| *** tpb has joined #timvideos | 00:00 | |
| CarlFK | nancy_: ok, soon is now :p | 00:01 |
|---|---|---|
| CarlFK | nancy_: ftp a.img.cloudns.cx | 00:02 |
| cr1901_modern | mithro: Yaaaaay | 00:02 |
| CarlFK | nancy_: user: anonymous password: anonymous | 00:03 |
| mithro | cr1901_modern / CarlFK - need to test the builds work | 00:07 |
| CarlFK | doing that now. spend the last X hours making it easy to drop files onto my test box | 00:07 |
| CarlFK | the downloader is broken | 00:13 |
| cr1901_modern | mithro: You mean just do a build locally? | 00:16 |
| CarlFK | Did not find firmware firmware for target hdmi2usb for platform opsis at rev v0.0.4-190-gcacbf40 (found ) | 00:16 |
| mithro | https://www.irccloud.com/pastebin/CZPlZIeg/ | 00:37 |
| tpb | Title: Snippet | IRCCloud (at www.irccloud.com) | 00:37 |
| mithro | CarlFK: ^ | 00:37 |
| CarlFK | mithro: um... why did you show me that ? | 00:38 |
| mithro | CarlFK: See the channel lines | 00:39 |
| CarlFK | yes... | 00:39 |
| mithro | CarlFK: Please test the unstable line | 00:43 |
| cr1901_modern | mithro: What should be do about this? https://github.com/timvideos/litex-buildenv/pull/27 | 00:46 |
| tpb | Title: Merge in LiteX changes by cr1901 · Pull Request #27 · timvideos/litex-buildenv · GitHub (at github.com) | 00:46 |
| cr1901_modern | and what do you mean "need to test the builds work" | 00:46 |
| CarlFK | mithro: I am blocked: Channel unstable is at rev v0.0.4-192-g7a70488 ... python3 download-prebuilt-firmware.py --platform opsis --rev v0.0.4-192-g7a70488 | 00:46 |
| CarlFK | Found firmwares: | 00:47 |
| CarlFK | Did not find firmware firmware for target hdmi2usb for platform opsis at rev v0.0.4-192-g7a70488 (found ) | 00:47 |
| CarlFK | and it doesnt download anyting | 00:47 |
| CarlFK | http://paste.ubuntu.com/p/BngPXdN5jn/ both of those should work | 00:48 |
| tpb | Title: Ubuntu Pastebin (at paste.ubuntu.com) | 00:48 |
| CarlFK | this does work: python3 download-prebuilt-firmware.py --platform opsis --rev v0.0.4-122-gbc7feb5 | 00:49 |
| CarlFK | this is where the file should be, right? https://github.com/timvideos/HDMI2USB-firmware-prebuilt/tree/master/archive/master/v0.0.4-192-g7a70488/opsis/hdmi2usb/lm32 | 00:53 |
| tpb | Title: HDMI2USB-firmware-prebuilt/archive/master/v0.0.4-192-g7a70488/opsis/hdmi2usb/lm32 at master · timvideos/HDMI2USB-firmware-prebuilt · GitHub (at github.com) | 00:53 |
| mithro | CarlFK: hrm.... There should be an image file there... | 01:02 |
| mithro | CarlFK: When did the images stop getting generated? | 01:03 |
| mithro | They exist back at HDMI2USB-firmware-prebuilt/archive/master/v0.0.4-74-g807ed55 | 01:04 |
| CarlFK | i have a working v0.0.4-122-gbc7feb5 | 01:04 |
| mithro | They are at HDMI2USB-firmware-prebuilt/archive/master/v0.0.4-157-ga112978/arty/base/lm32/ | 01:04 |
| CarlFK | edid is still broken, but I captured a test pattern | 01:05 |
| mithro | CarlFK: On what version? | 01:05 |
| CarlFK | v0.0.4-122-gbc7feb5 | 01:05 |
| CarlFK | https://veyepar.nextdayvideo.com/static/temp/v0.0.4-165-g619dcf9-p.png | 01:06 |
| mithro | https://travis-ci.org/timvideos/HDMI2USB-litex-firmware/jobs/355140444#L4531 | 01:08 |
| CarlFK | wait.. that says 165 | 01:08 |
| mithro | CarlFK: Channel unstable is at rev v0.0.4-192-g7a70488 | 01:08 |
| CarlFK | sorry, cut/paste fail. i have a working v0.0.4-165-g619dcf9 | 01:08 |
| CarlFK | mithro: not sure what you want me to do? | 01:11 |
| mithro | CarlFK: EDID is fixed in v0.0.4-190 or later | 01:11 |
| mithro | Oh - I know what it is... | 01:12 |
| mithro | CarlFK: When https://travis-ci.org/timvideos/HDMI2USB-litex-firmware/builds/355181729 finishes, the firmware should be back in the prebuilt repo | 01:50 |
