Public bug reported:
Briefly: if one has a pair of systemd .socket/.service unit files to
start Xorg via socket activation, the first X11 client which triggers
this socket activation gets stuck (waits forever) , although Xorg starts
properly. A second X11 client will then work as expected. For mor
Hans et al.,
I'm observing GDM behaviour with non-root Xorg. I see that GDM by default
keeps the greeter session alive, switching to another VT in order to start
a new user session. Shouldn't it break for non-seat0 seats, since they
can't handle multiple sessions due to lack of VT support?
I thin
Robert,
I think he's asking about the case one installs xorg-server-
lts-{vivid,wily} on Trusty. In this case, the bug does apply (sort of)
to Trusty.
Enviado através do Boxer
Em 18/11/2015 18:11, Robert Ancell
escreveu:
14.04 LTS is currently running Xorg 1.15 so there should be no need to
br
** Branch linked: lp:~lbssousa/lightdm/guest-session-write-run-screen
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to lightdm in Ubuntu.
https://bugs.launchpad.net/bugs/1442611
Title:
Guest session can't write on /var/run/s
** Branch linked: lp:~lbssousa/lightdm/guest-session-allow-read-proc-
net-dev
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to lightdm in Ubuntu.
https://bugs.launchpad.net/bugs/1442609
Title:
Guest session can't read "/proc
OK, Robert!
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to lightdm in Ubuntu.
https://bugs.launchpad.net/bugs/1504049
Title:
apparmor rules too tight for chromium
Status in Light Display Manager:
Fix Committed
Status in
Could you please edit also the following directive in abstractions
/lightdm_chromium-browser? I cannot close Chromium remotely with
"killall chromium" from a guest session because it's currently
disallowed to receive SIGTERM.
signal (receive, send) set=("exists", "term") peer=/usr/lib/lightdm
/lig
Robert, the following snippet is used to retrieve current Xorg version
in http://bazaar.launchpad.net/~ltsp-upstream/ltsp/ldm-
trunk/view/head:/screen.d/ldm:
X -version 2>&1 | grep "X.Org X Server" | sed -n 's/X.Org X Server //p'
Maybe you can use it in a new function e.g.
x_server_local_minimal_
Robert, there may be a mistake here: "-listen tcp" was first introduced
in xorg-server 1.17.0 (vivid ships with 1.17.1), not 1.16.0.
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to lightdm in Ubuntu.
https://bugs.launchpad.net
Robert,
This is a sample of my /var/log/kern.log messages regarding xfce4
-netload-plugin attempts to read network traffic:
Apr 9 14:46:34 localhost kernel: [ 786.952187] audit: type=1400
audit(1428601594.953:805): apparmor="DENIED" operation="open"
profile="/usr/lib/lightdm/lightdm-guest-sessi
Is this related?
http://cgit.freedesktop.org/plymouth/commit/?id=84eb4381db85877a9a56b35994e6c10d43e46ebe
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to lightdm in Ubuntu.
https://bugs.launchpad.net/bugs/1371651
Title:
Da
Try to enable vesafb at boot -- append e.g. vga=771 or vga=788 to grub
command line. I think it should be better with option video=uvesafb:...,
but this one requires /sbin/v86d to be installed.
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which i
A similar problem was observed with VirtualBox at bug #1365336. I guess
the root of both problems can be the same: in recent VirtualBox
versions, no framebuffer device is being exposed by the kernel "a
priori" in a guest Ubuntu-daily system.
Have you tried to boot on a VMware virtual machine?
--
It also happens with me. In my multiseat system, if I logout on a son-
seat0 seat (which has allow-user-switching=false by default), lightdm
fails to start a greeter on that seat:
DEBUG: Seat: Display server stopped
DEBUG: Seat: Active display server stopped, starting greeter
DEBUG: Seat: Stopping
Here are my test results on latest Lubuntu 14.10 daily build booted in
live mode on both VirtualBox 4.3.14 and VMware Player 6.0.3:
VirtualBox: FAILED (no framebuffer devices available, probably required
kernel module "vboxvideo" is missing in the Live DVD). Same results as
@elfy
VMWare Player: P
According with your "loginctl seat-status seat0" outputs, it seems your
live session lacks framebuffer devices (/dev/fb*) for VirtualBox/VMware.
Just for comparison, my Lubuntu system installed in a VirtualBox VM does
have a framebuffer device, so I suspect some needed kernel module for
providing f
@mdeslaur I was unable to reproduce your error here. I've just installed
and updated Lubuntu 14.10beta1 in a VMware Player VM and everything goes
fine. Could you provide some more info about your VM (lightdm logs, Xorg
logs, etc.)?
--
You received this bug notification because you are a member of
I've crated a topic in NVIDIA DevZone Forums about shipping this udev
rule. Someone have pointed out about the possibility of no kernel
devices (like /dev/dri/card*) being available, since they are not
created by NVIDIA drivers, but by other kernel modules.
