[Ffmpeg-devel-irc] ffmpeg-devel.log.20150820

burek burek021 at gmail.com
Fri Aug 21 02:05:03 CEST 2015

[00:33:19 CEST] <cone-247> ffmpeg 03Mariusz SzczepaDczyk 07master:bf5b2f9df876: lavf/ftp: implement move and delete callbacks
[01:57:43 CEST] <llogan> michaelni: what's the logo email about?
[02:11:35 CEST] <michaelni> llogan, you should have gotten the original mail
[02:11:59 CEST] <michaelni> ffmpeg won in some category in some magazine ...
[02:12:29 CEST] <michaelni> i can forwrd the origina mail if you dont have it
[02:14:57 CEST] <llogan> sure. i don't see it
[02:17:45 CEST] <michaelni> llogan, forwarded, i hope you like reading  html
[02:22:33 CEST] <llogan> i see ocassional spams for magazine/business awards. i assumed this was one, but i guess it is ok.
[02:23:49 CEST] <llogan> i missed it earlier because it went into spam and i was too lazy to root through it all
[02:25:49 CEST] <llogan> anyway i sent a SVG logo for them to convert to gif then upscale to 8 inches.
[02:54:33 CEST] <atomnuker> sooo yeah, turns out the aac MIPS encoder must have been broken for around 1.5 months now
[02:54:56 CEST] <atomnuker> both my and klaussfreire's fault since we both changed the function definitions
[02:55:17 CEST] <atomnuker> but the fact that no one even complained speaks volumes about how many people actually use it
[02:56:34 CEST] <rcombs> if the MIPS people care they should add tests
[03:02:43 CEST] <atomnuker> broken, as in, it shouldn't compile, no need to test for that
[03:08:35 CEST] <BBB> atomnuker: dont we have mips qemu fate stations?
[03:12:55 CEST] <atomnuker> looking at fate, we do
[03:13:40 CEST] <atomnuker> it seems aac-aref-encode and aac-ln-encode
[03:14:27 CEST] <atomnuker> actually on second though they probably compile fine
[03:15:04 CEST] <atomnuker> it's just that they put differently defined functions in function pointers
[03:15:09 CEST] <BBB> ah
[03:15:11 CEST] <BBB> right
[03:15:26 CEST] <BBB> yeah we should have fate tests that catch that
[03:15:27 CEST] <BBB> ..
[03:17:39 CEST] <atomnuker> hmm, that's odd, looking at the logs they don't even print warnings about the function pointers
[03:57:05 CEST] <cone-546> ffmpeg 03Michael Niedermayer 07master:db91e0edb63a: avformat/mux: Update sidedata in ff_write_chained()
[07:52:13 CEST] <rcombs> "nothing helps a couple reconcile like a good lawsuit.  Against a third party, that is."
[08:55:49 CEST] <nevcairiel> atomnuker: personally i would've just disabled the mips optimized functions months ago instead of trying to fix them all the time, you two are doing massive changes to the encoder, let some mips person fix them once you are done
[08:57:23 CEST] <rcombs> +1
[09:06:06 CEST] <nevcairiel> also re hevc simd, if they dont get to their senses and actually backport finished simd instead of writing their own version of everything (which they won't, of course), i'll just not merge anything of it. if someone that knows how our hevc dsp is structured wants to adapt any of it to work for us in some way, they are welcome to it
[09:09:14 CEST] <nevcairiel> i just dont understand how someone can start writing something extremely complex like simd for hevc and not even consider if one of the other two places using the same decoder base doesn't have it already
[09:09:22 CEST] <nevcairiel> tiny bubbles everywhere
[09:10:19 CEST] <iive> you mean, they don't do that on purpose?
[09:10:28 CEST] <nevcairiel> it would be extremely stupid
[09:10:32 CEST] <nevcairiel> but maybe they are
[09:10:34 CEST] <nevcairiel> i dont know
[11:24:59 CEST] <ubitux> do we have an example of creating an AVFrame with custom data?
[11:25:29 CEST] <wm4> "custom data"?
