[Ffmpeg-devel-irc] ffmpeg-devel.log.20181023

[burek]

[00:00:21 CEST] <nevcairiel> git isnt very useful in showing that info
[00:01:24 CEST] <JEEB> yea, that's how it looks like
[00:01:44 CEST] <JEEB> if you don't have libav master around it won't show that one of them is part of that
[00:03:02 CEST] <jamrial> yeah, one is from libav
[00:03:16 CEST] <jamrial> the merge commit is a separate commit than the merged commit in question
[00:03:21 CEST] <JEEB> yup
[00:03:39 CEST] <jamrial> http://git.videolan.org/?p=ffmpeg.git;a=commit;h=0ff76ca86e0ea4dcf2b392c45f5fac8e5576bb0d
[00:20:24 CEST] <cone-636> ffmpeg 03Mark Thompson 07master:3143fe34f622: cbs_h264: Actually decompose end-of-sequence NAL units
[00:20:25 CEST] <cone-636> ffmpeg 03Mark Thompson 07master:e5fda1ff2896: cbs_h264: Fix SPS used for pic_timing timestamp
[00:20:26 CEST] <cone-636> ffmpeg 03Mark Thompson 07master:da6db843b80e: cbs_h2645: Allocate all internal buffers with padding
[00:21:39 CEST] <tmm1> jkqxz: thanks
[00:21:51 CEST] <jkqxz> :)
[00:22:59 CEST] <jkqxz> jamrial:  Was there any interest in the libaom tiles thing?  (Nobody replied.)
[00:23:24 CEST] <jamrial> if nobody replied then apply it
[00:23:52 CEST] <jamrial> and there is interest, of course. no other way to get multithreading encoding, or tile threading during decoding :p
[00:26:57 CEST] <jamrial> jkqxz: also, if you dont' have time to review the amf patch, then we should apply it. it's been in the list for four months now, and supposedly blocking further work
[00:27:17 CEST] <jamrial> its author is the only amf developer, so he's a defacto maintainer
[00:27:34 CEST] <jamrial> i'd wait until 4.1 branch is created in that case, though
[00:28:00 CEST] <jkqxz> Bleh, I really should look at that.
[00:28:39 CEST] <JEEB> I should ping the discontinuity logging patches again
[00:28:45 CEST] <JEEB> since I only got an OK to one of them I think
[00:29:21 CEST] <jkqxz> It's been at the bottom of my todo list for ages, and keeps getting pushed down because ugly proprietary stuff I don't really care much about (but would prefer not to be horrible).
[00:43:03 CEST] <jkqxz> jamrial:  prores cbs looks like it would be completely trivial to implement, since the header is fixed.  Dunno if anyone would ever care about the trace, probably not given how little it contains.
[00:51:01 CEST] <jamrial> jkqxz: alright. just saying since it'd be nice as a long term project to shift bitstream parsing in general to CBS
[00:51:57 CEST] <jamrial> considering that eventually a new parser api based on avpacket should be written
[00:52:56 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07master:e90f0ac334f2: avcodec/ilbcdec: Check startindex
[00:52:57 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07master:305e523105f6: avcodec/jpeg2000dec: Fix off by 1 error in JPEG2000_PGOD_CPRL handling
[01:08:17 CEST] <jkqxz> Yeah, I'm in favour of doing it, but I agree it shouldn't block the trivial metadata bsf.
[01:08:27 CEST] <jkqxz> Is there a spec anywhere for proper names?  It would probably only take half an hour to write.
[01:12:48 CEST] <jkqxz> (All I can find from Apple is an informationless thing saying how good it is and that you shouldn't use FFmpeg because it is not an Authorised Implementation and that anyone who uses it is therefore Very Naughty.)
