[Ffmpeg-devel-irc] ffmpeg-devel.log.20150217

[burek]

[00:16] <cone-225> ffmpeg.git 03Michael Niedermayer 07master:cb9d40451cb2: avcodec/ac3: add () to protect the macro arguments
[00:16] <cone-225> ffmpeg.git 03Michael Niedermayer 07master:e3f71423068b: avcodec/vc1dsp: add () to protect the arguments of the op* macros
[00:16] <cone-225> ffmpeg.git 03Michael Niedermayer 07master:61eb602d27af: avutil/log: add () to protect the argument of AV_LOG_C()
[00:16] <cone-225> ffmpeg.git 03Michael Niedermayer 07master:527f1d22a1a4: avformat/tedcaptionsdec: add () to protect the argument of ERR_CODE()
[00:30] <cone-225> ffmpeg.git 03Michael Niedermayer 07master:9cde41cdb511: avutil/version: Add () to protect the AV_VERSION_INT() arguments
[00:30] <cone-225> ffmpeg.git 03Michael Niedermayer 07master:cdee0c9d5fed: avutil/murmur3: Add () to protect the ROT() arguments
[00:31] <cone-225> ffmpeg.git 03Michael Niedermayer 07master:f8607cfb0a82: avutil/intmath: Add () to protect the ff_log2() argument
[00:31] <cone-225> ffmpeg.git 03Michael Niedermayer 07master:06309fc704dc: avutil/attributes: Add () to protect the AV_GCC_VERSION_AT_LEAST() arguments
[00:31] <cone-225> ffmpeg.git 03Michael Niedermayer 07master:52d1929000c3: avutil/aes: Add () to protect the ROT() arguments
[00:58] <cone-225> ffmpeg.git 03Michael Niedermayer 07master:0cb95f908298: swresample/resample_template: Add () to protect the arguments of the OUT() macro
[00:58] <cone-225> ffmpeg.git 03Michael Niedermayer 07master:f15c734be165: avfilter/vf_phase: Add () to protect DIFF()s arguments
[00:58] <cone-225> ffmpeg.git 03Michael Niedermayer 07master:2bae7b337030: avfilter/vf_fieldmatch: Add () to protect the arguments of the HAS_FF_AROUND() macro
[04:08] <jamrial> porting OpenHEVC's idct x86 intrinsics to yasm (or generally optimize hevc with assembly, arm/aarch64 included), and adding AESNI support to our aes module could be two new GSoC tasks
[04:08] <jamrial> kierank would surely approve of the latter :P
[04:21] <kierank> :)
[06:02] <Daemon404> aesni should be pretty straighrforward no?
[07:26] <sekon> Hello i want a dummy ffmpeg opt of type OPT_STRING
[07:27] <sekon> it will not be used anywhere in ffmpeg .. but it should not make ffmpeg crash
[07:27] <sekon> *crash as in segfault in av_free
[07:28] <sekon> also could i suggest adding some documetation regarding what HAS_ARG, OPT_EXIT etc actually mean in cmdutils_common_opts.h
[07:28] <sekon> ?
[07:29] <sekon> documentation/comments 
[08:07] <ubitux> string needs to be dupped
[09:15] <j-b> hello
[12:48] <cone-408> ffmpeg.git 03Christophe Gisquet 07master:d9293c776ed9: x86inc: Correctly warn on use of SSE2 instructions in SSE functions
[13:19] <cone-408> ffmpeg.git 03Christophe Gisquet 07master:bc8e044d8c32: x86/doc/Makefile: DBG=1 to preprocess external asm
[13:46] <wm4> wow claws-mail is such a piece of shit
[14:10] <compn> wm4 : compared to what? whats better ?
[14:10] <compn> sylpheed claws? :P
[14:13] <arwa> How do I set min and max values allowed by using AVExpr?
[14:20] <michaelni> good question, probably best by simply checking or cliping with av_clipd() the "double" value from evaluating the expression
[15:17] <cone-408> ffmpeg.git 03Carl Eugen Hoyos 07master:a9dc94c39c7e: Force linking against user32 if dxva2 is enabled.
