FS#18339 - [ghostscript] segfaults when printing test page

Attached to Project: Arch Linux
Opened by Eric Bollengier (ebl) - Monday, 15 February 2010, 12:00 GMT
Last edited by Andreas Radke (AndyRTR) - Sunday, 21 March 2010, 21:20 GMT
Task Type Bug Report
Category Packages: Extra
Status Closed
Assigned To Andreas Radke (AndyRTR)
Architecture x86_64
Severity Medium
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 8
Private No

Details

Description:

Since i'm using Ghoscript 8.71-1, i'm unable to print documents, i'm getting black page. Re-installing the
previous version (8.70-4) fix the problem. In dmesg, i can see the following message

gs[5826]: segfault at 7fafdce34000 ip 0000000000728932 sp 00007fff9b520d90 error 6 in gs[400000+549000]

Additional info:
* package version(s)
ghostscript 8.71-1

* printer info
Epson Stylus Color 860
Connection: usb://EPSON/Stylus%20COLOR%20680?serial=L56010012080716040
Defaults: job-sheets=none, none media=iso_a4_210x297mm

* driver info
CUPS+Gutenprint v5.2.5 (color)

Steps to reproduce:
Print a document or a test page in cups
This task depends upon

Closed by  Andreas Radke (AndyRTR)
Sunday, 21 March 2010, 21:20 GMT
Reason for closing:  Fixed
Additional comments about closing:  patch applied
Comment by Matthew Gyurgyik (pyther) - Wednesday, 17 February 2010, 22:40 GMT
This seems as if it is an upstream bug. You might want to check upstream for any similar issues.
Comment by Jaroslaw Siebert (y0g1) - Friday, 19 February 2010, 07:55 GMT
Hello,
here is the same:
gs[2669]: segfault at c4c402c ip 0836a62c sp bf91ebf0 error 4 in gs[8048000+487000]
gs[2678]: segfault at c4c402c ip 0836a62c sp bf87ad40 error 4 in gs[8048000+487000]
gs[2708]: segfault at c4c402c ip 0836a62c sp bfdec1b0 error 4 in gs[8048000+487000]
gs[2819]: segfault at c4c402c ip 0836a62c sp bfcc2840 error 4 in gs[8048000+487000]
ghostscript 8.71-1

I can print with ghostscript-8.70-4-i686.pkg.tar.gz


printer: Canon Pixma IP4200
driver: CUPS+Gutenprint v5.2.5 (color)

Comment by Matthew (piezoelectric) - Tuesday, 23 February 2010, 01:23 GMT
I don't see any bug reports upstream that discuss full black pages. Could there be something wrong with the PKGBUILD? I see not much changed in the build() section.
Comment by Andrzej Giniewicz (Giniu) - Friday, 26 February 2010, 17:54 GMT
Also got black pages with update to 8.71 - with Epson SX600FW - does anyone already have account there and could report? I don't like bugzilla at all, last time I used it, it took few days to send confirmation email and then it didn't work and few times like that, it's also very hard to find stuff in there, so I rather don't touch it.
Comment by Eugene (Infy) - Tuesday, 02 March 2010, 12:40 GMT
Same with Epson CX3900.
Comment by Jean-Damien Nappey (jdn06) - Tuesday, 02 March 2010, 17:42 GMT
Same with Epson Stylus D92
Comment by Dieter Rethmeyer (Dieter) - Tuesday, 02 March 2010, 20:56 GMT
... and Canon Pixma IP43300 with commercial turboprint.
Comment by Smith Dhumbumroong (zodmaner) - Wednesday, 03 March 2010, 16:18 GMT
Same here. Everything kind of job, including printing a test page, print out a black page.

Printer: Epson Stylus C90
Driver: Epson Stylus C68 - CUPS+Gutenprint v5.2.5 (color)

Ghostscript version 8.71-1.
Comment by Scott (firecat53) - Friday, 05 March 2010, 02:57 GMT
Same here with Canon Pixma mx-310 using gutenprint and mp-180 driver.
Comment by Andreas Radke (AndyRTR) - Monday, 08 March 2010, 22:49 GMT
please check the new ghostscript/foomatic packages in testing. if it's still broken please look here for a
possible fix: http://cvs.fedoraproject.org/viewvc/rpms/ghostscript/devel/

or ask upstream (mailing list).
Comment by Smith Dhumbumroong (zodmaner) - Tuesday, 09 March 2010, 11:06 GMT
Upgrade all related packages (ghostscript, foomatic-*) to latest version and switch to Epson Stylus C68 Foomatic/gutenprint-ijs.5.2 (color) driver, I'm able to print again.

Thank you for fixing this, Andy.
Comment by Smith Dhumbumroong (zodmaner) - Tuesday, 09 March 2010, 11:09 GMT
Just want to add that the problem still persists if I use the old driver (the Epson Stylus C68 - CUPS+Gutenprint v5.2.5 (color)).
Comment by Matthew (piezoelectric) - Wednesday, 10 March 2010, 17:54 GMT
This is not fixed, I just upgraded, printed a black page, downgraded, printed a correct page. I have an Epson Stylus CX6600.

As far as I can see, the only thing that was changed was some pdf2dsc patch. I don't see what that would have to do with a black page.
Comment by Andrzej Giniewicz (Giniu) - Wednesday, 10 March 2010, 18:04 GMT
well, installed foomatic stuff and selected ghostscript-ijs driver and it works for me too, no blank pages anymore
Comment by Matthew (piezoelectric) - Wednesday, 10 March 2010, 23:45 GMT
How does switching to foomatic _fix_ the problem? That's just a workaround is it not?
Comment by Matthew (piezoelectric) - Thursday, 11 March 2010, 00:23 GMT
Well, switching to foomatic resolves my printing issues. (Still don't see how it's not a workaround). But I'll quite complaining because the foomatic method seems to be superior: for the first time in years, reverse printing is working (which is absolutely fabulous).

Can I just confirm here that I need foomatic-db AND foomatic-db-engine? They don't depend on one another, but the driver for my printer only shows up when both are installed.
Comment by Andrzej Giniewicz (Giniu) - Thursday, 11 March 2010, 09:49 GMT
I agree that it's workaround, but personally I don't care as **it works** :)
Comment by Andreas Radke (AndyRTR) - Thursday, 11 March 2010, 18:31 GMT
Anyone with a printer that is still not working?
Comment by Eric Bollengier (ebl) - Friday, 12 March 2010, 12:44 GMT
I don't understand very well how to switch to foomatic, and with the current ghostscript package 8.71-2, it still segfaults.
Comment by Matthew (piezoelectric) - Friday, 12 March 2010, 22:50 GMT
Did you install foomatic-db and foomatic-db-engine? In the CUPS web interface, modify the printer, and when you get to the driver selection, scroll down until you find your printer. (Don't be confused by the fact that your current driver is listed at the top)
Comment by Eric Bollengier (ebl) - Saturday, 13 March 2010, 08:58 GMT
Ok thanks, i was missing a part of the installation, it doesn't segfault anymore. Now, as it's a workaround and the ghostscript package is still broken, i'm not sure that we have to close this issue...
Comment by Andreas Radke (AndyRTR) - Sunday, 14 March 2010, 21:46 GMT
It won't get fixed until you start to contact the upstream devs. I will close this one soon as ustream.
Comment by Andy McMillan (mcmillan) - Saturday, 20 March 2010, 19:25 GMT
Fixed in upstream svn.
http://bugs.ghostscript.com/show_bug.cgi?id=691203

Attached is a patch that fixed the current package for me.
   ghostscript-raster.patch (3.1 KiB)

Loading...