Korgi | yeah, these are all the standard questions it asks during the "configure the network" portion of fhe setup. i dont think i was prompted for an ip adress when going through the initial setup though, that seems to be soecific to recovery mode | 00:00 |
---|---|---|
Korgi | specific | 00:00 |
fsmithred | you must have asked for that, whether you noticed or not | 00:01 |
fsmithred | I'm gonna try that | 00:02 |
fsmithred | I got auto-configuration | 00:09 |
Korgi | well, i managed to get past all of the network configuration windows, and am now on the "enter rescue mode" screen. When i attempt to select the device to use as the root file system though, it just says mount failed | 00:18 |
djph | bleh, I think I did something bad somewhere along the way ... | 00:20 |
djph | ... where does "/run/user/$uid" come from and why can't I nuke it from orbit? | 00:22 |
fsmithred | maybe from elogind | 00:22 |
djph | hmm ... | 00:22 |
fsmithred | Korgi, are you sure you selected the right partition? | 00:22 |
Korgi | yes, just to make sure i went down the full list and it gave the same error message on each | 00:23 |
djph | oh boy; I may've made a real mess here today; haaahahahaha | 00:28 |
djph | oh, no we're good | 00:31 |
sadsnork | I have seen someone mention a pastebin service in here previously but forget the URL. I think it was specifically (1) public (2) free (3) ad-free and (4) javascript-free. Is that correct? And if anyone knows the URL could you please share it with me? :-) | 01:17 |
fsmithred | paste.debian.net | 01:17 |
sadsnork | tyvm fsmithred! | 01:18 |
sadsnork | Boy, that should not have been hard for me to remember! /me blushes | 01:18 |
Korgi | I am still not making any headway with this. At least devuan is installed now so i guess that was the biggest step. gonna have to toggle this wifi thing later, phone is about to die on me. thanks to those who helped me make it this far | 01:20 |
systemdlete2 | I know anything can cause any kind of problems. But specifically, if I use a cheap fan connected to SYS_FAN1, and the fan itself does not have variable speed, what are the chances of damage to the board? | 02:09 |
systemdlete2 | This is a Gigabyte mainboard | 02:09 |
systemdlete2 | Thanks if you know. | 02:09 |
systemdlete2 | One description I read online is that SYS_FAN1 is controllable, but SYS_FAN2 and SYS_PWR are not. | 02:10 |
systemdlete2 | is that true for all mainboards or just certain ones? The manual for the board does not seem to be very specific about this. | 02:11 |
systemdlete2 | (I dread having to use their support line...) | 02:11 |
systemdlete2 | nvm.... it really IS here in the manual | 02:16 |
aloo_shu | my 2ct, if the pins for speed control are not connected on the cheap fan's connector, the only problem I can see is it will run full speed when it runs, means it's louder, and using slightly more energy - but I'm deducing this rather that having experience | 02:24 |
Tenkawa | well a cheap fan wont have a 3rd pin pwm likely anyway | 02:26 |
Tenkawa | and some pwm fans are fixed rpm fans so just on/off | 02:27 |
mason | sadsnork: There's also termbin, which has a particularly good trick associated: cat /proc/cmdline | nc termbin.com 9999 yields https://termbin.com/np1u for example | 02:27 |
systemdlete2 | This board supports 3 or 4 pins on 4 pin connectors, per the manual. The fan I am hooking up is a 4-pin. It is an ultra quiet fan (though I don't think it is really that quiet, even paying $35 for it. It registers 20dBM extra.) | 02:30 |
systemdlete2 | but I haven't yet figured out if it is variable speed. | 02:30 |
systemdlete2 | I didn't think to research that when I bought it. | 02:30 |
tuxd3v | systemdlete2, all fans that are 3/4 pins have speed variation( pwm controlled ) | 02:31 |
systemdlete2 | There are 2 pin ones also? lol, idk | 02:31 |
tuxd3v | the diference between 3 pin and 4 pins, is that 4 pin ones have a feedback line, were you can read the speed of the fan ;) | 02:31 |
systemdlete2 | So it sounds like I need not worry about this at all | 02:32 |
tuxd3v | you should be able to see what rpms it is spining :) | 02:32 |
systemdlete2 | I mean, I can't blow the mainboard by messing up fan connections. | 02:32 |
tuxd3v | 3/4 pin fans are compatible | 02:32 |
tuxd3v | you need to be very imaginative to do that :) | 02:33 |
systemdlete2 | tuxd3v: I come from back in the old days when someone like me was let's say just a bit intimidated by hardware. | 02:33 |
systemdlete2 | ESD was a major problem. This board, by comparison, has some ESD protection built into it. | 02:33 |
