[Ffmpeg-devel-irc] ffmpeg-devel.log.20150312

burek burek021 at gmail.com
Fri Mar 13 02:05:02 CET 2015


[00:05:47 CET] <debianuser> it's 0RGB (x11grab) vs ARGB (bmp), but the difference is 5%CPU (x11grab at 60fps) vs 100%CPU (bmp at 60fps) :)
[00:31:09 CET] <cone-005> ffmpeg 03Andreas Cadhalpun 07master:72211a2af047: asfenc: fix leaking asf->index_ptr on error
[00:31:10 CET] <cone-005> ffmpeg 03Michael Niedermayer 07master:a277c545b2b5: Merge commit '72211a2af0470799c9611b9c1c7039ab14a34a0a'
[00:38:35 CET] <cone-005> ffmpeg 03Andreas Cadhalpun 07master:67142a8c721c: roqvideoenc: set enc->avctx in roq_encode_init
[00:38:36 CET] <cone-005> ffmpeg 03Michael Niedermayer 07master:3bedc99723a3: Merge commit '67142a8c721c7916c9ad2eb439c14d567aeb88c1'
[00:52:56 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:0701f7c5c3f9: dnxhdenc: fix mb_rc size
[00:52:57 CET] <cone-005> ffmpeg 03Dale Curtis 07release/0.7:c3ece52decaf: matroska: Fix use after free
[00:52:58 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:1b291e046630: avutil/lzo: Fix integer overflow
[00:52:59 CET] <cone-005> ffmpeg 03Reimar Döffinger 07release/0.7:3ab6c0c4ecfd: configure: add noexecstack to linker options if supported.
[00:53:00 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:794fcabfe8b8: avcodec/jpeglsdec: Check run value more completely in ls_decode_line()
[00:53:01 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:b5e661bcd2bb: avcodec/mjpegdec: check bits per pixel for changes similar to dimensions
[00:53:02 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:cd3c4d8c5522: avcodec/utils: Add case for jv to avcodec_align_dimensions2()
[00:53:03 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:73962e677d87: avcodec/mmvideo: Bounds check 2nd line of HHV Intra blocks
[00:53:04 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:7a5590ef4282: avcodec/pngdec: Check bits per pixel before setting monoblack pixel format
[00:53:05 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:fd36238c402d: avcodec/pngdec: Calculate MPNG bytewidth more defensively
[00:53:06 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:ef32bc8dde52: avcodec/gifdec: factorize interleave end handling out
[00:53:07 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:faf293a83af1: avcodec/qpeg: fix off by 1 error in MV bounds check
[00:53:08 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:5b2097626d0e: avcodec/smc: fix off by 1 error
[00:53:09 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:79891f5b87ba: avformat/mpegts: Check desc_len / get8() return code
[00:53:10 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:f3cdeeb070d4: avcodec/dxa: check dimensions
[00:53:11 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:f78f7eca1a5c: avcodec/wmaprodec: Fix integer overflow in sfb_offsets initialization
[00:53:12 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:30e8a375901f: avcodec/mjpegdec: Fix context fields becoming inconsistent
[00:53:13 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:cb1db92cca98: avcodec/pngdec: Check IHDR/IDAT order
[00:53:14 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:9f993d2d4ba2: swscale/x86/rgb2rgb_template: fix crash with tiny size and nv12 output
[00:53:15 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:1aedd7645a3d: avformat/aviobuf: Check that avio_seek() target is non negative
[00:53:16 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:fac6f744d817: avcodec/vmdvideo: Check len before using it in method 3
[00:53:17 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:ae82abbbcf0c: avcodec/flac_parser: fix handling EOF if no headers are found
[00:53:18 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:42f8a33c3c39: avformat/utils: Fix number suffixes in tb_unreliable()
[00:53:19 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:127a2902533c: smackerdemuxer: check some values before instead of just after malloc() Fixes Ticket777 Bug Found by: Diana Elena Muscalu
[00:53:20 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:8cf62b34e428: avformat/smacker: Fix number suffix
