armin | hmmmmmmmmm | 05:54 |
---|---|---|
armin | just noticed that #devuan-social doesn't exist | 05:54 |
armin | is it cool to post links to heavy metal videos on youtube in here? | 05:54 |
furrywolf | #debianfork is the social channel | 05:55 |
debdog | armin: there's #debianfork for OT stuff | 05:55 |
armin | debdog: a bit un-intuitive, but hey | 05:55 |
furrywolf | you're welcome to post heavy metal videos there, and I'll probably click many of them. :) | 05:55 |
debdog | I don't think whether it's intuitive or not is not the issue here. but it ain't mentioned in the topic | 06:01 |
debdog | hmm, prolly one negation too much there | 06:02 |
DocScrutinizer05 | [support] TOPICLEN=390 | 06:11 |
debdog | oh | 06:12 |
DocScrutinizer05 | !ot | 06:19 |
infobot | OT: please take any off-topic chat to #debianfork | 06:19 |
DocScrutinizer05 | quite funny when people ping you in /query but miss your reply thanks to their /ignore list X-P | 06:20 |
DocScrutinizer05 | debdog: you might revoke your request since some folks here neither notice my posts nor those of infobot, thus still might ponder how to "fix this issue" | 06:27 |
rsevero | Hi, I have devuan 2.0 installed in an encrypited LVM (LVM + LUKS) default installation. I'm having trouble on shutdown as, at the last shutdown step, just before "System will be halted now"message (which is show exactly before actually turning off the computer, the shutdown process halts with several messages "sda5_crypt still busy" and this keeps on going for something like 30 seconds. How can I fix this delay? | 12:39 |
rsevero | Or at least, how can I further troubleshoot it? | 12:40 |
gnarface | check the forum | 12:44 |
gnarface | i don't know the answer but i think it's a known issue... | 12:44 |
rsevero | gnarface: Ok. Thanks. | 12:46 |
rsevero | gnarface: My shutdown delay problem with LVM + LUKS was fixed manually applying the patch available at https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=792552#5 | 13:19 |
rsevero | Should I report this anywhere else? Open a bug at devuan? | 13:19 |
gnarface | if there's not one already open, it can't hurt | 13:22 |
gnarface | typically upstream patches their version and devuan inherits it though, i think | 13:23 |
rsevero | In fact I found one but apparently no action has been taken on it: https://bugs.devuan.org//cgi/bugreport.cgi?bug=237 | 13:39 |
rsevero | BTW, I can't find how to comment on a Devuan bug with the browser, only with an email. Is it the only way? | 13:39 |
furrymcgee | yes, I think it is the way to go, thank you for the bug report | 13:44 |
rsevero | Another annoying issue: with Devuan the HDMI secondary monitor connection doesn't work on my Asus laptop. With a live Ubuntu pen drive it works just fine. When I connect the HDMI with Devuan, I see no second monitor in the "Display" settings. This laptop has a dual Intel/NVidia video board. Suggestions? | 13:53 |
furrymcgee | Whats your hardware kernel and drivers? Does it work with debian? | 14:01 |
gnarface | rsevero: ubuntu includes the nvidia proprietary drivers and bumblebee by default, i guess. | 14:15 |
gnarface | devuan does not | 14:15 |
gnarface | but they're in the repos | 14:15 |
gnarface | you should be able to install, it's worked for many people. it's kinda hit and miss though, that hardware is a p.i.t.a. | 14:16 |
gnarface | many people have got it working, but this is another thing i can't help with specifically | 14:17 |
gnarface | you might be able to get help with the nouveau driver it's using by default in #nouveau though | 14:18 |
rsevero | furrymcgee: Here is lspci output: https://paste.ee/p/DvcKn | 14:18 |
rsevero | furrymcgee: I will prepare a live Debian pendrive to answer that. | 14:19 |
gnarface | also something called DRI_PRIME with nouveau might be a better experience than bumblebee in some cases, worth looking into | 14:19 |
gnarface | it depends a lot on the hardware | 14:19 |
rsevero | gnarface: ubuntu incluides nvidia`s proprietary drives but they where not used. And in Devuan I already tried nvidia proprietary drivers. With them I couldn't even get X up. | 14:20 |
