| Monday, February 2nd, 2015, 00:02 UTC | ||
| [00:02:30] | joki (joki!~joki@p5486178C.dip0.t-ipconnect.de) has joined #mythtv-users | |
| [00:19:16] | Warped (Warped!~Warped@108.85.160.119) has quit (Quit: ChatZilla 0.9.91.1 [Firefox 35.0.1/20150122214805]) | |
| [00:30:04] | lol768- (lol768-!lol7681@bnc.lol768.com) has left #mythtv-users ("WeeChat 0.4.2") | |
| [00:40:55] | Warped (Warped!~Warped@108.85.160.119) has joined #mythtv-users | |
| [00:49:09] | [R] ([R]!~rbox@unaffiliated/rbox) has joined #mythtv-users | |
| [00:50:18] | hR13_ (hR13_!~hR13@37.250.59.47.bredband.tre.se) has joined #mythtv-users | |
| [00:54:18] | hR13__ (hR13__!~hR13@2.71.247.32.mobile.tre.se) has quit (Ping timeout: 244 seconds) | |
| [01:30:44] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has joined #mythtv-users | |
| [01:49:22] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has quit (Quit: Textual IRC Client: http://www.textualapp.com/) | |
| [01:49:59] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has joined #mythtv-users | |
| [01:50:34] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has quit (Client Quit) | |
| [01:51:14] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has joined #mythtv-users | |
| [01:53:18] | mcmoyer is now known as Shoedazzle | |
| [01:57:28] | Warped (Warped!~Warped@108.85.160.119) has quit (Quit: ChatZilla 0.9.91.1 [Firefox 35.0.1/20150122214805]) | |
| [02:02:57] | Shoedazzle is now known as mcmoyer | |
| [02:04:03] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has quit (Quit: Textual IRC Client: http://www.textualapp.com/) | |
| [02:04:27] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has joined #mythtv-users | |
| [02:21:21] | [R] ([R]!~rbox@unaffiliated/rbox) has quit (Quit: Leaving) | |
| [02:34:25] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has quit (Quit: Computer has gone to sleep.) | |
| [02:35:25] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has joined #mythtv-users | |
| [02:36:52] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has quit (Client Quit) | |
| [02:43:24] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has joined #mythtv-users | |
| [02:50:29] | Warped (Warped!~Warped@108.85.160.119) has joined #mythtv-users | |
| [02:53:01] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has quit (Quit: Computer has gone to sleep.) | |
| [03:04:30] | dmfrey (dmfrey!~dmfrey@65-78-98-83.c3-0.smt-ubr1.atw-smt.pa.cable.rcn.com) has quit (Quit: Ex-Chat) | |
| [03:55:48] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has joined #mythtv-users | |
| [03:57:54] | dj_segfault (dj_segfault!~david@75-149-142-18-Connecticut.hfc.comcastbusiness.net) has quit (Ping timeout: 265 seconds) | |
| [04:04:13] | ubIx_ (ubIx_!~ulf@p5484CDFA.dip0.t-ipconnect.de) has joined #mythtv-users | |
| [04:07:25] | ubIx (ubIx!~ulf@p5B264146.dip0.t-ipconnect.de) has quit (Ping timeout: 256 seconds) | |
| [04:08:37] | moparisthebest (moparisthebest!~mitb@gateway/tor-sasl/moparisthebest) has quit (Ping timeout: 250 seconds) | |
| [04:14:22] | moparisthebest (moparisthebest!~mitb@unaffiliated/moparisthebest) has joined #mythtv-users | |
| [04:17:53] | SmallwoodDR82 (SmallwoodDR82!~Smallwood@unaffiliated/smallwooddr82) has joined #mythtv-users | |
| [04:18:30] | [R] ([R]!~rbox@unaffiliated/rbox) has joined #mythtv-users | |
| [04:27:39] | davidbrooke (davidbrooke!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has quit (Ping timeout: 264 seconds) | |
| [04:27:43] | davidbrooke_ (davidbrooke_!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has joined #mythtv-users | |
