[Ffmpeg-devel-irc] ffmpeg-devel.log.20151122

burek burek021 at gmail.com
Mon Nov 23 02:05:04 CET 2015


[00:55:34 CET] <rcombs> https://gist.github.com/80b377c2394085c13879 ~android~
[01:12:01 CET] <JEEB> rcombs: NDK?
[01:12:06 CET] <rcombs> yup
[01:12:07 CET] <JEEB> been a while since I did an ARM build
[01:12:16 CET] <rcombs> I'm building on the device
[01:12:30 CET] <rcombs> turns out this is tricky
[01:12:49 CET] <rcombs> apparently it would be easier if I was doing regular ARM, but this is aarch64
[01:13:13 CET] <rcombs> and gcc has a bunch of config stuff setup properly for 32-bit ARM but not aarch64
[01:13:45 CET] <rcombs> remind me, do shared libs usually have a `_start`?
[01:15:10 CET] <JEEB> I don't have any ARM libs on hand, but at least my zlib on this box doesn't seem to have such
[01:15:43 CET] <rcombs> gcc/ld seems to be expecting a _start in the crt even when linking shared libs
[01:16:03 CET] <rcombs> I'm wondering if it's confused about how shared libs work in general
[01:16:36 CET] <rcombs> thus why `-lavutil -lavutil` (why does this happen? I dunno, makefiles are weird) results in complaints about every symbol in libavutil showing up twice
[01:17:01 CET] <rcombs> which I could just silence with an ld switch but it'd probably just mask the underlying issue
[01:17:28 CET] <rcombs> also, I tried to build ffmpeg static and got complaints about textrels
[01:17:36 CET] <rcombs> from the dynamic linker
[01:19:12 CET] <rcombs> meanwhile, if you haven't seen this already: https://lkml.org/lkml/2015/11/18/9
[01:23:36 CET] <atomnuker> yeah, saw it
[01:23:40 CET] <atomnuker> hilarious
[01:26:46 CET] <atomnuker> " The history of "approved standards" is a proof to not to trust to so called "Security Authorities", "Experts" or "Cryptography Gods""
[01:26:57 CET] <atomnuker> https://github.com/ikizir/HohhaDynamicXOR/tree/master
[01:29:41 CET] <Gramner> not sure if troll or delusional
[01:46:37 CET] <rcombs> delusional troll?
[01:47:59 CET] <JEEB> unbreakable crypto, eh?
[01:48:23 CET] <JEEB> "everyone's proprietary crypto is unbreakable, get in the line sonny boy"
[01:51:03 CET] <cone-013> ffmpeg 03Carl Eugen Hoyos 07master:f5074dd39ca3: lavc/rscc: Fix colourspace for codec_tag RSCC.
[01:51:03 CET] <cone-013> ffmpeg 03Carl Eugen Hoyos 07master:ced39dc5ed3a: lavc/opusdec: Fix a memleak when reading invalid files.
[12:16:24 CET] <Fyr> another bug over 1200 not to be ever fixed.
[12:16:25 CET] <Fyr> ='(
[13:00:56 CET] <Daemon404> what a terrifying attachement name
[15:15:59 CET] <Daemon404> i geuss ill merge some more
[15:16:11 CET] Action: Daemon404 wonder if nevcairiel is preying on ocd
[15:37:34 CET] <cone-751> ffmpeg 03Martin Storsjö 07master:92d107a17127: xtea: Add functions for little endian mode
[15:37:34 CET] <cone-751> ffmpeg 03Martin Storsjö 07master:588b6215b4c7: rtmpcrypt: Do the xtea decryption in little endian mode
[15:37:34 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:e12f40367861: Merge commit '588b6215b4c74945994eb9636b0699028c069ed2'
[15:39:36 CET] <cone-751> ffmpeg 03Martin Storsjö 07master:62b4a6f1b9aa: rtmpcrypt: Provide the xtea keys in little endian format for consistency
[15:39:37 CET] <cone-751> ffmpeg 03Michael Niedermayer 07master:2fb6e1e754f3: segafilm: Fix current_sample after seeking and avio_seek return type
[15:39:38 CET] <cone-751> ffmpeg 03Andreas Cadhalpun 07master:0a8bff788b0a: dds: disable palette flag for compressed images
[15:39:39 CET] <cone-751> ffmpeg 03Andreas Cadhalpun 07master:9fd2bf09dbc6: hqx: correct type and size check of info_offset
[15:39:40 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:5b4be0c03947: Merge commit '62b4a6f1b9aa83d56701af226adda98faa5ede09'
[15:39:41 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:3c110b009b8e: Merge commit '9fd2bf09dbc630484d9e88a1d27f7e8508b70a2c'
[15:40:33 CET] <Daemon404> michaelni, i am not sure if i should merge 4bb1070c154e49d35805fbcdac9c9e92f702ef96
[15:40:41 CET] <Daemon404> it's a ffv1 change, maybe you should look
[15:54:39 CET] <michaelni> replacing 0/1/2 by names is fine the rest seems odd
[15:58:06 CET] <Daemon404> ogg how?