[15:19] <arwa> Okay, thanks :)
[15:20] <j-b> Why do you need user32?
[15:21] <arwa> No, It was for michaelni.
[15:21] <j-b> I'm asking Carl
[15:24] <arwa> Ohhh!!
[15:30] <compn> j-b : i'm guessing because vfwcap provided a header file , and when thats missing/disabled need to link against user32
[15:30] <compn> but thats just my guessssss
[15:30] Action: compn afk
[15:34] <j-b> vfwcap for dxva?
[15:34] <j-b> seriously?
[15:45] <j-b> -lole32 -lshlwapi -luuid should be enough
[15:45] <j-b> and shlwapi, I'm not sure
[15:47] <nevcairiel> one function needs user32 apparently
[15:49] <j-b> sure, but which one?
[16:08] <cone-408> ffmpeg.git 03Christophe Gisquet 07master:398f53191560: x86: hevc_mc: fewer xmm regs used in epel h/v
[16:09] <nevcairiel> j-b: GetShellWindow, i guess
[16:10] <j-b> ah, yes. But it's used in DxVA?
[16:10] <nevcairiel> not the hwaccel, only in ffmpeg_dxva2.c to create the d3d device
[16:11] <nevcairiel> (which needs a window ref)
[16:15] <wm4> of course ffmpeg.c deps are included in the libs
[16:15] <Daemon404> of course.
[16:16] <Daemon404> because carl.
[16:16] <nevcairiel> i think that part is only for ffmpeg itself
[16:16] <nevcairiel> it does say ffmpeg_libs
[16:17] <Daemon404> o ok
[16:53] <gerion> I just found out, that x265 could not be controlled by -crf but with -x265-params crf etc.. Is there a technically/design reason for this or is it just not implemented?
[16:53] <Daemon404> because -crf is a private option for x264
[16:53] <Daemon404> which shouldnt exist IMO
[16:54] <Daemon404> i thought about adding a similar thing for x265 but it's still a hack
[16:54] <gerion> with "shouldnt exist" you mean someway globally?
[16:54] <Daemon404> i dont think the x264 wrapper should have exposed it in such a way
[16:54] <Daemon404> but that's in the past, can't change it...
[16:55] <Daemon404> you're not the first to ask... i guess i should add it
[16:55] <Daemon404> begrudgingly.
[16:56] <gerion> ;) ok, would be more consistent
[16:56] <gerion> (aka intuitive)
[16:56] <Daemon404> it's not intuitive
[16:57] <Daemon404> consistent yes
[16:57] <gerion> if you know how to work with x264 you kind of automatically know how to work with x265
[16:57] <gerion> maybe intuitive is an uncorrect word for this
[16:58] <gerion> atm you have to read documention to control x265 even if you can control x264 through ffmpeg
[16:59] <Daemon404> that's probably a good thing.
[16:59] <gerion> ;)
[16:59] <Daemon404> because everyone uses x264 wrong.
[16:59] <Daemon404> also hevc is not the same as avc
[16:59] <Daemon404> you have to read its options anyway
[17:00] <Daemon404> crf should be fine, sicne ratecontrol is generally always available via high level options
[17:01] <Daemon404> ill send a patch shortly
[17:01] <gerion> thank you
[17:02] <gerion> why everyone uses x264 wrong?
[17:02] <Daemon404> because they dont understand the options.
[17:02] <gerion> cause of twopass/crf/specific bitrate confusion?
[17:02] <Daemon404> nah not ratecontrol
[17:02] <Daemon404> but stuff like ipratio/pbratio, etc
[17:05] <gerion> is this often set?
[17:05] <Daemon404> cargo culted stuff tends to be
[17:05] <Daemon404> from random ffmpeg "tutorials"
[17:06] <Daemon404> (or x264)
[17:06] <Daemon404> anyway, i argue one should be reading the docs for encoder specific options.
[17:07] <gerion> yeah ok, but if x265 and x264 behave in similar way regarding a specific option they could be mapped together
[17:08] <Daemon404> funny you menton that
[17:08] <Daemon404> because x264 and x265 have different 'scales' for crf.
