/usr/lib/cups/filter/pdftoraster failed
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| | GS-GPL |
Fix Released
|
High
|
||
| | ghostscript (Fedora) |
Fix Released
|
High
|
||
| | ghostscript (Ubuntu) |
High
|
Unassigned | ||
| | Lucid |
High
|
Unassigned | ||
Bug Description
Binary package hint: cups
I tried to print some file using Chromium. I got the message that printing has failed. After diagnosing, I get the message:
/usr/lib/
This is on a Canon iP4500 printer. Printing to the pdf printer still works. Also, a test page does work. So, I guess it has something to do with the printed file.
(which is created using Chromium and GMail:
- start chromium
- go to gmail
- select conversation
- click print in drop down menu)
I noticed a couple of things:
- Printing the file directly from Chromium does not work.
- Printing from Chromium to pdf and printing the created pdf does not work also.
- However, printing from Firefox does work!
- Another webpage is printed fine using Chromium
As the file contains some private information, I cannot attach it here.
ProblemType: Bug
Architecture: amd64
Date: Tue Mar 16 17:56:42 2010
DistroRelease: Ubuntu 10.04
InstallationMedia: Ubuntu 10.04 "Lucid Lynx" - Alpha amd64 (20100122)
Lpstat:
device for Canon-iP4500-
device for Canon-iP4500-
device for PDF: cups-pdf:/
MachineType: MICRO-STAR INTERNATIONAL CO.,LTD MS-7512
Package: cups 1.4.2-9
Papersize: letter
PpdFiles:
Canon-
PDF: Generic CUPS-PDF Printer
Canon-
ProcCmdLine: BOOT_IMAGE=
ProcEnviron:
LANGUAGE=
PATH=(custom, user)
LANG=en_US.utf8
SHELL=/bin/bash
ProcVersionSign
SourcePackage: cups
Uname: Linux 2.6.32-16-generic x86_64
dmi.bios.date: 03/09/2009
dmi.bios.vendor: American Megatrends Inc.
dmi.bios.version: V3.4
dmi.board.
dmi.board.name: P45 Neo2-FR (MS-7512)
dmi.board.vendor: MICRO-STAR INTERNATIONAL CO.,LTD
dmi.board.version: 1.0
dmi.chassis.
dmi.chassis.type: 3
dmi.chassis.vendor: MICRO-STAR INTERNATIONAL CO.,LTD
dmi.chassis.
dmi.modalias: dmi:bvnAmerican
dmi.product.name: MS-7512
dmi.product.
dmi.sys.vendor: MICRO-STAR INTERNATIONAL CO.,LTD
|
|
#53 |
Please attach the PPD for the queue you are using. You can find this in the /etc/cups/ppd directory.
|
|
#54 |
Created attachment 390356
ppd of the printer queue
|
|
#55 |
Thanks. This is some kind of regression in the ghostscript cups driver.
|
|
#56 |
Created attachment 390508
in.ps.bz2 (PostScript input for reproducer)
This file, along with the PPD, shows the problem.
1. bunzip2 in.ps.bz2
2. PPD=tp0.ppd gs -sDEVICE=cups -r600x600 -dSAFER -dNOPAUSE -dBATCH -sOutputFile=
3. rasterview out.rast
|
|
#57 |
*** Bug 565804 has been marked as a duplicate of this bug. ***
|
|
#58 |
ghostscript-
http://
|
|
#59 |
ghostscript-
|
|
#60 |
ghostscript-
http://
|
|
#62 |
ghostscript-
http://
|
|
#63 |
Using Fedora 12 and turboprint-
Update to ghostscript-
Dropped back to ghostscript-
Error in /var/log/messages :-
Feb 19 18:14:41 mail kernel: tpu[3334]: segfault at 64 ip 00d53bed sp b7809154 error 4 in libpthread-
Feb 19 18:14:41 mail abrtd: Directory 'ccpp-126660328
Feb 19 18:14:41 mail abrtd: Lock file '/var/cache/
Feb 19 18:14:42 mail abrt[3356]: saved core dump of pid 3323 (/usr/lib/
Feb 19 18:14:42 mail abrtd: Executable '/usr/lib/
Feb 19 18:14:42 mail abrtd: Corrupted or bad crash, deleting
|
|
#64 |
ghostscript-
If you want to test the update, you can install it with
su -c 'yum --enablerepo=
|
|
#65 |
Still only ghostscript-
sudo rpm -Uvh /data/ghostscri
error: Failed dependencies:
libgs.so is needed by ghostscript-
libgs.so.8 is needed by (installed) libspectre-
sudo yum provides /usr/lib/libgs.so.8
ghostscript-
Repo : installed
Matched from:
Other : Provides-match: /usr/lib/libgs.so.8
Which has nothing to do with libspectre. 8.70-1 installs fine. Should 8.70-2 really be checking for libgs.so ?
