Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
X logs me out abruptly
View unanswered posts
View posts from last 24 hours
View posts from last 7 days

Goto page 1, 2, 3  Next  
Reply to topic    Gentoo Forums Forum Index Desktop Environments
View previous topic :: View next topic  
Author Message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Fri Oct 27, 2017 9:01 pm    Post subject: X logs me out abruptly Reply with quote

I wonder if this belongs here.

As much as I could investigate, this only happened when I'm running VirtualBox 5.1.30, which I usually do, with a W10 vm.

As I said, suddenly I am at the login DM screen, sddm in this case, with all my open apps crashed. The logs don't tell much, except,:
Code:
Xorg.0.log:
[  3994.484] (II) Server terminated successfully (0). Closing log file.

kernel.log
Oct 27 17:39:08 mypc kernel: [ 2557.432495] SoftwareVsyncTh (7163) used greatest stack depth: 11392 bytes left

sddm.log
17:39:08.084] (II) HELPER: [PAM] Closing session
[17:39:08.088] (II) HELPER: [PAM] Ended.
[17:39:08.089] (II) DAEMON: Auth: sddm-helper exited successfully
[17:39:08.089] (II) DAEMON: Socket server stopping...
[17:39:08.089] (II) DAEMON: Socket server stopped.
[17:39:08.089] (II) DAEMON: Display server stopping...
[17:39:08.823] (II) DAEMON: Display server stopped.
[17:39:08.824] (II) DAEMON: Running display stop script  "/usr/share/sddm/scripts/Xstop"
[17:39:08.827] (II) DAEMON: Removing display ":0" ...
[17:39:08.827] (II) DAEMON: Adding new display on vt 7 ...
[17:39:08.827] (II) DAEMON: Loading theme configuration from ""
[17:39:08.827] (II) DAEMON: Display server starting...

This is important to me. I can't work on the machine without fixing this. Any help will be much appreciated.


Last edited by jssilva on Sat Oct 28, 2017 4:28 pm; edited 1 time in total
Back to top
View user's profile Send private message
krinn
Watchman
Watchman


Joined: 02 May 2003
Posts: 7470

PostPosted: Sat Oct 28, 2017 12:16 am    Post subject: Reply with quote

good news, well maybe bad for you, but the information you provide just show everything is ok.
either provide good logs (wgetpaste) or see if the cat has jump on the stop button
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Sat Oct 28, 2017 3:41 pm    Post subject: Reply with quote

krinn wrote:
good news, well maybe bad for you, but the information you provide just show everything is ok.
either provide good logs (wgetpaste) or see if the cat has jump on the stop button

Adopting your good-humor spirit, thank you for making me see that I'm the lucky owner of an unreliable system, where everything pretends to be ok but is not :D
Fortunately, I didn't delete my other partition where I keep my old system of another distro, with almost the same applications and versions, and which is running without any problems. Although being also an openrc system, I'm only migrating to Gentoo because of long-term worries.

Concerning the logs, the abrupt logout event took place at 2017.10.27-17:39:08 so I'm pasting the logs between the boot of that session at 16:56:42 and the reboot which took place immediately after the logout at 17:39:19 (I reboot from the DM).

The dhcp requests from 10.xxx.xxx.xxx, I think are coming from the virtualbox vm, which net adapter is configured as NAT, and may be relevant to the phenomenon.

Syslog: https://paste.pound-python.org/show/CgITVpfWUzBsMRbmabnt/
Kernel: https://paste.pound-python.org/show/EmwJUjwp3ELUyA6OLPsr/
Messages: https://paste.pound-python.org/show/fpTRUONnJzUqx1s0lCn0/
auth: https://paste.pound-python.org/show/3pTBUbQUlrlG5wbnJMc6/

The only suspicious thing I can see is in auth. Xorg.0.log is gone, but it looked clean as you can see in the op. If you need something else, please just ask.

By the way, thank you for helping.

PS: I don't have a cat, dog or baby; will a ghost do?
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Sat Oct 28, 2017 4:27 pm    Post subject: Reply with quote

Some more information: booted, waited, nothing happened. Launched VirtualBox and few minutes later, happened again. It is definitely linked to VB.

