| *** tpb has joined #yosys | 00:00 | |
| *** Degi has quit IRC | 00:15 | |
| *** Degi has joined #yosys | 00:16 | |
| *** emeb has quit IRC | 01:45 | |
| *** citypw has joined #yosys | 02:40 | |
| *** citypw has quit IRC | 03:44 | |
| *** citypw has joined #yosys | 05:27 | |
| *** danvet has joined #yosys | 06:06 | |
| *** vidbina has joined #yosys | 06:06 | |
| *** srk has quit IRC | 06:23 | |
| *** srk has joined #yosys | 06:24 | |
| *** vidbina_ has joined #yosys | 07:26 | |
| *** vidbina has quit IRC | 07:27 | |
| *** vidbina_ has quit IRC | 07:35 | |
| *** citypw has quit IRC | 07:44 | |
| *** vidbina_ has joined #yosys | 08:37 | |
| *** adjtm_ has quit IRC | 10:04 | |
| *** citypw has joined #yosys | 10:52 | |
| *** citypw has quit IRC | 11:05 | |
| *** citypw has joined #yosys | 12:06 | |
| corecode | gatecat: nevermind, latest git works great | 14:08 |
|---|---|---|
| corecode | hm, somehow i can't get "show" to run xdot | 14:19 |
| corecode | oh nm. the design is too big :) | 14:19 |
| *** solidtux has quit IRC | 14:23 | |
| *** solidtux has joined #yosys | 14:26 | |
| *** citypw has quit IRC | 14:31 | |
| corecode | how would i visualize the critical path? nextpnr prints a bunch of stuff, but it has been passed through ABC, etc. and now i can't figure out what logic it is | 14:38 |
| corecode | i guess there should be a way to select it in yosys and then visualize it | 14:39 |
| Lofty | corecode: if your Yosys is up to date there's an autoname pass run by default that names things a bit better | 15:39 |
| Lofty | But in general, synthesis through Yosys/ABC is one way, and recovering the initial logic that forms the critical path is...impractical at best | 15:40 |
| Lofty | The nextpnr output really is the thing you have to look at | 15:40 |
| corecode | i don't see any real useful names | 15:47 |
| corecode | i wish there was an easy way to take the nextpnr critical path and process it to stick it into yosys select | 15:47 |
| corecode | i don't quite know how to select cells from nets | 15:48 |
| *** SpaceCoaster has quit IRC | 15:51 | |
| Lofty | corecode: could you paste the nextpnr output? | 16:00 |
| corecode | yes | 16:00 |
| Lofty | And also, which version of Yosys/nextpnr are you using? | 16:01 |
| corecode | https://gist.github.com/e63ee806a1ace6c5da47692a0702b261 | 16:01 |
| corecode | yosys 0.9, nextpnr from git | 16:01 |
| Lofty | Update your Yosys | 16:01 |
| Lofty | 0.9 is ancient | 16:01 |
| corecode | seems to be the latest release tho | 16:02 |
| corecode | nevertheless, it would be good to develop a flow that takes the critical path information and turns it into a select sequence that then allows to dump/show the path | 16:04 |
| corecode | i'm just having problems phrasing the selects | 16:04 |
| corecode | you'd think it would be simple, with all the unique names | 16:04 |
| corecode | but it doesn't work | 16:04 |
| Lofty | corecode: Use Git master | 16:04 |
| Lofty | Yosys' "releases" might as well be snapshots | 16:05 |
| Lofty | 0.9 is like a year and a half old | 16:06 |
| cr1901_modern | ^Agreed. For want of less-loaded language, yosys "changes too fast" | 16:25 |
| *** roamingr1 has quit IRC | 16:25 | |
| corecode | hm, i can't figure out how to select the DFF that is part of the LC | 16:29 |
| corecode | i can only select the LUT | 16:29 |
| Lofty | For real: update your Yosys to current Git master, corecode, you'll find life much easier | 16:31 |
| corecode | oh i see | 16:38 |
| corecode | i think nextpnr merges LUTs and DFFs | 16:39 |
| corecode | but then only retains the LUT name, and calls it "LC" | 16:39 |
| corecode | and then I can't find the DFF anymore | 16:39 |
| Lofty | That's how iCE40 works, yeah | 16:48 |
| corecode | hmm, now nextpnr doesn't want to load yosys's json output anymore | 16:56 |
| corecode | ERROR: JSON module blackbox attribute value is not a number | 16:57 |
| corecode | it's true, it's a string | 16:57 |
| corecode | "blackbox": "00000000000000000000000000000001", | 16:57 |
| corecode | oh, old stuff in /usr/local | 17:11 |
| corecode | today is derp day | 17:11 |
| corecode | ah jolly good, now i'm back to ERROR: Unable to place cell 'LED_DRV_CUR' of type 'SB_LED_DRV_CUR' | 17:12 |
| corecode | and that's on yosys and nextpnr master | 17:12 |
| *** Nazara has quit IRC | 17:26 | |
| *** Nazara has joined #yosys | 17:34 | |
| *** davidlattimore has quit IRC | 17:59 | |
| *** davidlattimore has joined #yosys | 18:01 | |
| *** bwidawsk has quit IRC | 19:55 | |
| *** krispaul has quit IRC | 20:03 | |
| *** kristianpaul has joined #yosys | 20:04 | |
| corecode | thanks for entertaining my derps today | 20:06 |
| corecode | now everything seems to work again, just that nextpnr can't even place the design | 20:06 |
| corecode | previous version of yosys and nextpnr could, so something is up. but i'm at 87% LC utilization | 20:07 |
| *** bwidawsk has joined #yosys | 20:11 | |
| Lofty | You could try again with a different random seed. | 20:37 |
| corecode | i guess | 20:38 |
| corecode | but i think my design needs to be improved | 20:38 |
| *** jophish1 has joined #yosys | 20:44 | |
| *** jophish has quit IRC | 20:46 | |
| *** vidbina_ has quit IRC | 21:06 | |
| *** roamingr1 has joined #yosys | 21:13 | |
| *** pointfree has joined #yosys | 21:22 | |
| *** roamingr1 has quit IRC | 21:23 | |
| *** roamingr1 has joined #yosys | 21:28 | |
| *** adjtm has joined #yosys | 22:01 | |
| *** danvet has quit IRC | 22:07 | |
| *** vidbina_ has joined #yosys | 22:22 | |
| *** vidbina_ has quit IRC | 22:26 | |
| *** roamingr1 has quit IRC | 22:27 | |
| *** lf has quit IRC | 23:14 | |
| *** lf has joined #yosys | 23:14 | |
| *** asu has quit IRC | 23:30 | |
| *** asu has joined #yosys | 23:33 | |
| *** pointfree has quit IRC | 23:48 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!