[11:26:06 CEST] <ubitux> user data
[11:26:35 CEST] <ubitux> like, what's going to happen if i alloc a frame, set frame->data/linesize and feed it to lav*
[11:27:35 CEST] <ubitux> (and delete my data right away)
[11:33:44 CEST] <wm4> if you don't set the bufs to (essentially) invalid values, then with some luck the receiving API will either just use it for the duration of the function call, or make a copy if it's needed longer
[11:33:49 CEST] <wm4> that's how it should work
[11:40:33 CEST] <wm4> using av_frame_get_buffer and then just changing the data pointers is highly illegal
[11:43:04 CEST] <nevcairiel> if it needs to keep a copy and there is no valid bufs for ref counting, it'll make a copy, and thats that
[11:56:31 CEST] <durandal_1707> when is next release?
[12:19:46 CEST] <cone-497> ffmpeg 03Carl Eugen Hoyos 07master:9bd305be91fd: lavf/mpeg: Allow users to force audio codec MLP.
[12:34:06 CEST] <cone-497> ffmpeg 03Ganesh Ajjanagadde 07master:ecf6b26aa500: avformat/ftp: use correct enum type
[14:29:34 CEST] <cone-497> ffmpeg 03Janne Grunau 07release/2.4:02477323b92a: aac_parser: add required padding for GetBitContext buffer
[14:29:35 CEST] <cone-497> ffmpeg 03Janne Grunau 07release/2.4:a9f108bd78e8: ac3_parser: add required padding for GetBitContext buffer
[14:29:36 CEST] <cone-497> ffmpeg 03Janne Grunau 07release/2.4:aa3ec219e1a5: imc: add required padding for GetBitContext buffer
[14:29:37 CEST] <cone-497> ffmpeg 03Luca Barbato 07release/2.4:859ce02c9815: h263: Always check both dimensions
[14:29:38 CEST] <cone-497> ffmpeg 03Anton Khirnov 07release/2.4:c49b88b93bca: opusdec: properly handle mismatching configurations in multichannel streams
[14:29:39 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:80b1cc31b8f0: Merge commit '02477323b92aacdabe0a2d129eeb0c15fbd1ec9e' into release/2.4
[14:29:40 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:8b0a063ba912: Merge commit 'a9f108bd78e842a47ade2f7c8b22a1764d01d4e6' into release/2.4
[14:29:41 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:7da1e8a0d891: Merge commit 'aa3ec219e1a5cc0e96ddec6ea83312ec780448f5' into release/2.4
[14:29:42 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:e0f859e8fe14: Merge commit '859ce02c9815b492da627d3098548b4f69bbc80e' into release/2.4
[14:29:43 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:f141ac2d7028: Merge commit 'c49b88b93bca53c04f18d78c27dbf1dc6daea909' into release/2.4
[14:53:52 CEST] <cone-497> ffmpeg 03Ganesh Ajjanagadde 07release/2.4:35ab85884b66: swresample/dither: check memory allocation
[14:53:53 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:39430866566a: avformat/mxfenc: Accept MXF D-10 with 49.999840 Mbit/sec
[14:53:54 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:8498cf60b209: Revert "avformat/rtpenc: check av_packet_get_side_data() return, fix null ptr dereference"
[14:53:55 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:df2258f18d18: swresample/swresample: Cleanup on init failure.
[14:53:56 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:0081afeaa741: avcodec/atrac3plusdec: consume only as many bytes as available
[14:53:57 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:4121c1db15e1: avcodec/alsdec: Check for overread
[14:53:58 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:5709ac5c42a7: avcodec/adpcm: Check for overreads
[14:53:59 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:8fc8b3eebe95: avformat/ffmdec: Check ffio_set_buf_size() return value
[14:54:00 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:84cf7418618f: avcodec/jpeg2000dec: Check that coords match before applying ICT
[14:54:01 CEST] <cone-497> ffmpeg 03Simon Thelen 07release/2.4:d7682421528f: libavutil/channel_layout: Correctly return layout when channel specification ends with a trailing 'c'.
