FilipZ1 | rwp: As rrq mentioned, its the "Standard DHCP wirelsess setup", and before that, I have "auto lo", and "iface lo inet loopback" | 00:03 |
---|---|---|
rwp | I just now set up and tested the install guide chimaera network-configuration and it works here okay. | 00:03 |
FilipZ1 | rrq: How should I create this socket? | 00:04 |
FilipZ1 | rwp: It worked ok too to me, untill this happened | 00:04 |
FilipZ1 | It is rather a specific hardware related issue, I think. | 00:05 |
rwp | Information from my end of things: https://paste.debian.net/plain/1255884 | 00:09 |
rrq | the socket is created by the script but the directory is not | 00:29 |
rrq | the directory should be "drwxrx----" with root:netdev | 00:30 |
rrq | well, drwxr-x--- | 00:31 |
rrq | mmmm yes that directory gets created correctly on reboot, so I'm not sure this is a cause | 00:37 |
rrq | rwp: yor setup presents two "hurdles".. firstly that only root can interact with wpa_supplicant, and secondly that it's fixed to a single access point | 00:55 |
rrq | the first hurdle is supposed to be addressed with the ctrl_interface=... line in the conf, including a GROUP=netdev setting | 00:57 |
rrq | the second is addressed with the "update_config=1" conf setting and using the wpa-roam feature, which opens up for using wpa_gui and wpi_cli | 01:01 |
rrq | wpa_cli | 01:01 |
FilipZ1 | rwp: on my side: 1. regardless interfaces file, I described it earlier. 2. I wrote the errors from ifup wlan0 at the beginning. 3. How the wpa_supplicant process that tries to turn on look in my case: "wpa_supplicant -s -B -P /run/wpa_supplicant.wlan0.pid -i wlan0 -W -D nl80211,wext -c /etc/wpa_supplicant/wpa_supplicant.conf". 4. Nothing in /run/ | 01:02 |
FilipZ1 | contains "wpa" and nothing here changes when invoking ifup wlan0 | 01:02 |
rwp | rrq, It's not "my setup" it's the setup from network-configuration.html that you pointed me too. I just replicated that and verified that it was working. | 01:03 |
rrq | right | 01:04 |
rwp | My own setup does things differently. I use either the wpa_supplicant instructions from that package or I manually launch everything. | 01:04 |
rwp | But as far as needing root for it... Aren't we all root on our own systems? | 01:05 |
rwp | I have to drop afk again for a bit... | 01:05 |
rrq | FilipZ1: something is very odd... perhaps add "set -x" to /etc/wpa_supplicant/ifupdown.sh to get a detailed log of its invocation,... is there something "odd" with "/run" ? | 01:07 |
FilipZ1 | rwp: But it was not "iface wlan0 inet dhcp | 01:08 |
FilipZ1 | wpa-ssid myssidhere | 01:08 |
FilipZ1 | wpa-psk mypassphrasehere", but " | 01:08 |
FilipZ1 | "auto wlan0 | 01:08 |
FilipZ1 | iface wlan0 inet manual | 01:08 |
FilipZ1 | wpa-roam /etc/wpa_supplicant/wpa_supplicant.conf | 01:08 |
FilipZ1 | iface default inet dhcp" and that is how it looks im my case, so not a single fixed AP | 01:08 |
FilipZ1 | And /etc/wpa_supplicant/wpa_supplicant.conf contains exactly what is described on that page to have as well | 01:10 |
rrq | yes, and I have the same (though working :)) | 01:11 |
rrq | please check "mount | grep '/run '" .. should be tmpfs, rw, and mode=755 | 01:13 |
rrq | another thing is that I run wpa_gui as non-root user, with that user being in netdev group | 01:21 |
rrq | and note that when/if you add a user to a group, then that user must re-login for that to be in effect | 01:22 |
rrq | (you may also use "newgrp" for actually bwing in the group with children processes to that shell) | 01:24 |
FilipZ1 | I just copied the ifup wlan0 output with verbosity of set -x . I may share it later, after leaving this problem state | 01:26 |
rrq | just to confirm: ifup and ifdown are used by root (only) but wpa_gui may be run by a non-root user who is in netdev group | 01:28 |
FilipZ1 | Yes, it does work like this to me too. wpa_gui runs with the non-root privileges | 01:30 |
rrq | "ifup wlan0" __should__ work, esp following "ifdown --force wlan0" | 01:31 |
FilipZ1 | I did it like it, but with the previously mentioned errors, instead | 01:32 |
rrq | if you want wore and deeper logging, you'd use "strace -f ifup wlan0" .. that log would very likely hide a needle of actual problem cause somewhere :) | 01:34 |
rrq | wore=more | 01:34 |
rrq | hwo about "dpkg -l wpasupplicant" .. mine is 2:2.9.0-21 (amd64) | 01:37 |
FilipZ1 | "Mount | grep '/run'" output: "tmpfs on /run/ type tmpfs (rw,nosuid,nodev, noexec,realtime,size=1625372k,mode=755)" | 01:44 |
FilipZ1 | And there are 3 lines more, the last beginning with "portal" instead of "tmpfs". | 01:45 |
rrq | yes .. you missed the space '/run ' :) | 01:49 |
rrq | oh wait | 01:50 |
rrq | your grep would need '/run/ ' | 01:50 |
FilipZ1 | For the command beginning with "strace" I got the answer that it was not found. | 01:51 |
rrq | anyhow, that's fine | 01:51 |
FilipZ1 | There earlier it was "/run", not "/run/", I wrote that incorrectly | 01:52 |
rrq | ok. fine anyhow... hmm .. do you have netdev group in /etc/group? | 01:53 |
FilipZ1 | And that space shouldn't matter, right? | 01:54 |
FilipZ1 | I will check | 01:54 |
rrq | well that space in the grep was for avoiding mounts within /run, with | 01:54 |
FilipZ1 | Right, so I get that I wouldn't get those 4 additional lines, this way. :) | 01:55 |
FilipZ1 | 3* | 01:55 |
rrq | yes that was the idea :) | 01:55 |
rrq | I guess your non-root is in netdev so the group exists | 01:56 |
FilipZ1 | About the wpasupplicant version, my have the samu number | 02:00 |
rrq | mmm so other than file access problem there is the device access.. maybe there is a problem there | 02:02 |
FilipZ1 | I do have the "netdev" group. It does refer to my user name here | 02:02 |
rrq | do you have networkmanager or something odd like that running? | 02:02 |
FilipZ1 | I removed it earlier and did apt purge and apt autoclean | 02:04 |
