| brocashelm | any tips on replacing rsyslogd with busybox-syslogd? | 04:21 |
|---|---|---|
| gnarface | no, does it fail or something? | 04:22 |
| gnarface | also, why? | 04:23 |
| brocashelm | since it's a part of busybox and i'd like to explore that avenue more | 04:27 |
| gnarface | are you using busybox? | 04:27 |
| systemdlete | I think busybox syslog does not support rsyslog's features | 04:28 |
| systemdlete | (but maybe that isn't important for your case, idk) | 04:28 |
| brocashelm | no, but i'd like to switch over to it eventually | 04:29 |
| systemdlete | dancing around rsyslog's shortcomings and defects and poorly-documented features, I've been able to make rsyslogd do what I want. | 04:30 |
| rustyaxe | syslog-ng <3 | 06:30 |
| systemdlete | I would make the jump to syslog-ng, but from what I've been reading online, it seems it has not had the uptake compared to rsyslogd. Not sure if that's a good reason; after all, here we are using devuan... | 06:45 |
| systemdlete | while the rest of the world has marched on into systemd blindly and obediently | 06:45 |
| systemdlete | also, syslog-ng seems to have a propriety feel to it, judging by the syslog-ng.com web page. | 06:47 |
| systemdlete | (sure, there's a link to their open source version, but still...) | 06:47 |
| systemdlete | I've looked at the docs and even experimented with syslog-ng a bit, and the architecture is--in my opinion--superior to rsyslog's. I would not deter anyone from migrating to it. | 06:49 |
| systemdlete | update: That VM that I disabled 3D on is still without a desktop freeze. HOpefully, I do not speak too soon. | 06:54 |
| systemdlete | I might be filing a bug at vbox... | 06:54 |
| systemdlete | (they do warn that it is experimental, though a user would not think that by the lack of a warning in the GUI) | 06:55 |
| user24037 | i need to replug my keyboard/mouse switching between ttys after pm-suspend. what's up with that? | 09:04 |
| user24037 | started happening with daedalus | 09:09 |
| rrq | user24037: you may need to elaborate on your use case. "switching ttys" sound like you are running multiple DEs concurrently. | 09:50 |
| rrq | or does it concern console ttys? | 09:50 |
| user24037 | you've got it i think. i mean that i run $ startx on tty1 and tty2 and switch between them. | 10:38 |
| user24037 | i3wm is the desktop | 10:38 |
| rrq | non-root user I guess? different users? elogind or seatd ? | 10:46 |
| rrq | using pm-suspend on command line? on tty1/tty2 or separate? | 10:48 |
| rrq | one major change for daedalus (from chimaera0 is that X11 now includes device access via "seat mediation" .. not sure if that plays in here | 11:18 |
| user24037 | same non-root user;; the elogind service is running;; pm-suspend from within either desktop | 11:22 |
| user24037 | thanks for the questions and help by the way. i'm not sure about where in the stack the problem lies either | 11:23 |
| rrq | perhaps at USB level the devices gain new bus addresses on resume without the driver(s) being told that the old address have become invalid... does suspend involve shifting out the xhci module(s) ? | 11:26 |
| rrq | what happens if you suspend from tty3 console ? | 11:31 |
| rrq | my thought line for that is simply to remove the X11 "stack" from the suspend/resume activity | 11:33 |
| user24037 | everything just works suspending from tty3 console (^^; | 11:36 |
| rrq | good.. at least there's a possible workaround .. but it doesn't solve it | 11:37 |
| user24037 | when it doesn't work it persists in not working; i have to replug with every change indefinitely | 11:38 |
| user24037 | ya man! good read! | 11:38 |
| rrq | I suppose the next step would be to use seatd rather than elogind for X11 input mediation, to see if the issue remains | 11:40 |
| rrq | btw did you configure /etc/elogind/logind.conf wrt sleep? | 11:42 |
| * fsmithred is testing on a spare laptop | 11:44 | |
| fsmithred | It still works after pm-suspend here. Ran startx as same user twice, tty1 and tty2. elogind is installed. No display manager. | 11:45 |
