whitequark changed the topic of #yosys to: Yosys Open SYnthesis Suite: https://github.com/YosysHQ/yosys/ | Channel logs: https://libera.irclog.whitequark.org/yosys/ | Bridged to #yosys:matrix.org
tpb has quit [Remote host closed the connection]
tpb has joined #yosys
so-offish has joined #yosys
so-offish has quit [Ping timeout: 276 seconds]
<philtor> Why to the latest builds of oss-cad-suite not include nextpnr-gowin ?
<philtor> for example, I find one in the 20230210 release build, but not in later ones from 2024 or from today.
<philtor> s/to/do/
bjorkintosh has joined #yosys
bjorkintosh has quit [Changing host]
bjorkintosh has joined #yosys
whitequark[cis] has joined #yosys
<whitequark[cis]> nextpnr-gowin is deprecated and slated for removal, use nextpnr-himbaechel
<philtor> Ok, trying that. seems like several options like --family and --cst are deprecated in nextpnr-himbaechel
<philtor> I have (from the makefile for this example): nextpnr-himbaechel --json counter.json --write counter_pnr.json --freq 27 --device GW1NR-LV9QN88PC6/I5 --cst tangnano9k.cst
famubu[m] has joined #yosys
<famubu[m]> You can use like this instead: nextpnr-himbaechel --device GW1NR-LV9QN88PC6/I5 --vopt family=GW1N-9C --vopt cst=tangnano9k.cst
<philtor> ack... core dumpts. it seems I should just stick with the older nextpnr-gowin
<philtor> that's probably why they recommend the 2023 02 10 release build in the Lushay tang nano 9K tutorial
<philtor> ... back to that one
<philtor> ah, 2024_2_16 build of nextpnr-himbaechel works with that commandline. The one that core dumps is today's build.
<philtor> 2025_2_11
bjorkintosh has quit [Quit: "Every day, computers are making people easier to use." David Temkin]
<whitequark[cis]> nextpnr-gowin has known routing bugs that completely break designs
<philtor> I'll file the bug on nextpnr github
<philtor> I just had the idea that maybe the core dump resulted because I was using the counter.json created from that 2023 version of yosys, but just tried it with today's build's yosys and got the same result (core dump) from nextpnr-himbaechel
<philtor> Oh, sorry, looks like a dup. Maybe a problem with how they're building it in oss-cad-suite? (based on the comments in #1452)
<famubu[m]> I had been trying to build nextpnr from source to get the latest nextpnr-himbaechel. But got error saying it can't find a header file x86_64-linux-gnu/python3.9/pyconfig.h though it is present in  /usr/include/x86_64-linux-gnu/python3.9/. Any idea how the compiler can be made to use that header file?
<famubu[m]> I had mentioned it in this issue: https://github.com/YosysHQ/nextpnr/issues/1446
<whitequark[cis]> I've never seen this error; it also seems specific to your system
<whitequark[cis]> you could simply disable Python
<famubu[m]> How can I disable python? And what features will be unavailable if python is disabled?
<famubu[m]> I needed apycula for a gowin fpga
<famubu[m]> <philtor> "ah, 2024_2_16 build of nextpnr-..." <- Can we download already built executables from somewhere?
<whitequark[cis]> cmake -DBUILD_PYTHON=OFF
bjorkintosh has joined #yosys
bjorkintosh has joined #yosys
bjorkintosh has quit [Changing host]
xutaxkamay has quit [Ping timeout: 268 seconds]
xutaxkamay has joined #yosys
<famubu[m]> Thanks. I was able to build it with that.
<famubu[m]> I then tried build/nextpnr-himbaechel --json counter.json --freq 27 --write counter_pnr.json --device GW1NR-LV9QN88PC6/I5 -l pnr.log --report report.json --vopt family=GW1N-9C --vopt cst=tangnano9k.cst --placed-svg placed.svg --routed-svg routed.svg with a simple ring counter file.
<famubu[m]> But the placed.svg and routed.svg files are blank. Any idea why?
<famubu[m]> The input files are as mentioned at https://github.com/YosysHQ/nextpnr/issues/1446
<famubu[m]> Could it be because the design is too trivial?
lofty[m] has joined #yosys
<lofty[m]> I mean, they're purely aesthetic anyway
<famubu[m]> I kind of wanted to see what the images generated by nextpnr would look like.
<famubu[m]> The body of the svg file is this: `<rect x="0" y="0" width="100%" height="100%" stroke="#fff" fill="#fff"/>`