rrq | yes if something else has got some lock on the wlan0 device... is "iwconfig wlan0" ok? | 02:05 |
FilipZ1 | I don't know. What should be the answer if its ok? | 02:06 |
FilipZ1 | IEEE 802.11 ESSID:off/any Mode:Managed Access point: Not-Associated Tx-Power=22 dBm Retry short limit:7 RTS thr:off Fragment thr:off Power management:on | 02:10 |
FilipZ1 | With sudo it also prints "Encryption key:off" | 02:10 |
FilipZ1 | Oh, a "wlan0" before "IEEE", of course :) | 02:11 |
rrq | ok.. sorry I made an strace of my ifup .. got disconnected | 02:17 |
rrq | btw may I confirm that /etc/network/interfaces end with a newline | 02:19 |
FilipZ1 | Strace does not work to me. "Command not gound" | 02:19 |
rrq | not sure hoe that would stop wpa_supplicant from starting though | 02:19 |
FilipZ1 | Does the newline at the end of interfaces matter? | 02:20 |
rrq | install "strace" | 02:20 |
rrq | without it, the last line is not seen | 02:20 |
FilipZ1 | I do not have the newline at the end of interfaces file | 02:21 |
rrq | well, you can see the line, but programs often don't | 02:21 |
rrq | ok, add that newline and see if that makes any difference (keep it regardless) | 02:22 |
FilipZ1 | It still fails, so I guess it didn't | 02:24 |
FilipZ1 | But I keep it ofc | 02:25 |
rrq | right, when/if you explore an strace dump, keep in mind that the error cause is somewhere well before the end, because it is followed by "successful" recovery and error message actions | 02:26 |
rrq | if it's a file permission issue, then it'd be an "openat" call that fails when it should succeed | 02:27 |
FilipZ1 | I do not have the connection so I cannot install it, unless I restart my PC, leving this problem state | 02:27 |
rrq | true. but it seems very peculiar that it ran for a while and then stopped working | 02:28 |
FilipZ1 | I didn't create that directory in /run yet. Should I try it now? | 02:29 |
rrq | well, try that, although I suspect that's not a cause | 02:29 |
rrq | what does "df -h" say about /run ? | 02:30 |
* rrq biab | 02:31 | |
FilipZ1 | tmpfs 1.6G 1.8M 1.6G 1% /run | 02:32 |
FilipZ1 | "Filesystem Size Used Avail Use% Mounted on" | 02:33 |
FilipZ1 | What is "biab"? | 02:34 |
fsmithred | back in a bit (soon) | 02:39 |
FilipZ1 | From the ifupdown.sh content I can see that "set -x" was already here in some other place, I guess that it would be triggered by the addition of a proper flag. I will remove it now to not get additional output, since it shouldn't be needed with strace, right? | 02:42 |
FilipZ1 | Also, I think that I saw earlier the "Trace" verbosity flag, related to some network related deamon. Is it the same thing as "strace" here? | 02:45 |
anjan | hi yall, should jellyfin work on devuan? | 02:47 |
rwp | FilipZ1, I configured up the "roaming" configuration. Here is my result with that config. https://paste.debian.net/plain/1255889 | 03:00 |
rwp | anjan, I know nothing about jellyfin but it says it supports a bunch of software distros, does not say it requires systemd, so it should work on Devuan okay. | 03:01 |
rwp | Why don't you try it and then report back! :-) | 03:02 |
FilipZ1 | rrq: You wrote that the wpa_supplicant directory should be both in the group root and netdev, right? | 03:06 |
FilipZ1 | rrq: Creation of the wpa_supplicant directory didn't help | 03:15 |
FilipZ1 | With the 750 privileges, root owner, and netdev group owner. | 03:16 |
rrq | right user root, group netdev ... that group is needed for not-root user operating it via wpa_gui | 03:32 |
rrq | I wonder, is there an "old" wpa_supplicant running? | 03:37 |
rrq | rwp's setup is fine aprat from not allowing non-root user operating via wpa_gui | 03:39 |
rwp | I am just following the recipes from the posted wiki articles! And I don't think sudo is that hard to use. :-) | 03:46 |
rwp | Also I am pretty sure that roaming configuration doesn't need any triggering after it is online as wpa_supplicant then notices networks and automatically connects. | 03:47 |
rwp | I am using i3 with the i3status tray and by default it reports the network connection status. So I always can see connection status in the system tray area. | 03:48 |
rrq | yes, manual fiddling is only needed if you are roaming among unknown access points | 03:50 |
rwp | I set up multiple ones for the client as a test and it seems to automatically connect to any of them that it finds available. | 03:51 |
rwp | The priority seems rather random, probably just first seen is first tried. | 03:51 |
rwp | It also looks like this will solve a problem I have been seeing with my previous method of connecting to open access points. | 03:52 |
rrq | I think unfortunately FilipZ1's issue is at a more fundamental level; something different in his setup from ours | 03:52 |
rwp | I was previously using the Linux kernel internal driver saying "sudo iw dev wlan0 connect 'ssidnamehere'" and the kernel connecds. | 03:53 |
rwp | It works. But if there is any glitch then the kernel simply reports "connection lost" and drops. Needing a loop scriptlet to reconnect. | 03:53 |
rwp | The "roaming" config posted uses wpa_supplicant for it which knows to retry and should solve that issue of glitches dropping the connection. | 03:54 |
rwp | "> something different in his setup from ours" Yes. Those are always the most difficult to debug. Because the cause is always going to be unexpected. | 04:11 |
FilipZ1 | I want to download the package("strace", to be specific. The packsage manager doesn't tell about it having any more dependencies needing to be downloaded as well) separately, to upload it from another source to my pc, without the internet connection. Where exactly can I find a proper file to get? | 04:47 |
FilipZ1 | Nvm, found it | 04:51 |
rwp | If you need to do that often there is a package "apt-offline" designed for that type of airgap sneakernet transfer. | 04:54 |