rpm -qlp /data/ghostscri
/usr/lib/libgs.so.8
/usr/lib/
|
|
#66 |
Removed ghostscript and dependencies, e.g.turboprint using sudo yum remove ghostscript. Still complaining about libgs.so , so
sudo rpm -Uvh --nodeps ghostscript-
sudo rpm -Uvh turboprint-
This time nothing actually get sent to the printer and no error messages. Turboprint print monitor just shows "Printer: USB error". Reinstalled 8.70-2 and works fine, monitor says "Printer: idle".
|
|
#67 |
ghostscript-
http://
|
|
#68 |
Installed ghostscript-
|
|
#69 |
8.71-4.fc12 finally worked for me, too, older version segfaulted with turboprint
gs[18071]: segfault at 2 ip 00000002 sp bfd20b64 error 4 in gs[755000+1000]
gs[18077]: segfault at 9 ip 00000009 sp bfd861c4 error 4
gs[18114]: segfault at 2 ip 00000002 sp bfdd4264 error 4
gs[18119]: segfault at 9 ip 00000009 sp bff1b694 error 4
|
|
#70 |
ghostscript-
If you want to test the update, you can install it with
su -c 'yum --enablerepo=
|
|
#71 |
ghostscript-
|
|
#76 |
In ghostscript-
|
|
#77 |
ghostscript-
http://
|
|
#78 |
ghostscript-
http://
|
|
#79 |
ghostscript-
http://
| Ernst (ernst-blaauw) wrote : | #1 |
Here is a pdf (without too much private info) causing the error. Does this
work for you?
| Anton Kraus (done) wrote : | #3 |
I have the same problem with an Epson Stylus DX 4050.
Printing a test page works, but trying to print a pdf always fails with the same error.
| Ernst (ernst-blaauw) wrote : | #4 |
I've set the status to 'confirmed' because 2 users have this issue.
| Changed in cups (Ubuntu): | |
| status: | New → Confirmed |
| affects: | cups (Ubuntu) → ghostscript (Ubuntu) |
| Changed in gs-gpl: | |
| status: | Unknown → New |
| Anton Kraus (done) wrote : | #5 |
When I attached my printer (see above) to a Windows PC, I received an ink level warning. I then found out that Epson doesn't allow B/W printing as soon as one of the colour cartridges is empty, no matter how full the black cartridge is.
Could it be that the pdftoraster error is actually caused by the printer's refusal to activate itself?
| Ernst (ernst-blaauw) wrote : | #6 |
Here that's not the case, as some other pages print fine (like the page '
www.nu.nl': it is printed fine using Chromium)! Also, the same e-mail
message I attached above is printed fine using Firefox, so the hardware is
not the problem I think.
Btw, I have a Canon iP4500 and all cartridges have enough inkt left.
On Wed, Mar 17, 2010 at 19:51, Anton Kraus <email address hidden> wrote:
> When I attached my printer (see above) to a Windows PC, I received an
> ink level warning. I then found out that Epson doesn't allow B/W
> printing as soon as one of the colour cartridges is empty, no matter how
> full the black cartridge is.
>
> Could it be that the pdftoraster error is actually caused by the
> printer's refusal to activate itself?
>
> --
> /usr/lib/
> https:/
> You received this bug notification because you are a direct subscriber
> of the bug.
