rwp | There was success in the last two minutes? | 00:00 |
---|---|---|
joerg | ctrl+alt+F1. Still works for me when my X11 freezes and nothing seems to work anymore | 00:02 |
joerg | double ctrl+alt+BSp also works and kicks me to the login screen | 00:03 |
joerg | the culprit probably is my Graca | 00:04 |
rwp | I am suspecting things are locking up hard. In which case even the Magic SysRq keys would not work. https://en.wikipedia.org/wiki/Magic_SysRq_key | 00:05 |
joerg | sometimes there's a usually not noticed test: when you got an optical drive, insert a CD or DVD and check if it tries reading it | 00:15 |
joerg | also numlock/capslock LEDs in kbd are driven by a low level CPU based process | 00:16 |
rustyaxe | joerg: most kbd are usb nowadays tho | 00:29 |
joerg | rustyaxe: still the LEDs are not autonomously switched on/off locally inside kbd, it's a command sent from PC via USB to kbd afaik | 00:30 |
joerg | proof: connect the USB to a simple USB wallwart PSU and check if you coult toggle the LED state | 00:31 |
joerg | I think it won't work | 00:32 |
joerg | could* | 00:32 |
joerg | so when numlock LED still toggles when you hit the numlock key, odds are CPU is still running | 00:33 |
systemdlete | I've been having intermittent crashes of a chimaera system, but other people are reporting similar on other platforms also. The syslog shows a line of nulls (@^) right before the log restarts (after I reboot it of course) | 08:28 |
systemdlete | one explanation is that it is because some file system was not unmounted cleanly. But I really don't have any fs mounted other than the local ones, and I don't unmount those during uptime. | 08:29 |
systemdlete | At first, I thought maybe it might be a problem with rsyslog, but I don't see any local config changes or additions to the default configuration of rsyslog. | 08:30 |
systemdlete | I also note that there was a long pause in the syslog for a few hours before the crash. | 08:31 |
systemdlete | This type of crash (with the NULLS in the syslog) tends to happen for a while, then it goes away. I am not sure what I am doing differently when it starts doing this, or when it stops either. | 08:33 |
systemdlete | any thoughts on this would be appreciated. | 08:33 |
systemdlete | If it were an occasional freak thing, it would not be a problem, but like I said, this tends to happen in batches of crashes. | 08:35 |
systemdlete | I checked disk space. | 08:35 |
gnarface | systemdlete: how old is the power supply in the system? sounds like the type of thing that happens when a power supply or a PATA device is failing. | 09:23 |
gnarface | though i wouldn't rule out the usual suspects (NVidia drivers) | 09:24 |
gnarface | maybe over heating | 09:24 |
gnarface | does it usually happen when the system is idle, or under load, or no pattern? | 09:24 |
ballsxd | hello everyone | 10:07 |
ballsxd | im back with the same issue of my computer freezing and all processes stopping but hardware keeps working | 10:08 |
ballsxd | i can now confirm it isnt a memory issue | 10:08 |
systemdlete | gnarface, I don't remember exactly how old the PS is. | 10:09 |
systemdlete | and it has a PCI video card and do believe it is an NVidia | 10:10 |
systemdlete | The machine is designated as a test box. Typically it is under almost no load. 32 GB and running 1 or 2 VMs. | 10:11 |
systemdlete | Mostly sits idle. | 10:11 |
systemdlete | The PS is probably about 2 years old at this point. | 10:12 |
systemdlete | I have purchased a slew of diagnostic hardware, but I am not sure if I have the means to test a PSU | 10:12 |
systemdlete | I guess breadboarding the thing and plugging it into a power meter to see if it is drawing the right level of power. | 10:13 |
systemdlete | something like that? | 10:13 |
systemdlete | I think they do make a power tester that you can plug into on the PC side, rather than the wall power (or in my case, UPS power) | 10:14 |
systemdlete | Maybe 6 months ago or so, I ran memtest on the box for about a day just to see if there was bad memory, but I got no errors. | 10:15 |
systemdlete | One thing that could be an issue, though it has not been up to now: The MB was a used board I bought on Newegg (no warranty). It's a M5A78L | 10:16 |
systemdlete | Iirc, some of the PCI slots were suspect, but I may be thinking of another, nearly identical board. | 10:17 |
systemdlete | oh just duh! I have a digital power meter right here, with the connector for the board side. | 10:18 |
systemdlete | I forgot I bought that. | 10:18 |
systemdlete | gnarface: thanks. I guess I'll take that machine down and run a battery of hw tests, including checking MB slots (yeah, I have testers for those too) and the like. | 10:19 |
systemdlete | I suppose it couldn't hurt to test the memory again also. | 10:19 |
systemdlete | gnarface, there is one other thing, and I think you hinted at it. I noticed that every time I switched back to the testbox through my KVM switch, the desktop Display configurator would pop up. I would close it, then it would be open again when I returned to that machine. | 10:22 |
systemdlete | So that could be related to some nvidia or other video issue, as you suggested it might be. | 10:22 |
systemdlete | btw, it is a VM that is crashing. Sorry I should have mentioned that up front. VMs are so ubiquitous these days I forget they are a bit different than their hardware hosts. | 10:23 |
systemdlete | virtualbox, not qemu. | 10:25 |
systemdlete | But I am not sure whether that display config window was popping up from the crashing VM or the host. | 10:26 |
gnarface | systemdlete: when you memtest it again, make sure to also do the optional bit-fade test this time, and make sure all tests complete | 11:12 |
gnarface | if it were the host system crashing, it would also sound like something that could be caused by a failing motherboard bus or slot/port | 11:13 |
gnarface | if it's just a VM crashing, that seems less likely to me | 11:14 |
gnarface | also, if it was bad ram i'd expect some occasional segfaulting, rather than this one specific, repeated behavior | 11:14 |
gnarface | so i'd start to suspect the VM software itself then | 11:14 |
gnarface | but i'd also want to make sure the motherboard firmware was fully updated | 11:15 |
gnarface | i'd also strongly consider loading the appropriate cpu microcode package if i wasn't already | 11:16 |
gnarface | the power supply doesn't sound so very old, so unless it was super cheap i'm not sure it'd be my first suspect anymore | 11:17 |
gnarface | also i thought you were talking about the host system, and it would seem to follow that a bad power supply wouldn't be likely to affect just VMs this way without doing anything to the host, but who knows | 11:17 |
gnarface | if these vms are not headless though the nvidia stuff is still slightly plausible | 11:18 |
rustyaxe | Working on packaging up an out-of-tree driver for devuan, any documentation for this i should look at? Trying to get the radioberry driver into a state where us devuan users can enjoy it (right now it's finicky to build outsode of raspios and only works with 32bit kernel) | 11:37 |
brocashelm | rustyaxe: probably this: https://git.devuan.org/devuan/documentation/src/branch/master/maintainers/PackagingGuide.md | 11:45 |
brocashelm | and also, lurk in #devuan-dev | 11:46 |
rustyaxe | cool, thanks. There's a few bits here needing packaged up -- mostly just trying to make soemone else's project work on my machine. making some progress anyways.. got it wrapped up in dkms and installing on the 32bit raspios as a dkms module; now just gotta fix the handful of bugs that break it on 64bit and go to sleep :o | 11:50 |
adam_free2air | rustyaxe: there's also #devuan-arm ... :) | 18:37 |
adam_free2air | rustyaxe: also please consider hosting your work on the https://git.devuan.org forge | 18:40 |
rustyaxe | adam_free2air: Right now, ive forked it from upstream github repo, so i can send a pull request shortl | 18:40 |
rustyaxe | Ideally he'll upstream the changes so there'll only be one repo out there to avoid confusion of users | 18:41 |
adam_free2air | rustyaxe: OK - understood. | 18:41 |
rustyaxe | ham radio projects get ... interesting. This seems at least still being maintained, just not many developers for the software. I figured their RF hat will save me many hours of prototyping, so i'll spend a few hours working on things they've not gotten to yet | 18:42 |
adam_free2air | rustyaxe: wow. you're doing some very interesting work! | 18:43 |
rustyaxe | mostly just trying to make my new toy usable on devuan heh | 18:52 |
rustyaxe | i really dont care for raspios ;) | 18:53 |
ml303 | Hello | 18:59 |
ml303 | I'm having some trouble figuring out why my backlight keys no longer work. I use sxhkd to control the brightness level, when I hit the keys, nothing seems to happen. It's not my config that's off, I've checked with xev. I'm guessing that some daemon is at work here! Any tips on how to proceed, | 19:05 |
gnarface | ml303: did you install packages or run an update or something right before it broke? which release are you using? | 19:32 |
ml303 | I just install it around 5days back. So everything is up to date. I'm using daedalus. | 19:43 |
gnarface | is sxhkd still running? | 19:44 |
gnarface | there's probably a kernel module involved too | 19:44 |
gnarface | i can't imagine why it'd stop working without changing one of those two things | 19:44 |
gnarface | unless maybe there's a keyboard hotkey enable/disable you hit accidentally? | 19:45 |
gnarface | you sure you didn't upgrade anything before it broke? it was working after the install? | 19:47 |
gnarface | you should see if sxhkd has any logs | 19:47 |
ml303 | Is sxhkd is up and running, it's really strange. It doesn't work in both xfce and bspwm! | 19:47 |
gnarface | also check the output of dmesg | 19:47 |
gnarface | did you change window managers before it broke? | 19:47 |
ml303 | This is my first setup, I didn't work from the very beginning. In fact I even had trouble using xbacklight, had to set the boot param acpi_backlight=native, might that have something to do with it? | 19:49 |
gnarface | yea maybe | 19:50 |
gnarface | sorry, i'm not familiar with keyboard lighting so much | 19:51 |
gnarface | your boot params are probably in /etc/default/grub | 19:51 |
r00t_ | hi | 19:55 |
r00t_ | anyine here? | 19:55 |
r00t_ | Can you help our opensource project. We want the unreal map file types. | 19:55 |
r00t_ | http://sf.net/p/chaosesqueanthology/tickets/2/ | 19:55 |
r00t_ | .t3d and .unr file formats (C project, GPL licensed) | 19:55 |
r00t_ | 19:55 | |
r00t_ | that is all | 19:55 |
r00t_ | I just want opensource contributors | 19:55 |
r00t_ | please check it out | 19:55 |
r00t_ | http://sf.net/p/chaosesqueanthology/tickets/2/ | 19:56 |
ml303 | The keys are XF86MonBrightnessUp and XF86MonBrightnessDown that aren't working. I might have left out a needed xorg package maybe?, I installed xserver-xorg, xserver-xorg-video-intel and xinit. | 19:57 |
gnarface | yea that's possible too | 19:58 |
gnarface | do you have xserver-xorg-input-libinput? | 19:58 |
ml303 | Yes, checked, it's installed. | 20:00 |
gnarface | xserver-xorg-input-evdev? | 20:01 |
ml303 | Nope, lemme install it real quick | 20:02 |
gnarface | you might also have to load extra kernel modules, like uinput or uhid | 20:03 |
gnarface | or something like that | 20:03 |
ml303 | Ok, Installed and tested but didn't solve it. | 20:07 |
ml303 | I'm going to try a reboot without the acpi_backlight=native option, do some more tests | 20:09 |
friedhelm | Hi, | 20:11 |
friedhelm | had trouble with the backlight on my lenovo t410 laptop a while back. | 20:12 |
friedhelm | I had to put "acpi_backlight=bios". | 20:13 |
friedhelm | That solved it. | 20:13 |
friedhelm | Maybe that helpes. | 20:13 |
ballsxd | hey guys, im trying to install jellyfin by adding its repository from debian. is this supposed to work? because it doesnt | 22:35 |
ballsxd | i did manage to install grafana this way, so im curious | 22:36 |
fsmithred | I don't know what jellyfin is, but generally anything that works on debian works on devuan. | 22:38 |
fsmithred | we only fork a few package. Otherwise, everything is the same as debian. | 22:38 |
ballsxd | alright, so debian repositories aswell as .deb packages should work? | 22:40 |
brocashelm | correct | 22:41 |
brocashelm | but use official devuan repos only if you want to be on the safe side | 22:41 |
brocashelm | using an official debian repository might screw with your dependencies and pull in systemd | 22:41 |
brocashelm | but something like winehq's debian repository should be safe, because it's just wine you're installing | 22:42 |
brocashelm | avoid the more complicated repos like dmo/debian multimedia org (because of different dependencies) | 22:42 |
fsmithred | what error did you get? | 22:42 |
brocashelm | devuan's repos should have you covered 99.9% of the time, anyway | 22:42 |
ballsxd | interesting,so it would be "safer" to use .deb's? | 22:43 |
ballsxd | fsmithred: there was no error, the required packages just didnt show up in apt search | 22:43 |
brocashelm | i am not sure. it really depends on the package and what dependencies they are | 22:43 |
brocashelm | i have installed local deb files and everything is humming along just fine | 22:44 |
brocashelm | just avoid packages that have abnormal dependencies compared to the ones from devuan repos; that could easily break your system if not careful | 22:44 |
ballsxd | alright, ill try to work with that | 22:44 |
fsmithred | if you were using debs, then the missing packages might be in the jellyfin repo. | 22:45 |
fsmithred | make sure you're comfortable with rescuing a screwed up system before you do that. | 22:46 |
fsmithred | and use "--simulate" with apt before you do anything for real | 22:46 |
ballsxd | sure thing, ill make a backup right now | 22:49 |
brocashelm | i use timeshift for easy system backups. it's a great tool you can even use straight from cli | 22:50 |
brocashelm | sudo timeshift --create | 22:50 |
brocashelm | then you can run --restore if there was a problem | 22:50 |
schillingklaus | i am accustomed to using broken debian already 25 years ago | 22:51 |
brocashelm | frankendebian? | 22:52 |
brocashelm | that's the magic of it: even when you break it, it still works | 22:52 |
ballsxd | eh i have only 70 gigs left on my ssd and no spare usb stick :p | 22:53 |
schillingklaus | frankendebian is what you get when mixing repos wildly. using dpkg --force installs goes even beyond that | 22:53 |
ballsxd | so i guess ill have to wait with running dangerous experiments | 22:53 |
brocashelm | installing from experimental is the real hardcore debian | 22:54 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!