[01:20:37 CEST] <JEEB> yea, unfortunately they didn't open source it like ALAC (probably for some very obvious Business Reasons)
[01:41:43 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/4.0:b2e9d3da817f: avformat/utils: Fix integer overflow in discontinuity check
[01:41:44 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/4.0:bccddd7fcc14: avformat/utils: Never store negative values in last_IP_duration
[01:41:45 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/4.0:65b0caf47d73: avcodec/ra144: Fix integer overflow in add_wav()
[01:41:46 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/4.0:1061a2e2efa9: avcodec/h264_cavlc: Check mb_skip_run
[01:41:47 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/4.0:6e14fc4aa1a2: avcodec/mpeg4videodec: Fix typo in sprite delta check
[01:41:48 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/4.0:afebdc3ed789: avcodec/jpeg2000dec: Fix off by 1 error in JPEG2000_PGOD_CPRL handling
[01:41:49 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/4.0:938bc919eac4: Prepare 4.0.3 release
[01:51:52 CEST] <jkqxz> nevcairiel:  Did you ever have a machine which hit the Intel H.265 plugin segfault on load problem?  If so, does it still happen?  If not, any idea how to track down what was going there (when considering ignoring it)?
[02:13:01 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:b8fd13befef4: libavformat/oggparsevorbis: Fix memleak on multiple headers
[02:13:02 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:7fd80d91f72c: avformat/utils: Fix integer overflow of fps_first/last_dts
[02:13:03 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:81a6076e4b8e: avformat/oggparsetheora: Do not adjust AV_NOPTS_VALUE
[02:13:04 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:4df16ad1ef06: avformat/avidec: Fix integer overflow in cum_len check
[02:13:05 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:46f37c65abc0: avformat/oggparseogm: Fix undefined shift in ogm_packet()
[02:13:06 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:69344f628bf5: avformat/oggparseogm: Check lb against psize
[02:13:07 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:ad7c57f9db50: avformat/oggparsedaala: Do not adjust AV_NOPTS_VALUE
[02:13:08 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:30d40580c4e0: avformat/mov: Fix integer overflows related to sample_duration
[02:13:09 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:448cd0d0b341: avcodec/wmalosslessdec: Reset num_saved_bits on error path
[02:13:10 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:e62380038765: avcodec/nuv: Check for minimum input size for uncomprssed and rtjpeg
[02:13:11 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:f4fe7022583e: avcodec/nuv: rtjpeg with dimensions less than 16 would result in no decoded pixels thus reject it
[02:13:12 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:028af5acbe7b: avformat/mov: Check STSC and remove invalid entries
[02:13:13 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:7ccb9c37ac47: avcodec/get_bits: Make sure the input bitstream with padding can be addressed
[02:13:14 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:0721e3c1bd23: avcodec/tableprint_vlc: Fix build failure with --enable-hardcoded-tables
[02:13:15 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:03f891c05099: avcodec/wmalosslessdec: Fix null pointer dereference in decode_frame()
[02:13:16 CEST] <cone-636> ffmpeg 03heimdallr 07release/3.0:19379529a5b8: avcodec/imgconvert: Fix loss mask bug in avcodec_find_best_pix_fmt_of_list()
[02:13:17 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:1f648a098da0: avcodec/cscd: Error out when LZ* decompression fails
[02:13:18 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:98096645f28f: avcodec/aacdec_fixed: Fix integer overflow in apply_independent_coupling_fixed()
[02:13:19 CEST] <cone-636> ffmpeg 03Hendrik Schreiber 07release/3.0:da371c5b215d: swresample/swresample: Fix for seg fault in swr_convert_internal() -> sum2_float during dithering.
[02:13:20 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:b2be78f9d2de: libavcodec/rv34: error out earlier on missing references
[02:13:21 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:72b9ba8a5bfe: avcodec/aacsbr_fixed: Fix integer overflow in sbr_hf_assemble()
[02:13:22 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:c0bb5613d42f: avcodec/movtextdec: Check style_start/end
[02:13:23 CEST] <cone-636> ffmpeg 03Matt Wolenetz 07release/3.0:4b04da1e52da: lavc/libopusdec: Allow avcodec_open2 to call .close
[02:13:24 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:5aaa5bd1b04d: avcodec/mjpegdec: Check input buffer size.