gnarface | yikes | 14:20 |
rsevero | gnarface: I will take a look at DRI_PRIME. Thanks! | 14:20 |
gnarface | there are newer versions of the kernel and the drivers in ascii-backports, which might help some of your issues (ubuntu uses newer versions of everything by default too - normally this causes more bugs but i understand how that might not seem this way right now to you) | 14:22 |
rsevero | gnarface: I already updated my kernel:Linux version 4.18.0-3-amd64 (debian-kernel@lists.debian.org) (gcc version 7.3.0 (Debian 7.3.0-30)) #1 SMP Debian 4.18.20-2 (2018-11-23) | 14:25 |
debdog | rsevero: this might be a good start: http://us.download.nvidia.com/XFree86/Linux-x86_64/390.42/README/index.html esp. chapter 17 | 14:25 |
rsevero | It didn't help. | 14:25 |
rsevero | debdog: Will look at it. Thanks! | 14:25 |
rsevero | gnarface: I really want to get rid of systemd ;) Just have to find out how to fix this issue. | 14:27 |
buZz | 'devuan 2.0' is ascii, right? | 14:37 |
buZz | maybe just grabbing cryptsetup from backports would solve it | 14:37 |
rsevero | buZz: I will try it and report back. | 14:40 |
rsevero | buZz: It didn't work. With the ceres cryptsetup package I got the old shutdown delay back. And apparently things changed a lot as the is no /lib/cryptsetup/cryptdisks.functions file to be patched. | 14:57 |
buZz | not ceres, the backports one | 15:00 |
buZz | (is what i recommended) | 15:00 |
buZz | but ok | 15:00 |
gnarface | it would be a mistake, yea to use a mix of ceres and ascii packages | 15:00 |
rsevero | buZz: Sorry, will try. | 15:08 |
rsevero | gnarface: Ok. Right now I have a ceres kernel in a ascii system. Bad idea also? | 15:09 |
gnarface | yes, you should actually use the backports repo | 15:10 |
gnarface | and the devuan one, not the debian one | 15:10 |
buZz | rsevero: the name of ascii's backports repo is 'ascii-backports' | 15:11 |
buZz | like ; deb http://nl.deb.devuan.org/merged ascii-backports main non-free contrib | 15:11 |
rsevero | gnarface: Ok. | 15:11 |
rsevero | buZz: Great. | 15:11 |
rsevero | buZz: Even with this new source, I still get ascii packages and not ascii-backports ones. Yes, I've run "apt update" after changing sources. What am I missing? | 15:14 |
gnarface | "-t ascii-backports" on the command-line | 15:14 |
buZz | -changing- ? no you -add- it to sources.list | 15:14 |
buZz | apt search -t ascii-backports cryptsetup | 15:15 |
rsevero | buZz: Sorry for the bad description. I added this new source. | 15:15 |
buZz | alright | 15:15 |
rsevero | Running kernel 4.18 and cryptsetup from ascii-backports. The shutdown delay came back. But now I have a /lib/cryptsetup/cryptdisks.functions file to patch ;) | 15:27 |
KatolaZ | rsevero: this is a known bug IIRC | 15:28 |
KatolaZ | do you have an encrypted / ? | 15:28 |
rsevero | KatolaZ: Yes I know: https://bugs.devuan.org//cgi/bugreport.cgi?bug=237 Unfortunatelly the patch available in this bug reported hasn't got into available packages yet. | 15:29 |
rsevero | KatolaZ: Yes, everything encrypted except /boot | 15:30 |
KatolaZ | rsevero: the "bug" is that the shutdown script waits for all the crypted volumes to be unmounted | 15:30 |
KatolaZ | but / cannot be unmouted yet at that point | 15:30 |
rsevero | KatolaZ: With the patch available at the bug report, there is no hanging at the end of the shutdown process. | 15:31 |
KatolaZ | yeah I know | 15:31 |
rsevero | KatolaZ: Do you believe there is any unwanted side effect with the proposed patch? It seems to work just fine for me. | 15:32 |
KatolaZ | rsevero: it should be just fine | 15:35 |
rsevero | KatolaZ: So I believe it should be included in the available cryptsetup packages. It's really annoying to wait 30 extra seconds for something that can happen immediately ;) | 15:36 |
KatolaZ | rsevero: that would mean forking cryptsetup... | 15:41 |
KatolaZ | but yeah, we should probably do that | 15:41 |
rsevero | KatolaZ: Does the /lib/cryptsetup/cryptdisks.functions file come from upstream? | 15:44 |
KatolaZ | rsevero: yes | 15:44 |
KatolaZ | AFAIK | 15:45 |
KatolaZ | yes, it does | 15:45 |
