] <Daemon404> (still dont care for ticket spam tho)
[15:06] <Compn> sounds like someone doesnt visit the bug tracker :P
[17:27] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r6add8eb2ce 10ffmpeg/libavcodec/x86/h264dsp_init.c: 
[17:27] <CIA-56> ffmpeg: x86/h264dsp_init: put a HAVE_YASM back
[17:27] <CIA-56> ffmpeg: Should fix compilation on open solaris
[17:27] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[17:28] <CIA-56> ffmpeg: 03Derek Buitenhuis 07master * r3629d1c09d 10ffmpeg/libavcodec/dsputil_template.c: 
[17:28] <CIA-56> ffmpeg: dsputil_template: Reindent after recent change
[17:28] <CIA-56> ffmpeg: Forgot to keep it pretty in e62e455f7b4c5b148f5a289ba823dcc71d6ca389.
[17:28] <CIA-56> ffmpeg: Signed-off-by: Derek Buitenhuis <derek.buitenhuis at gmail.com>
[17:28] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:39] <CIA-56> ffmpeg: 03Michael Niedermayer 07master * r44f83930d9 10ffmpeg/libavcodec/x86/mpegaudiodec.c: 
[18:39] <CIA-56> ffmpeg: x86/mpegaudiodec: fix compilation failure on cygwin with undefined imdct36_blocks_avx
[18:39] <CIA-56> ffmpeg: this is a regression since today
[18:39] <CIA-56> ffmpeg: Found-by: beastd
[18:39] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[19:18] <burek> there might be a NOTE in ffmpeg's html documentation for -codec (or -c) option that says "-c copy" doesn't copy all the streams (as it might be expected by users) but it rather chooses the "best" audio and video stream and copies only that. i.e. -map 0 needs to be added to have a "true" copying of all streams
[19:19] <burek> that's not quite obvious from the documentation and is a common mistake when people use -c copy
[20:02] <ubitux> copy doesn't choose anything
[20:02] <ubitux> i mean the -codec option doesn't do any selection
[20:02] <ubitux> of streams
[21:47] <CIA-56> ffmpeg: 03Derek Buitenhuis 07master * r16c03f2e91 10ffmpeg/libavformat/file.c: 
[21:47] <CIA-56> ffmpeg: file: Add S_ISFIFO compatability macro
[21:47] <CIA-56> ffmpeg: Not all systems have S_ISFIFO.
[21:47] <CIA-56> ffmpeg: Signed-off-by: Derek Buitenhuis <derek.buitenhuis at gmail.com>
[21:47] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[22:01] <ubitux> beastd: heh so you finally sent that patch ;)
[22:02] <ubitux> beastd: btw, were you able to get in touch with the texi2html guy?
[22:30] <beastd> ubitux: hey! yes thanks for listening to my weird thoughts back then
[22:31] <beastd> ubitux: i will try to contact that texi2html guy next
[22:32] <beastd> VDD disrupted my flow, so I did not get too much after work besides being tired
[22:33] <ubitux> spice must flow!
[22:54] <CIA-56> ffmpeg: 03Reinhard Nissl 07master * r2474ca1a22 10ffmpeg/libavcodec/aacdec.c: 
[22:54] <CIA-56> ffmpeg: aacdec.c: fix some comments
[22:54] <CIA-56> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[23:13] <burek> ubitux, if "-c:v copy" select all video streams and "-c:a copy" selects all audio streams, it might be expected that "-c copy" would select all streams, right? but it doesn't
[23:14] <burek> that's why I think it might be a good idea to update the html docs to highlight that thing
[23:16] <ubitux> -c:v copy doesn't select all the video streams
[23:17] <ubitux> it just marks any video stream selected as "copy"
[23:17] <ubitux> it doesn't affect the stream matching
[23:17] <burek> so if you have input with 2 videos and you type: ffmpeg -i input.mp4 -c:v copy output.mp4
[23:17] <burek> it will not copy all video streams?
[23:18] <ubitux> i don't think so.
[23:18] <relaxed> No because ffmpeg be default doesn't select all streams.
[23:18] <relaxed> s/be/by/
[23:22] <burek> oh.. :/
[23:22] <relaxed> If the default was to select all the streams it would force people to learn -map.
[23:26] <burek> that means that all the tutorials/logs where we said that "ffmpeg -i input -c copy output" would simply remux the input, were actually wrong/incorrect
[23:26] <burek> because it cuts all the other audio/video streams, except for chosen ones
[23:26] <burek> right?
[23:27] <burek> well.. at least there is something to do in the next couple of days.. :)
[23:28] <beastd> ubitux: texinfo mail sent. thanks for bugging me ;)
[23:28] <ubitux> np :p
[23:33] Action: beastd goes to bed
[23:33] <beastd> bye...
[00:00] --- Mon Sep 10 2012


More information about the Ffmpeg-devel-irc mailing list