[02:13:25 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:065057c8a351: avcodec/truemotion2: Propagate out of bounds error from GET_TOK()
[02:13:26 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:cf321e93c4a9: avcodec/utils: Enforce minimum width also for VP5/6
[02:13:27 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:b591673c507d: avformat/utils: Check cur_dts in update_initial_timestamps() more
[02:13:28 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:9665d6258c6c: doc/APIchanges: Fix typos in hashes
[02:13:29 CEST] <cone-636> ffmpeg 03Rahul Chaudhry 07release/3.0:ad3ec05d036a: swresample/arm: remove unintentional relocation.
[02:13:30 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:5891d222ff6e: avcodec/cinepak: move some checks prior to frame allocation
[02:13:31 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:3009bf0be645: avcodec/cinepak: Skip empty frames
[02:13:32 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:e49e8259df91: avcodec/dfa: Check dimension against maximum
[02:13:33 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:a8640c87a70f: avcodec/dsicinvideo: Propagate errors from cin_decode_rle()
[02:13:34 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:3bf0a405f051: avcodec/dsicinvideo: Fail if there is only a small fraction of the data available that comprises a full frame
[02:13:35 CEST] <cone-636> ffmpeg 03Stephan Holljes 07release/3.0:73d07e320c4a: lavf/http.c: Free allocated client URLContext in case of error.
[02:13:36 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:142e1e3e9af8: avcodec/ffv1enc: Check that the crc + version combination is supported
[02:13:37 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:70c7c7392089: avcodec/vc1_block: simplify ac_val computation
[02:13:38 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:cc8710b5afeb: avcodec/elsdec: Fix memleaks
[02:13:39 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:aaa2d4a26523: avcodec/h263dec: Check slice_ret in mspeg4 slice loop
[02:13:40 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:baa4913e1272: avcodec/error_resilience: Fix integer overflow in filter181()
[02:13:41 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:04e69effbb31: avcodec/wavpack: Fix integer overflow in DEC_MED() / INC_MED()
[02:13:42 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:9505b285385e: avcodec/flac_parser: Fix infinite loop
[02:13:43 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:88e5a30cf835: avcodec/g2meet: Change order of operations to avoid undefined behavior
[02:13:44 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:6aae60cc7da8: avcodec/jpeg2000dec: Skip init for component in CPRL if nothing is to be done
[02:13:45 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:6850377b598c: avcodec/jpeg2000dec: Fix undefined shift in the jpeg2000_decode_packets_po_iteration() CPRL case
[02:13:46 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:00c18862a313: avcodec/g2meet: Check RGB upper limit
[02:13:47 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:1747563cead7: avcodec/fic: Check available input space for cursor
[02:13:48 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:a8fcb810c964: avcodec/dirac_dwt: Fix integer overflow in COMPOSE_DD97iH0 / COMPOSE_DD137iL0
[02:13:49 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:ffe4ffa4ef7a: avformat/mov: Only fail for STCO/STSC contradictions if both exist
[02:13:50 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:876d3ee86209: indeo4: Decode all or nothing of a band header.