| [04:47:39] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has quit (Quit: Computer has gone to sleep.) | |
| [04:51:39] | fetzerch_ (fetzerch_!~quassel@unaffiliated/fetzerch) has joined #mythtv-users | |
| [04:55:31] | fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has quit (Ping timeout: 245 seconds) | |
| [05:01:33] | buu_ is now known as buu | |
| [05:11:49] | andreaz (andreaz!~andre_000@p57922A48.dip0.t-ipconnect.de) has joined #mythtv-users | |
| [05:34:17] | davidbrooke_ (davidbrooke_!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has quit (Ping timeout: 250 seconds) | |
| [05:34:54] | davidbrooke (davidbrooke!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has joined #mythtv-users | |
| [05:51:23] | sl1ce_2g (sl1ce_2g!~johnathan@pool-72-74-164-209.bstnma.fios.verizon.net) has quit (Ping timeout: 244 seconds) | |
| [06:08:24] | [R] ([R]!~rbox@unaffiliated/rbox) has quit (Remote host closed the connection) | |
| [06:16:44] | SmallwoodDR82 (SmallwoodDR82!~Smallwood@unaffiliated/smallwooddr82) has quit (Quit: and I'm out...) | |
| [06:54:21] | jluttine (jluttine!~jluttine@dsl-hkibrasgw1-58c39e-156.dhcp.inet.fi) has joined #mythtv-users | |
| [06:54:36] | sl1ce_2g (sl1ce_2g!~johnathan@pool-72-74-19-126.bstnma.fios.verizon.net) has joined #mythtv-users | |
| [06:55:27] | jluttine: | hmm.. my livetv has suddenly stopped working and mythfilldatabase doesn't fetch program info.. any ideas what could be wrong? on livetv, i get error: "you should have received a channel lock by now. ..." |
| [06:56:11] | jluttine: | i can't remember changing any settings, and it has worked previously |
| [07:05:11] | jluttine (jluttine!~jluttine@dsl-hkibrasgw1-58c39e-156.dhcp.inet.fi) has quit (Remote host closed the connection) | |
| [07:11:24] | hpeter (hpeter!~hpeter@77-56-33-72.dclient.hispeed.ch) has joined #mythtv-users | |
| [07:12:01] | knightr__ (knightr__!~Nicolas@69-165-170-178.dsl.teksavvy.com) has joined #mythtv-users | |
| [07:12:34] | knightr_ (knightr_!~Nicolas@69-165-170-178.dsl.teksavvy.com) has quit (Read error: Connection reset by peer) | |
| [07:19:15] | jluttine (jluttine!~jluttine@dsl-hkibrasgw1-58c39e-156.dhcp.inet.fi) has joined #mythtv-users | |
| [07:19:27] | jluttine: | fyi, rebooting fixed |
| [07:20:17] | jluttine: | this happens every now and then that i have to reboot my computer because mythtv breaks. any ideas why? |
| [07:23:38] | Captain_Murdoch2 (Captain_Murdoch2!~cpinkham@c-67-170-1-245.hsd1.wa.comcast.net) has quit (Ping timeout: 246 seconds) | |
| [07:33:02] | Roklobsta (Roklobsta!~Roklobsta@ppp118-209-87-48.lns20.mel4.internode.on.net) has joined #mythtv-users | |
| [07:46:15] | Freejack (Freejack!~Freejack@68-190-10-237.dhcp.dctr.al.charter.com) has quit (Ping timeout: 264 seconds) | |
| [07:47:03] | Tobbe5178 (Tobbe5178!~asdf@2001:2002:d9d4:ce7e:5976:6a1b:4757:c3a6) has joined #mythtv-users | |
| [09:34:33] | stuartm: | did anyone in the UK record the superbowl on C4 HD? |
| [10:00:40] | hashbang (hashbang!~isajb@cse-ajb.cse.bris.ac.uk) has joined #mythtv-users | |
| [10:06:59] | Warped (Warped!~Warped@108.85.160.119) has quit (Ping timeout: 250 seconds) | |
| [10:11:28] | jimBaxter (jimBaxter!~jbaxter@jimbax.plus.com) has joined #mythtv-users | |
| [10:16:04] | Warped (Warped!~Warped@108.85.160.119) has joined #mythtv-users | |
| [11:13:05] | andreaz (andreaz!~andre_000@p57922A48.dip0.t-ipconnect.de) has quit (Quit: Leaving.) | |
| [11:59:00] | Roklobsta (Roklobsta!~Roklobsta@ppp118-209-87-48.lns20.mel4.internode.on.net) has quit (Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/) | |
