| jn__ | lx0: it's still going on, in any case :) | 00:02 |
|---|---|---|
| programmerjake | if you want to be able to fix your stuff: https://www.gofundme.com/f/lets-get-right-to-repair-passed | 03:03 |
| programmerjake | sorry it requires running non-free software... | 03:14 |
| danielp3344 | programmerjake: what kind of legislation is he proposing? | 04:04 |
| programmerjake | I don't know the exact details, but I'd expect him to be pretty good at designing it, he has run an independent mac repair shop for 10 years | 04:06 |
| programmerjake | I'd guess the exact bill isn't completely designed yet | 04:06 |
| danielp3344 | programmerjake: yeah but... seems like it would be easy to cause more harm than good | 04:06 |
| danielp3344 | I'd wanna see the actual bill before giving him money | 04:06 |
| programmerjake | sounds reasonable | 04:07 |
| danielp3344 | IMHO the only way to have hardware you can repair reasonably easily is have manufacturers who will release useful documentation | 04:07 |
| danielp3344 | and I don't see a way to legislate that | 04:07 |
| programmerjake | that, plus actually being able to buy replacement chips | 04:08 |
| danielp3344 | I don't actually see that as a problem | 04:10 |
| programmerjake | it totally is a big problem... | 04:10 |
| danielp3344 | I would be worried it would place undue strain on manufacturers to make them make stuff like that available to individuals | 04:11 |
| danielp3344 | the fact that they don't would seem to just indicate lack of demand | 04:11 |
| programmerjake | what could be a $100 repair turns into an $800 repair if you have to buy a whole replacement motherboard instead of just the chip that died | 04:11 |
| danielp3344 | (or some sort of restriction on competing companies) | 04:12 |
| programmerjake | the fact that they won't is due to apple and others paying them lots of money to prevent them from selling that particular chip to anyone other than apple | 04:12 |
| programmerjake | if it's too much hassle, they can just have a minimum order quantity that's big enough | 04:13 |
| programmerjake | other companies such as mouser will happily split that into smaller amounts | 04:14 |
| danielp3344 | But, if apple is known to make their devices hard to repair shouldn't people just learn not to buy from them | 04:14 |
| danielp3344 | ?* | 04:14 |
| danielp3344 | It seems to me like trying to legislate nasty companies into good ones isn't a good idea really | 04:14 |
| programmerjake | well, nearly every large company is doing the same thing as apple, so there just aren't any mass-market options that are right-to-repair friendly | 04:16 |
| danielp3344 | pine? | 04:16 |
| danielp3344 | they have excellent documentation also | 04:16 |
| programmerjake | last I checked pine isn't selling >100M phones | 04:16 |
| danielp3344 | no but they do sell them | 04:17 |
| danielp3344 | maybe if people start caring more about repairing their devices they'll become more popular | 04:17 |
| programmerjake | that's part of what this direct-ballot thing is about, I imagine part of the funds will be used to educate the public why they should care about right-to-repair and why it is a problem how it is now | 04:19 |
| danielp3344 | I mean I see ways this could harm legitimate companies | 04:19 |
| danielp3344 | while at the same time I don't support trying to legislate established corporations into behaving | 04:20 |
| programmerjake | well, when the companies are going out of their way to make their products unrepairable...that's their fault | 04:21 |
| danielp3344 | we just shouldn't buy from them | 04:21 |
| danielp3344 | honestly I have a hard time imagining how someone who truly cares about repairing stuff buying from apple | 04:22 |
| danielp3344 | err | 04:22 |
| danielp3344 | whatever | 04:23 |