tuxd3v | the 2 Pin fan permits speed to change( pwm controlled ), but you don't have any feedback of speed, etc | 02:34 |
systemdlete2 | We used the log tables in the back of the math textbook, ok? | 02:34 |
systemdlete2 | THAT old. | 02:34 |
tuxd3v | it just output a pwn signal, and its a blind thing, but it works.. but better a 3/4 pin, with preference for a 4 pin one :) | 02:34 |
systemdlete2 | Well, this baby's a 4 pinner. | 02:35 |
systemdlete2 | I'm still cautious when dealing with mainboards, though, even today. No sense taking unnecessary risks. | 02:35 |
tuxd3v | then it would output its speed per minute( rpm ) | 02:35 |
systemdlete2 | great. Thank you tuxd3v | 02:36 |
tuxd3v | and if you have lm-sersors installed 'sensors' will show you the speed it is rotating too :) | 02:36 |
systemdlete2 | I have lm-sensors, yes. | 02:36 |
tuxd3v | nice :) | 02:36 |
systemdlete2 | (always) | 02:36 |
tuxd3v | :) | 02:36 |
systemdlete2 | Can the lm-sensors data be viewed from a remote machine, or only locally? I mean, is there support for it? | 02:37 |
systemdlete2 | I guess one of the networked monitoring tools out there could do it. | 02:38 |
systemdlete2 | Only problem with those is they have become commercialized and thus braindead for the free offerings. I'm leery of those. | 02:39 |
Tenkawa | tuxd3v: that is not true btw that all 3 pin are variable | 02:40 |
Tenkawa | the 3 pin fan that rpi sells is a 5000 rpm fixed speed | 02:40 |
systemdlete2 | (I'm thinking of greenbone, e.g.) | 02:41 |
Tenkawa | just pwm just gives it the push/pull ability for gpio | 02:41 |
tuxd3v | Tenkawa, that can hapens, if the board has no pwm features, but if the signal is pwm powered the fan should adjust speed | 02:41 |
Tenkawa | I tried it on another board and it wouldnt chang | 02:42 |
tuxd3v | the case systemdlete2 was puting I believe that it was using a normal motherboard | 02:42 |
Tenkawa | e | 02:42 |
tuxd3v | it will not change if the signal is not a pwm one | 02:43 |
systemdlete2 | tuxd3v, Tenkawa, just for reference, this is a brand-new Gigabyte MB. I think it may be only few years old tech wise. | 02:43 |
systemdlete2 | The manual and FCC declaration are from 2016. | 02:44 |
systemdlete2 | That is relatively new. I admit I am running relatively ancient hardware | 02:44 |
tuxd3v | systemdlete2, yeah at least from the last 12-15 years they should be pwn controlled | 02:44 |
Tenkawa | do x86 boards even use normal pwm? | 02:44 |
aloo_shu | you could use ssh to monitor any number of parameters you choose, couldn't you | 02:44 |
Tenkawa | I thought they used acpi control fans | 02:44 |
tuxd3v | yes, sensorsgive you that information | 02:44 |
tuxd3v | 'sensors' | 02:44 |
systemdlete2 | Oh, tuxd3v, the manual is very clear about this! Definintely PWM controlled, even individual settings in the bios (I haven't looked yet though) | 02:44 |
Tenkawa | nice | 02:45 |
systemdlete2 | aloo_shu, wouldn't that be expensive to poll over ssh? | 02:45 |
systemdlete2 | or clumsy anyway | 02:45 |
tuxd3v | Tenkawa, yes they use for a long period of time.. but in the sbc world that is not the case ufortunatly :( | 02:45 |
tuxd3v | Y know that you have seen cases like that, and I alsosaw tons of them.. | 02:46 |
tuxd3v | but there are alreasy some vendors that at least provide a 2 pin pwm controlled fan.. | 02:46 |
systemdlete2 | Well, this fan is also brand new. So... | 02:46 |
systemdlete2 | I am not worried. I appreciate all of your feedback, everyone. | 02:47 |
Tenkawa | hmm.. I wonder why my gaming rig isnt pwm | 02:47 |
tuxd3v | Now the driver that majority( embeeded sbcs ) is using, is the pwm-fan, and it doesn't have feedback, not full functionality | 02:47 |
systemdlete2 | Seattle's Best Coffee? Or some other SBC? | 02:48 |
Tenkawa | indeed.. I think "that" is the problem | 02:48 |
systemdlete2 | This board has a lot of features, but I don't know if it can make coffee... yet. | 02:48 |
tuxd3v | For example in kernel 4.4 pwm-fan works ok, I even developed a program for RockPro64 for that: https://github.com/tuxd3v/ats | 02:49 |
tuxd3v | but in Mainline 'pwm-fan' only accepts values of 25{0..5}, with 250 being stopped :/ | 02:49 |
tuxd3v | it should go from 0-255 | 02:49 |
systemdlete2 | Single Board Computers. SoC iow | 02:50 |
Tenkawa | sudo hwinfo | grep -i pwm | 02:50 |
Tenkawa | na@OMENPC:~$ | 02:50 |
Tenkawa | yeah thats my hp omen gaming x86 | 02:51 |
Tenkawa | nothing | 02:51 |
tuxd3v | I don't know if that has changed, I myself have thinking in even fork that driver and gives it full functionality, but then when you are not paid, it needs to be done by love, and pine64 has not apreciated me with a new board to develop it, so the process stalled.. | 02:51 |
