GDM does not starts after April 10 system reconfigure

  • Done
  • quality assurance status badge
Details
6 participants
  • Guillaume Le Vaillant
  • Ludovic Courtès
  • Maxim Cournoyer
  • Rene
  • sirgazil
  • R Veera Kumar
Owner
unassigned
Submitted by
R Veera Kumar
Severity
normal
R
R
R Veera Kumar wrote on 16 Apr 2020 02:20
(address . bug-guix@gnu.org)
20200416002020.GA381@tulip
I am using guix.git master branch.

I did guix system reconfigure on April 10 with gnome, mate, xfce and enlightenment service-types.
It was working. And next day I again did that and gdm does not starts anymore.

Initial console says:

New session c1 of user gdm.
Removed session c1.


/var/log/messages says:

gdm: Child process -325 was already dead.


/var/log/gdm/greeter.log says:

(EE)
Fatal server error:
(EE) Cannot open log file "/var/lib/gdm/.local/share/xorg/Xorg.pid-330.log"
(EE)
(EE)
Please consult the The X.Org Foundation support
for help.
(EE)
Unable to run X server

Any help would be appreciated!

Veera
R
R
R Veera Kumar wrote on 16 Apr 2020 03:14
SLIM is working though with gnome, mate, xfce and enlightenment
(address . 40652@debbugs.gnu.org)
20200416011457.GA1178@tulip
I checked that the installed system is working with SLIM.

All gnome, mate, xfce and enlightenment are working.

Perhaps the gdm patch for 1.1.0 release is causing the problem, which
changes xsession path.

Regards,
Veera
R
R
R Veera Kumar wrote on 16 Apr 2020 04:58
Local build of gdm without gdm-xsession.patch still fails to start
(address . 40652@debbugs.gnu.org)
20200416025819.GA350@tulip
Tried local build of gdm using master branch without gdm-xsession.patch still
fails to start it.

Updated master to 8080c03d14ccdd7897a902966ea4aeea8dbfd359 commit
Date: Wed Apr 15 19:44:41 2020 +0200

And done guix system reconfigure still gdm failing to start.

Regards,
Veera
S
S
sirgazil wrote on 16 Apr 2020 06:19
GDM does not starts after April 10 system reconfigure
(name . 40652)(address . 40652@debbugs.gnu.org)
17181363460.fa3c65a997322.4555192071860562497@zoho.com
I just pull and configured my system and get the same problem.

New session c1 of user gdm.
Removed session c1.

"/var/log/messages" shows the following messages between some ntpd messages:

Apr 15 22:31:06 localhost gdm: Child process -537 was already dead.
Apr 15 22:35:29 localhost gdm: Child process -558 was already dead.
Apr 15 22:39:14 localhost gdm: Child process -538 was already dead.



$ guix describe
Generation 71 Apr 15 2020 21:34:09 (current)
sirgazil-x 66d4b67
branch: master
commit: 66d4b677875c84d0b7a946376cd4885f202094eb
guix 8080c03
branch: master
commit: 8080c03d14ccdd7897a902966ea4aeea8dbfd359


---
R
R
R
R Veera Kumar wrote on 16 Apr 2020 07:31
#36924 way solves the problem for me
(address . 40652@debbugs.gnu.org)
20200416053137.GA787@tulip
The solution in bug #36924 solved the problem in my system.

Remove /var/lib/gdm and make a empty one instead.

Thanks Rene!

Regards,
Veera
G
G
Guillaume Le Vaillant wrote on 16 Apr 2020 09:32
(address . 40652@debbugs.gnu.org)
87pnc7yjgd.fsf@yamatai
R Veera Kumar <vkor@vkten.in> skribis:

Toggle quote (9 lines)
> The solution in bug #36924 solved the problem in my system.
>
> Remove /var/lib/gdm and make a empty one instead.
>
> Thanks Rene!
>
> Regards,
> Veera

I don't know if it's related, but recently I had GDM crashes at boot
after reconfiguring a system using gdm-service-type (generation n) to
make it use slim-service-type instead (generation n+1), and then
reconfiguring to gdm-service-type again (generation n+2).