KatolaZ | it would be probably better to push the patch upstream, I guess | 15:45 |
rsevero | KatolaZ: I've just downloaded cryptsetup sources from https://gitlab.com/cryptsetup/cryptsetup/tree/master and I can't find the cryptdisks.functions in it. | 15:49 |
rsevero | Is there a way to check on devuan's package where the file cames from? | 15:49 |
buZz | dpkg -L cryptsetup , iirc | 15:50 |
rsevero | buZz: This command lists the files from the cryptsetup package. What I would like to knwo is: where did a specific file that is in the package cama from: from the upstream sources? From the packager himself? | 15:55 |
rsevero | What I would like to know is: where did a specific file that is in the package came from: from the upstream sources? From the packager himself? | 15:56 |
buZz | packager, most probably | 15:58 |
furrymcgee | you can download the debian package sources with apt source cryptsetup and compare with upstream | 15:59 |
furrymcgee | it needs a deb-src in sources.list | 15:59 |
rsevero | furrymcgee: Thanks! | 16:04 |
KatolaZ | rsevero: no need | 16:04 |
KatolaZ | look for the "Filename:" header | 16:04 |
KatolaZ | apt-cache show PACKAGE | grep "^Filename: " | cut -d "/" -f 2 | 16:05 |
KatolaZ | rsevero: ^^^ | 16:05 |
rsevero | KatolaZ: The package maintainer is the Debian Cryptsetup Team. I will try furrymcgee suggestion to discover where does the cryptdisks.functions file comes from. | 16:08 |
KatolaZ | rsevero: the package comes from Debian | 16:08 |
KatolaZ | I told you already | 16:08 |
KatolaZ | we haven't forked it | 16:09 |
KatolaZ | :) | 16:09 |
KatolaZ | and the Filename: header has a trace of that | 16:09 |
KatolaZ | if you are to file a bug, you should probably file it against the "cryptsetup" package | 16:09 |
KatolaZ | at bugs.debian.org | 16:10 |
rsevero | KatolaZ: I know, but some files came from upstream (cryptsetup) and others might come from the Debian maintainers. I have to know that so I open the bug on the right place: Debian OR cryptsetup. | 16:10 |
KatolaZ | rsevero: just open it on bugs.debian.org | 16:10 |
KatolaZ | if it's an upstream bug, it will be forwarded upstream | 16:10 |
rsevero | Ok. | 16:10 |
KatolaZ | the maintainers would know | 16:11 |
KatolaZ | (if you are willing to help them, even better I guess) | 16:11 |
KatolaZ | (meaning, helping with pointing to the problem) | 16:11 |
rsevero | KatolaZ: I have just read much of https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=792552 | 16:17 |
rsevero | KatolaZ: I believe cryptsetup must be forked by Devuan as it seems that the solution is systmed X SysV specific. | 16:19 |
KatolaZ | rsevero: it is not clear from the bug you posted though... | 16:24 |
KatolaZ | it refers to https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=791944 | 16:25 |
KatolaZ | but it is not clear how that is relevant | 16:26 |
KatolaZ | (it's a long thread...) | 16:26 |
rsevero | KatolaZ: Do you believe that there is a common solution, not systemd or SysV specific, for this issue? One that would make sense for Debian to apply? Both solutions that work for me (https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=792552#5 and https://bugs.devuan.org//cgi/bugreport.cgi?bug=237) make changes in a file that, AFAIU, is relevant only for SysV. | 16:28 |
KatolaZ | rsevero: if it's relevant for sysv and does not break anything in systemd, the Debian maintainers might be able to accept the patch? | 16:49 |
KatolaZ | (I am just guessing) | 16:49 |
KatolaZ | I don't know hows LUKS is managed by systemd | 16:49 |
rsevero | KatolaZ: I'm kind of lost here. How to I report a bug that I see on Devuan when doing a shutdown managed by SysV scripts to Debian that uses systemd? How do I give any details about this issue without installing Debian with LVM + LUKS? | 16:57 |
rsevero | How do I give and relevant details (from Debian maintainers perspective) about this issue? | 16:59 |
KatolaZ | rsevero: Debian ships systemd, but all packagers still have the obligation (by policy) to ship simple init scripts for their services | 17:02 |
KatolaZ | if the code is only used by the sysvinit script, any Debian user with sysvinit woudl experience the same peoblem right? | 17:02 |
