Saturday, 2017-07-22

*** tpb has joined #timvideos00:00
jeaafter a bit of digging, we think that it is using a different codec00:00
jeai'm going to work out what i need to pass to the ingest to get it to work00:01
jearight, i think i know00:03
jeavoctomix-ingest uses format=S16LE, but I think we need format=S32LE00:03
jeaso now I need to work out where that is being set00:03
CarlFKcore's ini00:04
CarlFKingest  gets the settings from core00:05
jeayes, I see it00:05
jeai'm guessing that needs a core restart?00:06
jeaok. next break00:06
CarlFKwhat is requiring format=S32LE ?00:07
jeathe usb audio device is outputting 32bit00:07
CarlFKhuh.  well, good news is changing that over here broke mine, I think the same:00:09
CarlFKError-Details: #1: gstbasesrc.c(2950): gst_base_src_loop (): /GstPipeline:pipeline0/GstAlsaSrc:audiosrc:00:10
jeayep, that is what I get00:14
jeaso looks like we have narrowed it down :D00:14
jeait wouldn't be a conference without something breaking00:14
CarlFKvys: where did you get hdmi2usb-mode-switch and openocd ?00:23
jeai got it by running the full install for the build environment from hdmi2usb-litex-firmware00:25
CarlFKgross :p00:25
jeait is a bit00:26
jeabut at least it should be 'documented' somewhere in the install script...00:26
CarlFKthe what script ?00:27
jeaconda install openocd=$OPENOCD_VERSION00:27
tpbTitle: HDMI2USB-litex-firmware/ at master · timvideos/HDMI2USB-litex-firmware · GitHub (at
jeawhich i run from the instructions here
tpbTitle: HDMI2USB-litex-firmware/ at master · timvideos/HDMI2USB-litex-firmware · GitHub (at
jeabut that is a full dev environment for the firmware00:28
CarlFKI have tried that enough times :p00:28
tpbTitle: Home · CarlFK/HDMI2USB-mode-switch Wiki · GitHub (at
CarlFKthat is spozed to work00:30
CarlFKmithro: symlinks00:33
jeatwice in one day :P00:34
CarlFK[m]has break happned?01:26
mithroCarlFK: where are the docs for the voctomix / ansible stuff?02:02
CarlFK[m]mithro (IRC):
tpbTitle: video-stack-deploy/ at master · CarlFK/video-stack-deploy · GitHub (at
CarlFK[m]mithro: do you have current goal of setting up a box, or just wondering what's going on?02:10
jeaCarlFK[m]: yep, break happened, and it is working now :)02:10
jeastill some slight issues, but they are minor02:11
CarlFK[m]jea: \o/02:11
jeaI think we should have a full project for docs of both the ansible and the hdmi2usb02:12
CarlFK[m]both you and xfxf started right off by ignoring the docs...02:14
CarlFK[m]"make a usb stick"  no, im doing this other thing...02:14
jeai did make a usb stick, and followed the instructions02:14
jeai also made modiciatoins to make it work for these machines02:15
CarlFK[m]you didn't use the usb stick02:15
jeabut my point is that everyone needs to work on docs, not just mithro for hdmi2usb02:15
vys100% agree02:15
jeaand i am happy to contribute to both02:17
tpbTitle: Home · CarlFK/HDMI2USB-mode-switch Wiki · GitHub (at
CarlFK[m]get that text somewhere formal02:17
CarlFK[m] seems like a good place02:18
CarlFK[m]don't hit the Edit button, it doesn't work.02:19
vysjea has quite a lot of kit now, I would be happy to spend a day every few weekends setting up a bunch of shtuff from scratch to ensure everything works as expected, and iron out any effects from strange things02:20
vyslike changing the service from root to normal user, which should be fine, but working out that it couldn't find some log file, so it created it as root, and then the user version can't open it, but it doesn't handle failing to open a log etc etc02:21
vysI had some dodgy fedora version of mithros bootstrap stuff, but then figured docker might be ok for doing things02:23
