MythLogBot@irc.freenode.net :: #mythtv

Daily chat history

Current users (51):

gregl, jya, k-man, KdW, markspieth, MythBuild, MythLogBot, MythNotifyBot, ooshlablu, Tobbe5178, toddejohnson_, Warped, xris, zbot, aloril, brfransen, ChanServ, eharris, enyc, frobnic, GreyFoxx, Hydroponx, jheizer, jpabq, kalamaja, MitchCapper, peper03, poptix, pppingme, rhpot1991, sphery, stuartm, tonsofpcs, _charly_, Anssi, davic, dym, ElmerFudd, gigem, ijc, knowledgejunkie, libsci, mkbloke, Panic, ramshadow, sdrik, tris, RokLobsta, kwizart, Maliuta_[m], buz48
Thursday, July 4th, 2019, 00:15 UTC
[00:15:23] mad_enz (mad_enz!~mad_enz@2606:6d00:135:9d84:4c13:ca2c:b3cf:a0f2) has quit (Read error: Connection timed out)
[00:15:56] mad_enz (mad_enz!~mad_enz@2606:6d00:135:9d84:3822:f2e6:bb38:542c) has joined #mythtv
[00:23:02] amessina (amessina!~amessina@unaffiliated/amessina) has quit (Remote host closed the connection)
[00:23:29] amessina (amessina!~amessina@unaffiliated/amessina) has joined #mythtv
[00:36:23] sdrik (sdrik!~sdrik@sdrik.gabriello.fr) has quit (Ping timeout: 268 seconds)
[00:37:35] sdrik (sdrik!~sdrik@sdrik.gabriello.fr) has joined #mythtv
[01:05:08] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[01:05:08] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[01:09:32] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 244 seconds)
[02:00:25] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[02:00:25] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[02:04:49] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 244 seconds)
[02:06:51] mad_enz (mad_enz!~mad_enz@2606:6d00:135:9d84:3822:f2e6:bb38:542c) has quit (Read error: Connection timed out)
[02:07:29] mad_enz (mad_enz!~mad_enz@2606:6d00:135:9d84:3822:f2e6:bb38:542c) has joined #mythtv
[02:22:20] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[02:22:20] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[02:29:05] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 264 seconds)
[03:16:22] mad_enz (mad_enz!~mad_enz@2606:6d00:135:9d84:3822:f2e6:bb38:542c) has quit (Read error: Connection timed out)
[03:31:11] hampton (hampton!~dhampton@mythtv/developer/hampton) has left #mythtv ()
[03:34:00] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[03:34:00] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[04:01:08] amessina (amessina!~amessina@unaffiliated/amessina) has quit (Quit: Konversation terminated!)
[04:38:50] Tobbe5178 (Tobbe5178!~asdf@2001:2002:51eb:d24e:197b:260d:4822:1aaf) has joined #mythtv
[05:52:56] Steve-Goodey (Steve-Goodey!~steve@2a00:23c5:7d90:bc01:d516:778c:7853:534c) has joined #mythtv
[05:52:56] Mode for #mythtv by ChanServ!ChanServ@services. : +v Steve-Goodey
[06:25:54] stuarta: jpabq: there is an rsync job which pulls themes across to ftp.osuosl.org twice a day ~4 am/pm GMT iirc
[06:53:07] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 244 seconds)
[07:07:02] markk__ (markk__!~mark@host86-166-192-244.range86-166.btcentralplus.com) has joined #mythtv
[07:12:19] markk__: hrm – had totally forgotten about the test video playback/setup wizard. Slightly odd that the help text says test 1080p content and the sample is actually 1920x816:)
[07:12:49] markk__: that wizard could do with maybe a little updating for HEVC and perhaps 4k
[07:13:31] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[07:13:32] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[07:18:17] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 264 seconds)
[07:23:33] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[07:23:33] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[07:46:12] stuarta: morning
[07:51:30] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 244 seconds)
[07:53:04] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[07:53:05] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[07:59:24] willcooke (willcooke!~willcooke@ubuntu/member/willcooke) has joined #mythtv
[08:02:36] Merlin83b (Merlin83b!~Daniel@office.34sp.com) has joined #mythtv
[09:03:56] markk__: peterbennett: looks like the pi openmax situation is as messy as ever. as far as I can tell – kodi uses MMAL – which is just a broadcom library that sits on top of openmax (pointless imho). so apparently new hevc decoding works for kodi/pi4 (though not sure if that refers to the other ffmpeg hevc patch)
[09:05:26] markk__: the other hevc FFmpeg pi patch is, I think, mostly neon optimisations for hevc decoding. apparently 45k lines of code:) I'd be surprised if that makes it into FFmpeg – though you never know
[09:15:14] ** stuarta peers curiously at #13467 **
[09:15:14] ** MythLogBot https://code.mythtv.org/trac/ticket/13467 **
[09:38:01] kwizart: gigem, markk__ there is a downstream patch "from pi" for that improves the mmal decoder for vlc, IMO openxil is deprecated for the pi (it was non-conformant from the start at least)
[09:42:06] markk__: kwizart: but mmal just sits on top of openmax – it was created to simplify the handling of openmax – mostly before most people got their heads around how openmax works
[09:45:46] kwizart: markk__, what makes you think mmal is related to openmax ?