| [12:04:42] | grumpydevil: | stuartm, when i look at mythmaster:6544 default status page, it also displays the Frontend Status. That includes a slave backend as if it is a frontend. Is that on purpose, and if so why? |
| [12:10:12] | stuartm: | it's not on purpose, best guess is that the slave backend is opening a playback or monitor connection, which it's currently assumed that only frontends would do |
| [12:11:30] | stuartm: | I don't have a slave so I've not be able to test the behaviour with one |
| [12:11:58] | stuartm: | I'd planned on tracking the state of slaves in addition to frontends, but I've not done it yet |
| [12:16:33] | Warped (Warped!~Warped@108.85.160.119) has quit (Quit: ChatZilla 0.9.91.1 [Firefox 35.0.1/20150122214805]) | |
| [12:19:09] | stuartm: | "-v network" logs from when the slave connects should show what type of connections it's creating to the master |
| [12:19:23] | stuartm: | that's on the "-v network" on the master |
| [12:20:57] | stuartm: | no idea why the slave would need to open any connection other than a SlaveBackend one ... but that's the only way it could be mistaken for a frontend |
| [13:41:55] | dmfrey (dmfrey!~dmfrey@65-78-98-83.c3-0.smt-ubr1.atw-smt.pa.cable.rcn.com) has joined #mythtv-users | |
| [13:44:15] | chompydile is now known as croccydile | |
| [14:33:01] | moparisthebest (moparisthebest!~mitb@unaffiliated/moparisthebest) has quit (Ping timeout: 256 seconds) | |
| [14:38:04] | davidbrooke (davidbrooke!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has quit (Ping timeout: 255 seconds) | |
| [14:39:18] | moparisthebest (moparisthebest!~mitb@gateway/tor-sasl/moparisthebest) has joined #mythtv-users | |
| [14:55:49] | Easy_Rider9999 (Easy_Rider9999!~Miranda@pD9E5DCE4.dip0.t-ipconnect.de) has joined #mythtv-users | |
| [15:11:10] | davidbrooke (davidbrooke!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has joined #mythtv-users | |
| [15:50:47] | sceo-laptop (sceo-laptop!~cwells@cpe-184-153-133-210.maine.res.rr.com) has joined #mythtv-users | |
| [16:06:16] | Jordack (Jordack!~Jordack@75-151-31-172-Michigan.hfc.comcastbusiness.net) has joined #mythtv-users | |
| [16:21:14] | superm1 (superm1!uid4318@ubuntu/member/superm1) has joined #mythtv-users | |
| [16:23:25] | RedPenguin (RedPenguin!~RedPengui@207-255-52-248-dhcp.jst.pa.atlanticbb.net) has quit (Ping timeout: 250 seconds) | |
| [16:25:26] | RedPenguin (RedPenguin!~RedPengui@207-255-52-248-dhcp.jst.pa.atlanticbb.net) has joined #mythtv-users | |
| [16:33:34] | hpeter (hpeter!~hpeter@77-56-33-72.dclient.hispeed.ch) has quit (Ping timeout: 252 seconds) | |
| [16:45:26] | materdaddy__ (materdaddy__!~mmrosko@wsip-70-164-99-62.sd.sd.cox.net) has quit (Read error: Connection reset by peer) | |
| [16:48:20] | davidbrooke (davidbrooke!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has quit (Ping timeout: 265 seconds) | |
| [16:56:01] | Easy_Rider9999 (Easy_Rider9999!~Miranda@pD9E5DCE4.dip0.t-ipconnect.de) has quit (Read error: Connection reset by peer) | |
| [16:58:50] | hpeter (hpeter!~hpeter@77-58-249-169.dclient.hispeed.ch) has joined #mythtv-users | |
| [16:59:11] | davidbrooke (davidbrooke!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has joined #mythtv-users | |
| [17:00:19] | hashbang (hashbang!~isajb@cse-ajb.cse.bris.ac.uk) has left #mythtv-users () | |
| [17:20:27] | grumpydevil (grumpydevil!~rudy@2001:610:73e:0:21b:21ff:fe22:b647) has quit (Ping timeout: 272 seconds) | |
| [17:38:27] | davidbrooke (davidbrooke!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has quit (Ping timeout: 264 seconds) | |
| [17:44:13] | Steve-Goodey (Steve-Goodey!~steve@host109-158-208-74.range109-158.btcentralplus.com) has joined #mythtv-users | |
