Thursday, 2023-08-17

*** sauce <sauce!> has quit IRC00:21
*** gnucode <gnucode!~gnucode@user/jab> has quit IRC02:12
*** gnucode <gnucode!~gnucode@user/jab> has joined #libre-soc02:12
*** gnucode <gnucode!~gnucode@user/jab> has quit IRC02:52
openpowerbot[irc] <sadoon[m]1> Build time is reduced by over 95% when I use -j72 ;)06:19
openpowerbot[irc] <sadoon[m]1> So yeah depends on the system06:19
openpowerbot[irc] <sadoon[m]1> uhhhhh 13 seconds single thread hello_world on my Alder Lake laptop o_O08:18
openpowerbot[irc] <sadoon[m]1> Looks like I'll be using my laptop more for this kind of stuff08:18
openpowerbot[irc] <sadoon[m]1> 11 seconds on 4 threads08:19
openpowerbot[irc] <sadoon[m]1> 9 seconds on 308:22
openpowerbot[irc] <sadoon[m]1> 8s to micropython on 308:26
*** ghostmansd <ghostmansd!> has quit IRC08:26
*** openpowerbot <openpowerbot!> has quit IRC09:08
*** octavius <octavius!> has joined #libre-soc10:18
octaviusVery impressive numbers10:18
*** openpowerbot <openpowerbot!> has joined #libre-soc10:35
markos_hey guys, I missed the meeting on Tuesday, what did I miss?10:54
octaviusSadoon/Luke/Toshaan talking about the SFFS building issues (as always software assuming VSX is present...)11:01
openpowerbot[irc] <sadoon[m]1> Basically :)11:01
octaviusI also was streaming microwatt simulation booting the linux kernel (only took 2 and a bit hours with THREADS=3 on i7 4th gen CPU)11:01
*** ghostmansd <ghostmansd!> has joined #libre-soc11:04
openpowerbot[irc] <sadoon[m]1> Huh! That's good!11:04
octaviusLuke mentioned issues with using Wishbone Bus, and why for future designs (Multi-Issue, Out-of-Order), we'll need to use something like AXI for the core/memory interconnect. AXI allows to check if a request (write or read) can be completed before actually committing to a request. Wishbone doesn't have such provisions11:04
openpowerbot[irc] <sadoon[m]1> So on my Alder Lake laptop it now took 1 hour to reach the 1.7 second mark with 3 threads as well11:05
openpowerbot[irc] <sadoon[m]1> I'm not sure how much longer it will take to reach the shell11:05
octaviusWhat does it say at 1.7s?11:06
openpowerbot[irc] <sadoon[m]1> I'm travelling but I made sure to download everything I need, I don't connect to wifi while travelling (obvious reasons) and my carrier doesn't allow tethering on roaming11:06
openpowerbot[irc] <sadoon[m]1> I forgot but I'll let you know once I wake up the laptop11:06
octaviusah fair enough, have a good trip :)11:06
openpowerbot[irc] <sadoon[m]1> Just arrived at the hotel11:06
openpowerbot[irc] <sadoon[m]1> Thanks!11:06
markos_octavius, oh that's cool, on the arty a7?11:15
markos_oh, that's a software simulation11:17
openpowerbot[irc] <sadoon[m]1> <openpowerbot> "[irc] <octavius> What does it..." <- [1.719667] io scheduler mq-deadline registered11:22
openpowerbot[irc] <sadoon[m]1> Exactly at 1h12m of runtime now11:22
openpowerbot[irc] <sadoon[m]1> Putting it back to sleep, will continue later so I don't miss the moment it reaches shell11:23
openpowerbot[irc] <sadoon[m]1> Nvm staying at the hotel for a few it's going fast now11:29
openpowerbot[irc] <sadoon[m]1> We should consider using karg "quiet" to make it boot faster when testing binaries, I'll do that anyways :p11:30
*** ghostmansd <ghostmansd!> has quit IRC11:35
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC11:40
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc11:40
openpowerbot[irc] <sadoon[m]1> Nice to see the Intel scheduler for linux actually not being stupid and purposefully using unshared threads, one per core all on the P-cores11:46
openpowerbot[irc] <sadoon[m]1> Perhaps an -march=native will also help out if that's not set already11:46
