[Ffmpeg-devel-irc] ffmpeg-devel.log.20161204

Mon Dec 5 03:05:02 EET 2016

[00:17:39 CET] <philipl> BtbN, nevcairiel: They updated the support matrix on the website to say cuvid can decode 10/12bit vp9 on "select" pascal chips. Doesn't say which, of course.
[00:18:14 CET] <nevcairiel> it says the same thing about 8k and that is available on all except the gp100 chip w hich isnt available for consumers
[00:18:15 CET] <wm4> woah
[00:18:26 CET] <philipl> Yeah. So that's what I'm hoping is true here too.
[00:18:30 CET] <philipl> g vs h hardware.
[00:18:39 CET] <wm4> what's the difference?
[00:18:42 CET] <nevcairiel> but we'll need to wait until the next sdk, the sdk docs have a proper feature matrix
[00:18:55 CET] <nevcairiel> with chips instead of just categories
[00:19:01 CET] <philipl> link?
[00:19:13 CET] <philipl> oh you're saying the next doc will have it.
[00:19:46 CET] <philipl> wm4: http://us.download.nvidia.com/XFree86/Linux-x86/375.20/README/vdpausupport.html
[00:19:57 CET] <nevcairiel> read NVDEC_VideoDecoder_API_ProgGuide.pdf
[00:20:08 CET] <philipl> According to that (the most complete description they offer of the hardware), 8k is specific to the 'H' hardware.
[00:20:12 CET] <nevcairiel> it lists the difference between GP100 and GPO10x
[00:20:20 CET] <nevcairiel> -O
[00:20:23 CET] <wm4> ah
[00:20:51 CET] <wm4> btw. are there any signs about nvidia still caring about vdpau?
[00:20:56 CET] <philipl> wm4: no.
[00:21:29 CET] <philipl> I can ask around whether they reassigned or hired a replacement engineer...
[00:23:39 CET] <wm4> I mean obviously they're keeping it working, and there have been bugs that disappeared in newer driver versions
[00:31:07 CET] <philipl> Yeah, but no evidence of actually adding new features.
[00:41:24 CET] <Compn> dont we have an nvidia contact ?
[00:41:31 CET] <Compn> a few devels worked there ...
[00:41:34 CET] <Compn> even vlc devs
[00:41:47 CET] <Compn> not asking for trade secrets or anything
[00:41:53 CET] <Compn> but maybe some winks and nods to whats going on
[00:47:51 CET] <nevcairiel> huge companies dont really do that
[00:48:24 CET] <nevcairiel> if you're not working right in that department, even finding this stuff out internally is super hard
[01:03:52 CET] <cone-430> ffmpeg 03Michael Niedermayer 07master:020d53ebdb58: avcodec/flac: Check bps in ff_flac_parse_streaminfo()
[01:03:52 CET] <cone-430> ffmpeg 03Michael Niedermayer 07master:6567c59c4903: avcodec/flac: forward errors from ff_flac_parse_streaminfo()
[01:03:52 CET] <cone-430> ffmpeg 03Michael Niedermayer 07master:c72fa4323498: avcodec/get_bits: Fix get_sbits_long(0)
[01:03:52 CET] <cone-430> ffmpeg 03Michael Niedermayer 07master:1f5630af51f2: avcodec/flacdec: Fix undefined shift in decode_subframe()
[01:29:10 CET] <jkqxz> nevcairiel:  How does the socket unblocking thing on Windows avoid racing with preemption before the recv call?
[01:33:16 CET] <jkqxz> (I mean, that's why the close is undefined on unices.  You can't distinguish "blocked in recv" from "preempted just before being blocked in recv", and the latter is undefined behaviour (you are calling recv on a random file descriptor, which might have been reused by the implementation) so the whole thing has to be undefined behaviour.)