[17:08] <Daemon404> so technically they don'y.
[17:08] <Daemon404> dont*
[17:08] <thardin> whyyy
[17:08] <Daemon404> thardin, it's the nature of crf
[17:08] <Daemon404> even 10bit x264 and 8bit x264 differ
[17:09] <Daemon404> it's an arbitrary scale
[17:09] <thardin> ah
[17:09] <Daemon404> you'd all know this if you read the docs!
[17:09] Action: Daemon404 runs
[17:09] <gerion> yay i noticed that, but x265 says: crf = quality based vbr the same x264 says
[17:10] <gerion> so the effect is quite similar
[17:10] <Daemon404> yeah but the scale is arbitrary
[17:10] <Daemon404> ratecontrol is more or less consistent between encoders. hence ill expose it as such
[17:11] <BtbN> Does ffmpeg have some code to write h264 nal bitstream?
[17:13] <gerion> what do you mean with ratecontrol? the normal bitrate settings? you have to be very experienced to know which bitrate of a specific codec is necessary to produce the wanted result
[17:13] <thardin> gerion: same goes for crf
[17:14] <Daemon404> i mean interface-wise gerion 
[17:14] <Daemon404> BtbN, what do you mean
[17:14] <BtbN> libva doesn't generate the sps/pps for you. You have to do that yourself...
[17:14] <Daemon404> eh
[17:14] <Daemon404> oh
[17:14] <Daemon404> libva.
[17:14] <Daemon404> well.
[17:15] <Daemon404> the way libx264 and libx265 write elementary stream (nals) is that the encoders both have options to produce sps and pps
[17:15] <Daemon404> and lavf just passes it through
[17:15] <Daemon404> in rawenc.c iirc
[17:15] <Daemon404> so you'd have to do it in the libavcodec end
[17:15] <Daemon404> also wtf kind of crap lib makes you generate sps/pps yourself? :/
[17:16] <gerion> thardin: more less, crf 20 or so with x264 generate more less similar results (not depending on the inpu), that the equivalent on x265 is not 20 but 18 or so (don't know the real value) is kind of strange
[17:16] <BtbN> It's next level stupid that libva doesn't do that. Specialy as it already has bit stream generation code in place. It also does generate large parts of the bitstream.
[17:16] <BtbN> But not the SPS/PPS
[17:16] <BtbN> I have no idea what the point if that is.
[17:16] <Daemon404> you would think the encoder would be best suited to generate it...
[17:16] <BtbN> You also have to pass the sps/pps bitstream into libva then, it has a special buffer for that.
[17:16] <Daemon404> wut
[17:16] <BtbN> Yes
[17:17] <Daemon404> thats retarded
[17:17] <BtbN> No wonder intel wrote its own wrapper for that lib
[17:17] <Daemon404> and intel's wrapper isn't good either from what i hear
[17:17] <BtbN> It's of questionable quality and future
[17:18] <Daemon404> just like all their foss
[17:18] <BtbN> I realy wonder who designed that libva encode api. Either i'm missing something, or this stuff is next level broken.
[17:19] <Daemon404> it sounds like it's literally a thin wrapper over direct hw calls
[17:19] <BtbN> I looked into the driver, and it does have h264 nal generation
[17:19] <Daemon404> oh.
[17:19] <BtbN> But not for the sps/pps
[17:19] <BtbN> like, wtf
[17:20] <BtbN> B frames also look like they are a lot of pain.
[17:20] <Daemon404> well x264 and 5 just pass me nals back
[17:20] <Daemon404> and it's 0 effort. so, sucks to be you?
[17:20] Action: Daemon404 runs
[17:20] <BtbN> so does nvenc
[17:21] <cone-408> ffmpeg.git 03Clément BSsch 07master:a00bab347518: avfilter/paletteuse: raise cache size from 64k to 512k
[17:21] <BtbN> nvenc is actualy a realy nice api. Does precisely what it's expected to do.
[17:25] <iive> BtbN: there are some bistream filters, e.g. annexb and stuff, would any of these help you?