Tenkawa | but its got 2 major turbines that have 2 turbines that do show up in acpi | 02:51 |
aloo_shu | would 'acpi' and 'pwm' even be mutually exclusive concepts? I would think of acpi as an abstract standard for _interfacing_ with power control elements, while pwm is a hw technique , I'd think software -> acpi -> bios hw driver -> hw chip -> pwm 'ed power, correct me please | 02:52 |
tuxd3v | Tenkawa, install lm-sensors :) | 02:52 |
tuxd3v | then run 'sensor-detect' :) | 02:52 |
tuxd3v | yes to all options.. | 02:52 |
tuxd3v | it will provide you with that sensors information and even more ;) | 02:52 |
Tenkawa | tuxd3v: how do y ou think I monitor my temps: | 02:53 |
Tenkawa | Package id 0: +50.0°C (high = +100.0°C, crit = +100.0°C) | 02:53 |
tuxd3v | so you already are using it :) | 02:53 |
Tenkawa | I've on;y been doing this for 30+ years | 02:53 |
Tenkawa | ok time to play some games then watch tv.. .cheers all | 02:55 |
onefang | fsmithred: where abouts is the refracta ISOs again? | 05:54 |
golinux | refracta.org | 05:58 |
golinux | He's sleeping by now. | 05:59 |
onefang | Thanks. | 05:59 |
golinux | This is more direct: https://get.refracta.org/files/ | 06:00 |
hiddener | hi, can't make window-htile and window-vtile work in cwm, other bindings work fine, except these two, did I miss anything? | 09:25 |
hiddener | bind-key 4-v window-vtile, and bind-key 4-c window-htile | 09:27 |
gnarface | hiddener: it's a slwo channel, stick around for a while though, someone might know | 10:01 |
gnarface | hiddener: *slow channel | 10:01 |
hiddener | nvm, the cwm version I have doesn't support it :) | 10:38 |
gnarface | hiddener: sometimes there is a newer version in beowulf-backports | 11:33 |
hiddener | gnarface: checked, it's added in version 6.7, it's not in unstable/ceres yet too | 11:44 |
hiddener | not sure, but it should work on 6.3.1 too, but it doesn't | 13:14 |
hiddener | not going to tiling mode | 13:17 |
armadillotx | Hello all! how can i do to start on nftables in the init of the system? Tor does it automatically, /etc/init.d/tor = DAEMON=/usr/bin/tor. But the script of nftables in init.d it's set DAEMON=none,because there is not link to /usr/bin. When i do sudo /etc/init.d/nftables status i get it "nft ruleset loaded: yes", and i can see is working, because i set my nftables rules with "counter" and i see how the number rise up. | 15:11 |
armadillotx | So, nftables, is running from the init of the system but it doesn't show it? | 15:11 |
gnarface | nftables is kernel functionality, so anything could load rules for it independent of any init.d script | 15:14 |
gnarface | tor probably requires some rules just to work right so it loads them itself would be my guess | 15:14 |
gnarface | there's nothing particularly special about these scripts other than the functions they define | 15:15 |
gnarface | you could easily mimic one for your own rules or just throw them in /etc/rc.local | 15:15 |
gnarface | if you're working with tor you'll have to be careful not to stomp on whatever it's doing too | 15:15 |
armadillotx | oh, ok.thank you very much. | 15:16 |
gnarface | no problem | 15:17 |
gnarface | there might be some other tools to load and manage rules in the repo, i dunno any names off the top of my head though | 15:17 |
armadillotx | tor+nftables are working fine :-) i use links2 with tor and it works fast | 15:18 |
armadillotx | they say tor makes the connection slow, but it's not in my case, it flies :-) | 15:19 |
armadillotx | ok, thank you, knowing about nftables is a kernel functionality is enough | 15:21 |
Tenkawa | does nftables have a load/restore script he/she could use like iptables does? | 15:26 |
Tenkawa | (I don't use nftables so I'm just throwing that out there) | 15:27 |
gnarface | i think it is present but the issue is that rules are being loaded without it running, at least that's how i read it... | 15:28 |
Tenkawa | ahh yes… that could be blocked with a chattr if they wanted to | 15:29 |
Tenkawa | that and a chmod could block it from being accessed and timestamp needed | 15:30 |
Tenkawa | strange that tor has it hooked in directly | 15:31 |
armadillotx | tor has all the settings to init, you don't have to do more than writte 4 lines in torrc | 15:36 |
armadillotx | you can see in the start of the system Daemon tor: done Not the same with nftables, for this i asked, Daemon nftables: none, appear. | 15:39 |
Tenkawa | oh | 15:42 |
Tenkawa | you need nftables to auto start and its not right? | 15:43 |
Tenkawa | (sorry.. still a bit sleepy) | 15:44 |