So I'm proposing this alternative udev r
@robert-ancell I believe not, but I'm not sure. This udev rule I propose
is a derivative of other ideas proposed by Ubuntu Multiseat team
members.
When that "master-of-seat" concept was introduced in systemd, Lennart
Poettering suggested that a suitable udev rule should be included in
upstream NIV
For future reference to nvidia-graphics-drivers maintainers: I suggest
to install a file named
/usr/lib/udev/rules.d/71-seat-nvidia.rules
with the sixth alternative proposed above:
SUBSYSTEM=="drm", KERNEL=="card[0-9]*", ATTRS{vendor}=="0x10de",
DRIVERS=="nvidia", TAG+="master-of-seat"
--
You
Errata:
5. Without this custom udev rule, seat0 will NEVER change his property
"CanGraphical" from "no" to "yes", so LightDM fails to spawn a greeter
on it.
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to lightdm in Ubuntu.
h
Well... Now I can tell the whole story:
1. Following http://www.freedesktop.org/wiki/Software/systemd/writing-
display-managers/ , LightDM gains automatic multiseat support in release
1.11.7. This support improves a bit in release 1.11.8, which now watches
on changes in property "CanGraphical" of
Please take these two last alternatives:
Fifth alternative:
SUBSYSTEM=="drm", KERNEL=="card[0-9]*", DRIVERS=="nvidia", TAG+="master-of-seat"
Sixth alternative:
SUBSYSTEM=="drm", KERNEL=="card[0-9]*", ATTRS{vendor}=="0x10de",
DRIVERS=="nvidia", TAG+="master-of-seat"
--
You received this bug not
Now that your system is working again, could you make some more tests?
Here are 4 alternative rules for your file
/etc/udev/rules.d/71-nvidia.rules. Could you please test each of them
and see if your system breaks?
First alternative (should work):
SUBSYSTEM=="drm", KERNEL=="card[0-9]*", TAG+="mas
Great! Now you have your NVIDIA graphics card tagged as [MASTER].
Could you check output of command "loginctl show-seat seat0" once again?
Does it show CanGraphical=yes now? Did lightdm goes back to normal?
--
You received this bug notification because you are a member of Ubuntu
Touch seeded pac
Not yet tagged as [MASTER]. Let's try again with something even more
simple.
Update your file /etc/udev/rules.d/71-nvidia.rules with this rule:
SUBSYSTEM=="drm", KERNEL=="card[0-9]*", TAG+="master-of-seat"
Double check if filename is correct (a common mistake from mine is
putting a .conf extensi
Could you put again the output of "loginctl seat-status seat0" after
applying your udev rule? Attach your file /var/log/Xorg.0.log after
startx failure, too.
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to lightdm in Ubuntu.
h
@elfy
OK, let's try again. Update your file /etc/udev/rules.d/71-nvidia.rules
with the following content (again, put all of them in a single line):
TAG=="seat", DEVPATH=="/devices/pci:00/:00:01.0/:01:00.0/*",
TAG+="master-of-seat"
Run "sudo udevadm trigger" and see if your device
/sy
@elfy
After creating that udev rule and loaded it, run "loginctl seat-status seat0"
once again and check if there's a device tagged with [MASTER].
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to lightdm in Ubuntu.
https://bug
@elfy
Create a file named /etc/udev/rules.d/71-nvidia.rules with the following
content (don't break lines):
SUBSYSTEM=="pci",
DEVPATH=="/devices/pci:00/:00:01.0/:01:00.0", TAG+="seat",
TAG+="master-of-seat"
Then run "udevadm trigger" (or just reboot your machine).
--
You received t
@elfy could you please attach here the output of following commands?
loginctl show-seat seat0
loginctl seat-status seat0
I've seen you're using nvidia proprietary drivers, right?
---
Robert Ancell, I suspect this is an example of that "critical case" I've
pointed out in bug #1590581: main grap
I've installed Lubuntu 14.10beta1 in a VirtualBox VM, executed "apt
full-upgrade", and rebooted. I could not be able to reproduce this
error: everything goes fine to me.
@elfy did you try to restart lightdm service after failure (sudo restart
lightdm)?
--
You received this bug notification becau
Reading your lightdm log, it seems LightDM cannot load seat0 for your
case.
Do you have any custom config file under /etc/lightdm/lightdm.conf.d (or
/etc/lightdm/lightdm.conf)? Specially, do you set option "start-default-
seat=false" explicitly?
--
You received this bug notification because you
Public bug reported:
Please consider backporting the following patch introduced in v209:
http://cgit.freedesktop.org/systemd/systemd/commit/?id=3fdb2494c1e24c0a020f5b54022d2c751fd26f50
This patch reverts lazy session-activation on non-seat0 seats, which may
lead to inactive sessions for users lo
34 matches
Mail list logo