[02:13:51 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:3eff19d38806: avcodec/mlpdec: Only change noise_type if the related fields are valid
[02:13:52 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:0d5851101311: avcodec/mpeg4videoenc: Use 64 bit for times in mpeg4_encode_gop_header()
[02:13:53 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:fe37daf25dc6: avcodec/g723_1dec: Clip bits2 in both directions
[02:13:54 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:be77d7ba7879: oavcodec/aacpsdsp_template: Use unsigned for hs0X to prevent undefined behavior
[02:13:55 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:1f81818c6afe: avcodec/aacdec_fixed: use 64bit to avoid overflow in rounding in apply_dependent_coupling_fixed()
[02:13:56 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:243bdbde5718: avcodec/g2meet: ask for sample with overflowing RGB
[02:13:57 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:fe9c23bebbb2: avcodec/fic: Avoid some magic numbers related to cursors
[02:13:58 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:b4024909c121: avformat/mov: Break out early if chunk_count is 0 in mov_build_index()
[02:13:59 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:7f142e540231: avformat/mov: replace a value error by clipping into valid range in mov_read_stsc()
[02:14:00 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:bbd9f480f3e5: avcodec/amrwbdec: Fix division by 0 in find_hb_gain()
[02:14:01 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:f144d5eb4977: avcodec/opus_silk: Change silk_lsf2lpc() slightly toward silk/NLSF2A.c
[02:14:02 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:aee71463fb2f: avcodec/truemotion2: Fix overflow in tm2_apply_deltas()
[02:14:03 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:0c645bd73c7a: avcodec/mjpegdec: Fix integer overflow in ljpeg_decode_rgb_scan()
[02:14:04 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:07fd8627e5c2: avcodec/shorten: Sanity check nmeans
[02:14:05 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:50e55b81be23: avcodec/shorten: Fix multiple integer overflows
[02:14:06 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:23fdebaec579: avcodec/wavpack: Fix overflow in adding tail
[02:14:07 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:aa803d93bf89: avcodec/xwddec: Use ff_set_dimensions()
[02:14:08 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:18214e2a3f20: avcodec/h264_mc_template: Only prefetch motion if the list is used.
[02:14:09 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:ff75dc10fd27: avformat/mov: Only set pkt->duration to non negative values
[02:14:10 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:963915a4b416: fftools/ffmpeg: Fallback to duration if sample rate is unavailable
[02:14:12 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:626143903b1d: avutil/common: Fix undefined behavior in av_clip_uintp2_c()
[02:14:12 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:a1c125d2bda6: avcodec/dirac_dwt_template: Fix undefined behavior in interleave()
[02:14:13 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:eaefd3ada963: avcodec/aacdec_fixed: Fix undefined integer overflow in apply_independent_coupling_fixed()
[02:14:14 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:d862380718a1: avcodec/mjpegdec: Check for end of bitstream in ljpeg_decode_rgb_scan()
[02:14:15 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:1e067b7ed86c: avcodec/escape124: Check buf_size against num_superblocks
[02:14:16 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:662f7cef06d3: avcodec/cscd: Check output buffer size for lzo.
[02:14:17 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:181c3d82e902: avcodec/ra144: Fix integer overflow in ff_eval_refl()
[02:14:18 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:dd7d2770f331: avcodec/escape124: Fix spelling errors in comment
[02:14:19 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:405cfcae41cd: avcodec/jpeg2000dec: Check that there are enough bytes for all tiles
[02:14:20 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:3afdb1c8a752: avcodec/jpeg2000dec: Fixes invalid shifts in jpeg2000_decode_packets_po_iteration()
[02:14:21 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:21a65701581d: avcodec/shorten: Fix undefined addition in shorten_decode_frame()