[00:53:21 CET] <cone-005> ffmpeg 03wm4 07release/0.7:42b4ba4a8ae2: avformat/mpc8: fix hang with fuzzed file
[00:53:22 CET] <cone-005> ffmpeg 03wm4 07release/0.7:a58b8e99c421: avformat/mpc8: fix broken pointer math
[00:53:23 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:c68fe8ac8b9d: avformat/mpc8: Use uint64_t in *_get_v() to avoid undefined behavior
[00:53:24 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:75b0cfcf105c: avcodec/mjpegdec: Check number of components for JPEG-LS
[00:53:25 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:f3a776a9433c: avformat/thp: Check av_get_packet() for failure not only for partial output
[00:53:26 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:3eb6eeaab0cd: avcodec/h264_ps: More completely check the bit depths
[00:53:27 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:2b8c9c1f7de8: avcodec/mjpegdec: Skip blocks which are outside the visible area
[00:53:28 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:286915cb846f: avformat/gxf: Use 64bit for res to avoid overflow
[00:53:29 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:20446996f367: avformat/vqf: Use 64bit for ret to avoid overflow
[00:53:30 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:2ddabbb40f0c: avcodec/x86/mlpdsp_init: Simplify mlp_filter_channel_x86()
[00:53:31 CET] <cone-005> ffmpeg 03Andreas Cadhalpun 07release/0.7:1979bfecc780: avcodec/a64multienc: fix use of uninitialized values in to_meta_with_crop
[00:53:32 CET] <cone-005> ffmpeg 03Dyami Caliri 07release/0.7:c681f86a324d: Fix buffer_size argument to init_put_bits() in multiple encoders.
[00:53:33 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:bd92af66a1f0: avcodec/utils: Align YUV411 by as much as the other YUV variants
[00:53:34 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:5333edcc8691: avcodec/tiff: move bpp check to after "end:"
[00:53:35 CET] <cone-005> ffmpeg 03Andreas Cadhalpun 07release/0.7:a83b9f363c24: roqvideoenc: set enc->avctx in roq_encode_init
[00:53:36 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:98b0751e895e: h264/cabac: check loop index
[00:56:14 CET] <cone-005> ffmpeg 03Michael Niedermayer 07release/0.7:6f4cda4a6b66: Update for 0.7.17
[01:22:26 CET] <cone-005> ffmpeg 03Michael Niedermayer 07n0.7.17:HEAD: Merge commit '67142a8c721c7916c9ad2eb439c14d567aeb88c1'
[03:06:15 CET] <cone-005> ffmpeg 03Michael Niedermayer 07master:35fad1e9c90b: avutil/buffer: Avoid moving the AVBufferRef to a new place in memory in av_buffer_realloc()
[03:06:16 CET] <cone-005> ffmpeg 03Michael Niedermayer 07master:26d81b5703ea: avutil/buffer: Avoid moving the AVBufferRef to a new place in memory in av_buffer_make_writable()
[04:28:28 CET] <cone-005> ffmpeg 03Yayoi 07master:d5232d4717ec: avfilter/colormatrix:add slice threading
[09:55:40 CET] <user1111> Hello
[09:56:11 CET] <Guest16402> Hello
[09:56:24 CET] <Guest16402> I get error when encoding
[11:09:53 CET] <ninten> fflogger, ping
[11:13:00 CET] <ninten> anyone knows irc use name of thilo borgmann
[11:37:24 CET] <cbsrobot> ninten: I thinks he's not on irc - send him an email
[11:50:08 CET] <gb27> Are we supposed to consider XDR format while doing the B44 compression.How often we deal with the files in XDR format?
[12:20:30 CET] <Compn> gb27 : which xdr format and b44 compression ?
[12:20:31 CET] <Compn> ehe
[12:20:42 CET] <Compn> oh openexr
[12:20:46 CET] <Compn> i dont know.
[12:34:11 CET] <cone-219> ffmpeg 03Diego Biurrun 07master:9c8074050745: configure: Mark qsv subsystem as not selectable on the command line
[12:34:12 CET] <cone-219> ffmpeg 03Michael Niedermayer 07master:7d63f10c37cd: Merge commit '9c8074050745184a61059ad56045ad711299e33d'
[12:47:46 CET] <cone-219> ffmpeg 03Diego Biurrun 07master:b97f427fb56b: lavf: Explicitly convert types at function pointer assignment
[12:47:47 CET] <cone-219> ffmpeg 03Michael Niedermayer 07master:22af79a9c88f: Merge commit 'b97f427fb56bdbf7de8e2d4b72d01114aa6eceda'
[12:50:21 CET] <Daemon404> perhaps that mail was a bit too angry.
