*** tpb has joined #timvideos | 00:00 | |
kflux | it's okay. i have some rudimentary experience with Linux build environments, so if someone could point me into the right direction about how to update the firmware of my opsis, that'd be already very helpful. | 00:00 |
---|---|---|
mithro | kflux: The "VideoMixerSoC" doesn't actually do any video mixing at the moment, it is more of a "VideoMatrixSoC" | 00:01 |
mithro | kflux: ssk1328 started the work needed to do color correct video mixing as a GSoC student last year | 00:02 |
mithro | kflux: We are hoping ot merge that work after we finish the current major refactor | 00:02 |
mithro | kflux: The stuff for the refactor is going on at https://github.com/mithro/HDMI2USB-litex-firmware/ | 00:02 |
tpb | Title: GitHub - mithro/HDMI2USB-litex-firmware: Opsis SoC based on LiteX (at github.com) | 00:02 |
mithro | kflux: If you want to set up a build environment, I'd recommend using the instructions at https://github.com/mithro/HDMI2USB-litex-firmware/blob/nextgen/scripts/README.md for now | 00:03 |
tpb | Title: HDMI2USB-litex-firmware/README.md at nextgen · mithro/HDMI2USB-litex-firmware · GitHub (at github.com) | 00:03 |
mithro | kflux: Where in the world are you located? | 00:03 |
kflux | mithro: berlin | 00:03 |
mithro | kflux: Did we meet at 33C3? | 00:05 |
kflux | mithro: no unfortunately i couldn't come to 33C3 | 00:06 |
mithro | kflux: Damn, I was in Berlin for ~3 days after NYE, should have tried to meet up | 00:06 |
kflux | mithro: :) | 00:07 |
kflux | mithro: ok. so i'll try to set up a dev env for the opsis on my ubuntu 16.10 and flash the latest firmware. the one that came with the opsis seems a bit buggy. | 00:08 |
kflux | mithro: or is there any other way to flash the latest one? | 00:09 |
mithro | kflux: The HDMI2USB-mode-switch tool can be used to flash some of the prebuilt firmware, but we don't have any prebuilt firmware for the nextgen stuff yet | 00:09 |
mithro | jea: Have you switched to looking at the streaming-system stuff yet? | 00:10 |
mithro | kflux: https://github.com/timvideos/HDMI2USB-mode-switch | 00:10 |
tpb | Title: GitHub - timvideos/HDMI2USB-mode-switch: Tool for switching boards supported by HDMI2USB firmware between multiple different modes (programming, webcam, etc). (at github.com) | 00:10 |
kflux | mithro: allright. but the latest prebuilt one should be better than the one the opsis came with, right? | 00:11 |
mithro | kflux: It has a /different/ set of issues :-P -- fixing the major ones have required the nextgen stuff | 00:12 |
jea | mithro: about to do streaming now | 00:12 |
kflux | mithro: ok. thanks a lot for the info and updates. | 00:13 |
mithro | kflux: Really need more developers, so much we want to do but so little time to do it.... | 00:14 |
jea | micolous: you might also find this useful http://pastebin.ubuntu.com/23789653/ | 00:35 |
tpb | Title: Ubuntu Pastebin (at pastebin.ubuntu.com) | 00:35 |
micolous | Hmmm... you should really consider writing a patch for modemmanager | 00:40 |
micolous | I've done it before for another drivce where it was overly zealous | 00:40 |
micolous | They're rpretty responsive too :) | 00:41 |
mithro | tumbleweed: https://launchpad.net/~timvideos/+archive/ubuntu/fpga-support | 00:43 |
tpb | Title: Support packages for FPGA development : “TimVideos.Us - Live event streaming” team (at launchpad.net) | 00:43 |
mithro | micolous: We have udev rules which should prevent modem-manager from opening the opsis and stuff | 00:43 |
micolous | Yeah, just referring to jea's paste | 00:44 |
micolous | But you should upstream them (tm) | 00:44 |
micolous | Just trying to find the list of USB devices again | 00:44 |
