12:58 p.m.
I keep three PCs on my desk, running Fedora behind a KVM switch: #1 is
my present biggest fastest, #2 its predecessor, and #3 that one's
predecessor; I try try keep them as similar as is feasible.
They've all been having troubles for months, which have only gotten worse.
Under F27, they kept filling up with some sort of cruft, to the point
of refusing dnf upgrade; but when I found any of the cruft, it was in
places where I dared not lay about me with a cyber-battleaxe. I jumped
to F28 the day of release.
At this point, none of the three is usable at all without some dodge or
other.
#2 and #3 have each been wiped, twice, with different releases of DBAN,
followed each time with a netinstall of F28 with Mate, and then dnf
upgrade daily. All their boot messages show green OK. Those end with
"Started Light Display Manager."
Usually they proceed to a login box in front of the F28 display of blue
light-conducting fibers. That accepts my password, churns a little way,
and ends with a monitor full of blue horizontal lines, plus a few white
lines in the middle and at the bottom; there is nothing legible.
By doing Alt-Ctrl-F2, I get a display which lets me log in as root; I
keep doing dnf upgrades, and trying startx, every day. I also reboot
whenever there's a kernel change.
Startx always flashes a couple times, then fails with a few lines of
text, beginning with an attempt to adopt my monitor size (called "1920
1080" -- not 1920x1080),losing contact, and ending by saying that the
Xserver "terminated successfully."
I'd like to try replacing LDM with something else; how do I do that??
(I like Mate well, and would prefer not to swap it out,too)
Finally, #1 has two F28 kernels -- which do no better than on #2 nor #3
-- and a rescue kernel from F26. Sic. Twenty-six, not 27 nor 28. But
that rescue kernel does support almost my whole GUI (I miss Pan badly.),
and I don't have to descend to the console.
#1 got F28 by upgrading with dnf from F27, and still has a lot of stuff
on it that I'd rather copy to #2 and #3 directly than by way of solid
state storage, if I dare risk malware. (I don't know where else all the
cruft could have come from, apart from malware, but I hope someone here
does.)
--
Beartooth Staffwright, Not Quite Clueless Power User
Remember I know little (precious little!) of where up is.
1:38 p.m.
I use SLiM. I read somewhere that there is no further development and it will stop working
eventually but happily Fedora still packages it so as long as that holds, I guess I am
fine.
https://github.com/iwamatsu/slim
Ranjan
On Sun, 20 May 2018 13:58:36 -0400 Beartooth <beartooth(a)comcast.net> wrote:
I keep three PCs on my desk, running Fedora behind a KVM switch: #1 is
my present biggest fastest, #2 its predecessor, and #3 that one's
predecessor; I try try keep them as similar as is feasible.
They've all been having troubles for months, which have only gotten worse.
Under F27, they kept filling up with some sort of cruft, to the point
of refusing dnf upgrade; but when I found any of the cruft, it was in
places where I dared not lay about me with a cyber-battleaxe. I jumped
to F28 the day of release.
At this point, none of the three is usable at all without some dodge or
other.
#2 and #3 have each been wiped, twice, with different releases of DBAN,
followed each time with a netinstall of F28 with Mate, and then dnf
upgrade daily. All their boot messages show green OK. Those end with
"Started Light Display Manager."
Usually they proceed to a login box in front of the F28 display of blue
light-conducting fibers. That accepts my password, churns a little way,
and ends with a monitor full of blue horizontal lines, plus a few white
lines in the middle and at the bottom; there is nothing legible.
By doing Alt-Ctrl-F2, I get a display which lets me log in as root; I
keep doing dnf upgrades, and trying startx, every day. I also reboot
whenever there's a kernel change.
Startx always flashes a couple times, then fails with a few lines of
text, beginning with an attempt to adopt my monitor size (called "1920
1080" -- not 1920x1080),losing contact, and ending by saying that the
Xserver "terminated successfully."
I'd like to try replacing LDM with something else; how do I do that??
(I like Mate well, and would prefer not to swap it out,too)
Finally, #1 has two F28 kernels -- which do no better than on #2 nor #3
-- and a rescue kernel from F26. Sic. Twenty-six, not 27 nor 28. But
that rescue kernel does support almost my whole GUI (I miss Pan badly.),
and I don't have to descend to the console.
#1 got F28 by upgrading with dnf from F27, and still has a lot of stuff
on it that I'd rather copy to #2 and #3 directly than by way of solid
state storage, if I dare risk malware. (I don't know where else all the
cruft could have come from, apart from malware, but I hope someone here
does.)
--
Beartooth Staffwright, Not Quite Clueless Power User
Remember I know little (precious little!) of where up is.
_______________________________________________
users mailing list -- users(a)lists.fedoraproject.org
To unsubscribe send an email to users-leave(a)lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
--
Important Notice: This mailbox is ignored: e-mails are set to be deleted on receipt.
Please respond to the mailing list if appropriate. For those needing to send personal or
professional e-mail, please use appropriate addresses.
2:46 p.m.
On Sun, 20 May 2018 13:58:36 -0400
Beartooth <beartooth(a)comcast.net> wrote:
I'd like to try replacing LDM with something else; how do I
do that?? (I like Mate well, and would prefer not to swap it out,too)
I use openbox (it's in the Fedora repositories) with lxde. I usually
use startx from multiuser, but openbox is configured as the dm in lxde.
There was a thread here recently about how to switch desktop managers.
I ignored it since it didn't apply to me. Sorry. :-) Maybe someone
else will pipe up.
2:56 p.m.
On Sun, 20 May 2018 13:58:36 -0400
Beartooth <beartooth(a)comcast.net> wrote:
Forgot this:
Finally, #1 has two F28 kernels -- which do no better than on
#2 nor #3 -- and a rescue kernel from F26. Sic. Twenty-six, not 27
nor 28. But that rescue kernel does support almost my whole GUI (I
miss Pan badly.), and I don't have to descend to the console.
There was a thread about how to update the rescue kernel on a system
here recently as well. Courtesy of Samuel Sieb:
"""
You delete the rescue initramfs from /boot. Then the next time you
install a new kernel, it will create a new one. Possibly reinstalling
the existing kernel would work as well.
"""
As far as pan goes, going to koji, getting a version that works, and
downgrading seems to be the only recourse. There is a ticket open for
the problem, but I don't think upstream is active on pan anymore.
https://bugzilla.redhat.com/show_bug.cgi?id=1546863
3:02 p.m.
On Sun, 20 May 2018 12:46:22 -0700
stan wrote:
switch desktop managers.
I ignored it since it didn't apply to me.
Pretty simple really. Just do something like:
systemctl disable gdm.service
systemctl enable kdm.service
If you don't have the one you want installed, you
can do something like:
[root@zooty ~]# dnf install /usr/lib/systemd/system/xdm.service
Last metadata expiration check: 0:00:00 ago on Sun 20 May 2018 04:00:32 PM EDT.
Dependencies resolved.
================================================================================
Package Arch Version Repository Size
================================================================================
Installing:
xorg-x11-xdm x86_64 1:1.1.11-16.fc28 fedora 173 k
Transaction Summary
================================================================================
Install 1 Package
Total download size: 173 k
Installed size: 460 k
Is this ok [y/N]:
Presumably that will drag in any dependencies it need to properly
run that flavor login manager.