| fsmithred | what else can I do to break it? | 11:46 |
| rrq | USB kb+mouse? | 11:46 |
| fsmithred | yeah, I'll try that in a couple minutes. Need more coffee. | 11:46 |
| rrq | I think that's an important colour in this painting... because USB devices do change bus address upon resume; though I would have thought udev kicks in to do magic about that | 11:47 |
| rrq | it seems the OP effect is that the X11 in focus doesn't get told to renew its input devices | 11:48 |
| rrq | though hmm X11 should be told about the new devices and act upon that... hmm | 11:50 |
| rrq | user24037: you use udev (from eudev) I assume? and you don't have ay Xorg.conf declaring inputs? | 11:51 |
| fsmithred | having trouble testing here - usb keyboard didn't work in the laptop | 12:00 |
| fsmithred | back soon | 12:02 |
| fsmithred | back | 12:08 |
| fsmithred | after suspen, then I switched to the other display and the keyboad did not work there. The laptop trackpad and keyboard did work. | 12:09 |
| fsmithred | shit | 12:09 |
| fsmithred | after suspend, the keyboard worked | 12:10 |
| fsmithred | but only on the display that I used to suspend. | 12:10 |
| rrq | same with both elogind and seatd ? | 12:11 |
| fsmithred | I didn't try removing elogind | 12:11 |
| fsmithred | oh yeah, replug fixed it when it was dead. | 12:12 |
| rrq | it's sufficient to install seatd since it is the preferred mediation | 12:12 |
| rrq | (I think :) | 12:12 |
| fsmithred | brb | 12:14 |
| rrq | USB kb working on first X11 and not second is odder than OP's issue (not working on first and thus cannot switch to check second) | 12:18 |
| fsmithred | adding seatd seemed to fix it. I could not remove elogind without major destruction. | 12:18 |
| fsmithred | Did anything interesting happen while I was gone? | 12:19 |
| rrq | nope | 12:19 |
| rrq | ok; first indication would be that this issue hides in the Napoleon soup of elogind | 12:20 |
| fsmithred | hm, the laptop that failed already has seatd installed (and elogind) | 12:22 |
| rrq | or might possibly be due to some dbus messaging race (which the seatd option also avoids) | 12:22 |
| rrq | and it used seatd? | 12:23 |
| fsmithred | how do I know which it is using? | 12:24 |
| rrq | check Xorg.0.log ... it makes noises unless seatd is used | 12:24 |
| rrq | when opening the graphics device | 12:25 |
| rrq | (i.e. when it initializes) | 12:25 |
| fsmithred | yeah, it says it used seatd | 12:26 |
| rrq | ok. perhaps you can try with seatd stopped ? | 12:27 |
| rrq | restarting all X11 | 12:28 |
| fsmithred | I had to reboot to get the usb keyboard to work on console | 12:34 |
| rrq | (that's new) | 12:36 |
| fsmithred | ok, now the usb keyboard does not wake it up from suspend | 12:37 |
| fsmithred | but laptop keyboard does wake it up, and then usb key works on both displays | 12:38 |
| fsmithred | that's with elogind turned off | 12:38 |
| fsmithred | shit, no it's not. | 12:38 |
| rrq | :) | 12:39 |
| fsmithred | elogind-daemon is running. | 12:39 |
| fsmithred | disabling elogind in sysv-rc-conf didn't do what I wanted it to do | 12:39 |
| rrq | and seatd ? | 12:39 |
| fsmithred | oh, I turned off seatd, not elogind | 12:40 |
| fsmithred | yeah, seatd not running | 12:40 |
| fsmithred | should I reverse those? | 12:40 |
| rrq | right, so apart from wakeup the usb keyboard works fine when elogind is used.. yes try it with seatd and without elogind (though with seatd, elogind is shadowed for X11) | 12:42 |
| * rrq needd to go .. back later | 12:44 | |
| fsmithred | I'll leave notes | 12:44 |
| fsmithred | Slow logging in - takes a long time to get the prompt after giving password, also takes a long time for second desktop to come up. | 12:46 |
| fsmithred | with seatd running and elogind disabled, usb key does not wake it up from suspend. Laptop key does and then usb key works. I use usb key to switch back to first desktop and keyboard works there to type text in terminal, but then ctrl-alt-f2 does not switch back to second display. | 12:49 |
| fsmithred | Time for breakfast. | 12:49 |