mithro | You want to configure your computer with 192.168.100.100 | 00:44 |
mithro | micolous: https://github.com/timvideos/HDMI2USB-mode-switch/tree/master/udev | 00:45 |
tpb | Title: HDMI2USB-mode-switch/udev at master · timvideos/HDMI2USB-mode-switch · GitHub (at github.com) | 00:45 |
mithro | _florent_: What does --gdb-passthrough do with flterm? | 00:47 |
micolous | Yeah, looks like I had a lot of the old development environment on here, I just nuked ISE tools as they were 20GB | 00:48 |
*** hyades has quit IRC | 00:51 | |
*** nrossi has joined #timvideos | 01:01 | |
*** CarlFK has joined #timvideos | 01:04 | |
*** ChanServ sets mode: +v CarlFK | 01:04 | |
CarlFK | micolous: https://github.com/lca2017/symposion/blob/master/symposion/schedule/views.py#L191-L246 | 01:05 |
tpb | Title: symposion/views.py at master · lca2017/symposion · GitHub (at github.com) | 01:05 |
CarlFK | micolous: https://github.com/CarlFK/veyepar/blob/master/dj/scripts/addeps.py#L1021-L1080 | 01:08 |
tpb | Title: veyepar/addeps.py at master · CarlFK/veyepar · GitHub (at github.com) | 01:08 |
mithro | _florent_: So, I think I've got your underflow counter merged, the freq counter merged and the fx2 i2c_hack merged | 01:08 |
jea | micolous: https://github.com/xfxf/lca2017-av/issues/22 | 01:31 |
tpb | Title: Import the schedule from LCA into the streaming system · Issue #22 · xfxf/lca2017-av · GitHub (at github.com) | 01:31 |
mithro | _florent_: Your frequency counter appears to be returning weird results here... | 01:38 |
mithro | _florent_: input1: 0x0 (@ 74.67 MHz) | 01:38 |
mithro | Also... | 01:39 |
mithro | _florent_: input0: 0x0 (@ 0. 0 MHz) | 01:39 |
mithro | 0.<space>0 MHz :-P | 01:40 |
jea | mithro, _florent_: I'll fix the thing above | 01:54 |
*** puck`` is now known as puck | 01:56 | |
mithro | jea: The space bit, not the frequency counter reporting 74MHz when nothing connected - right? | 02:13 |
jea | mithro: yes, the space bit | 02:15 |
jea | Shopping list: https://github.com/xfxf/lca2017-av/issues/31 | 02:20 |
tpb | Title: Shopping List 13 Jan · Issue #31 · xfxf/lca2017-av · GitHub (at github.com) | 02:20 |
*** kflux has quit IRC | 02:20 | |
*** Hammy has joined #timvideos | 02:31 | |
Hammy | Hey Tim! | 02:31 |
Hammy | Bunnie used ther Serdes in 4-bit mode, where you can attach two serdes to a pin, allowing you to tune the I/O delays to work out the best phase. Then has to go through an 8 to 10 gearbox. | 02:33 |
*** Hammy has quit IRC | 02:44 | |
*** CarlFK has quit IRC | 03:07 | |
tumbleweed | we're online again \o/ | 03:07 |
*** CarlFK has joined #timvideos | 03:14 | |
*** ChanServ sets mode: +v CarlFK | 03:14 | |
mithro | https://opsis.hdmi2usb.tv/img/cypress-fx2/i2c.png | 03:57 |
mithro | _florent_: | 03:57 |
_florent_ | mithro: for the frequency counter, I had the same behaviour with the new and the old one | 04:19 |
_florent_ | mithro: not sure I broke something | 04:19 |
_florent_ | mithro: at least I don't think so | 04:20 |
mithro | Okay | 04:35 |
xfxf | http://pastebin.com/mbCkX6wU | 04:40 |
tpb | Title: "IODisplayEDID" = <00ffffffffffff00061019a00000000030160104a5211578026fb - Pastebin.com (at pastebin.com) | 04:40 |
jea | xfxf: details from that EDID http://pastebin.com/vJpi4uVx | 04:42 |
tpb | Title: Extracted contents: header: 00 ff ff ff ff ff ff 00 serial number: - Pastebin.com (at pastebin.com) | 04:42 |
mithro | jea: Is that the same as the other black magic ones? | 04:43 |
mithro | jea: Need a 720p60 modeline from that :-P | 04:43 |
jea | mithro: I'll check | 04:43 |
mithro | jea: Are you upstairs with micolous working on streaming stuff? | 04:44 |
jea | mithro: yeah | 04:45 |
micolous | Yeah | 04:45 |