| CarlFK | mithro: yay | 01:52 |
| mithro | CarlFK: What ever happened to the Google summer of code announcement you were going to do? | 02:02 |
| CarlFK | mithro: which one? ... I submitted a few things, I don think they got posted ' | 02:03 |
| mithro | CarlFK: You never responded to my comments on your pull request -> https://github.com/timvideos/hdmi2usb.tv/pull/24 | 02:04 |
| tpb | Title: GSoC announcement post. by CarlFK · Pull Request #24 · timvideos/hdmi2usb.tv · GitHub (at github.com) | 02:04 |
| CarlFK | mithro: I think I stopped when I saw Generally LGTM! | 02:05 |
| CarlFK | and really, just post it as is. | 02:06 |
| CarlFK | you might be right about the quotes, but it isn't worth the time to process it | 02:06 |
| mithro | CarlFK: fine I'll do it myself | 02:09 |
| CarlFK | mithro: https://travis-ci.org/ Running for 1 hr 47 min 27 sec | 03:37 |
| CarlFK | is that 'too long' or does it take that long for all the builds ? | 03:37 |
| mithro | CarlFK: Look at https://travis-ci.org/timvideos/HDMI2USB-litex-firmware | 03:43 |
| CarlFK | 20 min times about 100 - got it :p | 03:44 |
| *** rohitksingh_work has joined #timvideos | 03:44 | |
| mithro | CarlFK: Looks like they finished... | 03:51 |
| mithro | bblr - going to get dinner... | 03:52 |
| CarlFK | mithro: woo! Downloading to: image-gateware+bios+firmware.v0.0.4-194-ge14edee.opsis.hdmi2usb.lm32.bin | 03:52 |
| nancy_ | @CarlFK sir, can i get connected to ftp ? i got disconnected , till the time you cam back sir ! | 04:00 |
| CarlFK | nancy_: im here, it should be available now? | 04:01 |
| nancy_ | Though i mailed you sir ! | 04:01 |
| CarlFK | nancy_: yes, but please use ftp - it puts the file on the box I use for testing | 04:05 |
| nancy_ | ok sir | 04:06 |
| nancy_ | i got connected | 04:16 |
| nancy_ | should i transfer the file ? | 04:22 |
| CarlFK | yes | 04:22 |
| CarlFK | put image-gateware+bios+firmware-nancy-v2.bin | 04:23 |
| CarlFK | or some new name so we know which version I am working with | 04:23 |
| nancy_ | oh i already send you sir , did saw the last thread | 04:27 |
| nancy_ | image-gateware+bios+firmware-nancy-v1.bin | 04:28 |
| CarlFK | nancy_: I see image-gateware+bios+firmware.bin 2216328 | 04:32 |
| CarlFK | I assume that is what I should test now | 04:32 |
| nancy_ | yes sir | 04:35 |
| CarlFK | um.. did you just rename it? | 04:37 |
| CarlFK | doing this now: hdmi2usb-mode-switch -v --flash-image image-gateware+bios+firmware-nancy-v3.bin | 04:38 |
| nancy_ | yes | 04:39 |
| CarlFK | nancy_: good news - it worked.. kinda. | 04:44 |
| nancy_ | cool! | 04:45 |
| CarlFK | there are different builds - some don't have all the features. yours only does hdmi in and out, but no serial console | 04:45 |
| CarlFK | but thats is a config option somewhere, or look in a different dir | 04:45 |
| nancy_ | some dont have all features ! it means some scripts were not run ? | 04:46 |
| CarlFK | we do see that you have all the pieces in place | 04:46 |
| nancy_ | for serial console ? | 04:47 |
| nancy_ | drivers problem ? | 04:48 |
| CarlFK | It isn't really a problem - it isn't there | 04:49 |
| CarlFK | because it is optional, and what you gave me does not have that option | 04:49 |
| CarlFK | targets: base, hdmi2usb, net, video | 04:52 |
| CarlFK | I think you gave me video. see if you can find hdmi2usb | 04:52 |
| nancy_ | hdmi2usb ? | 04:54 |
| nancy_ | i think i need to run some script for it | 04:55 |
| CarlFK | I think set a var, and re-run make | 04:56 |
| nancy_ | make gateware | 05:05 |
| *** sb0 has quit IRC | 05:15 | |
| nancy_ | @CarlFK sir i got the file , but coulnt put it via ftp | 05:31 |