[02:14:22 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:438e848b65c2: avcodec/indeo4: Check for end of bitstream in decode_mb_info()
[02:14:23 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:ee8c6566e2ab: avformat/rmdec: Do not pass mime type in rm_read_multi() to ff_rm_read_mdpr_codecdata()
[02:14:24 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:6cadf46dff14: avformat/pva: Check for EOF before retrying in read_part_of_packet()
[02:14:25 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:94edbf464c00: avformat/mms: Add missing chunksize check
[02:14:26 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:469503ac1de3: avformat/movenc: Check that frame_types other than EAC3_FRAME_TYPE_INDEPENDENT have a supported substream id
[02:14:27 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:4ead5a947003: avcodec/mjpegdec: Check for odd progressive RGB
[02:14:28 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:790e6fead078: avformat/movenc: Check input sample count
[02:14:29 CEST] <cone-636> ffmpeg 03James Almer 07release/3.0:c3831e89efce: libwebpenc_animencoder: add missing braces to struct initialization
[02:14:30 CEST] <cone-636> ffmpeg 03James Almer 07release/3.0:0d4a11d0a947: avformat/utils: fix mixed declarations and code
[02:14:31 CEST] <cone-636> ffmpeg 03Rahul Chaudhry 07release/3.0:328ab7f0a1ca: swresample/arm: rename labels to fix xcode build error
[02:14:32 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:0c9c4c3103c2: avcodec/dirac_dwt_template: Fix signedness regression in interleave()
[02:14:33 CEST] <cone-636> ffmpeg 03Simon Thelen 07release/3.0:e7dc6231e263: avcodec/imgconvert: fix possible null pointer dereference
[02:14:34 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:d7d237a441cb: avcodec/dvdsub_parser: Init output buf/size
[02:14:35 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:961eeb4035c3: avcodec/dvdsub_parser: Allocate input padding
[02:14:36 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:60493f388ca7: avutil/pixfmt: Document chroma plane size for odd resolutions
[02:14:37 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:d7d54f3fd5a6: swresample/swresample: Fix input channel count in resample_first computation
[02:14:38 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:057cfa4200e4: avcodec/diracdec: Prevent integer overflow in intermediate in global_mv()
[02:14:39 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:86dfce06e3ac: avcodec/dirac_dwt_template: Fix several integer overflows in horizontal_compose_daub97i()
[02:14:40 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:cc82d22289c9: avcodec/diracdec: Change frame_number to 64bit as its a 32bit from the bitstream and we also have a -1 special case
[02:14:41 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:9f18b056da9b: avcodec/diracdec: Check bytes count in else branch in decode_lowdelay() too
[02:14:42 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:f81fd4c524e7: avcodec/qtrle: Check remaining bytestream in qtrle_decode_XYbpp()
[02:14:43 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:f6e570aa011b: avcodec/aacpsdsp_template: Fix integer overflow in ps_stereo_interpolate_c()
[02:14:44 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:f7778c2ce954: avcodec/mpegaudio_parser: Initialize poutbuf*
[02:14:45 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:5eeaaa29fae0: avcodec/shorten: Check verbatim length
[02:14:46 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:f1425b389a3b: avcodec/shorten: Fix integer overflow in residual/LPC combination
[02:14:47 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:0ea54cae2182: avformat/rmdec: Fix EOF check in the stream loop in ivr_read_header()
[02:14:48 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:4babf70c7efb: avformat/mlvdec: read_string() received unsigned size, make the argument unsigned
[02:14:49 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:172bb520c98f: avformat/nsvdec: Do not parse multiple NSVf
[02:14:50 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:22f743e60201: avcodec/snowdec: Fix integer overflow with motion vector residual
[02:14:51 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:c2ac8d314778: avcodec/vb: Check for end of bytestream before reading blocktype
