Ubuntu
plymouth package

mountall assertion failure breaks boot process

Bug #570289 reported by Andreas Schiffer
16
This bug affects 2 people
Affects Status Importance Assigned to Milestone
plymouth (Ubuntu)
Fix Released
High
Steve Langasek

Bug Description

Binary package hint: mountall

I have just upgraded from Ubuntu Karmic to Ubuntu Lucid.
I have a system that has a few LUKS file systems running (e.g. home and swap), so that during the boot process a password is requested and it takes a few seconds until all LUKS devices are opened and mounted.
With karmic this went all fine.

Now after the upgrade to lucid, mountall does not seem to be able to cope with my setup and terminates with an ABRT signal which makes the init process stop and fall back to the root command prompt (no GUI yet available). If I enter "mountall" again manually on the root command prompt, then the system finishes booting and runs without further problems.

With mountall version 2.14 the only way to get my system running is to manually enter "mountall" on the root command prompt - as described above.
With mountall version 2.13, I also get dumped to the root command prompt, but a few seconds later the system finishes booting without manual interaction from my side. So I will stay with 2.13 until this ugly bug is fixed.

The error messages that I get before getting dumped to the root command prompt are:
> mountall: ./ply_boot_client.c:451: ply_boot_client_process_pending_request: Assertion ply_list_get_length failed.
> init: mountall main process killed by ABRT signal

Also I get the error message "could not write bytes: Broken pipe" a few times, but I'm not sure if this is related to my problem.

Related branches

lp:ubuntu/lucid/plymouth
Scott James Remnant (Canonical) (canonical-scott)
affects: mountall (Ubuntu) → plymouth (Ubuntu)
Changed in plymouth (Ubuntu):
assignee: nobody → Steve Langasek (vorlon)
Steve Langasek (vorlon)
Changed in plymouth (Ubuntu):
status: New → Triaged
importance: Undecided → High
Revision history for this message
Steve Langasek (vorlon) wrote :

Hmm; at first blush I thought this was a regression caused by the fix for bug #559761, but now I'm not so sure, since ply_event_loop_process_pending_events doesn't appear to call ply_boot_client_process_pending_request. Andreas, can you help us get a full backtrace of this mountall crash?

Changed in plymouth (Ubuntu):
status: Triaged → Incomplete
Revision history for this message
Andreas Schiffer (andreas-schiffer) wrote :

Sure, I can reproduce it and get you a backtrace.
Unfortunately I never had to create a backtrace on Linux before, so I don't know how to do it.
Do you have a short list of steps to take or something like a wiki page that explains how to do it?

Revision history for this message
Launchpad Janitor (janitor) wrote :

This bug was fixed in the package plymouth - 0.8.2-2ubuntu2

---------------
plymouth (0.8.2-2ubuntu2) lucid; urgency=low

  * Don't call ply_boot_client_process_pending_requests on flush, because
    ply_event_loop_process_pending_events already calls this for us when the
    fd is writable - so calling it again causes a race condition and an
    assertion failure if _process_pending_events has already cleared the
    event queue for us. LP: #570289.
 -- Steve Langasek <email address hidden> Tue, 27 Apr 2010 09:35:00 +0100

Changed in plymouth (Ubuntu):
status: Incomplete → Fix Released
Revision history for this message
Steve Langasek (vorlon) wrote :

Andreas,

We should have this bug fixed now. You should find after manually bringing your system up and installing the new plymouth packages that have just been published that this is resolved for you. Thanks for taking the time to report this bug - not a pretty bug to have run into the week of the release, but your report was instrumental in us getting it fixed!

Revision history for this message
Andreas Schiffer (andreas-schiffer) wrote :

I upgraded the plymouth packages and got back to version 2.14 of mountall and everything runs fine now. I tried re-booting three times and had no problems (before the bug occurred with every boot).
So I can confirm that this bug is fixed for me.
Thanks a lot, Steve, for getting this fixed so fast!

Revision history for this message
chef (adotei) wrote :

I can confirm that after upgrading today (27/02/10) and rebooting at least 3 times, the error/halt did not occur. The problem seems to have been rectified in the latest plymouth update.
Thanks guys.

Revision history for this message
chef (adotei) wrote :

sorry, got the date wrong. Should be 27/01/10.

Revision history for this message
chef (adotei) wrote :

Damn, got it wrong again. Should be 27/04/10.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Subscribing...

Other bug subscribers

Remote bug watches

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