gnarface | alright, now that i got the other mystery figured out with qemu now i have to do something about audio | 04:57 |
gnarface | the alsa device shows up in the guest but i just get errors trying to play to it | 04:57 |
gnarface | wtf? | 04:57 |
gnarface | possibly also something stupid simple like the gateway thing | 04:58 |
FilipZ1 | I would like to upload a file from my Android phone, but my system doesn't detect it as the external storage through the USB. Does somebody have an idea how to fix that? | 05:17 |
rrq | you have "adb" installed? | 05:17 |
rrq | on the "my system" | 05:18 |
FilipZ1 | I don't | 05:18 |
rrq | ok then it's complicated... | 05:18 |
FilipZ1 | Then how can I do it? | 05:19 |
rrq | I guess "my system" doesn't have any networking ability atm | 05:20 |
rrq | any physcal sdcard transfer option? | 05:21 |
rwp | I am assuming you plugged your phone into your "my system"? Android dropped support for USB Mass Storage years ago. | 05:23 |
rwp | You have to use MTP and an MTP transfer client now. | 05:23 |
rrq | gnarface: you have the -audiodev argument setup? | 05:35 |
rrq | maybe -audidev alsa,id=default | 05:36 |
gnarface | yea, i had added it and tried a few variations despite that i didn't think it was necessary last time | 05:36 |
gnarface | still no joy | 05:36 |
gnarface | last time i just had this: -device intel-hda -device hda-duplex | 05:37 |
gnarface | don't remember how well i tested it though | 05:37 |
gnarface | now i have this: | 05:37 |
gnarface | -audiodev alsa,id=snd0,in.mixing-engine=off,out.mixing-engine=off,in.fixed-settings=off,out.fixed-settings=off | 05:37 |
gnarface | -device intel-hda -device hda-duplex,audiodev=snd0 | 05:37 |
gnarface | (also had tried it without all the mixing turned off, no luck) | 05:37 |
gnarface | speaker-test says something about a broken pipe | 05:38 |
rrq | maybe needs -soundhw something as well? | 05:38 |
gnarface | hmm, i had not tried -soundhw even though i saw it mentioned. i thought it was deprecated or something because it's no longer in the manpage | 05:39 |
gnarface | yea, my version says "-soundhw: invalid option" | 05:39 |
gnarface | i'm doing this on ceres | 05:39 |
gnarface | rrq: what do you know about possible -machine options for this? there's an example at the top of the man page that goes: "qemu-system-x86_64 some.img -audiodev <backend>,id=<name> -machine pcspk-audiodev=<name>" but neither pcspk-audiodev nor anything else that looks obviously audio related shows up in the man page section for -machine nor in the output of "qemu-system-x86_64 -machine help" | 05:48 |
gnarface | nowhere else online has any example for this mentioned using -machine in conjunction with audio though | 05:48 |
gnarface | or at all for that matter | 05:49 |
rrq | I would guess that shows how to confgure the emulated PC speaker to go through a certain host backend | 05:52 |
rrq | (I'm running chimaera) | 05:53 |
rrq | use maybe "-soundhw all" (I think it failed due to the ":") | 05:54 |
rrq | or "-soundhw hda" is fine | 05:55 |
rrq | ah use '-device intel-hda -device hda-duplex' instead | 05:55 |
rrq | that would be to establish an emulated sound card, and then you'll need "-audiodev blaha" to declare the host side | 05:57 |
rrq | long time since I used sound in qemu | 05:59 |
rrq | but the host end typically plays to the name "default" doesn't it? | 06:00 |
rrq | maybe that's the out.device setting for -audiodev | 06:00 |
rrq | qemu is happy with eg "-audiodev alsa,id=snd0 -device intel-hda -device hda-duplex,audiodev=snd0" | 06:09 |
rrq | or eg "-audiodev alsa,id=snd0 -device AC97,audiodev=snd0" | 06:11 |
FilipZ1 | I am trying to pass the output of strace to a file, but it just doesn't write to it, instead only printing it in the terminal. | 06:26 |
rrq | use strace -f ifup wlan0 >& LOG ... then eventually ctrl-C on that | 06:27 |
rrq | strace dumps to stderr | 06:27 |
FilipZ1 | Thanks! It works now. | 06:28 |
rrq | gnarface: seems like "-audiodev alsa,id=snd0 -device AC97,audiodev=snd0" is good .. I just trialled it with chimaera live iso | 06:32 |
rrq | (on a chimaera host) | 06:33 |
FilipZ | I tried to paste it in https://paste.debian.net , but I got "Length of code is not allowed to exceed 150kB", so I tried to open that file, to copy a part of it into one file, and the second into the other, but there was something wrong with its encoding, and its size increased several times with this. How should I do it properly? :) | 06:47 |
FilipZ | Also, a big part of characters from the output looks unrecognizable to me. | 06:51 |
rrq | use: split -l 1000 logfile XX | 06:52 |
rrq | gives you a succession of XXnnn files with 1000 lines in | 06:52 |
rrq | maybe 1000 is small | 06:52 |
rrq | or I can set up a different "publishing place" if you like | 06:54 |
FilipZ | Most of these printed-out characters are displayed as the question marks on the background of some black form, with use of the encoding accessible to me. Why would it be like this? | 06:55 |
FilipZ | the encodings accessible to me* | 06:55 |
rrq | hmmm shouldn't be "most" | 06:57 |
rrq | first line should be like: "execve("/sbin/ifup", ["ifup", "wlan0"], ..." | 06:57 |
FilipZ | "split -l 1000 logfile XX" only creates a single file for me. | 07:02 |
rrq | could you do "hexdump -C logfile | head -n 2" and copy those 2 lines? | 07:04 |
rrq | mmm without cut&paste might be hardish | 07:05 |
FilipZ | "00000000 77 70 61 5f 73 75 70 70 6c 69 63 61 6e 74 3a 20 |wpa_supplicant: | | 07:05 |
FilipZ | 00000010 2f 73 62 69 6e 2f 77 70 61 5f 73 75 70 70 6c 69 |/sbin/wpa_suppli| | 07:05 |
FilipZ | " | 07:05 |
rrq | ok.. so you straced wpa_supplicant rather than ifup .. fair enough .. how large is the file? | 07:06 |
rrq | use: wc logfile | 07:07 |
FilipZ | 189.2 KiB | 07:07 |