[14:54:02 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:402f83298463: avdevice/lavfi: do not rescale AV_NOPTS_VALUE in lavfi_read_packet()
[14:54:03 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:6ff5c4cd49ec: avcodec/h264_slice: Use AVFrame diemensions for grayscale handling
[14:54:04 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:670832e2a273: takdec: ensure chan2 is a valid channel index
[14:54:05 CEST] <cone-497> ffmpeg 03Deliang Fu 07release/2.4:d54ca4167c48: avformat: Fix bug in parse_rps for HEVC.
[14:54:06 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:22349b9f7d4f: ffmpeg_opt: Check for localtime() failure
[14:54:07 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:cd83ff5d4c70: mov: abort on EOF in ff_mov_read_chan
[14:54:08 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:18aef7c07560: libopenjpegdec: check existence of image component data
[14:54:09 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:8c8406462bb5: arm: only enable setend on ARMv6
[14:54:10 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:9da5ba1f57a4: vp9: change type of tile_size from unsigned to int64_t
[14:54:11 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:13a9a0c1a492: vp8: change mv_{min,max}.{x,y} type to int
[14:54:12 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:c5dd6fefd41e: sonic: set avctx->channels in sonic_decode_init
[14:54:13 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:5302adb32367: h264: er: Copy from the previous reference only if compatible
[14:54:14 CEST] <cone-497> ffmpeg 03wm4 07release/2.4:9d0aa058c997: avio: fix potential crashes when combining ffio_ensure_seekback + crc
[14:54:15 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:7de7d81ee55a: postproc: fix unaligned access
[14:54:16 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:e588252a429c: avcodec/dpxenc: implement write16/32 as functions
[14:54:17 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:933ae708c5fc: matroskadec: validate audio channels and bitdepth
[14:54:18 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:b3745ce8c261: matroskadec: check audio sample rate
[14:54:19 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:00772989041a: swr: Remember previously set int_sample_format from user
[14:54:20 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:60a3bd625f4d: swscale/rgb2rgb_template: Disable shuffle_bytes_2103_c on big endian
[14:54:21 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:2f1bff169000: swscale/rgb2rgb_template: Implement shuffle_bytes_0321_c and fix shuffle_bytes_2103_c on BE
[14:54:22 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:bada03d74634: swscale/rgb2rgb_template: Fix signedness of v in shuffle_bytes_2103_c()
[14:54:23 CEST] <cone-497> ffmpeg 03Sebastien Zwickert 07release/2.4:d50d11c56cd4: vda: unlock the pixel buffer base address.
[14:54:24 CEST] <cone-497> ffmpeg 03James Almer 07release/2.4:1795bef7c736: swscale/x86/rgb2rgb_template: add missing xmm clobbers
[14:54:25 CEST] <cone-497> ffmpeg 03James Almer 07release/2.4:286e14667fe4: swscale/x86/rgb2rgb_template: fix signedness of v in shuffle_bytes_2103_{mmx,mmxext}
[14:54:26 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:925adad3e074: ffmpeg: Do not use the data/size of a bitstream filter after failure
[14:54:27 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:a47bc9a05601: swscale/swscale_unscaled: Fix rounding difference with RGBA output between little and big endian
[14:54:28 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:2789d15114cf: wmavoice: limit wmavoice_decode_packet return value to packet size
[14:54:29 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:e35c5348900a: avcodec/pngdec: Only allow one IHDR chunk
[14:54:30 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:3dfadef52216: avfilter/vf_transpose: Fix rounding error
[14:54:31 CEST] <cone-497> ffmpeg 03James Zern 07release/2.4:c6e16ec711c2: vp9/update_prob: prevent out of bounds table read
[14:54:32 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:47a5cde6ac03: avcodec/h264_slice: Use w/h from the AVFrame instead of mb_w/h
[14:54:33 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:514d0e29c8cd: avcodec/aacsbr: check that the element type matches before applying SBR
