| sadoon[m] | Hey, I was typing in text so I couldn't tell you then: debian bookworm works extremely well on my talos and is ultra stable even at this alpha stage, you should probably look into it | 00:02 |
|---|---|---|
| sadoon[m] | gnucode: | 00:02 |
| sadoon[m] | It's a matter of a few months till the official release but most everything is frozen already | 00:03 |
| sadoon[m] | Mirrroring bullseye binary and source for the main suite now, shouldn't take too long (hopes) | 00:12 |
| *** gnucode <gnucode!~gnucode@user/jab> has joined #libre-soc | 00:36 | |
| *** gnucode <gnucode!~gnucode@user/jab> has quit IRC | 02:46 | |
| *** lxo <lxo!~lxo@gateway/tor-sasl/lxo> has quit IRC | 04:51 | |
| *** lxo <lxo!~lxo@gateway/tor-sasl/lxo> has joined #libre-soc | 04:51 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has quit IRC | 06:24 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has joined #libre-soc | 06:24 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has quit IRC | 06:29 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.164.60> has joined #libre-soc | 06:30 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.164.60> has quit IRC | 07:10 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has joined #libre-soc | 07:11 | |
| lkcl | toshywoshy, the other one which opens up HPC is the combination of EXTRA-V, Snitch, ZOLC and SVP64. | 08:34 |
| programmerjake | lkcl: note openpower-isa.git now uses https://ftp.libre-soc.org/power-instruction-analyzer/output-for-v0.2.0-7-g95fdd1c.json.gz | 08:34 |
| programmerjake | so if/when you clean up ftp, please don't remove it | 08:34 |
| lkcl | that'll need removing | 08:35 |
| lkcl | (the dependence) | 08:35 |
| lkcl | remember that pia needs to be retired and replaced | 08:35 |
| programmerjake | why? you don't need to download or run pia at all | 08:35 |
| lkcl | this has been discussed and does not need repeating | 08:36 |
| programmerjake | (unless you want to reproduce it) | 08:36 |
| lkcl | exactly | 08:36 |
| lkcl | and that forces a dependence on a tool and a compiler that has already been discussed that it was a severe and costly mistake to depend on and use | 08:36 |
| lkcl | thus it shall be removed | 08:36 |
| lkcl | the test api is to be used for creating and running programs that produce ExpectedResults | 08:37 |
| lkcl | by creating Makefiles and c/assembler binaries that can be directly run on a POWER9 system | 08:37 |
| programmerjake | it doesn't need any python bindings code which was the main issue, i'd expect pia to compile and run unmodified using debian bullseye packages | 08:38 |
| lkcl | you are ignoring - yet again - a decision that has already been made | 08:38 |
| programmerjake | literally just apt install cargo; cargo run | 08:38 |
| lkcl | and in a disrespectful way ignoring that decision | 08:38 |
| lkcl | rust is out | 08:39 |
| lkcl | cargo is out | 08:39 |
| lkcl | pia is out | 08:39 |
| lkcl | you have been told this multiple times | 08:39 |
| programmerjake | well, at least please don't remove it until i fix the ca/ov issues since i'm using it as an easy low-effort reference meanwhile | 08:39 |
| lkcl | indeed. | 08:40 |
| lkcl | easy for you | 08:40 |
| programmerjake | after that the new test can be removed/converted | 08:40 |
| lkcl | there is money available to do the Test API to get it to auto-generate Makefiles | 08:40 |
| lkcl | (from the ExpectedResults objects) | 08:40 |
| programmerjake | i'm using it to create expected results and programs to run the simulator on, i don't want to have to do several weeks of work before i can fix the simulator | 08:42 |
| programmerjake | i added the tests since our current ca/ov tests are woefully inadequate, i found like 10 insns we get wrong results on (assuming my test isn't buggy, i manually verified a few test cases) | 08:45 |
| programmerjake | hmm, what do you think of me translating pia (minus the python bindings) to c or c++? i expect it to take 2-3 days...it should generate identical json | 08:49 |
| programmerjake | that would happen at some later point | 08:50 |
| programmerjake | maybe more than 3 days... | 08:51 |
| lkcl | no i totally get it | 09:17 |
| lkcl | it's fine - many other things to do | 09:18 |
| lkcl | the goal with the TestAPI is that ExpectedResults become the canonical source of information/correctness | 09:18 |
| lkcl | toshywoshy, this is where the HPC rubber hits the road https://libre-soc.org/openpower/sv/SimpleV_rationale/ | 09:19 |
| lkcl | and there's a list of EU collaborators https://www.etp4hpc.eu/ | 09:20 |
| lkcl | the potential reductions in power consumption are 1 getting on for 2 orders of magnitude (!) | 09:21 |
| lkcl | by performing near-memory pre-filtering. | 09:21 |
| lkcl | requires 100% Deterministic Schedules Node/Graph/Memory-walking | 09:21 |
| lkcl | but that is pretty much literally the definition of all and any mapreduce algorithm ever invented | 09:22 |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has quit IRC | 12:38 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.164.210> has joined #libre-soc | 12:39 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.164.210> has quit IRC | 12:54 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has joined #libre-soc | 12:54 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has quit IRC | 15:47 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.54.132> has joined #libre-soc | 16:04 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.54.132> has quit IRC | 16:08 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has joined #libre-soc | 16:09 | |
| sadoon[m] | Currently http://soulserv.xyz/debian-ppc hosts the sid snapshot of ppc64 and powerpc as of today, signed with debian-ports' keys | 19:53 |
| sadoon[m] | Same idea will be applied for sffs, will have a repo for the | 19:54 |
| sadoon[m] | a- debian bullseye source packages mirror | 19:54 |
| sadoon[m] | b- debian bullseye source packages repo for packages that have been modified in order to work with sffs (might not even need this) | 19:54 |
| sadoon[m] | c- debian bullseye sffs binary repo | 19:54 |
| sadoon[m] | Starting with ppc and ppc64 because it's a personal project of mine and I can afford to make mistakes there and learn as opposed to sffs which I want to be perfect | 19:55 |
| lkcl | sadoon[m], fantastic | 20:19 |
| sadoon[m] | :D | 20:20 |
| sadoon[m] | I desperately need to update my website lol | 20:20 |
| sadoon[m] | It's stuck in 2021 | 20:20 |
| sadoon[m] | Seems hugo is all the craze for static sites, might look into it | 20:23 |
| lkcl | dude i run static html pages | 20:23 |
| lkcl | hand-edited! | 20:23 |
| lkcl | http://lkcl.net/reports/ | 20:23 |
| sadoon[m] | Heh, I hate html editing | 20:24 |
| sadoon[m] | python3-markdown was great for converting markdown to html | 20:24 |
| lkcl | i learned the trick of using <pre> </pre> around plaintext a looong time ago | 20:24 |
| sadoon[m] | It's what I used for very long | 20:24 |
| lkcl | do consider ikiwiki, copying the bootstrap template that we added | 20:25 |
| sadoon[m] | Will do | 20:25 |
| sadoon[m] | Thanka | 20:25 |
| sadoon[m] | Thanks* | 20:25 |
| lkcl | i eradicated with prejudice all but the bare minimum of javascript already | 20:25 |
| sadoon[m] | I'm set on keeping mine without any js, it's unnecessary for my use case | 20:31 |
| *** lkcl <lkcl!lkcl@freebnc.bnc4you.xyz> has quit IRC | 22:33 | |
| *** lkcl <lkcl!lkcl@freebnc.bnc4you.xyz> has joined #libre-soc | 22:43 | |
| *** josuah <josuah!~josuah@46.23.94.12> has joined #libre-soc | 22:56 | |
| *** josuah <josuah!~josuah@46.23.94.12> has quit IRC | 22:57 | |
| *** josuah <josuah!~josuah@46.23.94.12> has joined #libre-soc | 23:22 | |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!