| lfh | I'm trying to get the live beowulf media running, but it hangs after grub on nothing specific. I am trying run from a usb that is created by simply dd'ing the iso unto it. USB is good, it works with BSD distros no problem. However I haven't been able to make it work with new linux distros either (I had hoped to find a different to help debug this so I could get devuan). Is there anything I can do | 00:40 |
|---|---|---|
| lfh | to get more verbose error messages? I already tried acpi=off and nomodeset, no dice. | 00:40 |
| lfh | The last line before it hangs is "(uninitialized): registered PHC clock" for my network card, in failsafe mode. Seemingly nothing erroneous, but it scross too fast to get a good look. Keyboard is powered but unresponsive at that point. | 00:50 |
| lfh | In non-failsafe mode (whether or not loading into ram), the last line is instead some stuff about nouveau (nothing seemingly erroneous), then [drm] Driver supports precise vblank timestamp query. and then it hangs. | 00:52 |
| fsmithred | lfh, is this a uefi system? | 00:57 |
| fsmithred | oh well, I have to go. | 01:03 |
| mason | Argh. I need a newer libsane, but to backport it I need a newer debhelper-compat. Sigh. | 01:08 |
| tsadok | Been there. In fact, at one point I traced my "I need newer libraries" tree clean down to glibc, at which point I went "nope" and stuck with the version I had until upgrade. | 01:10 |
| tsadok | Actually, that specific occasion might not have been on Debian/Devuan. That might've been on FreeBSD 5, which I used for a bit before sarge came out. Because woody was too ancient even for me. | 01:11 |
| mason | tsadok: I think what I might do is just compile from source. Not a huge tree... It's pretty much just going to be xsane. | 01:11 |
| tsadok | Oh, yes, if it's only ABI compat that's broken, compiling from source is the easy way. | 01:12 |
| mason | stow for the win | 01:12 |
| tsadok | You don't start tracing down the need-newer-lib-version dependencies unless the API has changed. | 01:12 |
| tsadok | I think the occasion I'm thinking of, involved wanting to upgrade to the dev version of Inkscape, which used a dev version of Gtk, which used a dev version of glibc, or something like that. | 01:13 |
| tsadok | gtk2 vs 3, that sort of thing. | 01:14 |
| tsadok | Or it might've involved SDL; I'm not longer certain, it's been a few years. | 01:14 |
| mason | Wow, this is aggressive: "fatal: not a git repository" | 01:15 |
| lfh | fsmithred: yes, it is a uefi system. | 01:15 |
| mason | Ah, that was just the orig that Debian ships. Never mind. Anyway, I'll take this to -offtopic if I have further noise to emit. | 01:16 |
| tsadok | Hmm, UEFI causes all sorts of boot-related weirdness, but you said that a FreeBSD live system boots, so presumably it's not the "We Have Locked Down The Things And Thou Shalt Not Leave Microsoft" issue. | 01:16 |
| mason | Heh, randomly, I have a couple boxes here Devuan boots but FreeBSD won't boot. | 01:18 |
| tsadok | Could be something related to device ID issues, but I am kind of out of my depth there. | 01:18 |
| lfh | mason: yeah, that's the more usual scenario. | 01:18 |
| tsadok | I have _tried_ to avoid UEFI, but that's getting hard to do these days. | 01:18 |
| tsadok | If Devuan boots but FreeBSD doesn't, you can use LILO or grub to chainload BSD. | 01:19 |
| tsadok | Not sure about doing the reverse. | 01:19 |
| lfh | The problem is in the initramfs or similar, so grub won't help here (I get to grub just fine) | 01:20 |
| mason | tsadok: The bootloader comes up, but kernel load fails. | 01:20 |
| tsadok | Ah. Hmm. | 01:20 |
| tsadok | Are you using a separate /boot partition, or is it just part of the root fs? | 01:21 |
| mason | tsadok: Installer, in my case. I don't get to where I've set up a system. | 01:21 |
| lfh | The live media itself is what fails | 01:21 |
| mason | Nice, very parallel. | 01:21 |
| lfh | heh | 01:22 |
| tsadok | Ah. | 01:22 |
| tsadok | I see. | 01:22 |
| tsadok | Weird trick that might work: take the HD out, plug it into different hardware, install Devuan, then put the HD back in the hardware you want to use. | 01:22 |
| tsadok | You wouldn't *expect* this to work (unless you're an old BeOS user maybe), but these days it often does. | 01:23 |
| mason | I'd thought about doing just that in my case, but I just ran with Devuan instead. | 01:25 |
| lfh | yeah, I might not have a choice but to try. | 01:25 |
| jonadab | It triggers my "this is icky" radar, but it has _worked_ the last two or three times I've had to try it. | 01:25 |
| mason | Hooray, scanning achieved. | 02:05 |
| mason | I gave up on backporting the Debian package and just built it all from source. | 02:05 |
| mason | bbiab | 02:05 |
