[Ffmpeg-devel-irc] ffmpeg-devel.log.20170606
burek
burek021 at gmail.com
Wed Jun 7 03:05:02 EEST 2017
[00:11:13 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:88fabd801693: avcodec/aacsbr_fixed: Fix multiple runtime error: shift exponent 170 is too large for 32-bit type 'int'
[00:11:14 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:7d61e21953f9: avcodec/sbrdsp_fixed: fix runtime error: left shift of 1 by 31 places cannot be represented in type 'int'
[00:11:15 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:11e2bdd53e5a: avcodec/mlpdsp: Fix runtime error: signed integer overflow: -24419392 * 128 cannot be represented in type 'int'
[00:11:16 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:45c9fe61d932: avcodec/takdec: Fix runtime error: left shift of negative value -63
[00:11:17 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:0155d1c1cb61: avcodec/aac_defines: Fix: runtime error: left shift of negative value -2
[00:11:18 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:ea4b807c0a13: avcodec/takdec: Fix runtime error: signed integer overflow: 8192 * 524308 cannot be represented in type 'int'
[00:11:19 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:5ceb27b0eb7a: avcodec/vmnc: Check location before use
[00:11:20 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:8bc5e90bb218: avcodec/mpeg4videodec: Check for multiple VOL headers
[00:11:21 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:dd373b4027d6: avcodec/aacdec_fixed: Fix runtime error: shift exponent 34 is too large for 32-bit type 'int'
[00:11:22 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:32ac3f1b1f91: avcodec/mjpegdec: Fix runtime error: signed integer overflow: -32767 * 130560 cannot be represented in type 'int'
[00:11:23 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:5148395e86e6: avcodec/ivi_dsp: Fix multiple runtime error: left shift of negative value -71
[00:11:24 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:52e470097aee: avcodec/jpeglsdec: Check get_bits_left() before decoding a picture
[00:11:25 CEST] <cone-248> ffmpeg 03Max Justicz 07release/2.8:87c13e4e88ec: avcodec/sanm: Fix uninitialized reference frames
[00:11:26 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:75d32ff24d30: avcodec/jpeg2000dec: Check tile offsets
[00:11:27 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:345f29613583: avcodec/jpeg2000dec: Fix copy and paste error
[00:11:28 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:5529a6781709: avcodec/smc: Check remaining input
[00:11:29 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:134171355020: avcodec/aacdec_fixed: Fix runtime error: signed integer overflow: -2147483648 * -1 cannot be represented in type 'int'
[00:11:30 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:6025edc0c7c6: avutil/internal: Do not enable CHECKED with DEBUG
[00:11:31 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:fc449d70cf4d: avformat/mux: Fix copy an paste typo
[00:11:32 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:50fb6addc9ce: avcodec/ra144dec: Fix runtime error: left shift of negative value -17
[00:11:33 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:59436dc8f852: avcodec/mlpdec: Do not leave invalid values in matrix_out_ch[] on error
[00:11:34 CEST] <cone-248> ffmpeg 03Kevin Mark 07release/2.8:fde04ca71800: doc/filters: Clarify scale2ref example
[00:11:35 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:259582feaa7e: avcodec/ivi_dsp: Fix runtime error: left shift of negative value -2
[00:11:36 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:492099f9c49c: avcodec/sbrdsp_template: Fix: runtime error: signed integer overflow: 849815297 + 1315389781 cannot be represented in type 'int'
[00:11:37 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:8fbe1f634fdf: avcodec/libfdk-aacdec: Correct buffer_size parameter
[00:11:38 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:19556586d508: avcodec/wnv1: More strict buffer size check
[00:11:39 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:7b1d93ce9ca2: avcodec/aacdec_fixed: Fix multiple runtime error: shift exponent 127 is too large for 32-bit type 'int'
[00:11:40 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:0103b3ea924c: avcodec/jpeg2000dec: Check tile offsets more completely
[00:11:41 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:65f38d1285f3: avcodec/jpeg2000: Fix runtime error: signed integer overflow: 4185 + 2147483394 cannot be represented in type 'int'
[00:11:42 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:1d88ec2fb0e6: avcodec/snow: Fix runtime error: signed integer overflow: 1086573993 + 1086573994 cannot be represented in type 'int'
[00:11:43 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:5ee31596fa90: avcodec/aacdec_fixed: Fix runtime error: left shift of 1 by 31 places cannot be represented in type 'int'
[00:11:44 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:3fe7d4c403fd: avcodec/webp: Fixes null pointer dereference
[00:11:45 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:ba3bc229407a: avcodec/aac_defines: Add missing () to AAC_HALF_SUM() macro