As much as I can see, it is simultaneous with this auth.log: Unregistered Authentication Agent for unix
Code:
Oct 28 17:05:37 mypc polkitd[6195]: Unregistered Authentication Agent for unix-session:2 (system bus name :1.30, object path /org/gnome/PolicyKit1/AuthenticationAgent, locale en_GB.utf8) (disconnected from bus)
Oct 28 17:05:37 mypc sddm-helper: pam_unix(sddm:session): session closed for user myuser
Oct 28 17:05:39 mypc sddm-helper: pam_unix(sddm-greeter:session): session opened for user sddm by (uid=0)
Oct 28 17:05:51 mypc sddm-helper: pam_unix(sddm-greeter:session): session closed for user sddm


Does anybody know how I can trace this, please?
Back to top
View user's profile Send private message
hhfeuer
Apprentice
Apprentice


Joined: 28 Jul 2005
Posts: 185

PostPosted: Sat Oct 28, 2017 5:25 pm    Post subject: Reply with quote

From dmesg:
radeon: switched on
Means, you're switching from integrated to discrete GPU which logs you out. This doesn't tell why this gets switched, of course.
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Sat Oct 28, 2017 6:21 pm    Post subject: Reply with quote

hhfeuer wrote:
From dmesg:
radeon: switched on
Means, you're switching from integrated to discrete GPU which logs you out. This doesn't tell why this gets switched, of course.

I wished that was it, but it's not. That's not the cause, it's a consequence. It's a service that I made which, at boot, switches to the intel GPU and shuts-off radeon to save power. And, at shut-off, or suspend, or logout, switches both GPU's ON to avoid shut-off delays. That's what you are seing, it's called vgaswitcheroo if you want to google for it.

Anyway, thanks for your interest.
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Fri Nov 03, 2017 10:25 pm    Post subject: Reply with quote

Well, one week past, the current status is that I've been running the system intermitently, trying to pin-point the problem.

I already know it's not related to bluetooth because I blacklisted the module. It also happened once that I had VirtualBox turned off. On the last event, the logs are totally silent, which I find absolutely amazing, with this exception on auth.log:
Code:
Nov  3 21:49:56 mypc polkitd[6551]: Unregistered Authentication Agent for unix-session:12 (system bus name :1.115, object path /org/gnome/PolicyKit1/AuthenticationAgent, locale en_GB.utf8) (disconnected from bus)


It's not possible to trust this system, saving my work every minute, always stressed by never knowing when it's going to happen, fearing that my filesystem won't resist next kick-out, I guess you got the picture.

So, I'm now parking the partition and go back to my former distro next door. I'm sure I would have liked Gentoo. Thank you for all your help. As I said somewhere else in this forum, it's good to know you're not alone in the dark.
Back to top
View user's profile Send private message
Ant P.
Watchman
Watchman


Joined: 18 Apr 2009
Posts: 6920

PostPosted: Fri Nov 03, 2017 11:02 pm    Post subject: Reply with quote

The thing here is that X was working perfectly fine all along. It's exiting cleanly once the only thing keeping it running — your choice of DE, GNOME — has crashed. You're right that the system is untrustworthy, but you'd get further if your blame was aimed correctly to begin with. For one thing, you're running it without systemd which isn't an upstream-supported configuration (nor will it be on whatever distro you go back to). You should also get those constant CPU overheating warnings in dmesg looked at.
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Sat Nov 04, 2017 12:08 am    Post subject: Reply with quote

Ant P. wrote:
— your choice of DE, GNOME —

Thank you for your advice.

However, it's based on a wrong assumption.

I wonder where you got the idea that I'm running Gnome; I'm not, my DE is Xfce4 and so I'm not worried with upstream support (of gnome/systemd).

Moreover, systemd is out of the question for me. That's one of the reasons that brought me to Gentoo.
Back to top
View user's profile Send private message
LIsLinuxIsSogood
Veteran
Veteran


Joined: 13 Feb 2016
Posts: 1179

PostPosted: Sat Nov 04, 2017 12:30 am    Post subject: Reply with quote

Quote:
I'm running VirtualBox 5.1.30, which I usually do, with a W10 vm.


Blame Microsoft!!!! Just kidding, you can probably blame polkit.

The thing about using any desktop management or environment (same thing) is it doesn't matter if it's Xfce or something else, they all still rely on outdated and generally not that reliable software when integrating.

Polkit in auth log seems f*d and my suggested solution is to question whether the benefit of a particular session login is necessary for the services your are running. For example, can the VM be operated without any authentication with the linux system and thereby somehow it won't cause this crash. Sometimes software has some settings like this in order to bypass an otherwise faulty system like that.

