aptd crashed with SIGSEGV in debListParser::LoadReleaseInfo()

StacktraceTop:
 debListParser::LoadReleaseInfo (this=0x7fff6acb9ea0, FileI=..., File=<optimized out>, component=<optimized out>) at deb/deblistparser.cc:860
 debPackagesIndex::Merge (this=0x2bfaf20, Gen=<optimized out>, Prog=<optimized out>) at deb/debindexfile.cc:342
 BuildCache (Gen=..., Progress=0x7fff6acbafe0, CurrentSize=@0x7fff6acbaf00, TotalSize=83177485, Start=..., End=...) at pkgcachegen.cc:1067
 pkgCacheGenerator::MakeStatusCache (List=<optimized out>, Progress=0x7fff6acbafe0, OutMap=0x29bdf50, AllowMem=192) at pkgcachegen.cc:1261
 pkgCacheFile::BuildCaches (this=0x29bdf40, Progress=0x7fff6acbafe0, WithLock=<optimized out>) at cachefile.cc:80

Status changed to 'Confirmed' because the bug affects multiple users.

synaptic crashes quite regularly due to this

Raising the importance to High because of the high number of duplicates, as all packages using libapt seem to be affected.

This started on 2011-09-20, that is the day when 0.8.16~exp5ubuntu10 got uploaded.

But 0.8.16~exp5ubuntu10 got not accepted before the 22th, so that is unlikely.

Aptdaemon got update on the 19th to 0.43+bzr697 but the changes there do not look releated at all.

Plus synaptic (e.g. bug #859355) is affected as well so its unlikely that aptd itself is the problem.

Can someone try to reproduce this problem using a terminal (e.g. gnome-terminal) please:
It would be nice to know if
$ sudo apt-cache gencaches -o Debug::pkgCacheGen=true
crashes.

$ sudo apt-cache gencaches -o Debug::pkgCacheGen=true

does not crash on my machine

On 26/09/11 11:14, Michael Vogt wrote:
> Plus synaptic (e.g. bug #859355) is affected as well so its unlikely
> that aptd itself is the problem.
>
> Can someone try to reproduce this problem using a terminal (e.g. gnome-terminal) please:
> It would be nice to know if
> $ sudo apt-cache gencaches -o Debug::pkgCacheGen=true
> crashes.
>
Did not crash for me. (Latest updates checked and applied imediately
before test).

sudo apt-cache gencaches -o Debug::pkgCacheGen=true

gives

...loads of lines simialr to those below...
en
(/var/lib/apt/lists/archive.canonical.com_ubuntu_dists_natty_partner_i18n_Translation-en):
Has NO packages
Checking PkgFile http://archive.canonical.com/ubuntu/ natty/partner
Translation-en_GB
(/var/lib/apt/lists/archive.canonical.com_ubuntu_dists_natty_partner_i18n_Translation-en%5fGB):
Has NO packages
Checking PkgFile /var/lib/dpkg/status: with ID 52 is valid
pkgcache.bin is valid - no need to build anything
Reading package lists... Done

Hope this helps

Chris

This bug was fixed in the package apt - 0.8.16~exp5ubuntu12

---------------
apt (0.8.16~exp5ubuntu12) oneiric; urgency=low

  [ David Kalnischkies ]
  * apt-pkg/deb/deblistparser.cc:
    - fix crash when the dynamic mmap needs to be remapped during
      LoadReleaseInfo (LP: #854090)
 -- Michael Vogt <email address hidden> Mon, 26 Sep 2011 13:31:11 +0200

I just saw this again after updating and rebooting.

apt-cache madison reports:
apt | 0.8.16~exp5ubuntu12

So I do have the version which was supposed to contain a fix.

just got it today after couple of hundred updates

This is fixed in apt now, I close the update-manager task.

This bug was fixed in the package update-manager - 1:0.152.20

---------------
update-manager (1:0.152.20) oneiric; urgency=low

  * DistUpgrade/DistUpgradeQuirks.py:
    - increase the default cache size on a multiarch system to
      avoid potential crash in natty apt (LP: #854090)
  * DistUpgrade/DistUpgradeController.py, UpdateManager/Core/utils.py:
    - do not leak password from sources.list entries into the logfile
      (LP: #839094)
  * UpdateManager/UpdateManager.py:
    - do not crash if a package can not be put into "install" state,
      instead, just keep the old (unmarked) state (LP: #850482)
  * UpdateManager/DistUpgradeFetcher.py:
    - fix crash for changed gtk2 -> gtk3 API (LP: #859862)
  * UpdateManager/backend/InstallBackendAptdaemon.py:
    - remove debug output (LP: #855495)
 -- Michael Vogt <email address hidden> Fri, 30 Sep 2011 16:09:55 +0200

It happened again today. How is this fixed?

I've gotten a report of this in 12.04: https://bugs.kde.org/show_bug.cgi?id=299800

Since it doesn't seem to be fixed, I'm re-opening the bug.

We can see this still occurring regularly with later releases of Ubuntu at the Ubuntu Error Tracker.

https://errors.ubuntu.com/problem/1859029e9f0d88ba0e587718572a0d57e56f72c6

Bug 957231 is the most recent occurrence of this crash.

Should this be marked Fixed now that Bug #957231 is fixed?

Marking as fixed, as bug 957231 is marked fixed.