Comment 15 for bug 376193

This is neither a problem of CUPS nor of OpenOffice.org. If you have another printer, the job will most probably print without problems. The problem is that the printer driver executable, hpijs crashes.

One can test it easily without CUPS, Foomatic, and OpenOffice.org. One simply runs the Ghostscript command line as logged in the error_log manually, using an arbitrary PDF file with more than one page:

gs -dFirstPage=1 -q -dBATCH -dPARANOIDSAFER -dQUIET -dNOPAUSE -sDEVICE=ijs -sIjsServer=hpijs -dDEVICEWIDTHPOINTS=612 -dDEVICEHEIGHTPOINTS=792 -sDeviceManufacturer="HEWLETT-PACKARD" -sDeviceModel="deskjet 5600" -dDuplex=true -dTumble=false -r300 -sIjsParams=Quality:Quality=4,Quality:ColorMode=0,Quality:MediaType=0,Quality:PenSet=2,Quality:SpeedMech=1,PS:MediaPosition=7 -dIjsUseOutputFD -sOutputFile=- 3-pages-test.pdf > test.prn

After some error messages a Segmentation Fault is reported. /var/log/syslog contains:

Jun 22 21:37:02 till-laptop hpijs[29685]: prnt/hpijs/hpijs.cpp 381: unable to set key=Quality:SpeedMech, value=1
Jun 22 21:37:03 till-laptop last message repeated 16 times
Jun 22 21:37:03 till-laptop kernel: [625096.106277] hpijs[29685]: segfault at 18 ip 00007f7fc4a7ed88 sp 00007fff62356c40 error 4 in libc-2.9.so[7f7fc4a05000+168000]

Once it seems that hpijs gets supplied a parameter which is not available, at least for the currently selected printer. Second, hpijs (not Ghostscript) crashes with a segmentation fault.

The segmentation fault is perhaps a consequence of the wrong parameter, but in such a case hpijs should exit with a decent error message. For example it could be made to exit at the first "unable to set key=Quality:SpeedMech, value=1".

So this is an upstream bug of HPLIP.

eapache, can you remove your print queue and create it again to see whether this makes the problem go away?

eapache, can you try a live CD of Karmic to see whether the bug disappears there?