rrq | ok... so "grep openat logfile > XX" wouldn't be too large to paste | 07:08 |
FilipZ | I did "sudo strace -f ifup wlan0". I don't get why it straced wpa_supplicant instead of ifup. | 07:08 |
rrq | yes seems peculiar, would be safer with "sudo -- strace -f ifup wlan0" to kae sure that sude doesn't eat the "-f" | 07:10 |
rrq | kae=make | 07:10 |
FilipZ | I get: grep filelocation binary file matches , and the file created is single, and 0 byte in size | 07:11 |
rrq | ok, "grep -a openat logfile > XX" | 07:11 |
FilipZ | That was from it | 07:12 |
FilipZ | Oh, right | 07:12 |
FilipZ | That's a single file, 260 bytes long now. | 07:13 |
FilipZ | openat(AT_FDCWD, "/run/network/.ifstate.lock", O_RDWR|O_CREAT|O_APPEND|O_CLOEXEC, 0666) = 4 | 07:13 |
FilipZ | openat(AT_FDCWD, "/run/network/ifstate", O_RDWR|O_CREAT|O_APPEND|O_CLOEXEC, 0666) = 5 | 07:13 |
FilipZ | openat(AT_FDCWD, "/run/network/.ifstate.tmp", O_WRONLY|O_CREAT|O_TRUNC, 0666) = 6 | 07:13 |
rrq | hmm something funny has been happening with that logging | 07:14 |
rrq | could try again into a new file to see if that then is "the same"? | 07:20 |
rrq | "strace" I mean | 07:20 |
rrq | I'll have to play in real life a bit; some hours... hope you'll have it sorted when I'm back :) | 07:30 |
FilipZ | I am afraid it wouldn't be helpful for now, as I restarted the system and it is not in the same problem state. Or could it help still? | 07:31 |
FilipZ | It will probably happen again after some time, though. | 07:32 |
joerg | maybe useful? >> -o filename Write the trace output to the file filename rather than to stderr. filename.pid form is used if -ff option is supplied. << | 07:46 |
FilipZ | I tried earlier to send those messages, but they didn't pass through: | 08:51 |
FilipZ | After I did "sudo -- strace -f ifup wlan0", when the connection broke again, it didn't finished with the error, but was printing output all the time, until I have stopped it after a couple of minutes. It wasn't the same problem state though, but the lighter issue, where I can still fix it by doing ifdown, ifup on the wlan0 interface. | 08:52 |
FilipZ | I don't know why but I have the dhclient proces with the exact same arguments turned on twice. | 08:52 |
FilipZ | It is not how it should be, right? | 08:52 |
FilipZ | Beside that, the bigger issue, with that previous problem state appeared later again, and I did "sudo -- strace -f ifup wlan0" of it. | 08:55 |
FilipZ1 | Later I will also try the configuration recommended by rwp . Could it help in such case? | 09:08 |
unixbsd | hello | 12:07 |
unixbsd | where to find hearder to 5.10 7 amd64? | 12:07 |
djph | unixbsd: huh? | 12:07 |
unixbsd | termbin.com/tux9 Linux DEVUAN 5.10.0-7-amd64 | 12:09 |
unixbsd | i need the headers deb | 12:09 |
unixbsd | http://security.debian.org/pool/updates/main/l/linux/ <--- not there | 12:09 |
unixbsd | i need 5.10 7 common and amd64 for that 5.10.0-7-amd64 from the devuan website. | 12:10 |
djph | what version of devuan are you running? | 12:10 |
unixbsd | devuan kernel from teh live cdrom | 12:10 |
djph | ... so .. chimaera? | 12:11 |
unixbsd | yeah | 12:11 |
unixbsd | where to find those debs�? | 12:12 |
djph | you do know how to use apt, right? | 12:14 |
unixbsd | sure | 12:14 |
unixbsd | not there | 12:14 |
djph | 5.10.0-16 is certainly in the chimaera repos | 12:15 |
unixbsd | i need 5.10.0-7 | 12:16 |
djph | afaik, that's not part of chimaera | 12:16 |
unixbsd | it is on the cdrom for live or instalaltion | 12:17 |
djph | least per packages.devuan.org the lowest available is 5.10.0-12 | 12:17 |
djph | and yes, the live-image is going to be behind the curve | 12:17 |
unixbsd | well i didnt dream it is from chimaera installation. | 12:17 |
rrq | seems to be something here: https://software.pureos.net/package/bin/landing/linux-headers-5.10.0-7-common | 12:18 |
djph | that's how live-images tend to work; they're not respun for every kernel patch | 12:18 |
djph | what are you really trying to do here, unixbsd ? | 12:18 |
unixbsd | i need my deb | 12:18 |
unixbsd | onl.y two. common and amd64 | 12:19 |
unixbsd | links termbin.com/dxru <--- this si the /boot | 12:19 |
djph | unixbsd: *sigh* look, the live-image is ... weir ... what are you trying to do that you think you need those headers? | 12:19 |
unixbsd | because it works fine with my DELL ... other sometimes reboots the PC weirdly unexpeectilddely. | 12:20 |
djph | ... "kernel headers" won't change that at all. | 12:20 |
djph | they're for compiling software ... | 12:20 |
unixbsd | on it i need virtualbox so i need headers | 12:21 |
djph | .. you can't run vbox on the live image ... | 12:22 |
unixbsd | it is a custom live | 12:22 |
djph | *sigh* good luck then. | 12:23 |
unixbsd | inddeed | 12:23 |
unixbsd | lets look if into the devuan_chimaera_4.0.2_amd64_desktop-live.iso i can find those those two debs | 12:24 |
djph | ... or since you're customizing it ... why not just use a newer kernel ...? | 12:25 |
thegoat | unixbsd, the headers from 5.10.0-7 and 5.10.0-16 will be virtually the same, so whatever you need to do with the 7 version should be doable with the 16 version | 12:25 |
djph | I mean, your attempt to run vbox on a live session is still quite likely to be disappointing ... | 12:25 |
thegoat | 16 is just the more up-to-date with bug fies | 12:25 |
thegoat | fixes* | 12:25 |
Horrific-Helenah | djph: Upgrading to a newer kernel most the time doesn't provide any benefits. | 12:27 |
Horrific-Helenah | Hell, ISPs still stick 2.4 on routers and they do just fine. | 12:27 |
djph | Horrific-Helenah: yeah well, he's complaining he can't find something older than what's in the repos for his custom-respin of the live iso (so... just go with what's readily available...) | 12:28 |
unixbsd | ok then i apt-get the 16 !! | 12:29 |
unixbsd | thanks | 12:29 |
unixbsd | i will remake a live ... | 12:30 |