[17:26] <BtbN> Don't think so, i have to generate the entire thing from scratch
[17:26] <BtbN> I already have code that does that, but it's like 2000 lines
[17:27] <BtbN> https://github.com/BtbN/vlc-vaapi-enc/blob/master/vlc-h264-vaapi-enc.c#L622 well, not 2000, but still annoyingly long
[17:36] <iive> :|
[17:41] <wm4> haha
[17:42] <wm4> obviously you should use gstreamer
[17:43] <BtbN> btw., in gstreamer, they internaly write a wrapper around libva
[17:43] <BtbN> so the actual code doesn't have to deal with the worst of it
[17:44] <wm4> kind of like ffmpeg has hwaccel for decoding?
[17:44] <BtbN> No, more like libyami-light
[19:43] <cone-408> ffmpeg.git 03Xiaohan Wang 07release/2.2:a9602c6cfbe6: matroskadec: Fix read-after-free in matroska_read_seek()
[19:43] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:e5b878b37f64: Merge commit 'a9602c6cfbe6fa06ff97ad01c0ffa9ad5ccff30f' into release/2.2
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:56976346d989: avformat/utils: Fix number suffixes in tb_unreliable()
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:e61a7c9b4517: swresample/dither: Cleanup number suffixes
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:56e07e4caf48: avcodec/dxtory: Use LL instead of L number suffix
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:d0ed336d8a4a: avformat/matroskadec: Fix number suffixes
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:ab17d11310a9: avformat/smacker: Fix number suffix
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:cd5c78c804d1: avformat/omadec: fix number suffix
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:004cdd8b1538: avcodec/h264_cabac: use int instead of long for mbb_xy
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:4f2299b6b55f: avcodec/mpegvideo_enc: Fix number suffixes in rc_buffer_size calculation
[19:50] <cone-408> ffmpeg.git 03wm4 07release/2.2:c7dc73a6c333: avformat/tta: fix crash with corrupted files
[19:50] <cone-408> ffmpeg.git 03wm4 07release/2.2:3d8c51d69916: avformat/mpc8: fix hang with fuzzed file
[19:50] <cone-408> ffmpeg.git 03wm4 07release/2.2:79d86b844f08: avformat/mpc8: fix broken pointer math
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:b902eab45a79: avformat/mpc8: Use uint64_t in *_get_v() to avoid undefined behavior
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:b0d387308525: avcodec/mjpegdec: Check escape sequence validity
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:4c246c65bfa2: avcodec/mjpegdec: Check number of components for JPEG-LS
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:bd9c755b22ff: swscale/utils: Limit filter shifting so as not to read from prior the array
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:7c3a3d47cf7f: avformat/thp: Check av_get_packet() for failure not only for partial output
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:eca1e3dcc847: avcodec/h264_ps: More completely check the bit depths
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:a3dca10470ec: avcodec/h264: Be more strict on rejecting pps_id changes
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:0c9d465e98c7: avcodec/h264: Be more strict on rejecting pps/sps changes
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:68d83bc3a692: avutil/opt: Fix types used to access AV_OPT_TYPE_PIXEL_FMT
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:8323f0944283: avutil/opt: Fix type used to access AV_OPT_TYPE_SAMPLE_FMT
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:e6093f5b85b3: avcodec/h264_slice: Do not change frame_num after the first slice
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:0afe061f2877: avcodec/h264_slice: Check picture structure before setting the related fields
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:f0526bc21eb4: avcodec/h264_slice: ignore SAR changes in slices after the first
[19:50] <cone-408> ffmpeg.git 03Carl Eugen Hoyos 07release/2.2:b1b69baa0110: lavc/aarch64: Do not use the neon horizontal chroma loop filter for H.264 4:2:2. (cherry picked from commit 4faea46bd906b3897018736208123aa36c3f45d5)
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:0ee5f46fd1d1: avcodec/mjpegdec: Skip blocks which are outside the visible area
[19:50] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:2dca276dcb9c: avcodec/arm/videodsp_armv5te: Fix linking failure with "g++ -shared -D__STDC_CONSTANT_MACROS -o test.so ... libavcodec.a"
[19:59] <Daemon404> michaelni, fyi your "24 bit support" for wmapro is entirely broken
[19:59] <Daemon404> it is massive distorted.