[01:58:13 CET] <cone-430> ffmpeg 03Michael Niedermayer 07release/2.8:518beeb72f53: Changelog: Update
[01:58:15 CET] <cone-430> ffmpeg 03Michael Niedermayer 07release/2.8:3af916db3705: avformat/utils: Check start/end before computing duration in update_stream_timings()
[01:58:15 CET] <cone-430> ffmpeg 03Michael Niedermayer 07release/2.8:f202fefdb0ac: avformat/oggparsespeex: Check frames_per_packet and packet_size
[01:58:16 CET] <cone-430> ffmpeg 03Michael Niedermayer 07release/2.8:5790ce62732f: avcodec/flacdsp_template: Fix undefined shift in flac_decorrelate_indep_c
[01:58:17 CET] <cone-430> ffmpeg 03Michael Niedermayer 07release/2.8:31d46dc97dc3: avcodec/flacdec: Fix signed integer overflow in decode_subframe_fixed()
[01:58:18 CET] <cone-430> ffmpeg 03Michael Niedermayer 07release/2.8:1e512388ee98: avformat/ffmdec: Check media type for chunks
[01:58:19 CET] <cone-430> ffmpeg 03Michael Niedermayer 07release/2.8:a5989651f075: avcodec/get_bits: Fix get_sbits_long(0)
[01:58:20 CET] <cone-430> ffmpeg 03Michael Niedermayer 07release/2.8:3f8bb78f3ed6: avcodec/flacdec: Fix undefined shift in decode_subframe()
[01:58:21 CET] <cone-430> ffmpeg 03Michael Niedermayer 07release/2.8:fb93771072cf: Changelog: update
[02:23:43 CET] <KGB> [13FFV1] 15dericed opened pull request #25: Use rfc formatting (06master...06use-rfc-formatting) 02https://git.io/v1WAP
[03:16:00 CET] <cone-430> ffmpeg 03Michael Niedermayer 07n2.8.9:HEAD: avcodec/flacdec: Fix undefined shift in decode_subframe()
[17:05:49 CET] <KGB> [13FFV1] 15michaelni pushed 1 new commit to 06master: 02https://git.io/v1l8z
[17:05:49 CET] <KGB> 13FFV1/06master 144253d12 15dericed: fix error in code wrapping
[21:13:10 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:86f5b73429d8: avcodec/utils: Clear MMX state before returning from avcodec_default_execute*()
[21:13:10 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:9128ef030208: avcodec/mpegvideo_enc: Clear mmx state in ff_mpv_reallocate_putbitbuffer()
[21:13:10 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:bab7d72d4645: avformat/mxfdec: Check size to avoid integer overflow in mxf_read_utf16_string()
[21:13:10 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:d54aa2de1761: avcodec/interplayvideo: Check side data size before use
[21:13:10 CET] <cone-754> ffmpeg 03Ronald S. Bultje 07release/3.0:642cd5de4de6: vp9: change order of operations in adapt_prob().
[21:13:11 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:8cca6b06ec6c: avcodec/dvdsubdec: Fix buf_size check
[21:13:11 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:9ec40eacb169: avcodec/dvdsubdec: Fix off by 1 error
[21:13:12 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:35498718474f: avcodec/8bps: Check side data size before use
[21:13:12 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:9cab9b878fd4: avcodec/cinepak: Check side data size before use
[21:13:13 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:e0dd617ed7f6: avcodec/idcinvideo: Check side data size before use
[21:13:14 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:adc86ad32f04: avcodec/kmvc:  Check side data size before use
[21:13:15 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:136fc7084e16: avcodec/msrle:  Check side data size before use
[21:13:17 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:51aa6d09872d: avcodec/qtrle:  Check side data size before use
[21:13:18 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:1bc7215e8a44: avcodec/qpeg:  Check side data size before use
[21:13:19 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:7affe32738d8: avcodec/msvideo1: Check side data size before use
[21:13:20 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:8fd3293fa148: avcodec/rawdec: Check side data size before use
[21:13:20 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:4251dc6d7e7d: avcodec/tscc:  Check side data size before use
[21:13:22 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:bc63205ae2eb: avcodec/sunrast: Fix input buffer pointer check
[21:13:23 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:083919e9c459: avcodec/movtextdec: Fix potential integer overflow
[21:13:24 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:3f9a24f06d5d: avcodec/movtextdec: Fix tsmb_size check==0 check
[21:13:25 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:27fbf8f110f6: avcodec/movtextdec: Add error message for tsmb_size check
[21:13:25 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:8ad4e1c1cb74: avcodec/ituh263dec: Avoid spending a long time in slice sync
[21:13:26 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:63caf18787d5: avcodec/rv40: Test remaining space in loop of get_dimension()
[21:13:28 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:2caee4857687: avformat/mpeg: Adjust vid probe threshold to correct mis-detection
[21:13:29 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:ff01dbb6ec88: avformat/idroqdec: Check chunk_size for being too large
[21:13:30 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:6bee6ef0c2ec: avcodec/flac_parser: Update nb_headers_buffered
[21:13:31 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:66e8f87ebc78: avformat/utils: Check start/end before computing duration in update_stream_timings()
[21:13:32 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:e9003828dd7e: avformat/oggparsespeex: Check frames_per_packet and packet_size
[21:13:33 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:2f26f3de7801: avcodec/flacdsp_template: Fix undefined shift in flac_decorrelate_indep_c
[21:13:33 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:2d66fbc853ff: avcodec/flacdec: Fix signed integer overflow in decode_subframe_fixed()
[21:13:35 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:9f7eb718e9d3: avformat/ffmdec: Check media type for chunks
[21:13:36 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:66bf84e2c4e1: avcodec/get_bits: Fix get_sbits_long(0)
[21:13:37 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:2c6792b0c683: avcodec/flacdec: Fix undefined shift in decode_subframe()
[21:24:32 CET] <cone-754> ffmpeg 03Michael Niedermayer 07release/3.0:6e1bc747df86: Update for version 3.0.5
[21:27:41 CET] <cone-754> ffmpeg 03James Almer 07release/3.0:d111c9ce139c: avcodec/rawdec: check for side data before checking its size
[00:00:00 CET] --- Mon Dec  5 2016