Horrific-Helenah | djph: Oh | 12:47 |
Horrific-Helenah | unixbsd: So why do you need to remaster Devuan because of that problem? | 12:47 |
Horrific-Helenah | And depending on what you expect of a remaster/new distro, it could be a damn nightmare to maintain. | 12:48 |
Horrific-Helenah | Also I have just turned up so I apologise if I'm not understanding things. | 12:48 |
unixbsd | djph: nope 16 not working | 12:50 |
unixbsd | but 7 works fine | 12:50 |
unixbsd | pc is a dell | 12:50 |
unixbsd | 8 core | 12:51 |
Horrific-Helenah | Okay so which software is this? I'm trying to see if being on version 7 could be a unnecessary workaround rather than a solution to a problem you're having with version 13. | 12:51 |
Horrific-Helenah | You want virtualbox? | 12:52 |
unixbsd | me yes | 12:52 |
Horrific-Helenah | Have you tried other things too say qemu? Also are you trying to run virtualbox in a live system? That's going to be tediously slow and you won't like it! | 12:52 |
Horrific-Helenah | PC brand means nothing, it doesn't help us, it doesn't help you to find a solution nor is the number of CPU cores as VMs can run on single core hosts. What would more help us is say your BIOS settings, for example does your CPU support VT-x or the AMD equivalent and have you enabled it and things like this? | 12:56 |
unixbsd | on my dell computer, the kernel 7 worsk better than 16 visibly. | 12:57 |
djph | suuuure it does. | 12:59 |
djph | I mean, barring something between -8 and -16 disabling spectre/meltdown/speculative execution bugs that got some big news a while back ... | 13:00 |
unixbsd | my cpu is a intel inside core i7vpro | 13:02 |
Horrific-Helenah | Wait... you're using older kernels?! You probably haven't even thought about updating the headers... | 13:02 |
Horrific-Helenah | Also, why are you downgrading kernel versions? That could run you into some vulnerabilities you're unaware of, you should check advisory boards before you even think about doing that. | 13:02 |
unixbsd | well, if it stops other way is to vanilla. but it is too long to do. | 13:03 |
Horrific-Helenah | I think you're trying to work around issues without realising it rather than resolve them with a solution, I think if I remember, there is a module called "vbox", have you checked it's available in your kernel? Have you checked that you have updated your linux headers? Are you compiling kernels? | 13:04 |
Horrific-Helenah | Is the module loaded? | 13:04 |
Horrific-Helenah | djph: Did unixbsd even post any logs? | 13:04 |
Horrific-Helenah | I don't see any. | 13:04 |
djph | Horrific-Helenah: honestly, it was something about needing a specific header because apparently headers make the system more stable ... and then ohwaitno i need headers for this live CD so I can run vbox | 13:05 |
Horrific-Helenah | unixbsd: May I see some virtualbox logs please? | 13:06 |
unixbsd | i try again... maybe my media. | 13:08 |
Horrific-Helenah | Have you heard of the XY problem? Please read this https://xyproblem.info/ It's basically where you ask for X but really you want Y and we don't know you want Y because you asked for X so please try producing X. | 13:08 |
Horrific-Helenah | *so please try producing Y. | 13:08 |
unixbsd | ok, it seems to work better. i changed media. 16 seems to work... ufff!! i try vbox now | 13:09 |
unixbsd | holy shit | 13:10 |
unixbsd | it crashed | 13:10 |
unixbsd | i cannot install vbox on chimaera | 13:10 |
Horrific-Helenah | And that's another communication issue is that I asked for logs but then you've just gone about it your own way and so people are wasting their time with you because we've put energy into asking you questions and you are not answering them. I'm sorry to be harsh but this looks to people like you are rejecting their support and as a result people would rather help someone else than you. | 13:11 |
unixbsd | i took of course vbox 11 bullseye | 13:11 |
Horrific-Helenah | So you say that it doesn't install on chimaera but this is a communication issue because you didn't state why, you didn't produce a log. | 13:12 |
Horrific-Helenah | We can't help you unless you help us. | 13:12 |
djph | unixbsd: you cannot install vbox _ON A LIVE SESSION_ | 13:13 |
unixbsd | i give up | 13:13 |
Horrific-Helenah | djph: Ah | 13:13 |
Horrific-Helenah | Well there's your answer. May I ask why you want to spin up VMs on a live system? | 13:13 |
unixbsd | anyhow try to run vbox on chimera on a DELL pc that would be intresting. on mine it seems that vbox dpkg -i does not work. someohw. | 13:13 |
djph | Horrific-Helenah: I mean, I could entirely be missing key info here (like he installed it) | 13:13 |
Horrific-Helenah | djph: I'm missing a lot of information so your point is valid. | 13:14 |
djph | unixbsd: yes or no, you are trying this from a live-image of some sort? | 13:14 |
Horrific-Helenah | unixbsd: Forget about your computer branding, it's just a label. | 13:14 |
djph | they're probably all made in the same factory these days anyway :D | 13:16 |
Horrific-Helenah | Yep and not only that, most hardware bugs would be to do with things like chipset which companies such as Intel and AMD produce. | 13:18 |
unixbsd | i will make you a dmesg | 13:18 |
Horrific-Helenah | unixbsd: That's a start but we need as many relevant logs as you can find. | 13:18 |
Horrific-Helenah | As for dpkg, you just need to give us its error output. | 13:19 |
djph | Horrific-Helenah: I think starting with confirmation that "yes, really this is a live session" is kinda a good starting point ... | 13:19 |
Horrific-Helenah | Also why are you using dpkg for Devuan packages? | 13:19 |
Horrific-Helenah | That could put your system in a mess... | 13:19 |
Horrific-Helenah | Also what djph said. | 13:19 |
