Ubuntu
udev package

udev: Upgrading udev hoses my devfs

Bug #11561 reported by Debian Bug Importer
4
Affects Status Importance Assigned to Milestone
udev (Debian)
Fix Released
Unknown
udev (Ubuntu)
Fix Released
High
Unassigned

Bug Description

Automatically imported from Debian bug report #288043 http://bugs.debian.org/288043

See original description
Revision history for this message
Debian Bug Importer (debzilla) wrote :

Automatically imported from Debian bug report #288043 http://bugs.debian.org/288043

Revision history for this message
Debian Bug Importer (debzilla) wrote :
Download full text (3.7 KiB)

Message-ID: <email address hidden>
Date: Fri, 31 Dec 2004 15:17:35 -0500
From: Daniel Jacobowitz <email address hidden>
To: Debian Bug Tracking System <email address hidden>
Subject: udev: Upgrading udev hoses my devfs

Package: udev
Version: 0.050-2
Severity: critical
Justification: breaks the whole system

I have devfs mounted on /dev. I didn't deliberately install udev, it was
pulled in by dependencies. Here's what happened when I upgraded:

Setting up udev (0.050-2) ...
Populating the new /dev filesystem temporarily mounted on
/tmp/fileOaRTAz/...
ln: accessing `//tmp/fileOaRTAz/fd': Not a directory
dpkg: error processing udev (--configure):
 subprocess post-installation script returned error exit status 1

It leaves my devfs moved to /.dev, a messed up tmpfs mounted on /dev (I'm
not sure why - it returned "/dev: Not a directory") and a garbage tmpfs
mounted in /tmp/fileOaRTAz. This means no ptys available.

If devfs is mounted, is there any point in making udev try to take over
/dev? I had to uninstall udev and gnome-desktop-environment to get it to
leave me alone; I would have to mess with the system configuration to get it
to stop mounting devfs and I want to do it on my schedule, not udev's.

-- Package-specific info:
-- /etc/udev/rules.d/:
/etc/udev/rules.d/:
total 0
lrwxrwxrwx 1 root root 19 Dec 31 15:05 cd-aliases.rules -> ../cd-aliases.rules
lrwxrwxrwx 1 root root 20 Dec 31 15:05 compat-full.rules -> ../compat-full.rules
lrwxrwxrwx 1 root root 14 Dec 31 15:05 devfs.rules -> ../devfs.rules
lrwxrwxrwx 1 root root 12 Dec 31 15:04 z_hal-plugdev.rules -> ../hal.rules

-- /sys/:
/sys/block/fd0/dev
/sys/block/hda/dev
/sys/block/hda/hda1/dev
/sys/block/hda/hda2/dev
/sys/block/hda/hda3/dev
/sys/block/hda/hda5/dev
/sys/block/hda/hda6/dev
/sys/block/hda/hda7/dev
/sys/block/hdb/dev
/sys/block/hdc/dev
/sys/block/hdc/hdc1/dev
/sys/block/hdc/hdc2/dev
/sys/block/hdc/hdc5/dev
/sys/block/hde/dev
/sys/block/hde/hde1/dev
/sys/block/hde/hde2/dev
/sys/block/hdg/dev
/sys/block/hdg/hdg1/dev
/sys/block/hdg/hdg2/dev
/sys/block/md0/dev
/sys/block/ram0/dev
/sys/block/ram1/dev
/sys/block/ram10/dev
/sys/block/ram11/dev
/sys/block/ram12/dev
/sys/block/ram13/dev
/sys/block/ram14/dev
/sys/block/ram15/dev
/sys/block/ram2/dev
/sys/block/ram3/dev
/sys/block/ram4/dev
/sys/block/ram5/dev
/sys/block/ram6/dev
/sys/block/ram7/dev
/sys/block/ram8/dev
/sys/block/ram9/dev
/sys/class/cpuid/cpu0/dev
/sys/class/cpuid/cpu1/dev
/sys/class/drm/r128/dev
/sys/class/input/event0/dev
/sys/class/input/event1/dev
/sys/class/input/mice/dev
/sys/class/input/mouse0/dev
/sys/class/misc/agpgart/dev
/sys/class/misc/psaux/dev
/sys/class/misc/rtc/dev
/sys/class/misc/tun/dev
/sys/class/msr/msr0/dev
/sys/class/msr/msr1/dev
/sys/class/sound/audio/dev
/sys/class/sound/controlC0/dev
/sys/class/sound/controlC1/dev
/sys/class/sound/controlC2/dev
/sys/class/sound/controlC3/dev
/sys/class/sound/dsp/dev
/sys/class/sound/midi/dev
/sys/class/sound/mixer/dev
/sys/class/sound/timer/dev
/sys/class/sound/unknown6/dev
/sys/class/usb/lp0/dev

-- Kernel configuration:
 isapnp_init not present.