rsevero | KatolaZ: So I should open a bug at Debian saying something like: "When shuting down Devuan I have this issue...". Is that it? | 17:03 |
rsevero | KatolaZ: Are there Debian users that use sysinit? I thought that this wasn't an option with Debian anymore. | 17:03 |
KatolaZ | rsevero: ? | 17:04 |
KatolaZ | sure there are | 17:04 |
KatolaZ | it's a bit cumbersome to use | 17:04 |
KatolaZ | you can't choose it at install time | 17:04 |
KatolaZ | and most of the stuff which depends on systemd does not work | 17:04 |
muep | it has been an option all the timem just not a default setup | 17:04 |
KatolaZ | but *in principle* you can still use sysvinit in Debian | 17:04 |
KatolaZ | (at least so far) | 17:05 |
muep | basically the reverse of the situation in wheezy, where install-time you would always get sysvinit but could install systemd from the repos if that's preferred | 17:05 |
KatolaZ | well, with the addition that now a lot more stuff depends on systemd | 17:06 |
KatolaZ | (see session management) | 17:06 |
KatolaZ | and simply does not work if you don't have systemd | 17:06 |
KatolaZ | or if you don't have a replacement | 17:06 |
KatolaZ | (like elogind) | 17:06 |
rsevero | KatolaZ: If you say so. I will open a bug in Debian talking about my issue with Devuan. I really don't think Debian maintainers will consider it something for them to solve but let's see. | 17:06 |
KatolaZ | rsevero: it's worth trying I guess | 17:07 |
KatolaZ | maybe also post the patch | 17:07 |
KatolaZ | and/or reference the Devuan bug | 17:07 |
KatolaZ | if they don't want/cannot act on it | 17:07 |
KatolaZ | we could fork the package | 17:08 |
muep | to get attention of the debian maintainer, it would be way better to reproduce it with actual debian | 17:09 |
rsevero | muep: I bet so but I'm really not in the mood to install a whole Debian with LVM + LUKS just to report this bug... | 17:17 |
rsevero | To Debian | 17:17 |
rsevero | KatolaZ: Email sent to submit@bugs.devuan.org Let's see... | 17:22 |
rsevero | Oh, dammed. It went to devuan. I thought reportbug would send it to Debian... | 17:22 |
KatolaZ | rsevero: np | 17:23 |
KatolaZ | rsevero: it looks like the script has changed in unstable | 17:27 |
ibanja | I am receiving a "WARNING: Failed to connect to lvmetad." on boot, followed by "Volume group 'desktop' not found"... The problem is it keeps looping this message. a bunch of times before moving on and booting. I beleive the problem is I don't have a volume named 'desktop'. Where is it getting this from? | 22:21 |
ibanja | I just replaced a failing hd, by restoring from fsarchives onto the new LVM partitioned drive. I renamed the VG (volume group) from "desktop" to "vg_desktop" so there wouldn't be a conflict when I hook up my dying hd to retrieve a few non-backed up files. | 22:22 |
ibanja | so it's apparently a remnant from the previous hd, but where is the information stored, so I can correct this? | 22:23 |
ibanja | the fstab is correct | 22:23 |
ibanja | syslog entry (presumably when boot sequence resumes): lvmetad[2345]: Failed to accept connection errno 11. | 22:38 |
rsevero | Maybe someone can help me with this video issue: with Devuan the HDMI secondary monitor connection doesn't work on my Asus laptop. With a live Ubuntu pen drive it works just fine. When I connect the HDMI with Devuan, I see no second monitor in the "Display" settings. This laptop has a dual Intel/NVidia video board. I have kernek 4.18.0 from backports. Just tested in Debian 9.6.0 live and it doesn't work there also. I have no xorg.conf file at all. Does | 22:57 |
rsevero | anyboy have a clue why it works just fine with a Live Xubuntu but not in Devuan? | 22:57 |
golinux | rsevero: Devuan does not install proprietary video drivers by default. They should be in non-free | 23:02 |
rsevero | golinux: Forgot to mention, I already tried the proprietary drivers. No change. And I checked Ubuntu: it works with the open source drivers. | 23:03 |
golinux | Is there an xorg.conf in Ubuntu? | 23:08 |
azarus | If one can still install Debian without systemd: https://wiki.debian.org/systemd#Installing_without_systemd , what's the point of Devuan? | 23:12 |
azarus | (i tried those instructions, they seem to work OK) | 23:13 |
rsevero | golinux: No, even in the Live ISO it works fine. Without a xorg.conf file. | 23:14 |
rsevero | azarus: Those instructions are for Jessie (the old release). AFAIK they don't work anymore with Stretch. | 23:15 |
rsevero | The current Debian release. | 23:16 |
azarus | I believe I tried with stretch | 23:16 |
azarus | But that was a while ago, so maybe you're right | 23:16 |
rsevero | azarus: The point I think is "what you call: it worked". AFAIK, several things that depend on systemd in Stretch will stop working or not work quite well. | 23:17 |
azarus | I suppose so, yep | 23:17 |
azarus | It did feel pretty hacky with systemd remnants all over | 23:18 |
rsevero | azarus: I think these are the minor issues. The bigger ones are actual features not working like, I think, even graphic login. | 23:19 |
molio | Hey | 23:20 |
azarus | Don't really need that, so \_(ツ)_/ | 23:20 |
rsevero | :) | 23:20 |
molio | Anybody here knows a way to get the package "libxfce4panel-1.0"? It's a requirement for building dockbarx, a Xfce plugin. | 23:21 |
golinux | <azarus> . . . you asked: If one can still install Debian without systemd: https://wiki.debian.org/systemd#Installing_without_systemd , what's the point of Devuan? | 23:36 |
azarus | yep | 23:36 |
golinux | Let me answer. Because there are applications that will not work if systemd is not installed | 23:36 |
azarus | I guess to to compile flags and the such | 23:37 |
golinux | So byt not using systemd your freedom is denied. | 23:37 |
azarus | I actually don't use systemd on a daily basis, nor do I use linux on a daily basis | 23:37 |
golinux | Yes. And that's what Devuan does so that everything works without systemd | 23:37 |
azarus | There are plenty of other systemd-free distros that I have tried and run for a while | 23:38 |
azarus | Gentoo and Alpine Linux being the primary ones | 23:38 |
golinux | Then maybe you shouldn;t be offering advice on the Devuan help channel. | 23:38 |
azarus | ... me, offering advice? When? | 23:38 |
golinux | If one can still install Debian without systemd: https://wiki.debian.org/systemd#Installing_without_systemd , what's the point of Devuan? | 23:39 |
azarus | I see that as a question, and not advice. | 23:39 |
golinux | That indicates thet debian w/o systemd is equal to devuan and that is not true. | 23:40 |
azarus | It was a legitimate question, since the last time I tried, it had worked. I never told anyone to do as I did. | 23:40 |
golinux | It is built on the lie that those two are equal so spreading wrong information about devuan | 23:40 |
azarus | A "lie"? I simply asked a question. | 23:41 |
golinux | That depends on your definition of "work". The noose is tightening with each new release. | 23:41 |
azarus | You are interpreting my legitimate question as 100 other things. | 23:41 |
golinux | Debian was even going to drop support for sysvinit. | 23:41 |
azarus | Yep. It felt pretty hacky. | 23:42 |
golinux | Your question was based on a false premise. | 23:42 |
azarus | (to use sysvinit on modern debian9 | 23:42 |
azarus | debian) | 23:42 |
azarus | * | 23:42 |
golinux | Because you don't know much about devuan, it seems | 23:42 |
azarus | Correct! That's why I asked. | 23:42 |
azarus | Is that so wrong? | 23:43 |
golinux | I read that as you were encouraging rsevero to just use debian. | 23:44 |
golinux | to solve his problem. | 23:44 |
azarus | Well then you interpreted my question as not a question. | 23:44 |
golinux | I often pose suggestions as a question | 23:45 |
azarus | I joined at 23:12 CET, the first thing I read rsevero say was "golinux: No, even in the Live ISO it works fine. Without a xorg.conf file." | 23:45 |
azarus | I did not know, nor do I now know of rsevero's issue. | 23:45 |
golinux | OK. | 23:46 |
golinux | Wonder if he's still here . . . | 23:46 |
golinux | rsevero: ^^^ | 23:46 |
molio | Nevermind, got it. | 23:59 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!