micolous | Comfier chairs | 04:45 |
micolous | And you won't infect us :P | 04:45 |
vystrela | is mithro sick again? | 04:46 |
mithro | Still? | 04:46 |
jea | mithro: it is a different EDID. this one has extra modes in it, mainly for different types of 1920x1080 | 04:49 |
mithro | That makes some sense | 04:49 |
mithro | We need to get edid decoding into the firmware :-P | 04:49 |
jea | it also has valid checksums, unlike the other one! | 04:49 |
jea | mithro: raised this for that https://github.com/mithro/HDMI2USB-litex-firmware/issues/14 | 04:51 |
tpb | Title: Add EDID decode to firmware · Issue #14 · mithro/HDMI2USB-litex-firmware · GitHub (at github.com) | 04:51 |
jea | will do it after lca i think | 04:52 |
jea | let me extract a 720p60 line | 04:52 |
jea | mithro: their detailed 720p60 line matches our built in one | 04:55 |
mithro | Okay | 04:56 |
mithro | So just setting "m 9" should be the exact same signal? | 04:56 |
jea | yep | 04:57 |
jea | and they also have equivalent of our "m 10" in the CEA extension block detailed timings | 04:57 |
*** tvCommitBot has joined #timvideos | 05:05 | |
tvCommitBot | [streaming-system] micolous pushed 1 new commit to master: https://git.io/vM2TM | 05:05 |
tvCommitBot | streaming-system/master b2ca4bc Michael Farrell: Add symposion schedule scraper | 05:05 |
*** tvCommitBot has left #timvideos | 05:05 | |
micolous | oh great, the data for the miniconfs is crazy | 05:28 |
micolous | they have this "kind" field which there's special kinds unique to each miniconf stream | 05:29 |
* micolous gets out a little black book | 05:29 | |
CarlFK | micolous: I'll send you the side channel email | 05:33 |
CarlFK | mithro: "license": "CC BY", https://github.com/lca2017/symposion/blob/master/symposion/schedule/views.py#L213 | 05:34 |
tpb | Title: symposion/views.py at master · lca2017/symposion · GitHub (at github.com) | 05:34 |
*** nrossi has quit IRC | 05:40 | |
mithro | xfxf / tumbleweed: https://github.com/xfxf/lca2017-av/issues/33 | 05:44 |
tpb | Title: Add `rmmod uvcvideo; modprobe uvcvideo` to the vocto capture script... · Issue #33 · xfxf/lca2017-av · GitHub (at github.com) | 05:44 |
mithro | USB-C - YAY! - http://ewen.mcneill.gen.nz/blog/entry/2016-12-22-dell-xps-9360-hdmi-output/ | 05:56 |
tpb | Title: Dell XPS 13 (9360): HDMI Output (at ewen.mcneill.gen.nz) | 05:56 |
mithro | micolous / jea: How goes things? I could use someone to put together a quick Python script for me (to test firmware on the Opsis) | 06:19 |
*** hyadez has quit IRC | 06:22 | |
jea | mithro: streaming is basically working locally now. just need to fix the irc part, but that can be done later | 06:24 |
mithro | jea: \o/ | 06:25 |
mithro | How is the schedule importing stuff going? | 06:25 |
mithro | The streaming-system really wants no gaps in the schedule | 06:25 |
micolous | schedule was sorta working, then I found more crazy edge cases when I got the mobile programme going | 06:28 |
micolous | 06:28 | |
micolous | https://github.com/micolous/lca-mobile-programme/blob/lca2017/schedule.js#L228-L249 | 06:28 |
tpb | Title: lca-mobile-programme/schedule.js at lca2017 · micolous/lca-mobile-programme · GitHub (at github.com) | 06:29 |
jea | mithro / micolous - I presume the streaming system just needs to have a schedule for the full day for each room? | 06:39 |
jea | is there any good way to see the full schedule that the system has, or not? | 06:39 |
micolous | I normally test it by hacking some part of the code to change what the current time is | 06:40 |
*** tvCommitBot has joined #timvideos | 06:40 | |
tvCommitBot | [streaming-system] micolous pushed 1 new commit to master: https://git.io/vM2qa | 06:40 |
tvCommitBot | streaming-system/master 1757c7a Michael Farrell: symposion: drop out garbage entries | 06:40 |
*** tvCommitBot has left #timvideos | 06:40 | |
micolous | ^^ that should get rid of some junk entries | 06:40 |
micolous | And also fix a couple of the miniconf streams intros | 06:40 |
mithro | xfxf: Did you see this? https://github.com/xfxf/lca2017-av/issues/32 | 06:44 |
tpb | Title: Investigate the segfaults in `libv4l2.so` that are occuring on the machines · Issue #32 · xfxf/lca2017-av · GitHub (at github.com) | 06:44 |
mithro | jea / micolous: I need a script which does something like "make clean; make gateware; make gateware-load; checks on the serial port that the firmware booted; save build with status somewhere; repeat" | 06:50 |
jea | mithro: ok. what is the best way to check on the serial port? and what counts as saving a build? | 06:52 |
mithro | jea: Moving the directory somewhere with a name liked "build-12-failing" | 06:52 |
jea | mithro: the entire build directory? and do you just want failing builds, or all builds? | 06:53 |
mithro | jea: All builds | 06:54 |
jea | ok | 06:54 |
mithro | jea: For the serial - I think probably using an external ftdi IC is the best idea - which means it should stay around even when the Opsis restarts and stuff | 06:55 |
jea | mithro: um, ok. where do i find one of these? | 06:55 |
mithro | I have one setup here | 06:55 |
jea | ok, i'll come down | 06:56 |
mithro | xfxf: https://github.com/mithro/HDMI2USB-litex-firmware/blob/nextgen/firmware/ci.c#L275 | 06:59 |
tpb | Title: HDMI2USB-litex-firmware/ci.c at nextgen · mithro/HDMI2USB-litex-firmware · GitHub (at github.com) | 06:59 |
mithro | xfxf: That is the status code | 06:59 |
mithro | xfxf: https://github.com/mithro/HDMI2USB-litex-firmware/blob/nextgen/firmware/ci.c#L380 | 06:59 |
tpb | Title: HDMI2USB-litex-firmware/ci.c at nextgen · mithro/HDMI2USB-litex-firmware · GitHub (at github.com) | 06:59 |
*** nrossi has joined #timvideos | 07:00 | |
*** hyades has joined #timvideos | 07:09 | |
mithro | _florent_: https://github.com/mithro/HDMI2USB-litex-firmware/blob/mithro-is-stupid/gateware/opsis_i2c.py and https://github.com/mithro/HDMI2USB-litex-firmware/blob/mithro-is-stupid/gateware/i2c.py | 07:18 |
tpb | Title: HDMI2USB-litex-firmware/opsis_i2c.py at mithro-is-stupid · mithro/HDMI2USB-litex-firmware · GitHub (at github.com) | 07:18 |
*** _florent_ has quit IRC | 07:36 | |
*** _florent_ has joined #timvideos | 07:37 | |
mithro | _florent_: You can see the way I have changed things at https://github.com/mithro/HDMI2USB-litex-firmware/commit/5aff4a1e0744c8da3b9e2928aa01114055306310 | 07:37 |
tpb | Title: Refactoring TOFE/Opsis stuff a bit. · mithro/HDMI2USB-litex-firmware@5aff4a1 · GitHub (at github.com) | 07:37 |
*** Niharika has quit IRC | 07:37 | |
mithro | _florent_: Did https://github.com/m-labs/misoc/blob/master/misoc/cores/i2c.py get merged into litex now? | 07:41 |
tpb | Title: misoc/i2c.py at master · m-labs/misoc · GitHub (at github.com) | 07:41 |
*** Niharika has joined #timvideos | 07:41 | |
xfxf | mithro: https://www.alibaba.com/product-detail/Thin-mini-itx-Mini-PC-Case_60287315701.html | 08:06 |
tpb | Title: Thin Mini Itx Mini Pc Case A02-h56 In Stock Aluminum Alloy,Horizontal,For Embedded Computer/braeborn System/htpc/mini Desktop - Buy Mini Itx Aluminum Case,Thin Mini Itx Case,Aluminum Case Product on Alibaba.com (at www.alibaba.com) | 08:06 |
mithro | xfxf: I think you'll find these are bigger then they look | 08:17 |
xfxf | i'm getting 'RUNTIME>' as my hdmi2usb prompt now (my changes appear to work though), did i do a stupid or is that expected? | 08:18 |
mithro | xfxf: https://goo.gl/photos/zZUyxo9Zr763Nh8p8 | 08:19 |
mithro | xfxf: That is expected | 08:19 |
xfxf | ok, cool | 08:19 |
mithro | xfxf: I forgot we changed the prompt, been such a long time since that happened :) | 08:20 |
xfxf | has 'debug on' changed, or did i stuff something up in the command parsing? | 08:22 |
jea | xfxf: I suspect it has changed. I don't see it in help, but I haven't checked the code yet | 08:39 |
*** CarlFK has quit IRC | 08:41 | |
tumbleweed | jea: https://github.com/m-labs/milkymist/blob/master/tools/flterm.c | 08:42 |
tpb | Title: milkymist/flterm.c at master · m-labs/milkymist · GitHub (at github.com) | 08:42 |
*** CarlFK has joined #timvideos | 09:05 | |
*** ChanServ sets mode: +v CarlFK | 09:05 | |
jea | mithro: this is a script that might work http://pastebin.ubuntu.com/23791391/ | 09:42 |
tpb | Title: Ubuntu Pastebin (at pastebin.ubuntu.com) | 09:42 |
*** hyades has quit IRC | 09:51 | |
jea | mithro: what device does the opsis show up as? /dev/...? | 10:57 |
mithro | If you have udev rules installed | 10:57 |
mithro | then /dev/hdmi2usb/by-num/opsis0/.... | 10:57 |
jea | mithro: not sure if i have the udev rules. i don't see any opsis at the moment, not sure if that could be to do with the TOFE board being plugged in? | 10:58 |
micolous | mithro: I have a chromeos device here if you want to test | 11:18 |
mithro | micolous: Oh, yes! Do you want to come plug it into an Opsis and see if it spazs out? | 11:19 |
micolous | is 'make flash' supposed to be missing? | 11:42 |
micolous | docs, and the makefile itself, refers to it as a valid target, but it appears to be missing | 11:43 |
CarlFK | tumbleweed: how/where did you implement tims mplayer/vlc/all the things? | 12:02 |
CarlFK | seems like a good place to put python-wxgtk3.0 and my clocky? | 12:02 |
mithro | micolous: It will reappear shortly | 12:10 |
mithro | micolous: There is an flash-opsis command | 12:10 |
micolous | ok | 12:10 |
micolous | I'm doing some tweaks to the pattern generator | 12:10 |
mithro | micolous: We can load a pattern from your computer using the ethernet now too | 12:11 |
micolous | I'm going to make those git variables all defines instead | 12:11 |
micolous | So they're easier to use in other places | 12:11 |
micolous | I've also improved the font rendering a little bit by adding some whitespace | 12:12 |
mithro | micolous: Which git variables? | 12:12 |
micolous | the ones in version_data.c | 12:12 |
jea | mithro: have you tried out the ipython etherbone script yet? | 12:12 |
mithro | micolous: Please don't make them defines, they were changed to the way they are now because otherwise linking/rebuilding doesn't work as you would expect | 12:13 |
micolous | k | 12:14 |
mithro | micolous: Otherwise anything which uses the version stuff gets rebuilt *every* run | 12:14 |
mithro | micolous: They are globals though, so you should be able to to refer to them | 12:14 |
micolous | I was trying to avoid sprintf | 12:15 |
mithro | build/opsis_XXXXX_lm32/software/firmware/version_data.h | 12:15 |
mithro | micolous: Trying to put the git revision info in the pattern output? :-P | 12:15 |
micolous | yep | 12:15 |
micolous | It's working. | 12:16 |
mithro | micolous: Yay :-) | 12:20 |
micolous | http://imgur.com/0tVgO0N | 12:22 |
tpb | Title: Imgur: The most awesome images on the Internet (at imgur.com) | 12:22 |
micolous | Note that all the text is drawn with a little extra whitespace as well | 12:23 |
micolous | I notice when the text reaches the green bar, it had a really harsh edge on it | 12:23 |
micolous | But then I cleaned that up a bit | 12:23 |
micolous | I tried the chromebook... looks like it is only happy at 1024x768 | 12:28 |