[12:50:26 CET] <Daemon404> but i stand by my assertions.
[13:06:49 CET] <Compn> ah you guys, always arguing :P
[13:07:16 CET] <Compn> except you both agree, you just cant see it
[13:07:37 CET] <Daemon404> angry children on the internet is hardly novel
[13:07:47 CET] <Daemon404> now where's my diaper
[13:20:32 CET] <Daemon404> Compn, dual decoders *and* dual encoders
[13:20:36 CET] <Daemon404> double the insanity
[13:20:37 CET] <Daemon404> (prores)
[13:23:16 CET] <Compn> o i forgot encoder :P
[13:23:17 CET] <Compn> ehe
[13:23:34 CET] <Compn> but prores is where no one wanted to work together
[13:23:57 CET] <Compn> so instead of arguing with michael, you should encourage patch author to work with other patch author...
[13:24:07 CET] <Daemon404> privately, i have
[13:24:11 CET] <Compn> fair enough.
[13:24:15 CET] <Compn> :)
[13:38:15 CET] <wm4> so how do I select the video codec when using ffprobe?
[13:38:23 CET] <wm4> ffprobe -h is completely useless
[13:39:27 CET] <Daemon404> -c:v ?
[13:39:29 CET] <Daemon404> i dunno
[13:39:45 CET] <Daemon404> isnt the point to, well, probe
[13:40:15 CET] <wm4> it's also useful to dump properties of decoded frames
[13:40:30 CET] <wm4> "Failed to set value 'h264' for option 'c:v': Option not found"
[13:41:33 CET] <Daemon404> i guess you have a case where it fails to find the format:
[13:41:33 CET] <Daemon404> ?
[13:41:53 CET] <wm4> no
[13:42:15 CET] <Daemon404> then why do you even need to specify the codec
[13:42:29 CET] <wm4> without the option it works; but I don't want to select the default decoder for the given video codec
[13:42:46 CET] <Daemon404> ah.... 
[13:42:51 CET] <Daemon404> probably unsupported :D
[13:42:57 CET] <wm4> bleh
[13:43:04 CET] <wm4> the ffprobe docs claim it's supported
[13:43:24 CET] <wm4> (but the suggested options don't work either)
[13:43:58 CET] <Daemon404> lul
[13:44:59 CET] <wm4> it's probably spliced in "generic" documentation for all command line tools
[13:56:59 CET] Action: Daemon404 sends out more trolls to non-ffmpeg MLs
[13:58:34 CET] <Compn> ffprobe -f h264 file ?
[13:58:35 CET] <Compn> :P
[13:58:50 CET] <Daemon404> no that sets it to the h264 elementary stream format
[13:58:55 CET] <Daemon404> not codec
[13:59:08 CET] <Compn> correct
[13:59:11 CET] <Compn> but it does work ;P
[13:59:16 CET] <Compn> unlike -codec :P
[13:59:18 CET] <Compn> :PPPPPPPPPPPPPp
[13:59:49 CET] <Compn> ffprobe bug
[14:27:00 CET] <nevcairiel> Daemon404 is on fire today
[14:28:01 CET] <nevcairiel> also, I should totally take the dca patch from libav and rebase it to ffmpeg to make that discussion shut up =P
[14:28:10 CET] <wm4> more like he sets everything on fire
[14:28:23 CET] <Daemon404> i repeat
[14:28:30 CET] <Daemon404> [12:06] <@Daemon404> angry children on the internet is hardly novel
[14:28:30 CET] <Daemon404> [12:07] <@Daemon404> now where's my diaper
[14:29:10 CET] <nevcairiel> hm the mail with the latest patch is base64 encoded
[14:29:13 CET] <nevcairiel> i wonder if git-am likes that
[14:29:16 CET] <nevcairiel> i guess it even might
[14:43:38 CET] <Daemon404> [13:26] <+nevcairiel> Daemon404 is on fire today <-- result so far is lots of people angry at me
[14:43:44 CET] <Daemon404> and im just sitting here, eating pizza.