[02:14:52 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:23c5d8d9ce7d: avcodec/hq_hqa: Check remaining input bits in hqa_decode_mb()
[02:14:53 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:fceedd4de9b2: avcodec/ra144: Fix undefined integer overflow in add_wav()
[02:14:54 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:fcab1b996a21: avcodec/h264_refs: Document last if() in ff_h264_execute_ref_pic_marking()
[02:14:55 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:1049ccec3449: avcodec/dvdsubdec: Avoid branch in decode_run_8bit()
[02:14:56 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:1ed6224cdf28: avcodec/shorten: Fix bitstream end check in read_header()
[02:14:58 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:6161aade4c8d: avcodec/zmbv: Update decomp_len in raw frames
[02:14:58 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:648b904662c2: avcodec/zmbv: Check that the decompressed data size is correct
[02:14:59 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:58cb3ad576cb: avcodec/mpeg4videodec: Fix undefined shift in get_amv()
[02:15:00 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:9b95c4740d23: avcodec/dvdsubdec: Sanity check len in decode_rle()
[02:15:01 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:4b14c3ed7857: avcodec/unary: Improve get_unary() docs
[02:15:02 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:51404bb4f613: avformat/utils: Fix integer overflow in discontinuity check
[02:15:03 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:6a5b0a3c7513: avformat/utils: Never store negative values in last_IP_duration
[02:15:04 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:b15db639a5ca: avcodec/ra144: Fix integer overflow in add_wav()
[02:15:05 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:f9cfdf2baefd: avcodec/h264_cavlc: Check mb_skip_run
[02:15:06 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:9c1bb7e8debd: avcodec/mpeg4videodec: Fix typo in sprite delta check
[02:15:07 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:dd5232c83847: avcodec/jpeg2000dec: Fix off by 1 error in JPEG2000_PGOD_CPRL handling
[02:31:55 CEST] <cone-636> ffmpeg 03Michael Niedermayer 07release/3.0:1dbf2bc7a958: Update for 3.0.12
[03:51:27 CEST] <tmm1> jkqxz: i had to use https://0x0.st/s6Ks.diff to get the cbs_misc patchset to compile with master. is there a better value to pass in for subscripts?
[09:46:13 CEST] <nevcairiel> jkqxz: i dont remember encountering that particular problem, but i havent messed with qsv hevc encoding much
[10:51:04 CEST] <rcombs> I'm working on an I-D to add arbitrary channel layout support to Opus (I'll probably send it for FLAC as well)
[10:52:46 CEST] <rcombs> I'm intending to use the channel "indexes" from AV_CH_*, which afaik are from MS's dwChannelMask; anyone know if that's defined anywhere generic I can reference, or should I link to MS's WAVEFORMATEXTENSIBLE spec, or copy it wholesale into an I-D?
[10:56:26 CEST] <nevcairiel> whats an I-D
[10:58:58 CEST] <rcombs> precursor to an RFC
[10:59:07 CEST] <rcombs> proposed update to Opus
[11:02:05 CEST] <LigH> Hi.
[11:02:43 CEST] <LigH> Excuse the slightly indirect question...
[11:03:08 CEST] <LigH> Can you imagine any change in libavformat which may have broken compatibility to mplayer?
[11:03:25 CEST] <nevcairiel> original mplayer is so old, any change can break it =p
[11:03:37 CEST] <LigH> Since about 3 weeks it fails due to redefinition of pollfd.
[11:04:27 CEST] <LigH> Sounds like I have to abandon all hope for mencoder.
[11:04:43 CEST] <durandal_1707> why you still use it?
[11:05:13 CEST] <LigH> It has some more concenient CLI syntax in a few special cases.
[11:05:19 CEST] <LigH> *convenient
[11:06:37 CEST] <LigH> Well, OK, thanks.
[11:06:41 CEST] <JEEB> pollfd? that sounds like !FFmpeg thing
[11:07:04 CEST] <LigH> If you are interested in details: https://trac.mplayerhq.hu/ticket/2350
[11:07:28 CEST] <JEEB> anyways, mplayer always had a funky relationship with FFmpeg :P did they ever fix building with external FFmpeg
[11:07:29 CEST] <LigH> winsock2 vs. 3dostr
[11:08:11 CEST] <LigH> or os_support
[11:09:05 CEST] <JEEB> anyways, without further looking I guess this is just mplayer's build system failing with current FFmpeg or mingw-w64 specifically :P
[11:09:12 CEST] <LigH> Seems to depend on MSYS2/MinGW; I heard that it builds under Linux/MinGW instead...
