Wednesday, 2017-11-01

*** tpb has joined #timvideos00:00
CarlFK[m]I have 4 files?  what do I do next?  [email protected]:~/$ ls00:06
CarlFK[m]firmware.bin  firmware.fbi  top.bin  top.bit00:06
CarlFK[m]Bus 003 Device 005: ID 04e2:1410 Exar Corp. XR21V1410 USB-UART IC00:06
CarlFK[m]Bus 003 Device 004: ID 1443:0007 Digilent Development board JTAG00:06
tumbleweedload-gateware with top.bit, load-firmware with firmware.bin, load the fx2 firmware, and see if it works00:13
tumbleweed(without powercycling)00:13
*** aps has quit IRC00:14
*** jea has quit IRC00:23
*** jea has joined #timvideos00:24
CarlFK[m]FileNotFoundError: [Errno 2] No such file or directory: 'openocd': 'openocd'00:30
CarlFK[m]"everything you need" isn't everything I need :p00:31
tumbleweedI guess you conda-ed, but didn't add the conda stuff to your PATH?00:31
CarlFK[m]I followed the readme :p00:34
CarlFK[m]sudo ~/HDMI2USB-mode-switch/conda/bin/hdmi2usb-mode-switch -v --load-gateware top.bit00:36
CarlFK[m]that seemed to work00:36
CarlFK[m]load-firmware - --load-softcpu-firmware ?00:36
CarlFK[m]    raise NotImplemented("Not yet finished...")00:37
CarlFK[m]TypeError: 'NotImplementedType' object is not callable00:37
CarlFK[m]I suspect  I need flterm00:38
tumbleweedI suspect so too00:38
tumbleweed--kernel firmware.bin00:38
CarlFK[m]what about --kernel ?00:39
tumbleweedit's an argument to flterm00:39
* tumbleweed is actually going to go home now, and can poke at an atlys too00:39
CarlFK[m]oh.. but mode swich should be doing that?00:39
tumbleweedmodeswitch doesn't know how to do that, as you can see00:40
CarlFK[m]well, it wasn't installed the first time00:40
CarlFK[m]but yeah, installed, same error00:40
CarlFK[m]where did this come from? /dev/ttyXRUSB000:46
mithromode-switch can't "load" softcpu stuff yet - it can however flash it...00:47
tpbTitle: GitHub - scanlime/flipsyfat: FPGA-based emulator to assist with guessing bootloader SD card filenames (at
CarlFK[m]After flashing, the board will need to be power cycled.00:58
CarlFK[m]um.. do I do that or (07:12:02 PM) tumbleweed (IRC): (without powercycling)00:58
CarlFK[m]Ill do the "without"00:59
CarlFK[m] sudo ~/HDMI2USB-mode-switch/conda/bin/hdmi2usb-mode-switch -v --load-fx2-firmware firmware.bin00:59
CarlFK[m]subprocess.CalledProcessError: Command '['fxload', '-t', 'fx2lp', '-D', '/dev/bus/usb/003/006', '-I', '/home/juser/', '-v']' returned non-zero exit status 254.01:00
tumbleweedOK, home with an atlys...01:10
tumbleweedCarlFK[m]: hrm, worked for me01:13
CarlFK[m]what did?01:17
CarlFK[m]I need to run to dinner for about an hour - any interst in sshing to my box?01:17
tumbleweedI mean, it boots, and video is borked01:18
tumbleweedso, I have to keep digging01:19
CarlFK[m]did you see I ansibled  a r&d box?01:19
tumbleweednope, but yay?01:19
tpbTitle: video-stack-deploy/roles/rnd at master · CarlFK/video-stack-deploy · GitHub (at
tumbleweedthere's some nasty stuff there (I mean, the things it has to do)01:22
tumbleweedbut yay :)01:22
CarlFKbb in an hour or so01:26
* tumbleweed tries git bisection again. We have to dig into the nextgen branch...01:30
mithrotumbleweed: No guarantee you can the old nextgen stuff :-P02:26
CarlFK[m]mithro No guarantee  ... what?02:42