[14:46:36 CET] <wm4> pizza is unhalthy
[14:46:41 CET] <wm4> unhealthy even
[14:47:09 CET] <Daemon404> thats why im washing it down with coffee
[14:47:11 CET] <Daemon404> to be extra healthy
[14:52:02 CET] <ubitux> btw, someone might want to tell luca there is also a tutorial on writing filters in ffmpeg
[14:52:10 CET] <ubitux> if he wants to save more time
[14:52:24 CET] <ubitux> he just has to strip a few sections for the stuff that's not present in libav
[14:53:16 CET] <Daemon404> ubitux, link
[14:53:43 CET] <ubitux> http://git.videolan.org/?p=ffmpeg.git;a=blob_plain;f=doc/writing_filters.txt;hb=HEAD
[14:56:40 CET] <ubitux> (just mentioning that since he's saying there is no documentation, which is not quite true)
[14:56:50 CET] <ubitux> anyway, just my 2cts
[14:58:41 CET] <cone-219> ffmpeg 03Rainer Hochecker 07master:31816eae3201: hevc: delay ff_thread_finish_setup for hwaccel
[15:25:10 CET] <cone-219> ffmpeg 03Michael Niedermayer 07master:b7cb8b3d436f: avcodec/h264_mb: Fix undefined shifts
[15:53:04 CET] <cone-219> ffmpeg 03Michael Niedermayer 07master:27191b82de0b: avcodec/vp9: Fix undefined shifts in decode_frame_header()
[16:12:26 CET] <ninten> anyone know at what tym durandal_1707 comes online generally !!
[16:12:57 CET] <ninten> btw durandal_1707 is Paul B mahol
[16:13:01 CET] <Daemon404> not sure... he is in CET
[16:13:08 CET] <Daemon404> you may email him too
[16:13:23 CET] <ninten> thanks !
[16:17:47 CET] <ganeshadi1> Hey what happened to the bayes colorspace project? I swear I read that in the idea's page yesterday!
[16:21:30 CET] <ganeshadi1> Anyone?
[16:24:42 CET] <ganeshadi1> Is no one online!
[16:28:17 CET] <wm4> no
[16:28:57 CET] <ganeshadi1> Hey what happened to the baye color space to RGB project?
[16:29:34 CET] <ganeshadi1> @+wm4?
[16:29:54 CET] <wm4> how should I know
[16:30:11 CET] <wm4> we have bayes colorspace pixelformats, but AFAIK nothing uses it
[16:30:36 CET] <ganeshadi1> There was GSOC project to add the encoder to rebase patches related to it!
[16:30:44 CET] <wm4> except cinedec.c and mlvdec.c
[16:30:49 CET] <ganeshadi1> I saw it yesterday.. 
[17:21:23 CET] <nevcairiel> maybe you looked at a page from another year? the basic bayer stuff was implemented already, so its not a task for this year
[17:21:50 CET] <nevcairiel> we keep the pages from previous years for documentaiton purposes
[17:27:46 CET] <ganeshadi1> @-nevcairiel yes it was from last year .. my bad...
[17:30:09 CET] <ubitux> but if you're interested in bayer, maybe you could add some support for some raw photo formats
[17:31:00 CET] <ubitux> like, i wonder if DNG uses such pixel format
[17:31:54 CET] <ubitux> some samples @ http://kenrockwell.com/leica/m9/sample-photos-3.htm
[17:31:57 CET] <ubitux> maybe i should open a ticket..
[17:33:01 CET] <ubitux> this is relevant because some alternative firmwares of canon typically outputs this
[17:33:04 CET] <ubitux> and can record videos
[17:33:16 CET] <ubitux> so you get interesting raw videos...
[17:33:54 CET] <ubitux> i haven't been able to fix the firmware for my compact thought so i can't provide a real case sample so far
[17:40:43 CET] <cone-219> ffmpeg 03James Cowgill 07master:a251aa1a350a: mips/asmdefs: use _ABI64 as defined by gcc
[17:41:04 CET] <ubitux> here we go..
[17:41:17 CET] <Daemon404> dont you mean DNG
[17:42:26 CET] <ubitux> oups
[17:42:45 CET] <ubitux> thx
[17:43:06 CET] <Daemon404> i wouldnt be surprised if vlc had "dmg" support
[17:43:09 CET] <Daemon404> i mean, it has rar support.