[11:09:34 CEST] <JEEB> would be surprised, unless the build system is using a different config.h :P
[11:09:50 CEST] <nevcairiel> dont they use a pre-defined config.h
[11:09:54 CEST] <nevcairiel> no wonder that breaks all the time
[11:09:56 CEST] <JEEB> yes
[11:09:59 CEST] <JEEB> I would guess that
[11:10:00 CEST] <LigH> So I am back to my most discouraging question: Who should I ask, at all? ... ;-)
[11:10:14 CEST] <JEEB> go2 #mplayer ? or whatever is their IRC channel
[11:10:28 CEST] <JEEB> at this point I think mplayer is a single or two person hobby project
[11:10:30 CEST] <LigH> #mplayerdev ... {tumbleweed/crickets}
[11:10:51 CEST] <LigH> penvado sleeps in both.
[11:10:56 CEST] <JEEB> my condolences then
[11:10:59 CEST] <LigH> *pengvado
[11:11:28 CEST] <LigH> debianuser tried to help me but got stuck.
[11:12:14 CEST] <LigH> OK, back to different tasks. Bye.
[11:27:11 CEST] <cone-100> ffmpeg 03Paul B Mahol 07master:b7b3e8c5c759: avfilter: add vibrance filter
[11:51:09 CEST] <ubitux> durandal_1707: is it matching adobe photoshop vibrance filter?
[11:55:08 CEST] <durandal_1707> ubitux: i dont have adobe photoshop
[11:58:44 CEST] <ubitux> ok
[12:00:52 CEST] <durandal_1707> ubitux: but it should behave similar for >= 0 intensity and other options untouched, just different scaling
[12:54:49 CEST] <January> michaelni: could you UL https://0x0.st/s6-6.h264 as fate/h264/crew_cif_timecode-2.h264
[13:58:11 CEST] <cone-100> ffmpeg 03Zhong Li 07master:0e3d7d845d9b: lavc/qsvenc_jpeg: add async_depth support
[13:58:12 CEST] <cone-100> ffmpeg 03Zhong Li 07master:92c25963e8b6: lavc/qsvenc_jpeg: set a default quality
[16:59:58 CEST] <kierank> January: gonna push your changes
[17:00:14 CEST] <January> timecode? sure
[17:01:31 CEST] <cone-100> ffmpeg 03Devin Heitmueller 07master:4241e44a3c01: lavc/h264: create AVFrame side data from H.264 timecodes
[17:01:32 CEST] <cone-100> ffmpeg 03Josh de Kock 07master:76a8c3b522bf: h264/pic_timing: support multiple timecodes
[17:01:33 CEST] <kierank> waiting...
[17:01:33 CEST] <cone-100> ffmpeg 03Josh de Kock 07master:0ac82eed6f9c: lavfi/vf_showinfo: support displaying S12M timecode sidedata
[17:01:34 CEST] <cone-100> ffmpeg 03Josh de Kock 07master:3f2103ec73c9: fate: add h264 timecode test
[17:01:49 CEST] <kierank> that took a while
[17:42:22 CEST] <kierank> January: looks like fate is passing
[19:59:12 CEST] <cone-100> ffmpeg 03Dave Rice 07master:110b4a491859: avdevice/sdl2: add option to set window position
[19:59:13 CEST] <cone-100> ffmpeg 03Matsuzawa Tomohiro 07master:c2ac3b8e6a04: avformat/libsrt: add several options supported in srt 1.3.0
[20:44:43 CEST] <kierank> does anyone else have the ability to add to fate?
