| *** tpb has joined #symbiflow | 00:00 | |
| *** ASHR has joined #symbiflow | 00:31 | |
| *** xtro has quit IRC | 00:40 | |
| *** xtro has joined #symbiflow | 00:57 | |
| *** vup2 is now known as vup | 01:07 | |
| *** citypw has joined #symbiflow | 01:57 | |
| *** bjorkint0sh has joined #symbiflow | 02:55 | |
| *** bjorkintosh has quit IRC | 02:58 | |
| *** citypw has quit IRC | 03:49 | |
| *** citypw_ has joined #symbiflow | 03:49 | |
| *** Degi_ has joined #symbiflow | 03:56 | |
| *** Degi has quit IRC | 03:58 | |
| *** Degi_ is now known as Degi | 03:58 | |
| *** ASHR has quit IRC | 04:19 | |
| *** ym has quit IRC | 04:48 | |
| *** ym has joined #symbiflow | 04:48 | |
| *** ASHR has joined #symbiflow | 06:07 | |
| *** hansfbaier has joined #symbiflow | 07:14 | |
| *** QDX45 has quit IRC | 07:43 | |
| *** drawkula has joined #symbiflow | 08:06 | |
| *** yeti has quit IRC | 08:09 | |
| *** drawkula is now known as yeti | 08:10 | |
| *** ASHR has quit IRC | 08:11 | |
| *** xtro has quit IRC | 08:14 | |
| *** citypw_ has quit IRC | 08:49 | |
| *** ovf has quit IRC | 09:14 | |
| *** ovf has joined #symbiflow | 09:15 | |
| *** unrznbl[m] has quit IRC | 09:17 | |
| *** xobs has quit IRC | 09:17 | |
| *** Niklas[m]2 has quit IRC | 09:21 | |
| *** abeljj[m] has quit IRC | 09:21 | |
| *** promach3 has quit IRC | 09:21 | |
| *** hansfbaier has quit IRC | 09:43 | |
| *** SmutLord^ has quit IRC | 09:58 | |
| *** SmutLord^ has joined #symbiflow | 09:58 | |
| *** abeljj[m] has joined #symbiflow | 10:09 | |
| *** mkru has joined #symbiflow | 10:23 | |
| *** mkru has quit IRC | 10:29 | |
| *** promach3 has joined #symbiflow | 10:34 | |
| *** xobs has joined #symbiflow | 10:34 | |
| *** unrznbl[m] has joined #symbiflow | 10:34 | |
| *** Niklas[m]1 has joined #symbiflow | 10:34 | |
| *** sadoon_albader has joined #symbiflow | 12:12 | |
| *** sadoon_albader has quit IRC | 12:13 | |
| *** gromero has quit IRC | 12:47 | |
| *** gromero has joined #symbiflow | 12:48 | |
| *** gromero_ has joined #symbiflow | 12:51 | |
| *** gromero has quit IRC | 12:51 | |
| *** ym_ has joined #symbiflow | 12:52 | |
| *** gromero has joined #symbiflow | 12:54 | |
| *** ym has quit IRC | 12:55 | |
| *** gromero_ has quit IRC | 12:57 | |
| *** infinite_recursi has joined #symbiflow | 13:15 | |
| infinite_recursi | Is there a difference between iceprog and iceprogduino wrt hx8k olimex board | 13:16 |
|---|---|---|
| infinite_recursi | ? | 13:16 |
| infinite_recursi | I do have iceprog in my installed icestorm but do not have iceprogduino. The makefile has command, iceprogduino | 13:17 |
| *** infinite_recursi has quit IRC | 13:39 | |
| *** gromero_ has joined #symbiflow | 14:18 | |
| *** gromero has quit IRC | 14:19 | |
| *** kraiskil has joined #symbiflow | 16:48 | |
| *** kraiskil has quit IRC | 17:17 | |
| lambda | turns out arch-defs takes a pretty insane amount of RAM to build, I guess I'll have to use binaries :p are there any more canonical download links for the tarballs mentioned in https://symbiflow-examples.readthedocs.io/en/latest/getting-symbiflow.html#toolchain-installation? | 17:37 |
| tpb | Title: Getting SymbiFlow SymbiFlow examples documentation (at symbiflow-examples.readthedocs.io) | 17:37 |
| sf-slack | <acomodi> lambda: there are latest valid links, generated only if all CIs went green | 17:39 |
| sf-slack | <acomodi> lambda: https://console.cloud.google.com/storage/browser/symbiflow-arch-defs-gha;tab=objects?authuser=0&prefix=&forceOnObjectsSortingFiltering=false | 17:39 |
| tpb | Title: Google Cloud Platform (at console.cloud.google.com) | 17:39 |
| lambda | acomodi: interesting, is it possible to get at that information without a google account though? | 17:47 |
| sf-slack | <acomodi> I think that using GCS APIs should let you get there | 17:50 |
| sf-slack | <acomodi> even without an account | 17:50 |
| lambda | no, doesn't look like unauthorized requests can get you anything - and those files don't have public URLs either way | 18:02 |
| litghost | lambda: https://storage.googleapis.com/symbiflow-arch-defs-gha/symbiflow-xc7a50t_test-latest should be a fully public URL | 18:13 |
| litghost | The URL in that file should also be fully public, but might not be? | 18:14 |
| lambda | yeah that works, how did you get that URL? https://console.cloud.google.com/storage/browser/_details/symbiflow-arch-defs-gha/symbiflow-xc7a50t_test-latest shows "Public URL: Not applicable" | 18:15 |
| tpb | Title: Google Cloud Platform (at console.cloud.google.com) | 18:15 |
| litghost | There is a "Public URL" and an "Authenticated URL" | 18:16 |
| litghost | From the GCP bucket display | 18:16 |
| litghost | If the documentation is pointing to the "Authenticated URL", that is a mistake | 18:16 |
| litghost | For reference the "Authenticated URL" for that same file is https://storage.cloud.google.com/symbiflow-arch-defs-gha/symbiflow-xc7a50t_test-latest?authuser=0 | 18:17 |
| tpb | Title: Sign in - Google Accounts (at storage.cloud.google.com) | 18:17 |
| lambda | right, I was looking at the GCP bucket, but it's not showing me any public URLs | 18:17 |
| litghost | So the bucket was intented to expose public URLs, but if that isn't what is happening, that is problematic | 18:17 |
| litghost | Yay debugging | 18:18 |
| -_whitenotifier- [symbiflow-docs] litghost opened issue #429: Documentation should provide directions to get latest arch-defs? - https://git.io/Jt3ob | 18:21 | |
| *** xtro has joined #symbiflow | 19:31 | |
| *** ayazar1 has joined #symbiflow | 19:32 | |
| *** ayazar1 has quit IRC | 19:49 | |
| *** ayazar1 has joined #symbiflow | 19:49 | |
| *** ayazar1 has quit IRC | 19:55 | |
| *** kraiskil has joined #symbiflow | 20:20 | |
| *** alexhw has quit IRC | 20:25 | |
| lambda | where does prjxray-config come from (https://github.com/SymbiFlow/symbiflow-arch-defs/blob/master/xc/xc7/toolchain_wrappers/symbiflow_synth#L93)? I can't find it in prjxray or prjxray-db. | 20:26 |
| litghost | It's a way for packaging to provide the path to prjxray-db, one minute | 20:44 |
| litghost | https://github.com/litex-hub/litex-conda-eda/blob/df253ae4d38761b3c5d6a618099ceeaa79669a7d/bit/prjxray-db/build.sh#L19-L20 | 20:46 |
| lambda | oh, that's fairly straightforward :p | 20:47 |
| *** tpb has joined #symbiflow | 21:29 | |
| *** xtro has quit IRC | 21:39 | |
| *** nickoe has joined #symbiflow | 21:54 | |
| nickoe | hello | 21:54 |
| *** kraiskil has quit IRC | 22:20 | |
| nickoe | Why does the symbiflow-examples and symbiflow-arch-defs have different ways to "init" the dependenceis, as far as I can see both use conda stuff. The examples project uses miniconda directly while the arch-defs project has "make env" and does magic in there. | 22:25 |
| nickoe | can't the examples and arch-defs use the same shared install env? | 22:27 |
| litghost | They are the same under the hood in the sense that it downloads miniconda and creates an environment | 22:36 |
| litghost | You don't have to use the wrappers for setting up the env, and the other parts of the build system doesn't actually care how the environment is created | 22:37 |
| litghost | If you invoke cmake (for arch-defs) within an environment with the right dependencies, it should just "work" | 22:37 |
| litghost | Same more or less for symbiflow-examples | 22:38 |
| nickoe | I am just slightly confused by the simarilatires yet differences | 22:40 |
| nickoe | I am trying to see if I can get an example working for the basys3 | 22:40 |
| litghost | If you want to simplest example, I would just follow the instructions from symbiflow-examples | 22:42 |
| litghost | That should work, we actually test those instructions in CI | 22:42 |
| litghost | If you want to try re-use a conda env, you might get off the beaten path a bit | 22:42 |
| litghost | But you should be able to re-use the environment if you wanted too | 22:43 |
| nickoe | running TARGET="basys3" make -C counter_test now | 22:43 |
| nickoe | yay, there is a counter runnig | 22:44 |
| nickoe | It describes https://symbiflow-examples.readthedocs.io/en/latest/building-examples.html | 22:47 |
| tpb | Title: Building example designs SymbiFlow examples documentation (at symbiflow-examples.readthedocs.io) | 22:47 |
| nickoe | oh the gpio sutf is for another board | 22:47 |
| nickoe | Is the litex demo too big for the basys3 board os it just becasue it has no ethrenet? | 22:48 |
| nickoe | litghost: Do you know anything about the picosoc example? Can one build c firmware for it? | 22:55 |
| nickoe | I don't see any c file in the example | 22:55 |
| litghost | The firmware is currently encoded in verilog: https://github.com/SymbiFlow/symbiflow-examples/blob/476521430a7c8c33a5436cfaa87d933fdbfa882e/xc7/picosoc_demo/progmem.v#L27 | 23:04 |
| litghost | The tool for generating the verilog is here: https://github.com/SymbiFlow/symbiflow-arch-defs/blob/master/tests/9-soc/picosoc/hex2progmem.py | 23:05 |
| litghost | And the firmware is over here: https://github.com/SymbiFlow/symbiflow-arch-defs/blob/master/tests/9-soc/picosoc/firmware/Makefile | 23:05 |
| litghost | The idea in symbiflow-examples was to side step the firmware side to focus on the P&R side | 23:06 |
| litghost | I believe the litex demo is just the ethernet port | 23:06 |
| litghost | the arty and basys both use a similiar size fabric | 23:06 |
| nickoe | litghost: ah, pretty cool I was wondering a moment about that array in that progmem.v, but I guesssed it was some rom code. | 23:09 |
| nickoe | so in principle the litex exsample should aslo work wiht uart console? | 23:09 |
| *** SmutLord^ has quit IRC | 23:09 | |
| *** SmutLord^ has joined #symbiflow | 23:10 | |
| litghost | I also believe the arty has DDR, and the basys3 doesn't | 23:10 |
| litghost | But don't quote me on that | 23:10 |
| nickoe | probably | 23:11 |
| litghost | Ya, I just double checked | 23:11 |
| litghost | The whole idea behind the LiteX Linux SoC really relies on external DRAM | 23:11 |
| litghost | I bet you could finagle the Litex Linux SoC demo on the basys3, but you'd run out of BRAM really fast | 23:12 |
| nickoe | anyway, it looks like I could figure outhow to run the examples. I guess next step is to figoure out how to properly simulate the counter example and get the logic analyzer plots. Is it only gtk wave people use? | 23:12 |
| litghost | gtk wave only shows outputs I believe | 23:13 |
| litghost | verilator or iverilog is required to generate the waveforms | 23:13 |
| nickoe | litghost: I am pretty new to FPGA, although it was years back I learned some | 23:13 |
| nickoe | So first I need to get used to the workflow and actually plot some veriflog bits together. | 23:14 |
| nickoe | the examples site does not describe simulation | 23:14 |
| litghost | So pre-synthesis simulation definitely works. Post-synthesis and post-place and route simulation is hit or miss. We have an issue to improve the situation | 23:14 |
| litghost | https://github.com/SymbiFlow/symbiflow-arch-defs/issues/1734 | 23:14 |
| nickoe | mmm, I have no idea if I want pre or post synth sim | 23:15 |
| nickoe | I just want to know if I verilog is any good :S | 23:15 |
| litghost | Pre-synthesis simulation is probably fine then | 23:15 |
| litghost | http://iverilog.icarus.com/ | 23:16 |
| nickoe | Can you please remind me what the real difference is? Do you get timging skew with post synth or something? I guess post synth must add some more details to the sim? | 23:16 |
| litghost | Pre-synthesis is in the abstract verilog machine, post-synthesis everything has been lowered to real FPGA primitivies | 23:17 |
| *** QDX45 has joined #symbiflow | 23:17 | |
| nickoe | Some friends made some small modules they test with verilator tb | 23:17 |
| nickoe | so my idea was to try to plot those in a proper project | 23:17 |
| litghost | Post-place and route typically includes actually delays, but you can do without too | 23:17 |
| nickoe | actually wanted to do it in vivdo, but then I saw that vivado prjects were a mess in git, so I though I might just try symbiflow and see how far I can get. | 23:18 |
| litghost | Great, let us know how it goes! | 23:19 |
| litghost | FYI, vivado projects aren't too awful in git, but we are always excited to have people try out the toolchain | 23:21 |
| litghost | The key is to just re-create the project in tcl everytime | 23:21 |
| litghost | Example: https://github.com/SymbiFlow/prjxray/blob/master/minitests/hamsternz-displayport/Makefile | 23:21 |
| nickoe | yeah, that I also figured by some of the top hits on google, but just exporting it and such gave absolute paths to the verilog files and I sorta just gave up | 23:24 |
| nickoe | But just given tht symbiflow is fater to install,... if it works it works :D | 23:24 |
| litghost | :D | 23:25 |
| nickoe | and is FOSS sooo +++ | 23:25 |
| litghost | Word of warning, QoR from symbiflow is going to be worse than Vivado, so don't push the critical path if you can avoid it | 23:25 |
| nickoe | This is for an SDR transmit only radio, I don't think that will be that demanding to the artix, but I don't reeallly know as I have littel experince with it. | 23:26 |
| litghost | Good luck! | 23:26 |
| nickoe | But the plan is just to pump stuff to an external DAC and get it to do what we want as a start. | 23:26 |
| litghost | That should definitely be doable | 23:27 |
| nickoe | but as I understand the toolchain, I could use the vivado pnr stuff if needed? | 23:27 |
| litghost | So we use yosys on for synthesis, and you can have yosys output an edif for consumption with Vivado PnR | 23:27 |
| nickoe | great | 23:28 |
| nickoe | it is getting late here, but I will try to give it a stab tomorrow agian | 23:28 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!