>
| Till Kamppeter (till-kamppeter) wrote : | #7 |
Anton Kraus, it is for sure not a hardware problem, as if I run the file which Ernst has attached through the Ghostscript command line in http://
|
|
#80 |
ghostscript-
If you want to test the update, you can install it with
su -c 'yum --enablerepo=
|
|
#81 |
ghostscript-
If you want to test the update, you can install it with
su -c 'yum --enablerepo=
| Anton Kraus (done) wrote : | #8 |
I see. So the ink levels are an additional problem with my printer, but not the cause of the pdftoraster error discussed here.
| James Walmsley (james.walmsley) wrote : | #9 |
I have the same problem, with an Epson RX425. I don't think its a printer problem, something is failing converting the PDF to raster.
Can't print my dissertation! I think this must be a regression, because it seemed to work a few days ago, before I did an update.
James
| sboetsch (sebastien-boetsch) wrote : | #10 |
I don't want to change the bug status but since the 10.04 update of tonight (I guess it is related to package hplip-cups version 3.10.2-1ubuntu2). I can now print with my Epson Stylus DX4250 (before I was ony able to print the test page).
The error has gone.
Thanks guys!
|
|
#82 |
ghostscript-
| Till Kamppeter (till-kamppeter) wrote : | #11 |
HPLIP is only for HP printers, what has solved your problem was probably the recent update of the Ghostscript package.
| Daniel J Blueman (danielblueman) wrote : | #12 |
These PDFs may also be relevant test-cases - with the Epson R220 CUPS filter, I see ghostscript hit with SEGV:
http://
ftp://ftp.
http://
http://
http://
| Ernst (ernst-blaauw) wrote : | #13 |
Yesterday, I had to print again from Chrome. Now, I did not get a error
message beforehand, but only after the page was printed - which was, instead
of an e-mail, a totally black page. The error message was not clear to me,
it just seem to say 'finished printing page 1'.
$ apt-cache policy ghostscript
ghostscript:
Installed: 8.71.dfsg.
Candidate: 8.71.dfsg.
Version table:
*** 8.71.dfsg.
500 http://
100 /var/lib/
On Mon, Mar 29, 2010 at 13:32, Daniel J Blueman <email address hidden>wrote:
> These PDFs may also be relevant test-cases - with the Epson R220 CUPS
> filter, I see ghostscript hit with SEGV:
>
> http://
> ftp://ftp.
> http://
> http://
>
> http://
>
> --
> /usr/lib/
> https:/
> You received this bug notification because you are a direct subscriber
> of the bug.
>
| Ernst (ernst-blaauw) wrote : | #14 |
This pdf: http://
total black page.
On Wed, Mar 31, 2010 at 10:22, Ernst Blaauw <email address hidden> wrote:
> Yesterday, I had to print again from Chrome. Now, I did not get a error
> message beforehand, but only after the page was printed - which was, instead
> of an e-mail, a totally black page. The error message was not clear to me,
> it just seem to say 'finished printing page 1'.
>
> $ apt-cache policy ghostscript
> ghostscript:
> Installed: 8.71.dfsg.
> Candidate: 8.71.dfsg.
> Version table:
> *** 8.71.dfsg.
> 500 http://
> 100 /var/lib/
>
>
> On Mon, Mar 29, 2010 at 13:32, Daniel J Blueman <email address hidden>wrote:
>
>> These PDFs may also be relevant test-cases - with the Epson R220 CUPS
>> filter, I see ghostscript hit with SEGV:
>>
>> http://
>> ftp://ftp.
>> http://
>> http://
>>
>> http://
>>
>> --
>> /usr/lib/
>> https:/
>> You received this bug notification because you are a direct subscriber
>> of the bug.
>>
>
>
| Till Kamppeter (till-kamppeter) wrote : | #15 |
Ernst, I have tried it out by printing it into a PDF file with evince first (to get an Evince-rerendered version) and try to render that one directly with Ghostscript. Ghostscript errors out before sending the first pixel of the CUPS Raster output to the driver. So probably the driver interprets the missing input data as black pixels instead of erroring out by itself.
