Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Already on GitHub? Sign in to your account

systemctl status dlm.service failure (2016-11-27 build) #42

Closed
markd89 opened this Issue Nov 27, 2016 · 23 comments

Comments

Projects
None yet
5 participants

markd89 commented Nov 27, 2016 edited

I uninstalled then reinstalled with the new script.

FYI: There was an error during the script ln: failed to create symbolic link ‘/lib/modules/3.16.0-4-amd64/build/Kconfig’: because the link already existed. The install seemed to complete OK, but I re-removed the link manually, uninstalled and reinstalled and didn't see the error the second time.

After rebooting, checked the service and could not get it to start. Here's what I did:

root@ux305:/home/mark# systemctl status dlm.service
● dlm.service - DisplayLink Manager Service
Loaded: loaded (/lib/systemd/system/dlm.service; enabled)
Active: activating (auto-restart) (Result: exit-code) since Sun 2016-11-27 15:01:32 PST; 330ms ago
Process: 6199 ExecStart=/opt/displaylink/DisplayLinkManager (code=exited, status=1/FAILURE)
Process: 6197 ExecStartPre=/sbin/modprobe evdi (code=exited, status=0/SUCCESS)
Main PID: 6199 (code=exited, status=1/FAILURE)

Nov 27 15:01:32 ux305 systemd[1]: Unit dlm.service entered failed state.
root@ux305:/home/mark# systemctl start dlm.service
root@ux305:/home/mark# systemctl status dlm.service
● dlm.service - DisplayLink Manager Service
Loaded: loaded (/lib/systemd/system/dlm.service; enabled)
Active: activating (auto-restart) (Result: exit-code) since Sun 2016-11-27 15:01:53 PST; 3s ago
Process: 6218 ExecStart=/opt/displaylink/DisplayLinkManager (code=exited, status=1/FAILURE)
Process: 6216 ExecStartPre=/sbin/modprobe evdi (code=exited, status=0/SUCCESS)
Main PID: 6218 (code=exited, status=1/FAILURE)

Nov 27 15:01:53 ux305 systemd[1]: Unit dlm.service entered failed state.

dmesg:
[ 213.740710] usb 2-1: new SuperSpeed USB device number 4 using xhci_hcd
[ 213.756999] usb 2-1: New USB device found, idVendor=17e9, idProduct=ff03
[ 213.757003] usb 2-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[ 213.757005] usb 2-1: Product: MB168B+
[ 213.757006] usb 2-1: Manufacturer: DisplayLink
[ 213.757007] usb 2-1: SerialNumber: E7LMTF092982
[ 264.229608] usb 2-1: USB disconnect, device number 4
[ 275.858429] usb 2-1: new SuperSpeed USB device number 5 using xhci_hcd
[ 275.875154] usb 2-1: New USB device found, idVendor=17e9, idProduct=ff03
[ 275.875163] usb 2-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[ 275.875167] usb 2-1: Product: MB168B+
[ 275.875171] usb 2-1: Manufacturer: DisplayLink
[ 275.875174] usb 2-1: SerialNumber: E7LMTF092982

Contributor

redjamesg commented Dec 2, 2016

This is related to #43

markd89 commented Dec 3, 2016

@adnanhodzic - would it be possible to have access to an older version of your script which downloads an older version of the Displaylink stuff? Ideally, we'd want to use the latest, but if that's difficult maybe running the older stuff until you have time to dig into the newer stuff more would be a workaround?

I appreciate all of your work on this :-)

Thanks,
Mark

Owner

AdnanHodzic commented Dec 4, 2016 edited

@markd89 @redjamesg with upcoming Stretch freeze and hence new stable release just around the corner. I really don't have time to spend on debugging issues on old kernel.

So as a workaround, instead of using old code (unsupported by both me and upstream) could you please do following. Install 4.7 kernel from Jessie backports:

1. Add to sources.list:

deb http://ftp.debian.org/debian jessie-backports main

2. Update sources and install kernel 4.7

sudo apt-get update && sudo apt-get install -t jessie-backports linux-image-4.7.0-0.bpo.1-amd64

After this simply run (latest) displaylink-debian install/re-install.

