Plymouth doesn't show messages sent before the splash screen is visible

Guess what? Without that line in /etc/fstab, booting works fine. This is still a bug, though, because whatever handled the removal of that mount point should have removed it from the fstab automatically. Don't know why it hasn't happened to others yet, but alpha 2 is only a day old, so we'll see how it unfolds.

Don't know if it happens to others, but if it does, it's an important upgrade bug.

Removing the line in fstab does not work for me.

I have the same problem, but in my case this line was added by me. I need it to get Virtualbox (closed version) running with usb support. This does not work anymore, since I had to remove this fstab entry as well. What would be the correct entry for lucid?

Finally, someone with the same problem! :)

"I have the same problem, but in my case this line was added by me. I need it to get Virtualbox (closed version) running with usb support."

Maybe some underlying device system in Ubuntu was changed so for USB support, that line in the fstab is no longer needed (because mountall is no longer the process for USB devices?), and it being there actually causes problems.

Javier, could you please post a comment with the contents of /etc/fstab attached?

Same here: I also added said line to /etc/fstab in order to get VirtualBox working with USB devices. Removing the line lets me boot Lucid just fine, although there are some other issues yet to be solved.

I had this problem with kernels > 2.6.32-9-generic, and trying today after a bunch of updates my older kernels failed as well. After finding this bug report I removed the line in /etc/fstab and that fixed the issue. I never entered that line manually, I'm not sure what added it.

That problem happend to me too.
In my case that line was added because of VirtualBox.

Virtualbox for me too. Lucid boots and Virtualbox USB now works without the fstab entry.

Same problem here.
Annoyingly, I wasn't even able to get into a recovery console because of that error, so I had to boot from another device to correct fstab.

also, I need to get USB in the VirtualBox worked well, please advice

My Virtualbox 3.14 USB just kept working with no action after removing the fstab entry. If your Virtualbox USB broke after removing the fstab entry, check your Virtualbox version. My fstab entry was from an old Virtualbox version maybe a year or two ago prior to an upgrade.

Have the same problem, forgot I added this line to fstab

Alexey, this is the tip from the Virtualbox user forum. You can run the following command in terminal before launching Virtualbox:

sudo hald --daemon=no

Virtualbox should then be able to recognize your USB devices. It works for me.

No, I did that, removed the last line from /etc/fstab, and ran `sudo hald --daemon=no`, and USB still not working for me.

That's the error info. I got when opening the Settings window.

"Could not access USB on the host system, because neither the USB file system (usbfs) nor the DBus and hal services are currently available. If you wish to use host USB devices inside guest systems, you must correct this and restart VirtualBox.
The USB Proxy Service could not be started, because neither the USB file system (usbfs) nor the hardware information service (hal) is available."

Result Code:
VBOX_E_HOST_ERROR (0x00BB0008)
Component:
Host
Interface:
IHost {e380cbfc-ae65-4fa6-899e-45ded6b3132a}
Callee:
IMachine {99404f50-dd10-40d3-889b-dd2f79f1e95e}

Let's not make this fix about solving USB and Virtualbox, it's about Lucid not booting when "none /proc/bus/usb usbfs" is in fstab which could cause problems with upgrades.

Concerning Virtualbox, I had the same error message, but now it works, maybe the Lucid updates fixed it. No need for putting "none /proc/bus/usb usbfs" in fstab anymore.

I am affected too. Any illegal mounting point blocked me from booting. This is a very serious issue.

does it work when you delete the line from the fstab via boot cd?

yeah, it does!

I had the same issue and I had to boot from a CD, mount the filesystem and edit the fstab and comment out the usb entry. After that I was able to boot.

Same here. Had to boot from livecd to comment out this specific usb fs line:

#none /proc/bus/usb usbfs devgid=121,devmode=664 0 0

Same here, commented out that line permits normal boot...

Note that calling magic sysrq K (CTRL+ALT+SysRq+K) brings you a root shell prompt. No need to boot from a live cd.

I rebooted with init=/bin/bash on the kernel parameters and commented out the relevant fstab from there.

I updated from Karmic to Lucid about 12 hours ago and searching the cause for Lucid not booting took a lot of time.
It turned out the problem was this very "usbfs" which I put into fstab to get USB working in VirtualBox. And unlike the description of this bug there was no error message hinting to USB; last thing i saw in recovery-mode was how my ext3 filesystems got mounted and then the booting process just stopped.

I'm pretty sure a large percentage of VirtualBox users edited fstab the same way and all these users will be left with a black screen after trying to boot their updated system, which might even cause some of them to do a complete reinstall of their system.
I suggest that /etc/fstab is parsed during the update process and any lines containing "usbfs" as filesystem should be commented out.

Affecting mountall as that's the service responsible for mounting it. (and stops the boot process if it fails)

Couldn't this simply be solved by adding the following bit of shell-fu (or similar) to the upgrade scripts:
sed 's/\(none \/proc\/bus\/usb usbfs.*\)/#\1/g' /etc/fstab

btw. this would implement up-whatever's more general solution
sed 's/\(.*usbfs.*\)/#\1/g' /etc/fstab

Edit:
Added backup, exclusion of comments and inline editing:
sed -i.old 's/^\([^#].*usbfs.*\)/#\1/g' /etc/fstab

Tom, that would work on an upgrade but that is not a real fix, mountall should not hang, there maybe other non usbfs hanging cases.

Thanks

Thanks for the report.

In this situation (an unknown mount in /etc/fstab), mountall should display an error message on the Plymouth splash screen and prompt allowing you to skip this fileystem or drop to a maintenance shell. Apparently this isn't working for you.

I need some debugging information to be able to understand why this isn't working.

First please provide the output of "ls -l /lib/plymouth/themes"

Secondly edit /etc/init/mountall.conf and add "--debug" to the exec line, after booting attach /var/log/boot.log to this bug

Thanks very much

For the record, this is what you're *supposed* to see ;-)

Also does the following fix this:

  echo FRAMEBUFFER=y | sudo tee /etc/initramfs-tools/conf.d/splash
  sudo update-initramfs -u

Here's what's on my system (Karmic => Lucid):

dphillips@GOLIATH:~$ ls -l /lib/plymouth/themes/
total 4
lrwxrwxrwx 1 root root 34 2010-04-15 11:36 default.plymouth -> /etc/alternatives/default.plymouth
drwxr-xr-x 2 root root 29 2010-04-15 09:52 details
lrwxrwxrwx 1 root root 31 2010-04-15 11:24 text.plymouth -> /etc/alternatives/text.plymouth
drwxr-xr-x 2 root root 4096 2010-04-15 09:53 ubuntu-logo
drwxr-xr-x 2 root root 33 2010-04-15 10:48 ubuntu-text

I don't have a splash screen (just a text boot-up sequence). I'll provide --debug output and a check of the above-referenced initramfs change shortly.

Scott,

Why is this reassigned to plymouth? The plymouth aspect of this bug is the same as bug #563621, but even fixing that won't give us a prompt in this case because of the virtual-filesystems->udev->plymouth-splash dependency chain?

On Tue, 2010-04-20 at 04:46 +0000, Steve Langasek wrote:

> Why is this reassigned to plymouth? The plymouth aspect of this bug is
> the same as bug #563621, but even fixing that won't give us a prompt in
> this case because of the virtual-filesystems->udev->plymouth-splash
> dependency chain?
>
Because I think the only fix is to start plymouth in the initramfs

Scott
--
Scott James Remnant
<email address hidden>

Tested a few times with nightly build from 19. and looks like being fixed. However, I tried only a few times and before the problem did not occur always.

Um, no, this fix hasn't been uploaded yet.

Probably due to other changes probability just changed - will your fix be included in the next nightly build?

Should be included in the next nightly build, but there isn't another nightly build until Thursday night because we're currently in freeze for the release candidate preparations. (Also there are still a couple of other bugs to sort out in plymouth and mountall before uploading.)

just installed 10.04 RC and latest virtualbox, got this error

Result Code:
VBOX_E_HOST_ERROR (0x00BB0008)
Component:
Host
Interface:
IHost {e380cbfc-ae65-4fa6-899e-45ded6b3132a}
Callee:
IMachine {99404f50-dd10-40d3-889b-dd2f79f1e95e}

USB support not working. any help ?

squiddy,
bugs are not the proper replace to request support, specially with issues which are not related with the bug being reported.

Since for USB support you need the closed source version please ask for support at their forum:
http://forums.virtualbox.org/

Thanks

This bug was fixed in the package plymouth - 0.8.2-2ubuntu1

---------------
plymouth (0.8.2-2ubuntu1) lucid; urgency=low

  * src/main.c: if the splash screen isn't up yet, queue message requests
    instead of discarding them. LP: #507881.
  * src/client/ply-boot-client.c: some replies may be sent out of order
    because they depend on user input, so pay attention to the message type
    when picking the handler instead of handing the response to the first
    handler in the list; without this, cancelling fsck in mountall will
    never work. LP: #562811.
  * src/client/ply-boot-client.c: instead of trying to read from the server
    pipe if there are any outstanding requests, call
    ply_event_loop_process_pending_events() which already knows whether we
    can read from the pipe. LP: #559761.
  * add the pixel display bpp symbols to libplymouth2.symbols with a correct
    version, so that packages using them don't wind up with overly-strict
    dependencies on libplymouth2.
 -- Steve Langasek <email address hidden> Sun, 25 Apr 2010 16:15:37 +0100

This bug was fixed in the package mountall - 2.14

---------------
mountall (2.14) lucid; urgency=low

  [ Scott James Remnant ]
  * Flush updates to Plymouth before emitting Upstart events, in case
    the event kills Plymouth. LP: #559761.
  * Don't mark a filesystem "nodev" just because it's got "none" in the
    device column; this will block the "virtual-filesystems" event which
    is the one that can't use Plymouth to prompt. LP: #507881.
  * When cancelling in-progress fsck, don't deference the NULL mount
    record. LP: #562811.
  * mountall is missing a very important line of code that increases the
    udev buffer size; without this it's possible we may miss events during
    busy periods. LP: #561390.

  [ Steve Langasek ]
  * If we're not marking all nodev filesystems as virtual, we need to
    at least mark our placeholder filesystem entries (type=none && dev=none)
    this way.
 -- Steve Langasek <email address hidden> Sun, 25 Apr 2010 23:36:01 +0100

Appears to not be fixed with plymouth 0.8.2-2ubuntu1 and mountall 2.14.

A message is briefly displayed on the screen about a bad filesystem and a
root shell being launched.... but then the screen resolution changes, and a
root shell appears, with a blank space about the prompt. The actual message
is barely a flicker, so users are dumped into a root prompt with effectively
no explaination.

On Sun, Apr 25, 2010 at 23:45, Launchpad Bug Tracker <
<email address hidden>> wrote:

> This bug was fixed in the package mountall - 2.14
>
> ---------------
> mountall (2.14) lucid; urgency=low
>
> [ Scott James Remnant ]
> * Flush updates to Plymouth before emitting Upstart events, in case
> the event kills Plymouth. LP: #559761.
> * Don't mark a filesystem "nodev" just because it's got "none" in the
> device column; this will block the "virtual-filesystems" event which
> is the one that can't use Plymouth to prompt. LP: #507881.
> * When cancelling in-progress fsck, don't deference the NULL mount
> record. LP: #562811.
> * mountall is missing a very important line of code that increases the
> udev buffer size; without this it's possible we may miss events during
> busy periods. LP: #561390.
>
> [ Steve Langasek ]
> * If we're not marking all nodev filesystems as virtual, we need to
> at least mark our placeholder filesystem entries (type=none && dev=none)
> this way.
> -- Steve Langasek <email address hidden> Sun, 25 Apr 2010 23:36:01
> +0100
>
> ** Changed in: mountall (Ubuntu Lucid)
> Status: Fix Committed => Fix Released
>
> --
> Plymouth doesn't show messages sent before the splash screen is visible
> https://bugs.launchpad.net/bugs/507881
> You received this bug notification because you are a direct subscriber
> of the bug.
>

--
Jeremy Nickurak -= Email/XMPP: <email address hidden> =-

Appears to not be fixed with plymouth 0.8.2-2ubuntu1 and mountall 2.14.

A message is briefly displayed on the screen about a bad filesystem and a root shell being launched.... but then the screen resolution changes, and a root shell appears, with a blank space about the prompt.  The actual message is barely a flicker, so users are dumped into a root prompt with a read-only filesystem, with effectively no explaination.

Even the flickering message isn't really sufficient. You get:

"General error mounting filesystems.
A maintenance shell will now be started.
CONTROL-D will terminate this shell and reboot the system.
root@hostname:~# ..."

... with none of the actual errors that caused the problem, so I have no idea what's failing. Is there another log where this information could end up?

You are describing a completely separate bug. Please file a separate bug report.

In this case, it seems to be an ntfs-3g partition causing the problem.

Commenting out the line:

"/dev/sda1 /mnt/c ntfs defaults 0 0"

in /etc/fstab gets it booting again. Once the system is booted, I can uncomment it and mount it fine.

No way of telling what the error is currently.

The ntfs-fails-to-mount bug is different, certainly.

Is the plymouth-and-mountall-don't-report-any-error-messages issue the same bug though?

Filed https://bugs.launchpad.net/ubuntu/+source/plymouth/+bug/570848 anyways.

Jeremy,
the plymouth-and-mountall-don't-report-any-error-messages symptom which lead to this bug report is not related to the one you are describing, on the bug case from this report there was no message sent to the user at all. I mean not even reporting a general mount failure.

Thanks

To get VirtualBox 3.1.6 working in Lucid, add the user on vboxusers groups and restart your pc.

i have VirtualBox 3.1.6 on Lucid RC with my user added to the vboxusers groups and don't work!!

VirtualBox OSE or PUEL?

I don't know why, but OSE don't have USB, so I replaced it by the VirtualBox 3.1.6 PUEL from Karmic and it's works fine.

Yesterday, I upgraded from koala to lucid and had also boot problems. By removing the usbfs lines in /etc/fstab, the boot problems are solved. But is it normal the location '/proc/bus/usb' disappeared since upgrading?
And I am still not available to let usb working in virtualbox, before removing the lines in fstab, and after removing. Closed version of VB, I am added to the group vboxusers, and tried 'sudo hald --daemon=no'... Nothing seems to work. Does anyone have a possible sollution?

Apperently, in Lucid 'hald' doesn't start automatically, so you need to execute it yourself by
sudo hald
before using VB, then usb works fine!

I was having problem getting any kernel past 2.6.31 to boot with the virtualbox fstab line in place, and while I am aware of the sudo hald --daemon=no claimed fix, I haven't needed it.

The Sun/Oracle Virtualbox 3.2 beta for lucid appears to deal with usb just fine with the
# none /proc/bus/usb usbfs auto,devmode=0666 0 0
commented out as shown.... and usb appears to work correctly, I just printed from Virtualbox with an XP guest.

Note: looks like the menu didn't get updated correctly, put in the icon properties file:
/usr/bin/VirtualBox

While the mounting problem appears to be one that should be solved, it appears to no longer be a problem for Virtualbox users ... comment out the infamous line above and install the Virtualbox 3.2.0 BETA1 for Lucid, change the command in the icon as above, and open Virtualbox.

For more information, go to the end of this discussion:
http://forums.virtualbox.org/viewtopic.php?f=1&t=27820