[17:43:22 CET] <Daemon404> (or was taht only mpc?)
[17:43:35 CET] <wm4> no vlc also has it
[17:43:42 CET] <Daemon404> o ok
[17:44:00 CET] <wm4> and I stole it from them too
[18:05:45 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:7248e735599b: avcodec: fix aac/ac3 parser bitstream buffer size
[18:05:46 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:ced4e9fdbba7: avformat/m4vdec: Check for non startcode 00 00 00 sequences in probe
[18:05:47 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:348b87b9bd51: avcodec/ac3enc_template: fix out of array read
[18:05:48 CET] <cone-219> ffmpeg 03Reimar Döffinger 07release/0.10:ef803afa769b: configure: add noexecstack to linker options if supported.
[18:05:49 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:1704a7b67dce: avcodec/jpeglsdec: Check run value more completely in ls_decode_line()
[18:05:50 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:bb6a8a0509d5: avcodec/mjpegdec: check bits per pixel for changes similar to dimensions
[18:05:51 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:bd3a28e8b65c: avcodec/utils: Add case for jv to avcodec_align_dimensions2()
[18:05:52 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:19d0c9e9934b: avcodec/mmvideo: Bounds check 2nd line of HHV Intra blocks
[18:05:53 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:7c1150bf05cd: avcodec/tiff: more completely check bpp/bppcount
[18:05:54 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:7238c744de42: avcodec/pngdec: Check bits per pixel before setting monoblack pixel format
[18:05:55 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:677da723005e: avcodec/pngdec: Calculate MPNG bytewidth more defensively
[18:05:56 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:e0ed766f2a4d: avcodec/qpeg: fix off by 1 error in MV bounds check
[18:05:57 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:27d82d83741b: avformat/mpegts: Check desc_len / get8() return code
[18:05:58 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:017226fdf9eb: avcodec/dxa: check dimensions
[18:05:59 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:d327f673f934: avcodec/dnxhddec: treat pix_fmt like width/height
[18:06:00 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:3296e30d372d: avcodec/h264_slice: Clear table pointers to avoid stale pointers
[18:06:01 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:ec640e10b249: avcodec/wmaprodec: Fix integer overflow in sfb_offsets initialization
[18:06:02 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:aebfcf7d6258: avcodec/mjpegdec: Fix context fields becoming inconsistent
[18:06:03 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:0b5d644839c5: avcodec/pngdec: Check IHDR/IDAT order
[18:06:04 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:a5c43d7c87b5: avformat/rmdec: Check codec_data_size
[18:06:05 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:5a1061b4ad41: swscale/x86/rgb2rgb_template: fix crash with tiny size and nv12 output
[18:06:06 CET] <cone-219> ffmpeg 03wm4 07release/0.10:39a6977354de: avformat/matroskadec: fix handling of recursive SeekHead elements
[18:06:07 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:adf2f2166e7e: avformat/aviobuf: Check that avio_seek() target is non negative
[18:06:08 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:87ec3c615689: avcodec/vmdvideo: Check len before using it in method 3
[18:06:09 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:1d1cc267e6e7: swscale: increase yuv2rgb table headroom
[18:06:10 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:cd4827dfd43a: avcodec/indeo3: use signed variables to avoid underflow
[18:06:11 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:87a716991d84: avcodec/indeo3: ensure offsets are non negative
[18:06:12 CET] <cone-219> ffmpeg 03wm4 07release/0.10:ce219702c346: avcodec/dvdsubdec: fix out of bounds accesses
[18:06:13 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:766c1cbeb4a1: avformat/mov: fix integer overflow in mov_read_udta_string()
[18:06:14 CET] <cone-219> ffmpeg 03Johan Andersson 07release/0.10:b3242c0f2293: cmdutils: update copyright year to 2015.