I just did this on on Debian Jessie 8.6 and it worked like a charm.

markd89 commented Dec 8, 2016

Thanks Adnan. I can't upgrade the kernel right now because it will break something else (VMWare Player Unity Mode), which I need for work. It's possible for me to hack on that and probably get it working, then I could upgrade the kernel... or I could hack on one of the older revs of this project and see if I can make it work with the current kernel. Probably I'll do the former as like you said - a newer kernel is coming soon anyway to mainstream support. Thanks again for all your help and creating the project.

Owner

AdnanHodzic commented Dec 8, 2016

@markd89 most welcome!

@AdnanHodzic AdnanHodzic closed this Dec 8, 2016

saljut7 commented Dec 9, 2016 edited

@adnanhodzic thanks alot I followed your workaround installing linux-image-4.7.0-0.bpo.1-amd64. Running now on Linux T60 4.7.0-0.bpo.1-amd64 #1 SMP Debian 4.7.8-1~bpo8+1 (2016-10-19) x86_64 GNU/Linux.
Unfortunately it didn't change anything for me. I couldn't activate dlm.service until I applied libstdc++.so.6.0.22 workaround in addition to kernel upgrade.

Just to confirm my steps: when dlm.service was finally active I went on with

$ xrandr --listproviders
Providers: number : 2
Provider 0: id: 0x77 cap: 0xf, Source Output, Sink Output, Source Offload, Sink Offload crtcs: 2 outputs: 3 associated providers: 0 name:radeon
Provider 1: id: 0x41 cap: 0x2, Sink Output crtcs: 1 outputs: 1 associated providers: 0 name:modesetting

and

$ xrandr --setprovideroutputsource 1 0

Using arandr I could activate DVI-1-1 (my DisplayLink adapter) and could arrange my 3 display (1 with DisplayLink, one Notebook display and one VGA attached ot Notebook). The command line setup in my case is:

#!/bin/sh
xrandr --output VGA-0 --mode 1280x1024 --pos 2680x0 --rotate normal --output LVDS --mode 1400x1050 --pos 1280x0 --rotate normal --output DVI-1-1 --mode 1280x1024 --pos 0x0 --rotate normal --output DVI-0 --off

But if I execute/apply these settings my system "crashes" back to login screen as reportet in systemctl status dlm.service failure #37 by markd89. Did I miss a certain step with the suggested workarounds?

markd89 commented Dec 10, 2016

fyi: That's the same behavior I had with the older kernel when using the workaround...

Owner

AdnanHodzic commented Dec 10, 2016

@markd89 @saljut7 in latest commit I made "libstdc++.so.6 hack" mentioned in #37 part of the installer.

However since packages in Jessie are now so old, you'll still have to install "libstdc++.so.6" from Stretch repo in order for everything to work.

@saljut7 following that logic ^, I'd suggest you also upgrade "xserver-xorg" packages to Stretch. I know this helped in one of the issues in past (couldn't find it for reference)

Problem you're facing is that you're using old release with already outdated packages. I have personally never used pure "stable" on my desktop environments. I've always used "Apt-Pinning" to have certain packages from different releases without putting my stability in question.

I wrote a blog post on this topic which you might find useful: "Taking control over Debian and its package repositories"

Other option would just to simply upgrade to Stretch.

saljut7 commented Dec 10, 2016

thx @adnanhodzic

I just ask because upgrading the kernel was not without problems (forgot that I had VirtualBox installed) do you think that there is a chance that upgrading xserver-xorg from 1:7.7+7 to 1:7.7+18 will help?