armadillotx | i would like it appear like Daemon nftables: done, like tor, but no, at least it load and works. Is only a silly thing. | 15:44 |
Tenkawa | afk… on phone | 15:44 |
GyrosGeier | Tenkawa, I believe someone ported the iptables-persistent package to nftables | 15:47 |
armadillotx | i had problems with nftables to start service when i decided to install and run Devuan with Runit. I couldn't make it work.So i reinstalled Devuan with SysVinit again and it's working this way. | 15:47 |
GyrosGeier | the problem with the iptables stuff is that the tables aren't a service, so there is no notion of "running" | 15:48 |
GyrosGeier | in iptables-persistent, I used an init script because sysvinit doesn't have a notion of "running" either, so that maps well | 15:48 |
GyrosGeier | (or rather, sysv-rc doesn't have a notion of "running") | 15:49 |
armadillotx | is sysv-rc different of sysvinit? the first time i see it :-) | 15:50 |
GyrosGeier | sysvinit is the core init with /etc/inittab | 15:51 |
GyrosGeier | sysv-rc is /etc/init.d, which is called from the inittab as a one-shot entry on runlevel change | 15:51 |
armadillotx | ok | 15:51 |
GyrosGeier | classical init has service supervision in the inittab, and that is used for getty | 15:52 |
GyrosGeier | then we drop service supervision for the actual services because they are more complex than getty and blindly restarting them is more likely to be harmful | 15:53 |
GyrosGeier | then systemd people come and complain that there is no service supervision :3 | 15:53 |
armadillotx | i thought to try Devuan with Open-rc, but at the end i decided to go to something works for me. | 15:55 |
Tenkawa | systemd *shudder* | 15:55 |
fys | I thought offensive words were banned on Freenode? | 15:57 |
fys | jk jk | 15:57 |
Tenkawa | lol | 15:57 |
armadillotx | about systemd, i hate to see the dir of systemd in /etc/ is possible to delete it? :-D | 15:58 |
armadillotx | i mean, without any risk for the system. | 15:59 |
onefang | I'm guessing that some packages have systemd unit files that get installed there, even if not used coz they also have sysvinit scripts. | 16:00 |
fys | ^ | 16:00 |
fys | You can use dpkg-query to find out. | 16:00 |
fys | or something more modern i don't use | 16:00 |
onefang | So you could delete it, but it might just get recreated. | 16:00 |
armadillotx | i supposed.:-( | 16:01 |
fys | The very fact your system doesn't USE systemd means that the files aren't even used, so no - it *SHOULDN'T* have any negative consequences (ymmv, no warranties) | 16:01 |
fys | :) | 16:01 |
Tenkawa | are these pkgs coming from devuan? | 16:02 |
Tenkawa | I assume no | 16:02 |
Tenkawa | if they were I would assume the pkg maintainers would be taking the non compliant files out | 16:03 |
onefang | Also if you use any sort of file integrity system that checks files against the packages they came from, expect lots of complaints about missing files. | 16:03 |
Ankokukishi | its called shitstainD guys, remember the name | 16:03 |
fys | Tenkawa: How are they "not compliant"? | 16:03 |
armadillotx | i supposed it was because Devuan comes from Debian it keeps some dir there. | 16:03 |
armadillotx | systemDick | 16:03 |
Tenkawa | fys: because devuan doesn't use systemd | 16:03 |
fys | By default. | 16:03 |
Ankokukishi | because guys like fsmithred have actual morals in computing we have a safe space away from garbage bloat like shitstaind | 16:04 |
fys | If I would get it from ./configure --prefix=/usr make make install, i want it | 16:04 |
Tenkawa | fys: no.. at all | 16:04 |
fys | Tenkawa: I can install systemd on my machine if I want.. I don't but I could. | 16:05 |
fys | Removing files for vanity is silly. | 16:05 |
Tenkawa | fys: systemd is not "packaged" officially in devuan by any context | 16:05 |
fys | Didn't say it was. | 16:05 |
Tenkawa | vanity?? | 16:05 |
onefang | There are not that many of us Devuan devs. Most packages are exactly the same as the Debian package. If a package comes with both sysvinit and systemd stuff, but works fine without systemd, then we don't bother making any changes. | 16:06 |
fys | If the authors of those software packages intend to include the systemd files it's not the devuan packagers job to make sure you don't have an /etc/systemd directory. | 16:06 |
brocashelm | the magic of amprolla plus the blacklists (to filter out systemd packages) and 200+ package forks to remove/alter some lines of code for inits | 16:06 |
Ankokukishi | when i can get more time, i wanna help out in any way i can. Start learning how to code so i can help the devs of devuan | 16:07 |
brocashelm | you will know it's a devuan fork if it says "devuan" in the version | 16:07 |
