| *** tpb has joined #timvideos | 00:00 | |
| *** rohitksingh has joined #timvideos | 04:27 | |
| mithro | CarlFK: pong - I replied to your email | 05:28 |
|---|---|---|
| CarlFK | mithro: thanks | 05:32 |
| CarlFK | mithro: "Flash the firmware onto the Atlys board" I am liking that | 05:32 |
| mithro | CarlFK: I actually spent the weekend working on your problem | 05:32 |
| CarlFK | lol | 05:32 |
| CarlFK | thank you. | 05:32 |
| mithro | CarlFK: The problem is that nobody has done "Flash the firmware onto the Atlys board" bit before | 05:32 |
| CarlFK | oh. | 05:33 |
| CarlFK | ok, lets skip that | 05:33 |
| mithro | CarlFK: In theory it should be 3 commands to get that working | 05:33 |
| mithro | CarlFK: but nobody has tested it | 05:33 |
| mithro | CarlFK: the find_board.py program I was working on over the weekend scans your system for a Atlys (or Opsis) board in any "state" and then should eventually help you get it into the state you want. | 05:37 |
| CarlFK | mithro: that sounds like what I want | 05:37 |
| CarlFK | mithro: at ps1 when experimenting around with Jay's chromebook, at least once I power cycled the Atlys, then spent what seemed like a min or two getting it back up | 05:38 |
| mithro | CarlFK: If you know you always want the device in a certain configuration, then udev rules are much easier than this script | 05:38 |
| CarlFK | I am hoping that can be cut down to under 30 seconds | 05:38 |
| mithro | CarlFK: because a udev rule is effectively "when device of type X is plugged in, run command Y" | 05:39 |
| mithro | So you end up with 3 udev rules | 05:39 |
| mithro | (a) When Atlys is plugged in, upload the programming firmware | 05:40 |
| mithro | (b) When programming firmware is found, program the device | 05:40 |
| mithro | (c) When programmed device is found, upload the final .hex file | 05:40 |
| CarlFK | mithro: would't that be the same as: power cycld Atlys, run script that does all 3? | 05:41 |
| mithro | CarlFK: no - because you have to write the machinery which does the "detect the programming has finished" and "wait for the device to come back" and then "figure out where the hell linux put it" stuff | 05:42 |
| CarlFK | ah I think I understand | 05:43 |
| mithro | Where as udev does all that auto-magically for you | 05:43 |
| CarlFK | so udev takes care of the timing and some ID's get passed in | 05:43 |
| mithro | CarlFK: yeah | 05:44 |
| mithro | CarlFK: so - I think the first step is getting flcli onto your machines | 05:45 |
| CarlFK | mithro: where/what is flcli? | 05:53 |
| mithro | https://github.com/makestuff/flcli | 05:54 |
| tpb | Title: makestuff/flcli · GitHub (at github.com) | 05:54 |
| mithro | CarlFK: you want the first half of https://gist.github.com/makestuff/7486988 I think | 05:54 |
| tpb | Title: Build FPGALink and the cksum VHDL example, and use flcli to write data to an Atlys board, and then read data back from it. · GitHub (at gist.github.com) | 05:54 |
| CarlFK | mithro: I have pealed that apart.. just a sec... | 05:55 |
| mithro | This part | 05:55 |
| mithro | https://www.irccloud.com/pastebin/pCP3SrBJ/ | 05:55 |
| tpb | Title: Pastebin | IRCCloud (at www.irccloud.com) | 05:55 |
| CarlFK | mithro: any idea why that release? there are 2 more recent ones: https://github.com/makestuff/flcli/releases | 06:00 |
| tpb | Title: Releases · makestuff/flcli · GitHub (at github.com) | 06:00 |
| mithro | CarlFK: get the latest release | 06:01 |