| [17:48:22] | davidbrooke (davidbrooke!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has joined #mythtv-users | |
| [18:31:07] | ctmjr (ctmjr!~chuck@pdpc/supporter/active/ctmjr) has joined #mythtv-users | |
| [18:36:46] | Freejack (Freejack!~Freejack@68-190-10-237.dhcp.dctr.al.charter.com) has joined #mythtv-users | |
| [18:54:00] | ctmjr (ctmjr!~chuck@pdpc/supporter/active/ctmjr) has quit (Remote host closed the connection) | |
| [18:57:42] | ctmjr (ctmjr!~chuck@pdpc/supporter/active/ctmjr) has joined #mythtv-users | |
| [19:02:14] | ctmjr (ctmjr!~chuck@pdpc/supporter/active/ctmjr) has quit (Remote host closed the connection) | |
| [19:32:24] | jimBaxter (jimBaxter!~jbaxter@jimbax.plus.com) has quit (Quit: Quiting) | |
| [19:33:58] | davidbrooke (davidbrooke!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has quit (Ping timeout: 252 seconds) | |
| [19:34:19] | davidbrooke (davidbrooke!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has joined #mythtv-users | |
| [19:35:03] | grumpydevil (grumpydevil!~rudy@2001:610:73e:0:21b:21ff:fe22:b647) has joined #mythtv-users | |
| [19:44:10] | andreaz (andreaz!~andre_000@p57922A48.dip0.t-ipconnect.de) has joined #mythtv-users | |
| [19:46:12] | sphery_ (sphery_!~mdean@mythtv/developer/sphery) has quit (Quit: leaving) | |
| [19:46:36] | sphery (sphery!~mdean@mythtv/developer/sphery) has joined #mythtv-users | |
| [19:46:36] | Mode for #mythtv-users by ChanServ!ChanServ@services. : +v sphery | |
| [20:01:25] | esperegu (esperegu!~quassel@ip-213-124-221-170.ip.prioritytelecom.net) has joined #mythtv-users | |
| [20:04:38] | esperegu: | how is the number of encoders defined? I have 4 cards with 2 record options per card but instead of 8 encoders I suddenly got 16. I suppose something changed and wondered how it now is working..... |
| [20:19:10] | justinh: | if you're using trunk builds... maybe it's different. I think the multirec code is undergoing big changes |
| [20:19:38] | justinh: | but you're right.. it should be number of cards * number of simultaneous recordings (default 2) |
| [20:32:49] | devinheitmueller (devinheitmueller!~devinheit@cpe-74-71-114-159.nyc.res.rr.com) has joined #mythtv-users | |
| [20:33:29] | devinheitmueller: | “All the RPis (Mark 1 and 2) have the OpenMax video API (Which is what |
| [20:33:30] | devinheitmueller: | XMBC uses). Is it just a matter of adapting that (for which the XBMC |
| [20:33:31] | devinheitmueller: | guys wrote a command-line video player to help with their integration, |
| [20:33:32] | devinheitmueller: | it's up on github somewhere) for MythTV, or am I missing something?” |
| [20:33:38] | devinheitmueller: | ^ Understatement of the month. |
| [20:36:07] | devinheitmueller: | Yeah, just add support for OpenMAX! No problem! |
| [20:47:43] | superm1 (superm1!uid4318@ubuntu/member/superm1) has quit (Quit: Connection closed for inactivity) | |
| [20:50:52] | hpeter (hpeter!~hpeter@77-58-249-169.dclient.hispeed.ch) has quit (Quit: hpeter) | |
| [20:56:03] | hpeter (hpeter!~hpeter@77-58-249-169.dclient.hispeed.ch) has joined #mythtv-users | |
| [20:58:21] | hpeter (hpeter!~hpeter@77-58-249-169.dclient.hispeed.ch) has quit (Client Quit) | |
| [20:59:02] | Jordack (Jordack!~Jordack@75-151-31-172-Michigan.hfc.comcastbusiness.net) has quit (Quit: My mother never saw the irony in calling me a son-of-a-bitch.) | |
| [21:01:50] | justdave (justdave!~dave@unaffiliated/justdave) has quit (Read error: Connection reset by peer) | |
| [21:07:45] | justdave (justdave!~dave@unaffiliated/justdave) has joined #mythtv-users | |