The problem was that the 'gdm' user id number (or group id number) was
not the same in generations n and n+2, which prevented GDM from
accessing the '/var/lib/gdm' directory.

Changing the permissions with "chown -R gdm.gdm /var/lib/gdm" or
deleting '/var/lib/gdm' to force the creation of a new one with correct
permissions allowed GDM to work correctly again.
-----BEGIN PGP SIGNATURE-----

iHUEARYKAB0WIQQkUwKxurH4Z/3KlryPP4Yfgut6mgUCXpgKAgAKCRCPP4Yfgut6
mtmBAP95hrxX5hbG//wwQf2dLSWVj+ax1Rq9QwdvFma092G+YgEA/OZFytx8EOuu
d4W26EVGRE7MSxJdcCZN75k/4bN2BQw=
=3inm
-----END PGP SIGNATURE-----

L
L
Ludovic Courtès wrote on 16 Apr 2020 10:43
(name . Guillaume Le Vaillant)(address . glv@posteo.net)(address . 40652@debbugs.gnu.org)
877dyfkegp.fsf@gnu.org
Hi Guillaume,

Guillaume Le Vaillant <glv@posteo.net> skribis:

Toggle quote (9 lines)
> I don't know if it's related, but recently I had GDM crashes at boot
> after reconfiguring a system using gdm-service-type (generation n) to
> make it use slim-service-type instead (generation n+1), and then
> reconfiguring to gdm-service-type again (generation n+2).
>
> The problem was that the 'gdm' user id number (or group id number) was
> not the same in generations n and n+2, which prevented GDM from
> accessing the '/var/lib/gdm' directory.

When did that happen?

Commit a43e9157ef479e94c19951cc9d228cf153bf78ee (Sep. 2019) supposedly
ensures that /var/lib/gdm has proper ownership.

Thanks,
Ludo’.
G
G
Guillaume Le Vaillant wrote on 16 Apr 2020 10:57
(name . Ludovic Courtès)(address . ludo@gnu.org)(address . 40652@debbugs.gnu.org)
87o8rryfih.fsf@yamatai
Ludovic Courtès <ludo@gnu.org> skribis:

Toggle quote (21 lines)
> Hi Guillaume,
>
> Guillaume Le Vaillant <glv@posteo.net> skribis:
>
>> I don't know if it's related, but recently I had GDM crashes at boot
>> after reconfiguring a system using gdm-service-type (generation n) to
>> make it use slim-service-type instead (generation n+1), and then
>> reconfiguring to gdm-service-type again (generation n+2).
>>
>> The problem was that the 'gdm' user id number (or group id number) was
>> not the same in generations n and n+2, which prevented GDM from
>> accessing the '/var/lib/gdm' directory.
>
> When did that happen?
>
> Commit a43e9157ef479e94c19951cc9d228cf153bf78ee (Sep. 2019) supposedly
> ensures that /var/lib/gdm has proper ownership.
>
> Thanks,
> Ludo’.

I think it was around 2 weeks ago.
-----BEGIN PGP SIGNATURE-----

iHUEARYKAB0WIQQkUwKxurH4Z/3KlryPP4Yfgut6mgUCXpgd9gAKCRCPP4Yfgut6
mlsjAQDgeGj5sSrhzEIn+3B/Q+kGVUPEVQuMPWrKrpbK4j485gEA5ARBBD8qe3Ra
fVyEqNXnDjiwviuaKE7wk9AvZJnShgE=
=YR+e
-----END PGP SIGNATURE-----

G
G
Guillaume Le Vaillant wrote on 16 Apr 2020 11:24
(name . Ludovic Courtès)(address . ludo@gnu.org)(address . 40652@debbugs.gnu.org)
87mu7bye95.fsf@yamatai
Guillaume Le Vaillant <glv@posteo.net> skribis:

Toggle quote (25 lines)
> Ludovic Courtès <ludo@gnu.org> skribis:
>
>> Hi Guillaume,
>>
>> Guillaume Le Vaillant <glv@posteo.net> skribis:
>>
>>> I don't know if it's related, but recently I had GDM crashes at boot
>>> after reconfiguring a system using gdm-service-type (generation n) to
>>> make it use slim-service-type instead (generation n+1), and then
>>> reconfiguring to gdm-service-type again (generation n+2).
>>>
>>> The problem was that the 'gdm' user id number (or group id number) was
>>> not the same in generations n and n+2, which prevented GDM from
>>> accessing the '/var/lib/gdm' directory.
>>
>> When did that happen?
>>
>> Commit a43e9157ef479e94c19951cc9d228cf153bf78ee (Sep. 2019) supposedly
>> ensures that /var/lib/gdm has proper ownership.
>>
>> Thanks,
>> Ludo’.
>
> I think it was around 2 weeks ago.

Concerning the service extensions of gdm-service-type, is it guaranteed
that %gdm-activation will be run after %gdm-accounts and not before?
If it's not the case it could explain the problem...
-----BEGIN PGP SIGNATURE-----

iHUEARYKAB0WIQQkUwKxurH4Z/3KlryPP4Yfgut6mgUCXpgkVgAKCRCPP4Yfgut6
mh8HAQDcTyzfd0U8WgSDd3se3QirTUntpe2rlbo/z9LT9QHvQgEAh3SZrg1jOHuF
cnQr+OBFCcWE/Ttqvcs7KQo99OF8SQ8=
=AA+U
-----END PGP SIGNATURE-----

S
S
sirgazil wrote on 16 Apr 2020 16:05
GDM does not starts after April 10 system reconfigure
(name . 40652)(address . 40652@debbugs.gnu.org)
171834f66e9.c92ff7e8102434.8055446173045915124@zoho.com
Yes, Rene, that worked.

I deleted "/var/lib/gdm" and also ".local/share" and ".cache", which were 4.7 GiB and 2.9 GiB respectively.

Reading bug #36924 now I remember I've been through this before.
L
L
Ludovic Courtès wrote on 16 Apr 2020 23:03
Re: bug#40652: #36924 way solves the problem for me
(name . Guillaume Le Vaillant)(address . glv@posteo.net)(address . 40652@debbugs.gnu.org)
87eesnf8jd.fsf@gnu.org
Guillaume Le Vaillant <glv@posteo.net> skribis:

Toggle quote (31 lines)
> Guillaume Le Vaillant <glv@posteo.net> skribis:
>
>> Ludovic Courtès <ludo@gnu.org> skribis:
>>
>>> Hi Guillaume,
>>>
>>> Guillaume Le Vaillant <glv@posteo.net> skribis:
>>>
>>>> I don't know if it's related, but recently I had GDM crashes at boot
>>>> after reconfiguring a system using gdm-service-type (generation n) to
>>>> make it use slim-service-type instead (generation n+1), and then
>>>> reconfiguring to gdm-service-type again (generation n+2).
>>>>
>>>> The problem was that the 'gdm' user id number (or group id number) was
>>>> not the same in generations n and n+2, which prevented GDM from
>>>> accessing the '/var/lib/gdm' directory.
>>>
>>> When did that happen?
>>>
>>> Commit a43e9157ef479e94c19951cc9d228cf153bf78ee (Sep. 2019) supposedly
>>> ensures that /var/lib/gdm has proper ownership.
>>>
>>> Thanks,
>>> Ludo’.
>>
>> I think it was around 2 weeks ago.
>
> Concerning the service extensions of gdm-service-type, is it guaranteed
> that %gdm-activation will be run after %gdm-accounts and not before?
> If it's not the case it could explain the problem...

‘%gdm-activation’ would throw an exception if the “gdm” user didn’t
exist, so apparently it’s run before the activation snippet of
‘account-service-type’ (the ordering guarantee is not explicit.)

Hmm I wonder what I’m missing then. Would you like to try again?

Now, I think we should generalize this chown thing and apply it to all
the user accounts. ‘user-homes’ would chown recursively if needed or
use the newfangled shiftfs, like systemd-homed does¹.

Thoughts?

Ludo’.

G
G
Guillaume Le Vaillant wrote on 17 Apr 2020 11:06
(name . Ludovic Courtès)(address . ludo@gnu.org)(address . 40652@debbugs.gnu.org)
87y2qucwgn.fsf@yamatai
Ludovic Courtès <ludo@gnu.org> skribis:

Toggle quote (6 lines)
> ‘%gdm-activation’ would throw an exception if the “gdm” user didn’t
> exist, so apparently it’s run before the activation snippet of
> ‘account-service-type’ (the ordering guarantee is not explicit.)
>
> Hmm I wonder what I’m missing then. Would you like to try again?

I tried again and I wasn't able to reproduce the problem.
Maybe I did something weird with my config last time, but I can't
remember what it could have been...

Toggle quote (9 lines)
> Now, I think we should generalize this chown thing and apply it to all
> the user accounts. ‘user-homes’ would chown recursively if needed or
> use the newfangled shiftfs, like systemd-homed does¹.
>
> Thoughts?
> Ludo’.
>
> ¹ https://systemd.io/HOME_DIRECTORY/

A recursive chown for system accounts (with their home directory
somewhere in '/var') sounds like a good idea.

For user accounts (in '/home'), I guess it could be slightly annoying if
a user wants to set a specific group id to some of their files and if it
gets set back to the 'users' group at each system reconfiguration.
However it's probably not a very common use case, and if we only change
the files' uid, they could end up with an invalid gid anyway.
-----BEGIN PGP SIGNATURE-----

iHUEARYKAB0WIQQkUwKxurH4Z/3KlryPP4Yfgut6mgUCXplxqAAKCRCPP4Yfgut6
mhOVAPwM1/J6kBrTMoAlBEPRZlCAhLbUzOF4MsDJ0YyEgkzb7gD9FaF5S2dHHRvc
zu1CQOoGNn5yUp2L7RDt9NiyvMIVrwQ=
=UVAQ
-----END PGP SIGNATURE-----

L
L
Ludovic Courtès wrote on 18 Apr 2020 18:40
(name . Guillaume Le Vaillant)(address . glv@posteo.net)(address . 40652@debbugs.gnu.org)
87wo6c7nny.fsf@gnu.org
Hi,

Guillaume Le Vaillant <glv@posteo.net> skribis:

Toggle quote (12 lines)
> Ludovic Courtès <ludo@gnu.org> skribis:
>
>> ‘%gdm-activation’ would throw an exception if the “gdm” user didn’t
>> exist, so apparently it’s run before the activation snippet of
>> ‘account-service-type’ (the ordering guarantee is not explicit.)
>>
>> Hmm I wonder what I’m missing then. Would you like to try again?
>
> I tried again and I wasn't able to reproduce the problem.
> Maybe I did something weird with my config last time, but I can't
> remember what it could have been...

OK.

Toggle quote (18 lines)
>> Now, I think we should generalize this chown thing and apply it to all
>> the user accounts. ‘user-homes’ would chown recursively if needed or
>> use the newfangled shiftfs, like systemd-homed does¹.
>>
>> Thoughts?
>> Ludo’.
>>
>> ¹ https://systemd.io/HOME_DIRECTORY/
>
> A recursive chown for system accounts (with their home directory
> somewhere in '/var') sounds like a good idea.
>
> For user accounts (in '/home'), I guess it could be slightly annoying if
> a user wants to set a specific group id to some of their files and if it
> gets set back to the 'users' group at each system reconfiguration.
> However it's probably not a very common use case, and if we only change
> the files' uid, they could end up with an invalid gid anyway.

Right. The recursive chown would only happen if the home directory
itself has the wrong UID though, so that would still let you fiddle with
ownership of the files within it. Worth trying!

Thanks,
Ludo’.
M
M
Maxim Cournoyer wrote on 10 Jun 2022 02:40
Re: bug#40652: GDM does not starts after April 10 system reconfigure
(name . R Veera Kumar)(address . vkor@vkten.in)(address . 40652-done@debbugs.gnu.org)
87czfhgxkd.fsf_-_@gmail.com
Hello,

R Veera Kumar <vkor@vkten.in> writes:

Toggle quote (4 lines)
> The solution in bug #36924 solved the problem in my system.
>
> Remove /var/lib/gdm and make a empty one instead.

Thanks for letting us know.

Closing since #36924 is still open.

Thanks,

Maxim
Closed
?