Can you please provide an error_log as described in the "CUPS error_log" section of https:/
If you need your files printed, try to avoid evince. Use Adobe Reader, Okular, XPDF, gv, direct printing with "lpr", ... for printing PDF files.
| Ernst (ernst-blaauw) wrote : | #16 |
I created the log with the file I provided above (org_chromium_
As stated above, after printing the file (or actually: during the printing),
I get a GUI saying that there was a printing error and that I can diagnose
it (Still, I get a totally black page :-O). Some versions of gs ago, this
popup came before the printing, that saved a lot of ink :-).
*
*
On Thu, Apr 1, 2010 at 10:13, Till Kamppeter <email address hidden>wrote:
> Ernst, I have tried it out by printing it into a PDF file with evince
> first (to get an Evince-rerendered version) and try to render that one
> directly with Ghostscript. Ghostscript errors out before sending the
> first pixel of the CUPS Raster output to the driver. So probably the
> driver interprets the missing input data as black pixels instead of
> erroring out by itself.
>
> Can you please provide an error_log as described in the "CUPS error_log"
> section of https:/
>
> If you need your files printed, try to avoid evince. Use Adobe Reader,
> Okular, XPDF, gv, direct printing with "lpr", ... for printing PDF
> files.
>
> --
> /usr/lib/
> https:/
> You received this bug notification because you are a direct subscriber
> of the bug.
>
| Till Kamppeter (till-kamppeter) wrote : | #17 |
According to the error_log the printing process behaves as I thought. At the beginning of the page Ghostscript crashes and there is no output to the next filter any more. Even without getting input, the rastertogutenprint filter and the backend continue working and they seem to fill the missing input with zeros which means black in your case.
According to the Ghostscript command line in your error_log you are using "Grayscale" as "Color Model" setting. Can you try other settings here and see whether they work?
| Changed in ghostscript (Ubuntu): | |
| status: | Confirmed → Incomplete |
| Ernst (ernst-blaauw) wrote : | #18 |
Also with color model "RGB", I get a black page. However, I don't get an error. Here is the error log. I removed all the lines already present in my previous log.
| Changed in ghostscript (Ubuntu): | |
| status: | Incomplete → Confirmed |
| Ernst (ernst-blaauw) wrote : | #19 |
I've set the status to 'confirmed' because I still get the black page.
| Till Kamppeter (till-kamppeter) wrote : | #20 |
It has the same error (you find the error near the beginning of the error_log). Ghostscript crashes before sending the first pixels of the page bitmap. Driver and backend continue running again with the driver not receiving input data.
Can you try the following now: Choose a color model where zero means white, CMYK, CMY, or KCMY. Does it also give black output pages?
| Ernst (ernst-blaauw) wrote : | #21 |
I just tried KCMY - I get a totally white page. That's a cheaper error :-).
On Fri, Apr 2, 2010 at 18:13, Till Kamppeter <email address hidden>wrote:
> It has the same error (you find the error near the beginning of the
> error_log). Ghostscript crashes before sending the first pixels of the
> page bitmap. Driver and backend continue running again with the driver
> not receiving input data.
>
> Can you try the following now: Choose a color model where zero means
> white, CMYK, CMY, or KCMY. Does it also give black output pages?
>
> --
> /usr/lib/
> https:/
> You received this bug notification because you are a direct subscriber
> of the bug.
>
|
|
#83 |
ghostscript-
|
|
#84 |
ghostscript-
| nicoduck (nico-vthadden) wrote : | #22 |
i have the same problem, but only when printing more then 3 pages:
i printed a pdf, set it to print 2 pages on one sheet of paper and the printer stops after the third sheed of paper (6th page in pdf)
then selecting page 7-12 works without problems (printing 7-14 stops after page 12) - always printing 2 pages on one sheet of paper
syslog always says
Apr 27 13:09:08 xaser kernel: [58984.827174] gs[30796]: segfault at 7fffe5fb11c0 ip 00007fdcac2edd55 sp 00007fffebe10168 error 4 in libgs.so.
only the positions are different (where the segfault is)
I have a canon pixma ip4200 printer using CUPS+Gutenprint v5.2.5
| Changed in ghostscript (Ubuntu): | |
| importance: | Undecided → High |
| milestone: | none → lucid-updates |
| Till Kamppeter (till-kamppeter) wrote : | #23 |
SRU for Lucid prepared and uploaded to the -proposed queue, waiting for approval.
The fix makes all files mentioned in this bug report rendering correctly and probably also fixes many other reported Ghostscript crash problems (please mark other bugs as duplicate of this one if the proposed SRU fixes them).
| Changed in ghostscript (Ubuntu Lucid): | |
| status: | Confirmed → Fix Committed |
| Changed in gs-gpl: | |
| status: | New → Fix Released |
Accepted ghostscript into lucid-proposed, the package will build now and be available in a few hours. Please test and give feedback here. See https:/
| tags: | added: verification-needed |
| Leo Milano (lmilano) wrote : | #25 |
Hi all
I just tested the hpcups driver from the proposed repository in my system (Kubuntu Lucid, HP Deskjet F380), and the bug is fixed. I no longer get the black background.
Great work everyone involved! Cheers,
-- Leo
| tags: |
added: verification-done removed: verification-needed |
| Till Kamppeter (till-kamppeter) wrote : | #26 |
Leo Milano's original bug report is bug 553573.
| jmrbachiller (jmrbachiller) wrote : | #27 |
I'm having problem with my Epson SX210.
I've upgraded to Lucid recently (reinstalling all the so) With karmic I was able to print without problem (well, configuring my printer as SX100 and having very poor printing speeds).
When printing from OpenOffice.org Writer my printer ejects one paper and nothing is printed. No error messages.
I exported the document to PDF and try to print from Evince. A message box is shown, the diagnose utility notifies the status messasge "/usr/lib/
I've tried downgrading to ghostscript from karmic, with no results.
I have printed a GIF file (I probe it two times, and no problems), but when trying with other image (JPEG or GIF), the printing fails again.
I've attached the PDF document properties.
| Till Kamppeter (till-kamppeter) wrote : | #28 |
jmrbachiller, can you try out the fix proposed in comment #24 in this bug report?
| jmrbachiller (jmrbachiller) wrote : | #29 |
I've installed
ghostscript_
libgs8_
ghostscript-
ghostscript-
from lucid-proposed as indicated. No changes at all :(
Thanks for all the work!
PD: the error message when printing from OO-Writer is "/usr/lib/
| jmrbachiller (jmrbachiller) wrote : | #30 |
God news!
I've discovered that aparently the problems with my ODT document and the PDF (generated from it) were generatad by OTF the fonts included. This looks like a bug in OO.org + new OTF support.
When I uses the two OTF fonts installed the printing fails as I described above. When I replace the font everything works fine.
I've tried to print several images (JPG, PNG) withoout problems.
So, sorry for my incomplete tests of yesterday (I was tired of testing), because I didn't test the image files, and only try with the ODT and PDF with the OTF fonts.
The new revisions from lucid-proposed seems to solve my problems!
Than you!
PD: I can give the logs from gs, when trying
# gs -sDEVICE=cups -dNOPAUSE -dBATCH -sOutputFile=out bad_otf_file.pdf
or similar.
| jmrbachiller (jmrbachiller) wrote : | #31 |
And sorry again for my devilish english...
(God news! ...Yes, directly from heaven!)
| Till Kamppeter (till-kamppeter) wrote : | #32 |
jmrbachiller, so please report a new bug on the OTF fonts. Thank you.
| Djoko (djk-srs) wrote : | #33 |
Thanks for all the effort that has been done :)
I have upgrade with proposed repo's, but my Epson T20E still can't produce any output.
No error message and no error_log recorded.
regards
Djoko Srs
| Till Kamppeter (till-kamppeter) wrote : | #34 |
Djoko, your problem is most probably not the problem described in this bug report. Please post a new bug report. Thanks.
| Lars Volker (lv) wrote : | #35 |
jmrbachiller, could you please give some description on which packages to install from lucid-proposed, to get the OTF font inclusion problem fixed?
| jmrbachiller (jmrbachiller) wrote : | #36 |
Lars,
the packages installed from lucid-proposed repo were:
ghostscript_
libgs8_
ghostscript-
ghostscript-
This upgrade have solved some of my problems. I can now print as in karmic (using SX100 indeed SX210, and obtaining very slow speeds with text and images, but I can print! I think it twice before buy next printer, Epson surely loses one client)
But the upgrade didn't repair the problems with OTF fonts in OpenOffice documents and the PDF generated from them.
Jose Manuel.
| Till Kamppeter (till-kamppeter) wrote : | #37 |
jmrbachiller, thank you for your testing, please report a new bug for the OTF font problem, this is independent.
| Lars Volker (lv) wrote : | #38 |
The OTF problem is already known upstream and will be fixed in OpenOffice 3.2.1. I also filed a bug to watch the process. See #577043.
| Till Kamppeter (till-kamppeter) wrote : | #39 |
Link to OTF font bug: bug 577043
| Martin Pitt (pitti) wrote : | #40 |
Copied to maverick and lucid-updates.
| Changed in ghostscript (Ubuntu): | |
| status: | Fix Committed → Fix Released |
| Changed in ghostscript (Ubuntu Lucid): | |
| status: | Fix Committed → Fix Released |
| echo-e (jamesrudee) wrote : | #41 |
I am up to date with the ghostscript packages listed above, but I am encountering some variant of this bug. i.e. I get the
/usr/lib/
| Till Kamppeter (till-kamppeter) wrote : | #42 |
echo-e, your problem is most probably not the one described here. Please post a new bug report and follow the instructions under "CUPS error_log" on https:/
| James Clemence (jvc26) wrote : | #43 |
/usr/lib/
J
| Till Kamppeter (till-kamppeter) wrote : | #44 |
James Clemence's problem is handled in bug #628030.
| Petr Lacek (petr-lacek) wrote : | #45 |
I had the same problem with my printer Canon MP520, when I was printing a bank statement - in PDF, using Document Viewer 2.30.3 in Gnome, Ubuntu Lucid.
I have solved the problem by saving the document to PostScript format, then opening the PostScript file (in the Viewer too) and printing it. All was OK. It may not be a solution for all such problems but you can try it.
Good luck! Petr
| Hunter Laux (jolting) wrote : | #46 |
pdftoraster is used only when using the cups raster API. This only effects printers that don't understand postscript. Not all printers use this. Cups uses poppler to convert from pdf to ps. The problem is pdftoraster uses ghostscript. Ghostscript has problems with TONS of pdf files. That is why cups switched and is now using poppler. The pdftoraster is part of ghostscript.
The solution is simple. Convert from pdf to ps with poppler and then convert from ps to raster with ghostscript.
| tszanon (tszanon) wrote : | #47 |
I've got the same error message (pdftoraster failed), but with a different printer (Epson Stylus TX420W). It happens no matter what I try to print. Even the test page fails. Should I report a new bug? How can I help?
| Till Kamppeter (till-kamppeter) wrote : | #48 |
tszanon, please report a new bug and attach the error_log of a failed job as described in "CUPS error_log" on https:/
| Changed in gs-gpl: | |
| importance: | Unknown → High |
| Hacky (dahacky) wrote : | #49 |
I'm always getting this error when i try to print the attached file.
| Hacky (dahacky) wrote : | #50 |
I'm always getting this error when i try to print the attached file.
| David Tombs (dgtombs) wrote : | #51 |
Hacky, this particular issue has been fixed. Please report a new bug with 'ubuntu-bug cups' and mention your printer model.
| Changed in ghostscript (Fedora): | |
| importance: | Unknown → High |
| status: | Unknown → Fix Released |
Description of problem:
After updating to ghostscript-
Version-Release number of selected component (if applicable):
ghostscript-
cups-1.
turboprint-
How reproducible:
Always
Steps to Reproduce:
1. Update to ghostscript-
2. Print something (e.g. a PDF file or a test page).
Actual results:
The printed page is mostly black with only some vertical stripes and patterns on it.
Expected results:
The page is printed correctly.
Additional info:
I am using turboprint as driver for my Canon Pixma IP 4000. I don't know if this is a bug in ghostscript or an incompatibility with turboprint. Anyhow I could reproduce this behaviour on two Fedora 12 x86_64 machines.