mithrocan build02:42
tumbleweedif it takes > 1hr: give up, right?02:55
CarlFK[m]tumbleweed (IRC): is the cpu spinning?02:56
tumbleweedtrying to route02:56
mithrotumbleweed: On the older firmware before nextgen >2h02:56
CarlFK[m]I thought it went idle.  which made no sense.02:56
tumbleweedI was suprised how fast the nextgen build was03:01
CarlFK[m]can you walk me though how you tested ?03:05
tpbTitle: Index of /stefanor/tmp/atlys-16341a1 (at
CarlFK[m]low priority if you are actively trying to .. do just about anything03:05
tumbleweedhdmi2usb-mode-switch --load-gateware top.bit03:11
tumbleweedflterm --port /dev/ttyVIZ0 --kernel firmware.bin03:12
tumbleweedtype serialboot03:12
tumbleweed/sbin/fxload -t fx2lp -I ~/git/HDMI2USB-firmware-prebuilt/archive/master/v0.0.2-81-gd0d3aea/atlys/hdmi2usb/hdmi2usb.hex -D /dev/hdmi2usb/by-num/atlys0/usbdev03:12
tumbleweedthat was about it03:12
CarlFK[m]ah, I did everything with mode-switch03:44
tpbTitle: : autoEdit (at
mithro"Fast text based video editing"04:54
mithroBunch of extra info at
tpbTitle: An open source tool for enabling faster, easier editing of video interviews - Vox Product Blog (at
CarlFK[m]that's inginious05:09
mithroAnyone tried siteleaf before?
*** sb0 has quit IRC09:38
*** sb0 has joined #timvideos09:38
*** sb0 has quit IRC10:29
*** aps has joined #timvideos10:45
*** rohitksingh has joined #timvideos11:24
*** sb0 has joined #timvideos11:41
shorneI created a pr for some useful stuff13:42
tpbTitle: Support for in.tftpd + qemu GDB by stffrdhrn · Pull Request #11 · mithro/HDMI2USB-litex-firmware · GitHub (at
*** rohitksingh1 has joined #timvideos14:28
*** rohitksingh has quit IRC14:30
cr1901_modernmithro: Packages arrived15:01
rohitksingh1mithro: ping15:17
*** aps has quit IRC15:55
*** sb0 has quit IRC16:18
*** sb0 has joined #timvideos16:49
*** CarlFK has quit IRC17:29
mithrorohitksingh1: pong!17:48
mithrorohitksingh1: I'm in US/Pacific timezone currently so it might be hard to get IRC working.17:49
mithrorohitksingh1: feel free to send me an email17:50
rohitksingh1mithro: oh okay, sure!17:51
*** CarlFK has joined #timvideos18:22
*** ChanServ sets mode: +v CarlFK18:22
tumbleweedCarlFK[m]: want to try another image?18:39
CarlFKtumbleweed: sure thing18:40
tpbTitle: Index of /stefanor/tmp/atlys-840e634e (at
*** rohitksingh1 has quit IRC19:08
CarlFK[m]weird. things don't work well plugged into a usb3 port.  im confused.20:04
CarlFK[m]weird. things don't work well plugged into a usb3 port.  im confused.20:04
CarlFKtwice even20:10
tumbleweedso confused20:10
tumbleweedI have been able to crash USB ports on my machine before, when hacking on hdmi2usb hardware20:11
tumbleweedtry suspend & resume, or even reboot20:11
CarlFK[email protected]:~/bin$ dmesg|grep "xr_usb_serial_set_reg Error" |wc20:31
CarlFK     34     204    258420:31
CarlFKI plugged it into usb2 and its working20:31
CarlFK[m]tumbleweed (IRC): for testing your stuff, do I still do:   /sbin/fxload -t fx2lp -I ~/git/HDMI2USB-firmware-prebuilt/archive/master/v0.0.2-81-gd0d3aea/atlys/hdmi2usb/hdmi2usb.hex -D /dev/hdmi2usb/by-num/atlys0/usbdev20:36
tumbleweedCarlFK[m]: yes20:48
* CarlFK[m] adds 60 - [timvideos, HDMI2USB-firmware-prebuilt]20:50