[15:58:07 CET] <Daemon404> er
[15:58:08 CET] <Daemon404> odd*
[16:10:59 CET] <michaelni> changing 012 to names is cosmetic, the other changes are not cosmetic. removing exporting the coder_type might be ok assuming nothing uses it, but i dont know if nothig uses it.  The other removed check looks like it was written with a coder_type == -1 for auto, i dont know if the removial is better than adding a auto=-1 value
[16:12:05 CET] <michaelni> its up to you if you want to merge it completely, split it or merge just parts
[16:13:25 CET] <Daemon404> michaelni, it's documenetd as encoder-only
[16:13:32 CET] <Daemon404> technically its use was never valid
[16:13:55 CET] <Daemon404> i cant see anything using it... maybe mplayer
[16:14:18 CET] <michaelni> yes, you are correct, technically its use was never valid
[16:14:34 CET] <Daemon404> ill grep the usual players and google
[16:15:13 CET] <michaelni> i see nothing in mplayer
[16:15:43 CET] <Daemon404> me neither
[16:40:50 CET] <cone-751> ffmpeg 03Michael Niedermayer 07release/2.6:48d388b03336: Update for 2.6.5
[17:44:36 CET] <Daemon404> michaelni, i cant figure out why this check was remove: av_log(avctx, AV_LOG_ERROR, "bits_per_raw_sample of more than 8 needs -coder 1 currently\n");
[17:44:48 CET] <Daemon404> the commit message, and ML 'review' has no details
[17:44:58 CET] <Daemon404> im erring on leaving it, but i can;t decide.
[17:48:55 CET] <Daemon404> ah there it is... ok nvm i figured it out
[17:49:00 CET] <Daemon404> the define happened to be 0.
[18:09:07 CET] <cone-751> ffmpeg 03Vittorio Giovara 07master:4bb1070c154e: ffv1: Explicitly name the coder type
[18:09:08 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:bba2488f0749: Merge commit '4bb1070c154e49d35805fbcdac9c9e92f702ef96'
[18:19:55 CET] <kruzin> anyone around?
[18:20:09 CET] <Daemon404> maybe
[18:21:12 CET] <kruzin> @Daemon404, would you possibly be able to guide me with an issue I'm facing with ffmpeg on android?
[18:21:26 CET] <kruzin> I believe its a small issue, maybe I'm missing a step or two
[18:25:18 CET] <Daemon404> ?
[18:30:44 CET] <BBB> kruzin: dont ask to ask, just ask. and user questions should go to #ffmpeg, not here
[18:30:53 CET] <kruzin> Heres my issue to be exact, everything works just fine in a new project I create, even on 64bit processors,  but when I try to run it in my existing project, somehow, the ffmpeg fails to execute
[18:30:54 CET] <kruzin> oh
[18:30:59 CET] <BBB> kruzin: this channel is for ffmpeg development, that is, development of ffmpeg, not development with ffmpeg
[18:31:05 CET] <kruzin> sorry, my bad, they pointed it here.. sorry :/
[18:31:13 CET] <BBB> no prob :)
[18:31:14 CET] <wm4> who did?
[18:31:21 CET] <JEEB> the Fyr guy
[18:31:48 CET] <kruzin> my bad.. sorry once again :D
[18:35:29 CET] <Daemon404> wut
[18:35:30 CET] <Daemon404> remote: -Deny-          ERROR: update contains a merge.
[18:35:30 CET] <Daemon404> remote: deny
[18:35:30 CET] <Daemon404> remote: error: hook declined to update refs/heads/master
[18:35:33 CET] <Daemon404> this is a new one
[18:35:39 CET] <Daemon404> of coruse it contains a merge, im merging libav!
[18:35:49 CET] <Daemon404> nevcairiel || michaelni ?
[18:38:01 CET] <Daemon404> ah... i wonder...
[18:39:56 CET] <Daemon404> i have signed-off-by in one commit instead of merged-by
[18:40:31 CET] <cone-751> ffmpeg 03Vittorio Giovara 07master:79d89cf2f4b6: flacenc: Clamp user-supplied min/max prediction orders
[18:40:32 CET] <cone-751> ffmpeg 03Vittorio Giovara 07master:eda183287489: travis: Enable OSX integration
[18:40:33 CET] <cone-751> ffmpeg 03Luca Barbato 07master:5c30ae1a09b6: dvdsubdec: Validate the RLE offsets
[18:40:34 CET] <cone-751> ffmpeg 03Anton Khirnov 07master:21ab6fd01b91: lavf: document that passing valid timestamps to muxers is now mandatory
[18:40:35 CET] <cone-751> ffmpeg 03John Stebbins 07master:7f4ec4364bc4: avformat: expose av_stream_new_side_data helper
[18:40:36 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:5ebaf1e1cbee: Merge commit '79d89cf2f4b62eeb653fd8139041c87e75f7ca65'
[18:40:37 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:df8e24be5da2: Merge commit 'eda183287489b2c705843aa373a19c4e46fb2fec'
[18:40:38 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:cde68661bfa5: Merge commit '5c30ae1a09b66179e16694f6137658023ed1fef3'
[18:40:39 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:2bf93dfd9654: Merge commit '21ab6fd01b9107d7171fc1a77c2bfacff845a97a'
[18:40:40 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:ad317c94c5fa: Merge commit '7f4ec4364bc4a73036660c1c6a3c4801db524e9e'
[18:40:41 CET] <Daemon404> yes that was it.