| programmerjake | that doesn't work since the vast majority of consumers only realize they care about repairability after they bought from apple | 04:23 |
| danielp3344 | Is that a problem? | 04:23 |
| danielp3344 | I mean legislating to save people from themselves is another thing I don't like :P | 04:23 |
| danielp3344 | most people will have plenty of chances to buy electronic devices anyway (especially if they buy from apple_ | 04:24 |
| danielp3344 | )* | 04:24 |
| programmerjake | not from apple --- typo :P I meant from any company that's like that | 04:24 |
| danielp3344 | I guess the same argument still applies no? | 04:24 |
| programmerjake | well, at the very least we should make sure the public is informed...currently apple and others can manage to protray themselves as repairable because they do something like sell you the motherboard of a $1500 computer for $1400 and say you can replace the motherboard if your $5 battery charging chip dies | 04:27 |
| programmerjake | you can check out louis rossmann's youtube channel for more details ... unfortunately I don't know of any decent quick summaries... | 04:30 |
| Chips4Makers | Status: I did setup cocotb testbench running svf file. But am using iverilog and that is too slow. If I put the chip in reset I can very fast simulate the get of IDCODE, if I put the chip out of reset it took more than 12 hours to simulate 4 clock cycles of the CPU and 30 are needed to siimulate IDCODE. Am using old CPU but still. | 09:02 |
| Chips4Makers | In order to use verilator with coctb a more recent version is neeeded than provided by debian/Ubuntu. I get compilation errors. Will try existing docker images next to get it going in the short term. | 09:02 |
| Chips4Makers | Will update bug today but need to fix some DRC errors on standard cells and IO first. | 09:02 |
| lkcl | danielp3344: no... they just buy yet another product and throw the broken one in landfill. except in China where they take it to people who work in the factory or know people who do, and get it repaired that way | 11:51 |
| lkcl | i started the EOMA initiative to solve this one for "the-rest-of-the-world" | 11:52 |
| lkcl | Chips4Makers, how do we convert phlib (or other cell library) VBE into VHDL/VST? | 16:06 |
| lkcl | i see in alliance there is the sxlib cell library, that has had VHDL added to it | 16:13 |
| lkcl | found it | 16:24 |
| lkcl | vasy | 16:24 |
| lkcl | vasy -s -I vbe {filename.vbe} outputwithoutextension | 16:41 |
| lkcl | Chips4Makers, https://git.libre-soc.org/?p=soc-cocotb-sim.git;a=tree;f=ls180;hb=HEAD | 18:53 |
| lkcl | i got fed up of trying to do it via Makefiles | 18:53 |
| lkcl | instead wrote two python scripts | 18:53 |
| lkcl | one to call vasy in the chroot, to get nsxlib VBE converted to VHD | 18:53 |
| lkcl | the second to build the object files using ghdl | 18:54 |
| lkcl | vasy outputted in standard VHDL format, not Alliance format | 18:55 |
| klys | https://paste.debian.net/1191835/ | 19:02 |
| klys | see you in a bit after lunch | 19:03 |
| Chips4Makers | @lkcl: The c4m-flexlib standard cells are delivered with verilog and vhdl behavior models. | 19:05 |
| programmerjake | you also need to install ieee754fpu.git and nmutil.git from git.libre-soc.org | 19:06 |
| programmerjake | klys ^ : some others too -- I think there's a list somewhere | 19:07 |
| klys | thx | 19:15 |
| lkcl | Chips4Makers: when you say "delivered with", how is that done? there's nothing in here https://gitlab.com/Chips4Makers/c4m-flexcell/-/tree/master/c4m/flexcell | 19:21 |
| lkcl | and i'm not seeing any other directories here https://gitlab.com/Chips4Makers/c4m-flexcell/-/tree/master/c4m/flexcell | 19:22 |
| lkcl | klys: the scripts to install the dependencies and the libraries are here | 19:22 |