| mithro | I just googled flcli and clicked the first link :P | 06:01 |
| CarlFK | wget -qO- http://tiny.cc/msbil ... ../scripts/msget.sh makestuff/flcli <- very convoluted way of making some dirs and doing some git clone | 06:02 |
| mithro | CarlFK: I was trying to see if there is a way to compile flcli as a static brinary | 06:02 |
| mithro | CarlFK: it pulls dependencies | 06:03 |
| CarlFK | mithro: I think it is just: | 06:03 |
| CarlFK | git clone https://github.com/makestuff/common.git | 06:03 |
| CarlFK | mkdir apps; cd apps | 06:03 |
| tpb | Title: makestuff/common · GitHub (at github.com) | 06:03 |
| CarlFK | git clone https://github.com/makestuff/flcli.git | 06:03 |
| CarlFK | git clone https://github.com/makestuff/fx2loader.git | 06:03 |
| tpb | Title: makestuff/flcli · GitHub (at github.com) | 06:03 |
| tpb | Title: makestuff/fx2loader · GitHub (at github.com) | 06:03 |
| CarlFK | (and a few more empty dirs that the tar file create) | 06:04 |
| mithro | Using the upstream tar will continue to work as he moves things around | 06:04 |
| CarlFK | bleck - pull a copy local and get tumbleweed to package it :D | 06:05 |
| mithro | CarlFK: I've been trying to convince someone to package makestuff things for a while | 06:09 |
| CarlFK | mithro: looks like I got it to build: | 06:16 |
| CarlFK | juser@dc10b:~/temp/apps/flcli$ ./lin.x64/rel/flcli -h | 06:16 |
| CarlFK | FPGALink Command-Line Interface Copyright (C) 2012-2014 Chris McClelland | 06:16 |
| mithro | CarlFK: great! | 06:17 |
| CarlFK | sec... let me confirm | 06:17 |
| mithro | SUBSYSTEM=="usb", ACTION=="add", ATTRS{idVendor}=="XXXX", ATTRS{idProduct}=="YYYY", RUN+="ZZZZ" | 06:18 |
| *** rohitksingh has quit IRC | 06:18 | |
| mithro | CarlFK: on a machine which has the complete toolchain on it | 06:20 |
| CarlFK | mithro: do we need fx2loader too? | 06:20 |
| mithro | CarlFK: do a "make gateware; PROG=fpgalink make load-gateware; make fx2-firmware" | 06:20 |
| mithro | CarlFK: the make load-gateware command will fail if you don't have an Atlys plugged in - but it should still generate the xsvf file you need | 06:21 |
| mithro | CarlFK: hrm - actually - flcli might already have the logic for waiting in it.... | 06:25 |
| *** travis-ci has joined #timvideos | 06:29 | |
| travis-ci | [timvideos/HDMI2USB-misoc-firmware/flterm-fix#161] (61ccda0): The build passed. (https://travis-ci.org/timvideos/HDMI2USB-misoc-firmware/builds/81333351) | 06:29 |
| *** travis-ci has left #timvideos | 06:29 | |
| CarlFK | mithro: somewhere in make deps it does: | 06:29 |
| CarlFK | wget -qO libargtable2.tgz --no-check-certificate https://github.com/makestuff/libargtable2/archive/dev.tar.gz | 06:29 |
| CarlFK | wget -O argtable2.tgz http://downloads.sourceforge.net/project/argtable/argtable/argtable-2.13/argtable2-13.tar.gz | 06:29 |
| tpb | Title: Download argtable from SourceForge.net (at downloads.sourceforge.net) | 06:29 |
| CarlFK | mithro: when you say "compile flcli as a static brinary" do you mean no shared object files? | 06:36 |
| mithro | Yeah | 06:36 |
| CarlFK | k - I didn't do that. juser@negk:~/makestuff/apps/flcli/lin.x64/rel$ ls | 06:36 |
| CarlFK | flcli libbuffer.so libfpgalink.so libs.txt libargtable2.so liberror.so libfx2loader.so libusbwrap.so | 06:36 |
| mithro | CarlFK: yeah - I don't know how to do that | 06:37 |
| mithro | https://www.irccloud.com/pastebin/KRBXScIG/ | 06:37 |
| tpb | Title: Pastebin | IRCCloud (at www.irccloud.com) | 06:37 |
| mithro | CarlFK: that script works for me locally | 06:37 |