# apt-cache policy xserver-xorg
xserver-xorg:
  Installed: 1:7.7+7
  Candidate: 1:7.7+7
  Version table:
     1:7.7+18 0
        750 http://ftp.de.debian.org/debian/ testing/main amd64 Packages
         50 http://ftp.de.debian.org/debian/ unstable/main amd64 Packages
 *** 1:7.7+7 0
        995 http://ftp.de.debian.org/debian/ jessie/main amd64 Packages
        995 http://ftp.de.debian.org/debian/ stable/main amd64 Packages
        100 /var/lib/dpkg/status
Owner

AdnanHodzic commented Dec 10, 2016

xserver-xorg is only a transitional package and doesn't provide the driver itself, I was referring to "xserver-xorg-core" and "xserver-xorg-video-*" (whichever you're using).

saljut7 commented Dec 10, 2016 edited

@adnanhodzic thanks alot for clarifying. UPgrading with apt install -t jessie-backports xserver-xorg-video-intel=2:2.99.917+git20161105-1~bpo8+1 was no problem but for apt install -t testing xserver-xorg-core=2:1.18.4-2 I expect a repo mix-and-match hit parade:

 # apt install -t testing xserver-xorg-core=2:1.18.4-2
Reading package lists... Done
Building dependency tree       
Reading state information... Done
Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:

The following packages have unmet dependencies:
 xserver-xorg-core : Depends: xserver-common (>= 2:1.18.4-2) but 2:1.16.4-1 is to be installed
                     Depends: libdbus-1-3 (>= 1.9.14) but 1.8.20-0+deb8u1 is to be installed
                     Depends: libgcrypt20 (>= 1.7.0) but 1.6.3-2+deb8u2 is to be installed
                     Recommends: libpam-systemd but it is not going to be installed
E: Unable to correct problems, you have held broken packages.

So I think it would be better to upgrade completely to testing. Thanks alot for your article http://foolcontrol.org/?p=1708. I'm already using apt-pinning but Imy experience with mixing repos is "so-so". Would you recommend me a special "upgrading to Debian stesing" just for the case you know a good (How to upgrade Debian stable (Wheezy) to testing (Jessie)? or Update Debian 8.3 Jessie "Stable" to "Testing" or "Unstable" for more up-to-date edition look to easy :) ).

Owner

AdnanHodzic commented Dec 10, 2016

It looks easy because it is easy :)

As long as you have correct sources, it's easy as running apt-get dist-upgrade. But if you're doing this for the first time, please follow: https://www.linkedin.com/pulse/update-debian-83-jessie-stable-testing-unstable-more-edition-bagyono

saljut7 commented Dec 11, 2016 edited

@adnanhodzic but is the procedure described in https://www.linkedin.com/pulse/update-debian-83-jessie-stable-testing-unstable-more-edition-bagyono really the same with configured apt-pinning?

I just want to confirm because https://www.debian.org/releases/testing/amd64/release-notes/ch-upgrading.en.html says...

4.2.2. Disabling APT pinning

If you have configured APT to install certain packages from a distribution other than stable (e.g. from testing), you may have to change your APT pinning configuration (stored in /etc/apt/preferences and /etc/apt/preferences.d/) to allow the upgrade of packages to the versions in the new stable release. Further information on APT pinning can be found in apt_preferences(5).

...which is vague because headling says "disabling" but explanation text just "you may have to change Apt pinning configuration".

@AdnanHodzic AdnanHodzic reopened this Dec 13, 2016

Owner

AdnanHodzic commented Dec 13, 2016 edited

Sorry, but now you're asking my help with administering a Debian box, which is definitely out of this projects scope. But since I'm in a good mood, I'll help you out :)

If you're going to upgrade to Stretch, unless you want to use packages from Sid/Experimental, then you don't need pinning anymore and with dist-upgrade it can get you into trouble.

So just do this:

  1. disable pinning
    sed -i 's/^\([^#]\)/#\1 /etc/apt/preferences

  2. change every occurrence of jessie with stretch
    sed -e "s/jessie/stretch/g" /etc/apt/sources.list

2.5. make sure you do the same with any other file under /etc/apt/sources.d/

  1. update package list and upgrade
    sudo apt-get update && sudo apt-get dist-upgrade

That's it, you're running on Stretch.