| [21:12:19] | ctmjr (ctmjr!~chuck@pdpc/supporter/active/ctmjr) has joined #mythtv-users | |
| [21:22:01] | clever: | devinheitmueller: i have attempted to add omx support to mplayer |
| [21:22:24] | clever: | the main issue, is that omx tries to do everything, from decoding to display, and even the container if you let it |
| [21:23:16] | clever: | there is no simple way to feed in h264 and get frames out, and the only way to do that is async, ffmpeg doesnt play nicely with the async nature |
| [21:25:01] | clever: | devinheitmueller: it would almost be simpler to bypass ffmpeg entirely, and for mythtv to just run the omx sample code bare, but then you still have to deal with av sync, pause, seeking, and gui overlay |
| [21:25:56] | davidbrooke (davidbrooke!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has quit (Ping timeout: 245 seconds) | |
| [21:26:28] | davidbrooke (davidbrooke!~quassel@c-174-50-212-175.hsd1.ky.comcast.net) has joined #mythtv-users | |
| [21:28:46] | clever: | i think the best project to look at for proper ffmpeg support is the VLC/omx integration, they had frame handles fully working which would fit ffmpeg the best |
| [21:38:19] | Quantum` (Quantum`!~quassel@gateway/tor-sasl/quantum/x-07081428) has joined #mythtv-users | |
| [21:39:24] | Quantum`: | Will Myth compile with Qt5? |
| [21:41:32] | devinheitmueller: | clever: Yeah, I’m intimately familiar with OMX. I’ve done a bunch of gStreamer work using it. |
| [21:41:37] | devinheitmueller: | Agreed, it’s a PITA. |
| [21:42:46] | devinheitmueller: | Also doesn’t help that every OMX implementation works a little different. |
| [21:45:29] | Steve-Goodey (Steve-Goodey!~steve@host109-158-208-74.range109-158.btcentralplus.com) has quit (Quit: Konversation terminated!) | |
| [21:45:44] | clever: | main issue i ran into, is that i tried to cheat the omx and ffmpeg api, my h264 frame decode function would nop a few times |
| [21:46:03] | clever: | and then begin returning frames from about -15 frames ago, so all input came out late |
| [21:46:21] | clever: | ffmpeg then tried to do b-frame re-ordering, after i shifted things, and it just came out a total mess |
| [21:46:45] | devinheitmueller: | Yeah, the initial nop was probably because it was waiting for the first GOP. |
| [21:46:46] | clever: | now that i think of it, omx may have also does the b-frame re-ordering (the cause of the huge latency) |
| [21:47:14] | clever: | so it was fixing the b-frame order twice, and the 2nd pass was offset and shuffling the wrong frames about |
| [21:47:37] | devinheitmueller: | I would be surprised if BCM’s OMX hardware did re-ordering. That’s usually done by software. Possible that they baked it into the OMX stack above the silicon though. |
| [21:47:48] | clever: | yeah |
| [21:48:01] | clever: | the firmware blob |
| [21:48:13] | clever: | plus the added cpu usage to copy everything to arm, between several arm buffers, and then back to dispmanx at the end |
| [21:48:13] | devinheitmueller: | Latency in decode is usually a byproduct of too much prebuffering feeding the decoder. |
| [21:49:06] | clever: | the VLC fix looks the best, get a frame handle for the future decoding result, and then pass that around in ffmpeg, claiming its a custom pixel format |
| [21:49:22] | devinheitmueller: | Yeah, the usual suspects. Doing it to “work” usually isn’t too bad, but getting it to zero-copy can be a real b***ch. |
| [21:49:23] | Quantum`: | No one's tried compiling Myth with Qt5? |
| [21:49:24] | clever: | and then after it goes thru all of the ffmpeg buffering, OMX is done, and you can spend that handle in render |