But, if you are after making such a system work then I would focus on the elogin daemon since that is giving you many errors for your settings, also this one:

Oct 27 17:02:39 mypc pkexec[7685]: myuser: Executing command [USER=root] [TTY=unknown] [CWD=/] [COMMAND=/usr/sbin/xfpm-power-backlight-helper --set-brightness 6584]

And, keep your head up because the known problems are just beginning.
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Sat Nov 04, 2017 12:58 am    Post subject: Reply with quote

Thank you for your extense comment but I'm not sure I can extract a concrete course of action from it.

VirtualBox is running under myuser, with gid vboxusers, as advised. The W10 VM hosted by it, is sandboxed from the Gentoo system, is it not? I don't understand "can the VM be operated without any authentication with the linux system" probably because English is not my birth language.

Furthermore, I already said somewhere that it already happened, at least once, with VBox stopped. So, I think it could be more like a stress that VBox places on the system, but that can also happen under other circumstances.

And polkit, if it is suspicious, is there any way of getting rid of it?

And why is the setbrightness an error? Shouldn't it be run by pkexec? (polkit, isn't it)?

I feel a bit lost. My other distro, being binary canned, doesn't place all these questions. I begin to wonder if in fact Gentoo is for me. I'm a programmer, yes, but my code is 'c' firmware for very small chips, all under 50KB; this is not my league, I think.
Back to top
View user's profile Send private message
LIsLinuxIsSogood
Veteran
Veteran


Joined: 13 Feb 2016
Posts: 1179

PostPosted: Sat Nov 04, 2017 5:25 am    Post subject: Reply with quote

I can see one of the issues is that you maybe haven't had the experience with the X11 software. Make sure to take a look at the guide on the wiki, and then please provide the Xorg log from a time that vmbox crashed, if possible.
Also, perhaps running Vmbox as root is a good initial test to see if that works instead of running as "myuser".
Where the only issue there could be allowing root to access the display, but that is an easy problem to solve, see xauth or xhost for that one.
If the test is better, then the next step would be fairly simply done with a slightly different approach that doesn't rely on polkit, just opting to use a more direct thing to authenticate like sudo or gksudo (for graphical environment) to invoke a script that starts the program that way. That script might look like "env VAR1=x VAR2=y sudo exec /bin/vmbox" and would be stored in a .desktop file on your system.

Quote:
Moreover, systemd is out of the question for me. That's one of the reasons that brought me to Gentoo.

I tend to agree with this statement, for the camp that is all about don't change it if it works. Plus openrc does provide the scriptable interface to start and stop services in a much more elegant way than killing these things called daemon from a process tree(ewh!)

Quote:
The W10 VM hosted by it, is sandboxed from the Gentoo system, is it not?

I'm not sure I would agree with that, but then again since you didn't really get my point I won't go on. I realize my post may seem laze or off-topic even, but after spending a bunch of time learning to write the rules for the consolekit/dbus subsystem that is polkit, with programs like pkexec, I can honestly say they present no real benefit overall (except for sometimes it just looks good if it works).

Quote:
And polkit, if it is suspicious, is there any way of getting rid of it?

(if you are up to it) try removing it, and seeing how the Desktop manager responds. You can always re-install the package. And if you don't want to make that huge an effort, but still are willing to put time into it, then look at man polkit and man pkexec, as those will describe how to create such rules and where to store them for authenticating for the application that isn't working. It is still no guarantee that it fix the problem of Vmbox crashing the X server, but will help resolve the authenticating issue and removes that aspect of the equation.

Quote:
Thank you for your extense comment but I'm not sure I can extract a concrete course of action from it.

My belief is you might simply try and unmerge polkit and see how your DM responds. It will probably not like that, and some applications may just stop working, which is a possibility!! The better solution to this could be to switch to a session that is not dependent on consolekit, but this will take some definite catching up to speed about things in terms of replacing different aspects offfered through those (such as font, theming, and whatever else all those DM's seem to embody but which actually present problems for some some people more than providing any value for what it is worth, those are my thoughts). For what it is worth, with some basic approach to managing the GUI using a login manager, and window manager, that requires you to go through some further installations for things like a panel, pager and system monitor (I recommend conky) it can pretty much mimic the feel of it, and is much more friendly to you as the custom user of the system.
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Sat Nov 04, 2017 11:02 am    Post subject: Reply with quote

That's an extensive lecture! I read it several times, I'm still digesting, and I'm sure I'll read it many times more once I start to put your advices in practice; one can tell you have a deep knowledge of the subject. Thank you for the time you decided to spend with me.

After your advice, I went to take a peek on my former distro which is also openrc (actually I'm writing on it now), and it uses neither consolekit nor polkit (or policykit). So I guess it will be a good start to see how they do things, and try to mimic.