| lfh | just for a sanity check before getting the hdd out and operating on it, I checked the devuan boot medium on a second computer and it worked just fine. | 02:19 |
| lfh | definitely hardware related, then | 02:20 |
| rwp | I actually expect that installing onto a hard drive on a system will boot on another system. | 03:28 |
| rwp | As long as the initramfs MODULES variable is set to "most". If set to "dep" then maybe not if the system is too different. | 03:28 |
| rwp | lfh, If the problem is the live boot hang then you might try downloading different live boot media and seeing if they work or hang. Something like KNOPPIX. | 03:30 |
| rwp | And there is always a manual debootstrap install if other things fail but doing it manually means you have to drive through it with some skill. | 03:30 |
| gnarface | it's common to get a false positive on this test though; the graphical live images usually have several seconds of seeming stalled on a black screen right at the end of Xorg starting up | 03:36 |
| gnarface | it is really just SD read delay usually | 03:36 |
| gnarface | SD/USB | 03:36 |
| gnarface | (if you are using an optical drive or a USB key with a actual i/o access light on it though you should still see it blinking during that time) | 03:37 |
| gnarface | (it isn't really stalled unless nothing happens for several seconds and the traffic light isn't blinking either) | 03:37 |
| kreyren | How do i install `snapd` on devuan? | 08:42 |
| kreyren | trying to apti snapd but it can't find it | 08:42 |
| * kreyren < https://matrix.org/_matrix/media/r0/download/matrix.org/pnztjimVxAiCOeVblYeBjQlS/message.txt > | 08:42 | |
| kreyren | ah it's banned in devuan? | 08:45 |
| kreyren | x.x | 08:45 |
| kreyren | http://packages.devuan.org/bannedpackages.txt | 08:45 |
| Unit193 | It depends on systemd, of course it's banned. | 08:48 |
| gnarface | is there anything on there you really need that there's no existing alternatives for in the repo? | 08:57 |
| jason234 | I just made a devuan live ascii of 248 mb ;) base syystem. | 11:45 |
| furrymcgee | public? | 12:00 |
| ShorTie | ascii ?? | 12:08 |
| furrymcgee | #tinycorelinux is 16MB | 12:39 |
| gour | hello, i've installed 'unattended-upgrades' to automatically take care about system updates, but it does not seem to work - there are updates available, the service is not running, starting it manually does not show um, similar with 'status'...any clue what is required to make it working on devuan? (i'm on chimaera) | 14:25 |
| ShorTie | other then not a good idea to use, imho | 14:29 |
| luna_is_here | gour: AFAIK there is no daemon for unattended-upgrades. It is triggered by cron. Furthermore, by default only security updates are installed. | 14:42 |
| luna_is_here | At least it used to be that way a while ago. | 14:43 |
| gour | luna_is_here: thanks for the info | 15:43 |
| Tetralet | Hi, Is there any plan to put memtest86+ in install CD? | 15:53 |
| rwp | My biggest problem with unattended-upgrades is that it is mostly undocumented. But look through "dpkg -L unattended-upgrades | grep /etc/" and browse the configuration files for clues to how it works. | 15:53 |
| lfh | I got the disk out of the machine but it seems weirdly corrupted: I get 3 partitions, one is my encrypted root (all good there), and the other 2 I have no idea. One of them, when mounted, is an empty ext4 partition (only contains "lost+found"). The second I can't read at all, it says 'not a block device'... neither of them (nor the root under the crypt'd lvm) have an actual /boot (all empty). | 16:46 |
| lfh | What's going on? | 16:46 |
| lfh | I got the disk out of the machine but it seems weirdly corrupted: I get 3 partitions, one is my encrypted root (all good there), and the other 2 I have no idea. One of them, when mounted, is an empty ext4 partition (only contains "lost+found"). The second I can't read at all, it says 'not a block device'... neither of them (nor the root under the crypt'd lvm) have an actual /boot (all empty). | 16:46 |
| lfh | What's going on? | 16:46 |
| lfh | oops, sorry for the double post | 16:47 |
| lfh | also no efi partition despite this being an efi system... | 16:47 |
| fsmithred | lfh how did you install the system? live-iso or installer-iso, iinstalled on thie efi machine or installed on another and moved? Manual or automatic partitioning? | 17:26 |
| lfh | fsmithred: I'm actually trying to boot devuan from the live-iso on that machine, and trying to figure out why it hangs at boot. This is not devuan-specific: I also tried ubuntu, opensuse, arch and gentoo versions from 2016 up to yesterday and get the same result (also means it's not init system specific).The current system was an ubuntu install that got killed following an update due to severe | 17:31 |
| lfh | regressions in their initramfs-update tools. | 17:31 |
| fsmithred | can you comapre a working initramfs with a non-working one? | 17:32 |
| fsmithred | at the moment I can't remember the simple command to unpack initramfs. | 17:32 |
| fsmithred | unmkinitramfs | 17:33 |
| lfh | fsmithred: as I said, I see nothing at all in the putative boot partition, one of their tools may have wiped it altogether, or there's an additional hardware issue involved. In any case, what seems to be the problem is simply that the lvm2 and cryptsetup modules aren't correctly added to the initramfs. | 18:00 |
| lfh | But that is that and this is this: I'd like to figure out why live isos don't go past the early loading beyond grub (just hanging with no specific error or warning), so that I can install devuan on that device. | 18:01 |
| lfh | I managed to boot from the live medium. It was easy really, all I had to do was boot up with the hard drive disconnected..... it appears that an unknown bug in that distro is doing some really bad things to hard drives. Since it's probably still impossible to boot up with the drive inside, is there a good procedure to install devuan onto the drive and then plug it in? Can I just copy the relevant | 23:34 |
| lfh | iso contents into the root, chroot to complete setup and then go? | 23:34 |
| crashoverride | u w0t m8? | 23:42 |
| crashoverride | by "that distro" do you mean Devuan? | 23:43 |
| Xenguy | Sounds ominous | 23:47 |
| crashoverride | yep. | 23:47 |
| furrymcgee | maybe you want debootstrap and install grub | 23:48 |
| crashoverride | and that feeling is supported by the potentially condescending "all I had to do was boot up with the hard drive disconnected" | 23:48 |
| crashoverride | AFAICT every time I wasn't able to boot something, it wasn't due to the medium I meant to boot, but the crummy firmware (most of the time UEFI) I had to use to boot it. | 23:49 |
| crashoverride | so going to an OS's channel and starting what looks like a judgemental condescending complaint isn't the best way to go about "I was unable to correctly configure my motherboard's firmware, can you help please" | 23:50 |
| Xenguy | I try to keep my condescension classy | 23:52 |
| Xenguy | But then life's too short to judge anyway | 23:53 |
| Xenguy | I'm all for a welcoming attitude, like, c'mon in and have a pint | 23:53 |
| rwp | I would be inclined to run all of the SMART tools on this hard drive that is giving problems. Probably the drive is failing. Which is why otherwise normal things are failing. | 23:53 |
| crashoverride | yeah I mean, the drive could be fine, too. | 23:53 |
| crashoverride | a LOT of UEFI (especially early ones) implementations are REALLY bad. | 23:54 |
| crashoverride | and prevent you from booting *anything*. | 23:54 |
| crashoverride | (but windows) | 23:54 |
| rwp | Agreed. Early EFI boot often had problems. But maybe the BIOS has a setting to turn it off? Most of the early ones with the problems did. | 23:54 |
| rwp | It's often obscured as an option to boot Windows XP versus Windows Vista or something else similarly confusing. | 23:55 |
| crashoverride | and to be fair, even when you can boot the UEFI firmware, the fact that Devuan does not have the EFI/boot/bootx64.efi file also is causing problems on some implementations. | 23:55 |
| Xenguy | Since I'm a simple man who dislikes anything requiring real prowess, I avoid all things UEFI, and successfully so far | 23:56 |
| crashoverride | Xenguy: how is it running a 2 threads CPU clocked at 2 GHz that uses 150W of electricity? :D | 23:56 |
| * crashoverride hides | 23:56 | |
| crashoverride | "hey come on, at least I got multithreading! | 23:56 |
| rwp | But I would boot different boot media and then run "smartctl -t short /dev/sdX" on it (X being the drive letter assigned to it) and then "smartctl -l selftest /dev/sdX" to check it after it finishes. And then the same with -t long if that reports all okay. | 23:57 |
| Xenguy | crashoverride, I suppose I'd say I'm hardware-allergic mostly | 23:57 |
| crashoverride | rwp: I'm expecting us to continue driving the disucssion and lfh to /quit without adding another word. | 23:57 |
| Xenguy | I never learned about cars and engines either | 23:57 |
| crashoverride | hardware allergic? | 23:57 |
| crashoverride | you mean? | 23:57 |
| Xenguy | I really enjoy software, and all the cool utilities and such | 23:58 |
| crashoverride | you touch a mobo, you get spots? | 23:58 |
| crashoverride | also should we move this discussion to #devuan-offtopic ? | 23:58 |
| Xenguy | Just no interest in hardware, and therefore little knowledge of it | 23:58 |
| * ShorTie snickers | 23:58 | |
| Xenguy | Oh sorry, you're right | 23:58 |
| rwp | Maybe not allergic but just "hardware hesitant"? (Referring to "vaccine hesitancy" which is in the news now...) | 23:59 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!