4:33 p.m.
On Sun, 2018-05-20 at 16:02 -0400, Tom Horsley wrote:
On Sun, 20 May 2018 12:46:22 -0700
stan wrote:
> switch desktop managers.
> I ignored it since it didn't apply to me.
Pretty simple really. Just do something like:
systemctl disable gdm.service
systemctl enable kdm.service
IIRC KDM is deprecated, so SDDM (now the default for KDE/Plasma) would
probably be better.
poc
4:42 p.m.
On Sun, 20 May 2018 22:33:59 +0100
Patrick O'Callaghan wrote:
IIRC KDM is deprecated, so SDDM (now the default for KDE/Plasma)
would
probably be better.
No doubt so they could prevent me from changing any X server
options :-). That's why I switched to kdm in the first place:
https://bugzilla.redhat.com/show_bug.cgi?id=451562
(Now on the register of historic bugs :-).
4:48 p.m.
On 05/21/18 04:02, Tom Horsley wrote:
Pretty simple really. Just do something like:
systemctl disable gdm.service
systemctl enable kdm.service
In actuality you just need one command.
systemctl -f enable sddm (or whatever new dm you want)
-f = force which means "When used with enable, overwrite any existing conflicting
symlinks."
and no need to use ".service"
--
Conjecture is just a conclusion based on incomplete information. It isn't a fact.
7:36 p.m.
Allegedly, on or about 20 May 2018, Beartooth sent:
Under F27, they kept filling up with some sort of cruft, to the
point of refusing dnf upgrade; but when I found any of the cruft, it
was in places where I dared not lay about me with a cyber-battleaxe.
I jumped to F28 the day of release.
Not enough real information for anyone to diagnose.
I run systems installed new, not upgrades over prior releases. I don't
use special wiping tools to erase the hard drives before hand, I just
use the installer to delete prior partitions. I install maybe about
six more programs than the default workstations (I use the MATE
desktop). I use "dnf update" every few days, manually, to keep things
up to date. I don't allow any automatic background updating, nor any
other updating tool to run. I haven't do anything to clear the caches.
I don't find my systems filling up with unexpected files.
What I do notice is that with systems left running 24/7, that things do
slow down and reboots can become required every few weeks. There are
programs that just don't terminate every time when I log out. Various
system/desktop daemons, sometimes web browsers.
Startx always flashes a couple times, then fails with a few
lines of text, beginning with an attempt to adopt my monitor size
(called "1920 1080" -- not 1920x1080),losing contact, and ending by
saying that the Xserver "terminated successfully."
I've had a few dicey moments with a dying video card, that got swapped,
then I replaced the whole motherboard for something new (instead of
carrying on using something that's around 8 years old). Graphics
problems can be due to failing hardware, or just poor supported
hardware, and sometimes badly designed monitors (ones that don't
announce their specs, or are still in stand-by mode when the OS is
trying to probe them).
I'd like to try replacing LDM with something else; how do I do
that?? (I like Mate well, and would prefer not to swap it out,too)
I just went with MATE's own display manager. Unless you have a very
basic graphics set, or poorly supported one, it ought to manage the
job.
--
[tim@localhost ~]$ uname -rsvp
Linux 4.16.7-100.fc26.x86_64 #1 SMP Wed May 2 21:45:56 UTC 2018 x86_64
Boilerplate: All mail to my mailbox is automatically deleted.
There is no point trying to privately email me, I only get to see
the messages posted to the mailing list.
- Mwuu haha haaaaa haaaa, soon the world will be mine!
- Sir, you've got to take your finger off the intercom button.
10:09 a.m.
On Mon, 2018-05-21 at 10:06 +0930, Tim via users wrote:
There are
programs that just don't terminate every time when I log out. Various
system/desktop daemons, sometimes web browsers.
Possibly unrelated, but I sometimes - not always - find that after
updating (using dnf with the tracer plugin) I'm advised to restart my
session, then on logging out and in I'm suddenly forced out again after
a short time, usually less than a minute.
(Also, and only in this specific situation, Firefox forgets all the
windows it had open except one, but I had the same problem with
Chrome).
On logging in *again*, everything is fine. It's as if the session
manager isn't waiting long enough for things to terminate gracefully
before presenting the login screen, then when they do terminate it
panics. I haven't managed to narrow it down yet.
poc
10:10 a.m.
On Mon, 2018-05-21 at 16:09 +0100, Patrick O'Callaghan wrote:
On Mon, 2018-05-21 at 10:06 +0930, Tim via users wrote:
> There are
> programs that just don't terminate every time when I log out. Various
> system/desktop daemons, sometimes web browsers.
Possibly unrelated, but I sometimes - not always - find that after
updating (using dnf with the tracer plugin) I'm advised to restart my
session, then on logging out and in I'm suddenly forced out again after
a short time, usually less than a minute.
(Also, and only in this specific situation, Firefox forgets all the
windows it had open except one, but I had the same problem with
Chrome).
On logging in *again*, everything is fine. It's as if the session
manager isn't waiting long enough for things to terminate gracefully
before presenting the login screen, then when they do terminate it
panics. I haven't managed to narrow it down yet.
Forgot to mention that this is with KDE and SDDM.
poc
4:52 p.m.
Tim:
> There are programs that just don't terminate every time when
I log
> out. Various system/desktop daemons, sometimes web browsers.
Patrick O'Callaghan:
Possibly unrelated, but I sometimes - not always - find that after
updating (using dnf with the tracer plugin) I'm advised to restart my
session, then on logging out and in I'm suddenly forced out again
after a short time, usually less than a minute.
Software updates aren't part of my problem's situation. It can do it
when I haven't done any updates since the last reboot.
--
[tim@localhost ~]$ uname -rsvp
Linux 4.16.7-100.fc26.x86_64 #1 SMP Wed May 2 21:45:56 UTC 2018 x86_64
Boilerplate: All mail to my mailbox is automatically deleted.
There is no point trying to privately email me, I only get to see
the messages posted to the mailing list.
You can't have equality AND special treatment.
2:16 a.m.
On 05/21/18 23:09, Patrick O'Callaghan wrote:
On Mon, 2018-05-21 at 10:06 +0930, Tim via users wrote:
> There are
> programs that just don't terminate every time when I log out. Various
> system/desktop daemons, sometimes web browsers.
Possibly unrelated, but I sometimes - not always - find that after
updating (using dnf with the tracer plugin) I'm advised to restart my
session, then on logging out and in I'm suddenly forced out again after
a short time, usually less than a minute.
FWIW, if you investigate a little I think you'd find that for every user that logs-in
for a GUI session there will be about 9+ processes that are left running and remain
running. It seems these processes are then reused on subsequent logins by the same
uid. I suppose this may be done to speed-up subsequent logins.
So, those processes may potentially keep copies of libraries that have been updated
and (guessing) clashes happen with new processes with new libraries.
Therefore, I don't logout/login after updates. If anything, I reboot. In my case a
reboot takes about 12 seconds so I'm not bothered by it.
I was considering asking on the KDE list if there is anyway to tell the system to end
all users processes after logout.
(Also, and only in this specific situation, Firefox forgets all the
windows it had open except one, but I had the same problem with
Chrome).
On logging in *again*, everything is fine. It's as if the session
manager isn't waiting long enough for things to terminate gracefully
before presenting the login screen, then when they do terminate it
panics. I haven't managed to narrow it down yet.
[egreshko@acer ~]$ date
Tue May 22 15:15:23 CST 2018
[egreshko@acer ~]$ ps -eaf | grep 14:02 | grep maria | wc
9 83 828
[egreshko@acer ~]$ finger
Login Name Tty Idle Login Time Office Office Phone Host
egreshko Ed Greshko pts/1 May 22 14:03
(192.168.2.190)
and "maria" has been logged-out of the GUI for over an hour.
--
Conjecture is just a conclusion based on incomplete information. It isn't a fact.
4:34 a.m.
Allegedly, on or about 22 May 2018, Ed Greshko sent:
FWIW, if you investigate a little I think you'd find that for
every
user that logs-in for a GUI session there will be about 9+ processes
that are left running and remain running. It seems these processes
are then reused on subsequent logins by the same uid. I suppose this
may be done to speed-up subsequent logins.
I had wondered about that, but I'd find that usually when I log out,
that all my processes eventually disappear. Virtually straight away.
Sometimes, things would loiter around, and they'd prevent you from
logging back in. With things the login attempt would think you're
still logged in, or the desktop would never finish drawing up because
it was waiting for stalled files that never closed when you first
logged out, or there'd be no sound because pulseaudio had wedged.
--
[tim@localhost ~]$ uname -rsvp
Linux 4.16.7-100.fc26.x86_64 #1 SMP Wed May 2 21:45:56 UTC 2018 x86_64
Boilerplate: All mail to my mailbox is automatically deleted.
There is no point trying to privately email me, I only get to see
the messages posted to the mailing list.
- And how would you describe Windows?
- One man's trash is another man's treasure...
6:02 a.m.
On Tue, 2018-05-22 at 07:22 +0930, Tim via users wrote:
Tim:
> > There are programs that just don't terminate every time when I log
> > out. Various system/desktop daemons, sometimes web browsers.
Patrick O'Callaghan:
> Possibly unrelated, but I sometimes - not always - find that after
> updating (using dnf with the tracer plugin) I'm advised to restart my
> session, then on logging out and in I'm suddenly forced out again
> after a short time, usually less than a minute.
Software updates aren't part of my problem's situation. It can do it
when I haven't done any updates since the last reboot.
The only times I ever log out is after updating, so that may or may not
be relevant.
poc
6:12 a.m.
On Tue, 2018-05-22 at 15:16 +0800, Ed Greshko wrote:
On 05/21/18 23:09, Patrick O'Callaghan wrote:
> On Mon, 2018-05-21 at 10:06 +0930, Tim via users wrote:
> > There are
> > programs that just don't terminate every time when I log out. Various
> > system/desktop daemons, sometimes web browsers.
>
> Possibly unrelated, but I sometimes - not always - find that after
> updating (using dnf with the tracer plugin) I'm advised to restart my
> session, then on logging out and in I'm suddenly forced out again after
> a short time, usually less than a minute.
FWIW, if you investigate a little I think you'd find that for every user that
logs-in
for a GUI session there will be about 9+ processes that are left running and remain
running. It seems these processes are then reused on subsequent logins by the same
uid. I suppose this may be done to speed-up subsequent logins.
So, those processes may potentially keep copies of libraries that have been updated
and (guessing) clashes happen with new processes with new libraries.
Yes, I'm aware of that. However 'tracer' supposed to detect this,
either when called from the Shell or via the dnf plug-in. That's why
it's telling me to restart the session.
Therefore, I don't logout/login after updates. If anything, I
reboot. In my case a
reboot takes about 12 seconds so I'm not bothered by it.
It's fast here too, but I usually have a VM running that I may not want
to forcibly reboot.
poc
7:21 a.m.
On 05/22/18 19:28, Rex Dieter wrote:
Ed Greshko wrote:
> I was considering asking on the KDE list if there is anyway to tell the
> system to end all users processes after logout.
Set in /etc/systemd/logind.conf:
KillUserProcesses=yes
Thanks....
--
Conjecture is just a conclusion based on incomplete information. It isn't a fact.
7:26 a.m.
On Tue, 22 May 2018 06:28:21 -0500
Rex Dieter wrote:
Set in /etc/systemd/logind.conf:
KillUserProcesses=yes
Does that get all the "user mode daemon" stuff that runs under the
"/usr/lib/systemd/systemd --user" process?
Those hanging around forever always seemed to make a reboot take
forever, so I wrote a special program to find all the user daemon
process trees and kill them all off before a reboot (which seems
to make reboot infinitely faster).
http://tomhorsley.com/game/punch.html
7:42 a.m.
On 05/22/18 20:26, Tom Horsley wrote:
On Tue, 22 May 2018 06:28:21 -0500
Rex Dieter wrote:
> Set in /etc/systemd/logind.conf:
>
> KillUserProcesses=yes
Does that get all the "user mode daemon" stuff that runs under the
"/usr/lib/systemd/systemd --user" process?
Well, like any good tech, I just tried it and it did kill all of those. No traces of
the logged-out user remain.
--
Conjecture is just a conclusion based on incomplete information. It isn't a fact.
8:28 a.m.
On Tue, 22 May 2018 20:42:30 +0800
Ed Greshko wrote:
Well, like any good tech, I just tried it and it did kill all of
those. No traces of
the logged-out user remain.
Cool. I'll have to configure my system with that.
Of course I'll still need my reboot script to umount -l nfs filesystems
to keep systemd from spending 5 hours timing out trying to talk to
systems that have gone down :-).
9:51 a.m.
On Tue, 2018-05-22 at 06:28 -0500, Rex Dieter wrote:
Ed Greshko wrote:
> I was considering asking on the KDE list if there is anyway to tell the
> system to end all users processes after logout.
Set in /etc/systemd/logind.conf:
KillUserProcesses=yes
I already have that set, and have had it for a long time.
poc
10 a.m.
On 05/22/18 19:12, Patrick O'Callaghan wrote:
On Tue, 2018-05-22 at 15:16 +0800, Ed Greshko wrote:
> On 05/21/18 23:09, Patrick O'Callaghan wrote:
>> On Mon, 2018-05-21 at 10:06 +0930, Tim via users wrote:
>>> There are
>>> programs that just don't terminate every time when I log out. Various
>>> system/desktop daemons, sometimes web browsers.
>> Possibly unrelated, but I sometimes - not always - find that after
>> updating (using dnf with the tracer plugin) I'm advised to restart my
>> session, then on logging out and in I'm suddenly forced out again after
>> a short time, usually less than a minute.
> FWIW, if you investigate a little I think you'd find that for every user that
logs-in
> for a GUI session there will be about 9+ processes that are left running and remain
> running. It seems these processes are then reused on subsequent logins by the same
> uid. I suppose this may be done to speed-up subsequent logins.
>
> So, those processes may potentially keep copies of libraries that have been updated
> and (guessing) clashes happen with new processes with new libraries.
Yes, I'm aware of that. However 'tracer' supposed to detect this,
either when called from the Shell or via the dnf plug-in. That's why
it's telling me to restart the session.
I seriously doubt that tracer will be able to detect that some processes may survive
a log-outt/log-in event.
Do you run tracer again after logging in again?
> Therefore, I don't logout/login after updates. If anything,
I reboot. In my case a
> reboot takes about 12 seconds so I'm not bothered by it.
It's fast here too, but I usually have a VM running that I may not want
to forcibly reboot.
There is always something, isn't there?
--
Conjecture is just a conclusion based on incomplete information. It isn't a fact.
10:01 a.m.
Allegedly, on or about 22 May 2018, Tom Horsley sent:
Of course I'll still need my reboot script to umount -l nfs
filesystems to keep systemd from spending 5 hours timing out trying
to talk to systems that have gone down :-).
That's why I went with autofs, rather than fstab entries. They're only
mounted on demand, and automatically dismount after a little while.
And it only gets worse when you have several NFS mounts. I used to
just have a central server, and that wasn't too bad. But when you add
in NAS devices, or clients that have individual exports, it gets really
messy to handle shutdowns, or IP changes.
--
[tim@localhost ~]$ uname -rsvp
Linux 4.16.7-100.fc26.x86_64 #1 SMP Wed May 2 21:45:56 UTC 2018 x86_64
Boilerplate: All mail to my mailbox is automatically deleted.
There is no point trying to privately email me, I only get to see
the messages posted to the mailing list.
Television should really come with an intelligence knob. I've tried
adjusting the brightness, but it didn't help.
10:02 a.m.
On 05/22/18 22:51, Patrick O'Callaghan wrote:
On Tue, 2018-05-22 at 06:28 -0500, Rex Dieter wrote:
> Ed Greshko wrote:
>
>
>> I was considering asking on the KDE list if there is anyway to tell the
>> system to end all users processes after logout.
> Set in /etc/systemd/logind.conf:
>
> KillUserProcesses=yes
I already have that set, and have had it for a long time.
Oh, then it seems nothing I've written actually applies to your situation. Too
bad I
only just read this.
--
Conjecture is just a conclusion based on incomplete information. It isn't a fact.
11:33 a.m.
On Tue, 2018-05-22 at 23:00 +0800, Ed Greshko wrote:
>> So, those processes may potentially keep copies of libraries
that have been updated
>> and (guessing) clashes happen with new processes with new libraries.
> Yes, I'm aware of that. However 'tracer' supposed to detect this,
> either when called from the Shell or via the dnf plug-in. That's why
> it's telling me to restart the session.
I seriously doubt that tracer will be able to detect that some processes may survive
a log-outt/log-in event.
Naturally not, but it should (and AFAIK does) detect libraries being
updated and recommend that processes using them be restarted.
Do you run tracer again after logging in again?
I sometimes run it from a root console while the user session is logged
out, and make sure nothing out of date is still hanging around. This
doesn't appear to correlate with the problem, but I haven't been
systematic enough about it to be sure.
>> Therefore, I don't logout/login after updates. If
anything, I reboot. In my case a
>> reboot takes about 12 seconds so I'm not bothered by it.
> It's fast here too, but I usually have a VM running that I may not want
> to forcibly reboot.
>
There is always something, isn't there?
Yes, Murphy had something to say about that ...
poc
11:35 a.m.
On Tue, 2018-05-22 at 23:02 +0800, Ed Greshko wrote:
On 05/22/18 22:51, Patrick O'Callaghan wrote:
> On Tue, 2018-05-22 at 06:28 -0500, Rex Dieter wrote:
> > Ed Greshko wrote:
> >
> >
> > > I was considering asking on the KDE list if there is anyway to tell the
> > > system to end all users processes after logout.
> >
> > Set in /etc/systemd/logind.conf:
> >
> > KillUserProcesses=yes
>
> I already have that set, and have had it for a long time.
>
Oh, then it seems nothing I've written actually applies to your situation. Too bad
I
only just read this.
Never mind. I have actually brought it up before now (can't be bothered
to track down the post though it's probably on the KDE list) but no-one
else reacted so it doesn't appear to be common.
poc
11:56 a.m.
On 05/22/2018 09:33 AM, Patrick O'Callaghan wrote:
On Tue, 2018-05-22 at 23:00 +0800, Ed Greshko wrote:
>>> So, those processes may potentially keep copies of libraries that have been
updated
>>> and (guessing) clashes happen with new processes with new libraries.
>> Yes, I'm aware of that. However 'tracer' supposed to detect this,
>> either when called from the Shell or via the dnf plug-in. That's why
>> it's telling me to restart the session.
>
> I seriously doubt that tracer will be able to detect that some processes may survive
> a log-outt/log-in event.
Naturally not, but it should (and AFAIK does) detect libraries being
updated and recommend that processes using them be restarted.
> Do you run tracer again after logging in again?
I sometimes run it from a root console while the user session is logged
out, and make sure nothing out of date is still hanging around. This
doesn't appear to correlate with the problem, but I haven't been
systematic enough about it to be sure.
>>> Therefore, I don't logout/login after updates. If anything, I reboot.
In my case a
>>> reboot takes about 12 seconds so I'm not bothered by it.
>> It's fast here too, but I usually have a VM running that I may not want
>> to forcibly reboot.
>>
> There is always something, isn't there?
Yes, Murphy had something to say about that ...
And I just had lightdm screw up royally. First, upon login I got an
XFCE polkit error popup, which is singularly useless in telling you
anything you can troubleshoot with. Then, I had no access to the sound
hardware on my machine, nor could I enable/disable wireless, play
with firewall settings or anything else, as polkit thought I didn't have
permissions to do anything, nor did it ever pop up an authentication
dialog.
I wiped my xfce settings, reinstalled lightdm, xfce4-polkit and anything
else I could think of to no avail. I finally just punted, disabled
lightdm and enabled lxdm and everything started working again.
Very, VERY frustrating. Grrrrr!
----------------------------------------------------------------------
- Rick Stevens, Systems Engineer, AllDigital ricks(a)alldigital.com -
- AIM/Skype: therps2 ICQ: 22643734 Yahoo: origrps2 -
- -
- Admitting you have a problem is the first step toward getting -
- medicated for it. -- Jim Evarts (http://www.TopFive.com) -
----------------------------------------------------------------------
12:32 p.m.
On Tue, 22 May 2018 09:56:25 -0700 Rick Stevens wrote:
And I just had lightdm screw up royally. First, upon login I got an
XFCE polkit error popup, which is singularly useless in telling you
anything you can troubleshoot with. Then, I had no access to the sound
hardware on my machine, nor could I enable/disable wireless, play
with firewall settings or anything else, as polkit thought I didn't have
permissions to do anything, nor did it ever pop up an authentication
dialog.
This is not the fault of lightdm but of pam-kwallet. See this other
thread:
F27 problems with pam?
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
--
francis
2:05 p.m.
Not sure if it is related. I've seen messages about /run/user/1000 directory
not existing for a number of things in the message log.
Have found that if I open a terminal and ssh into the same machine with the
same user id, the directory is created and the error messages don't show up.
Thought of creating a start up process to automatically do this on login.
On 22 May 2018 at 19:32, Francis.Montagnac(a)inria.fr wrote:
From: Francis.Montagnac(a)inria.fr
To: Community support for Fedora users
<users(a)lists.fedoraproject.org>
Subject: Re: Could LightDM be bad?? (LONG)
Date sent: Tue, 22 May 2018 19:32:30 +0200
Send reply to: Community support for Fedora users
<users(a)lists.fedoraproject.org>
On Tue, 22 May 2018 09:56:25 -0700 Rick Stevens wrote:
> And I just had lightdm screw up royally. First, upon login I got an
> XFCE polkit error popup, which is singularly useless in telling you
> anything you can troubleshoot with. Then, I had no access to the sound
> hardware on my machine, nor could I enable/disable wireless, play
> with firewall settings or anything else, as polkit thought I didn't have
> permissions to do anything, nor did it ever pop up an authentication
> dialog.
This is not the fault of lightdm but of pam-kwallet. See this other
thread:
F27 problems with pam?
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
--
francis
_______________________________________________
users mailing list -- users(a)lists.fedoraproject.org
To unsubscribe send an email to users-leave(a)lists.fedoraproject.org
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
+------------------------------------------------------------+
Michael D. Setzer II - Computer Science Instructor (Retired)
mailto:mikes@guam.net
mailto:msetzerii@gmail.com
Guam - Where America's Day Begins
G4L Disk Imaging Project maintainer
http://sourceforge.net/projects/g4l/
+------------------------------------------------------------+
http://setiathome.berkeley.edu (Original)
Number of Seti Units Returned: 19,471
Processing time: 32 years, 290 days, 12 hours, 58 minutes
(Total Hours: 287,489)
BOINC@HOME CREDITS
ROSETTA 65534319.832076 | ABC 16613838.513356
SETI 109369103.846085 | EINSTEIN 141058465.999240
3:24 p.m.
On 05/22/2018 10:32 AM, Francis.Montagnac(a)inria.fr wrote:
On Tue, 22 May 2018 09:56:25 -0700 Rick Stevens wrote:
> And I just had lightdm screw up royally. First, upon login I got an
> XFCE polkit error popup, which is singularly useless in telling you
> anything you can troubleshoot with. Then, I had no access to the sound
> hardware on my machine, nor could I enable/disable wireless, play
> with firewall settings or anything else, as polkit thought I didn't have
> permissions to do anything, nor did it ever pop up an authentication
> dialog.
This is not the fault of lightdm but of pam-kwallet. See this other
thread:
F27 problems with pam?
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
Well, very that's very interesting and, yup, that appears to be what I'm
dealing with. Really odd that lightdm would have such a dependency. I
could see it if I were using KDE, but wow!
Looking at my dnf logs, pam-kwallet got updated (for me) on the 18th,
but I didn't log out until sometime on the 19th. The defecation hit the
impeller yesterday (the 21st) when I logged in for the first time after
the update. Grrrrrr!
I suppose I could just disable the pam_kwallet lines from the pam config
and try again with lightdm. Using lxdm works around the issue for now,
however, so I have a way out in case that's not the issue.
Thanks for the tip, Francis! Ya learn something new everyday.
----------------------------------------------------------------------
- Rick Stevens, Systems Engineer, AllDigital ricks(a)alldigital.com -
- AIM/Skype: therps2 ICQ: 22643734 Yahoo: origrps2 -
- -
- Admitting you have a problem is the first step toward getting -
- medicated for it. -- Jim Evarts (http://www.TopFive.com) -
----------------------------------------------------------------------
5:15 p.m.
On 05/22/2018 01:24 PM, Rick Stevens wrote:
On 05/22/2018 10:32 AM, Francis.Montagnac(a)inria.fr wrote:
>
> On Tue, 22 May 2018 09:56:25 -0700 Rick Stevens wrote:
>
>> And I just had lightdm screw up royally. First, upon login I got an
>> XFCE polkit error popup, which is singularly useless in telling you
>> anything you can troubleshoot with. Then, I had no access to the sound
>> hardware on my machine, nor could I enable/disable wireless, play
>> with firewall settings or anything else, as polkit thought I didn't have
>> permissions to do anything, nor did it ever pop up an authentication
>> dialog.
>
> This is not the fault of lightdm but of pam-kwallet. See this other
> thread:
>
> F27 problems with pam?
>
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
Well, very that's very interesting and, yup, that appears to be what I'm
dealing with. Really odd that lightdm would have such a dependency. I
could see it if I were using KDE, but wow!
Looking at my dnf logs, pam-kwallet got updated (for me) on the 18th,
but I didn't log out until sometime on the 19th. The defecation hit the
impeller yesterday (the 21st) when I logged in for the first time after
the update. Grrrrrr!
I suppose I could just disable the pam_kwallet lines from the pam config
and try again with lightdm. Using lxdm works around the issue for now,
however, so I have a way out in case that's not the issue.
Thanks for the tip, Francis! Ya learn something new everyday.
Follow up: Commented out the references to pam_kwallet.so and
pam_kwallet5.so in /etc/pam.d/lightdm and et Voila! I'm back to lightdm
and Xfce4 with polkit behaving itself.
Thanks again to Francis for pointing me in the right direction!
----------------------------------------------------------------------
- Rick Stevens, Systems Engineer, AllDigital ricks(a)alldigital.com -
- AIM/Skype: therps2 ICQ: 22643734 Yahoo: origrps2 -
- -
- Fear is finding a ".vbs" script in your Inbox -
----------------------------------------------------------------------
5:38 p.m.
On 05/23/18 04:24, Rick Stevens wrote:
On 05/22/2018 10:32 AM, Francis.Montagnac(a)inria.fr wrote:
> On Tue, 22 May 2018 09:56:25 -0700 Rick Stevens wrote:
>
>> And I just had lightdm screw up royally. First, upon login I got an
>> XFCE polkit error popup, which is singularly useless in telling you
>> anything you can troubleshoot with. Then, I had no access to the sound
>> hardware on my machine, nor could I enable/disable wireless, play
>> with firewall settings or anything else, as polkit thought I didn't have
>> permissions to do anything, nor did it ever pop up an authentication
>> dialog.
> This is not the fault of lightdm but of pam-kwallet. See this other
> thread:
>
> F27 problems with pam?
>
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
Well, very that's very interesting and, yup, that appears to be what I'm
dealing with. Really odd that lightdm would have such a dependency. I
could see it if I were using KDE, but wow!
FWIW, I have an F28/xfce system installed in a VM. It uses lightdm
root 776 1 0 05:54 ? 00:00:00 /usr/sbin/lightdm
root 876 776 0 05:54 tty1 00:00:10 /usr/libexec/Xorg -core -noreset :0
-seat seat0 -auth /var/run/lightdm/root/:0 -nolisten tcp vt1 -novtswitch -background none
root 1058 776 0 05:54 ? 00:00:00 lightdm --session-child 12 19
The only "pam" packages installed on the system are...
gnome-keyring-pam-3.28.2-1.fc28.x86_64
pam-1.3.0-10.fc28.x86_64
systemd-pam-238-8.git0e0aa59.fc28.x86_64
pam_krb5-2.4.13-9.fc28.x86_64
fprintd-pam-0.8.0-2.fc28.x86_64
So, it seems to me, something else you've installed on your system pulled in
pam-kwallet and it isn't lightdm.
Maybe try removing it to see what else it tries to remove?
--
Conjecture is just a conclusion based on incomplete information. It isn't a fact.
5:45 p.m.
On 05/22/2018 03:38 PM, Ed Greshko wrote:
On 05/23/18 04:24, Rick Stevens wrote:
> On 05/22/2018 10:32 AM, Francis.Montagnac(a)inria.fr wrote:
>> On Tue, 22 May 2018 09:56:25 -0700 Rick Stevens wrote:
>>
>>> And I just had lightdm screw up royally. First, upon login I got an
>>> XFCE polkit error popup, which is singularly useless in telling you
>>> anything you can troubleshoot with. Then, I had no access to the sound
>>> hardware on my machine, nor could I enable/disable wireless, play
>>> with firewall settings or anything else, as polkit thought I didn't have
>>> permissions to do anything, nor did it ever pop up an authentication
>>> dialog.
>> This is not the fault of lightdm but of pam-kwallet. See this other
>> thread:
>>
>> F27 problems with pam?
>>
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
> Well, very that's very interesting and, yup, that appears to be what I'm
> dealing with. Really odd that lightdm would have such a dependency. I
> could see it if I were using KDE, but wow!
FWIW, I have an F28/xfce system installed in a VM. It uses lightdm
root 776 1 0 05:54 ? 00:00:00 /usr/sbin/lightdm
root 876 776 0 05:54 tty1 00:00:10 /usr/libexec/Xorg -core -noreset :0
-seat seat0 -auth /var/run/lightdm/root/:0 -nolisten tcp vt1 -novtswitch -background
none
root 1058 776 0 05:54 ? 00:00:00 lightdm --session-child 12 19
The only "pam" packages installed on the system are...
gnome-keyring-pam-3.28.2-1.fc28.x86_64
pam-1.3.0-10.fc28.x86_64
systemd-pam-238-8.git0e0aa59.fc28.x86_64
pam_krb5-2.4.13-9.fc28.x86_64
fprintd-pam-0.8.0-2.fc28.x86_64
So, it seems to me, something else you've installed on your system pulled in
pam-kwallet and it isn't lightdm.
Maybe try removing it to see what else it tries to remove?
I did install KDE at one time, so it's probably from that. Regardless,
I just commented it out of the pam config. I could remove it, I suppose.
I don't expect to use KDE often, if at all.
----------------------------------------------------------------------
- Rick Stevens, Systems Engineer, AllDigital ricks(a)alldigital.com -
- AIM/Skype: therps2 ICQ: 22643734 Yahoo: origrps2 -
- -
- I'm telling you that the kernel is stable not because it's a -
- kernel, but because I refuse to listen to arguments like this. -
- -- Linus Torvalds -
----------------------------------------------------------------------
5:48 p.m.
On 05/23/18 06:45, Rick Stevens wrote:
On 05/22/2018 03:38 PM, Ed Greshko wrote:
> On 05/23/18 04:24, Rick Stevens wrote:
>> On 05/22/2018 10:32 AM, Francis.Montagnac(a)inria.fr wrote:
>>> On Tue, 22 May 2018 09:56:25 -0700 Rick Stevens wrote:
>>>
>>>> And I just had lightdm screw up royally. First, upon login I got an
>>>> XFCE polkit error popup, which is singularly useless in telling you
>>>> anything you can troubleshoot with. Then, I had no access to the sound
>>>> hardware on my machine, nor could I enable/disable wireless, play
>>>> with firewall settings or anything else, as polkit thought I didn't
have
>>>> permissions to do anything, nor did it ever pop up an authentication
>>>> dialog.
>>> This is not the fault of lightdm but of pam-kwallet. See this other
>>> thread:
>>>
>>> F27 problems with pam?
>>>
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
>> Well, very that's very interesting and, yup, that appears to be what I'm
>> dealing with. Really odd that lightdm would have such a dependency. I
>> could see it if I were using KDE, but wow!
> FWIW, I have an F28/xfce system installed in a VM. It uses lightdm
>
> root 776 1 0 05:54 ? 00:00:00 /usr/sbin/lightdm
> root 876 776 0 05:54 tty1 00:00:10 /usr/libexec/Xorg -core -noreset :0
> -seat seat0 -auth /var/run/lightdm/root/:0 -nolisten tcp vt1 -novtswitch -background
none
> root 1058 776 0 05:54 ? 00:00:00 lightdm --session-child 12 19
>
> The only "pam" packages installed on the system are...
>
> gnome-keyring-pam-3.28.2-1.fc28.x86_64
> pam-1.3.0-10.fc28.x86_64
> systemd-pam-238-8.git0e0aa59.fc28.x86_64
> pam_krb5-2.4.13-9.fc28.x86_64
> fprintd-pam-0.8.0-2.fc28.x86_64
>
> So, it seems to me, something else you've installed on your system pulled in
> pam-kwallet and it isn't lightdm.
>
> Maybe try removing it to see what else it tries to remove?
I did install KDE at one time, so it's probably from that. Regardless,
I just commented it out of the pam config. I could remove it, I suppose.
I don't expect to use KDE often, if at all.
Well, if the KDE/lightdm combination causes problems then I hope those running, or
attempting to run, that combination will BZ it.
--
Conjecture is just a conclusion based on incomplete information. It isn't a fact.
6:07 p.m.
On 05/22/2018 03:48 PM, Ed Greshko wrote:
On 05/23/18 06:45, Rick Stevens wrote:
> On 05/22/2018 03:38 PM, Ed Greshko wrote:
>> On 05/23/18 04:24, Rick Stevens wrote:
>>> On 05/22/2018 10:32 AM, Francis.Montagnac(a)inria.fr wrote:
>>>> On Tue, 22 May 2018 09:56:25 -0700 Rick Stevens wrote:
>>>>
>>>>> And I just had lightdm screw up royally. First, upon login I got an
>>>>> XFCE polkit error popup, which is singularly useless in telling you
>>>>> anything you can troubleshoot with. Then, I had no access to the
sound
>>>>> hardware on my machine, nor could I enable/disable wireless, play
>>>>> with firewall settings or anything else, as polkit thought I
didn't have
>>>>> permissions to do anything, nor did it ever pop up an authentication
>>>>> dialog.
>>>> This is not the fault of lightdm but of pam-kwallet. See this other
>>>> thread:
>>>>
>>>> F27 problems with pam?
>>>>
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
>>> Well, very that's very interesting and, yup, that appears to be what
I'm
>>> dealing with. Really odd that lightdm would have such a dependency. I
>>> could see it if I were using KDE, but wow!
>> FWIW, I have an F28/xfce system installed in a VM. It uses lightdm
>>
>> root 776 1 0 05:54 ? 00:00:00 /usr/sbin/lightdm
>> root 876 776 0 05:54 tty1 00:00:10 /usr/libexec/Xorg -core -noreset
:0
>> -seat seat0 -auth /var/run/lightdm/root/:0 -nolisten tcp vt1 -novtswitch
-background none
>> root 1058 776 0 05:54 ? 00:00:00 lightdm --session-child 12 19
>>
>> The only "pam" packages installed on the system are...
>>
>> gnome-keyring-pam-3.28.2-1.fc28.x86_64
>> pam-1.3.0-10.fc28.x86_64
>> systemd-pam-238-8.git0e0aa59.fc28.x86_64
>> pam_krb5-2.4.13-9.fc28.x86_64
>> fprintd-pam-0.8.0-2.fc28.x86_64
>>
>> So, it seems to me, something else you've installed on your system pulled in
>> pam-kwallet and it isn't lightdm.
>>
>> Maybe try removing it to see what else it tries to remove?
> I did install KDE at one time, so it's probably from that. Regardless,
> I just commented it out of the pam config. I could remove it, I suppose.
> I don't expect to use KDE often, if at all.
Well, if the KDE/lightdm combination causes problems then I hope those running, or
attempting to run, that combination will BZ it.
Bug reported by me:
https://bugzilla.redhat.com/show_bug.cgi?id=1581495
----------------------------------------------------------------------
- Rick Stevens, Systems Engineer, AllDigital ricks(a)alldigital.com -
- AIM/Skype: therps2 ICQ: 22643734 Yahoo: origrps2 -
- -
- The gene pool could use a little chlorine. -
----------------------------------------------------------------------
1:30 a.m.
Hi
On Tue, 22 May 2018 16:07:22 -0700 Rick Stevens wrote:
On 05/22/2018 03:48 PM, Ed Greshko wrote:
> On 05/23/18 06:45, Rick Stevens wrote:
>> On 05/22/2018 03:38 PM, Ed Greshko wrote:
>>> On 05/23/18 04:24, Rick Stevens wrote:
>>>> On 05/22/2018 10:32 AM, Francis.Montagnac(a)inria.fr wrote:
>>>>> On Tue, 22 May 2018 09:56:25 -0700 Rick Stevens wrote:
>>>> Really odd that lightdm would have such a dependency.
I don't think so: it should allow using any kind of session and thus
includes in its /etc/pam.d/lightdm file optional pam modules for Gnome
and KDE:
-session optional pam_gnome_keyring.so auto_start
-session optional pam_kwallet5.so
-session optional pam_kwallet.so
>>> So, it seems to me, something else you've installed
on your
>>> system pulled in pam-kwallet and it isn't lightdm.
Right.
>>> Maybe try removing it to see what else it tries to
remove?
As I said in the other thread: no other RPM depends ont it. It is thus
safe to remove the pam-kwallet RPM.
>> I did install KDE at one time, so it's probably from
that. Regardless,
>> I just commented it out of the pam config. I could remove it, I suppose.
>> I don't expect to use KDE often, if at all.
> Well, if the KDE/lightdm combination causes problems then I hope
> those running, or attempting to run, that combination will BZ it.
Bug reported by me:
Thanks. I should have done it :-(
--
francis
12:22 p.m.
Francis.Montagnac(a)inria.fr wrote:
On Tue, 22 May 2018 09:56:25 -0700 Rick Stevens wrote:
> And I just had lightdm screw up royally. First, upon login I got an
> XFCE polkit error popup, which is singularly useless in telling you
> anything you can troubleshoot with. Then, I had no access to the sound
> hardware on my machine, nor could I enable/disable wireless, play
> with firewall settings or anything else, as polkit thought I didn't have
> permissions to do anything, nor did it ever pop up an authentication
> dialog.
This is not the fault of lightdm but of pam-kwallet. See this other
thread:
F27 problems with pam?
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
I think it's too early to assign blame. There clearly *is* some sort of
problem between lightdm/pam-kwallet/polkit, but where exactly and how to
address it has yet to be determined.
Formal bug tracking the issue is here, if you want to follow along:
https://bugzilla.redhat.com/show_bug.cgi?id=1581495
-- Rex
1:34 p.m.
2018-05-22 13:56 GMT-03:00 Rick Stevens <ricks(a)alldigital.com>:
And I just had lightdm screw up royally. First, upon login I got an
XFCE polkit error popup, which is singularly useless in telling you
anything you can troubleshoot with. Then, I had no access to the sound
hardware on my machine, nor could I enable/disable wireless, play
with firewall settings or anything else, as polkit thought I didn't have
permissions to do anything, nor did it ever pop up an authentication
dialog.
OMG, just what happened to me. So lightdm is to blame?
I wiped my xfce settings, reinstalled lightdm, xfce4-polkit and
anything
else I could think of to no avail. I finally just punted, disabled
lightdm and enabled lxdm and everything started working again.
I think I had another DM around. Will switch and see.
--
Martín Marqués
It’s not that I have something to hide,
it’s that I have nothing I want you to see
4:51 p.m.
On 05/28/2018 11:34 AM, Martín Marqués wrote:
2018-05-22 13:56 GMT-03:00 Rick Stevens
<ricks(a)alldigital.com>:
>
> And I just had lightdm screw up royally. First, upon login I got an
> XFCE polkit error popup, which is singularly useless in telling you
> anything you can troubleshoot with. Then, I had no access to the sound
> hardware on my machine, nor could I enable/disable wireless, play
> with firewall settings or anything else, as polkit thought I didn't have
> permissions to do anything, nor did it ever pop up an authentication
> dialog.
OMG, just what happened to me. So lightdm is to blame?
No it's not lightdm that's to blame, it's an interaction between lightdm
and pam-kwallet.
Quick fix: Edit your /etc/pam.d/lightdm file and comment out any line
referencing pam_kwallet.so or pam_kwallet5.so, log out and back in and
you should be fine. For some weird reason, lightdm includes this in its
configuration. My version of the above file:
[root@golem4 xxx]# cat /etc/pam.d/lightdm
#%PAM-1.0
# Note that pam-kwallet 5.12.4-2.fc27 is severely broken and will
# prevent xfce-polkit from working, so the lines referencing it are
# commented out here.
auth [success=done ignore=ignore default=bad] pam_selinux_permit.so
auth required pam_env.so
auth substack system-auth
-auth optional pam_gnome_keyring.so
#-auth optional pam_kwallet5.so
#-auth optional pam_kwallet.so
auth include postlogin
account required pam_nologin.so
account include system-auth
password include system-auth
session required pam_selinux.so close
session required pam_loginuid.so
session optional pam_console.so
-session optional pam_ck_connector.so
session required pam_selinux.so open
session optional pam_keyinit.so force revoke
session required pam_namespace.so
-session optional pam_gnome_keyring.so auto_start
#-session optional pam_kwallet5.so
#-session optional pam_kwallet.so
session include system-auth
session optional pam_lastlog.so silent
session include postlogin
There is a bugzilla on this:
https://bugzilla.redhat.com/show_bug.cgi?id=1581495
and even the latest version of the pam-kwallet RPM does not fix it--not
even for F28.
> I wiped my xfce settings, reinstalled lightdm, xfce4-polkit and
anything
> else I could think of to no avail. I finally just punted, disabled
> lightdm and enabled lxdm and everything started working again.
I think I had another DM around. Will switch and see.
That's drastic, but as I said above, disabling PAM's use of kwallet
when using lightdm fixes the issue. Note I'm an Xfce4 user, so disabling
pam_kwallet* may have consequences on KDM/Plasma desktops that I don't
know about.
I see that lxdm's PAM config does NOT include pam_kwallet*.so in its
default PAM config (I've not tried tried using pam_kwallet* in an lxdm
config). So, if your KDE/Plasma desktop works as expected using lxdm,
then commenting out the pam_kwallet*.so stuff in the lightdm config
shouldn't affect your desktop either.
YMMV
----------------------------------------------------------------------
- Rick Stevens, Systems Engineer, AllDigital ricks(a)alldigital.com -
- AIM/Skype: therps2 ICQ: 22643734 Yahoo: origrps2 -
- -
- Microsoft Windows: Proof that P.T. Barnum was right -
----------------------------------------------------------------------
11:35 a.m.
On Tue, 22 May 2018 17:35:04 +0100, Patrick O'Callaghan wrote:
[....]
Never mind. I have actually brought it up before now (can't be
bothered
to track down the post though it's probably on the KDE list) but no-one
else reacted so it doesn't appear to be common.
Fwiw, the pan item in the latest dnf upgrade worked the trick :
PC #1 and PC #3 (but, oddly, not PC #2) now boot to their accustomed GUI,
and from there launch Pan. All this *without* (at least yet) replacing
LightDM.
I guess that makes me mathematically 2/3 of a happy camper; but
in practice it's a lot more, inasmuch I now I have not only Pan back, but
also a working backup for PC #1.
Many, many thanks! Blessed be the developers!
--
/home/btth/sig/nqc
3:57 p.m.
On Mon, 21 May 2018 05:48:24 +0800, Ed Greshko wrote:
In actuality you just need one command.
systemctl -f enable sddm (or whatever new dm you want)
-f = force which means "When used with enable, overwrite any existing
conflicting symlinks."
and no need to use ".service"
Bless you, SIR, bless you again, and bless you some more!
First I launched dnfdragora on the one machine that still gives
me a graphical interface (by running F26 Rescue -- 26 no typo, *not* 28),
and used the GUI to install as many display managers as I could find.
I noticed that lxdm seemed to be included in many groups. So I
did dnf install with the console (from Ctrl-Alt-F2) on my #2 machine,
which had been denying me any GUI, for love nor electrons. Then systemctl
-f enable lxdm, and rebooted.
IT WORKED!!
It's not quite working on #3 yet, but I bet it will be in a day
or three. I've already stopped feeling like I'm riding a bicycle along
the edge of a cliff. Champagne is definitely indicated.
--
Beartooth Implacable, Curmudgeon On Line
Who shall gather the smoke of the dead wood burning,
Or behold the flowing years from the Sea returning?
4:22 p.m.
On 06/22/2018 01:57 PM, Beartooth wrote:
On Mon, 21 May 2018 05:48:24 +0800, Ed Greshko wrote:
> In actuality you just need one command.
>
> systemctl -f enable sddm (or whatever new dm you want)
>
> -f = force which means "When used with enable, overwrite any existing
> conflicting symlinks."
>
> and no need to use ".service"
Bless you, SIR, bless you again, and bless you some more!
First I launched dnfdragora on the one machine that still gives
me a graphical interface (by running F26 Rescue -- 26 no typo, *not* 28),
and used the GUI to install as many display managers as I could find.
I noticed that lxdm seemed to be included in many groups. So I
did dnf install with the console (from Ctrl-Alt-F2) on my #2 machine,
which had been denying me any GUI, for love nor electrons. Then systemctl
-f enable lxdm, and rebooted.
IT WORKED!!
It's not quite working on #3 yet, but I bet it will be in a day
or three. I've already stopped feeling like I'm riding a bicycle along
the edge of a cliff. Champagne is definitely indicated.
And just a warning: The lightdm display manager has issues with
pam_kwallet and pam_kwallet5 if they're installed. If you decide to use
lightdm, edit the /etc/pam.d/lightdm file and comment out any lines
that reference pam_kwallet*.so.
Since you're using lxdm you should be fine, but I just wanted to put
that warning out there. The bugzilla I opened on it hasn't been resolved
yet to my knowledge.
----------------------------------------------------------------------
- Rick Stevens, Systems Engineer, AllDigital ricks(a)alldigital.com -
- AIM/Skype: therps2 ICQ: 226437340 Yahoo: origrps2 -
- -
- If the enemy's in range...so are you! -
----------------------------------------------------------------------
10:41 a.m.
New subject: Could LightDM be bad?? Ooops!
On Fri, 22 Jun 2018 20:57:41 +0000, Beartooth wrote:
On Mon, 21 May 2018 05:48:24 +0800, Ed Greshko wrote:
> In actuality you just need one command.
>
> systemctl -f enable sddm (or whatever new dm you want)
>
> -f = force which means "When used with enable, overwrite any existing
> conflicting symlinks."
>
> and no need to use ".service"
Bless you, SIR, bless you again, and bless you some more!
First I launched dnfdragora on the one machine that still gives
me a graphical interface (by running F26 Rescue -- 26 no typo, *not*
28),
and used the GUI to install as many display managers as I could find.
I noticed that lxdm seemed to be included in many groups. So I
did dnf install with the console (from Ctrl-Alt-F2) on my #2 machine,
which had been denying me any GUI, for love nor electrons. Then
systemctl -f enable lxdm, and rebooted.
IT WORKED!!
It's not quite working on #3 yet, but I bet it will be in a day
or three. I've already stopped feeling like I'm riding a bicycle along
the edge of a cliff. Champagne is definitely indicated.
It did work. I swear it did. But now it just keeps giving me a
login screen, letting me type, and (after Enter) going back to the login
screen.
Somebody got another favorite to replace lxdm??
--
Beartooth Staffwright, Not Quite Clueless Power User
Remember I know little (precious little!) of where up is.
11:58 a.m.
On Fri, 22 Jun 2018 14:22:13 -0700, Rick Stevens wrote:
[....]
And just a warning: The lightdm display manager has issues with
pam_kwallet and pam_kwallet5 if they're installed. If you decide to use
lightdm, edit the /etc/pam.d/lightdm file and comment out any lines that
reference pam_kwallet*.so.
Since you're using lxdm you should be fine, but I just wanted to put
that warning out there. The bugzilla I opened on it hasn't been resolved
yet to my knowledge.
Done, with thanks!
I also did the same with my #2, the one I thought yesterday had
decided to straighten up and fly right. But there the commenting out
didn't help. That PC is still console only.
--
Beartooth Staffwright, Not Quite Clueless Power User
Remember I know little (precious little!) of where up is.
1:19 p.m.
New subject: Could LightDM be bad?? Ooops!
Allegedly, on or about 24 June 2018, Beartooth sent:
But now it just keeps giving me a login screen, letting me type, and
(after Enter) going back to the login screen.
I remember things like that, in the past, being associated with the
wrong permissions for the /tmp directory. The graphics system wanted
to write a lock file, and would get upset when it couldn't.
$ ls -ldZ /tmp
drwxrwxrwt. 16 root root system_u:object_r:tmp_t:s0 400 Jun 25 03:48 /tmp
--
[tim@localhost ~]$ uname -rsvp
Linux 4.16.11-100.fc26.x86_64 #1 SMP Tue May 22 20:02:12 UTC 2018 x86_64
Boilerplate: All mail to my mailbox is automatically deleted.
There is no point trying to privately email me, I only get to see
the messages posted to the mailing list.
Well somebody had to eat the last biscuit.
You're just miffed that it wasn't you.
2133
days inactive
2168
days old
45 comments
14 participants
participants (14)
-
Beartooth -
Beartooth
-
Ed Greshko -
Ed Greshko -
Francis.Montagnac＠inria.fr -
Martín Marqués -
Michael D. Setzer II -
Patrick O'Callaghan -
Ranjan Maitra -
Rex Dieter -
Rick Stevens -
stan -
Tim -
Tom Horsley