[18:06:15 CET] <cone-219> ffmpeg 03wm4 07release/0.10:00cde0cddc56: avcodec/dvdsubdec: error on bitmaps with size 0
[18:06:16 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:c78ed1fb5d33: avcodec/flac_parser: fix handling EOF if no headers are found
[18:06:17 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:c00fd60ab345: avformat/utils: Fix number suffixes in tb_unreliable()
[18:06:18 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:1d73ad25dcb2: avformat/smacker: Fix number suffix
[18:06:19 CET] <cone-219> ffmpeg 03wm4 07release/0.10:376533e0cd86: avformat/mpc8: fix hang with fuzzed file
[18:06:20 CET] <cone-219> ffmpeg 03wm4 07release/0.10:e83c03aaf147: avformat/mpc8: fix broken pointer math
[18:06:21 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:d0a925ba2e7c: avformat/mpc8: Use uint64_t in *_get_v() to avoid undefined behavior
[18:06:22 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:5c0413aa85f7: avcodec/mjpegdec: Check escape sequence validity
[18:06:23 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:a16669f1e226: avcodec/mjpegdec: Check number of components for JPEG-LS
[18:06:24 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:3b90117a1f9a: avformat/thp: Check av_get_packet() for failure not only for partial output
[18:06:25 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:20d95a8ad32b: avcodec/h264_ps: More completely check the bit depths
[18:06:26 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:4640184c76f2: avcodec/mjpegdec: Skip blocks which are outside the visible area
[18:06:27 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:3d023f4343d6: avformat/gxf: Use 64bit for res to avoid overflow
[18:06:28 CET] <cone-219> ffmpeg 03Andreas Cadhalpun 07release/0.10:7232fc75e140: avcodec/a64multienc: fix use of uninitialized values in to_meta_with_crop
[18:06:29 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:07a9a43d5c1f: avcodec/snowdec: Fix ref value check
[18:06:30 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:0c1f8a784db9: avcodec/zmbv: Check len before reading in decode_frame()
[18:06:31 CET] <cone-219> ffmpeg 03Dyami Caliri 07release/0.10:c89645c3ef4b: Fix buffer_size argument to init_put_bits() in multiple encoders.
[18:06:32 CET] <cone-219> ffmpeg 03Andreas Cadhalpun 07release/0.10:f92e8cccf581: avformat/adxdec: check avctx->channels for invalid values
[18:06:33 CET] <cone-219> ffmpeg 03Andreas Cadhalpun 07release/0.10:6739df26a620: avformat/bit: check that pkt->size is 10 in write_packet
[18:06:34 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:7f99fae1ec29: avcodec/utils: Align YUV411 by as much as the other YUV variants
[18:06:35 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:25e2ef355d8a: avcodec/tiff: move bpp check to after "end:"
[18:06:36 CET] <cone-219> ffmpeg 03Andreas Cadhalpun 07release/0.10:fcabfc5590fe: roqvideoenc: set enc->avctx in roq_encode_init
[18:20:31 CET] <cone-219> ffmpeg 03Michael Niedermayer 07release/0.10:d6541bdbd7e2: Update for 0.10.16
[18:39:56 CET] <ameyamarathe18> Hi! This is Ameya from Pune, India. I would like to work on FFMpeg as my GSoC 2015 project.
[18:40:31 CET] <ameyamarathe18> I have started looking into the FFMpeg wiki.
[18:48:48 CET] <Compn> doh 
[18:48:51 CET] <Compn> was going to reply
[18:48:52 CET] <Compn> toolat
[18:51:49 CET] <Compn> we do want rar support though
[18:51:51 CET] <Compn> rar is good
[18:51:58 CET] <Compn> not good, but ... it exists.
[18:52:44 CET] <Compn> wm4 / ganeshadi1 : the bayer stuff is used in r3dcode camera i think. dunno its been a while.
[19:08:08 CET] <jamrial> commit 1d523ea8 broke aarch64
[19:09:07 CET] <jamrial> the original check in hevcdsp.c was ok. what needs changing is arm/Makefile
[19:36:02 CET] <jamrial> patch sent. a proper fix needed a bit more than just makefile
[19:52:54 CET] <Zeranoe> Postproc is now intigrated into FFmpeg, correct? Which would explain why I didn't get a .dll file for that lib?
[19:53:32 CET] <Zeranoe> Or actually, it must be GPL only
[19:53:52 CET] Action: michaelni just wanted to say --enable-gpl :)
[19:58:20 CET] <michaelni> jamrial, ffmpeg/libavcodec/hevcdsp.c:262:9: error: HAVE_ARM undeclared (first use in this function)
[19:58:57 CET] <jamrial> ah, should be ARCH_ARM
[19:59:02 CET] <jamrial> my bad
[19:59:53 CET] <jamrial> should i send a new patch?
[20:02:37 CET] <michaelni> if thats the only change needed, no
[20:04:30 CET] <Daemon404> deciding whether to shut down my MSVC static analysis FATE at Libav, or move it here
[20:04:37 CET] <Daemon404> would anyone actually check it?
[20:04:46 CET] <Daemon404> http://fate.libav.org/x86_64-windows-msvc11-appverifier/20150312153413/compile <-- output like this
[20:07:16 CET] <michaelni> i would probably look at it when i have time
[20:07:50 CET] <Daemon404> its currently running daily on weekdays, i'd probably make it run much more infrequently
[20:07:56 CET] <Daemon404> theyre very slow builds.
[20:08:26 CET] <michaelni> sure
[20:16:39 CET] <Daemon404> i see there is an http/2 project
[20:17:13 CET] Action: Daemon404 could backup mentor
[20:27:29 CET] <wm4> http2 sounds useful and probably not too hard
[20:27:46 CET] <wm4> though I wonder if it's really needed for a project like ffmpeg
[20:27:59 CET] <Daemon404> nothing media related uses it yet
[20:28:08 CET] <Daemon404> youtube might server DASH over it, not sure
[20:28:21 CET] <cone-219> ffmpeg 03James Almer 07master:d5addf155502: hevcdsp: fix compilation for arm and aarch64
[20:28:22 CET] <Daemon404> http2 is not hard, but annoying (hpack)
[20:28:31 CET] <Daemon404> i think its one of the more useful ideas.
[20:28:39 CET] <Daemon404> (not lolAPNG)
[20:28:44 CET] <wm4> definitely
[20:29:00 CET] <wm4> I'm just saying that http2 is no win for multimedia
[20:29:07 CET] <Daemon404> interestingly, libcurl didnt implement its own http2
[20:29:14 CET] <Daemon404> yeah.
[20:29:22 CET] <wm4> because old http has not much overhead with such long time bulk transfers
[20:29:44 CET] <Daemon404> i was thinking maybe it matters with fragments
[20:29:51 CET] <Daemon404> but no
[20:29:58 CET] <Daemon404> i cant really see it.
[21:05:48 CET] <cone-219> ffmpeg 03James Almer 07n0.10.16:HEAD: hevcdsp: fix compilation for arm and aarch64
[21:10:27 CET] <Plorkyeran> it'd make using a request per frame more reasonable if you had some reason to want to do that
[21:10:37 CET] <Plorkyeran> but there's plenty of remaining reasons why that'd be dumb
[21:11:29 CET] <BtbN> From what i heard http2 will horrible
[21:12:01 CET] <BtbN> It forces a huge bunch of logic on the server
[21:28:52 CET] <cone-219> ffmpeg 03Niels Möller 07master:78c892284150: dca: Document the permuted order in the lfe_fir_64 and lfe_fir_128 tables
[21:28:53 CET] <cone-219> ffmpeg 03Michael Niedermayer 07master:0bc2da6dcf17: Merge commit '78c892284150e12f6b08b287bdf7e62307c6985f'
[21:39:22 CET] <cone-219> ffmpeg 03Vittorio Giovara 07master:a72d93daa09f: mpegvideo_enc: Check AVCodecContext allocation
[21:39:23 CET] <cone-219> ffmpeg 03Michael Niedermayer 07master:9ecfe0b87834: Merge commit 'a72d93daa09ffbad2771f1450820941055eaf210'
[21:50:22 CET] <cone-219> ffmpeg 03Vittorio Giovara 07master:17c45d4d056d: libtheora: Check frame allocation
[21:50:23 CET] <cone-219> ffmpeg 03Michael Niedermayer 07master:a91c175db208: Merge commit '17c45d4d056d0e10ecb88b424ec9e68be398da5e'
[21:58:21 CET] <cone-219> ffmpeg 03Vittorio Giovara 07master:edca1dd552ef: xcbgrab: Check av_strdup() allocation
[21:58:22 CET] <cone-219> ffmpeg 03Michael Niedermayer 07master:98c387c9c1cd: Merge commit 'edca1dd552efa1ebef016ca9eff4ce6757605819'
[22:26:38 CET] <cone-219> ffmpeg 03Vittorio Giovara 07master:5a2645cafeca: nutdec: Prevent leaks on memory error
[22:26:39 CET] <cone-219> ffmpeg 03Michael Niedermayer 07master:8b3a681aaa7d: Merge commit '5a2645cafeca1c2207ac55cc831c3349572a82ed'
[22:44:15 CET] <joey_> Hello, my name is Aditya and I would like participate in GSOC, I am interested in the basic server for network protocols project.