[14:54:34 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:7fdc2ba3d46e: avcodec/aacsbr: Assert that bs_num_env is positive
[14:54:35 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:7ecaa736e781: avcodec/rawenc: Use ff_alloc_packet() instead of ff_alloc_packet2()
[14:54:36 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:9e52f6b98672: wmalosslessdec: avoid reading 0 bits with get_bits
[14:54:37 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07release/2.4:9463930faf9f: wmalosslessdec: reset frame->nb_samples on packet loss
[14:54:38 CEST] <cone-497> ffmpeg 03Chris Watkins 07release/2.4:3fb241210af6: oggparsedirac: check return value of init_get_bits
[14:54:39 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:c58b0d981ea7: avcodec/mpegvideo: Clear pointers in ff_mpv_common_init()
[14:54:40 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:cbc5d2bf30d0: avcodec/utils: use a minimum 32pixel width in  avcodec_align_dimensions2() for H.264
[14:54:41 CEST] <cone-497> ffmpeg 03Anton Khirnov 07release/2.4:a23a6bf06b99: bytestream2: set the reader to the end when reading more than available
[14:54:42 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:41fba53525f7: avcodec/alac: Clear pointers in allocate_buffers()
[14:54:43 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:f78573466d44: avcodec/sanm: Reset sizes in destroy_buffers()
[14:54:44 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:bd5cf1dd8f9e: avcodec/pthread_frame: check avctx on deallocation
[14:54:45 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:cf27b297ca0b: ffmpeg: Fix cleanup with ost = NULL
[14:54:46 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:634605f79e36: avformat/mov: Fix deallocation when MOVStreamContext failed to allocate
[14:54:47 CEST] <cone-497> ffmpeg 03Zhang Rui 07release/2.4:7a7ec3ccd9b6: avutil/fifo: Fix the case where func() returns less bytes than requested in av_fifo_generic_write()
[14:54:48 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:3bc20fe48c5e: swscale/utils: Clear pix buffers
[14:54:49 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:e3bacdbaadda: avcodec/pthread_frame: clear priv_data, avoid stale pointer in error case
[14:54:50 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:649f09c4cfe9: avfilter/af_aresample: Check ff_all_* for allocation failures
[14:54:51 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:24f1698758f0: avcodec/rv34: Clear pointers in ff_rv34_decode_init_thread_copy()
[14:54:52 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:7ae349a324f1: avcodec/diracdec: Check for hpel_base allocation failure
[14:54:53 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:8fe79605fdce: avcodec/vp8: Fix null pointer dereference in ff_vp8_decode_free()
[14:54:54 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:d2b0aae5e1b2: avcodec/vp8: Check buffer size in vp8_decode_frame_header()
[14:54:55 CEST] <cone-497> ffmpeg 03wm4 07release/2.4:2d582d142c18: rawdec: fix mjpeg probing
[14:54:56 CEST] <cone-497> ffmpeg 03wm4 07release/2.4:43956940ea20: rawdec: fix mjpeg probing buffer size check
[14:54:57 CEST] <cone-497> ffmpeg 03Emanuel Czirai 07release/2.4:33629ff60fb3: libavcodec/aacdec_template: Use init_get_bits8() in aac_decode_frame()
[14:54:58 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:595af5a036f4: avcodec/dcaenc: clear bitstream end
[14:54:59 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:eac75d405b97: avcodec/svq1enc: Check dimensions
[14:55:00 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:ac3358d73ad2: avcodec/flashsvenc: Correct max dimension in error message
[14:55:01 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:cbaa9ef0d601: avformat/mux: Update sidedata in ff_write_chained()
[14:56:43 CEST] <wm4> I didn't know we have someone maintaining releases
[15:01:19 CEST] <BBB> I still have 5 patches (or 4, I guess) awaiting review, anyone want to help?
[15:01:53 CEST] <BBB> https://github.com/rbultje/ffmpeg/commit/e5bb803bdee35d67e1ffd4cc001a155fdf120730, https://github.com/rbultje/ffmpeg/commit/dc8062fa28d3462fd952f6e27193b27d075be3bd, https://github.com/rbultje/ffmpeg/commit/304586445ffe85eae1a87b3a37fd38a0f290b87b and https://github.com/rbultje/ffmpeg/commit/87cb730a91e4e6cd08fe9ee152c1e0815a635ffb
[15:05:30 CEST] <wm4> send them to the ML?