micolous | ...and 800x600 (m 3) and 640x480 (m 0) | 12:29 |
micolous | Looks like it is completely ignoring EDID information from the opsis | 12:30 |
xfxf | micolous: it's a google chromebook, fix it :P | 12:43 |
micolous | xfxf: http://imgur.com/h5yjWY3 | 13:14 |
tpb | Title: Imgur: The most awesome images on the Internet (at imgur.com) | 13:14 |
xfxf | i prefered the rickroll | 13:14 |
micolous | lol | 13:14 |
xfxf | can we make it read random emails off the presenters computer and display them? | 13:15 |
micolous | I'll just include a QR code generator... | 13:15 |
micolous | :facepalm: | 13:15 |
micolous | I added colour! | 13:15 |
micolous | Now when the firmware crashes in the middle of the conference, it's a great advertising opportunity :P | 13:16 |
micolous | If I recall correctly, some people photographed hdmi2usb units and shared them on social media | 13:16 |
micolous | When they were at the colour bar screen | 13:16 |
micolous | Now I just want font scaling | 13:16 |
jea | https://chromium.googlesource.com/chromiumos/platform/factory/+/master/py/probe/functions/edid.py | 13:22 |
tpb | Title: py/probe/functions/edid.py - chromiumos/platform/factory - Git at Google (at chromium.googlesource.com) | 13:22 |
micolous | jea: cheers | 13:23 |
jea | what it is doing will work, just | 13:23 |
jea | interestingly it is just looking at the resolution, nothing else | 13:23 |
jea | so i have no idea how it knows what dotclock to use, etc | 13:24 |
jea | there is this part too https://chromium.googlesource.com/chromiumos/platform/factory/+/master/py/gooftool/edid.py | 13:37 |
tpb | Title: py/gooftool/edid.py - chromiumos/platform/factory - Git at Google (at chromium.googlesource.com) | 13:37 |
jea | I also like this look of this, but I have a feeling it was never included from the comments | 13:50 |
jea | https://bugs.chromium.org/p/chromium/issues/detail?id=395947 | 13:50 |
jea | if it is the right licence, then it might be worth using for edid.tv | 13:50 |
jea | mithro: https://github.com/joeladdison/HDMI2USB-litex-firmware/commit/362811237008f81a883fca0f722edb4a559debd3 | 13:53 |
tpb | Title: Add scripts for running gateware build non-stop · joeladdison/HDMI2USB-litex-firmware@3628112 · GitHub (at github.com) | 13:53 |
jea | that is your script for doing builds | 13:53 |
*** CarlFK has quit IRC | 13:59 | |
*** Doralitze has joined #timvideos | 14:47 | |
*** Doralitze has quit IRC | 14:54 | |
*** hyades has joined #timvideos | 15:37 | |
*** hyadez has joined #timvideos | 17:31 | |
*** Doralitze has joined #timvideos | 17:48 | |
*** hyades has quit IRC | 17:51 | |
*** sb0 has quit IRC | 19:09 | |
*** nrossi has quit IRC | 19:30 | |
*** CarlFK has joined #timvideos | 19:37 | |
*** ChanServ sets mode: +v CarlFK | 19:37 | |
CarlFK | anyone awake? | 20:03 |
*** kflux has joined #timvideos | 21:13 | |
*** hyades has joined #timvideos | 21:15 | |
*** Doralitze has quit IRC | 21:42 | |
tumbleweed | CarlFK: yes | 21:43 |
tumbleweed | I thought I heard you knocking, but it was just the wind, I think | 21:43 |
*** sb0 has joined #timvideos | 22:03 | |
micolous | I'm awake | 22:09 |
tumbleweed | the door is open (for now, until the gale force winds come back) | 22:35 |
CarlFK | im toasting! | 22:43 |
tumbleweed | o_O | 22:43 |
tumbleweed | good news: all the machines seem to have made it through the night | 22:43 |
CarlFK | yay | 22:46 |
CarlFK | now burning raisin bread | 22:47 |
*** CarlFK has quit IRC | 23:07 | |
*** CarlFK has joined #timvideos | 23:21 | |
*** ChanServ sets mode: +v CarlFK | 23:21 | |
*** hyades has quit IRC | 23:21 | |
*** kflux has quit IRC | 23:59 |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!