| [21:49:49] | devinheitmueller: | You might stand a better chance using the native API rather than OMX. That’s ultimately what the VLC dev did. |
| [21:50:01] | clever: | devinheitmueller: zero-copy seems to be the whole point of the omx chain, its passing a pool of buffers back and forth between gpu modules |
| [21:50:04] | clever: | so it doesnt have to copy at all |
| [21:50:04] | devinheitmueller: | (I discussed it at length with him at VideoLAN Dev Days a couple of months ago) |
| [21:50:10] | clever: | but ffmpeg doesnt play with that nicely |
| [21:50:21] | clever: | native api? |
| [21:50:37] | devinheitmueller: | Yeah, there’s a native alternative to the OMX interface. That’s what VLC is using nowadays. |
| [21:50:43] | clever: | ah |
| [21:51:01] | clever: | havent heard anything about that |
| [21:51:17] | devinheitmueller: | The whole point of pipelines is to support zero-copy, but all too often it doesn’t actually work. In many cases because certain modules don’t support operating on the buffers. |
| [21:51:40] | devinheitmueller: | I’ve got a fun case where I wanted to do H.264 decode via OMX, then rotation in software, then output via OMX. The two buffer copies makes the whole thing a non-starter. |
| [21:51:48] | Quantum` (Quantum`!~quassel@gateway/tor-sasl/quantum/x-07081428) has left #mythtv-users ("http://quassel-irc.org - Chat comfortably. Anywhere.") | |
| [21:52:06] | clever: | i think dispmanx supports rotation |
| [21:52:09] | clever: | and yuv420 |
| [21:52:41] | devinheitmueller: | In the case of the Pi, that’s likely to be the case. |
| [21:52:52] | devinheitmueller: | (my OMX work was on a different platform) |
| [21:52:58] | clever: | i had started with a -vo dispmanx module for mplayer, which was setup to limit itself to 1 copy, but the stuff i learned in v3d could get that down to zero copy |
| [21:54:14] | devinheitmueller: | “Someone wrote the crystalhd decoder because they wanted it.” |
| [21:54:22] | devinheitmueller: | ^ Unfortunately a bit of a misleading statement. |
| [21:55:12] | clever: | devinheitmueller: do you know about the threadx malloc stuff? |
| [21:55:25] | clever: | how it keeps heap fragmentation down without an mmu |
| [21:55:57] | devinheitmueller: | Not really. |
| [21:56:31] | clever: | basicaly, all memory allocation are refered to by an opaque token |
| [21:56:34] | devinheitmueller: | I can speak in generalities about how heap managers try to minimize fragmentation (typically through pools), but don’t really no anything about threadx. |
| [21:56:45] | clever: | and the OS is free to defrag the heap by moving data arround all it wants |
| [21:57:07] | clever: | when you want to operate on data, you lock the token, which returns the current physical address |
| [21:57:16] | clever: | and when your done, you unlock the token, and its free to move it again |
| [21:57:55] | devinheitmueller: | A clever trick, if you’re the only one who ever accesses the memory. I would imagine that can cause some real headaches though if you have any structures containing pointers. |
| [21:58:19] | clever: | you would also replace all pointers with opaque tokens |
| [21:58:26] | clever: | and repeat the same trick while following them |
| [21:58:35] | devinheitmueller: | Wouldn’t surprise me though if there were simple use cases where it was applicable though. |
| [21:58:48] | clever: | the entire palmOS did the same thing |
| [21:58:57] | devinheitmueller: | Sounds like a nightmare trying to track down obsolete references. |
| [21:59:04] | clever: | all apps shared the ram, and it had to defrag it to keep large free blocks |