[00:11:46 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:a7878fe2478e: avcodec/ra144: Fix runtime error: signed integer overflow: 11184810 * 404 cannot be represented in type 'int'
[00:11:47 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:33aed7bb767a: avcodec/ra144: Fix runtime error: signed integer overflow: -2449 * 1398101 cannot be represented in type 'int'
[00:11:48 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:b1777d92f9e2: avcodec/truemotion2: Fix runtime error: left shift of 1 by 31 places cannot be represented in type 'int'
[00:11:49 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:bbc6bfae376d: avcodec/truemotion2: Fix passing null pointer to memset()
[00:11:50 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:20b9b1fd77d4: avcodec/jpeg2000dec: Use ff_set_dimensions()
[00:11:51 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:8d1cd5fa08dd: avcodec/ansi: Fix frame memleak
[00:11:52 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:9a9247855633: avcodec/wavpack: Fix runtime error: signed integer overflow: 24 * -2147483648 cannot be represented in type 'int'
[00:11:53 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:3a56bbe001fc: avcodec/wavpack: Check float_shift
[00:11:54 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:464288cc5ea8: avcodec/acelp_pitch_delay: Fix runtime error: value 4.83233e+39 is outside the range of representable values of type 'float'
[00:11:55 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:bb7df2232873: avformat/avidec: Limit formats in gab2 to srt and ass/ssa
[00:11:56 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:b62a5c83fcc5: avcodec/cavsdec: Fix runtime error: signed integer overflow: 59 + 2147483600 cannot be represented in type 'int'
[00:11:57 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:9c0c749c334d: avcodec/pnm: Use ff_set_dimensions()
[00:11:58 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:368aa6aac74f: avcodec/ra144: Fixes runtime error: signed integer overflow: 7160 * 327138 cannot be represented in type 'int'
[00:11:59 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:041ad6d2bb44: avcodec/hevc_ps: Fix runtime error: signed integer overflow: 2147483628 + 256 cannot be represented in type 'int'
[00:12:00 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:653de8249dab: avcodec/cinepak: Check input packet size before frame reallocation
[00:12:01 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:d1390a8670db: avcodec/wavpack: Fix runtime error: signed integer overflow: 2013265955 - -134217694 cannot be represented in type 'int'
[00:12:02 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:6ebdffe73c42: avcodec/wavpack: Fix runtime error: shift exponent 32 is too large for 32-bit type 'int'
[00:12:03 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:ad600ec88633: avcodec/aacps: Fix runtime error: left shift of 1073741824 by 1 places cannot be represented in type 'INTFLOAT' (aka 'int')
[00:12:04 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:195b5b6cea07: avcodec/ac3dec_fixed: Fix runtime error: left shift of 419 by 23 places cannot be represented in type 'int'
[00:12:05 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:6636b35ef0d0: avcodec/pafvideo: Check packet size and frame code before ff_reget_buffer()
[00:12:06 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:268d84a74231: avcodec/hevc_ps: Fix runtime error: index 32 out of bounds for type 'uint8_t [32]'
[00:12:07 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:591a3d263247: avutil/softfloat: Fix sign error in and improve documentation of av_int2sf()
[00:12:08 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:a20e70835290: avcodec/qdrw: Fix null pointer dereference
[00:12:09 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:e447d3143faa: avformat/hls: Check local file extensions
[00:12:10 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:dbe771e50de5: avcodec/cavs: Fix runtime error: signed integer overflow: -12648062 * 256 cannot be represented in type 'int'
[00:12:11 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:05d7c10fd30f: avcodec/tiff: Avoid loosing allocated geotag values
[00:12:12 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:fb95f1d9d6d7: avcodec/mjpegdec: Check that reference frame matches the current frame
[00:12:13 CEST] <cone-248> ffmpeg 03Michael Niedermayer 07release/2.8:ffab459e4e49: Update for 2.8.12
[01:26:11 CEST] <cone-248> ffmpeg 03James Almer 07master:c1503cd95e20: fate: add fate-adts-id3v1-demux
[02:48:38 CEST] <Zeranoe> Should this https://ffmpeg.zeranoe.com/forum/viewtopic.php?f=13&p=12244#p12242 be addressed and removed? These builds appear to be including fdk-aac through delayed loading.
[03:40:39 CEST] <cone-248> ffmpeg 03James Almer 07n2.8.12:HEAD: fate: add fate-adts-id3v1-demux
[06:37:51 CEST] <nevcairiel> that doesnt exactly fix the license situation, so they would be producing non-free builds
[08:06:32 CEST] <kierank> Classic unfinished drive by commit:https://blog.savoirfairelinux.com/en-ca/2017/ffmpegs-technological-challenges-to-further-shape-tr-03-standards-for-broadcasters/?noredirect=en_CA
[08:54:39 CEST] <kevmark> 549acc9995 and 01718dc0df?
[08:57:38 CEST] <kevmark> Are you referring to them adding unfinished code they have no apparent intention of maintaining?
[09:08:10 CEST] <mateo`> tmm1: hello, i'll start testing your patch tomorrow (i'm currently on holiday)
[10:08:36 CEST] <kierank> kevmark: exactly
[10:09:23 CEST] <kevmark> kierank, I'm sure it'll get a CVE later
[10:09:33 CEST] <kierank> probably
[12:23:18 CEST] <cone-969> ffmpeg 03Stefano Sabatini 07master:dc70ea8c193a: lavc/mpegvideo_enc: allow low_delay for non MPEG2 codecs depending on strict_std_compliance
[14:12:53 CEST] <cone-969> ffmpeg 03Ilia Valiakhmetov 07master:73d9a9a6af5d: libavcodec/vp9: ipred_dl_32x32_16 avx2 implementation
[14:13:19 CEST] <BBB> atomnuker: ^^ done
[14:14:10 CEST] <atomnuker> nice
[15:18:59 CEST] <J_Darnley> BBB: Thank you so very much.
[15:19:09 CEST] <BBB> yw dude
[15:27:23 CEST] <jamrial> tdjones, atomnuker: nice seeing vorbisenc getting some love :)
[15:42:49 CEST] <tdjones> Yeah, it's my GSoC project, so hopefully it gets closer to a usable state over the next several weeks
[15:48:14 CEST] <atomnuker> jamrial: just found out how to pipe anything from decoders to filters in under 20 lines - https://pars.ee/temp/comp_l_ff.webm
[15:49:12 CEST] <atomnuker> this eliminates any guesswork when optimizing transient detection or bit distribution
[15:50:04 CEST] <atomnuker> its not hacky either, its just avframes with side data
[15:50:48 CEST] <atomnuker> if I only had the idea to do this earlier
[15:56:20 CEST] <atomnuker> I bet no one has ever thought of doing that, this is like discovering steam power
[15:57:14 CEST] <atomnuker> and with just a single command I can find that my transient detector is too sensitive to high frequencies so I need to window my per-band transient contribution
[15:58:26 CEST] <wm4> doing what?
[16:00:06 CEST] <atomnuker> rolling fft graph with coloring depending on random codec data
[16:00:21 CEST] <atomnuker> (red == transient frame in this case)
[16:45:06 CEST] <Compn> atomnuker : make a howto on trac ?
[16:55:39 CEST] <cone-969> ffmpeg 03Michael Niedermayer 07master:e4efd41b83e7: avcodec/takdec: Fix multiple runtime error: signed integer overflow: 637072 * 4096 cannot be represented in type 'int'
[16:55:40 CEST] <cone-969> ffmpeg 03Michael Niedermayer 07master:c4360559ee2a: avcodec/pafvideo: Fix assertion failure
[16:55:41 CEST] <cone-969> ffmpeg 03Michael Niedermayer 07master:18bca25adbae: avcodec/mpeg4videodec: Fix runtime error: signed integer overflow: 53098 * 40448 cannot be represented in type 'int'
[17:29:36 CEST] <cone-969> ffmpeg 03Vittorio Giovara 07master:d9909b11d9b0: vf_colorspace: Add support for gbr color space
[17:36:52 CEST] <kierank> BBB: !!!
[17:37:03 CEST] <BBB> wut?
[17:37:12 CEST] <kierank> "This makes the output of dct-test exact:"
[17:37:28 CEST] <kierank> nice work
[17:37:44 CEST] <BBB> \o/
[18:19:24 CEST] <masikh> Hello dear all, I found (and fixed) 2 devision by zero errors in ffserver. I'd like to give back the code to the community, but hell I don't know how to do a git mail patch and alike.
[18:19:36 CEST] <masikh> maybe someone might like to help me on this?
[18:21:25 CEST] <masikh> ffserver.c: Line: 469 -> change to -> if (cur_time == drd->time1 || cur_time == 0 || drd->time1 == 0)
[18:22:06 CEST] <tmm1> there is some documentation on https://ffmpeg.org/developer.html#Submitting-patches
[18:23:28 CEST] <masikh> Yes, I read, I didn't understoot
[18:23:31 CEST] <masikh> hence I am here!
[18:23:34 CEST] <masikh> hello!
[18:24:37 CEST] <tmm1> do you know how to use git?
[18:24:37 CEST] <masikh> ffserver.c: Line: 1910 (around) change to ->
[18:24:42 CEST] <masikh> if (st->time_base.num != 0) {
[18:24:42 CEST] <masikh> snprintf(parameters, sizeof(parameters),
[18:24:42 CEST] <masikh> "%dx%d, q=%d-%d, fps=%d", st->codecpar->width,
[18:24:42 CEST] <masikh> st->codecpar->height, st->codec->qmin, st->codec->qmax,
[18:24:42 CEST] <masikh> st->time_base.den / st->time_base.num);
[18:24:43 CEST] <masikh> } else {
[18:24:43 CEST] <masikh> snprintf(parameters, sizeof(parameters),
[18:24:44 CEST] <masikh> "%dx%d, q=%d-%d, fps=0", st->codecpar->width,
[18:24:44 CEST] <masikh> st->codecpar->height, st->codec->qmin, st->codec->qmax);
[18:24:45 CEST] <masikh> }
[18:24:57 CEST] <masikh> that's it
[18:25:23 CEST] <masikh> So maybe tmm1, you DO understand how to make these patches?
[18:25:55 CEST] <masikh> Cause I don't, but I'd like these fixes being given to the community
[18:26:21 CEST] <masikh> So if you could help me that would be splendid!
[18:26:47 CEST] <tmm1> i am trying to help you
[18:26:52 CEST] <masikh> Cool!
[18:27:10 CEST] <tmm1> you need to make a git commit first. visit https://github.com/FFmpeg/FFmpeg/blob/master/ffserver.c and click the pencil icon
[18:27:14 CEST] <tmm1> and make your changes there
[18:27:28 CEST] <tmm1> pasting coding into IRC is considered spam
[18:29:26 CEST] <masikh> ok, working on it 1 moment please
[18:32:49 CEST] <BBB> kierank: dont forget Im equally happy someone finally wrote a real SSE2 version of the idct, or ported it from inlien assembly to real world stuff
[18:32:58 CEST] <masikh> Ok, done and now?
[18:32:59 CEST] <cone-969> ffmpeg 03Michael Niedermayer 07master:850cbd496f8d: avformat/utils: Slightly un-clutter code in determinable_frame_size() by using a switch
[18:33:00 CEST] <cone-969> ffmpeg 03Kevin Mark 07master:482566ccc3fd: libavutil/eval: Add round function to expression parser
[18:33:53 CEST] <BBB> $ grep volatile ~/Projects/ffmpeg/libavcodec/x86/*.c|cut -d: -f1|uniq|wc -l
[18:33:53 CEST] <BBB> 19
[18:33:56 CEST] <tmm1> masikh: now you can email the link to your commit to ffmpeg-devel at ffmpeg.org
[18:33:57 CEST] <BBB> kierank: ^^
[18:34:36 CEST] <kierank> If there's anything we care about then I guess we can but I doubt it
[18:34:53 CEST] <kierank> Need to move on to other stuff
[18:35:18 CEST] <masikh> Do you mean this link? -> https://github.com/masikh/FFmpeg/commit/50326ef5683a85eef73b3db628c6dbacc51a8856
[18:36:33 CEST] <tmm1> masikh: yes, or you can add .patch at the end like https://github.com/masikh/FFmpeg/commit/50326ef568.patch
[18:36:49 CEST] <masikh> Ok, will do
[18:36:57 CEST] <masikh> Thank you very much for the help!
[18:42:17 CEST] <masikh> @tmm1: Mail away! And now we wait, I guess!
[18:58:13 CEST] <cone-969> ffmpeg 03Tyler Jones 07master:34c52005605d: vorbisenc: Fix memory leak on errors
[21:11:25 CEST] <BBB> kierank: understood :-p
[21:11:39 CEST] Action: kierank looks
[21:12:26 CEST] <kierank> mpeg-2 dequant I guess
[21:12:36 CEST] <kierank> but that might be one of those riced dequants
[21:12:39 CEST] <kierank> which really need removing
[21:13:30 CEST] <kierank> if (!(s->avctx->flags & AV_CODEC_FLAG_BITEXACT))
[21:13:30 CEST] <kierank> s->dct_unquantize_mpeg2_intra = dct_unquantize_mpeg2_intra_mmx;
[21:13:37 CEST] <kierank> yeah non bitexact mpeg2 is really pointless in this day and age
[21:16:54 CEST] <iive> i'm more puzzled, why it is not bitexact
[21:34:37 CEST] <jamrial> iive: probably doing some tricks that makes calculations slightly off for the sake of speed
[21:43:59 CEST] <iive> there are 2 C functions, the second one is called bitexact. the difference is that it does the odd/even thing from the standard.
[22:01:52 CEST] <tmatth> out of curiousity, any reason "round(x)" isn't implemented as an expression? https://www.ffmpeg.org/ffmpeg-utils.html#Expression-Evaluation
[22:04:10 CEST] <tmatth> nevermind, apparently that got added today (!)
[22:18:55 CEST] <kevmark> tmatth, ;)
[00:00:00 CEST] --- Wed Jun 7 2017
More information about the Ffmpeg-devel-irc
mailing list