Horrific-Helenah | I keep forgetting you are using a live session but you will just be wasting your time and our time, it's not going to work, you simply will never get vbox to run in a live session. | 13:21 |
unixbsd | anyhow we must use the bulleye vbox deb file, that is maybe not clean | 13:21 |
Horrific-Helenah | And even if you do, it's going to eat your RAM because everything, your distro, your software, vbox is loading fully into RAM. | 13:22 |
Horrific-Helenah | unixbsd: That is unclean indeed. It's hacky. | 13:22 |
djph | Horrific-Helenah: it has to be a chatbot. | 13:22 |
Horrific-Helenah | Oh dear | 13:22 |
* Horrific-Helenah stops talking | 13:22 | |
djph | Horrific-Helenah: it's the only thing that makes sense anymore | 13:23 |
Horrific-Helenah | djph: I'm going to help people who actually want help and not people who go around in circles. | 13:23 |
unixbsd | dmesg log... here www.termbin.com/3g33 <- vbox crashes on it, with bbox bulleye virtualbox-6.1_6.1.38-153438_Debian_bullseye_amd64.deb | 13:24 |
unixbsd | the pc turns off and black while dpkg -i this deb | 13:24 |
djph | unixbsd: yes or no, this is a live session? | 13:24 |
Horrific-Helenah | Are you sure that's the right vboxdrv for your kernel? I see "module verification failed: signature and/or required key missing - tainting kernel" | 13:25 |
Horrific-Helenah | That's a mismatch. | 13:25 |
unixbsd | ah it is installed seems like | 13:25 |
Horrific-Helenah | That's irrelevant, I can see it's installed. | 13:25 |
unixbsd | but xset mayeb. .. i dont know. realy. i messed up my installation. | 13:25 |
Horrific-Helenah | This is how modules work, any changes to the kernel can break them so the modules have to be kept up to date with the kernel or they won't work. You can't just take a kernel and try loading modules for a different kernel version. | 13:26 |
unixbsd | www.termbin.com/3g33 seems anyhow better with a kernel 7 rather than 16. | 13:29 |
Horrific-Helenah | unixbsd: It's the same one you posted earlier... | 13:29 |
Horrific-Helenah | unixbsd: Please read this: https://xyproblem.info/ | 13:30 |
Horrific-Helenah | In full. | 13:30 |
Horrific-Helenah | It will help you in life. | 13:30 |
Horrific-Helenah | unixbsd: I'm going to stop talking to you until you read it in full and your questions about problems show you have. | 13:32 |
Horrific-Helenah | One example is, you are still going for Y but djph told you Y won't work and that you need to do X but you refuse to do X and keep resorting to Y, Y in this case is running vbox in a live session, djph stated X which is what you want for vbox to work which is to run it in an installed session. | 13:33 |
unixbsd | anyhow if you find the 5.10 7 amd64 deb files, please think about me. | 13:37 |
Horrific-Helenah | unixbsd: Did you read it? | 13:37 |
unixbsd | not yet., i wil. | 13:38 |
fsmithred | oracle vbox is the only choice in chimaera. I've got it installed here. | 13:43 |
fsmithred | ii virtualbox-6.1 6.1.34-150636.1~Debian~bullseye amd64 Oracle VM VirtualBox | 13:44 |
fsmithred | 5.10.0-18-amd64 | 13:45 |
onefang | Only coice for what? I have qemu installed on my Chimaera desktop. | 13:45 |
Horrific-Helenah | What onefang said. | 13:45 |
fsmithred | only choice for virtualbox. It's not in the repo | 13:45 |
onefang | Ah. | 13:46 |
fsmithred | yes, qemu is there and I use that, too. | 13:46 |
Horrific-Helenah | qemu would be better in a live session than virtualbox anyway, qemu is a lot lighter, virtualbox for example has a LOT of GUI baggage to drag around. | 13:47 |
fsmithred | I'm using vbox because it's easier to do bridged network, but it is s l o w compared to qemu. | 13:47 |
Horrific-Helenah | fsmithred: Well what that is, is you are not used to it. It's easy for me but I understand. :) | 13:48 |
onefang | I have qemu bridged network on my server, no problems. | 13:48 |
unixbsd | indeed | 13:48 |
unixbsd | same here | 13:48 |
unixbsd | bridge is hard | 13:48 |
unixbsd | vbox works with networking very well | 13:48 |
Horrific-Helenah | Again, it's that you are not used to it, bridged networking becomes easy to do once you understand what it actually does and how it's too be configured and why it has to be configured that way. | 13:49 |
fsmithred | well, maybe when I get back home we can go over my tap/tun configs and you can tell me how to do it better. If I run two VMs, they get the same address if they are set for dhcp | 13:49 |
Horrific-Helenah | fsmithred: We would be happy to help. :) | 13:49 |
Horrific-Helenah | Well someone will be. | 13:50 |
fsmithred | you gonna be around later this week?> | 13:50 |
Horrific-Helenah | Well I'm away on the 7th. | 13:50 |
onefang | I suspect the two VMs are not being set to use different MACs. Off the top of my head. | 13:51 |
Horrific-Helenah | unixbsd: Did you read about the XY problem? | 13:51 |
fsmithred | good call, onefang. I did not set mac addresses. | 13:51 |
Horrific-Helenah | You most likely don't have to set it but best to rule things out. :) | 13:52 |
onefang | But I'm using fixed IPs and a single VM on my bridged server. So never tried your use case. | 13:52 |
fsmithred | if I set static IP in the guest, it works. But I run a lot of live-isos and those are all set for dhcp. | 13:53 |
onefang | From my notes "QEMU uses 52:54:00:12:34:56 for all VMs" unless you tell it otherwise. | 13:53 |
fsmithred | checking... | 13:54 |
Horrific-Helenah | Oh, I forgot about that, onefang | 13:55 |
Horrific-Helenah | Yes, that's bound to run into issues. :D | 13:55 |
fsmithred | confirmed | 13:55 |
fsmithred | so if I have separate static ip addresses set up for separate tap devices in /etc/network/interfaces, I just need to add some mac addresses there? Or does it have to be in the qemu command? | 13:56 |
onefang | In the qemu comand. | 13:57 |
Necrodiver | since everyones talking about qemu, any ideas on how to create a vm for winblows 95 in qemu, if its even possible? | 13:57 |