| shenki | o/ | 06:38 |
| CarlFK | mithro: here is my clone and build script: http://paste.ubuntu.com/12512720/ | 06:38 |
| tpb | Title: Ubuntu Pastebin (at paste.ubuntu.com) | 06:38 |
| mithro | CarlFK: okay | 06:39 |
| mithro | CarlFK: well you need to get the .xsvf and the .hex file from somewhere too | 06:39 |
| mithro | hey shenki | 06:39 |
| shenki | mithro: we were talking about a kernel port last night | 06:40 |
| CarlFK | mithro: I was just simplifying the convoluted wget tar thing | 06:40 |
| shenki | mithro: what was the architecture? | 06:40 |
| mithro | shenki: yes - lm32 is the architecture - take a look at https://wiki.debian.org/HelmutGrohne/rebootstrap and read the lm32 section | 06:41 |
| tpb | Title: HelmutGrohne/rebootstrap - Debian Wiki (at wiki.debian.org) | 06:41 |
| mithro | shenki: the other architecture we could support is or1k | 06:41 |
| mithro | shenki: see the or1k of the above thing too | 06:41 |
| mithro | shenki: does that help? | 06:43 |
| shenki | mithro: yeah, that jogged the memory | 06:44 |
| shenki | mithro: reading up on the mmu | 06:44 |
| CarlFK | mithro: looks like those three flcli commands.. except I need to enable the encoder cuz it isn't the default | 06:54 |
| CarlFK | ^^^ 3 commands worked | 06:55 |
| shenki | mithro: | 07:24 |
| shenki | $ make make lm32-firmware make load-lm32-firmware make fx2-firmware make load-fx2-firmware make view | 07:24 |
| shenki | LONELY BOY | 07:24 |
| shenki | err | 07:24 |
| shenki | $ make lm32-firmware | 07:24 |
| shenki | cd build/misoc && python3 make.py -X ../.. -t atlys_hdmi2usb -Ot firmware_filename ../../firmware/lm32/firmware.bin -Op programmer impact build-headers | 07:24 |
| shenki | /bin/bash: line 0: cd: build/misoc: No such file or directory | 07:24 |
| shenki | Makefile.hdmi2usb:22: recipe for target 'lm32-firmware' failed | 07:24 |
| shenki | make: *** [lm32-firmware] Error 1 | 07:24 |
| shenki | mithro: did i miss a step? | 07:24 |
| mithro | shenki: that looks like you didn't do a get-env.sh correctly | 07:33 |
| shenki | mithro: ok | 07:38 |
| mithro | shenki: get-env should have downloaded build/misoc | 08:15 |
| *** wanig has quit IRC | 10:04 | |
| *** wanig has joined #timvideos | 10:04 | |
| mithro | shenki: did you fix it? | 12:05 |
| mithro | Can everyone sign up and skull / like / etc https://hackaday.io/project/7772-numato-opsis-fpga-based-open-video-platform ? | 14:08 |
| tpb | Title: Numato Opsis: FPGA-based open video platform Hackaday.io (at hackaday.io) | 14:08 |
| CarlFK | mithro: if you are still up, have anything for me to test on clean boxes? | 14:43 |
| mithro | CarlFK: are you happy with creating the .xsvf / .hex file and the above flash script? | 14:47 |
| CarlFK | mithro: kinda. I am a little fuzzy on why those files can't be created by someone else | 14:48 |
| mithro | CarlFK: because you need to be self sufficient and useful member of society | 14:49 |
| CarlFK | mithro: you need to watch the Debconf talk on volunteers | 14:51 |
| CarlFK | and I want to streamline the process of getting the Atlys up | 14:51 |
| CarlFK | I don't care if you want me to hack on the code. that isn't happening any time soon | 14:52 |
| mithro | CarlFK: the files will be generated from travis-ci soon | 14:55 |
| CarlFK | mithro: are the bit/bin files used to create something (hex and ?) that is specific to the board? like Atlys / Numato ? | 14:55 |
| mithro | CarlFK: I explained it in the reply to your email | 14:56 |