| lkcl | https://git.libre-soc.org/?p=dev-env-setup.git;a=blob;f=hdl-dev-repos;hb=HEAD | 19:22 |
| lkcl | and here | 19:23 |
| lkcl | https://git.libre-soc.org/?p=dev-env-setup.git;a=blob;f=install-hdl-apt-reqs;hb=HEAD | 19:23 |
| lkcl | if you'd prefer to do that in a chroot, the script to do that is here | 19:23 |
| lkcl | https://git.libre-soc.org/?p=dev-env-setup.git;a=blob;f=mk-deb-chroot;hb=HEAD | 19:23 |
| lkcl | annoyingly, you *may* need to remove the git tag on c4m-jtag | 19:24 |
| lkcl | git tag -d 24jan2021ls180 | 19:24 |
| Chips4Makers | lkcl: They are part of the release package, see for example FreePDK45 release package: https://gitlab.com/Chips4Makers/c4m-pdk-freepdk45/-/releases/v0.0.1 | 19:26 |
| lkcl | okaaay excellent | 19:26 |
| lkcl | is there one for nsxlib? | 19:27 |
| Chips4Makers | No, there you will need to convert VBE files. | 19:28 |
| lkcl | Jean-Paul hasn't completed the doDesign for FreePDK so i am... yes, that's what the scripts are for :) | 19:28 |
| lkcl | i am currently running into "vdd not found" | 19:28 |
| Chips4Makers | As said plan is to get libresoc synthesized and P&Red with FreePDK45 release package, but are going to do cocotb for 0.18 first. | 19:29 |
| lkcl | that needs Jean-Paul to complete the doDesign.py file | 19:30 |
| lkcl | there only exists a doDesign.py file for nsxlib | 19:30 |
| lkcl | and for the NDA'd TSMC018 | 19:30 |
| lkcl | not for FreePDK45 | 19:30 |
| lkcl | no doDesign.py https://git.libre-soc.org/?p=soclayout.git;a=tree;f=experiments9/symbolic;hb=HEAD | 19:31 |
| Chips4Makers | I'll likely do my own version of a pnr.py file for FreePDK45; don't like the alliance-check-toolkit setup. | 19:31 |
| lkcl | great. you have commit rights to soclayout (i will make sure) | 19:31 |
| lkcl | yes you do | 19:32 |
| lkcl | i am currently fighting with VST files from P&R | 19:32 |
| lkcl | they are in Alliance subset | 19:32 |
| lkcl | running vasy -s -I vst ls180.vst ls180 | 19:33 |
| Chips4Makers | Regarding c4m-jtag. I may upstream some of your changes. Can I rebase master branch and for push for libre-soc repo afterwards or should I merge new upstream in master. | 19:33 |
| lkcl | this *should* convert from Alliance VST subset into standard VHDL | 19:33 |
| Chips4Makers | ..for push.. -> ..force push.. | 19:34 |
| lkcl | yes go for it. errr... force push no please don't | 19:34 |
| lkcl | that causes serious problems: everyone has to go through an extremely inconvenient procedure to restore to latest master | 19:34 |
| Chips4Makers | OK will merge new upstream master then in libresoc git master then. | 19:35 |
| lkcl | appreciated | 19:35 |
| lkcl | do you have any suggestions on how to solve the conversion of VST to VHDL using alliance vasy command? it doesn't understand "VSS" or "VDD" | 19:36 |
| lkcl | or should we ask Jean-Paul | 19:36 |
| Chips4Makers | I'm no gitolite user but I think you can disallow force push in, especially for master branch. | 19:36 |
| Chips4Makers | Don't have time to bother with that ATM, sorry. | 19:36 |
| lkcl | Chips4Makers: ack. i will ask JP. | 19:37 |
| lkcl | cc you because it's relevant, you can't run cocotb ghdl simulations if the VST files are "faulty" because of syntax errors with unknown "VDD" or "VSS" in them | 19:37 |
| programmerjake | hmm, check a2i or a2o's source code, iirc they also have vdd and vss in them, maybe you can just insert the missing declaration | 19:46 |
| programmerjake | lkcl ^ | 19:46 |
| lkcl | programmerjake: ahh good call | 19:56 |
| * klys now has output from ./mkpinmux.sh | 20:45 | |
| klys | looks like vhdl or something. does this go into verilator? | 20:47 |
| lkcl | klys, yes. we need to double-check that coriolis2's modifications do npt alter the functionality | 21:55 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!