Happy hacking, and google is your friend :)

saljut7 commented Dec 15, 2016 edited

@adnanhodzic at first I'm very sorry for bothering you with my last question. In fact it is not part of this issue tracker to help with dist upgrades (but to be honest: Google is not my friend and I try to avoid any kind of (not only) Google surveillance services as much as possible - unfortunately it means sometimes I don't benefit from very good Google search results: I tried to find some upgrading-to-Debian-testing-with-configured-apt-pinning-tutorial, of course). So thank you even more for providing helping steps (I promise to help others in same way).

Anywaya at first I tried upgrading on an identical machine running with Debian8 (just wihtout apt-pinning as on my main machine). Uprading to testing was fine but results with debian-displaylink were worse than before:

Reinstalling (because I''ve tested displaylink-debian on this Notebook before) there was an error report:

Performing post install steps

ln: failed to create symbolic link '/opt/displaylink/libstdc++.so.6': No such file or directory
sed: can't read /lib/systemd/system/dlm.service: No such file or directory
Failed to enable unit: File dlm.service: No such file or directory

After uninstalling, rebooting, installing (this time with an error which seems not to be critical)...

Preparing for install

Archive:  DisplayLink_Ubuntu_1.2.65.zip
  inflating: 1.2.65/displaylink-driver-1.2.65.run  
  inflating: 1.2.65/LICENSE          
Creating directory displaylink-driver-1.2.65
Verifying archive integrity... All good.
Uncompressing DisplayLink Linux Driver 1.2.65  100%  
ln: failed to create symbolic link '/lib/modules/4.8.0-1-amd64/build/Kconfig': File exists

...and rebooting dlm.service was running at startup finally. But then:

$ xrandr --listproviders
Providers: number : 2
Provider 0: id: 0x44 cap: 0x0 crtcs: 2 outputs: 3 associated providers: 0 name:ATI Mobility Radeon X1400 @ pci:0000:01:00.0
Provider 1: id: 0x2c3 cap: 0x2, Sink Output crtcs: 1 outputs: 1 associated providers: 0 name:modesetting

$ xrandr --setprovideroutputsource 1 0
X Error of failed request:  BadValue (integer parameter out of range for operation)
  Major opcode of failed request:  140 (RANDR)
  Minor opcode of failed request:  35 (RRSetProviderOutputSource)
  Value in failed request:  0x44
  Serial number of failed request:  16
  Current serial number in output stream:  17

BadValue (integer parameter out of range for operation) was not reported in context of displaylink-debian, was it?
rotation is not the problem in this case because arandr@testing does not even list the attached display as under Debian Stable.

Owner

AdnanHodzic commented Feb 18, 2017 edited

Marking this issue as obsolete and having it closed.

Please update to latest version of displaylink-debian tool, and please re-open this or create a new issue if the problem persists.

saljut7 commented Feb 19, 2017 edited

Hi @adnanhodzic
I've tried again on a fresh installed Debian8.7 and still get "failed state" error:

User@Debian8.7x64:~$ sudo systemctl start dlm.service
User@Debian8.7x64:~$ sudo systemctl status dlm.service
● dlm.service - DisplayLink Manager Service
   Loaded: loaded (/lib/systemd/system/dlm.service; enabled)
   Active: activating (auto-restart) (Result: exit-code) since Sun 2017-02-19 11:28:54 CET; 4s ago
  Process: 1368 ExecStart=/opt/displaylink/DisplayLinkManager (code=exited, status=1/FAILURE)
  Process: 1365 ExecStartPre=/bin/sh -c modprobe evdi || (dkms install evdi/1.3.52 && modprobe evdi) (code=exited, status=0/SUCCESS)
 Main PID: 1368 (code=exited, status=1/FAILURE)