vysthere are things to investigate, see if they have any drawbacks, or if they allow anything exciting02:23
mithroCarlFK: I have a bunch of laptops I would like to setup for recording kits.02:31
mithroCarlFK: but I'm also frustrated at your demands for docs02:32
mithroCarlFK: and you are complaining "you guys are doing things not the way I set things up" when you are doing the exact same thing to me. I have very good docs for setting up the dev environment.02:33
CarlFK[m]I have tired the dev env docs, they don't work.02:34
mithroWhere is the bug report for that?02:34
mithroI know of a few small issues I need to fix, but no idea what you are stuck on.02:35
CarlFK[m]mostly I don't want a dev envirment - I just want to flash from prebuilt02:36
mithroWhich I want to support, but it's a lower priority - specially since nobody seems to be contributing to that02:37
mithroCarlFK: I know you can write python, I know you can improve mode-switch02:38
CarlFK[m]mithro: there is a bit of a chicken and egg thing here02:38
mithroI also know you can write docs02:39
CarlFK[m]it starts with you making a symlink02:39
CarlFK[m]so that the docs that use is stand a chance at working02:40
mithroWhy don't you contribute some code to mode-switch which downloads the pre-built firmware, then that doesn't even matter02:42
CarlFK[m]a) time b) I don't think anyone knows what it should be downloading02:43
mithroStart with some hard coded URLs02:43
CarlFK[m]see A02:43
mithroIf that makes it easier02:43
mithroCarlFK: I'm pretty time limited too02:44
CarlFK[m]I am hoping that the symlink thing is 'quick'02:45
CarlFK[m]if it is a problem, then lets have a wiki page of what URL we should be using02:45
mithroCarlFK: I'm not sure what you think the Domino symlink thing is going to get you?02:45
CarlFK[m]I want this to go to the build you want me to test # cd HDMI2USB-firmware-prebuilt/opsis/firmware/testing02:46
CarlFK[m]you have links in there, they point to things, and you say "don't use that version"02:47
CarlFK[m]I think jea or vys  used cd HDMI2USB-firmware-prebuilt/archive/master/v0.0.3-710-g947321b/opsis/hdmi2usb/lm3202:48
mithroUse the latest version number that works02:49
CarlFK[m]sounds like maintaining the links is too hard/time consuming?02:50
mithroUse branch master unless we have customised the firmware specifically for your conference02:50
vyswhat is this symlink stuff?02:51
mithroI'm going to make the unstable version point to the latest, stable/testing need someone to update manually after actually doing some testing02:51
mithroI may get to that today02:52
CarlFK[m]vys (IRC):
tpbTitle: HDMI2USB-firmware-prebuilt/opsis/firmware at master · timvideos/HDMI2USB-firmware-prebuilt · GitHub (at
mithroBut I need to fix some PCB schematics and get the Opsis build working again first02:53
mithroAnd and like 5 emails02:53
mithroAll of which I'm sure that nobody else can do02:54
mithroAnd a bunch which are blocking people which are actually been sending patches02:55
mithroOh, also need to review GSoC students work and check things for PyCon AU02:56
mithroAnd figure out why the or1k version locks up so I can unblock the Linux porting short02:57
mithroAnd write a talk for PyCon AU02:57
mithroAnd put in a proposal to try to LCA2018 for a Miniconf around the HDMI2USB stuff02:58
mithroAnd a whole bunch of other things that I'm forgetting02:58
mithroAnd I have a full time job which I actually do need to do work for03:00
CarlFK[m]vys (IRC): can you confirm that this is the build you used?  v0.0.3-710-g947321b (which I took from HDMI2USB-firmware-prebuilt/archive/master/v0.0.3-710-g947321b/opsis/hdmi2usb/lm32 which saw here earlier03:02
vysthat is the build03:03
vysCarlFK[m]: ^03:03
vysand then `opsis-mode-switch --verbose --flash-gateware=/home/me/Downloads/flash.bin`03:05
vyswell, I downloaded mine from the github prebuilt web server thing03:07
vysbut that is how I flashed it03:07
vysI did run `opsis-mode-switch --verbose --mode=jtag` and then wait like 60 seconds03:08
vysbecause if you try to flash it too quickly it doesn't work03:10
vysthis info was pulled from /HDMI2USB-litex-firmware/targets/opsis/Makefile.mk03:10
CarlFK[m]mithro: hdmi2usb-mode-switch --flash-gateware flash.bin errors:03:34
CarlFK[m]subprocess.CalledProcessError: Command '['openocd', '-f', 'board/numato_opsis.cfg', '-c', 'init; jtagspi_init 0 /usr/share/hdmi2usb-mode-switch/hdmi2usb/firmware/spartan6/opsis/bscan_spi_xc6slx45t.bit; jtagspi_program /home/juser/temp/HDMI2USB-firmware-prebuilt/archive/master/v0.0.3-710-g947321b/opsis/hdmi2usb/lm32/flash.bin 0x0; exit']' returned non-zero exit status 103:34
tpbTitle: Ubuntu Pastebin (at
CarlFK[m]hdmi2usb-udev is already the newest version (0.0.0+git20161016-1+ppa1).03:35
CarlFK[m]but.. I have not re-installed after people have been doing things on it03:35
mithroCarlFK[m]: I have not use the packages and have no idea if they work03:35
mithroCarlFK[m]: Get the code at then type "make conda"03:36
CarlFK[m]mithro (IRC): looks like it worked: Finished processing dependencies for hdmi2usb.modeswitch==0.0.0.post18403:40
mithroCarlFK[m]: the last output should be a PATH you have to set to run the command03:40
tpbTitle: Ubuntu Pastebin (at
mithroCarlFK[m]: hrm.. "export PATH=$PWD/conda/bin:$PATH" then "which hdmi2usb-mode-switch"03:43
mithroCarlFK[m]: Okay, give that a try03:43
CarlFK[m][email protected]:~/temp/HDMI2USB-firmware-prebuilt/archive/master/v0.0.3-710-g947321b/opsis/hdmi2usb/lm32$ hdmi2usb-mode-switch --flash-gateware flash.bin03:44
CarlFK[m]WARNING:root:unbind-helper not found, will have to run as root!03:44
mithroCarlFK[m]: and?03:44
CarlFK[m]ah, more stuff now...03:45
CarlFK[m]Warn : Bypassing JTAG setup events due to errors03:45
CarlFK[m]Error: Unknown flash device (ID 0x00818181)03:45
mithroyou probably want --verbose03:45
tpbTitle: HDMI2USB-litex-firmware/ at master · timvideos/HDMI2USB-litex-firmware · GitHub (at
tpbTitle: Ubuntu Pastebin (at
mithroCarlFK[m]: ?03:47
tpbTitle: HDMI2USB-litex-firmware/ at master · timvideos/HDMI2USB-litex-firmware · GitHub (at
CarlFK[m]WARNING:root:unbind-helper not found, will have to run as root!03:48
CarlFK[m]should I be sudo all this?03:48
tpbTitle: Check for common issues when flashing · Issue #33 · timvideos/HDMI2USB-mode-switch · GitHub (at
mithroCarlFK[m]: No, I need to fix that warning to only appear when it actually needs root03:49
mithroCarlFK[m]: it was written back before I had another solution for not needing root03:49
mithroCarlFK[m]: Also, the last time we tried this I seem to recall you had installed a bunch of random crap which was fucking with mode-switch03:52
mithroCarlFK[m]: And I've yet to add a "detect Carl's screwed up set up and complain" mode to mode-switch03:53
CarlFK[m]i've let at least one person do fpga things on this box03:54
CarlFK[m]takes about 15 min to reinstall the os - want me to?03:54
mithroCarlFK[m]: if it continues to not work03:55
mithroCarlFK[m]: Can you check for the following before you do03:57
mithroDoes /lib/firmware/ixo-usb-jtag/ exist?03:59
CarlFKls: cannot access '/lib/firmware/ixo-usb-jtag/': No such file or directory04:00
mithroWhat about /etc/udev/rules.d/99-hdmi2usb-permissions.rules ?04:00
mithroCarlFK[m]: And you are testing on an Opsis right? Not an Atlys?04:00
CarlFKyes opsis04:00
CarlFK'/etc/udev/rules.d/99-hdmi2usb-permissions.rules': No such file or directory04:01
CarlFK/etc/udev/rules.d/ is empty04:01
mithroCarlFK[m]: Did you follow the doc I linked above for the error?04:01
CarlFKone of them, im getting ready to wipe the os04:02
mithroCarlFK[m]: Can you do the one I linked above and paste its output?04:03
CarlFK[m]this?  hdmi2usb-mode-switch --mode=serial; hdmi2usb-mode-switch --mode=jtag04:04
mithroCarlFK[m]: Yes -- put a sleep 5 in there please04:04
CarlFK[m]mithro (IRC):
tpbTitle: Ubuntu Pastebin (at
mithroCarlFK[m]: and what happens when you try flash again?04:12
CarlFK[m]is htis the magic "no worky try again value?"  DNA = 110000001100000011000000110000001100000010011101100100010 (0x181818181813b2204:13
mithroCarlFK[m]: hrm...04:13
CarlFK[m]it errored:
tpbTitle: Ubuntu Pastebin (at
mithroCarlFK[m]: okay - so try the serial / jtag switch again04:16
CarlFK[m]mithro (IRC): k - looks the same as last time04:17
vysCarlFK[m]: I was getting that04:21
vysmostly 18s with some reasonable stuff at the end04:21
vysI literally cycled to serial and then back to jtag, and left it for like 60 seconds04:21
vysreplugging the usb cable after putting it in jtag mode worked for me, possibly04:22
*** rohitksingh has joined #timvideos04:25
mithroCarlFK[m]: well, I think I have the symlink thing working05:17
mithroCarlFK[m]: Will know in an hour or two05:17
mithroGoing to get lunch, will be back later05:17
mithrorohitksingh: ping?10:01
rohitksinghmithro: pong10:01
mithrorohitksingh: How goes your USB module?10:02
rohitksinghmithro: Work-in-progress towards pushing to Github :p10:03
mithrorohitksingh: I have something for you!10:07
mithrorohitksingh: As well, how can I order you little adapter thing?10:08
rohitksinghmithro: let me just publish that order from should be in my oshpark account10:09
tpbTitle: USB IDs · timvideos/HDMI2USB Wiki · GitHub (at
mithrorohitksingh: Past Tim got USB IDs for when we eventually got the SoftUTMI stuff working...10:14
mithrorohitksingh: You can use them for your stuff...10:30
mithrorohitksingh: Ordered!10:31
rohitksinghmithro: Awesome! \o/10:49
rohitksinghbtw I'll use 0x1d50/0x60b8 vid/pid (unconfigured device) during testing...although during testing it doesn't really matter which IDs one uses10:50
*** rohitksingh has quit IRC13:15
mithroCarlFK[m] / CarlFK: The unstable symlink is now updating16:41
CarlFK[m]mithro (IRC): yay!16:42
CarlFK[m]mithro (IRC): less than 5 min ago I finely got a new ubuntu box up16:42
mithroThe stable / testing symlinks should be updated manually by people16:42
CarlFK[m]mithro (IRC): does this nick you?  or does the (irc) mess that up16:43
mithroIf you say mithro in any form it will cause me to get pinged16:43
mithroAnyway, I'm out now16:44
CarlFK[m]mithro (IRC): can you give me 10 min to build on my new box?16:45
CarlFK[m]i relaize it is about 3am...16:45
CarlFK[m]mithro (IRC): make conda still dosn't end with the PATH= thing you mentioned - want me to open an issue?16:48
*** cr1901 has quit IRC17:37
*** CarlFK has quit IRC21:20
*** cr1901 has joined #timvideos21:55
*** CarlFK has joined #timvideos23:17
*** ChanServ sets mode: +v CarlFK23:17
*** CarlFK has quit IRC23:18
mithroCarlFK[m]: please do23:24
mithroCarlFK[m]: if you dont open a bug, there is a 90% chance it will get forgotten23:25
*** CarlFK has joined #timvideos23:25
*** ChanServ sets mode: +v CarlFK23:25
mithroCarlFK[m]: now I have the unstable link working, I have a task for you. I would like you to update the testing link any time you successfully test the unstable version. Could you do that for me?23:26
mithroWhen we successfully use a firmware version at an evebt, we should update the stable link23:27
mithro(if we where happy with the firmware performance)23:27
CarlFKkinda - "successfully test" isn't too well defined - unless you are ok with it shows some hint of working23:27
mithroCarlFK: does that make sense?23:27
mithroCarlFK: I'm happy for you to figure out what level you are comfortable with23:28
mithroCarlFK: if you're okay with, I ran a test for 2 mins and it worked with vocto23:29
CarlFKk - that makes sense23:29
mithroCarlFK: if you instead want to wait till you've recorded something at PS1 or something, I'm also happy with that23:30
*** deep-book-gk_ has joined #timvideos23:31
mithroCarlFK: testing should be something we are happy to consider running at our next "big" event with maybe a bit more testing23:31
mithroCarlFK: as you are the person who is most active in doing recording, it would be really good for you to figure out the level your happy with23:32
mithroBut we really need testing to be as close to as unstable as possible23:32
mithroThe further it is out by, the harder it is too fix bugs23:33
mithroCarlFK: can you do that for me?23:33
CarlFKunstable and testing are easy23:33
*** deep-book-gk_ has left #timvideos23:33
CarlFKmithro: lets see if I can actually flash...23:33
CarlFKI think stable is: we have tested it enough we are ready to risk it23:34
mithroStable should be what we are willing to let "other" people use23:35
CarlFKmaybe we need 4... the 4th being 'used it, fall back to this when we discover the other one isn't really that great'23:35
mithroStable should be "fall back to this if testing isn't working"23:35
CarlFKanyway.. I have a show in 5 days and 100 tings teetering on falling over.. like I have no idea where or who is doing what with audio23:36
mithroCarlFK: if you update the link once a month, that would be awesome23:36
mithroCarlFK: development on the firmware is pretty slow at the moment23:37
mithroCarlFK: so, I dub thee, "release manager" ;-)23:37
CarlFKit would be awesome if I can get to the point of updating it 'ever' ;)23:38
mithroCarlFK: I want to get you flashing things23:39
mithroCarlFK: but I do have to run now23:39
CarlFKI am at ps1 with a ubuntu 16.4 and an Atlys...23:39
CarlFKrats - how soon can you be back?23:39
mithroCarlFK: I don't regularly test atlys boards23:39
mithroCarlFK: I'll be in & out today23:40
CarlFKI need an hour window in the next 6 hours to pack up all this crap and bring it home23:40
mithroJust chat at me and I'll reply asap23:40
CarlFKk - I'll get back home asap where the opsis is23:40
mithrovys: might also be able to help23:41
mithroCarlFK: I would like to keep supporting the Atlys until atleast the end of next year23:42
CarlFKmithro: vys: I'll be back in 30 - 60 min.  then I will be at home where the  ubuntu box and Opsis is23:45
mithroCarlFK: I also want to sort out how to flash the FX2 on the Atlys so it operates like a Opsis (ie the firmware is under the FPGA control and you don't need anything on the host)23:45
mithroAnd get the CDC serial port working on the Atlys again so you only need a single USB cable and not the broken exart / vizzini driver23:46
CarlFKmithro: I have an Atlys with one broken usb connector -  does that make it useless for recording ?23:46
CarlFKI guess it depends on which broke23:47
mithroCarlFK: if the broken one is the UART yes, if the prog one, no23:47
CarlFKwe can worry about that later.23:47
mithroWith paddatrapper working on the FX2 stuff now, hopefully we can get that under control23:53
mithroWhich should fix the need to load/unload the UVC module23:55

Generated by 2.13.1 by Marius Gedminas - find it at!