EDIT: The last statement is just partly true: it doesn't use consolekit but does use polkit. However, xfce4-power-manager which is crucial for my system, doesn't depend on consolekit or systemd like gentoo's, it depends on upower, libxfce4ui, libnotifier and hicolor-icon-theme. Why can't gentoo do the same?

Thank you once more, I'll be posting my progress.
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Mon Nov 06, 2017 11:01 am    Post subject: Reply with quote

I spent all my weekend on this thing. First of all, let me tell you that, although this is my free-time, I have other things that I want and must do on my free-time besides tuning my gentoo system to be able to use it as my main workstation. That's why I've decided to definitely shut-off and park the system. It will stay just as an experimental campus.

So, I started by investigating the possibility of trashing polkit, which I concluded to be almost impossible and still have a useful system (I don't want to go back to windows 3.1). I would have to find ways around NetworkManager, xfce4-power-manager, udisks and a lot more things that I need/want in my system.

Then I turned on the USE debug flag and recompiled more than 200 packages, just to find out that VBox wouldn't start with, I think, qt debug on. And VBox is one of suspects.

After experimenting several things, I lost track of the system and had to restore a partition image that I had and follow from there.

So, turned on debug on polkit and derivatives, pam, xfce4-power-manager, bluez, pm-utils, dbus, elogind and consolekit. Launched VBox and waited for the crash. It was another disappointment, the logs are almost silent about it, despite all the debug turned on that, for instance, causes dozens o terminal log lines when I sudo. The logs say only, like before,
auth:
Code:

Nov  6 09:59:53 mypc polkitd[6495]: Unregistered Authentication Agent for unix-session:2 (system bus name :1.29, object path /org/gnome/PolicyKit1/AuthenticationAgent, locale en_GB.utf8) (disconnected from bus)
Nov  6 09:59:53 mypc sddm-helper: pam_unix(sddm:session): session closed for user myuser

daemon:
Code:

Nov  6 09:59:53 mypc bluetoothd[6616]: Endpoint unregistered: sender=:1.41 path=/MediaEndpoint/A2DPSource
Nov  6 09:59:53 mypc bluetoothd[6616]: Endpoint unregistered: sender=:1.41 path=/MediaEndpoint/A2DPSink
Nov  6 10:00:11 mypc init: Switching to runlevel: 6

At least on this system, the only package that depends on consolekit, one of my suspects, is xfce4-power-manager, which ebuild requires either that or systemd. But it must be possible to do it otherwise because my former distro does not require either. So, I leave a suggestion to the devs to investigate the possibility of getting rid of consolekit.

As I said, I'm shuting-off now and go back to my former distro (also openrc but mostly canned binaries). I won't mark the thread as solved because it isn't.

Thank you all, again
Back to top
View user's profile Send private message
LIsLinuxIsSogood
Veteran
Veteran


Joined: 13 Feb 2016
Posts: 1179

PostPosted: Mon Nov 06, 2017 9:26 pm    Post subject: Reply with quote

What happened to the kernel settings you needed to change? Anything come of it?

Quote:
I don't want to go back to windows 3.1


Why is that your only other option, if you are living in the 1990's then I would understand. This is the 10's almost the 20's though.

Quote:
I would have to find ways around NetworkManager, xfce4-power-manager, udisks and a lot more things that I need/want in my system.


In some way shape or form this is really the issue that you are facing, because yes some other programs that really on them are helpful and you don't want to remove those programs. Anway, there was probably no need to reinstall a bunch of packages I don't see how that would have helped. Or hurt actually. But I do apologize for putting you through that!

My guess would be that you could do something to bypass the polkit rules for the VM you are using. Let me look into it a bit more and I will tell you what I find out in terms of alternatives for relying on *kit with VirtualBox.

In the meantime, could you please post these as well:

emerge --info app-emulation/virtualbox

Thanks!

EDIT: Have you tried QEMU? It is the application I am using and it may not provide any drawbacks, as another open source and very solid virtual machine emulator. Have a look here: https://wiki.gentoo.org/wiki/QEMU


Last edited by LIsLinuxIsSogood on Mon Nov 06, 2017 10:56 pm; edited 1 time in total
Back to top
View user's profile Send private message
LIsLinuxIsSogood
Veteran
Veteran


Joined: 13 Feb 2016
Posts: 1179

PostPosted: Mon Nov 06, 2017 9:37 pm    Post subject: Reply with quote

Did you do the step from the wiki to add your user to the group, vboxusers?

