| lkcl | octavius, awesome - paid already! https://bugs.libre-soc.org/show_bug.cgi?id=763 | 00:14 |
|---|---|---|
| lkcl | remember to update your page https://libre-soc.org/andreym/ so you can keep track | 00:15 |
| lkcl | feel free to cheat if you like by running the budget-sync program | 00:16 |
| octavius | Yep, I was amazed as well! Really rapid turn-around. I update the wiki as well | 00:19 |
| *** A_Dragon is now known as NotRegistered | 02:36 | |
| *** NotRegistered is now known as A_Dragon | 02:36 | |
| programmerjake | lkcl: your search+replace was a little overactive: https://git.libre-soc.org/?p=soc.git;a=blob;f=src/soc/experiment/icache.py;h=57b738c6641bbe7e7975fe16e6d6fd71957f7268;hb=57b738c6641bbe7e7975fe16e6d6fd71957f7268#l74 | 04:07 |
| programmerjake | i added grev to soc, just waiting on the formal proof to finish running, it takes at least 20min with no indication of doing anything | 05:42 |
| programmerjake | finally finished (it finished a while ago)! | 06:05 |
| lkcl | programmerjake, yeah i was in the middle of doing that one (by hand) and must have done "git commit -a" not "git commit" | 11:17 |
| lkcl | fantastic | 11:17 |
| lkcl | yeah that sounds pretty high but nothing like as bad as if using a multiply operator | 11:18 |
| lkcl | oh you added it to the shift_rot proof, yeah that one iirc already takes a horrendous amount of time | 11:42 |
| lkcl | btw if you wanted to get quite a lot of $EUR quite quickly, completing the shift_rot formal correctness proofs would be a good way to do it | 11:44 |
| lkcl | 3 people have tried doing it already, there's something really skewy going on, some input combinations that are "wiggling about" and we havne't been able to nail them down | 11:45 |
| lkcl | they're combinations of inputs to the module that are not actually used (and never will be) by actual operations | 11:46 |
| lkcl | but unless you actually tell a Formal Correctness Proof that (with Assumes) it happily finds them for you sigh | 11:47 |
| programmerjake | i can work on the shift/rot proof | 21:47 |
| lkcl | that would be brilliant | 22:21 |
| * lkcl has a VERSA_ECP5 running at 50mhz with Jean Thomas' DDR3 PHY/Controller | 22:21 | |
| lkcl | which is great news | 22:21 |
| lkcl | it means it stands a chance of booting linux on a 45k LUT4 FPGA | 22:22 |
| programmerjake | yay! | 22:58 |
| programmerjake | lkcl, do you need to do anything else for the nlnet gigabit router grant, or are you just waiting on nlnet? | 23:00 |
| lkcl | it needs to go to michiel | 23:07 |
| programmerjake | if there's something you haven't got around to yet for that, can you do that soon? thx! | 23:08 |
| rsc | What's the plan for the gigabit router in detail? Making it a multi-port router? | 23:21 |
| programmerjake | https://libre-soc.org/nlnet_2021_crypto_router/ | 23:25 |
| programmerjake | mostly that we're building a router as an excuse to improve cryptographic algorithms on libre-soc. i'd assume the router will have at least 2 ethernet ports...iirc it was planned to have 4 tho I can't find any references at the moment and i may be misremembering | 23:27 |
| programmerjake | also to get libre-soc onto skywater130 iirc | 23:28 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!