[18:40:53 CET] <Daemon404> message could have been a bit clearer
[18:40:54 CET] <JEEB> :)
[18:42:17 CET] <cone-751> ffmpeg 03John Stebbins 07master:79ae1e630b47: avcodec: Define side data type for fallback track
[18:42:18 CET] <cone-751> ffmpeg 03John Stebbins 07master:ca489564ae15: movenc: add fallback audio track tref support
[18:42:19 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:fe20e34b139f: Merge commit '79ae1e630b476889c251fc905687a3831b43ab5e'
[18:42:20 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:64db3639af7b: Merge commit 'ca489564ae150bc41764f175b88151d883e69ae0'
[18:49:12 CET] <Daemon404> q
[18:49:25 CET] <JEEB> !
[19:01:11 CET] <cone-751> ffmpeg 03Anton Khirnov 07master:d80f0a4ad634: mpevideo_enc: disallow multiple slices for h261 and flv
[19:01:12 CET] <cone-751> ffmpeg 03Anton Khirnov 07master:f6c94457b44f: mpegvideo_enc: enable rtp_mode when multiple slices are used
[19:01:13 CET] <cone-751> ffmpeg 03Anton Khirnov 07master:4d8f536b5354: qsvenc: print the actual video parameters used by MSDK
[19:01:14 CET] <cone-751> ffmpeg 03Anton Khirnov 07master:3edac01f79a0: qsvenc: fix setting maxrate for VBR
[19:01:15 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:819a99ef21e9: Merge commit 'd80f0a4ad634b3949b91b85f21fd608c0cddeef7'
[19:01:16 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:09f9ef8536ef: Merge commit 'f6c94457b44f41d900cd0991857f54e1f0ccedd6'
[19:01:17 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:98540a2c7d79: Merge commit '4d8f536b535487063a08609636e712ad86d2ad54'
[19:01:18 CET] <cone-751> ffmpeg 03Derek Buitenhuis 07master:4613b4df31f6: Merge commit '3edac01f79a08635d8d2c08e9410651d9a330d61'
[19:01:34 CET] <Daemon404> drawing ever closer to parity...
[19:03:16 CET] Action: Daemon404 wonders if "Ivan Uskov" is on irc
[19:29:46 CET] <cone-751> ffmpeg 03Michael Niedermayer 07master:764900d6458a: avcodec/dvdsubdec: Fix types for offset1/2 in ff_dlog()
[19:34:21 CET] <Daemon404> damn.
[20:34:20 CET] <RiCON> oh, rtmpcrypt fixes have been merged, cool
[20:45:11 CET] <BtbN> Anyone happens to have a somewhat direct link where i can get the mfx windows sdk? The Intel Website is like some kind of modern maze
[22:15:15 CET] <cone-751> ffmpeg 03Ganesh Ajjanagadde 07master:7324af651f60: avformat/mov: use hypot()
[22:24:50 CET] <cone-751> ffmpeg 03Ganesh Ajjanagadde 07master:18b4016431d9: avfilter/vf_fspp: use M_SQRT1_2, M_SQRT2
[22:24:51 CET] <cone-751> ffmpeg 03Ganesh Ajjanagadde 07master:94565f379196: avcodec/ac3: use M_SQRT1_2, M_SQRT2
[22:24:52 CET] <cone-751> ffmpeg 03Ganesh Ajjanagadde 07master:824ba897bdd1: avcodec/cos_tablegen: use M_PI instead of actual literal
[22:24:53 CET] <cone-751> ffmpeg 03Ganesh Ajjanagadde 07master:ccf3c694032b: avcodec/dct32_template: use M_SQRT1_2 instead of actual literal
[22:24:54 CET] <cone-751> ffmpeg 03Ganesh Ajjanagadde 07master:61a1ca13ea9f: avcodec/ppc/fdctdsp: use M_SQRT2 instead of ad-hoc SQRT2
[22:57:59 CET] <cone-751> ffmpeg 03Ganesh Ajjanagadde 07master:82c5f3178930: avcodec/dct32_template: fix typo
[23:02:41 CET] <cone-751> ffmpeg 03Ganesh Ajjanagadde 07master:90ccfdb65bc0: Revert "avcodec/dct32_template: fix typo"
[23:02:52 CET] <ubitux> damn derp madness.
[23:24:40 CET] <Compn> The expansion here is of a ludicrous
[23:24:40 CET] <Compn> length and gives no value whatsoever.
[23:49:42 CET] <kierank> michaelni: that's a bug right?
[23:49:43 CET] <kierank> https://github.com/FFmpeg/FFmpeg/blob/release/2.7/libavcodec/error_resilience.c#L1222
[23:49:46 CET] <kierank> it assumes 4:2:0?
[00:00:00 CET] --- Mon Nov 23 2015


More information about the Ffmpeg-devel-irc mailing list