-- System Information:
Debian Release: 3.1
  APT prefers unstable
  APT policy: (500, 'unstable')
...

Read more...

Revision history for this message
In , Marco d'Itri (md) wrote : Re: Bug#288043: udev: Upgrading udev hoses my devfs

On Dec 31, Daniel Jacobowitz <email address hidden> wrote:

> If devfs is mounted, is there any point in making udev try to take over
> /dev? I had to uninstall udev and gnome-desktop-environment to get it to
Not sure. So far it has been enabled on devfs systems too, but only
after a reboot, and I do not see a good reason to not do it.
I do not understand why postinst fails on your system, apparently it
happens when "ln /tmp/$TEMPDEV/fd /proc/self/fd" is run.

> It leaves my devfs moved to /.dev, a messed up tmpfs mounted on /dev (I'm
> not sure why - it returned "/dev: Not a directory") and a garbage tmpfs
> mounted in /tmp/fileOaRTAz. This means no ptys available.
I can't see how this could be possible, the commands which do this are
executed *after* ln:

  # -n is used because mount gets very confused by --bind (see #282205)
  mount -n --bind /dev/pts/ $TEMPDEV/pts/
  mount -n --bind /dev/shm/ $TEMPDEV/shm/
  mount -n --bind /dev /.dev
  mount -n --move $TEMPDEV /dev

I think that your description of the events is not fully correct.
Do you mind trying again? :-)

BTW, if you want to disable udev all you need to do is put "exit 0"
in the init script (and clean up after the postinst fuckage).

--
ciao,
Marco

Revision history for this message
Debian Bug Importer (debzilla) wrote :

Message-ID: <email address hidden>
Date: Fri, 31 Dec 2004 22:03:53 +0100
From: <email address hidden> (Marco d'Itri)
To: Daniel Jacobowitz <email address hidden>, <email address hidden>
Subject: Re: Bug#288043: udev: Upgrading udev hoses my devfs

--uAKRQypu60I7Lcqm
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Dec 31, Daniel Jacobowitz <email address hidden> wrote:

> If devfs is mounted, is there any point in making udev try to take over
> /dev? I had to uninstall udev and gnome-desktop-environment to get it to
Not sure. So far it has been enabled on devfs systems too, but only
after a reboot, and I do not see a good reason to not do it.
I do not understand why postinst fails on your system, apparently it
happens when "ln /tmp/$TEMPDEV/fd /proc/self/fd" is run.

> It leaves my devfs moved to /.dev, a messed up tmpfs mounted on /dev (I'm
> not sure why - it returned "/dev: Not a directory") and a garbage tmpfs
> mounted in /tmp/fileOaRTAz. This means no ptys available.
I can't see how this could be possible, the commands which do this are
executed *after* ln:

  # -n is used because mount gets very confused by --bind (see #282205)
  mount -n --bind /dev/pts/ $TEMPDEV/pts/
  mount -n --bind /dev/shm/ $TEMPDEV/shm/
  mount -n --bind /dev /.dev
  mount -n --move $TEMPDEV /dev

I think that your description of the events is not fully correct.
Do you mind trying again? :-)

BTW, if you want to disable udev all you need to do is put "exit 0"
in the init script (and clean up after the postinst fuckage).

--=20
ciao,
Marco

--uAKRQypu60I7Lcqm
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFB1b65FGfw2OHuP7ERAuaWAJ0YitDncmwukzLpw0kgtO/9+SzC9ACghZcd
XqYeJlFqnNWEczVoo3KzyIY=
=TFpg
-----END PGP SIGNATURE-----

--uAKRQypu60I7Lcqm--

Revision history for this message
In , Daniel Jacobowitz (dan) wrote :

On Fri, Dec 31, 2004 at 10:03:53PM +0100, Marco d'Itri wrote:
> On Dec 31, Daniel Jacobowitz <email address hidden> wrote:
>
> > If devfs is mounted, is there any point in making udev try to take over
> > /dev? I had to uninstall udev and gnome-desktop-environment to get it to
> Not sure. So far it has been enabled on devfs systems too, but only
> after a reboot, and I do not see a good reason to not do it.
> I do not understand why postinst fails on your system, apparently it
> happens when "ln /tmp/$TEMPDEV/fd /proc/self/fd" is run.
>
> > It leaves my devfs moved to /.dev, a messed up tmpfs mounted on /dev (I'm
> > not sure why - it returned "/dev: Not a directory") and a garbage tmpfs
> > mounted in /tmp/fileOaRTAz. This means no ptys available.
> I can't see how this could be possible, the commands which do this are
> executed *after* ln:
>
> # -n is used because mount gets very confused by --bind (see #282205)
> mount -n --bind /dev/pts/ $TEMPDEV/pts/
> mount -n --bind /dev/shm/ $TEMPDEV/shm/
> mount -n --bind /dev /.dev
> mount -n --move $TEMPDEV /dev
>
> I think that your description of the events is not fully correct.
> Do you mind trying again? :-)

It appears to be a problem with tmpfs in 2.6.10-rc1, maybe. Witness:

root@nevyn:/home/drow# mount -t tmpfs tmpfs /mnt/spazmount
root@nevyn:/home/drow# ls /mnt/spazmount
ls: /mnt/spazmount: Not a directory

Aha:
# CONFIG_TMPFS is not set

I don't know why the tmpfs mount succeeds even though tmpfs is
disabled. tmpfs also appears in /proc/filesystems. That's really
nasty! It has something to do with the fact that tmpfs is used for
in-kernel filesystems.

Should we be detecting this problem?

--
Daniel Jacobowitz

Revision history for this message
In , Browaeys-alban (browaeys-alban) wrote : udev had to conflict with devfsd
Download full text (3.1 KiB)

Package: udev
Version: 0.050-2
Followup-For: Bug #288043

I though udev conflicts with devfsd ... isn't it a policy violation as
they both write the same files ?

I don't feel the bug has something to do with udev or "devfs" (kernel
part) as none of those manage pts nowadays.

Alban

-- Package-specific info:
-- /etc/udev/rules.d/:
/etc/udev/rules.d/:
total 0
lrwxrwxrwx 1 root root 19 2004-12-05 01:12 cd-aliases.rules -> ../cd-aliases.rules
lrwxrwxrwx 1 root root 13 2004-12-05 01:13 udev.rules -> ../udev.rules
lrwxrwxrwx 1 root root 12 2004-12-30 22:39 z_hal-plugdev.rules -> ../hal.rules

-- /sys/:
/sys/block/hda/dev
/sys/block/hda/hda1/dev
/sys/block/hda/hda2/dev
/sys/block/hda/hda3/dev
/sys/block/hda/hda4/dev
/sys/block/hda/hda5/dev
/sys/block/hda/hda6/dev
/sys/block/hda/hda7/dev
/sys/block/hdc/dev
/sys/block/hdc/hdc1/dev
/sys/block/hdc/hdc2/dev
/sys/block/hdc/hdc3/dev
/sys/block/hdc/hdc5/dev
/sys/block/ram0/dev
/sys/block/ram10/dev
/sys/block/ram11/dev
/sys/block/ram12/dev
/sys/block/ram13/dev
/sys/block/ram14/dev
/sys/block/ram15/dev
/sys/block/ram1/dev
/sys/block/ram2/dev
/sys/block/ram3/dev
/sys/block/ram4/dev
/sys/block/ram5/dev
/sys/block/ram6/dev
/sys/block/ram7/dev
/sys/block/ram8/dev
/sys/block/ram9/dev
/sys/class/cpuid/cpu0/dev
/sys/class/graphics/fb0/dev
/sys/class/i2c-dev/i2c-0/dev
/sys/class/i2c-dev/i2c-1/dev
/sys/class/i2c-dev/i2c-2/dev
/sys/class/input/event0/dev
/sys/class/input/event1/dev
/sys/class/input/mice/dev
/sys/class/input/mouse0/dev
/sys/class/misc/agpgart/dev
/sys/class/misc/fbsplash/dev
/sys/class/misc/nvram/dev
/sys/class/misc/psaux/dev
/sys/class/misc/rtc/dev
/sys/class/nvidia/nvidia0/dev
/sys/class/nvidia/nvidiactl/dev
/sys/class/ppp/ppp/dev
/sys/class/sound/admmidi1/dev
/sys/class/sound/adsp/dev
/sys/class/sound/amidi1/dev
/sys/class/sound/audio/dev
/sys/class/sound/controlC0/dev
/sys/class/sound/controlC1/dev
/sys/class/sound/controlC2/dev
/sys/class/sound/controlC3/dev
/sys/class/sound/dmmidi1/dev
/sys/class/sound/dsp/dev
/sys/class/sound/midi1/dev
/sys/class/sound/midiC1D0/dev
/sys/class/sound/midiC1D1/dev
/sys/class/sound/midiC1D2/dev
/sys/class/sound/midiC1D3/dev
/sys/class/sound/mixer1/dev
/sys/class/sound/mixer/dev
/sys/class/sound/pcmC0D0c/dev
/sys/class/sound/pcmC0D0p/dev
/sys/class/sound/pcmC0D1c/dev
/sys/class/sound/pcmC0D2p/dev
/sys/class/sound/seq/dev
/sys/class/sound/sequencer2/dev
/sys/class/sound/sequencer/dev
/sys/class/sound/timer/dev

-- Kernel configuration:
 isapnp_init not present.

-- System Information:
Debian Release: 3.1
  APT prefers unstable
  APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: i386 (i686)
Kernel: Linux 2.6.10
Locale: LANG=fr_FR@euro, LC_CTYPE=fr_FR@euro (charmap=ISO-8859-15)

Versions of packages udev depends on:
ii hotplug 0.0.20040329-16 Linux Hotplug Scripts
ii initscripts 2.86-5 Standard scripts needed for bootin
ii libc6 2.3.2.ds1-20 GNU C Library: Shared libraries an
ii makedev 3.3.8.2-0 Creates device files in /dev
ii sed 4.1.2-8 The GNU sed stream editor

-- debconf information:
  udev/devfs-warning:
  udev/reboot-warning:...

Read more...

Revision history for this message
In , Browaeys-alban (browaeys-alban) wrote : udev can fake devfs device file names

Package: udev
Version: 0.050-2
Followup-For: Bug #288043

just a note : the migration from devfs to udev does not require any
change to the "system" configuration" (be it fstab, cd ripper, lilo or else).

Udev can emulate devfs device names by replacing the symlink in
/etc/udev/rules.d/ :
from udev.rules -> ../udev.rules
to devfs.rules -> ../devfs.rules

(you can also let both rules active to make the system transition)

CHeers
Alban

-- System Information:
Debian Release: 3.1
  APT prefers unstable
  APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: i386 (i686)
Kernel: Linux 2.6.10
Locale: LANG=fr_FR@euro, LC_CTYPE=fr_FR@euro (charmap=ISO-8859-15)

Versions of packages udev depends on:
ii hotplug 0.0.20040329-16 Linux Hotplug Scripts
ii initscripts 2.86-5 Standard scripts needed for bootin
ii libc6 2.3.2.ds1-20 GNU C Library: Shared libraries an
ii makedev 3.3.8.2-0 Creates device files in /dev
ii sed 4.1.2-8 The GNU sed stream editor

-- debconf information excluded

Revision history for this message
Debian Bug Importer (debzilla) wrote :

Message-ID: <email address hidden>
Date: Sat, 1 Jan 2005 01:12:52 -0500
From: Daniel Jacobowitz <email address hidden>
To: Marco d'Itri <email address hidden>
Cc: <email address hidden>
Subject: Re: Bug#288043: udev: Upgrading udev hoses my devfs

On Fri, Dec 31, 2004 at 10:03:53PM +0100, Marco d'Itri wrote:
> On Dec 31, Daniel Jacobowitz <email address hidden> wrote:
>
> > If devfs is mounted, is there any point in making udev try to take over
> > /dev? I had to uninstall udev and gnome-desktop-environment to get it to
> Not sure. So far it has been enabled on devfs systems too, but only
> after a reboot, and I do not see a good reason to not do it.
> I do not understand why postinst fails on your system, apparently it
> happens when "ln /tmp/$TEMPDEV/fd /proc/self/fd" is run.
>
> > It leaves my devfs moved to /.dev, a messed up tmpfs mounted on /dev (I'm
> > not sure why - it returned "/dev: Not a directory") and a garbage tmpfs
> > mounted in /tmp/fileOaRTAz. This means no ptys available.
> I can't see how this could be possible, the commands which do this are
> executed *after* ln:
>
> # -n is used because mount gets very confused by --bind (see #282205)
> mount -n --bind /dev/pts/ $TEMPDEV/pts/
> mount -n --bind /dev/shm/ $TEMPDEV/shm/
> mount -n --bind /dev /.dev
> mount -n --move $TEMPDEV /dev
>
> I think that your description of the events is not fully correct.
> Do you mind trying again? :-)

It appears to be a problem with tmpfs in 2.6.10-rc1, maybe. Witness:

root@nevyn:/home/drow# mount -t tmpfs tmpfs /mnt/spazmount
root@nevyn:/home/drow# ls /mnt/spazmount
ls: /mnt/spazmount: Not a directory

Aha:
# CONFIG_TMPFS is not set

I don't know why the tmpfs mount succeeds even though tmpfs is
disabled. tmpfs also appears in /proc/filesystems. That's really
nasty! It has something to do with the fact that tmpfs is used for
in-kernel filesystems.

Should we be detecting this problem?

--
Daniel Jacobowitz

Revision history for this message
Debian Bug Importer (debzilla) wrote :
Download full text (3.3 KiB)

Message-ID: <email address hidden>
Date: Sat, 1 Jan 2005 07:21:58 +0100
From: <email address hidden>
To: Debian Bug Tracking System <email address hidden>
Subject: udev had to conflict with devfsd

Package: udev
Version: 0.050-2
Followup-For: Bug #288043

I though udev conflicts with devfsd ... isn't it a policy violation as
they both write the same files ?

I don't feel the bug has something to do with udev or "devfs" (kernel
part) as none of those manage pts nowadays.

Alban

-- Package-specific info:
-- /etc/udev/rules.d/:
/etc/udev/rules.d/:
total 0
lrwxrwxrwx 1 root root 19 2004-12-05 01:12 cd-aliases.rules -> ../cd-aliases.rules
lrwxrwxrwx 1 root root 13 2004-12-05 01:13 udev.rules -> ../udev.rules
lrwxrwxrwx 1 root root 12 2004-12-30 22:39 z_hal-plugdev.rules -> ../hal.rules

-- /sys/:
/sys/block/hda/dev
/sys/block/hda/hda1/dev
/sys/block/hda/hda2/dev
/sys/block/hda/hda3/dev
/sys/block/hda/hda4/dev
/sys/block/hda/hda5/dev
/sys/block/hda/hda6/dev
/sys/block/hda/hda7/dev
/sys/block/hdc/dev
/sys/block/hdc/hdc1/dev
/sys/block/hdc/hdc2/dev
/sys/block/hdc/hdc3/dev
/sys/block/hdc/hdc5/dev
/sys/block/ram0/dev
/sys/block/ram10/dev
/sys/block/ram11/dev
/sys/block/ram12/dev
/sys/block/ram13/dev
/sys/block/ram14/dev
/sys/block/ram15/dev
/sys/block/ram1/dev
/sys/block/ram2/dev
/sys/block/ram3/dev
/sys/block/ram4/dev
/sys/block/ram5/dev
/sys/block/ram6/dev
/sys/block/ram7/dev
/sys/block/ram8/dev
/sys/block/ram9/dev
/sys/class/cpuid/cpu0/dev
/sys/class/graphics/fb0/dev
/sys/class/i2c-dev/i2c-0/dev
/sys/class/i2c-dev/i2c-1/dev
/sys/class/i2c-dev/i2c-2/dev
/sys/class/input/event0/dev
/sys/class/input/event1/dev
/sys/class/input/mice/dev
/sys/class/input/mouse0/dev
/sys/class/misc/agpgart/dev
/sys/class/misc/fbsplash/dev
/sys/class/misc/nvram/dev
/sys/class/misc/psaux/dev
/sys/class/misc/rtc/dev
/sys/class/nvidia/nvidia0/dev
/sys/class/nvidia/nvidiactl/dev
/sys/class/ppp/ppp/dev
/sys/class/sound/admmidi1/dev
/sys/class/sound/adsp/dev
/sys/class/sound/amidi1/dev
/sys/class/sound/audio/dev
/sys/class/sound/controlC0/dev
/sys/class/sound/controlC1/dev
/sys/class/sound/controlC2/dev
/sys/class/sound/controlC3/dev
/sys/class/sound/dmmidi1/dev
/sys/class/sound/dsp/dev
/sys/class/sound/midi1/dev
/sys/class/sound/midiC1D0/dev
/sys/class/sound/midiC1D1/dev
/sys/class/sound/midiC1D2/dev
/sys/class/sound/midiC1D3/dev
/sys/class/sound/mixer1/dev
/sys/class/sound/mixer/dev
/sys/class/sound/pcmC0D0c/dev
/sys/class/sound/pcmC0D0p/dev
/sys/class/sound/pcmC0D1c/dev
/sys/class/sound/pcmC0D2p/dev
/sys/class/sound/seq/dev
/sys/class/sound/sequencer2/dev
/sys/class/sound/sequencer/dev
/sys/class/sound/timer/dev

-- Kernel configuration:
 isapnp_init not present.

-- System Information:
Debian Release: 3.1
  APT prefers unstable
  APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: i386 (i686)
Kernel: Linux 2.6.10
Locale: LANG=fr_FR@euro, LC_CTYPE=fr_FR@euro (charmap=ISO-8859-15)

Versions of packages udev depends on:
ii hotplug 0.0.20040329-16 Linux Hotplug Scripts
ii initscripts 2.86-5 Standard scripts needed for bootin
ii libc6 2.3.2.ds1-20 GNU C Library: Shared libraries...

Read more...

Revision history for this message
Debian Bug Importer (debzilla) wrote :

Message-ID: <email address hidden>
Date: Sat, 1 Jan 2005 07:33:55 +0100
From: <email address hidden>
To: Debian Bug Tracking System <email address hidden>
Subject: udev can fake devfs device file names

Package: udev
Version: 0.050-2
Followup-For: Bug #288043

just a note : the migration from devfs to udev does not require any
change to the "system" configuration" (be it fstab, cd ripper, lilo or else).

Udev can emulate devfs device names by replacing the symlink in
/etc/udev/rules.d/ :
from udev.rules -> ../udev.rules
to devfs.rules -> ../devfs.rules

(you can also let both rules active to make the system transition)

CHeers
Alban

-- System Information:
Debian Release: 3.1
  APT prefers unstable
  APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: i386 (i686)
Kernel: Linux 2.6.10
Locale: LANG=fr_FR@euro, LC_CTYPE=fr_FR@euro (charmap=ISO-8859-15)

Versions of packages udev depends on:
ii hotplug 0.0.20040329-16 Linux Hotplug Scripts
ii initscripts 2.86-5 Standard scripts needed for bootin
ii libc6 2.3.2.ds1-20 GNU C Library: Shared libraries an
ii makedev 3.3.8.2-0 Creates device files in /dev
ii sed 4.1.2-8 The GNU sed stream editor

-- debconf information excluded

Revision history for this message
In , Marco d'Itri (md) wrote : Re: Bug#288043: udev had to conflict with devfsd

On Jan 01, <email address hidden> wrote:

> I though udev conflicts with devfsd ... isn't it a policy violation as
> they both write the same files ?
Which files? What are you talking about?

--
ciao,
Marco

Revision history for this message
In , Marco d'Itri (md) wrote : Re: Bug#288043: udev: kernel allow mounting non-working tmpfs

severity 288043 important
retitle 288043 kernel allow mounting non-working tmpfs
tag 288043 sarge sid help upstream
thanks

On Jan 01, Daniel Jacobowitz <email address hidden> wrote:

> It appears to be a problem with tmpfs in 2.6.10-rc1, maybe. Witness:
>
> root@nevyn:/home/drow# mount -t tmpfs tmpfs /mnt/spazmount
> root@nevyn:/home/drow# ls /mnt/spazmount
> ls: /mnt/spazmount: Not a directory
>
> Aha:
> # CONFIG_TMPFS is not set
>
> I don't know why the tmpfs mount succeeds even though tmpfs is
> disabled. tmpfs also appears in /proc/filesystems. That's really
> nasty! It has something to do with the fact that tmpfs is used for
> in-kernel filesystems.
>
> Should we be detecting this problem?
Yes, but I have no clue about how to do it. /proc/filesystems does not
work either: see #275874 for details.
That bug log reports that it has been fixed in debian kernels, but
apparently the patch has not been merged by Linus yet.

--
ciao,
Marco

Revision history for this message
Debian Bug Importer (debzilla) wrote :

Message-ID: <email address hidden>
Date: Sat, 1 Jan 2005 11:14:19 +0100
From: <email address hidden> (Marco d'Itri)
To: Alban Browaeys <email address hidden>,
 <email address hidden>
Subject: Re: Bug#288043: udev had to conflict with devfsd

--bg08WKrSYDhXBjb5
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Jan 01, <email address hidden> wrote:

> I though udev conflicts with devfsd ... isn't it a policy violation as
> they both write the same files ?
Which files? What are you talking about?

--=20
ciao,
Marco

--bg08WKrSYDhXBjb5
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFB1nf7FGfw2OHuP7ERAq89AKCYKSX0iAj0XgjO6CrErmkAy86QHQCcDHdZ
a4/9lleVK7SP9IM5qrwy4e8=
=lfQO
-----END PGP SIGNATURE-----

--bg08WKrSYDhXBjb5--

Revision history for this message
Debian Bug Importer (debzilla) wrote :

Message-ID: <email address hidden>
Date: Sat, 1 Jan 2005 11:32:00 +0100
From: <email address hidden> (Marco d'Itri)
To: Daniel Jacobowitz <email address hidden>, <email address hidden>
Cc: <email address hidden>, <email address hidden>
Subject: Re: Bug#288043: udev: kernel allow mounting non-working tmpfs

--RASg3xLB4tUQ4RcS
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

severity 288043 important
retitle 288043 kernel allow mounting non-working tmpfs
tag 288043 sarge sid help upstream
thanks

On Jan 01, Daniel Jacobowitz <email address hidden> wrote:

> It appears to be a problem with tmpfs in 2.6.10-rc1, maybe. Witness:
>=20
> root@nevyn:/home/drow# mount -t tmpfs tmpfs /mnt/spazmount
> root@nevyn:/home/drow# ls /mnt/spazmount
> ls: /mnt/spazmount: Not a directory
>=20
> Aha:
> # CONFIG_TMPFS is not set
>=20
> I don't know why the tmpfs mount succeeds even though tmpfs is
> disabled. tmpfs also appears in /proc/filesystems. That's really
> nasty! It has something to do with the fact that tmpfs is used for
> in-kernel filesystems.
>=20
> Should we be detecting this problem?
Yes, but I have no clue about how to do it. /proc/filesystems does not
work either: see #275874 for details.
That bug log reports that it has been fixed in debian kernels, but
apparently the patch has not been merged by Linus yet.

--=20
ciao,
Marco

--RASg3xLB4tUQ4RcS
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFB1nwgFGfw2OHuP7ERAiEzAJ9jWxJd2FGzAHDY5k7vuyWuVyaENgCcC8mo
j9c47MUBsJaBE+qn5FubmMc=
=EDPU
-----END PGP SIGNATURE-----

--RASg3xLB4tUQ4RcS--

Revision history for this message
In , Browaeys-alban (browaeys-alban) wrote : Re: Bug#288043: udev had to conflict with devfsd

>> On Dec 31, Daniel Jacobowitz <email address hidden> wrote:
>> I have devfs mounted on /dev. I didn't deliberately install udev, it
>> was pulled in by dependencies

Le samedi 01 janvier 2005 à 11:14 +0100, Marco d'Itri a écrit :
> On Jan 01, <email address hidden> wrote:
>
> > I though udev conflicts with devfsd ... isn't it a policy violation as
> > they both write the same files ?
> Which files? What are you talking about?

I respond to the first part of the bug report (the fact that udev
install was done even though devfs (mostly devfsd as we cannot conflict
with the kernel part) was there.

devfsd and udev both mount /dev (and create files inside). I am wrong
telling they overwrite their /dev ?
And i am pretty confident devfsd is there as jason told :
"It leaves my devfs moved to /.dev" and only devfsd populates the /dev
part . Mounting devfs is not enough for that.

this bug reports two independant problems : a missing conflict (devfsd
and udev) and a kernel bug (tmpfs). I may be wrong , this was merely to
point out that the first part is easier to fix .
 And the second may be only a normal bug if it appears only whith a
"broken" install : kernel missing tmpfs, or anything else .

Dunno for sure, though i m surprised by the which files conflicts ...

Ciao
Alban

Revision history for this message
In , Marco d'Itri (md) wrote :

On Jan 01, Alban browaeys <email address hidden> wrote:

> devfsd and udev both mount /dev (and create files inside). I am wrong
> telling they overwrite their /dev ?
Yes. No conflict is needed, if udev is installed then devfsd will not be
started (no matter if devfs had been mounted by the kernel or not).

--
ciao,
Marco

Revision history for this message
Debian Bug Importer (debzilla) wrote :

Message-Id: <1104602070.5479.29.camel@localhost>
Date: Sat, 01 Jan 2005 18:54:30 +0100
From: Alban browaeys <email address hidden>
To: Marco d'Itri <email address hidden>
Cc: <email address hidden>
Subject: Re: Bug#288043: udev had to conflict with devfsd

>> On Dec 31, Daniel Jacobowitz <email address hidden> wrote:
>> I have devfs mounted on /dev. I didn't deliberately install udev, it
>> was pulled in by dependencies

Le samedi 01 janvier 2005 =E0 11:14 +0100, Marco d'Itri a =E9crit :
> On Jan 01, <email address hidden> wrote:
>=20
> > I though udev conflicts with devfsd ... isn't it a policy violation as
> > they both write the same files ?
> Which files? What are you talking about?

I respond to the first part of the bug report (the fact that udev
install was done even though devfs (mostly devfsd as we cannot conflict
with the kernel part) was there.

devfsd and udev both mount /dev (and create files inside). I am wrong
telling they overwrite their /dev ?
And i am pretty confident devfsd is there as jason told :
"It leaves my devfs moved to /.dev" and only devfsd populates the /dev
part . Mounting devfs is not enough for that.

this bug reports two independant problems : a missing conflict (devfsd
and udev) and a kernel bug (tmpfs). I may be wrong , this was merely to
point out that the first part is easier to fix .
 And the second may be only a normal bug if it appears only whith a
"broken" install : kernel missing tmpfs, or anything else .=20

Dunno for sure, though i m surprised by the which files conflicts ...

Ciao
Alban

Revision history for this message
Debian Bug Importer (debzilla) wrote :

Message-ID: <email address hidden>
Date: Sat, 1 Jan 2005 18:51:40 +0100
From: <email address hidden> (Marco d'Itri)
To: Alban browaeys <email address hidden>
Cc: <email address hidden>
Subject: Re: Bug#288043: udev had to conflict with devfsd

--17pEHd4RhPHOinZp
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Jan 01, Alban browaeys <email address hidden> wrote:

> devfsd and udev both mount /dev (and create files inside). I am wrong
> telling they overwrite their /dev ?
Yes. No conflict is needed, if udev is installed then devfsd will not be
started (no matter if devfs had been mounted by the kernel or not).

--=20
ciao,
Marco

--17pEHd4RhPHOinZp
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFB1uMsFGfw2OHuP7ERAvXDAJ9momr2hoBKUnbz7O4Mw1kuZ787pgCgktur
D1+a02zxkjkiEgjGt6uUaxg=
=bz6a
-----END PGP SIGNATURE-----

--17pEHd4RhPHOinZp--

Revision history for this message
In , Marco d'Itri (md) wrote : Bug#288043: fixed in udev 0.050-3

Source: udev
Source-Version: 0.050-3

We believe that the bug you reported is fixed in the latest version of
udev, which is due to be installed in the Debian FTP archive:

udev_0.050-3.diff.gz
  to pool/main/u/udev/udev_0.050-3.diff.gz
udev_0.050-3.dsc
  to pool/main/u/udev/udev_0.050-3.dsc
udev_0.050-3_i386.deb
  to pool/main/u/udev/udev_0.050-3_i386.deb

A summary of the changes between this version and the previous one is
attached.

Thank you for reporting the bug, which will now be closed. If you
have further comments please address them to <email address hidden>,
and the maintainer will reopen the bug report if appropriate.

Debian distribution maintenance software
pp.
Marco d'Itri <email address hidden> (supplier of updated udev package)

(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing <email address hidden>)

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Format: 1.7
Date: Sat, 1 Jan 2005 23:57:26 +0100
Source: udev
Binary: udev udev-udeb
Architecture: source i386
Version: 0.050-3
Distribution: unstable
Urgency: high
Maintainer: Marco d'Itri <email address hidden>
Changed-By: Marco d'Itri <email address hidden>
Description:
 udev - /dev/ management daemon
Closes: 287225 287959 287959 288043
Changes:
 udev (0.050-3) unstable; urgency=high
 .
   * Fixed the regexp in preinst which caused cd-aliases.rules to not be
     enabled on upgrades. (Closes: #287225)
   * Included scsi_id.config as an example. (Closes: #287959)
   * Modified postinst and the init script to fail gracefully if the kernel
     mounts a non-working tmpfs. (Closes: #288043)
   * Added patch scsi_id_tmp: makes scsi_id create its temporary devices in
     /dev instead of /tmp, which is read only when udevstart is run.
     (Closes: #287959)
Files:
 b69e8cd7e3c643bd8013dab465a47451 561 admin extra udev_0.050-3.dsc
 91fa0d5c85e180af10bb89526df57be3 30207 admin extra udev_0.050-3.diff.gz
 dcb93cfae428ace5b00f0d8f90cbaf7a 243980 admin extra udev_0.050-3_i386.deb

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFB18HrFGfw2OHuP7ERAqzAAJwKJ+ReTJDJbzY2G6bsc9QK9g8izwCfb+D5
w7NHmZQxLbfQcE3aN2n7kag=
=zEXo
-----END PGP SIGNATURE-----

Revision history for this message
Debian Bug Importer (debzilla) wrote :

Message-Id: <email address hidden>
Date: Sun, 02 Jan 2005 05:32:03 -0500
From: Marco d'Itri <email address hidden>
To: <email address hidden>
Subject: Bug#288043: fixed in udev 0.050-3

Source: udev
Source-Version: 0.050-3

We believe that the bug you reported is fixed in the latest version of
udev, which is due to be installed in the Debian FTP archive:

udev_0.050-3.diff.gz
  to pool/main/u/udev/udev_0.050-3.diff.gz
udev_0.050-3.dsc
  to pool/main/u/udev/udev_0.050-3.dsc
udev_0.050-3_i386.deb
  to pool/main/u/udev/udev_0.050-3_i386.deb

A summary of the changes between this version and the previous one is
attached.

Thank you for reporting the bug, which will now be closed. If you
have further comments please address them to <email address hidden>,
and the maintainer will reopen the bug report if appropriate.

Debian distribution maintenance software
pp.
Marco d'Itri <email address hidden> (supplier of updated udev package)

(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing <email address hidden>)

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Format: 1.7
Date: Sat, 1 Jan 2005 23:57:26 +0100
Source: udev
Binary: udev udev-udeb
Architecture: source i386
Version: 0.050-3
Distribution: unstable
Urgency: high
Maintainer: Marco d'Itri <email address hidden>
Changed-By: Marco d'Itri <email address hidden>
Description:
 udev - /dev/ management daemon
Closes: 287225 287959 287959 288043
Changes:
 udev (0.050-3) unstable; urgency=high
 .
   * Fixed the regexp in preinst which caused cd-aliases.rules to not be
     enabled on upgrades. (Closes: #287225)
   * Included scsi_id.config as an example. (Closes: #287959)
   * Modified postinst and the init script to fail gracefully if the kernel
     mounts a non-working tmpfs. (Closes: #288043)
   * Added patch scsi_id_tmp: makes scsi_id create its temporary devices in
     /dev instead of /tmp, which is read only when udevstart is run.
     (Closes: #287959)
Files:
 b69e8cd7e3c643bd8013dab465a47451 561 admin extra udev_0.050-3.dsc
 91fa0d5c85e180af10bb89526df57be3 30207 admin extra udev_0.050-3.diff.gz
 dcb93cfae428ace5b00f0d8f90cbaf7a 243980 admin extra udev_0.050-3_i386.deb

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFB18HrFGfw2OHuP7ERAqzAAJwKJ+ReTJDJbzY2G6bsc9QK9g8izwCfb+D5
w7NHmZQxLbfQcE3aN2n7kag=
=zEXo
-----END PGP SIGNATURE-----

Revision history for this message
Martin Pitt (pitti) wrote :

I just merged udev 0.050-3, so I guess this can be closed.

Bug Watch Updater (bug-watch-updater)
Changed in udev:
status: Unknown → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.