[22:46:08 CET] <joey_> @george are you online sir? 
[22:49:22 CET] <debianuser> joey_: Please, wait a few minutes. Developers don't read the chat every minute, you may have to wait for ~5-15 minutes before someone who knows the answer will read your question.
[22:50:37 CET] <joey_> @debianuser I just posted the second message to see if I got his IRC handle correctly .... 
[22:52:39 CET] <debianuser> (I don't know the answer, sorry. :( I've seen them mentioning some qualification tasks... But I'm not ffmpeg developer, I'm just looking through its code trying to understand why my code runs so slow...)
[22:53:24 CET] <Mavrik> profilers ftw :)
[22:53:41 CET] <Mavrik> google cpu profiler is pretty useful for that
[22:54:05 CET] Action: debianuser doesn't even know who to ping when someone asks about gsoc...
[22:54:53 CET] <root670> probably best to email the gsoc admin
[22:55:17 CET] <root670> i'm in the same boat, looking to get a little more information about a currently unmentored project idea
[22:56:32 CET] <joey_> @root670 I guess I will have to .. its 3:40 AM here .. too sleepy to wait ... All the best to you :)
[22:57:12 CET] Action: debianuser 'd seen Compn trying to answer some gsoc question before, and someone else...
[23:00:30 CET] <debianuser> Mavrik: heh. The code (pastebin.com/7CRExR4H) gets ~180fps on its own, but when inside libswscale/ (pastebin.com/zsePueme) it slows down to 50fps. Valgrind/callgrind says that 94% of instructions are executed inside that function. How could that be...
[23:00:38 CET] <iive> maybe it is good idea the topic to contain some link to the relevant wiki entry.
[23:01:34 CET] <iive> i think there was a dedicated maillist for GSoC too.
[23:02:20 CET] <iive> https://trac.ffmpeg.org/wiki/SponsoringPrograms/GSoC/2015
[23:03:28 CET] <Mavrik> debianuser, that's a very good question... cache problem or maybe some kind of memcpy?
[23:03:35 CET] <Mavrik> which call takes the most time?
[23:04:29 CET] <debianuser> Mavrik: according to valgrind it's the call to my ff_rgb32toyv12_c() function. :)
[23:04:57 CET] <Mavrik> hrmpf :D
[23:05:34 CET] <Mavrik> the other thing to look at are intel peformance counters
[23:05:36 CET] <Mavrik> cache misses, etc.
[23:06:00 CET] <iive> linux kernel `perf` tool could also be useful.
[23:25:53 CET] <ubitux> michaelni: do you want me to fix it?
[23:26:28 CET] <cone-219> ffmpeg 03Vittorio Giovara 07master:c3bd1d60af97: formats: Check memory allocations
[23:26:29 CET] <cone-219> ffmpeg 03Michael Niedermayer 07master:9f8f2bcab626: Merge commit 'c3bd1d60af97e8d2568dac9fcce7bdabb4ff93c8'
[23:27:38 CET] <michaelni> ubitux, fix what ? but yes whatever it is you talk about yes please fix it :)
[23:28:47 CET] <ubitux> michaelni: lavfi/formats.c refs/realloc/assert thing
[23:29:19 CET] <michaelni> yes, sure :)
[23:39:21 CET] <ubitux> michaelni: done
[23:39:40 CET] <ubitux> i'm currently adding the necessary checks in the filters
[23:41:51 CET] <ubitux> i could probably raise ENOMEM if f is NULL btw..
[23:41:57 CET] <ubitux> this will simplify a bunch of things
[23:42:04 CET] <jamrial> michaelni: your aarch64 qemu box seems to have run out of space (latest fate run didn't even finish compiling)
[23:45:05 CET] <michaelni> jamrial, should be fixed
[00:00:00 CET] --- Fri Mar 13 2015


More information about the Ffmpeg-devel-irc mailing list