[15:17:23 CEST] <cone-497> ffmpeg 03Michael Niedermayer 07release/2.4:98f167202262: Update for 2.4.11
[15:22:57 CEST] <BBB> wm4: they are there, unresponded-to
[15:23:03 CEST] <wm4> send them again
[15:50:30 CEST] <Compn> did someone already bring up the audible decoder license header ?
[15:50:42 CEST] <Compn> i dont want to read 50 replies to the patch :\
[15:51:26 CEST] <Compn> oh yes durandal_1707 did 
[15:51:30 CEST] <Compn> good, i dont have to reply then
[16:51:49 CEST] <ubitux> anyone opposed to a c muxer? :)
[17:11:09 CEST] <wm4> a what
[17:20:15 CEST] <kierank> ubitux: a what
[17:20:54 CEST] <ubitux> kierank: a .c file muxer, with "unsigned char *buf = { ... } " and actually other meta
[17:21:01 CEST] <ubitux> for direct inclusion within code
[17:21:05 CEST] <ubitux> gimp does that for images typically
[17:21:29 CEST] <ubitux> (seems i'm going to hell soon)
[17:21:33 CEST] <Compn> it muxes c code
[17:21:37 CEST] <Compn> from what kind of input ? :P
[17:21:54 CEST] <ubitux> let's say rgb rawvideo
[17:21:56 CEST] <nevcairiel> probably something silly as raw rgb or something
[17:21:57 CEST] <nevcairiel> see
[17:21:58 CEST] <nevcairiel> :D
[17:22:00 CEST] <Compn> lol
[17:22:01 CEST] <ubitux> :D
[17:22:28 CEST] <ubitux> seems nevcairiel wanted to do that as well but never dare to propose it
[17:22:33 CEST] <ubitux> i'll take that as a ack!
[17:22:44 CEST] <nevcairiel> no, i dont have any clue what you would want to use that for
[17:22:53 CEST] <ubitux> stupid shit obviously
[17:23:14 CEST] <nevcairiel> for the one time in a  lifetime i would need it, i would just dumb the rawvideo with a hex editor and wrap { } around it =p
[17:23:19 CEST] <nevcairiel> dump*
[17:23:29 CEST] <ubitux> yeah, i muxed to rawvideo and used xxd -i
[17:41:54 CEST] <wm4> ubitux: that wouldn't be a muxer, but a protocol
[17:44:41 CEST] <nevcairiel> i dont know, its kind of a container for raw formats
[17:54:39 CEST] <wm4> it just wraps a byte stream, nothing else is special about it
[18:10:08 CEST] <oreWizard> someone use ffmpeg with c#?
[18:12:08 CEST] <kierank> ubitux: my joylent finally arrived
[18:13:49 CEST] <podman> Is there any reason why -map_metadata -1 wouldn't remove menu metadata?
[20:35:15 CEST] <cone-497> ffmpeg 03Lou Logan 07master:538ea9a20976: doc/indevs: Remove non-existent decklink option examples
[20:56:36 CEST] <durandal_1707> so nobody against?
[21:03:33 CEST] Action: durandal_1707 started debate with inocent patch
[21:22:35 CEST] <cone-497> ffmpeg 03Ganesh Ajjanagadde 07master:b772847eb04e: avformat/webmdashenc: fix uninitialized variable
[21:31:17 CEST] <cone-497> ffmpeg 03Ganesh Ajjanagadde 07master:c3052c9e8a91: avcodec/aacsbr_fixed: remove unused variable
[22:07:29 CEST] <cone-497> ffmpeg 03Ludmila Glinskih 07master:e23a01c06df3: api-seek-test: first version
[22:42:19 CEST] <cone-497> ffmpeg 03Andreas Cadhalpun 07master:96f5fdebc3c6: graph2dot: use avfilter_pad_get_name accessor function
[23:56:23 CEST] <cone-497> ffmpeg 03Ganesh Ajjanagadde 07master:e8319f602e94: avcodec/metasound_data: comment out unused variable
[00:00:00 CEST] --- Fri Aug 21 2015

More information about the Ffmpeg-devel-irc mailing list