openpowerbot[irc] <sadoon[m]1> 1h32m in, line TERM=linux11:46
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC12:05
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc12:06
openpowerbot[irc] <sadoon[m]1> 1h55m login prompt12:12
openpowerbot[irc] <sadoon[m]1> That's worse than I expected but still very good12:12
openpowerbot[irc] <sadoon[m]1> There's about a second of delay typing characters which is also very good, expected much much worse12:14
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC12:37
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc12:38
lkclsadoon, fantastic! only one second is extremely responsive, it means you're getting echo back to /bin/busybox with only around... 10,000 instructions executed13:19
openpowerbot[irc] <sadoon[m]1> Hey that's much better than the 5,000 you were talking about yesterday huh13:20
lkclif busybox also includes chroot and/or pivot_root you can rebuild the cpio disk (remember my warning about paths being absolute) with debian (or gentoo) /bin/bash plus dynamic libraries in a subdirectory.13:20
openpowerbot[irc] <sadoon[m]1> Maybe a low end but modern Intel machine can be fitted in my collection for these cases13:21
lkclok 5,000 / sec.  you can count them by looking at the pc logs (also an option to the verilator program)13:21
openpowerbot[irc] <sadoon[m]1> This might unironically bring me back to my overclocking days13:21
lkclif you only have the one machine available please don't13:21
openpowerbot[irc] <sadoon[m]1> Nah this is a laptop13:22
openpowerbot[irc] <sadoon[m]1> Running at 3.8ghz max13:22
lkclnot bad.13:22
openpowerbot[irc] <sadoon[m]1> Imagine what a 5.5GHz machine with DDR5 7000 or whatever it was13:22
openpowerbot[irc] <sadoon[m]1> But anyways I'm hoping that simulation won't be a big part of what we need to do13:23
lkclerr duh :)13:23
openpowerbot[irc] <sadoon[m]1> I'd rather get it done on an FPGA and that at least to some degree appreciates extra threads13:23
lkclalthough if snapshots *do* work then you can resume from known-good and save 2 hours13:23
openpowerbot[irc] <sadoon[m]1> Oh speaking of which, almost filled my SSD yes13:24
openpowerbot[irc] <sadoon[m]1> I noticed and deleted them13:24
openpowerbot[irc] <sadoon[m]1> (Kept the most recent 40 or so)13:24
lkclthe fact that it is an initramfs may make that impossible... *unless*... you reserve a *second* memory area and work out how to "mount that into memory"13:24
lkcli did warn you!13:24
lkclit is completely unnecessary.13:25
lkcllike... *really* unnecessary.13:25
openpowerbot[irc] <sadoon[m]1> Yeah hehe13:25
openpowerbot[irc] <sadoon[m]1> Anyways Luke I do have an important question before moving forward13:25
lkclthe only reason i was saving that frequently was because it was taking so long and i had an O(N^2) iterative investigation (where N was getting to *18 hours*)13:26
openpowerbot[irc] <sadoon[m]1> Debian is mostly built, few failures of not so necessary stuff (systemd, not needed for chroot)... (full message at <>)13:27
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC13:27
lkclsadoon that's too long13:27
openpowerbot[irc] <sadoon[m]1> Oh my bad13:27
openpowerbot[irc] <sadoon[m]1> IRC13:27
lkclcan you please re-type so that it does not trigger " (full message at <>)"13:27
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc13:27
openpowerbot[irc] <sadoon[m]1> Debian is mostly built, few failures of not so necessary stuff (systemd, not needed for chroot)13:27
openpowerbot[irc] <sadoon[m]1> Gentoo is done And objdump testing works13:28
lkclok so where is the "reproducible build" script for debian?13:28
openpowerbot[irc] <sadoon[m]1> But as far as qemu goes I really don't think it'll be possible in a timely manner13:28
openpowerbot[irc] <sadoon[m]1> It'll be on the repos before September13:28
openpowerbot[irc] <sadoon[m]1> All of it13:28
openpowerbot[irc] <sadoon[m]1> I'm just worried about qemu13:28
lkclno, you need to commit it _now_13:28
lkcldon't be.13:28
lkclif it is too much then it is too much13:29
lkclplease read the email i just sent.13:29
openpowerbot[irc] <sadoon[m]1> One sec13:29
lkclyou should never have under *any* circumstances been working "privately"13:29
lkcli have raised this a number of times13:29
lkcl*everything* that you do has to be public, published, and available *at the time it is done*.13:29
lkclno private notes "that you will publish later"13:29
lkclno private repositories "that you will publish later"13:30
lkclno private conversations "that you have to go through IRC or email and put into the bugtracker later after several days of searching"13:30
lkclthis is *extremely important* and i have asked several times13:31
lkcl"i'll publish it when it's finished" is a massive red flag and an extreme risk13:31
openpowerbot[irc] <sadoon[m]1> Oh oh dw at all, not using github gitlab or even my own site13:31
openpowerbot[irc] <sadoon[m]1> didn't even use git yet, everything I did is basically already in attachments on the bugreport13:31
openpowerbot[irc] <sadoon[m]1> It's just that I need to put it together on the git repo for libre-soc13:32
lkcltoshywoshy did the NGI POINTER work, but did it privately. and hasn't published it, so you're now having to re-discover it and ask him once a week (tuesday) for hints13:32
openpowerbot[irc] <sadoon[m]1> For debian the whole thing is just the build script and the dpkg-buildflags, and finally the repo script which I am finishing13:33
lkclthen you should indeed have used git13:33
openpowerbot[irc] <sadoon[m]1> I already published what is finished13:33
lkcland you should be committing that repo script and the build script *as you make modifications*13:33
openpowerbot[irc] <sadoon[m]1> Yes I should have, sorry for that13:33
lkclwith commit messages saying "fixed this because of error Y"13:33
lkclnow if i ask you to remember what you did, can you reproduce it?13:33
openpowerbot[irc] <sadoon[m]1> Luckily yes, but I understand your concern13:34
openpowerbot[irc] <sadoon[m]1> I'll commit everything tonight13:34
openpowerbot[irc] <sadoon[m]1> Before working on anything else13:34
lkclit is an "evidence trail" that the EU Auditor will want to see - they may not understand it but they are going to be extremely nervous about paying out EUR 5,000 for something that only has one commit and no evidence that it took EUR 5,000 worth of work to do it13:34
lkcl(which *NLnet* need the confidence to be able to explain and justify)13:35
octaviuslkcl, are these scripts going to go into "dev-env-setup" repo, or will sadoon need a new repo on
lkcloctavius, that's up to all of you to decide.13:35
octaviusIf a new one, then someone with permissions needs to create a blank repo and give sadoon permissions13:35
lkclthat'll be me. see the email i just sent, which explains the process.13:36
octaviusThen let's make a new repo, not sure what a good name would be13:36
openpowerbot[irc] <sadoon[m]1> I prefer a separate repo so I can't mess anything up with dev-scripts, they're very essential13:36
lkcl(and is another thing to go on the list)13:36
lkclsadoon, that's what branches are for13:36
openpowerbot[irc] <sadoon[m]1> In that case then yes it can be done in dev-scripts13:36
lkcldon't be nervous: if you are, get over it. you cannot do any "damage" unless you do something really stupid like a force-master push13:36
openpowerbot[irc] <sadoon[m]1> If you guys prefer13:37
lkclyes because it is a developer script13:37
lkcli'll add you to write perms13:37
octaviusAh perfect, sadoon, make sure to commit to;a=summary13:37
openpowerbot[irc] <sadoon[m]1> So something like ./build-debian and ./build-gentoo which would automate it all eh?13:37
openpowerbot[irc] <sadoon[m]1> Alright sounds good but I suspect it'll be multiple scripts13:38
octaviusDon't worry about multiple scripts, as long as the process can be followed from start to finish13:38
lkclsadoon, done. you can check with "ssh"13:38
openpowerbot[irc] <sadoon[m]1> Alright, then after that a wiki page of course13:38
lkcloctavius: it's down to sadoon to decide what he feels is the best way to manage what he is doing13:39
openpowerbot[irc] <sadoon[m]1> There was something I was getting to with the qemu thing13:39
lkcleach of us can advise but it is his responsibility13:39
lkclqemu, yes13:39
lkclas long as you have a *reasonable justification* because you did the investigation *and wrote up a report*13:40
lkcl(i.e. you actually *tried* to do EUR 2,500 worth of work, but the "trying" itself turned out to *be* EUR 2,500 worth of work)13:40
lkclthen that's "good enough" and we can apply for the next grant13:40
openpowerbot[irc] <sadoon[m]1> It'd be much easier for me to actually port microwatt to an intel FPGA so we get good performance and that would be perfect for testing and be a definitive answer to "is it really compliant with no VSX"13:40
octaviusMake sure to add a new libre-soc wiki page documenting how to run the scripts, I'm guessing under the HDL_workflow. You can copy the format of the other tutorial pages (like microwatt)13:40
lkcl"port to an intel FPGA".... does that include the money required to get the FOSS-Intel-FPGA toolchain compiled13:41
openpowerbot[irc] <sadoon[m]1> Yeah I spent weeks trying to get qemu to behave so I guess that's that13:41
lkcland the dev-scripts written to automate the building of that FOSS intel FPGA toolchain?13:42
openpowerbot[irc] <sadoon[m]1> Do we absolutely need a FOSS toolchain for testing this?13:42
lkclso document what it was that you did13:42
lkcland why it did not work13:42
lkclohhh yes.13:42
openpowerbot[irc] <sadoon[m]1> Will do13:42
lkclthis is a LIBRE project not an OPEN project13:42
lkcllibre **MEANS** libre.13:42
openpowerbot[irc] <sadoon[m]1> Unfortunately my libre FPGA still didn't arrive. Sigh13:43
openpowerbot[irc] <sadoon[m]1> I'll look into the libre options13:43
lkcli only added arty-a7-100t to the list of FPGA boards when reports that nextpnr-xilinx was working.13:43
octaviuslkcl, but does this requirement apply everywhere? ls180 asic was implemented in a TSMC library, but we have a libre equivalent using and different pdk13:43
lkcldigilent arty a7 100t plus 2 Quad HyperRAM PMODs from 1bitsquared.de13:43
openpowerbot[irc] <sadoon[m]1> Maybe things have changed since.. November 2022 when I ordered the ULX13:43
lkcloctavius: ls180 was implemented *BY ME* using *FOSS* cell libraries13:44
openpowerbot[irc] <sadoon[m]1> openpowerbot: I'll order it no questions asked if it works13:44
openpowerbot[irc] <sadoon[m]1> dw about the costs I'd rather just get it working ASAP13:44
lkclany coincidence of *LIP6* having happened to have replaced those *FOSS* cell libraries with *TSMC* cell libraries is entirely not my - our - concern13:45
lkclthey work. don't get the single PMOD by mistake13:45
lkcland don't get the arty-35t by mistake either13:45
lkclit's *specifically*13:45
octaviusAh ok, makes sense13:45
openpowerbot[irc] <sadoon[m]1> Perfect13:45
lkclQTY2of 1bitsquared Quad HyperRAMs13:46
lkclthey're on the wiki somewhere13:46
lkcl256 megabit. you may have to use the US 1bitsquared site13:47
octaviusSadly both EU and US stores are sold out of HyperRAM13:48
octaviusThat's why we had the conversation about ordering PCBAs13:49
lkcldamnit we may have to do that13:49
markos_would it not be worth doing the time investment to get the ddr3 working on the fpgas?13:49
octaviusShall I start thinking about that? We could probably get an affordable PCBA to do it13:50
openpowerbot[irc] <sadoon[m]1> Is ddr3 gonna benefit us when the core will probably run at 50mhz?13:51
lkclmarkos_, remember what i said about it being "really complicated"?13:52
lkcli spent *three months* on gram and associated firmware13:52
markos_that's why I said "invest"13:52
lkclit got so bad i had to stop it, do hyperram instead (which was 2 weeks flat, to 100% working)13:53
lkcland then assigned tplaten to attempt to solve it.13:53
markos_it's not something that's easy, but it's something that will be needed eventually13:53
lkcleighteen *months* later he has not been able to complete it13:53
markos_it's definitely doable, litex guys seem to have done it13:54
lkclit needs one person - a proper FPGA/dev/board-bringup/low-level engineer - to focus on it and nothing else13:54
markos_maybe we're missing something13:54
lkclno: they managed to botch it up and hack it together in such a way that the *only people* who can understand it is themselves13:54
lkclwhich guarantees that they get further contract work at very high rates13:55
markos_you just described half the foss projects out there :)13:55
lkclthe gate/LUT efficiency of what they've written is... extremely poor.13:55
markos_I don't disagree with the pmods solution, but it's a temporary solution13:55
lkclthis is a well-known fact btw13:56
markos_esp if you go the route of a more powerful fpga board13:56
lkclthat if i do not state "it is a well-known fact" the fuckers will attempt to make YET ANOTHER formal complaint to the OpenPOWER Foundation13:56
markos_I don't folow13:56
lkclso if you are a litex developer reading this: fuck off and don't you ever do that again when you know damn fucking well that the quality of your work is not up to scratch13:57
lkcli made "public statements" three years ago about the quality of litex - which everyone *privately* knew13:57
lkcland one of the developers of litex made a formal complaint about it to the OpenPOWER Foundation.13:58
lkcldavid had to get involved.13:58
lkclit was one of the many many attacks on this project that i haven't told you about in detail.13:58
lkcli need to go out in the sun.13:59
* lkcl later13:59
openpowerbot[irc] <sadoon[m]1> Bye!14:00
markos_lkcl, sorry about your trouble with litex, but that's not my point, even the worse project might have/use a trick that might point to a solution in another problem/project14:03
markos_tbh, I don't care about politics, if it's foss it's visible to everyone and quality of code is apparent to everyone that cares. People can judge by themselves. My point is that if they're doing it, we can do it too.14:07
lkclmarkos_, gram is a port of litex dram.15:05
lkcl(to nmigen)15:05
openpowerbot[irc] <sadoon[m]1> Ok finally got internet working by changing carriers temporarily and paying extra.. eh15:20
markos_understood, ok, in any case, just saying we might want to invest some extra time in this, in parallel to the hyperram pmods, which we have to buy anyways15:20
markos_also, bad news, the nexys video, out of the 4-pmod connectors, one is for analog signals only15:24
markos_so only 3 can be used for hyperram pmods, I don't know if there is a single pmod for hyperram15:25
openpowerbot[irc] <sadoon[m]1> lkcl sent you an email with my ssh key for the git repos15:35
openpowerbot[irc] <sadoon[m]1> nvm, stupid email issues.. probably gmail not liking something15:36
*** octavius <octavius!> has quit IRC17:03
*** octavius <octavius!~octavius@> has joined #libre-soc17:04
lkclsadoon: you already have.17:16
lkclssh-rsa AAAAB3NzaC1yc2EA.....bzaFAzjB6aflOQSq9t3Yc= sadoon@HaswellS7Devuan17:17
openpowerbot[irc] <sadoon[m]1> Ah, I still have it back at home I suppose, but not on this laptop17:21
openpowerbot[irc] <sadoon[m]1> Since I'm going to be away for almost a week it'd be better to replace it so I can actually commit while I'm on this short vacation17:22
openpowerbot[irc] <sadoon[m]1> And then I'd just keep using the new one17:22
openpowerbot[irc] <programmerjake> just add your laptop key, don't replace can have more than 1 ssh key17:23
lkcli'm not going to replace it, i'll add a second one17:23
openpowerbot[irc] <sadoon[m]1> another option yeah17:23
openpowerbot[irc] <programmerjake> unless your old one was compromised somehow?17:23
openpowerbot[irc] <sadoon[m]1> Nah17:23
lkclno i don't have it. please drop it here inline.17:24
openpowerbot[irc] <sadoon[m]1> It's just my old laptop which I still have17:25
openpowerbot[irc] <sadoon[m]1> Just not on me now17:25
openpowerbot[irc] <sadoon[m]1> I'll send the pubkey in this chat later since email is having some issues ugh17:25
openpowerbot[irc] <programmerjake> or if you're never going to use your old laptop to commit, in which case removing the old ssh key is a good idea17:25
lkclbtw is storing messages for you and *not* informing you that you are no longer using it17:25
lkclyou have several messages sent to you now that you've not responded to17:26
openpowerbot[irc] <sadoon[m]1> It's still mine17:26
openpowerbot[irc] <sadoon[m]1> I don't plan to remove it anytime soon for that reason specifically17:26
openpowerbot[irc] <sadoon[m]1> I don't want it to be used maliciously17:26
openpowerbot[irc] <programmerjake> i'm assuming you're referring to email and not ssh key?17:27
openpowerbot[irc] <sadoon[m]1> Yes the email but also true for the ssh key17:27
openpowerbot[irc] <sadoon[m]1> They're all still with me guys ,:)17:28
openpowerbot[irc] <sadoon[m]1> Gotta go afk for a while, ttyl17:28
openpowerbot[irc] <programmerjake> ttyl17:29
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC17:48
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc17:48
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC17:51
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc17:51
*** octavius <octavius!~octavius@> has quit IRC17:57
*** octavius <octavius!> has joined #libre-soc17:58
*** ghostmansd <ghostmansd!> has joined #libre-soc17:58
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC18:17
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc18:18
openpowerbot[irc] <sadoon[m]1> lkcl this is my ssh pubkey18:22
openpowerbot[irc] <sadoon[m]1> ssh-rsa18:22
openpowerbot[irc] <sadoon[m]1> ega8jZdg6nWX4XwDaNwURu1zIovl7UCAR2vCEois6SOqVAejpXZIJv2l5P+/LCtgeVtSvFc3l2qZIoHzSQpQOatoTlTX6WOgdBcjSE= sadoon@SwiftUnity18:22
openpowerbot[irc] <sadoon[m]1> AAAAB3NzaC1yc2EAAAADAQABAAABgQCgf+Tsrcigx4wJWpoXyYh2nRYKrAVJVA4nozaqdvR7UHYuKgdopY1w8P1UE3GSsLAn6byzPLy8zlUEDmQlTDwSJIJcIINyPYxPNdniKHbUbisiT2FyObx+SX83EayDFIxeOcsiy9hVy27SxopBYPZ/fFGFX99iKSRywi6HSUwPvOvcfD43k1BVLhNtT56DRc6fmrpoc32as5R2Y4uHkoHj6vSNFaeRGwjFO+i0QjLbrvl8LDw72NHOLz0db5kXvVsX2B0vJw4miD4ccotFI6Z/te8P+fRp3KqgVBrgxibg9rVLrsSx/FN8aZSc/L4bgOrnro80UQKwetwU5uNyWiquQ4wmSAxL+Ekctp5m0jJBQfAeFJpIrnfaxWhEVi18:22
openpowerbot[irc] <sadoon[m]1> If I have any more email issues it's basically out of my control, I set up everything I could but some things are managed by my ISP unfortunately..18:23
openpowerbot[irc] <sadoon[m]1> In which case I'll just switch to my protonmail account but I hope it won't be necessary18:23
openpowerbot[irc] <programmerjake> since irclog or the matrix bridge reordered the parts of the ssh key, here it is again:18:30
openpowerbot[irc] <programmerjake> echo -n ssh-rsa\ AAAAB3NzaC1yc2EAAAADAQABAAABgQCgf+Tsrcigx4 # idx=018:30
openpowerbot[irc] <programmerjake> echo -n wJWpoXyYh2nRYKrAVJVA4nozaqdvR7UHYuKgdopY1w8P1UE3GS # idx=5018:30
openpowerbot[irc] <programmerjake> echo -n sLAn6byzPLy8zlUEDmQlTDwSJIJcIINyPYxPNdniKHbUbisiT2 # idx=10018:30
openpowerbot[irc] <programmerjake> echo -n FyObx+SX83EayDFIxeOcsiy9hVy27SxopBYPZ/fFGFX99iKSRy # idx=15018:30
openpowerbot[irc] <programmerjake> echo -n wi6HSUwPvOvcfD43k1BVLhNtT56DRc6fmrpoc32as5R2Y4uHko # idx=20018:30
openpowerbot[irc] <programmerjake> echo -n Hj6vSNFaeRGwjFO+i0QjLbrvl8LDw72NHOLz0db5kXvVsX2B0v # idx=25018:31
openpowerbot[irc] <programmerjake> echo -n Jw4miD4ccotFI6Z/te8P+fRp3KqgVBrgxibg9rVLrsSx/FN8aZ # idx=30018:31
openpowerbot[irc] <programmerjake> echo -n Sc/L4bgOrnro80UQKwetwU5uNyWiquQ4wmSAxL+Ekctp5m0jJB # idx=35018:31
openpowerbot[irc] <programmerjake> echo -n QfAeFJpIrnfaxWhEViZ4zp5ZvavOLj3osoX/7EMdxyyaCjjsEe # idx=40018:31
openpowerbot[irc] <programmerjake> echo -n ga8jZdg6nWX4XwDaNwURu1zIovl7UCAR2vCEois6SOqVAejpXZ # idx=45018:31
openpowerbot[irc] <programmerjake> echo -n IJv2l5P+/LCtgeVtSvFc3l2qZIoHzSQpQOatoTlTX6WOgdBcjS # idx=50018:31
openpowerbot[irc] <programmerjake> echo -n $'E= sadoon@SwiftUnity\n' # idx=55018:31
openpowerbot[irc] <programmerjake> sort by idx18:31
openpowerbot[irc] <sadoon[m]1> Oh boy, let me just upload this to my site and share the link18:32
openpowerbot[irc] <sadoon[m]1> Just in case18:32
openpowerbot[irc] <sadoon[m]1>
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC18:34
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc18:35
openpowerbot[irc] <sadoon[m]1> Debian sffs guide is complete, scripts coming soon possibly by end of next week
octaviusBrilliant! Thanks for updating us Sadoon. I like the draft wiki page, didn't realise you actually need a POWER system (but now makes sense)19:00
openpowerbot[irc] <sadoon[m]1> I mostly reused stuff from the microwatt guide because it was quite good :)19:01
openpowerbot[irc] <sadoon[m]1> You could cross compile and then run in qemu user mode as a chroot but it's very tedious and can break things19:02
openpowerbot[irc] <sadoon[m]1> Which is why Debian's policy is to not to do that ever19:02
openpowerbot[irc] <sadoon[m]1> But you don't need a GUI at all so using the POWER server you guys have is going to absolutely be enough if you want to try it19:02
openpowerbot[irc] <sadoon[m]1> Gentoo SFFS wiki page is also ready
openpowerbot[irc] <sadoon[m]1> It's so much easier than debian it makes me sad19:18
octavius" POWER server you guys have" ah yes, I keep forgetting about it19:32
octaviusWould be nice to go through that flow19:32
octaviusI ran Gentoo a few times, and I'm still fond of the idea of having all the source code on your system19:33
octaviusAnd customise the build flags to your exact setup19:33
octaviusBut for interoperability, not so much19:33
openpowerbot[irc] <sadoon[m]1> I plan to keep the source code as well as the binaries on the tarball just to solve that problem19:40
openpowerbot[irc] <sadoon[m]1> Meaning the software will never change until we decide19:40
openpowerbot[irc] <sadoon[m]1> Software versions*19:41
*** gnucode <gnucode!~gnucode@user/jab> has joined #libre-soc20:04
*** gnucode <gnucode!~gnucode@user/jab> has quit IRC21:04
octaviusAh nice21:37
*** octavius <octavius!> has quit IRC21:40
*** gnucode <gnucode!~gnucode@user/jab> has joined #libre-soc22:24
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC22:48
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc22:58

Generated by 2.17.1 by Marius Gedminas - find it at!