| Summary: | Landscape prints as truncated portrait | ||||||
|---|---|---|---|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Jonathan Dieter <jdieter> | ||||
| Component: | cups | Assignee: | Tim Waugh <twaugh> | ||||
| Status: | CLOSED CURRENTRELEASE | QA Contact: | Fedora Extras Quality Assurance <extras-qa> | ||||
| Severity: | unspecified | Docs Contact: | |||||
| Priority: | unspecified | ||||||
| Version: | 16 | CC: | caolanm, dtardon, erack, jpopelka, ltinkl, mstahl, sbergman, twaugh | ||||
| Target Milestone: | --- | ||||||
| Target Release: | --- | ||||||
| Hardware: | x86_64 | ||||||
| OS: | Linux | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | Doc Type: | Bug Fix | |||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2012-02-29 16:06:55 UTC | Type: | --- | ||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Attachments: |
|
||||||
|
Description
Jonathan Dieter
2012-02-24 06:09:23 UTC
Isn't this cups, not us that's doing the truncation ? Sounds like #740496 Can we get the output of... rpm -q cups ghostscript-cups please On the desktops: $ rpm -q cups ghostscript-cups cups-1.5.2-1.fc16.x86_64 ghostscript-cups-9.04-7.fc16.x86_64 On the print server: # rpm -q cups ghostscript-cups cups-1.4.2-44.el6.x86_64 package ghostscript-cups is not installed Though we have the same problem printing directly from the desktops to an attached printer. (All the printers are HP if that has anything to do with anything.) Could this be bug #768811? Try using the hpcups driver instead of hpijs (see bug #768811 comment #10, obviously choose the appropriate model instead of what's written there). Ok, that was an interesting couple of hours. As mentioned above we have two printers attached to a CentOS server that the desktops print to. Switching the driver for both printers to hpcups from hplip caused landscape documents printed from Fedora 16 using PDF as the printer language to completely fail to print. The print jobs just hang on the Fedora desktops, with ps showing the ipp job just sitting there sleeping. Portrait documents print immediately. As for the printer attached directly to the Fedora 16 desktop, it's using the foo2zjs driver and attempting to switch to hpcups or hplip resulted in hung print jobs. I even tried to run hp-setup -i and it still didn't work. Reverting to foo2zjs allowed me to print again, but with the same original bug of landscape jobs being truncated. If I'm doing something stupid, please let me know. If you want me to set up a Fedora print server instead of CentOS, I'll happily do that (it's just a VM). And if I need to split this into a couple of bug reports (obviously foo2zjs isn't in Fedora, though I'd love to get that system running with hpcups) I'll happily do that too. FWIW, the printers are: Attached to print server HP LaserJet 1505n HP LaserJet P2015n Attached to the desktop: HP LaserJet P1606dn Let's concentrate on the P1505n and P2015n for the moment. It sounds like the jobs are hitting some sort of error when the server processes them. Could you try running 'cupsctl --debug-logging' on the print server, and sending it a landscape job again? Then look in /var/log/cups/error_log -- you should see some sort of error message somewhere (errors have 'E' at the beginning of the line), and the lines leading up to that will help to give that some context. (Yes, aware this is CentOS -- but it might be possible to identify a similar bug in Fedora or Red Hat Enterprise Linux based on debugging it...) Created attachment 566068 [details]
CUPS error log
This is the trimmed output of:
# cat /var/log/cups/error_log | grep "^E" -B 100 -A 220
Let me know if you need any more
Here are the interesting bits: D [27/Feb/2012:17:07:06 +0200] [Job 2146] printer-state-reasons=�marker-supply-low-warning Wonder what those strange characters are doing there? D [27/Feb/2012:17:07:02 +0200] [Job 2146] prnt/hpcups/HPCupsFilter.cpp 493: cupsRasterOpen failed, fd = 0 D [27/Feb/2012:17:07:02 +0200] PID 19686 (/usr/lib/cups/filter/hpcups) stopped with status 1! E [27/Feb/2012:17:07:02 +0200] PID 19685 (/usr/lib/cups/filter/pstoraster) crashed on signal 11! This looks like the pstoraster process, i.e. ghostscript, crashed causing truncated output; that in turn caused hpcups to fail because it hadn't finished reading the CUPS Raster header. It would be interesting to see whether Fedora 16 also fails in this way. My guess is that it won't, and that bug #639593 is the root cause of the CentOS 6 failure. Needinfo: whether pstoraster crashes in Fedora 16 too. I've just moved the printer server over to a Fedora 15 VM, but the printers themselves are powered down for the night. I'll test in the morning and see if they work in landscape. Thanks for the guidance! Ok, that did the trick. I'm able to print just fine in both portrait and landscape using Fedora 15 as the print server. I noticed that you reverted the patch in bug #639593. Is there a working patch that can be applied to ghostscript-8.70? Again, a huge thank you for your help. I think you want a26ed9754118ab5a6027346f2e4d0c41077dfc1c, but I expect it depends on earlier patches. Closing as CURRENTRELEASE as currently supported versions of Fedora work correctly. Thanks again for your help. |