adam_free2air | hey all. trying to build devuan/ceres with rpi-img-builder and getting error making rootfs with duplicate base-files package https://pastebin.com/uHeYs1hn any tips? | 18:21 |
---|---|---|
adam_free2air | (using main branch on a AMD64/x86_64 system running on Ubuntu Focal) | 18:22 |
Tenkawa | just a sec let me take a look | 18:24 |
Tenkawa | well thats a new one... | 18:25 |
Tenkawa | let me see if I can recreate it with a debian host | 18:26 |
Tenkawa | (dont have ubuntu one handy atm) | 18:27 |
Tenkawa | pi 4? | 18:30 |
Tenkawa | (image) | 18:30 |
adam_free2air | yup | 18:30 |
Tenkawa | ok.. trying a test now | 18:31 |
Tenkawa | wont take very long on this machinr | 18:31 |
Tenkawa | er machine | 18:31 |
adam_free2air | it's a newly prepped ubuntu host - i've been building on debian host so far using the debian64 branch. | 18:32 |
Tenkawa | will tell me if it is a ubuntu or general os problem though | 18:32 |
adam_free2air | but debian & gcc-10 prompted me to try diff recommended ubuntu build os. | 18:33 |
adam_free2air | trying beowulf now on same host. | 18:33 |
Tenkawa | this will tell me a few things though.. if I need to I can build a ubuntu vm fairly quick to debug the builder | 18:34 |
adam_free2air | btw, command issued for ceres that produced that output was 'make all' | 18:35 |
Tenkawa | you didnt run the builder menu? | 18:36 |
Tenkawa | you run make config then make menu to build an image | 18:37 |
adam_free2air | make config then make all | 18:39 |
Tenkawa | make all can be "iffy" | 18:40 |
Tenkawa | brb.. need to go close garage door | 18:40 |
adam_free2air | as preflight, installed packages as per README, and did a make ccompile as a sanity check .. | 18:41 |
adam_free2air | hmm ... "iffy" :D ok. will let beowulf 'make all' iffy build run to end then try make menu instead ... | 18:43 |
Tenkawa | ok back.. | 18:47 |
Tenkawa | make all can make assumptions | 18:48 |
Tenkawa | and skip targets if conditions are not exactly right | 18:48 |
Tenkawa | (depending on how well Makefiles are written) | 18:49 |
Tenkawa | adam_free2air: can you find the debootstrap.log for that build? | 18:52 |
Tenkawa | thats where the real error is probably going to be…. from what I'm reading this is a masked error | 18:53 |
adam_free2air | yup. here https://pastebin.com/1K6kF8PF | 18:56 |
adam_free2air | actually, beowulf just did the same thing. | 18:57 |
Tenkawa | just a sec | 19:00 |
Tenkawa | yeah its colliding pkgs... | 19:02 |
Tenkawa | now the q is why | 19:02 |
adam_free2air | so perhaps my setup or something specific to devuan (been building off branches for so long) ... | 19:02 |
adam_free2air | ok. you too. | 19:03 |
Tenkawa | thats why I want to see what happens when I build an image | 19:03 |
Tenkawa | mine is building right now | 19:04 |
adam_free2air | ah, misunderstood. thought yours might have finished. | 19:04 |
Tenkawa | no… I didnt increase the kernel jobs enough so its running slower than it "could" | 19:05 |
adam_free2air | until yesterday, hacky bash scripts with jenkins have been building nightly with no problem off the debian7 & devuan64 branches using 'make all' | 19:06 |
Tenkawa | hm | 19:06 |
Tenkawa | strange | 19:07 |
adam_free2air | but this was also using a (slightly dated) fork for controlled builds. | 19:07 |
Tenkawa | last night was when mine started working again heheh | 19:07 |
adam_free2air | heh. | 19:08 |
Tenkawa | (5.13 were finally updated in foundations repo) | 19:08 |
adam_free2air | ah, do you know when? | 19:08 |
Tenkawa | I had been waiting for them to add 5.13 | 19:08 |
Tenkawa | its there now | 19:09 |
Tenkawa | finally | 19:09 |
Tenkawa | Linux cm4 5.13.1-cm4-v8 #1 SMP Wed Jul 7 14:52:42 EDT 2021 aarch64 GNU/Linux | 19:09 |
Tenkawa | (I manually added .1 from mainline "it only had 4 changes") | 19:10 |