Feb 19 11:28:54 Debian8.7x64 systemd[1]: Unit dlm.service entered failed state.
User@Debian8.7x64:~$ sudo systemctl start dlm.service
User@Debian8.7x64:~$ sudo systemctl status dlm.service
● dlm.service - DisplayLink Manager Service
   Loaded: loaded (/lib/systemd/system/dlm.service; enabled)
   Active: activating (auto-restart) (Result: exit-code) since Sun 2017-02-19 11:29:46 CET; 1s ago
  Process: 1503 ExecStart=/opt/displaylink/DisplayLinkManager (code=exited, status=1/FAILURE)
  Process: 1501 ExecStartPre=/bin/sh -c modprobe evdi || (dkms install evdi/1.3.52 && modprobe evdi) (code=exited, status=0/SUCCESS)
 Main PID: 1503 (code=exited, status=1/FAILURE)

User@Debian8.7x64:~$ sudo journalctl -xn
-- Logs begin at Sun 2017-02-19 11:27:36 CET, end at Sun 2017-02-19 11:31:37 CET. --
Feb 19 11:31:35 Debian8.7x64 mtp-probe[1765]: checking bus 5, device 6: "/sys/devices/pci0000:00/0000:00:1d.7/usb5/5-6/5-6.1"
Feb 19 11:31:35 Debian8.7x64 mtp-probe[1765]: bus: 5, device: 6 was not an MTP device
Feb 19 11:31:36 Debian8.7x64 sudo[1669]: pam_unix(sudo:session): session closed for user root
Feb 19 11:31:36 Debian8.7x64 systemd[1]: Starting DisplayLink Manager Service...
-- Subject: Unit dlm.service has begun with start-up
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
-- 
-- Unit dlm.service has begun starting up.
Feb 19 11:31:36 Debian8.7x64 systemd[1]: Started DisplayLink Manager Service.
-- Subject: Unit dlm.service has finished start-up
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
-- 
-- Unit dlm.service has finished starting up.
-- 
-- The start-up result is done.
Feb 19 11:31:36 Debian8.7x64 DisplayLinkManager[1788]: /opt/displaylink/DisplayLinkManager: ./libstdc++.so.6: version `GLIBCXX_3.4.21' not foun
Feb 19 11:31:36 Debian8.7x64 systemd[1]: dlm.service: main process exited, code=exited, status=1/FAILURE
Feb 19 11:31:36 Debian8.7x64 systemd[1]: Unit dlm.service entered failed state.
Feb 19 11:31:37 Debian8.7x64 sudo[1794]: User : TTY=pts/0 ; PWD=/home/User ; USER=root ; COMMAND=/bin/journalctl -xn
Feb 19 11:31:37 Debian8.7x64 sudo[1794]: pam_unix(sudo:session): session opened for user root by User(uid=0)

Well if it is still necessary to upgrade the kernel I would suggest to update the readme related to...

Regardless of which kernel version you're using.

...and better say which kernel version is needed.

Anyway thanks alot for this great script!

Owner

AdnanHodzic commented Feb 20, 2017

@saljut7 resolution to your problem has been explained as part of Troubleshooting most common issues

saljut7 commented Feb 26, 2017 edited

@adnanhodzic thanks a lot. I followed the instruction but ended up like last time with a crashing desktop environment after enabling the arandr configuration. Upgrading the kernel /xserver didn't help the last time. Maybe this is a similar problem to Crashing when starting X (so in the end a problem with my external monitor which is nothing special but a quite old 4:3 BENQ TFT display)?

@AdnanHodzic AdnanHodzic reopened this Feb 26, 2017

when i run

systemctl status dlm

I get the error:

Process: 3689 ExecStartPre=/bin/sh -c modprobe evdi || (dkms install evdi/1.3.52 && modprobe evdi) (code=exited, status=1/FAILURE)

It was a combo of config for intel cards that @gitowiec suggested as well as uefi settings in bios. but I got it working! Could you put this info in the post installation guide?

Contributor

redjamesg commented May 29, 2017

It is already there

saljut7 commented Jun 24, 2017 edited

[edit: I think my post was not related to dlm.service so I opened a new issue here]

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment