print system hangs when printing

[Hadmut Danisch]

Binary package hint: cups

Hi,

I have a HP Laserjet 1320 and a Laserjet 1300. The old problem from former ubuntu releases still exists with a completely fresh installed 10.04: When printing (especially from xpdf or evince) the system hangs: Printer starts blinking and never prints, job hangs forever.

There is definitely something wrong in the processing queue, which causes a problem somewhere in the flow of the document. And even if it prints, something is missing at the top and the bottom of the page. The printing logic is still awful. I don't even see where to start debugging. (and the problem I've reported some time ago that the lpadmin path for finding ppd files is wrong is also not fixed)

Looks as if ubuntu printing has been abandoned. It is a nightmare. The only way to reliable print is to convert everything manually into ps and use lpr -l , but it's not the user's task to manually do what's the job of the print system.

ProblemType: Bug
DistroRelease: Ubuntu 10.04
Package: cups 1.4.3-1
ProcVersionSignature: Ubuntu 2.6.32-22.33-generic 2.6.32.11+drm33.2
Uname: Linux 2.6.32-22-generic x86_64
Architecture: amd64
Date: Sun May 16 22:09:32 2010
InstallationMedia: Ubuntu 10.04 LTS "Lucid Lynx" - Release amd64 (20100427.1)
Lpstat:
 device for Deskjet-F2100-series: hp:/usb/Deskjet_F2100_series?serial=CN7CK4T2NP04TK
 device for HP-LaserJet-1300: socket://192.168.160.1:9100
 device for hp-LaserJet-1320-series: hp:/usb/hp_LaserJet_1320_series?serial=00CNM1G08790
 device for Laser: ///dev/null
 device for PDF: cups-pdf:/
MachineType: Gigabyte Technology Co., Ltd. GA-MA78G-DS3H
Papersize: a4
PpdFiles:
 PDF: Generic CUPS-PDF Printer
 HP-LaserJet-1300: HP LaserJet 1300 Series Postscript (recommended)
 hp-LaserJet-1320-series: HP LaserJet 1320 series Postscript (recommended)
 Deskjet-F2100-series: HP Deskjet f2100 Series hpijs, 3.10.2
ProcCmdLine: BOOT_IMAGE=/vmlinuz-2.6.32-22-generic root=/dev/mapper/raid2-ubuntu64 ro quiet splash
ProcEnviron:
 PATH=(custom, user)
 LANG=en_US.utf8
 SHELL=/bin/tcsh
SourcePackage: cups
dmi.bios.date: 04/03/2008
dmi.bios.vendor: Award Software International, Inc.
dmi.bios.version: F2
dmi.board.name: GA-MA78G-DS3H
dmi.board.vendor: Gigabyte Technology Co., Ltd.
dmi.board.version: x.x
dmi.chassis.type: 3
dmi.chassis.vendor: Gigabyte Technology Co., Ltd.
dmi.modalias: dmi:bvnAwardSoftwareInternational,Inc.:bvrF2:bd04/03/2008:svnGigabyteTechnologyCo.,Ltd.:pnGA-MA78G-DS3H:pvr:rvnGigabyteTechnologyCo.,Ltd.:rnGA-MA78G-DS3H:rvrx.x:cvnGigabyteTechnologyCo.,Ltd.:ct3:cvr:
dmi.product.name: GA-MA78G-DS3H
dmi.sys.vendor: Gigabyte Technology Co., Ltd.

[Hadmut Danisch (hadmut)]

As far as I could find by now the backends (usb, network,...) are dying with a segfault after reading

/usr/share/hplip/data/models/models.dat

[Till Kamppeter (till-kamppeter)]

The printing system consists of many programs coming from different packages. Can you find out via the /var/log/syslog file which executable segfaults?

[Till Kamppeter (till-kamppeter)]

Can you provide an error_log following the instructions in the "CUPS error_log" section of https://wiki.ubuntu.com/DebuggingPrintingProblems? Thanks.

[Hadmut Danisch (hadmut)]

No, not really.

Maybe the segfault came because I tried to run the programs manually (with mimiking this silly ARGV[0] calling method), but without the required environment variables. It was the hp usb driver which died after reading the models.dat, but maybe this was caused by me. The fact that these programs can segfault by being called without proper environment is a sign of poor quality.

Debugging ubuntu-cups is a nightmare, since it is not obvious what is going on and not trivial to fetch the outputs of the several stages. I'm having these printing problems since around 9.04, each time hoping that the next release would fix it.

There should be a way to have a copy of the in and output of each stage for debugging, and there should be a documentation of what happens at all. It is not obvious what happens with the print job after submission.

My current guess is that somewhere in the processing of the postscript stuff something goes wrong, because I am under the impression (but not sure) that the problems occur more often when printing pdf files (from xpdf or evince), which means that the problem depends on the contents of the file.

Even when the printer prints, in these cases top and bottom are cropped often. I therefore tend to process everything manually from the command line into the final postscript and use lpr -l

I've seen that ubuntu-cups print processing generates a postscript and runs it through a filthy shell script (cpdftocps) with heaps of sed scripts and other programs. This is probably a bad idea. Too many things can go wrong when processing postscript with sed scripts. About fifteen years ago I ran into a similar problem where I tried to process postscript with such methods. It sometimes worked, sometimes not. I found that some of the sed/awk/... scripts were triggering on usual postscript commans, which accedently occured in char-encoded pictures as well. It's definitely a very bad idea to process Postscript with things like sed. That should be abandoned.

I'd propose to throw that shell script away and write a single program in ruby or whatever that first properly parses the postscribt, dissects into header and pages, and then performs well defined operations.

Unfortunately I am quite busy at the moment and can't afford to spent hours and days into debugging print system with which's internals i am not familiar with.

If it at least was possible to simulate all steps cups performs from the command line easier, that would help.

[Hadmut Danisch (hadmut)]

That's the requested error log

I started to print a five page PDF file with evince, that always causes the printer to print four pages and then remain blinking. Unfortunately, the contents of the file are confidential, so I can't post it here.

regards
Hadmut

[Sesivany (jiri-eischmann)]

This bug may be duplicate of this one:
https://bugs.launchpad.net/ubuntu/+source/cups/+bug/634058

I'm not sure, therefore I don't want to mark it as duplicate.

[Launchpad Janitor (janitor)]

[Expired for hplip (Ubuntu) because there has been no activity for 60 days.]