fys | libvirt-daemon-system-sysv/unstable 7.0.0-3+devuan2 all such as? | 16:08 |
fys | Never noticed that before tbqh - but good to know. | 16:08 |
fys | Makes sense. | 16:08 |
Ankokukishi | like refracta, right? :) | 16:09 |
Ankokukishi | i use refracta personally | 16:09 |
fys | What makes refactra different from standard devuan? | 16:10 |
Ankokukishi | not a whole lot, i might midnight commander installed default, and like 20mb lighter or something lol | 16:11 |
Ankokukishi | and the icons/backgrounds, but thats superficial | 16:12 |
fys | I use Openbox anyways | 16:12 |
onefang | apt install mc # one of the first things I do on any new system. lol | 16:12 |
brocashelm | there are some differences; doesn't ship with metapackages installed (so you won't take a whole system down over one package you wanted gone), uses alsa instead of pulseaudio, custom sysrq keys, some default programs in place such as geany and abiword (instead of libreoffice). it's meant to be lighter than a standard devuan. its championing feature being the refractatools so you can make your own distro and write to a usb drive | 16:13 |
Ankokukishi | i would install geany anyway | 16:14 |
Ankokukishi | even though i use vim more | 16:14 |
Ankokukishi | just as a second choice | 16:14 |
Ankokukishi | oh yeah i forgot about refractatools | 16:15 |
Ankokukishi | and refracta2usb | 16:15 |
Ankokukishi | take your entire workflow on the go | 16:15 |
brocashelm | the isos (devuan and refracta beowulf 64-bit with xfce) are also different in size. refracta is 834 mb, whereas devuan is 1.2 gb | 16:15 |
Ankokukishi | ive read lots of mint users using it to write their mint setups to USB | 16:16 |
brocashelm | so refracta is aimed at being more lightweight in packages while being flexible enough to become its own os if you treat it right | 16:16 |
Ankokukishi | yeah | 16:16 |
Ankokukishi | fsmithred is the guy behind refracta | 16:16 |
brocashelm | refracta also comes with firmware-tools directory in ~/ so you could install wifi and other non-free drivers from debs without having to connect online | 16:17 |
Ankokukishi | wooo that i didnt know | 16:17 |
brocashelm | *firmware.beowulf | 16:18 |
Ankokukishi | but wifi is dangerous, not gonna bother lol | 16:18 |
brocashelm | so in case you don't have access to ethernet, you have some local debs to play with | 16:18 |
Ankokukishi | oh yeah i see that folder now | 16:18 |
Ankokukishi | i try to keep as much non-free stuff off my pc | 16:20 |
brocashelm | so it has a lot to offer | 16:20 |
brocashelm | yeah you can check with vrms | 16:20 |
Ankokukishi | yeah only thing non free is libdvd-pkg | 16:20 |
brocashelm | and removing non-free and contrib from your source.list as ultimatum | 16:20 |
brocashelm | but things like amd gpu drivers might be necessary (like in my case with dual monitors) | 16:21 |
Ankokukishi | i dont game on my pc or do anything that requires a graphics card like that | 16:21 |
brocashelm | due to debian/devuan not using non-free kernel blobs by default | 16:21 |
Ankokukishi | i sit and chat, and watch shit on bitchute or odysee. gaming is reserved for ancient things like ps2s and psps lol | 16:22 |
brocashelm | the only thing keeping debian/devuan from being fsf-approved is their repos officially host non-free packages (as long as you edit your /etc/apt/sources.list) and maybe the default kernel not being compiled as libre-only | 16:23 |
Ankokukishi | ah | 16:23 |
Ankokukishi | thats kinda overboard, but at least devuan is like right on the edge of approval, meaning its 99.9% libre :) | 16:24 |
Ankokukishi | my friend was using mint, then switched to refracta | 16:24 |
Ankokukishi | then told me "HEY TRY REFRACTA" | 16:24 |
Ankokukishi | tried it on my previous laptop first, loved it | 16:24 |
Ankokukishi | then switched my daily driver over to it | 16:24 |
brocashelm | https://www.gnu.org/distros/common-distros.html | 16:25 |
onefang | Think you have wandered off the support channel stuff and into #devuan-offtopic. | 16:25 |
Ankokukishi | lol | 16:25 |
Ankokukishi | oops | 16:25 |
mason | Ankokukishi: Hey, please change your quit message. | 16:37 |
aloo_shu | ^^ | 16:40 |
iv4nshm4k0v | Re. /etc/systemd , I have path-exclude=/etc/systemd/system/* in my /etc/dpkg/dpkg.cfg.d/zz-no-sd . | 16:40 |
aloo_shu | you really can be a cool guy without that, or 'shitstain' strongwordery | 16:44 |
aloo_shu | it's not neccessary, and we can build a better network of mutually helpful foss users/devs | 16:44 |