[09:46:32] kwizart: to me is a totally different API (but serving the same purpose)
[09:55:20] markk__: kwizart: https://github.com/techyian/MMALSharp/wiki/What-is-MMAL%3F
[09:55:27] markk__: that was my understanding
[10:04:32] kwizart: yep, I've hit this also, but I think he's wrong. If that would be the case, any libmmal*so would be linked to libopenmaxil.so, and it doesn't seem the case
[10:04:32] markk__: though now I'm starting to doubt it. I'm sure I saw an MMAL/IL mapping header when i was working on some pi camera stuff a while back...
[10:05:41] markk__: but more usefully – for some reason I never realised FFmpeg had a MMAL decoder – we should be switching to that – but I'm not sure we get the EGL integration for direct rendering:(
[10:08:14] stuarta: life is never simple :(
[10:16:41] kwizart: with that said I wonder how it would work with pi4 since it will uses "upstream" v3d driver (mesa/libdrm) as primary official driver, but does it means mmal is still relevant there over v4l2-requests api ?
[10:17:11] markk__: hrm – looks like there is egl functionality – but will it be useable with ffmpeg...
[10:17:30] markk__: what a mess:)
[10:18:51] stuarta: kwizart: i suspect that may apply to running raspbian buster on earlier pi's as well
[10:22:36] kwizart: yep, seems like earlier pi can use the vc4 driver (mesa/libdrm), with decent performances, but there the default remains the raspberry-pi-vc driver
[10:27:13] markk__: bugger – really don't know where to go with pi for the render branch now...
[10:29:58] stuarta: markk__: target buster
[10:30:15] stuarta: by the time we release this, buster should be common
[10:32:52] markk__: stuarta: but openmax/mmal/v3d (i.e. drm)? we could use a drm implementation
[10:33:13] markk__: warpme_: any view? you probably have a better idea than most
[10:33:23] stuarta: good question
[10:35:56] kwizart: I think kodi went the drm way (IIRC)
[10:36:33] kwizart: if not a full egl/gles driver
[10:37:07] markk__: and egl fullscreen v X11 might be a pain. I suspect EGL fullscreen will break with the new drivers (on the Qt side that is) – although maybe the default/non-broadcom specific EGLFS will work
[10:37:47] kwizart: https://archive.fosdem.org/2018/schedule/event/kodi/
[10:39:41] stuarta: just reading the overview that looks interesting
[11:00:26] stuarta: markk__: ^^^ this video is worth a look
[11:07:32] markk__ (markk__!~mark@host86-166-192-244.range86-166.btcentralplus.com) has quit (Ping timeout: 244 seconds)
[11:14:03] markk__ (markk__!~mark@host86-166-192-244.range86-166.btcentralplus.com) has joined #mythtv
[11:16:29] markk__: stuarta: yeah  – that was something on my roadmap – but not in the near future. drm prime that is. and of course keeps warpme_ happy:) an in 2 minds now whether to get a pi 4 – as that would be my simplest route to getting something that has DRM support to develop/test on.
[11:17:30] markk__: I'm thinking existing code – extended for EGL interop – for Pi < 4 and try and get some drm/prime up and running for >= 4
[11:17:59] markk__: anyone know if drm prime needs x running?
[12:12:02] hampton (hampton!~dhampton@pool-68-134-213-224.bltmmd.fios.verizon.net) has joined #mythtv
[12:12:02] hampton (hampton!~dhampton@mythtv/developer/hampton) has joined #mythtv
[12:12:02] hampton (hampton!~dhampton@pool-68-134-213-224.bltmmd.fios.verizon.net) has quit (Changing host)
[12:12:02] Mode for #mythtv by ChanServ!ChanServ@services. : +v hampton
[12:16:43] stuarta: markk__: i'm going to drop a raspbian buster on a fresh SD card and see how that goes on the rpi3
[12:18:22] markk__: stuarta: can you check whether the dri module is loaded by default? i.e. is there anything under /dev/dri?
[12:20:40] stuarta: it's not when set to legacy (as mine was), reconfigured it and now the build has finished, i shall reboot it to see
[12:25:27] stuarta: markk__: yup, have /dev/dri/card0 and /dev/dri/by-path/platform-soc:gpu-card -> ../card0
[12:25:42] stuarta: (ie. 2nd one is a symlink back to card0)
[12:43:18] markk__ (markk__!~mark@host86-166-192-244.range86-166.btcentralplus.com) has quit (Ping timeout: 272 seconds)
[12:44:46] markspieth2 (markspieth2!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[12:44:46] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth2
[12:47:02] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 244 seconds)
[12:48:26] markk__ (markk__!~mark@host86-166-192-244.range86-166.btcentralplus.com) has joined #mythtv
[12:49:29] markspieth2 (markspieth2!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 264 seconds)
[13:12:43] markk__ (markk__!~mark@host86-166-192-244.range86-166.btcentralplus.com) has quit (Ping timeout: 246 seconds)
[13:29:17] markk__ (markk__!~mark@host86-166-192-244.range86-166.btcentralplus.com) has joined #mythtv
[13:34:17] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[13:34:17] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[13:38:42] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 244 seconds)
[13:55:04] willcooke (willcooke!~willcooke@ubuntu/member/willcooke) has quit (Remote host closed the connection)
[14:07:36] willcooke (willcooke!~willcooke@ubuntu/member/willcooke) has joined #mythtv
[14:12:49] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[14:12:49] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[14:19:00] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 244 seconds)
[14:20:56] ghoti (ghoti!~paul@glphon2233w-grc-09-184-145-52-216.dsl.bell.ca) has joined #mythtv
[14:21:54] peterbennett: I see mmal for raspberry pi and ffmpeg supports it. Is it new? I never noticed it before.
[14:22:12] peterbennett: The fact that ffmpeg supports mmal is a strong reason for going that way.
[14:24:12] peterbennett: I am not sure if it is just a layer on openmax, various pages say that is what it is.
[14:25:29] stuarta: it's not new, it's been there since the rpi's inception, AFAICT
[14:26:56] peterbennett: I think maybe the ffmpeg support fir mmal is new
[14:28:23] peterbennett: Or maybe I just never looked for it, since LVR developed using OpenMAX.
[14:46:11] stuarta: it's entirely possible the support in ffmpeg is new
[14:46:46] willcooke (willcooke!~willcooke@ubuntu/member/willcooke) has quit (Remote host closed the connection)
[14:49:20] markk__ (markk__!~mark@host86-166-192-244.range86-166.btcentralplus.com) has quit (Ping timeout: 272 seconds)
[14:56:01] willcooke (willcooke!~willcooke@ubuntu/member/willcooke) has joined #mythtv
[15:00:31] Warped (Warped!~Warped@unaffiliated/warped) has joined #mythtv
[15:01:37] enyc (enyc!~enyc@muddle.enyc.org.uk) has quit (Ping timeout: 250 seconds)
[15:05:31] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[15:05:32] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[15:07:46] enyc (enyc!~enyc@muddle.enyc.org.uk) has joined #mythtv
[15:07:48] peterbennett: jpabq: Steppes on v30 is fine now, thanks :)
[15:14:41] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 264 seconds)
[15:46:02] warpme_ (warpme_!~warpme_@89-70-137-1.dynamic.chello.pl) has joined #mythtv
[15:57:01] warpme_: hi guys!
[15:57:56] warpme_: saw discussion on rpi4 video and generally video playback stuff....
[15:58:57] warpme_: I?m a bit off video topic as recently was busy to getting minimyth2 to booting on allwinner h6 soc...