CarlFK[m]I'm surprised I didn't have that already20:51
CarlFK[m]tumbleweed (IRC): um.. in one term I have: BIOS> serialboot21:04
CarlFK[m]Booting from serial... ... Telnet listening on port 2321:04
tpbTitle: Ubuntu Pastebin (at
CarlFK[m]in a 2nd:  [email protected]:~/$ /sbin/fxload -t fx2lp -I ~/HDMI2USB-firmware-prebuilt/archive/master/v0.0.2-81-gd0d3aea/atlys/hdmi2usb/hdmi2usb.hex -D /dev/hdmi2usb/by-num/atlys0/usbdev21:05
CarlFK[m]it just returned to a prompt v.quick.  is that expected ?21:05
CarlFK[m]tumbleweed (IRC): uptime 44 min!!!21:18
CarlFK[m]sorry, all the stuff is on a different floor, so had to run up and down stairs to see it21:19
* tumbleweed was out for coffee21:35
tumbleweedCarlFK[m]: is the video over USB corrupted?21:35
CarlFK[m]um.. I have to look up a viewer :p21:46
tumbleweedmplayer tv://1 -tv device=/dev/video121:46
CarlFK[m]tumbleweed (IRC): looks like it :(21:51
CarlFK[m]nothing but this over and over [mjpeg @ 0x7feb57764560]No JPEG data found in image21:51
CarlFK[m]Error while decoding frame!21:51
tumbleweedCarlFK[m]: no, it should look corrupted but play fine21:52
tumbleweedCarlFK[m]: encoder on?21:52
CarlFK[m]um.. I have half forgotten  how to test, and half just doing whatever you say :p21:53
tumbleweedtype encoder on, then21:53
tumbleweedyou reminded me that it was off by default in thepast21:53
CarlFK[m]um.. I still have this flterm prompt21:54
tumbleweedyeah, in there21:54
CarlFK[m]oh, I guess that is my prmpt21:54
CarlFK[m]RUNTIME>x c p e21:55
CarlFK[m]Connecting pattern to encoder21:55
CarlFK[m]that turns it on, right?21:55
* CarlFK[m] runs down staris again 21:55
tumbleweedand encoder on21:55
CarlFK[m]damm, didnt' see that in time21:56
CarlFK[m]RUNTIME>encoder on21:56
CarlFK[m]Enabling encoder21:56
CarlFK[m]back down stairs21:56
* CarlFK[m] uploaded an image: file1509573347605.jpg (6966KB) <>21:57
* CarlFK[m] uploaded an image: file1509573393539.jpg (7208KB) <>21:58
tumbleweedmithro: that basically means that the atlys has been broken since atlys support was added to nextgen22:16
mithroThat would have been my guess22:17
tumbleweed840e634e is only about 8 commits after atlys support was added22:17
tumbleweedand it built (earlier ones I tried didn't)22:17
tumbleweedso, I'm back to being interested about strategies in testing22:18
xfxfDon’t update the gateware? ;)22:20
* xfxf ducks22:20
CarlFKI don't understand something.    it generates a pattern that goes into a frame buffer (is that correct term here?)22:27
CarlFKthat is used to generate?  a signal to send to hdmi out - which works fine22:28
CarlFKit is also encoded to a jpeg - which I am guessing is where the problem is, but I don't see how this is different between atlys and opsis22:29
tumbleweedthe problem we are seeing here is cleary in the encoder or USB stack or the interfaceb etween them22:29
CarlFKI have trouble understanding how the usb stack could cause problem without making the jpeg invalid22:33
CarlFKARG!!! I made the ansible tweeks to the pxe branch22:37
CarlFKon neat. I made a new branch to pu them in.22:40
CarlFK[m]and still mucked it up.  bunch of mv's ended up in the wrong branch.23:01
*** CarlFK has quit IRC23:38

Generated by 2.13.1 by Marius Gedminas - find it at!