| [21:59:11] | clever: | https://github.com/raspberrypi/firmware/wiki/ . . . e-mem-handle |
| [21:59:21] | clever: | this mailbox call will return the opaque token for a dispmanx image resource |
| [22:00:13] | clever: | https://github.com/raspberrypi/firmware/wiki/ . . . #lock-memory |
| [22:00:16] | devinheitmueller: | Ok, sure. I can see such a use case working when you’re really talking about large data structures representing things like video buffers. It’s one of the reason that Windows did everything with Handles. |
| [22:00:23] | clever: | and this api will lock it and return the current physical address |
| [22:00:49] | clever: | using those 2 calls, i was able to get the current location of a buffer, and directly render to it |
| [22:01:15] | clever: | if you then mmap that phys address somehow, and tie it into the DRM of mplayer, you can get zero-copy rendering |
| [22:01:25] | devinheitmueller: | Ok, so this isn’t really a malloc replacment. It’s a generic construct for video memory referencing. Make sense. |
| [22:01:59] | clever: | i think any large object on the GPU firmware is allocated with that, to keep fragmentation under control |
| [22:02:25] | devinheitmueller: | Gotcha. |
| [22:02:34] | devinheitmueller: | Yeah, I’ve seen similar constructs for other platforms. |
| [22:03:03] | clever: | palmos did it because every app had to share the limited physical ram without an mmu |
| [22:03:14] | clever: | fragmentation would basicaly lead to allocate failure globaly |
| [22:03:15] | devinheitmueller: | In many cases it’s not just about minimizing fragmentation. It’s about cache flushing when the memory is shared between different cores. |
| [22:03:48] | devinheitmueller: | For example, you do a bunch of stuff on the memory with the ARM core. You want to make sure the data gets flushed before instructing the GPU core to operate on the memory. |
| [22:04:01] | clever: | yeah, the gpu also has 2 VC4 cores |
| [22:04:10] | clever: | the arm/gpu stuff is done entirely different |
| [22:04:20] | devinheitmueller: | The alternative is to use completely uncached memory, which s**ks for performance. |
| [22:04:24] | clever: | there are 2 copies of every buffer, and the omx api has a function to just copy it across the gap |
| [22:05:14] | clever: | ive also looked into the kernel code that implements that, it has to decode the paging tables, and convert the 1 buffer refernece (virt addr+size) into an array of phys addr+size |
| [22:05:18] | devinheitmueller: | I would be surprised if it were really doing a copy. More likely the same memory is addressible by both cores, and it’s mostly a matter of coordination. |
| [22:05:21] | clever: | because the MMU can fragment your buffer in phys ram |
| [22:05:38] | clever: | both sides do have unlimited access to all ram |
| [22:06:17] | devinheitmueller: | An odd design choice if that’s what they’re realy doing. Other platforms I’ve seen simply require preallocation of contiguous memory. Much simpler assuming you’re prepared to set aside the memory at startup (which most embedded platforms will) |
| [22:07:00] | clever: | pretty sure all of the input buffers for the h264 bitstreams are allocated with plain old malloc in c++ |
| [22:07:22] | devinheitmueller: | That isn’t too surprising. The compressed bitstream isn’t where the real I/O happens anyway. |
| [22:07:28] | clever: | yep |
| [22:07:38] | devinheitmueller: | A few megabits/second compared to tens of Mb/sec. |
| [22:07:47] | clever: | and you are right, when i was doing the v3d drivers, i spent a lot of time on the kernel driver to allocate contiguous phys memory and map it to the userspace apps |
| [22:07:52] | ctmjr (ctmjr!~chuck@pdpc/supporter/active/ctmjr) has quit (Remote host closed the connection) | |
| [22:08:51] | clever: | my plan was to hide all phys memory addresses from userspace, so you purely work on tokens that refer to chunks of memory |
| [22:11:42] | pfeffer (pfeffer!~pfeffer@ip-62-143-159-185.hsi01.unitymediagroup.de) has joined #mythtv-users | |
| [22:11:56] | pfeffer: | hi |
| [22:12:22] | pfeffer: | i am facing problems with mythfilldatabase |
| [22:12:54] | pfeffer: | the problem is that the .xmltv file is missing |
| [22:13:04] | pfeffer: | how can I create it? |
| [22:13:36] | pfeffer: | 2015-02–02 23:09:32.035123 I XMLTV config file is: /media/Daten/Roberts/.mythtv/de_duesseldorf_epgdata.xmltv |
| [22:13:36] | pfeffer: | 2015-02–02 23:09:33.241524 E FillData: XMLTV grabber returned error code 1 |
| [22:15:46] | amessina (amessina!~amessina@unaffiliated/amessina) has joined #mythtv-users | |
| [22:25:35] | disputin (disputin!~sean@65.102.58.17) has joined #mythtv-users | |
| [22:26:39] | stuartm: | pfeffer: run the xmltv script with --configure |
| [22:32:36] | sceo-laptop (sceo-laptop!~cwells@cpe-184-153-133-210.maine.res.rr.com) has quit (Quit: sceo-laptop) | |
| [22:34:12] | pfeffer: | stuartm: do you mean, for example: " tv_grab_eu_epgdata --configure"? |
| [22:35:07] | pfeffer: | stuartm: that one does not create the "de_duesseldorf_epgdata.xmltv" |
| [22:35:21] | pfeffer: | stuartm: which is missing |
| [22:35:30] | stuartm: | --config-file .mythtv/de_duesseldorf_epgdata.xmltv |
| [22:35:52] | stuartm: | see the --help output for more |
| [22:38:55] | pfeffer: | stuartm: ahh! that looks much better! :-) |
| [23:03:33] | Tobbe5178 (Tobbe5178!~asdf@2001:2002:d9d4:ce7e:5976:6a1b:4757:c3a6) has quit (Read error: Connection reset by peer) | |
| [23:08:36] | justinh: | hahaha. I was just wondering how long it'd take before there was a users ML thread about it. The pi 2 |
| [23:08:57] | disputin (disputin!~sean@65.102.58.17) has quit (Quit: Leaving.) | |
| [23:11:30] | pfeffer: | stuartm: thank you so much! – it seems to be working |
| [23:17:51] | devinheitmueller: | justinh: I like the guy who recommended using the Qt Multimedia framework, but then acknowledged that he had no experience using it. |
| [23:33:09] | nfgdrummer211 (nfgdrummer211!4904ed1b@gateway/web/freenode/ip.73.4.237.27) has joined #mythtv-users | |
| [23:33:26] | nfgdrummer211 (nfgdrummer211!4904ed1b@gateway/web/freenode/ip.73.4.237.27) has quit (Client Quit) | |
| [23:39:53] | materdaddy (materdaddy!~mmrosko@wsip-70-164-99-62.sd.sd.cox.net) has joined #mythtv-users | |
| [23:40:16] | mcmoyer (mcmoyer!~textual@99-32-86-245.lightspeed.ftwotx.sbcglobal.net) has joined #mythtv-users | |
| [23:42:51] | pigeon_ is now known as pigeon | |
| [23:43:30] | sceo-laptop (sceo-laptop!~cwells@cpe-184-153-133-210.maine.res.rr.com) has joined #mythtv-users | |
| [23:48:15] | pmmd (pmmd!~pmmd@2a00:1630:2:f00::4518:f271) has quit (Ping timeout: 276 seconds) | |
| [23:53:56] | pmmd (pmmd!~pmmd@2a00:1630:2:f00::4518:f271) has joined #mythtv-users | |
| [23:54:54] | disputin (disputin!~sean@65.102.58.17) has joined #mythtv-users | |
| [23:55:20] | disputin1 (disputin1!~sean@65.102.58.17) has joined #mythtv-users | |
| [23:55:42] | disputin (disputin!~sean@65.102.58.17) has quit (Read error: Connection reset by peer) | |
| [23:56:38] | joki (joki!~joki@p5486178C.dip0.t-ipconnect.de) has quit (Ping timeout: 245 seconds) | |
IRC Logs collected by
BeirdoBot.
Please use the above link to report any bugs.