[19:59] <Daemon404> massively*
[20:00] <cone-408> ffmpeg.git 03Michael Niedermayer 07release/2.2:36cfee3adc70: Update for 2.2.13
[20:02] <michaelni> Daemon404, is there a ticket for this ? how can it be reproduced ? is it a regression?
[20:03] <Daemon404> its not a regression because the support was never real
[20:03] <Daemon404> it was an untested hack afaict
[20:03] <Daemon404> every 24bit wmapro file i play is distorted
[20:03] <Daemon404> let me see if i can get a file.
[20:05] <michaelni> i only remember a recent partial 24bit fix for wmalossless not wmapro
[20:06] <Daemon404> sorry
[20:06] <Daemon404> i meant wma lossless
[20:14] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:03dab49a1267: avcodec/vp56: Add () to protect TRANSPOSE()s arguments
[20:25] <cone-408> ffmpeg.git 03Diego Biurrun 07master:f8c1719771dc: dca: Remove separate header for DCA ExSS
[20:25] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:d2668ae47269: Merge commit 'f8c1719771dc4ac2e13e6bc8bf741854a30e3a86'
[20:47] <cone-408> ffmpeg.git 03Vittorio Giovara 07master:5dc47a2bd52e: matroskaenc: Validate chapter start and end times
[20:48] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:d302853bcabf: Merge commit '5dc47a2bd52e375ed742c45d08356b45098f458d'
[20:48] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:a4cd057bc7dd: avformat/matroskaenc: Use the correct data type for the chapter times
[20:58] <cone-408> ffmpeg.git 03Vittorio Giovara 07master:9d3b752fceb0: graphparser: Check av_get_token() memory error
[20:58] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:9d5017a20c26: Merge commit '9d3b752fceb0f2a42cac7c2a1109b0629823c99f'
[21:08] <cone-408> ffmpeg.git 03Vittorio Giovara 07master:81688e68f93f: avconv: Check rc_override memory allocation
[21:08] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:15cd2a93e7f6: Merge commit '81688e68f93f3142e2093f1a3d226edaeb179992'
[21:16] <cone-408> ffmpeg.git 03Vittorio Giovara 07master:266f241193b2: avplay: Check frame allocation inside video_thread()
[21:16] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:32b2b7bd8c5a: Merge commit '266f241193b2fa8c99bb8b1f007c66bedd3b7d97'
[21:23] <cone-408> ffmpeg.git 03Vittorio Giovara 07master:733f4b05f0e1: avplay: Check format allocation inside decode_thread()
[21:23] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:79216a189b4e: Merge commit '733f4b05f0e120ddd0393b23f2b6d9106cf922e4'
[21:35] <cone-408> ffmpeg.git 03Vittorio Giovara 07master:cb70a93ca1e7: lavc: Document interaction between avcodec_open2() and decoding routines
[21:35] <cone-408> ffmpeg.git 03Vittorio Giovara 07master:bb7701684c71: lavf: Document the codec context initialization
[21:35] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:2980c8e7d34e: Merge commit 'cb70a93ca1e7fae1b5dfb65ad149d311a65206a4'
[21:35] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:563ed912eaac: Merge commit 'bb7701684c7170dfd5f33a3c7d827265844619f2'
[21:42] <cone-408> ffmpeg.git 03Vittorio Giovara 07master:d89e58f53939: dct-test: Fix initialization syntax
[21:42] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:beae6ac5db9d: Merge commit 'd89e58f53939d91a88aa7042c94d9d1cd364da52'
[21:44] <iive> finally I see some libav merges. I was starting to wonder is it me that always miss them, or there aren't that many anymore.