fsmithred | I guess you would need iso files of the installer disk(s) | 13:57 |
onefang | I've never tried that. I have a Windows 8.1 VM. | 13:57 |
fsmithred | I have win xp that I can run in vbox or qemu | 13:58 |
Necrodiver | i have this cd that was designed for 95/98 that i want to play but wine wont let me use it even when i go through the trouble to set up the win95/98 stuff for it | 13:58 |
onefang | And yes, I did install it from the installer disk. Then moved it to real hardware with dd, with no issues, and recently moved it back to a VM. lol | 13:58 |
fsmithred | I don't recall how I made that VM, as it was over 10 years ago | 13:58 |
Necrodiver | oof | 13:59 |
Horrific-Helenah | Necrodiver: It's even possible to run things a lot older than that such as System V UNIX | 13:59 |
Necrodiver | in qemu? | 13:59 |
Horrific-Helenah | Necrodiver: Yes | 14:00 |
Necrodiver | maybe the commands i used for qemu werent right maybe? | 14:00 |
Horrific-Helenah | Well it's about the hardware you're setting the VM to have for example Windows 95 isn't going to like the VirtIO stuff unless there are drivers for it on the VirtIO disc. | 14:01 |
Necrodiver | i dunno i tried to follow some tutorials on youtube for some easy stuff like peppermintOS devuan edition and ran into various errors along the way so i just gave up | 14:01 |
fsmithred | that should work. | 14:01 |
Horrific-Helenah | and also the amount of RAM you assign, Windows 95, I forgot what its maximum RAM it can support is, but there is a ceiling. | 14:01 |
Necrodiver | yeah i think it was like 512 mb | 14:01 |
Horrific-Helenah | That sounds like it. | 14:02 |
Necrodiver | i could get peppermintOS to run but no sound | 14:02 |
Necrodiver | and couldnt figure out what i was doing wrong | 14:02 |
fsmithred | that would be a qemu command issue | 14:02 |
onefang | Qemu does sometimes deprecate command line options, so you sometimes have to track down "How do I do that in my versions, which is different from the version i found described on the web?". | 14:02 |
Necrodiver | yeah figured | 14:02 |
Necrodiver | best way is porbably to just look at the tutorials/faqs on the qemu site im guessing? | 14:03 |
fsmithred | and experience | 14:03 |
Necrodiver | when i get some more time ill have to mess with it | 14:03 |
Necrodiver | starting a new job tomorrow so | 14:03 |
onefang | And keep notes in your qemu scripts, so you don't forget how you fixed it last time. | 14:03 |
Necrodiver | thats something i need to learn to do, is make qemu scripts | 14:04 |
fsmithred | example: qemu-system-x86_64: warning: '-soundhw hda' is deprecated, please use '-device intel-hda -device hda-duplex' instead | 14:04 |
Necrodiver | i was doing everything through commands | 14:04 |
fsmithred | I use the deprecated option because the recommended one doesn't work | 14:04 |
fsmithred | I learned scripting so I wouldn't have to remember commands | 14:05 |
onefang | I have learned 100 languages in my programming career. I learned scripting coz I needed to write some scripts. B-) | 14:06 |
Necrodiver | i think because im still on the newer side of linux, i prefer to use the commands just so i can at least try to understand how they function | 14:07 |
fsmithred | good idea | 14:08 |
onefang | Well shell scripts are just a list of commands with "#!/bin/sh" as the first line. So you are most of the way there. B-) | 14:08 |
Necrodiver | lol | 14:09 |
Necrodiver | im weird i like the hard way for some reason | 14:10 |
Necrodiver | i do have a odt file for copy pasting commands i need on the fly | 14:10 |
Horrific-Helenah | That's not weird, personal choice. :) | 14:13 |
rrq | when it comes to qemu networking, I prefer the vde option as it especially lets me run the qemu boxes non-root | 14:18 |
FilipZ | rrq: could we continue now? I have the new logs from the strace, as this issue happened again. | 14:35 |
rrq | ok. it's late here, so let's sort it quickly ;) | 14:36 |
rrq | could you also first paste your /etc/network/interfaces somewhere? | 14:37 |
FilipZ | hexdump -C logfile | head -n 2 output from that new logs is: | 14:39 |
FilipZ | 00000000 77 70 61 5f 73 75 70 70 6c 69 63 61 6e 74 3a 20 |wpa_supplicant: | | 14:39 |
FilipZ | 00000010 2f 73 62 69 6e 2f 77 70 61 5f 73 75 70 70 6c 69 |/sbin/wpa_suppli| | 14:39 |
rrq | that's still very very strange since strace starts with "ifup" | 14:40 |
FilipZ | rrq: I can, but could you set up this publishing place, you mentioned before? It would be easier with the bigger file | 14:40 |
rrq | the interfaces file isn't large | 14:40 |
FilipZ | I did it with "sudo -- strace -f ifup wlan0" like you recommended before. I did it also when it wasn't crashing like this yet, but that file is very large, because it printed it for many minutes, untill I stopped it, as there were no error messages, apparently. | 14:42 |
rrq | the log you could upload wht "nc -w 2 108.61.184.72 12000 < logfile" | 14:43 |
rrq | then it's available at https://borta.rrq.id.au/download/logfile | 14:44 |
FilipZ | Interfaces file content: https://paste.debian.net/1255937/ | 14:44 |
rrq | do you have any files in /etc/network/interfaces.d/ ? | 14:45 |
FilipZ | I do not | 14:45 |
rrq | ok.. then the head of /etc/wpa_supplicant/wpa_supplicant.conf ... but note that that file contains your wireless access credentials | 14:46 |
rrq | you might not want to publish those | 14:46 |
rrq | I just want to sight the initial lines, before any "network" block | 14:47 |
FilipZ | https://paste.debian.net/1255939/ here. I changed the credentials here | 14:49 |
rrq | ok so that looks all perfect | 14:50 |
rrq | the backlog said something about multipl dhclient .. is that still so? use: pgrep -a dhclient | 14:52 |