[20:44:45 CEST] <kierank> apart from `md 
[20:44:46 CEST] <kierank> michaelni: 
[20:52:42 CEST] <jamrial> jkqxz: https://0x0.st/s6Nh.ivf https://pastebin.com/j94dYarf
[20:54:55 CEST] <jamrial> redundant frame header obu. seen_frame_header is correctly set to true when it shows up (as per the spec), but cbs is not handling it right
[21:01:44 CEST] <michaelni> kierank, you mean add fate samples ? there should be many people who have the rights, but few use them, the process to upload is described in doc/fate.texi
[21:02:43 CEST] <kierank> michaelni: can you give January access
[21:09:04 CEST] <michaelni> January, i realize now (looking at fate.ffmpeg.org) that you pushed the test with noone uploading the sample, i saw the push and thought someone else uploaded the sample already so i didnt, will double check that the sample isnt there and then uplaod
[21:15:27 CEST] <michaelni> January, crew_cif_timecode-2.h264 uploaded. I didnt realize it earlier as i was testing your code and so had manually added the file locally but then when i was done testing saw you already pushed and assumed without checking that the sample was uploaded. Please make sure when you push some new test that all samples it needs have been uploaded by someone before
[21:18:39 CEST] <michaelni> and ideally samples should be uploaded 24h before the code, there where some fate clients syncing the samples only once a day
[21:41:31 CEST] <kierank> JEEB: lotharkript_ says we can use oss-fuzz do to the h264 bitexact fuzzing
[22:49:42 CEST] <atomnuker> peloverde: there's some aac decoder layout patches on the ML, will you have the time to look at them anytime soon?
[22:50:00 CEST] <peloverde> yeah, I should have time
[22:50:12 CEST] <peloverde> I'll take a look now
[23:28:33 CEST] <jkqxz> jamrial:  Hmm, yeah.  I think I must have misread that when writing it (copy as a verb rather than a noun?), since it currently doesn't look for anything there.
[23:28:46 CEST] <jkqxz> Dealing with it is not very fun, since parsing it again would require all of the reference state to be the same as it was at the beginning of the first frame header.
[23:30:25 CEST] <jkqxz> Maybe we could save a reference to the first frame header OBU along with bounds in the context, and then just compare the bits if we see another frame header?
[23:30:54 CEST] <jkqxz> Or save all of the state before the frame header so that it can be restored if there is another one.
[23:34:29 CEST] <jkqxz> nevcairiel:  I thought it might have been you who hit it, since you mention it in that old thread.  I can't remember where the report came from now.
[23:36:18 CEST] <jkqxz> I guess if we don't even know exactly what the problematic case actually is now then just enabling hardware is probably ok.  Maybe some poor sod with an old OEM driver and be screwed over by it, but oh well.
[23:36:43 CEST] <nevcairiel> its not like the software variant is probably all that good for them either
[23:38:15 CEST] <jamrial> jkqxz: a decoder is meant to ignore the redundant frame header (the spec states seen_frame_header must be 1 when those show up)
[23:38:16 CEST] <jkqxz> Well, they probably don't have the software one either.  The evil case is that any application trying to probe the supported codecs will segfault.
[23:38:38 CEST] <jamrial> they show up in between tile groups, when error resilence is used
[23:38:49 CEST] <jamrial> they are meant to be bitexact to the actual frame header
[23:39:40 CEST] <jkqxz> Perhaps the interesting question is - how should they appear in trace output?
[23:40:46 CEST] <jamrial> cbs should fully parse them, so the whole seen_frame_header logic should be used only to make sure the bitstream is correct but not to skip parsing an obu
[23:41:18 CEST] <jamrial> "Redundant Frame Header" would look nicer than "Frame Header"
[23:42:01 CEST] <jamrial> and being in its own function you can make it have a different seen_frame_header check than for frame and frame header obus
[23:45:06 CEST] <jamrial> the spec requires seen_frame_header to be 0 when you parse a frame header obu, but 1 when you parse a redundant frame header obu. otherwise the bitstream is invalid
[23:46:55 CEST] <jkqxz> In theory we could be parsing a stream which has lost some packets and therefore the redundant frame header is the one actually being used, though.
[23:51:45 CEST] <peloverde> atomnuker: I reviewed 3/3. the others look encoder centric?
[23:58:01 CEST] <atomnuker> yeah, I dislike how they're adding even more custom layouts for no good reason, so I'll respond to that soon
[23:59:03 CEST] <kierank> michaelni: who has fate write access?
[23:59:10 CEST] <kierank> durandal_1707: do you have fate write access?
[00:00:00 CEST] --- Wed Oct 24 2018