| CarlFK | nancy_: why not ftp? | 05:32 |
| nancy_ | put hdmi2usb local: hdmi2usb remote: hdmi2usb hdmi2usb: not a plain file. | 05:32 |
| CarlFK | hdmi2usb doesn't look like the right filename | 05:39 |
| nancy_ | ok sir | 05:41 |
| nancy_ | i need to send bin file ? | 05:42 |
| CarlFK | yes | 05:42 |
| nancy_ | i send you sir , kindly review | 05:44 |
| nancy_ | Is it not working ? | 05:59 |
| CarlFK | nancy_: what is the filename? | 06:01 |
| nancy_ | zero.bin | 06:02 |
| CarlFK | it is only 32 bytes | 06:02 |
| nancy_ | yes , i checked sir | 06:05 |
| CarlFK | I don't think it will do much | 06:07 |
| nancy_ | sir i want to ask you something | 06:08 |
| CarlFK | yes? | 06:08 |
| nancy_ | i worked on verilog | 06:08 |
| nancy_ | and i checked the whole thing of how HDMI2USB works ! | 06:08 |
| nancy_ | so the development part that is supposed to be done in python is done on migen ? | 06:09 |
| CarlFK | I think so... I don't do that - I just flash and use it | 06:10 |
| nancy_ | ok sir | 06:13 |
| *** sb0 has joined #timvideos | 06:24 | |
| CarlFK | I am off to bed - wont have much time tomorrow unless I get up extra early. might happen. | 07:20 |
| *** nancy_ has quit IRC | 07:42 | |
| *** CarlFK has quit IRC | 07:49 | |
| *** futarisIRCcloud has joined #timvideos | 08:25 | |
| *** bunnie has joined #timvideos | 09:31 | |
| bunnie | Can anyone point me to documentation on streams in migen? | 09:31 |
| *** olasd has quit IRC | 11:01 | |
| *** olasd has joined #timvideos | 11:17 | |
| jea | CarlFK[m]: https://github.com/joeladdison/streaming-ui | 11:35 |
| tpb | Title: GitHub - joeladdison/streaming-ui: Streaming UI (at github.com) | 11:35 |
| jea | finally finished dealing with urgent work support stuff after two weeks, so could clean it up and push to github | 11:37 |
| cr1901_modern | bunnie: There aren't any, really. I don't personally use the streams (unless interfacing to existing code) | 11:42 |
| *** ChanServ sets mode: +v thaytan_ | 12:01 | |
| *** thaytan_ is now known as thaytan | 12:01 | |
| *** nancy has joined #timvideos | 12:08 | |
| *** neoweb has joined #timvideos | 12:09 | |
| *** CarlFK has joined #timvideos | 12:15 | |
| *** ChanServ sets mode: +v CarlFK | 12:15 | |
| *** nancy has quit IRC | 12:33 | |
| *** sb0 has quit IRC | 12:35 | |
| *** rohitksingh_work has quit IRC | 12:38 | |
| *** thaytan has quit IRC | 13:24 | |
| *** thaytan has joined #timvideos | 13:35 | |
| *** ChanServ sets mode: +v thaytan | 13:35 | |
| *** sb0 has joined #timvideos | 13:44 | |
| bunnie | jea: thanks, i was referring to the migen "stream" hardware construct, not the application-level streaming of video. But thanks for the response. | 13:50 |
| bunnie | cr1901_modern: thanks. Yes, I'm not trying to write code using it. I'm trying to modify existing code that uses a stream and I'm having a hell of a time getting it to do what I think it should. | 13:51 |
| *** shivm has joined #timvideos | 14:00 | |
| *** shivm28 has joined #timvideos | 14:02 | |
| *** shivm has quit IRC | 14:05 | |
| *** rohitksingh has joined #timvideos | 14:27 | |
| *** rohitksingh has quit IRC | 14:36 | |
| *** rohitksingh has joined #timvideos | 14:37 | |
| CarlFK | jea: yay!!! | 15:08 |
| CarlFK | jea or someone - can you reference that code on the timvideos github idea issue? | 15:09 |
| CarlFK | extra credit for getting it under the timvideos project | 15:09 |
| CarlFK | im heading to the airport now, so no time for antoer 6 hours | 15:09 |
| _florent_ | bunnie: there is no documentation, but feel free to ask if you have questions | 15:11 |
| _florent_ | bunnie: stream are just specific records, with valid/ready/first/last/data/param signals | 15:12 |
| _florent_ | bunnie: that's a bit similar to streaming in AXI | 15:12 |
| *** anant3110 has joined #timvideos | 15:34 | |
| *** nancy has joined #timvideos | 15:37 | |
| anant3110 | Hey, i saw the chat with shivm_ here. I have access to a Spartan 6 Atlys board. There shouldn't be a problem right? | 15:40 |
| *** rohitksingh has quit IRC | 15:42 | |
| *** rohitksingh has joined #timvideos | 16:16 | |
| *** CarlFK has quit IRC | 16:29 | |
| *** anant3110 has quit IRC | 16:44 | |
| *** rohitksingh has quit IRC | 16:49 | |
| bunnie | _florent_: so streams are largely abstract objects that don't result in generated hardware? For example, I see several places in litevideo where a stream.Endpoint is created, some bits of signals are assigned, but not enough really to do anything. It seems like...elsewhere, the streams then get patched together. But in the simplest case is a stream just a fancy way to bundle together wires, or do they always result in e.g. a FIFO being instantiated of some | 17:14 |
| bunnie | like if I just made two stream.Endpoints that point to identical descriptors and the .connect() them together, is that basically the same as doing self.comb += on a whole bunch of signals? | 17:15 |
| _florent_ | bunnie: yes that's it | 17:16 |
| bunnie | or does it do something more like implement flow control and buffering by default | 17:16 |
| bunnie | ah ok. so then if I took the simple case of streams connected together and then inserted in the code a call to .AsyncFIFO() it would then magically stick an async FIFO between the source and sink, but *only* if I did that call | 17:17 |
| bunnie | i see functions like upconverter, downconverter, strideconverter inside the stream.py file...how do those come into play? | 17:18 |
| *** akhil_singh has joined #timvideos | 17:18 | |
| _florent_ | bunnie: yes, stream are not doing fancy stuff, that's just useful to connect bunch of signals together | 17:18 |
| _florent_ | also connect handle the fact that signals have directions | 17:19 |
| *** akhil_singh has quit IRC | 17:19 | |
| _florent_ | ex: valid/first/last/data are Master to Slave, ready is Slave to Master | 17:19 |
| _florent_ | converters are useful to convert datawidths of streams | 17:20 |
| _florent_ | ex: you have a 8 bits datas each cycle and want 16 bits data every two cycle, you will use a upconverter 8 to 16 | 17:20 |
| bunnie | is the upconverter automatically created when the data widths are different or do I need to specify it? | 17:21 |
| bunnie | and "source" is always "master", "sink" is always "slave"? | 17:22 |
| _florent_ | "master" and "slave" depends from where you see things: inside the module or outside the module | 17:24 |
| _florent_ | the upconverter is not created automatically | 17:25 |
| _florent_ | you need to instanciate it | 17:25 |
| bunnie | ah interesting. so what's the convention for master vs. slave when inside or outside the module? | 17:30 |
| *** Roedy256 has joined #timvideos | 17:34 | |
| mithro | hey bunnie! | 17:39 |
| mithro | bunnie: I saw your hack for scoping a CPU -- I've been meaning to send you a message about that, you did that the hardest way I could think of :-P | 17:42 |
| mithro | bunnie: I was also going to say you should be reaching out to _florent_ a bit more | 17:46 |
| mithro | bunnie: Wish I had more time to help you but got a deadline in 4 weeks :-/ | 17:46 |
| mithro | https://opensource.googleblog.com/2018/03/coding-your-way-into-cinemas.html | 17:46 |
| tpb | Title: Coding your way into cinemas | Google Open Source Blog (at opensource.googleblog.com) | 17:46 |
| *** akhil_singh has joined #timvideos | 17:47 | |
| _florent_ | bunnie: outside the module: source of the module is where data comes from, sink where data goes | 17:48 |
| akhil_singh | mithro: After reading form various papers according to me we should use PRBS-31 for our case that will generate the data sequences and then we can use CRC method to check for bit errors. I looked at advantages of it. First advantage is that this sequence will be very high speed so a ideal sequence we want to transceivers with. Secondly the overhead is only of xor gates and it can test many inputs in parallel. | 17:51 |
| akhil_singh | mithro: is this correct ? | 17:51 |
| mithro | akhil_singh: I'm afraid not -- your still thinking about it from the wrong direction | 17:51 |
| akhil_singh | mithro | 17:51 |
| mithro | akhil_singh: First - put together a list of potential applications we might want to use transceivers for | 17:52 |
| akhil_singh | mithro: Like ethernet etc ? | 17:53 |
| mithro | akhil_singh: Then put together a short description of the important aspects of each of the signals which are sent - IE How big is their word size, how fast do they operate, do they have some type of encoding, why do they use this encoding | 17:53 |
| mithro | akhil_singh: Probably how they get clocking information is probably important too | 17:54 |
| akhil_singh | mithro: okay, will check it. | 17:56 |
| akhil_singh | mithro: Also can you tell about configuring the transceiver. Should we change the equalization parameters and what are the other things that we should be able to change ? | 17:58 |
| mithro | akhil_singh: That is something that you should be telling me | 17:58 |
| akhil_singh | mithro: okay, sorry for asking this question. | 18:00 |
| shivm28 | mithro: Hi, I read a few mails from last year's archive in OpenSOCDebug. A figure was proposed in those mails: https://docs.google.com/drawings/d/1AmjG0FFkTv_wLnw1iKEqU-Do72TsChtH6hlmAS98At8/edit | 18:03 |
| mithro | akhil_singh: It's perfectly fine to ask questions and I encourage you to do so! | 18:03 |
| tpb | Title: Unbenannte Zeichnung - Google Drawings (at docs.google.com) | 18:03 |
| mithro | I'm afraid I have to go now | 18:04 |
| mithro | I'll won't be back for at least 4ish hours | 18:04 |
| shivm28 | Can we discuss it tomorrow around same time? | 18:05 |
| *** nancy has quit IRC | 18:05 | |
| *** akhil_singh has quit IRC | 18:16 | |
| bunnie | _florent_: thanks. Another question, if I have the time. I've got video data going from an input to an output via the framebuffers, but it seems like the video is coming across in groups of 16 pixels wide, and within those 16 pixels the video order is reversed. I feel like this must be one of those problems that's not uncommon and easily resolved ... like a misconfiguration of burst order to DDR or something. | 18:29 |
| bunnie | *if you have the time ;) | 18:29 |
| bunnie | mithro: when it comes to Python, it seems I'm always doing it the hardest way possible. My brain is simply not wired for Python. :P | 18:33 |
| *** nancy has joined #timvideos | 18:45 | |
| shivm28 | florent: Hi, I am doing a bit of research to add support for Basys3 to litex-buidenv for some basic testing. Can you please guide me? | 18:49 |
| *** harshgugale has joined #timvideos | 19:04 | |
| *** smartguy has joined #timvideos | 19:06 | |
| *** shivm28 has quit IRC | 19:10 | |
| *** smartguy has quit IRC | 19:22 | |
| *** harshgugale has quit IRC | 19:32 | |
| *** akhil_singh has joined #timvideos | 22:33 | |
| akhil_singh | mithro: Are you free ? | 22:38 |
| *** CarlFK has joined #timvideos | 23:05 | |
| *** ChanServ sets mode: +v CarlFK | 23:05 | |
| akhil_singh | mithro: Ping | 23:56 |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!
