Segmentation Fault at boot - Q4OS 32bit

MC · 2017-01-18 14:37

After the recent upgrades, I now get a segmentation fault at boot, but the system still boots and runs fine. I've done some web-searching about this, but I'm not finding anything about when this happen at boot.

After the grub, it goes to a new screen and says:

loading q4os operating system...
loading initial ramdisk...
loading, please wait...
segmentation fault
q4os: clean (and then shows  the files/blocks)

Then boots without any problems.

Thanks in advance for any help.

q4osteam · 2017-01-18 14:56

There were multiple Debian 8.7 upgrades recently, so it looks like a related glitch. Please run in terminal:
$ sudo journalctl
and check the log for errors and segfaults. You could post it here as well.

MC · 2017-01-18 15:08

I'm not sure what I'm looking for, unless it's the line "Runtime journal is using 8.0M (max allowed 78.4M, trying to leave 117.6M fre".

See attached.

EDIT: The first attachment was incomplete. Here is the full result. (Attached)

Last edited by MC (2017-01-18 15:31)

Attachments:

journalctl.txt, Size: 90.87 KiB, Downloads: 994

q4osteam · 2017-01-18 15:32

You need to scroll the journal with 'Page Down' key to view contents. Errors and significant messages are colored. To write to a text file just run:
$ sudo journalctl > journalctl.txt
and attach it here again.

MC · 2017-01-18 15:33

I just re-attached the full output in my previous post.

q4osteam · 2017-01-18 15:44

MC wrote:

I just re-attached the full output in my previous post.

It is not still complete, please run this command:
$ sudo journalctl > journalctl.txt
and attach it here once more.

MC · 2017-01-18 15:54

q4osteam wrote:

MC wrote:
I just re-attached the full output in my previous post.
It is not still complete, please run this command:
$ sudo journalctl > journalctl.txt
and attach it here once more.

Attachments:

journalctl.txt, Size: 100.12 KiB, Downloads: 910

q4osteam · 2017-01-18 16:13

See lines 577-578

input: AT Translated Set 2 keyboard as /devices/platform/i8042/serio0/input/input0
loadkeys[59]: segfault at 4b49 ip 00004b49 sp bfab2bd8 error 14 in loadkeys[8048000+11000]

We assume, there could be some kind of a bug in 'loadkeys'. Please try to run following:
$ sudo systemctl status kbd.service
$ sudo systemctl restart kbd.service
and report back.

MC · 2017-01-18 16:25

mc@q4os-desktop:~$ sudo systemctl status kbd.service
● kbd.service - LSB: Prepare console
Loaded: loaded (/etc/init.d/kbd)
Active: active (exited) since Wed 2017-01-18 08:20:50 MST; 1min 15s ago
Process: 14078 ExecStop=/etc/init.d/kbd stop (code=exited, status=0/SUCCESS)
Process: 14087 ExecStart=/etc/init.d/kbd start (code=exited, status=0/SUCCESS)

Jan 18 08:20:49 q4os-desktop kbd[14087]: Setting console screen modes.
Jan 18 08:20:50 q4os-desktop systemd[1]: Started LSB: Prepare console.
mc@q4os-desktop:~$ sudo systemctl restart kbd.service
mc@q4os-desktop:~$

q4osteam · 2017-01-18 17:42

You could try to reconfigure console keyboard layout:
$ sudo dpkg-reconfigure keyboard-configuration
and try it again.

The other option is to disable kbd:
$ sudo systemctl disable kbd.service
$ sudo systemctl mask kbd.service

MC · 2017-01-18 18:15

q4osteam wrote:

You could try to reconfigure console keyboard layout:
$ sudo dpkg-reconfigure keyboard-configuration
and try it again.

mc@q4os-desktop:~$ sudo dpkg-reconfigure keyboard-configuration
[sudo] password for mc:
update-rc.d: warning: start and stop actions are no longer supported; falling back to defaults
update-rc.d: warning: start and stop actions are no longer supported; falling back to defaults
mc@q4os-desktop:~$

I then shutdown and restarted. No change.

q4osteam wrote:

The other option is to disable kbd:
$ sudo systemctl disable kbd.service
$ sudo systemctl mask kbd.service

mc@q4os-desktop:~$ sudo systemctl disable kbd.service
[sudo] password for mc:
Synchronizing state for kbd.service with sysvinit using update-rc.d...
Executing /usr/sbin/update-rc.d kbd defaults
insserv: warning: current start runlevel(s) (empty) of script `kbd' overrides LSB defaults (S).
insserv: warning: current stop runlevel(s) (S) of script `kbd' overrides LSB defaults (empty).
Executing /usr/sbin/update-rc.d kbd disable
insserv: warning: current start runlevel(s) (empty) of script `kbd' overrides LSB defaults (S).
insserv: warning: current stop runlevel(s) (S) of script `kbd' overrides LSB defaults (empty).
mc@q4os-desktop:~$ sudo systemctl mask kbd.service
mc@q4os-desktop:~$

Again, shutdown and restart. No change.

q4osteam · 2017-01-18 19:33

So, it could be a bug in Linux kernel, hardware related. The best way now, would be to boot Q4OS Live CD (1.8.2 version) and other live Debian CD (both with the current kernel Debian 3.16.39-1) and check for this segfault. If it arises for both, make a Debian bugreport https://www.debian.org/Bugs/Reporting . We guess, this glitch could be resolved in some of the next Debian updates, we would recommend you to wait, if possible.

Other option could be to use kernel from Backports, please see http://www.q4os.org/dqa011.html#system.5

MC · 2017-01-18 21:18

q4osteam wrote:

We guess, this glitch could be resolved in some of the next Debian updates, we would recommend you to wait, if possible.

No rush at the moment. It does boot and run as it should. When I get a chance, I'll try running a couple of live disks to see what happens, and probably change the kernel to see if it helps. If I get some results that merit sharing, I'll post back here.

Thanks for your help trying to solve this.

MC · 2017-01-18 23:58

OK, here's the results:

First, since the only Q4OS discs I have are "installers" I ran qinst-kernel-bpo.

I booted into 4.8.0-0.bpo.2-686-pae. After the grub screen, the next screen read:

Loading q4os operating system...
Loading initial ramdisk...
loading, please wait...
modprobe: module microcode not found in modules.dep
q4os: clean (...files/blocks...)

Then, it boots right into the desktop. No problem.

Rebooted into 3.16.0-4-686-pae and still receive the "segment fault" message as stated before. ( But still boots into the desktop.)

--------------------

Downloaded Q4OS 1.8.2 Live 32bit/burnt to disc.

Booted and entered "Live". It ran for a second or two after the grub, then aborted due to "kernel panic".
Rebooted and entered Failsafe, same results; kernel panic.

Went to DW and saw Debian has a new release, 8.7.1. Downloaded the 32bit version/burnt to disc/ran it live/ no problem. HOWEVER, I forgot to check to make sure it was 3.16.39-1 before downloading. While In the live Debian, I ran uname -r: 3.16.0-4-686-pae.

------------------

I don't know if any of this helps, but I'm just going to keep running it for now, and like you said, wait for the next updates.

~ If it does decided to go "tires-up" I still have my Q4OS 64bit desktop.

Last edited by MC (2017-01-19 00:06)

q4osteam · 2017-01-19 07:43

MC wrote:

modprobe: module microcode not found in modules.dep

You could install microcode module this way:
$ sudo apt-get install intel-microcode

MC wrote:

I forgot to check to make sure it was 3.16.39-1 before downloading. While In the live Debian, I ran uname -r: 3.16.0-4-686-pae.

Just run:
$ uname -a
it should print the Debian kernel version.

MC · 2017-01-19 22:18

I installed intel-microcode, powered off, booted, and it still shows not found.

So, I looked in /etc/modprobe.d and it has been blacklisted.

The .conf says:

The microcode module attempts to apply a microcode update when
it autoloads. This is not always safe, so we block it by default.

I guess at this point, my basic question is; even though there is a sigfault or the missing modprobe, my computer is running as normal, so do either of these events do any harm, possible breakage?

q4osteam · 2017-01-20 08:56

MC wrote:

I guess at this point, my basic question is; even though there is a sigfault or the missing modprobe, my computer is running as normal, so do either of these events do any harm, possible breakage?

No, there is no possible breakage. The segfault occurs in kernel 'loadkeys' routine, so the only possible issue could be the defaulting to console keyboard setup in early kernel console mode. It is not related to any higher level console setup, as it is provided by another 'Systemd' module.

Microcode could be applied for some CPUs to upgrade internal instructions set, however it has a absolutely minor impact to performance. You can very safely ignore microcode module.

MC · 2017-01-20 23:20

Thanks for the info above. I'm going to keep running it "as is"...and later this year, I will replace it with Scorpion 64bit.

Thanks again for all your help.

MC

Q4OS Forum

#1 2017-01-18 14:37

Segmentation Fault at boot - Q4OS 32bit

#2 2017-01-18 14:56

Re: Segmentation Fault at boot - Q4OS 32bit

#3 2017-01-18 15:08

Re: Segmentation Fault at boot - Q4OS 32bit

#4 2017-01-18 15:32

Re: Segmentation Fault at boot - Q4OS 32bit

#5 2017-01-18 15:33

Re: Segmentation Fault at boot - Q4OS 32bit

#6 2017-01-18 15:44

Re: Segmentation Fault at boot - Q4OS 32bit

#7 2017-01-18 15:54

Re: Segmentation Fault at boot - Q4OS 32bit

#8 2017-01-18 16:13

Re: Segmentation Fault at boot - Q4OS 32bit

#9 2017-01-18 16:25

Re: Segmentation Fault at boot - Q4OS 32bit

#10 2017-01-18 17:42

Re: Segmentation Fault at boot - Q4OS 32bit

#11 2017-01-18 18:15

Re: Segmentation Fault at boot - Q4OS 32bit

#12 2017-01-18 19:33

Re: Segmentation Fault at boot - Q4OS 32bit

#13 2017-01-18 21:18

Re: Segmentation Fault at boot - Q4OS 32bit

#14 2017-01-18 23:58

Re: Segmentation Fault at boot - Q4OS 32bit

#15 2017-01-19 07:43

Re: Segmentation Fault at boot - Q4OS 32bit

#16 2017-01-19 22:18

Re: Segmentation Fault at boot - Q4OS 32bit

#17 2017-01-20 08:56

Re: Segmentation Fault at boot - Q4OS 32bit

#18 2017-01-20 23:20

Re: Segmentation Fault at boot - Q4OS 32bit

Board footer