| CueXXIII | you could use chvt remotely to switch between different virtual consoles and thus x sessions | 13:14 |
| rrq | fsmithred: is it a very recent daedalus install? and specifically which kernel? | 14:26 |
| user24037 | rrq: eudev is installed and my xorg.conf https://pastebin.com/Hhak7k8H | 14:32 |
| soolapimo | hello, I'm running Ceres and since yesterday I've been having problems upgrading packages due to signatures not being verified anymore. I think that gpg or something broke? I saw that I had to restart the daemon to use it again after an update | 14:33 |
| soolapimo | this is what I'm facing: https://paste.debian.net/plain/1317391 | 14:33 |
| soolapimo | Anyone knows how can I "force" this update without verification so that it gets back on track? | 14:34 |
| rrq | user24037: that xorg.conf seems "harmless" wrt this issue. I guess you should test also without that xorg.conf, but probably will not make a difference. | 14:38 |
| rrq | could you share your Xorg.0.log ? eg via https://transfer.rrq.au (or whereever) | 14:40 |
| rrq | as you maye have seen, fsmithred has got some similar USB keyboard issues... | 14:42 |
| rrq | user24037: btw how do you trigger resume? does it wake by a USB keyboard key press? | 14:43 |
| rrq | specifically how it looks directly at resume when the USB keyboard fails ... you might capture that by copying it on the same command line with pm-suspend (with semi-colon) ... and then maybe show the additional log after replugging the keyboard | 14:56 |
| rrq | .. or just the latter; the full log immediately after having replugged the keyboard | 14:58 |
| rrq | soolapimo: apt-get has --allow-unauthenticated which might help you | 15:04 |
| rrq | user24037: there's updated xserver-xorg-core in daedalus-security that might help (unless you are using that already)... (I have a vague memory that it came about due to a USB keyboard issue) | 15:08 |
| fsmithred | rrq, 6.1.0-18-686 on asus EEE installed last August. | 15:16 |
| rrq | version of xserver-xorg-core ? | 15:17 |
| fsmithred | 2:21.1.7-3+deb12u2devuan1 | 15:18 |
| fsmithred | not the one in daedalus-security | 15:18 |
| rrq | ok; could you install from daedalus-security to see if the isue remains? it's that package abd | 15:19 |
| rrq | xserver-common | 15:19 |
| fsmithred | upgrading now | 15:19 |
| fsmithred | almost half way there | 15:25 |
| CueXXIII | soolapimo: you could try apt -o Acquire::AllowInsecureRepositories=true update | 15:29 |
| fsmithred | once again, it looks like I will have to reboot to get the usb keyboard to work in console. Going through 'init 1' is not enough. | 15:33 |
| rrq | mmm that's not X11 issue | 15:35 |
| fsmithred | ok. after restarting eudev and repluging the keyboard, it works. Had to do both. | 15:37 |
| fsmithred | should I bother with x11 tests? | 15:37 |
| rrq | afair, the USB keyboard issue got resolved by commit b813afc2b, leading to package version devuan/2%21.1.8-1devuan2 .. and eventually rolled into that devuan-security version | 15:38 |
| rrq | perhaps we hear from user24037 if the issue remains with devuan-security those pacakges | 15:40 |
| fsmithred | ok, after suspend, usb keyboard works on both desktops (:0 and :1) but only after I use the laptop keyboard to wake it up. | 15:41 |
| fsmithred | that's with both elogind and seatd active | 15:41 |
| rrq | fair enough I think; wakeup is something closer to h/w while userland (and Xorg) remains suspended | 15:43 |
| * rrq bedtime | 15:43 | |
| fsmithred | sleep well | 15:44 |
| user24037 | $ startx now is busted. can't change tty afterward;; the xorg.log file contains @^ | 17:51 |
| user24037 | saw that seatd and libseat1 are running in an older log that had ASCII text | 17:51 |
| user24037 | Desktop starts but can't get input registering | 17:57 |
| fsmithred | user24037, what was the last change you made to the system? | 18:01 |
| fsmithred | rrq is sleeping now. Leave some notes for us. | 18:02 |
| fsmithred | I'm going outside | 18:02 |
| user24037 | xserver-xorg-input-libinput got removed (^^; pardon | 18:06 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!