If you did then I don't think that polkit should be necessary for your VM, sorry to have this epiphany so late in the process, I should have checked to see the progress sooner on your end.

Code:
root #gpasswd -a <user> vboxusers
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Mon Nov 06, 2017 11:14 pm    Post subject: Reply with quote

Thank you once again for helping.
LIsLinuxIsSogood wrote:
What happened to the kernel settings you needed to change? Anything come of it?

I almost didn't touch the kernel, only on bluetooh setting, mostly following the wiki, because it was wrong.

Quote:
I don't want to go back to windows 3.1

Why is that your only other option, if you are living in the 1990's then I would understand. This is the 10's almost the 20's though.

Now you made me smile. If you didn't understand that was rhetoric, a way of getting the message through, is because English is not my birth language. The fact is I started on msdos (edit: actually before, on cp/m), and then windows 3.1. I might still have a set of installation diskettes laying around.

What I meant is that my system without NetworkManager, udisks, xfce4-power-manager and some other tools dependent on polkit, is as useable as windows 3.1, well, almost, and that's of course my view.
Code:
# equery d polkit
 * These packages depend on polkit:
gnome-base/gconf-3.2.6-r4 (policykit ? sys-auth/polkit)
gnome-base/gvfs-1.32.1 (policykit ? sys-auth/polkit)
gnome-extra/nm-applet-1.8.2 (>=sys-auth/polkit-0.96-r1)
gnome-extra/polkit-gnome-0.105-r1 (>=sys-auth/polkit-0.102)
net-firewall/ufw-frontends-0.3.2-r5 (policykit ? sys-auth/polkit)
net-misc/modemmanager-1.6.4 (policykit ? >=sys-auth/polkit-0.106[introspection])
net-misc/networkmanager-1.8.4 (policykit ? >=sys-auth/polkit-0.106)
net-print/hplip-3.17.10 (policykit ? sys-auth/polkit)
net-wireless/blueman-2.0.4 (policykit ? sys-auth/polkit)
sys-apps/accountsservice-0.6.45 (sys-auth/polkit)
sys-auth/consolekit-1.2.0 (policykit ? >=sys-auth/polkit-0.110)
sys-auth/elogind-234.4 (policykit ? sys-auth/polkit)
sys-auth/polkit-qt-0.112.0_p20160416-r1 (>=sys-auth/polkit-0.103)
sys-block/gparted-0.29.0 (policykit ? sys-auth/polkit)
sys-fs/udisks-1.0.5-r1 (>=sys-auth/polkit-0.110)
sys-fs/udisks-2.7.3 (>=sys-auth/polkit-0.110)
xfce-base/xfce4-session-4.13.0 (policykit ? >=sys-auth/polkit-0.102)
xfce-extra/xfce4-power-manager-1.6.0-r1 (policykit ? >=sys-auth/polkit-0.112)

Quote:
I would have to find ways around NetworkManager, xfce4-power-manager, udisks and a lot more things that I need/want in my system.

In some way shape or form this is really the issue that you are facing, because yes some other programs that really on them are helpful and you don't want to remove those programs. Anway, there was probably no need to reinstall a bunch of packages I don't see how that would have helped. Or hurt actually. But I do apologize for putting you through that!

First of all, you don't have to apologize for wanting to help me, much on the contrary, I'm very grateful. And then, the compilation became necessary to turn-on the debug USE flag on all those packages, in the hope that one of them decided to tell me what was happening. That was the way I saw it.
Quote:
My guess would be that you could do something to bypass the polkit rules for the VM you are using. Let me look into it a bit more and I will tell you what I find out in terms of alternatives for relying on *kit with VirtualBox.

As I said, I tried. But VirtualBox refused to start with qt (I think) in debug mode. And then, when I started trying to unset the debug flag just on the packages VBox wanted, the system was already so twisted that I had to restore the image.

Anyway, later on the system crashed with VBox fully stopped (but not the loaded modules) and that makes me think that it's not the culprit.

Quote:
In the meantime, could you please post these as well:

emerge --info app-emulation/virtualbox

Yes, by all means: https://paste.pound-python.org/show/n13nWlEfuaHrn34VjOdD/

And, concerning your next post, yes I did join vboxusers right after emerging the package:
Code:
$ groups
sys adm lp wheel uucp audio cdrom video cdrw usb input users vboxusers lpadmin wireshark scanner plugdev myuser

Please note that I'm not in the disk group because I read that this is a good security measure.
And concerning polkit, if VBox does not require it, well it is not important because I'm convinced I can't live without it, or else, back to windows 3.1 :D

I've decided to use my former distro to work, mostly during the day, and to use the Gentoo partition in the evening when a crash is not so problematic. And right now I'm conducting an experiment, using just startx to take sddm out of the equation.