[21:48] <cone-408> ffmpeg.git 03Hugo Beauzée-Luyssen 07master:da7e31a240f2: rmdec: Check memory allocations from ff_rm_alloc_rmstream()
[21:48] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:575ed4fe37fd: Merge commit 'da7e31a240f276836a0b90ca6c0714181b353cc4'
[22:00] <cone-408> ffmpeg.git 03Hugo Beauzée-Luyssen 07master:a8765852158e: movenc: Check memory allocations
[22:00] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:80c5dc56104d: Merge commit 'a8765852158ecb2ae34895fa35ff51dc95c186f9'
[22:00] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:95165f7c1b53: avformat/movenc: Fix stts_entries allocation check
[22:01] <akira4> ubitux, I used ./ffmpeg -i test.ass -flag2 +rawtextsub out.srt to transcode ass subtitles to srt. Is it the right way?
[22:02] <ubitux> yes
[22:02] <cbsrobot_> ubitux: what do you mean with "string needs to be dupped" ?
[22:03] <cbsrobot_> is there an example ?
[22:04] <ubitux> print_format in ffprobe
[22:14] <joeflateau> Is there a magic way of using tee to output hls and rtmp from one encode? I can't seem to get the bsf option right
[22:18] <BtbN> nginx-rtmp can do that
[22:18] <BtbN> You stream to it via rtmp, and it distributes the stream via rtmp and hls
[22:21] <cone-408> ffmpeg.git 03Hugo Beauzée-Luyssen 07master:3035d21b5a77: nutdec: Check memory allocations
[22:21] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:880cee004418: Merge commit '3035d21b5a77df0e9531df397fc06d59488996c7'
[22:29] <cone-408> ffmpeg.git 03Federico Tomassetti 07master:e51f22122d23: swscale: Check memory allocations
[22:29] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:2a1b79d7e659: Merge commit 'e51f22122d23589e93ac4f0b3e570bb925755915'
[22:38] <cone-408> ffmpeg.git 03Vittorio Giovara 07master:ec17782e17de: dvdsubdec: Check memory allocations
[22:38] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:3d5c0ba816b9: Merge commit 'ec17782e17de1e8501ca213e276dfe5412ff1d11'
[22:46] <cone-408> ffmpeg.git 03Vittorio Giovara 07master:31dc73e92a96: vorbisdec: Check memory allocations
[22:46] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:7a7b77e6c97a: Merge commit '31dc73e92a96f08d07650c0e7d31c0b9a1465d46'
[22:55] <cone-408> ffmpeg.git 03Vittorio Giovara 07master:fe0f4e56577a: avpacket: Check buffer reference
[22:55] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:62aec0c2d5b3: Merge commit 'fe0f4e56577a2dbd373bb50b0ae2d49e69d822fc'
[23:05] <cone-408> ffmpeg.git 03Himangi Saraogi 07master:18f4fa251b0e: aac_adtstoasc_bsf: Check extradata memory allocation
[23:05] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:38e867f6c08e: Merge commit '18f4fa251b0eb36392839f5bf6180f280dc04d8d'
[23:06] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:2e786bc49898: avcodec/aac_adtstoasc_bsf: Clear extradata_size on allocation failure
[23:26] <cone-408> ffmpeg.git 03Himangi Saraogi 07master:a1e2c47cd4fd: libxvid: Return meaningful error messages
[23:26] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:d128794fbc5e: Merge commit 'a1e2c47cd4fdaa64beda7e6dfa623b65c46f5012'
[23:34] <cone-408> ffmpeg.git 03Himangi Saraogi 07master:f3e045263e44: qdm2: Return meaningful error codes
[23:34] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:41572ac698b3: Merge commit 'f3e045263e445c4ab54d85ecae359494cd96a3e2'
[23:42] <cone-408> ffmpeg.git 03Himangi Saraogi 07master:42c8f92e2fa3: wmv2: Return meaningful error codes
[23:42] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:d06dce21a454: Merge commit '42c8f92e2fa390fa17b03d37b4323ec0d721d4cd'
[23:51] <cone-408> ffmpeg.git 03Vittorio Giovara 07master:c78dc12118cb: riff: Support ProRes in avi (APCH fourcc)
[23:51] <cone-408> ffmpeg.git 03Michael Niedermayer 07master:97bd0e2a93f1: Merge commit 'c78dc12118cb5b2a69a17f38c69f77a42e89a9c4'
[00:00] --- Wed Feb 18 2015