aloo_shu | you see, if people are _telling_ you, instead of reporting straight away, that means they _did_ wait to see if you actually have something positive to say, they're telling _you_ instead of talking behind your back, and they _would_ like you to stay around - just not that thoughtlessness part, you can do better | 16:51 |
mason | Since I evidently got the syntax wrong I'll remove the lingering ban anyway and we can await a response. | 16:59 |
fsmithred | I'm glad you removed that ban. | 17:00 |
fsmithred | But I think the one on mikee was too weak. I forget what name he used. | 17:01 |
mason | Oh, I missed that. | 17:01 |
fsmithred | you did it | 17:01 |
Josh_2 | Hi, when booting I keep getting the error "failed to install nvidia" just before being prompted to unlock my encrypted partition | 17:02 |
fsmithred | mason pm | 17:03 |
Josh_2 | and now I'm in my DE nvidia-settings wont open etc | 17:03 |
Josh_2 | I tried reinstalling nvidia drivers | 17:04 |
Josh_2 | I hope my gpu isn't borked | 17:05 |
Josh_2 | lspci says that the nvidia kernel module is in use | 17:06 |
Tenkawa | Josh_2: when you run lsmod is nouveau running? | 17:08 |
Josh_2 | no just nvidia | 17:09 |
Tenkawa | hm interesting | 17:10 |
Josh_2 | I dont see any complaints about nvidia in dmesg | 17:15 |
Josh_2 | just after grub I get that pesky error | 17:16 |
Josh_2 | its never happened before | 17:16 |
Ankokukishi | mason: its a movie song title | 17:18 |
mason | Ankokukishi: Regardless. It needs to change. | 17:18 |
Josh_2 | ima try boot windows see if that works | 17:20 |
hiddener | :D | 17:29 |
Josh_2 | Okay I checked to see if my chip was actually working using windoge. It is | 17:41 |
Josh_2 | The error I am getting when booting is 'udevd[132]: Error running install command for nvidia" | 17:42 |
Josh_2 | this comes up before my prompt to unlock my disk | 17:43 |
Josh_2 | fixed it! | 18:16 |
Vall | Josh_2: how did you fix it? | 19:02 |
Josh_2 | Vall: by not ending up in runlevel 2 | 19:33 |
Josh_2 | OpenRC keeps taking forever to boot because it gets stuck configuring network devices when I have no ethernet cable plugged in. So I was ctrl + c that step which was putting me into runlevel 2 | 19:34 |
Josh_2 | I should fix that | 19:37 |
mason | Josh_2: I hit that one a while back, in Gentoo. The answer was evidently to rely on NetworkManager or have a dummy network service instead of it. | 20:10 |
fsmithred | Josh_2, the fix is to change 'allow-hotplug' to 'auto' in /etc/network/interfaces | 20:11 |
mason | That was going to be my other question - can ifupdown be the listed provider. | 20:11 |
fsmithred | yeah, or network-manager | 20:11 |
fsmithred | yeah, the edit changes control from udev to ifupdown | 20:11 |
mason | My assumption is that this is with OpenRC's native/built-in NIC management. | 20:11 |
fsmithred | I don't know if openrc changes it | 20:12 |
fsmithred | I've only dealt with the issue in sysvinit | 20:12 |
fsmithred | in/with | 20:12 |
mason | My experience was also in Gentoo, and hence likely not identical to what happens here. | 20:13 |
Josh_2 | mason: I have had the same issue with Gentoo as well | 20:27 |
mason | Josh_2: I think the best way to think of it is that it's by design, then. It's expected that you're a stable workstation or server and have ethernet, unless you use something "mobile" to manage changing connections. | 20:29 |
mason | Josh_2: Looking at history, maybe try rc_need="!net" in rc.conf ? | 20:31 |
mason | Josh_2: And then manage networks manually once you're booted. | 20:31 |
mason | That was advice for Gentoo, but it might apply here. (I haven't spun up an OpenRC system in Devuan ever, so I'm wildly stabbing here.) | 20:31 |
Josh_2 | I am using a laptop so yeh, not a stable workstation | 20:34 |
Josh_2 | I am looking at getting a stable workstation, I have no need for this machine | 20:38 |
mason | Josh_2: Well, but you want that box to work. I'd see if the rc.conf thing is applicable. | 20:46 |
mason | Or maybe ifupdown can be plugged in as a provider. Unsure. That'd be my choice if it's possible. | 20:46 |
mason | https://wiki.debian.org/OpenRC makes me think it ought to Just Work with ifupdown | 20:48 |
rwp | fsmithred, (Josh_2) Is it possible that it might be reversed and need allow-hotplug instead of auto? | 20:55 |
rwp | AIUI the auto system makes networking a hard requirement to boot. But allow-hotplug uses the event driven system, therefore optional. | 20:56 |
rwp | Generally I use auto with all servers and other always-on systems. But for actual hotplug networking USB dongles I use allow-hotplug. | 20:58 |
rwp | Josh_2, To summarize you have installed Beowulf Stable, Selected OpenRC, then had an error during boot last message was about nvidia? Is that correct? I would like to try to recreate it here if possible. | 21:01 |
Josh_2 | Yes I have an error at boot just after grub, before being prompted for my pass to decrypt my root part | 21:09 |
fsmithred | how could it do network setup before decrypting root? Are you sure about the cause of the delay? | 21:14 |
mason | fsmithred: It's possible to stuff networking into your initramfs. Think, for example, of running dropbear from initramfs to decrypt root. | 21:17 |
Tenkawa | yeah initramfs can do a lot of stuff | 21:18 |
fsmithred | yeah, ok. | 21:18 |
Tenkawa | initramfs + busybox can run a whole lightweight machine | 21:19 |
Tenkawa | if built right | 21:19 |
Josh_2 | different error | 21:20 |
Tenkawa | (+ a kernel of course) | 21:20 |
Josh_2 | rwp was asking about the nvidia error | 21:20 |
Tenkawa | Josh_2: what error you get this time? | 21:20 |
Josh_2 | Same error but because I didn't Ctrl + C when stuck waiting for networking I did end up in runlevel 2, so Nvidia worked despite the error | 21:20 |
Tenkawa | I run nvidia on my primary box | 21:21 |
Tenkawa | let me scroll up to see if I can see what you typed | 21:21 |
Tenkawa | oh… is it telling you waiting on network devices? | 21:22 |
mason | Josh_2: FWIW, I don't specify auto *or* allow-hotplug on my laptop. | 21:25 |
mason | Josh_2: I just bring it up by hand. That said, I don't know if you need to do anything to get ifupdown to work nowadays - like, telling OpenRC it's your network provider. | 21:25 |
Tenkawa | yeah I'd turn off anything auto for the moment to test | 21:26 |
Tenkawa | mason: I'm reading that that and also eudev could be doing it | 21:26 |
Tenkawa | er s/that that/that/ | 21:27 |
Tenkawa | afk… bbiafm | 21:28 |
Tenkawa | back | 21:43 |
rwp | Josh_2, So in addition to what I summarized additionally you created it as a full encrypted system, plus using an nvidia graphics card. | 21:44 |
rwp | Josh_2, Please remember that while you know what you did none of us know anything other than what we read. :-) | 21:44 |
Tenkawa | still seems to be an odd spot for it to break | 21:46 |
Tenkawa | and "circumstances" | 21:46 |
Tenkawa | hard drive related I could understand… but a video card/driver… thats a bit tougher unless its messed up the initrd | 21:47 |
Tenkawa | and its not handing off the unencrypt functions right due to a bad driver install | 21:48 |
Josh_2 | who what | 21:51 |
Josh_2 | its not broken, its working | 21:51 |
Josh_2 | i'm using it right now despite the error. I cocked up thinking that the error about nvidia was the cause the of the problem | 21:51 |
Josh_2 | but in fact it was me being impatient | 21:52 |
Tenkawa | ok.. good… but the delay still could be fixed probably | 21:53 |
Josh_2 | yeh | 21:53 |
Josh_2 | the delay is a pita | 21:53 |
Tenkawa | whats the last thing it says before it happens? | 21:54 |
rwp | Quite often the last message displayed is the last thing that worked and the problem is the next thing that hasn't printed a message yet. | 21:54 |
Tenkawa | that will help isolate which script is causing it | 21:54 |
rwp | That was me referring to the previous nvidia message. | 21:54 |
Tenkawa | rwp: or it could also possibly be the culprit | 21:55 |
rwp | Sorry Tenkawa our lines passed in flight. I hadn't seen your comment before mine. But afterward what I said sounded critical but I hadn't seen it yet. | 21:55 |
Tenkawa | I have a suspicion what it is | 21:55 |
Tenkawa | but I need output of what he is seeing | 21:56 |
Tenkawa | this is eerily familiar o= | 21:56 |
Tenkawa | er s/o=// | 21:56 |
Josh_2 | rwp: the nvidia error is "udevd[132]: Error running install command for nvidia" | 21:56 |
Tenkawa | Josh_2: do you see anything on the screen about waiting for root to mount? | 21:57 |
Tenkawa | or a timer counting down for a step | 21:58 |
Josh_2 | rwp: https://paste.debian.net/1194468/ | 21:58 |
Tenkawa | (if so on the timer that indicates a eudev or rc job is failing) | 21:59 |
Tenkawa | udevd.. hmm interesting | 22:00 |
Josh_2 | That is the nvidia error, however nvidia is working | 22:00 |
Josh_2 | Now I can't remember what OpenRC says when it pauses, its basically just saying waiting for network interfaces | 22:01 |
Josh_2 | and then it sits for 15 minutes before giving up and continuing | 22:01 |
fsmithred | encrypted lvm? | 22:01 |
Tenkawa | good q | 22:02 |
Tenkawa | i know he said it was luks | 22:02 |
Josh_2 | yes | 22:03 |
Josh_2 | encrypted lvm with luks | 22:03 |
fsmithred | and was it installed from regular installer iso or from live? | 22:03 |
Josh_2 | installer | 22:03 |
fsmithred | installer if lvm, yes | 22:03 |