Thank you so much for your extended help.
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Tue Nov 07, 2017 12:12 am    Post subject: Reply with quote

Just crashed using startx from the console under myuser:
Xorg.0.log: https://paste.pound-python.org/show/5a98aJ54zYg4zHdQgKeZ/

This time X really crashed, as can be seen on the log bottom, the console reappeared but frozen showing only the begin of X terminal output after the launch. Only responsive key was power-off (one fast click, equivalent to Ctrl-Alt-Del, I think).

I was using VBox because I know from experience that this is the quickest way to crash; but, as I said, it already happened with VBox stopped.

But, looking at the other logs, namely daemon, one can see that the os was running and terminating orderly:
daemon.log
Code:
Nov  6 23:41:59 mypc ntpd[7848]: 999.999.999.999 local addr 192.168.100.11 -> <null>
Nov  6 23:42:00 mypc laptop-mode: disabled, not active
Nov  6 23:42:00 mypc laptop-mode: Failed to re-set power saving mode for wireless card
Nov  6 23:42:00 mypc bluetoothd[7601]: Terminating
Nov  6 23:42:00 mypc bluetoothd[7601]: Stopping SDP server
Nov  6 23:42:00 mypc bluetoothd[7601]: Exit
Nov  6 23:42:00 mypc NetworkManager[7513]: <info>  [1510011720.5698] device (FF:FF:FF:FF:FF:FF): state change: disconnected -> unmanaged (reason 'removed', internal state 'removed')
Nov  6 23:42:01 mypc acpid: exiting
Nov  6 23:42:01 mypc NetworkManager[7513]: <info>  [1510011721.4139] caught SIGTERM, shutting down normally.
Nov  6 23:42:01 mypc NetworkManager[7513]: <info>  [1510011721.4154] device (wlan0): state change: activated -> deactivating (reason 'unmanaged', internal state 'managed')
Nov  6 23:42:01 mypc NetworkManager[7513]: <info>  [1510011721.4157] manager:NetworkManager state is now DISCONNECTING

EDIT: As far as I can tell, the crash took place at 23:41:56, but why?
Syslog:
Code:
Nov  6 23:41:56 mypc shutdown[11015]: shutting down for system halt
Nov  6 23:41:57 mypc init: Switching to runlevel: 0
Nov  6 23:41:57 mypc shutdown[11031]: shutting down for system halt
Nov  6 23:41:59 mypc saned[7886]: bailing out, waiting for children...
Nov  6 23:41:59 mypc saned[7886]: bail_out: all children exited
Nov  6 23:41:59 mypc ntpd[7848]: ntpd exiting on signal 15 (Terminated)
Back to top
View user's profile Send private message
LIsLinuxIsSogood
Veteran
Veteran


Joined: 13 Feb 2016
Posts: 1179

PostPosted: Tue Nov 07, 2017 6:16 am    Post subject: Reply with quote

(Edited)

Please attach any of the missing Xorg log file, unless that's it.

Otherwise, as previously mentioned and we should recap everything at this point, because the Xorg log looks fine!

Perhaps the confusion is something about the loss of your display and unless you can find any other messages in the Xorg log (they usually say EE when there is an error)

NOTE: I did however see that the fonts and touchpad settings look messy, so later you can clean those up (but as for the cause of a crash those are highly unlikely, maybe the touchpad but probably not even that and for sure not without some further sign of error!).

Where does that leave everything?

Here's some ideas, the forum is here as a support and it is a good idea to take every suggestion seriously. Therefore, in response to an earlier suggestion you should certainly let it be known if you see MORE of those CPU temp issues. I had overlooked initially, but that is possibly indicative of hardware issues. The core temp for your CPU is easily monitored, and it should be possible to "benchmark" it. Also as far as your crash is concerned it is highly likely (now that I reviewed again the System log from OP you made) those are related. Can you look to see the cause of this last crash, in the syslog? If it has CPU temp, then you can bet it was that both times!

Let's add to the picture now...if the machine you are on is a laptop or desktop? Can you check inside to see if everything is secured in terms of fans, cpu, etc.

How old of a computer is it? Maybe it is time to upgrade the machine. But you did say there is another OS that works fine on it so what's the deal with that, idk.
Back to top
View user's profile Send private message
LIsLinuxIsSogood
Veteran
Veteran


Joined: 13 Feb 2016
Posts: 1179

PostPosted: Tue Nov 07, 2017 6:43 am    Post subject: Reply with quote

You need to start by checking the functioning of your graphics card. And then if it is all right then move on to the cpu to check that. These are the kinds of problems that can lead to crashes with no logs.
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Tue Nov 07, 2017 10:45 am    Post subject: Reply with quote

LIsLinuxIsSogood wrote:

Please attach any of the missing Xorg log file, unless that's it.

That's it. This morning I took one ended normally and started in the same conditions, by startx, for comparison: https://paste.pound-python.org/show/n6oMEsVDA4TZ4BURxRcN/

Crash logs:
auth.log: https://paste.pound-python.org/show/TSGpX5iR335PU3JrwbG8/
syslog: https://paste.pound-python.org/show/xgaTk00wwDyVTNtvxkLm/

Quote:
Otherwise, as previously mentioned and we should recap everything at this point, because the Xorg log looks fine!

Doesn't it end abruply? I mean, comparing with the one above.

Quote:
NOTE: I did however see that the fonts and touchpad settings look messy, so later you can clean those up (but as for the cause of a crash those are highly unlikely,

Yes, I confess some lazyness to fix that.

Quote:
Here's some ideas, the forum is here as a support and it is a good idea to take every suggestion seriously. Therefore, in response to an earlier suggestion you should certainly let it be known if you see MORE of those CPU temp issues.

And it was, as you can see in the logs. I'm sorry I didn't mention it. Again by lazyness, or too much to do, there were tons of dust piled between the fins of the heat permutators. I also took off the mainboard and heatsink and replaced the paste on the cpu and gpu. It now runs cool, as cool as a Mac can be. As you know, it's aluminium and the whole body is a heatsink.

Quote:
I had overlooked initially, but that is possibly indicative of hardware issues.
Let's add to the picture now...if the machine you are on is a laptop or desktop? Can you check inside to see if everything is secured in terms of fans, cpu, etc.

I don't think is hardware, the temperature is now normal and the problem didn't stop.
Moreover, the machine is perfect in terms of hardware.

Quote:
How old of a computer is it? Maybe it is time to upgrade the machine. But you did say there is another OS that works fine on it so what's the deal with that, idk.

The machine is a laptop, MacBook pro 8.2 early 2011 with SandyBridge i7-2675QM cpu @ 2.2GHz, 16GB Ram, 2TB ssd and I am against waste: "if it works, keep on using it".
Furthermore, it has on the other partitions: macos Sierra, Manjaro openrc and Xubuntu 14.04, all of them without any problems. A slight exception though, the Manjaro, which is my daily system, had once, just once on everyday use, the same problem about one month ago. But it works intensively perhaps 14 hours a day.

Concerning hardware, just to tune your trust on my observations, my professional life has been designing electronic equipment, hardware and firmware, although not so much these days; I do more management now. In the '70s (that's how old I am) I designed and built, at the component level (cpu, memories, crystals, i/o chips), not board level, my own 8080 cp/m computer with my bios and bootloader programmed in assembly; it was about the time of the Altair 8080, before the TRS80, and many years before the IBM PC. That was fun.

Thank you again for your extended help.


Last edited by jssilva on Tue Nov 07, 2017 11:41 am; edited 2 times in total
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Tue Nov 07, 2017 10:47 am    Post subject: Reply with quote

LIsLinuxIsSogood wrote:
You need to start by checking the functioning of your graphics card. And then if it is all right then move on to the cpu to check that. These are the kinds of problems that can lead to crashes with no logs.

As I said, there are three more os's that work on that gpu without problems.
Back to top
View user's profile Send private message
LIsLinuxIsSogood
Veteran
Veteran


Joined: 13 Feb 2016
Posts: 1179

PostPosted: Tue Nov 07, 2017 10:37 pm    Post subject: Reply with quote

Aha! It's a mac, this is why I've been terrified of trying to go through the installation on a tower that I have from 2005, not because the technology is outdated, which it isn't actually (still has enough RAM and processor power to do some heavy lifting, especially for the use of being a basic server for websites and stuff like that). But the idea of getting the right kernel settings, and configurations for things like X11 seem very difficult when it comes to using these generally not so generic machines. My suggestion is to continue researching the causes of it by looking particularly at either the kernel or X configuration file for the right settings for your particular macbook.

By the way, you are mentioning startx a lot, but have you checked in your home folders for the session files, like .xsession-errors could help, also just .Xsession and .xinitrc are common places to store things that will be run in the X environment. But more importantly, from the looks of this last log (and I hadn't seen it previously) but the keyboard settings is something I would pay especially close attention to as well.
Back to top
View user's profile Send private message
LIsLinuxIsSogood
Veteran
Veteran


Joined: 13 Feb 2016
Posts: 1179

PostPosted: Tue Nov 07, 2017 10:41 pm    Post subject: Reply with quote

Can you please include some information about the laptop's hardware, a good place to start would be:
Code:
lspci -v -s 00:02.0


and
Code:
lscpu
Back to top
View user's profile Send private message
jssilva
Tux's lil' helper
Tux's lil' helper


Joined: 27 Sep 2017
Posts: 106

PostPosted: Tue Nov 07, 2017 11:32 pm    Post subject: Reply with quote

LIsLinuxIsSogood wrote:
Can you please include some information about the laptop's hardware, a good place to start would be:

This is the integrated intel GPU i965 (uses i915 driver)
Code:
# lspci -v -s 00:02.0
00:02.0 VGA compatible controller: Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller (rev 09) (prog-if 00 [VGA controller])
   Subsystem: Apple Inc. 2nd Generation Core Processor Family Integrated Graphics Controller
   Flags: bus master, fast devsel, latency 0, IRQ 45
   Memory at b0000000 (64-bit, non-prefetchable) [size=4M]
   Memory at a0000000 (64-bit, prefetchable) [size=256M]
   I/O ports at 3000 [size=64]
   Capabilities: [90] MSI: Enable+ Count=1/1 Maskable- 64bit-
   Capabilities: [d0] Power Management version 2
   Capabilities: [a4] PCI Advanced Features
   Kernel driver in use: i915
   Kernel modules: i915


Code:
# lscpu
Architecture:        x86_64
CPU op-mode(s):      32-bit, 64-bit
Byte Order:          Little Endian
CPU(s):              8
On-line CPU(s) list: 0-7
Thread(s) per core:  2
Core(s) per socket:  4
Socket(s):           1
NUMA node(s):        1
Vendor ID:           GenuineIntel
CPU family:          6
Model:               42
Model name:          Intel(R) Core(TM) i7-2675QM CPU @ 2.20GHz
Stepping:            7
CPU MHz:             2195.032
CPU max MHz:         3100.0000
CPU min MHz:         800.0000
BogoMIPS:            4390.06
Virtualization:      VT-x
L1d cache:           32K
L1i cache:           32K
L2 cache:            256K
L3 cache:            6144K
NUMA node0 CPU(s):   0-7
Flags:               fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic popcnt tsc_deadline_timer aes xsave avx lahf_lm tpr_shadow vnmi flexpriority ept vpid xsaveopt dtherm ida arat pln pts

I think you'll also need to know the kernel parameters used. If the Mac firmware detects that macos is not being booted, but another os instead, it won't expose the intel gpu, but only the radeon (ati).
So, I used Refind as bootloader with the option "spoof_osx_version 10.9" which fools the firmware into thinking that macos is going to be booted. And then I turn on modeset on both gpus, so that they become available on boot:
Code:
rootfstype=ext4 net.ifnames=0 libata.force=noncq ipv6.disable=1 i915.modeset=1 radeon.modeset=1 elevator=noop vt.handoff=7

And finally, at boot, I have a service that switches to the intel gpu and turns-off the radeon to save power. It's known as vgaswitcheroo and you can find it in the kernel config as apple-gmux. I wrote something about it here (scroll down almost to the bottom): https://forums.gentoo.org/viewtopic-t-1070830-highlight-.html

~/.Xsession-errors show no activity in the last few days, as much as I can understand because it is not dated and the timestamp is old.
By the way, upgraded to kernel 4.13.11 using same config, started X from the console under my user about 6 hours ago, operated VBox and all the usual apps, and it still didn't crash. Coincidence?

Now, I want you to know that I deeply appreciated your help and learned a lot from it. But the problem is that I can't afford to spend any more time trying to fix this in a systematic way; instead, I'll take a peek once in a while and experiment something else. Moreover, the machine is working perfectly booting from the other partition with the standard kernel settings.
So, I'm really shutting down now. Thank you again very, very much for all your help and support.

However, if in the future you need to know more details of my setup to help you setup your server, by all means feel free to ask; I'll do whatever I can to help.

Best regards.


Last edited by jssilva on Tue Nov 07, 2017 11:56 pm; edited 2 times in total
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Desktop Environments All times are GMT
Goto page 1, 2, 3  Next
Page 1 of 3

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum