Saturday, 2023-07-15

*** tpb <[email protected]> has joined #yosys00:00
*** jn <jn!~quassel@user/jn/x-3390946> has quit IRC (Ping timeout: 240 seconds)01:57
*** jn <jn!~quassel@2a0a-a541-37ac-0-20d-b9ff-fe49-15fc.ipv6dyn.netcologne.de> has joined #yosys01:58
*** skipwich <skipwich!~skipwich@user/skipwich> has quit IRC (Quit: DISCONNECT)03:01
*** skipwich <skipwich!~skipwich@user/skipwich> has joined #yosys03:02
*** Forty-Bot <[email protected]> has joined #yosys03:11
*** pbsds <[email protected]> has quit IRC (Quit: The Lounge - https://thelounge.chat)07:46
*** bjorkintosh <bjorkintosh!~bjork@user/bjorkintosh> has quit IRC (Read error: Connection reset by peer)08:38
*** bjorkintosh <bjorkintosh!~bjork@2600:1700:5400:c80:f09e:4fd1:7164:3401> has joined #yosys08:38
*** pbsds <[email protected]> has joined #yosys09:14
*** Raito_Bezarius <Raito_Bezarius!~Raito@wireguard/tunneler/raito-bezarius> has quit IRC (Read error: Connection reset by peer)11:02
*** Raito_Bezarius <Raito_Bezarius!~Raito@wireguard/tunneler/raito-bezarius> has joined #yosys11:40
FL4SHKokay so this bug appears to be in sv2v14:02
tntis it ? What's wrong with the verilog ? It looked fine to me.14:03
FL4SHKthe problem is the order in which task arguments are called14:03
FL4SHKI had used named (.name(...)) arguments in the SV14:03
FL4SHKand the order wasn't right in the final Verilog14:04
tntOooohhhhh yeah, I completely missed that when reading the verilog.14:04
FL4SHKI just found it14:04
FL4SHKthis explains everything14:04
FL4SHKhm14:06
FL4SHKit appears that there is still an issue with yosys14:06
FL4SHKbecause even with not using named arguments14:06
FL4SHKit still assigns 'x14:06
FL4SHKwithout rst there's no issue14:07
FL4SHKbut with rst there is14:07
FL4SHKor hm?14:09
FL4SHKmaybe with rst there's still a bug14:10
FL4SHKI could have sworn I had put these in the right order14:10
FL4SHKtnt: so yeah it still doesn't work14:18
tnt:(15:00
FL4SHKtnt: it might only be a problem with Verilog output16:43
FL4SHKIs there a different way to output a netlist besides write_verilog?16:44
jix_didn't read all the backlog yet, so might be missing context, but write_rtlil writes a netlist in yosys's native format16:45
FL4SHKjix_: that sounds like something for me to try16:53
loftyFL4SHK: there's also write_json16:56
FL4SHKlofty: I'll give that a try17:08
*** cr1901_ <cr1901_!~cr1901@2601:8d:8600:911:8d73:996:e200:f429> has joined #yosys18:08
*** cr1901 <cr1901!~cr1901@2601:8d:8600:911:e4d2:571a:53b4:ed7a> has quit IRC (Ping timeout: 246 seconds)18:11
FL4SHKwhat does "cells not processed" mean?20:28
FL4SHKit looks like even the other output types have task input ports driven with 'x 20:29
FL4SHKI'm running yosys-nightly20:32
FL4SHKso I'll try another version20:32
FL4SHKappears to be a problem even with the Arch Linux yosys repo20:36
FL4SHKyosys version*20:36
jix_FL4SHK: it's set by the frontend and should be cleared by the hierarchy pass and AFAIK is basically just a marker that hierarchy still has to process stuff20:48
FL4SHKjix_: I see20:49
FL4SHKdo I need another pass besides `read_verilog` and `proc`?20:49
jix_you should run `hierarchy` between those20:50
FL4SHKgotcha20:51
FL4SHKthat explains the issue I was finding20:52
jix_the list of passes in `help synth` or alternatively `help prep` are complete examples of pass orders, and taking those (or a prefix) of them is usually a good starting point if you want to do something custom20:52
*** cr1901_ is now known as cr190120:57
FL4SHKit looks like <= doesn't work21:07
FL4SHKbut = does21:07
FL4SHKin a task21:07
FL4SHKoh wait no21:08
FL4SHKoh huh no it is <= that doesn't work21:09
FL4SHKI'm satisfied with this result21:21
*** nonchip <[email protected]> has quit IRC (Quit: https://quassel-irc.org - Chat comfortably. Anywhere.)22:15
*** nonchip <[email protected]> has joined #yosys22:15

Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!