[16:10:02] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[16:10:02] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[16:10:18] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has joined #mythtv
[16:10:18] Mode for #mythtv by ChanServ!ChanServ@services. : +v natanojl
[16:12:19] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Read error: Connection reset by peer)
[16:20:12] warpme_: but what i think generally regarding video stuff: first lets decompose issue into 2 major areas: decode and rendering. For decode – I personaly think we shoud go with functinality provided by current ffmpeg and rather avoid private decoders. Future here seems to be v4l2. For rendering I think myth should support 2 general methods: egl imports and drm prime. IIRC egl path exports frame from ffmpeg to gles3.2 surface and lests gpu do rendering
[16:20:12] warpme_: drm prime exports frame to drm plane and asks drm to do scaling, conversion and composing (i.e. video frame and osd). drm can do this in atomic updaes synced to crt vsync. drm is quite powerful as it can do rescaling, conversion in hardware. For both rendering methods we should go with dmabuf I think as it is allowing zero cpu copy. I was investigating drm-prime method sometime ago and on s905 got 720p 24fps content decoded+redered at 60fps
[16:20:12] warpme_: 8–10% of cpu total for decode+render. This was achieved with ffmpeg4.0.2+lrusak patches for ffmpeg use drm dmabuf exports (to do zerocopy frames from v4l2 buffers to drm planes) & v4l2 hwdecode on s905. Test code I used is here:https://github.com/Elyotna/ffmpeg-drm ffmpeg patch to enable dmabuf on ffmpeg: https://github.com/warpme/minimyth2/blob/mast . . . k-v4l2.patch
[16:22:20] warpme_: now – for rpi4 I think this drm-prime + ffmpeg-mmal might be doable as: drm-prime is already there. mmal aspect of ffmpeg can be hidden for myth in the same way like v4l2 decode can be hidden...
[16:29:56] warpme_: I personally think v4l2 & drm-prime is way to go for us – if we want to have myth playing video nicelly on arm soc world... (btw: this exacly route is only single route kodi is getting for future). iirc v19 of kodi will remove all private decoders, patches for speciffic hardware, etc – just for going with only ffmpeg + drm-prime. (nice thing it is foss so we can easy see code + ask ppl like lukas rusak or maximie jourdan for details)....
[16:30:47] stuarta: warpme_: did you see the link kwizart posted earlier? it was lukas' talk at fosdem18
[16:33:18] warpme_: yes. saw it sometime ago. another nice video is maxime on drm-prime: https://www.youtube.com/watch?v=pF6i19oscqg see at 16:03 onward. maxime nicelly explains drm-prime...
[16:53:11] warpme_ (warpme_!~warpme_@89-70-137-1.dynamic.chello.pl) has quit (Quit: warpme_)
[16:54:20] Merlin83b (Merlin83b!~Daniel@office.34sp.com) has quit (Quit: Leaving)
[17:38:49] willcooke (willcooke!~willcooke@ubuntu/member/willcooke) has quit (Quit: Do your hobbies)
[19:30:44] buz48 (buz48!4b83860d@75-131-134-13.static.opls.la.charter.com) has joined #mythtv
[19:31:15] buz48: hello world. I have a patch for mythburn.py
[19:31:39] buz48: where can I submit this patch?
[20:00:45] gigem: buz48: Open a ticket at https://code.mythtv.org/trac/ and attach your patch to it.
[20:25:11] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[20:25:11] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[20:32:05] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 264 seconds)
[20:55:11] Steve-Goodey (Steve-Goodey!~steve@2a00:23c5:7d90:bc01:d516:778c:7853:534c) has quit (Quit: Konversation terminated!)
[21:21:14] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[21:21:14] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth
[21:29:23] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 244 seconds)
[22:00:14] peterbennett (peterbennett!~pi@mythtv/developer/peterbennett) has quit (Quit: WeeChat 1.6)
[22:02:03] peterbennett (peterbennett!~pi@mythtv/developer/peterbennett) has joined #mythtv
[22:02:03] Mode for #mythtv by ChanServ!ChanServ@services. : +v peterbennett
[22:39:57] ghoti (ghoti!~paul@glphon2233w-grc-09-184-145-52-216.dsl.bell.ca) has quit (Read error: No route to host)
[22:45:55] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has quit (Ping timeout: 268 seconds)
[23:58:47] markspieth (markspieth!~yaaic@mythtv/developer/markspieth) has joined #mythtv
[23:58:47] Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth

IRC Logs collected by BeirdoBot.
Please use the above link to report any bugs.