Saturday, 2015-10-17

*** tpb has joined #timvideos00:00
*** xfxf1 has quit IRC01:18
mithroseaLne: my primary task today once I finish doing "chores" is to get the serial uart working on the Opsis when the HDMI2USB video firmware is loaded01:53
mithro_florent_ / CarlFK / xfxf: what was the outcome of the input issue?01:54
mithroG33KatWork: it would be good to figure out why your version of ISE is having problems compared to our versions. IE I've never been able to get ISE to timeout locally.02:07
*** travis-ci has joined #timvideos03:48
travis-ci[mithro/HDMI2USB/opsis-bringup#93] (c4babc3): The build passed. (
*** travis-ci has left #timvideos03:48
*** travis-ci has joined #timvideos03:51
travis-ci[mithro/HDMI2USB/opsis-bringup#92] (72854b1): The build passed. (
*** travis-ci has left #timvideos03:51
mithroI'm sure I figured out what the code here is doing ->
tpbTitle: HDMI2USB/app.c at master · timvideos/HDMI2USB · GitHub (at
mithroBut I can't seem to find any comment about it :/05:02
mithroHrm - I think it was
tpbTitle: fx2lib/cdc.h at cdc-usb-serialno-from-eeprom · mithro/fx2lib · GitHub (at
*** se6astian|away is now known as se6astian08:39
*** se6astian has left #timvideos08:40
*** Bertl is now known as Bertl_zZ09:26
*** se6astian has joined #timvideos10:24
seaLnemithro: sounds good, let me know if there is anything i can help test10:27
*** se6astian is now known as se6astian|away10:54
mithro_florent_: the system should romboot if there are errors with the memtest right?11:05
mithrowell, third time today I've screwed up my Linux USB stack causing me to need to reboot11:47
*** rohitksingh has joined #timvideos11:51
_florent_mithro: yes, it should only show the bios prompt11:53
mithro_florent_: so, if I connect to the device, and it is at the HDMI2USB> prompt, then memtest must have passed atleast once?11:53
mithro_florent_: so, I'm seeing that behavior on my opsis12:04
_florent_OK, I think there are issues when rebooting because some DMA are still activated12:05
mithro_florent_: that should be previous obvious in a verbose memtest right?12:05
mithro_florent_: It should be easy to see the DDR having an effect if the memtest output the malformed memory location+value ?12:09
_florent_yes, but the first thing to do is to desactivate all DMAs in the reboot command12:12
mithro_florent_: didn't we try that?12:12
_florent_not me, have you done it?12:13
mithro_florent_: btw - are you see the same issue locally after a reboot?12:13
_florent_IIRC it was working correctly here12:14
mithro_florent_: well - I turned off all the inputs+outputs before rebooting?12:14
mithroseaLne: as another person with an Opsis, do you want to see if you can reproduce?12:14
seaLnemithro: ok, what am i to try?12:20
_florent_seaLne: start the board, get the HDMI2USB> prompt, type "reboot"12:20
_florent_and see if you have errors in the memtest12:20
mithroseaLne: "make gateware; make load-gateware; opsis-mode-switch --mode=serial; make connect-lm32"12:21
seaLnewhich version of gateware?12:21
_florent_you can try the last12:21
mithroyou'll need my opsis-fixes branch for the serial port to work12:22
mithroseaLne: your handy with electronics stuff right?12:27
mithroseaLne: any chance you've got time to do a footprint for myself and Bertl?12:28
seaLnewhat of?12:28
mithroseaLne: we need it for a Opsis->TOFE<->TOFE->Opsis bridge and a Opsis->TOFE<->axiom bridge12:30
seaLnemonday ok?12:31
* seaLne is trying to remember where he hacked to change to using ACM112:32
seaLneah just Makefile.lm3212:32
seaLnereboot seem ok for me
tpbTitle: KDE Paste (at
mithroI have the CDC working on the Opsis12:34
mithro(in the HDMI2USB firmware)12:34
seaLnew00t :)12:34
mithro_florent_: if you can confirm that also works on your Opsis, I think it means there is something up with mine12:35
mithro_florent_: guess I should try the memtest which ships with the Xilinx EDK on the board and see if that fails too12:36
_florent_yes you can, but if you memtest is always passing OK the first time, it's probably an issue in the gateware/firmware not on the board12:39
mithro_florent_: oh true....13:00
*** travis-ci has joined #timvideos13:17
travis-ci[sealne/HDMI2USB-misoc-firmware/download-prebuilt#58] (cb96670): The build is still failing. (
*** travis-ci has left #timvideos13:17
*** travis-ci has joined #timvideos14:09
travis-ci[sealne/HDMI2USB-misoc-firmware/download-prebuilt#58] (cb96670): The build is still failing. (
*** travis-ci has left #timvideos14:09
seaLnemithro: would there be any issue with removing the make clean between make firmware and make gateware? it causes prebuilt upload in .travis/ to fail as it has deleted hdmi2usb.hex before trying to upload if after my more complete cleaning14:13
seaLnerelated to #11614:14
mithroseaLne: hrm?14:14
mithroseaLne: "make firmware; make clean; make gateware; make firmware; make upload" is what it effectively does, right?14:15
tpbTitle: Travis CI - Test and Deploy Your Code with Confidence (at
seaLnemake gateware doesn't seem to be redoing make firmware?14:18
mithroseaLne: not it shouldn't14:18
mithros/not/no, /14:18
seaLnedo you want the clean to make sure make gateware isn't relying on anything in make firmware?14:20
mithroseaLne: yes14:20
seaLnek that makes sense14:20
seaLneso the problem is me now deleting the .hex in make clean14:21
mithroseaLne: The problem is that it doesn't run "make firmware" again after the "make gateware"14:22
mithroseaLne: we want to run "make firmware" first because its fast and lets us see errors quickly14:22
seaLneit doesn't seem very elegant to have to rerun make firmware just to get one file to upload14:23
_florent_lm32-firmware is integrated in the gateware, so we need to build it first (but I'm not sure you are talking about that :)14:24
mithroseaLne: running "make firmware" before "make gateware" is just an early warning thing - if you delete that line everything should continue to work14:24
mithro_florent_: the gateware target is dependent on the lm32-firmware, hence that will work fine14:25
seaLnewill that still leave a generated hdmi2usb.hex file though? it doesn't seem to14:25
mithroseaLne: no, the gateware doesn't require the fx2 firmware14:28
seaLneah sorry getting firmwares mixed up in my head14:30
mithrowell, I have a fx2 firmware which lets you use the serial and webcam functionality at the same time14:31
mithrobut it also sometimes causes the linux usb stack to hard lock and require a reboot....14:31
CarlFKmithro: go to bed ;)14:39
CarlFKmithro: but first ... any idea where I can post questions about youtube api things?  like "can I get the status of just one video, not a ilst?"  because the docs lead me here:14:41
CarlFK  "Step 2: Check the video's status       Call the videos.list method to check the video's status."14:41
CarlFKwhich seems cumbersome.14:41
mithroseaLne: got a moment to try and replicate this?14:46
mithroOtherwise, I'm heading home to bed :P14:47
mithroseaLne: well, checkout my latest branch and "make load-fx2" should give you a UVC camera and CDC device which both works14:53
mithroseaLne: there are some weird quirks that I've yet to figure out14:53
mithroseaLne: like the first time you start the video stream you can't be connected to the serial port14:53
mithroseaLne: but after that it works fine !?14:53
mithroseaLne: and occasionally it'll lock up the USB stack.... so make sure you are okay with rebooting before running it14:54
mithroseaLne: the Opsis will appear as14:54
mithro[  574.141519] usb 3-4: New USB device found, idVendor=2a19, idProduct=544214:54
mithro[  574.141525] usb 3-4: New USB device strings: Mfr=1, Product=2, SerialNumber=314:55
mithro[  574.141529] usb 3-4: Product: HDMI2USB14:55
mithro[  574.141533] usb 3-4: Manufacturer: JANI14:55
mithro[  574.141536] usb 3-4: SerialNumber: ffff001ec0f1419b14:55
mithroThat serial number should equal the MAC address in your board14:55
mithro(IE it won't be the same as mine)14:55
mithrowell, I'm heading home + to bed!14:55
mithroCarlFK: btw - did you see my comments on the issue around input1 stuff?14:56
mithroIt's surprisingly warm outside for 2am15:01
CarlFKmithro: input1 - yes.  it works, but something was odd, like it didn't work at first.  but I think it was my hdmi source being dumb.  so I want to spend a little time before I give it a pass15:17
CarlFKmithro: can you check your Atlys jumpers against the picture at the bottom of
tpb<> (at
*** Bertl_zZ is now known as Bertl15:45
*** Bertl is now known as Bertl_oO16:23
*** travis-ci has joined #timvideos16:46
travis-ci[mithro/HDMI2USB-misoc-firmware/opsis-fixes#25] (ed90ee1): The build has errored. (
*** travis-ci has left #timvideos16:46
CarlFKtumbleweed: are you alive yet?17:39
*** se6astian|away is now known as se6astian18:36
*** rohitksingh has quit IRC20:34
*** se6astian is now known as se6astian|away21:11

Generated by 2.13.1 by Marius Gedminas - find it at!