View previous topic :: View next topic |
Author |
Message |
freifunk_connewitz Apprentice
Joined: 08 Feb 2006 Posts: 231
|
Posted: Mon Aug 11, 2014 7:47 am Post subject: [REOPENED] content of /etc/mtab: number for root? |
|
|
Hi,
it's been a while since this occurs: when issueing "mount" or "cat /etc/mtab" to see the actually mounted filesystems, it prints as the first two lines:
Quote: |
rootfs on / type rootfs (rw)
812 on / type ext3 (rw,noatime,nodiratime,commit=0)
|
what's with this "812"? shouldnt it say something like "/dev/sdaX"?
I found nothing bout it in the mount or fstab documentation...
Cheers,
FFCW
Last edited by freifunk_connewitz on Sun Jul 05, 2015 1:04 pm; edited 2 times in total |
|
Back to top |
|
|
Demonking n00b
Joined: 12 Mar 2007 Posts: 41
|
Posted: Mon Aug 11, 2014 1:14 pm Post subject: |
|
|
Hello freifunk_connewitz,
the first value could also be a virtual Filesystem from the Kernel like proc,cgroup_root or something like that.
This would look like this :
Code: |
proc on /proc type proc (rw,relatime)
tmpfs on /run type tmpfs (rw,nosuid,nodev,relatime,size=817364k,mode=755)
cgroup_root on /sys/fs/cgroup type tmpfs (rw,nosuid,nodev,noexec,relatime,size=10240k,mode=755)
|
But the 812 is strange, have you checked your fstab?
Maybe look what this command returns :
Code: | cat /proc/filesystems |
Greetz
Demonking _________________ I finally understand why Java's security model is so torturous:
It's to protect everyone else against people who write code like this.
The Answer to Life, the Universe, and Everything = 42 |
|
Back to top |
|
|
freifunk_connewitz Apprentice
Joined: 08 Feb 2006 Posts: 231
|
Posted: Mon Aug 11, 2014 2:09 pm Post subject: |
|
|
Hi Demonking,
thanks for your answer.
Demonking wrote: |
the first value could also be a virtual Filesystem from the Kernel like proc,cgroup_root or something like that.
|
Yeah, I do have all those mounted, too.
Quote: | But the 812 is strange, have you checked your fstab?
|
Most def. There, everything looks like it should.
Quote: | Maybe look what this command returns :
Code: | cat /proc/filesystems |
|
- a long list of possible filesystems.
By the way, I have the same thing on a laptop with a quite similar installation on it. Except that the number isn't 812 but something else.
What I first thought was that it could be some new feature of OpenRC or LVM or dm-crypt. But on both machines, not root, but only the home dir is encrypted, and I do not use use logical volumes.
LVM had to be installed anyway, because lilo, cryptsetup and some others depend on it. For whatever reason.
So, nobody has ever seen such a mtab?
Cheers,
freifunk connewitz |
|
Back to top |
|
|
krinn Watchman
Joined: 02 May 2003 Posts: 7470
|
Posted: Mon Aug 11, 2014 7:00 pm Post subject: |
|
|
Hi, victim #2
Look here, you'll get some explain about what is the number...
https://forums.gentoo.org/viewtopic-t-996732.html
Nobody have the solve to get back a saner naming. Waiting to see if more users get affected. |
|
Back to top |
|
|
freifunk_connewitz Apprentice
Joined: 08 Feb 2006 Posts: 231
|
Posted: Tue Aug 12, 2014 10:19 am Post subject: |
|
|
krinn,
thanks for the hint. that's the thread I was looking for.
fortunately for me, my good old LILO indeed does complain about "812 not found", too, when updating the bootsector, but still continues its job and succeeds.
so, should there be a bug filed against udev?
cheers,
ffcw |
|
Back to top |
|
|
krinn Watchman
Joined: 02 May 2003 Posts: 7470
|
Posted: Tue Aug 12, 2014 10:35 am Post subject: |
|
|
Without really knowing my feelings are more a kernel problem than udev (because /proc is not handle by udev but kernel)
But as you and the other user didn't expose any udev or kernel version. Things get hard to debug without info.
You indeed should filebug that (as i'm sure victim #3 exists but didn't notice it yet) |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54244 Location: 56N 3W
|
Posted: Tue Aug 12, 2014 4:31 pm Post subject: |
|
|
krinn,
... behind you :) _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
freifunk_connewitz Apprentice
Joined: 08 Feb 2006 Posts: 231
|
Posted: Mon May 11, 2015 6:02 pm Post subject: |
|
|
in my case some update silently solved the problem. /dev/root is displayed correctly in /proc/self/mountinfo, /proc/self/mounts or df -h. no hex numbers instead of /dev/root anymore.
my current setup:
lilo-24.0-r1
openrc-0.13.11
kernel 3.18.11
thanks for the help |
|
Back to top |
|
|
bstaletic Apprentice
Joined: 05 Apr 2014 Posts: 253
|
Posted: Mon May 11, 2015 6:17 pm Post subject: |
|
|
A few weeks ago I had the same problem, now everything says /dev/root except "mount". "mount" shows 823.
EDIT:
Kernel 4.0.2 (deblobed)
Lilo 24.1
eudev 3.0
openrc 0.15.1 |
|
Back to top |
|
|
krinn Watchman
Joined: 02 May 2003 Posts: 7470
|
Posted: Mon May 11, 2015 7:55 pm Post subject: |
|
|
NeddySeagoon wrote: | krinn,
... behind you |
From that time, i was really happy by that comment...
Now that some time has pass, i'm not really feeling really safe reading it anymore
Happy for you that your problem is solved freifunk_connewitz, thanks for the solved title edit, other users will enjoy your hint. |
|
Back to top |
|
|
freifunk_connewitz Apprentice
Joined: 08 Feb 2006 Posts: 231
|
Posted: Sun Jul 05, 2015 1:01 pm Post subject: |
|
|
this unfortunately has to be "reopened".
bstaletic, same here: everything says /dev/root except "mount". "mount" shows (in my case) 812.
kernel-3.18.12-gentoo
openrc-0.16.4
sys-fs/udev-216
virtual/udev-215
lilo 24.0-r1
and on top of it: on a freshly installed machine with kernel 4.0.5 everything (/proc/mount{s,info} and 'mount') shows the wrong info again: 803 instead of /dev/root (or /dev/sda3).
setup:
openrc 0.17
udev 216
virtual/udev 215
lilo 24.0-r1 |
|
Back to top |
|
|
freifunk_connewitz Apprentice
Joined: 08 Feb 2006 Posts: 231
|
Posted: Sun Jul 05, 2015 1:56 pm Post subject: |
|
|
as discussed here: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=766920, this bug disappears if you boot with UUID in your lilo.conf instead of /dev/sdXX. unfortunately, this is not an option for me because it only worked if I also began to use an initrd/initramfs. what I won't.
but krinn, Neddy, or anybody else, would you agree that it is lilo where I should file a bug report? |
|
Back to top |
|
|
bstaletic Apprentice
Joined: 05 Apr 2014 Posts: 253
|
Posted: Sun Jul 05, 2015 4:12 pm Post subject: |
|
|
Isn't bootloader supposed to only load the kernel? |
|
Back to top |
|
|
steveL Watchman
Joined: 13 Sep 2006 Posts: 5153 Location: The Peanut Gallery
|
Posted: Sun Jul 05, 2015 7:42 pm Post subject: |
|
|
You should still be able to use PARTUUID as that's done by the kernel, and does not require userland, nor FS support.
As for where the bug is, I'd be inclined to blame the recent "let's change mount from userland" "work" done by distro developers, specifically debian wrt mount/libmount/utab, and ofc we're following along "because coreutils". |
|
Back to top |
|
|
freifunk_connewitz Apprentice
Joined: 08 Feb 2006 Posts: 231
|
Posted: Tue Jul 07, 2015 10:05 am Post subject: |
|
|
SteveL, thanks for the hint, but lilo considers a direct PARTUUID entry in lilo.conf like
Code: | root=/dev/disk/by-partuuid/xxxxx |
as illegal.
but we can use instead:
Code: | append="root=/dev/by-partuuid/xxxxx" |
the funny thing is, if I do that, the output of mount or /proc/self/mountinfo shows no hex numbers anymore, but:
Quote: | PARTUUID=xxxxx / ext4 ... |
obviously, lilo uses exactly the information you give it with the append option above. having seen that, I thought, well, why not just tell lilo to use what I wanted there from the beginning, the real device path. so I changed lilo.conf again, replaced the PARTUUID-part by /dev/sda3 (thats where my root fs lives), rebooted, was happy.
this is my working lilo.conf (when booting the 'test' entry):
Quote: | # $Header: /var/cvsroot/gentoo-x86/sys-boot/lilo/files/lilo.conf,v 1.2 2004/07/18 04:42:04 dragonheart Exp $
# Author: Ultanium
#
# Start LILO global section
#
compact
lba32
# MBR to install LILO to:
boot = /dev/sda
menu-scheme=Wb
prompt
timeout=5
delay = 50
vga = normal
# End LILO global section
#
#
# Linux bootable partition config begins
#
default = Gentoo
image = /boot/vmlinuz-4.0.5-gentoo
root = /dev/sda3
label = Gentoo
read-only # read-only for checking
image = /boot/vmlinuz-4.0.5-gentoo
append="root=/dev/sda3"
label = test
read-only # read-only for checking
|
and this is the first output line of cat/proc/mountinfo:
Quote: | /dev/sda3 / ext4 rw,noatime,data=ordered 0 0 |
Finally, SteveL, I'm sorry, but I don't really get your last sentence. I just don't know enough about the development details. Do you mean it's gonna be a Gentoo bug? |
|
Back to top |
|
|
|