cupsd crashed with SIGSEGV in cupsArrayFind()

[pinzia]

Binary package hint: cups

I have a system (lucid Lynx 64bit) with 2 usb printers (ml2010 and ml1710), but now i use ML-1710. I created a group for this printer and I share with a windows 2000 machine (not with samba, but with cups and ipp). Often when i try to print (from localhost or from win2000) the printer don't print.

In terminal I see crash cupsd. If i restart cups, the printer print.
This crash-problem happens often

Sorry for my bad english

ProblemType: Crash
DistroRelease: Ubuntu 10.04
Package: cups 1.4.3-1ubuntu1.2
ProcVersionSignature: Ubuntu 2.6.32-24.42-generic 2.6.32.15+drm33.5
Uname: Linux 2.6.32-24-generic x86_64
Architecture: amd64
CrashCounter: 1
Date: Thu Sep 9 07:35:06 2010
ExecutablePath: /usr/sbin/cupsd
InstallationMedia: Ubuntu 10.04 LTS "Lucid Lynx" - Release amd64 (20100427.1)
Lpstat:
 device for ML-1710: usb://Samsung/ML-1710
 device for ML-2010: usb://Samsung/ML-2010
 device for Stampanti: ///dev/null
MachineType: HP ProLiant ML110 G5
Papersize: a4
PpdFiles:
 ML-2010: Samsung ML-2010 Foomatic/gdi
 ML-1710: Samsung ML-1710 Foomatic/gdi
ProcAttrCurrent: /usr/sbin/cupsd (enforce)
ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.32-24-generic root=UUID=cf44562a-c6f1-4a3d-8dfe-dbbf5d1bff9c ro quiet splash
ProcCmdline: /usr/sbin/cupsd -C /etc/cups/cupsd.conf
ProcCwd: /
ProcEnviron:
 PATH=(custom, no user)
 LANG=it_IT.UTF-8
SegvAnalysis:
 Segfault happened at: 0x7f89d2905b0f: mov (%rcx,%rsi,8),%rsi
 PC (0x7f89d2905b0f) ok
 source "(%rcx,%rsi,8)" (0x7f892b937e00) not located in a known VMA region (needed readable region)!
 destination "%rsi" ok
SegvReason: reading unknown VMA
Signal: 11
SourcePackage: cups
StacktraceTop:
 ?? () from /usr/lib/libcups.so.2
 cupsArrayFind () from /usr/lib/libcups.so.2
 ?? ()
 ?? ()
 ?? ()
Title: cupsd crashed with SIGSEGV in cupsArrayFind()
UserGroups:

dmi.bios.date: 10/09/2009
dmi.bios.vendor: HP
dmi.bios.version: O15
dmi.board.asset.tag: No Asset Tag
dmi.board.name: ProLiant ML110 G5
dmi.board.vendor: Wistron Corporation
dmi.board.version: NA
dmi.chassis.asset.tag: xeon
dmi.chassis.type: 7
dmi.chassis.vendor: HP
dmi.chassis.version: N/A
dmi.modalias: dmi:bvnHP:bvrO15:bd10/09/2009:svnHP:pnProLiantML110G5:pvrNA:rvnWistronCorporation:rnProLiantML110G5:rvrNA:cvnHP:ct7:cvrN/A:
dmi.product.name: ProLiant ML110 G5
dmi.product.version: NA
dmi.sys.vendor: HP

[Apport retracing service (apport)]

StacktraceTop:
 ?? () from /usr/lib/libcups.so.2
 cupsArrayFind () from /usr/lib/libcups.so.2
 ?? () from /usr/sbin/cupsd
 ?? () from /usr/sbin/cupsd
 ?? () from /usr/sbin/cupsd

[Jan-Marek Glogowski (jmglogow)]

I'm using a Lucid cups backport to Hardy having the same issue. I rebuild my backport with --enable-debug and got a useful backtrace. I have a VM, with a reproducible test case.

[Jan-Marek Glogowski (jmglogow)]

After some investigation, I found that cupsd always crashes in cupsdDeleteAllPrinters (cups_array_find) when send a SIGHUP (cups reload) and I have setup a class.

It always crashes with my attached config files. If I leave the classes.conf empty, it won't crash. As it crashes in cupsdFindPolicyOp I added a default Policy to cupsd.conf, but the crash persists.

I've attached the config without the default policy.

[Jan-Marek Glogowski (jmglogow)]

I've also opened http://www.cups.org/str.php?L3682+P0+S-2+C0+I0+E0+M20+Q

[Jan-Marek Glogowski (jmglogow)]

As you can see in the CUPS bug report, this is fixed in CUPS 1.4.4. I tested a Backport from Maverick, which doesn't crash on SIGHUP.

[Till Kamppeter (till-kamppeter)]

Marking as fixed as it is fixed in Maverick. If you need it fixed in Lucid, please open a Lucid task.

[contemno (contemno)]

I am still seeing this bug in Lucid. I'm new here, how would I open a Lucid task?

[Till Kamppeter (till-kamppeter)]

Everyone having this problem please update your system. The fix is in cups 1.4.3-1ubuntu1.6. See bug 857663.