FilipZ | "the log you could upload wht "nc -w 2 108.61.184.72 12000 < logfile"" I don't get that part. It contains a specific IP, or at least it looks like this to me, and with this exact content(logfile swapped for the actual file location ofc) it doesn't print anything | 14:52 |
FilipZ | rrq: please check the http://reisenweber.net/irclogs/libera/_devuan/_devuan.2022-10-04.log.html chat logs for this. I wrote about it earlier. I did notice in the taskmanager that I had 2 dhclient processes. | 14:54 |
FilipZ | I still have. | 14:55 |
FilipZ | But when I did ifdown, or when it crashed, one was left. | 14:55 |
rrq | ok; do ifdown and kill the one that's left | 14:56 |
FilipZ | It didn't dissapear, but when I did ifup wlan0 again, the third one appeared. :) | 14:58 |
FilipZ | With a higher pid. | 14:58 |
FilipZ | They are of different sizes, somehow. The new one have the highest size. The one with the lowest PID have the lowest size. | 14:59 |
rrq | ok; do "ifdown --force wlan0" then "pkill -9 dhclient" then ifup wlan0 | 14:59 |
rrq | the pkill command will kill all of them | 14:59 |
FilipZ | I did so. When I did ifup wlan0 2 of them appeared, and there was some other third process, but with a little different name, then it and one of the dhclient processes disappeared, and now there is just one. | 15:03 |
FilipZ | I think that it is even bigger than the previous 3. | 15:04 |
FilipZ | Uses up more memory, I meant. | 15:04 |
rrq | hmm could you paste output of "pstree" somewhere? it's a command from the "psmisc" package | 15:05 |
rrq | it shows running processes with parent-child relationship | 15:08 |
FilipZ | here: https://paste.debian.net/1255942/ | 15:08 |
FilipZ | Sorry. That was without sudo | 15:09 |
rrq | that's fine | 15:09 |
FilipZ | Beside this, what did you mean by "then it's available at https://borta.rrq.id.au/download/logfile"? At https://borta.rrq.id.au/ I got 502 Bad Gateway, and when I entered that exact url I only got an empty file. | 15:12 |
rrq | hmm https://borta.rrq.id.au/download/logfile is a file of 193418 bytes | 15:13 |
FilipZ | Oh, now I got this. I thought there was some error earlier. | 15:14 |
rrq | probably strace should have the "-o logfile" argument, to avoid gett its log mixed up with the program's stderr | 15:16 |
FilipZ | There was now, that I check this again. It was a 0 byte size file earlier. | 15:16 |
rrq | right; there is always some caching involved... but it's there now ... | 15:16 |
FilipZ | "nc -w 2 108.61.184.72 12000 < logfile" output with this is: (UNKNOWN) [108.61.184.72] 12000 (?) : Connection refused | 15:17 |
FilipZ | Did you asked me about that? | 15:17 |
rrq | no, sorry... my nc finished ... I meant that you need to rerun strace with "-o logfile" rather then the ">&logfile" bit | 15:19 |
rrq | (it looks like "sudo" makes the stderr logging messed up too much) | 15:20 |
onefang | Don't stay up past your bed time rrq. | 15:22 |
FilipZ | I make use strace with -o argument next, but it needs to be done just when that same problem appears again, right? | 15:22 |
rrq | yes; so ifup wlan0 worked fine after having killed all dhclient? | 15:23 |
FilipZ | After you asked me earlier? Yes, if it didn't, I would need to restart the system. | 15:24 |
FilipZ | I mean, if that error occurred again. | 15:25 |
rrq | next time you might try first with the same procedure (ifdown + kill any dhclient + ifup) | 15:28 |
rrq | that doesn;t explain why it breaks but might at least be a recovery path | 15:29 |
FilipZ | Alright. Then I will do this, and report its output with the -o option here. | 15:29 |
rrq | ok ... g'night | 15:30 |
onefang | G'night. | 15:30 |
FilipZ | good night :) | 15:34 |
gnarface | rrq: AC97 mode seems to be working fine, dunno what's wrong with the other one | 16:10 |
gnarface | thanks for the idea | 16:11 |
onefang | Think he's gone to sleep, it's after 1 AM in his time zone. | 16:11 |
gnarface | (i had picked intel-hda because that was what the motherboard actually has, odd that it wouldn't work) | 16:11 |
eyalroz | Hello Devuaners, | 17:10 |
eyalroz | when I apt-get dist-upgrade'd my daedalus today, I was told: | 17:10 |
eyalroz | dpkg: pipewire-media-session: dependency problems, but removing anyway as you requested: | 17:10 |
eyalroz | gnome-remote-desktop depends on pipewire-media-session | wireplumber; however: | 17:10 |
eyalroz | Package pipewire-media-session is to be removed. | 17:10 |
eyalroz | Package wireplumber is not installed. | 17:10 |
eyalroz | I think I want the remote-desktop app, and I did not ask for the media-session package to be removed, so not sure what's going on. | 17:10 |
gnarface | well there's a problem with the packages probably. it would have warned you about packages it's gonna remove though, you just gotta actually read it. | 17:13 |
gnarface | sometimes it's easy to miss if there are a lot of packages, it lists the removals in the middle | 17:14 |
eyalroz | Ok, just giving a heads-up in case the problem is not merely on my system | 17:28 |
eyalroz | Oh, it looks like libc was updated and g++ wad version-bumped... | 17:29 |
onefang | Also daedulus isn't stable yet, so you might get this sort of oddness every now and then. Try again the next day. | 17:33 |
onefang | The thing I find odd about that is that upstream Debian is switching from pulesaudio to pipewire for that version. Maybe they are half way in doing that? | 17:40 |
onefang | Then again, way past my bed time. G'night. | 17:41 |
rwp | eyalroz, Regarding running Testing, See the "Best practices for Testing users" here https://wiki.debian.org/DebianTesting | 18:15 |
FilipZ1 | rrq: after doing "sudo -- strace -f ifup wlan0 -o logfile" like you recommended, I got some errors that seem to tell, that it for some reason cannot write to the file, or doesn't detect it, even if present. | 23:38 |
FilipZ1 | Whats wrong with it, then? | 23:39 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!