| CarlFK | mithro: ok, so thats my understanding. how many boards are we (all people tim vidoes or whoever is working) working with right now? just the two, right? | 14:59 |
| mithro | CarlFK: the HDMI2USB.tv MiSoC firmware currently works with two boards and has 3 configurations | 15:00 |
| mithro | https://travis-ci.org/timvideos/HDMI2USB-misoc-firmware | 15:00 |
| tpb | Title: Travis CI - Test and Deploy Your Code with Confidence (at travis-ci.org) | 15:00 |
| CarlFK | mithro: 3 configurations = the states as the different parts are uploaded or enabled right? | 15:01 |
| mithro | CarlFK: no - as in base, hdmi2usb and hdmi2ethernet | 15:02 |
| mithro | CarlFK: the boards do have 3 states they can be in | 15:03 |
| mithro | CarlFK: which I've started calling "unconfigured, jtag, operating" | 15:03 |
| CarlFK | mithro: got it. | 15:04 |
| CarlFK | mithro: it seems my Atlys is working with 100% of the laptops I have tested with. can you do a release that I can use for pyconza so that a regression doesn't slip in? | 15:07 |
| CarlFK | and if I report any issues, there won't be any question what code I am using | 15:08 |
| mithro | CarlFK: btw having you understand how to generate firmware for the boards means you can help get someone like tumbleweed set up with building the firmware and he can fix issues while at your event | 15:41 |
| CarlFK | mithro: i seriously doubt there will be any fixing issues at the event | 15:42 |
| CarlFK | mithro: given how stable things seem to be, I would not want to risk making it worse | 15:43 |
| CarlFK | if some laptop doesn't work, we will spend maybe a min or two, then use the twinpact | 15:43 |
| mithro | CarlFK: I would hope that you corner the person in the evening and steal there laptop to do more extensive testing and figure out why it doesn't work | 15:47 |
| CarlFK | mithro: I wold try, but generally the event soaks up all my time | 15:48 |
| CarlFK | thats why tumbleweed needs to come to Node conf so there is someone with nothing to do :D | 15:49 |
| tumbleweed | :) | 15:59 |
| tumbleweed | https://en.wikipedia.org/wiki/Metric_system#/media/File:Metric_system_adoption_map.svg | 17:42 |
| tpb | Title: Metric system - Wikipedia, the free encyclopedia (at en.wikipedia.org) | 17:42 |
| tumbleweed | US, Burma, Liberia :P | 17:42 |
| tumbleweed | err, wrong place for this | 17:42 |
| tumbleweed | CarlFK: ^^ | 17:42 |
| CarlFK | tragic. | 17:43 |
| *** automatical has joined #timvideos | 18:08 | |
| *** ysionneau has joined #timvideos | 19:40 | |
| ysionneau | hi there! | 19:40 |
| ysionneau | congratz about the fund raising having great success! | 19:41 |
| CarlFK | :) | 20:14 |
| CarlFK | tumbleweed: do you have any hdmi cables to hook things up to your Atlys board? (I am thinking I should order a few and have them shipped to your address so you can pack them. | 20:27 |
| tumbleweed | I think I have one, plus a micro-HDMI cable that came with the board | 20:33 |
| tumbleweed | I'll check when I get home | 20:33 |
| ysionneau | is it possible to "subscribe" to the google group and receive the mails like a normal mailing list ? | 20:38 |
| ysionneau | I mean, with a non Google email address | 20:39 |
| CarlFK | ysionneau: um.. I think so | 20:39 |
| ysionneau | ah yes | 20:40 |
| ysionneau | I need to send an email to [email protected] | 20:40 |
| ysionneau | hummm failed | 20:42 |
| ysionneau | ah, done! | 20:42 |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!