adam_free2air | so what should VERSION be set to in userdata.txt if foundation is now 5.13 - should it be updated from 5.10.y? | 19:12 |
Tenkawa | for foundation its always version.y | 19:13 |
Tenkawa | so 5.13.y | 19:13 |
Tenkawa | etc | 19:13 |
adam_free2air | ok. thanks. | 19:13 |
Tenkawa | mainline would be 5.13 | 19:14 |
Tenkawa | specificly a version 5.13.x | 19:14 |
Tenkawa | 5.13.1 can be built mainline but its missing features | 19:15 |
Tenkawa | ok kernel is almost done… rootfs next | 19:16 |
Tenkawa | figures: | 19:30 |
Tenkawa | File has unexpected size (8631168 != 8631344). Mirror sync in progress? [IP: 2001:878:346::116 80] | 19:30 |
Tenkawa | ugggh | 19:30 |
* adam_free2air afk for dinner, back in an hour or so ... | 19:31 | |
Tenkawa | maybe the mirrors will be done by then | 19:31 |
adam_free2air | in make menu, which to select all or image? | 19:32 |
Tenkawa | go ahead and use all | 19:33 |
Tenkawa | I will use the individual steps here since I can add tracing into areas if it breaks on my side | 19:34 |
Tenkawa | need to wait for the mirrors to sync first here or apt will keep failing :( | 19:34 |
c0rnelius | adam_free2air: make cleanup | 20:17 |
c0rnelius | I don't currently have it set to clean its self after hitting an error when building the rootfs. | 20:18 |
c0rnelius | When I hit Devuan mirror errors I usually just reboot the host machine. It seems to be the only way I've found to fix the problem when it does the endless happening. | 20:23 |
Tenkawa | c0rnelius: he's not hitting the mirror error… I am | 20:25 |
Tenkawa | I hit it debugging his errror | 20:25 |
Tenkawa | error | 20:25 |
Tenkawa | I'm building 2 image scenarios to see if I can recreate what he got but yeah it looks like cleanup to me too | 20:26 |
c0rnelius | its cleanup | 20:26 |
Tenkawa | ok so we concur | 20:27 |
Tenkawa | did you see the news in the other channel? | 20:27 |
c0rnelius | I would need to write in a scenario and I've just been to lazy to do so. | 20:28 |
c0rnelius | there is also this - https://github.com/pyavitz/rootfs-tarballs/releases/tag/rootfs | 20:30 |
c0rnelius | you can just wget or aria2c which ever tarball ur looking for. | 20:30 |
c0rnelius | No matter what I do I can't get this damn N2+ to stop getting CPU errors when building kernels. Odd thing is, when I ran a few stress-ng tests it didn't fail. | 20:32 |
Tenkawa | c0rnelius: in a bit I have an idea then | 20:39 |
Tenkawa | I'll msg in the other chan when I get back | 20:39 |
Tenkawa | you running debian or ubuntu on there> | 20:44 |
Tenkawa | er ? | 20:44 |
c0rnelius | Ubuntu currently | 20:44 |
c0rnelius | I was considering trying Debian, but I can't really see why that would matter much. | 20:45 |
c0rnelius | Its all the same garbage really. | 20:45 |
Tenkawa | no its not | 20:45 |
c0rnelius | Pretty much. None of the other boards care. | 20:46 |
Tenkawa | its majorly diff the way the libraries call hardware | 20:46 |
Tenkawa | you should still try it | 20:46 |
Tenkawa | ubuntu has "always" been more hardware flaky here | 20:47 |
c0rnelius | In all fairness I hate Ubuntu I only use it for the GCC versions available for building. | 20:47 |
c0rnelius | BUt yeah I'll make a Debian img and see what happens. | 20:47 |
Tenkawa | might as well right? | 20:48 |
c0rnelius | Only thing I haven't tried yet is vendor uboot. | 20:48 |
Tenkawa | oh | 20:48 |
c0rnelius | But.. Yeah. Seems like that would not be the prob if its works on ur end. | 20:48 |
Tenkawa | yeah but I'm not running it so.... | 20:48 |
Tenkawa | exactly | 20:49 |
c0rnelius | The kid leaves on Saturday so I'll have more time to play with it. | 20:50 |
c0rnelius | At the moment I just run some quick test in the morning whilst he is still sleeping. | 20:51 |
Tenkawa | heheh | 20:52 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!