rwp | Why would it say "udev" instead of "eudev"? | 22:03 |
Tenkawa | rwp: thats what I was wondering too | 22:03 |
fsmithred | eudev supplies udev (in name) | 22:03 |
fsmithred | sorry, "provides" | 22:03 |
fsmithred | I didn't look yet... | 22:04 |
Josh_2 | I looked the nvidia error up, seems that all the forums links are do Devuan :P | 22:04 |
rwp | Provides is a dpkg package Depends, yes, but syslog is a program name logging. Shouldn't that be eudev? I'll go look at my systems... | 22:04 |
Tenkawa | I only use arm… but in arm it actually calls it eudev | 22:05 |
Tenkawa | [ 2.187474] udevd[109]: starting eudev-3.2.9 | 22:05 |
fsmithred | 560 ? Ss 0:02 /sbin/udevd --daemon | 22:06 |
Tenkawa | thats on arm64 devuan | 22:06 |
Tenkawa | let me look at ps | 22:06 |
Tenkawa | yeah just udev but in dmesg its eudev which is what he gave us | 22:07 |
Tenkawa | odd indeed | 22:07 |
Tenkawa | what release are you on btw | 22:07 |
Josh_2 | https://paste.debian.net/1194469/ | 22:08 |
rwp | And here everything I expect to be *udev is logged as "kernel:" so... no data from me. | 22:08 |
fsmithred | I don't see the error I was expecting. Something about "waiting 1000000 seconds for <device>" repeated many times | 22:08 |
Tenkawa | ok good | 22:08 |
Tenkawa | fsmithred: yeah I expected it too | 22:09 |
Tenkawa | and I "really" hate nvidia drivers ftr | 22:10 |
Tenkawa | cause me nothing but grief so I still wouldnt count them out as having broken something | 22:10 |
Josh_2 | Tenkawa: dw, next time I get a device it won't be nvidia | 22:10 |
rwp | If it is the free software nouveau then it is okay but the proprietary nvidia drivers have been trouble, depends upon the version. | 22:10 |
Josh_2 | but I've had this laptop 5 years | 22:10 |
Tenkawa | exactly | 22:10 |
Tenkawa | yeah my omen has a built in nvidia card | 22:11 |
rwp | It's almost new then! :-) (I am typing this on a 9 year old model) | 22:11 |
Tenkawa | mine is <2 yrs lol | 22:11 |
hiddener | 14 year here :] | 22:11 |
Josh_2 | I have a thinkpad x220 which is great, its a much better 'laptop' than this | 22:11 |
Tenkawa | its a server now | 22:11 |
rwp | I need to replace the fan in my beloved Thinkpad T60p and then it will be running again. | 22:12 |
Tenkawa | it made a better build server | 22:12 |
Josh_2 | unfortunately I like to play the occasional game, can't do that on my thinkpad | 22:12 |
rwp | I should have added that the laptop I was typing on was an X220. :-) | 22:12 |
Tenkawa | Josh_2: I do that on my MBP now | 22:12 |
Josh_2 | MBP? | 22:13 |
Tenkawa | Macbook pro | 22:13 |
Josh_2 | ah | 22:13 |
Tenkawa | I own 0 desktops | 22:13 |
Josh_2 | I use my laptop like a desktop, so I might as well get a desktop | 22:14 |
Tenkawa | oh.. i see what might be happening now | 22:16 |
Tenkawa | this thing might be running dkms rebuild on your kernel driver every reboot | 22:16 |
Tenkawa | Josh_2: check this if you can | 22:18 |
Tenkawa | do a sudo dpkg -l | grep nvidia | 22:18 |
Tenkawa | see how many entries you have | 22:18 |
Tenkawa | you might have legacy drivers + dkms packages confusing it | 22:19 |
Josh_2 | https://paste.debian.net/1194470/ | 22:19 |
Tenkawa | yikes | 22:19 |
Josh_2 | ruh oh | 22:20 |
Tenkawa | I ran a search on that error.. and thats what came up… its not a problem.. it just explains the slowness | 22:21 |
Tenkawa | there's a lot of excess nvidia software on there… some that is unnecessary | 22:22 |
Tenkawa | unfortunately without knowing your machine I'm not sure specificly which pkgs | 22:23 |
Tenkawa | even on my i7 with 12 threads it takes about 1.5 min at boot to build if I install a new set + kernel | 22:25 |
Josh_2 | the slowness is because it sits trying to configure the eth0 device when I have no ethernet cable plugged in | 22:26 |
Tenkawa | oh I can tell you a quick way to turn that off | 22:27 |
Tenkawa | which way are you using to configure your net devs? | 22:28 |
Tenkawa | turn auto off and use allow-hotplug | 22:29 |
Tenkawa | if you use interfaces | 22:30 |
Tenkawa | and /etc/network/if-up.d/00check-network-cable from https://packages.debian.org/ifupdown-extrahttps://packages.debian.org/ifupdown-extrahttps://packages.debian.org/ifupdown-extra to skip the interface if no cable is connected. | 22:30 |
Josh_2 | https://paste.debian.net/1194472/ I changed this earlier and I haven't rebooted since I changed it | 22:31 |
Tenkawa | er that script from ifupdown-extra pkg | 22:32 |
Tenkawa | you have to restart networking for those changes to take effect | 22:33 |
Tenkawa | and you need to make sure you have that extra file if you dont | 22:33 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!