ps2ascii crashs at start

Please can you explain how the crash occured. Can you reproduce it?

Try to obtain a backtrace by following the instructions on https://wiki.ubuntu.com/DebuggingProgramCrash.
(install packages 'ghostscript-dbgsym' and 'gdb' and see here for backtrace https://wiki.ubuntu.com/Backtrace)

Set to incomplete until we have a backtrace or more info.

ps2ascii is a shell-script. It calls ghostscript (gs) as an interpreter. The interpreted ps-file (which does the work) is called /usr/share/ghostscript/8.63/lib/ps2ascii.ps. The Interpreter itself (that is ghostscript) doesn't crash with a segmentation fault or so. So a backtrace won't bring anything to solve the problem.

The problem is probably with the ps2ascii.ps-file. Or the new ghostscript missinterpretes this file.

On my system (amd64) the problem occurs in every situation. ps2ascii doesn't work in any circumstances.

Just for your information: ps2epsi works identically to ps2ascii without any problems!
(the shellscript ps2epsi is a bit longer and complexer than ps2ascii, but it calls in the same manner ghostscript as interpreter with a ps2epsi.ps-file)

Hope, this helps you enough.

Since ps2ascii.ps-file didn't change to my working version in Hardy, the problem must probably be with the new interpreter.

Can you edit /usr/bin/ps2ascii removing the "-dDELAYBIND" from the "OPTIONS=..." line? Then try to use ps2ascii again and check whether you get correct results.

I tried this. The output is correct so far. But ghostscript can't quit (the last command "-c quit" in /usr/bin/ps2ascii) without a Segmenation fault (core dumped).

Because ghostscript-dbg or -dbgsym isn't in the repo, I don't think the core dump will help further.

But it is attached for your information.

I have installed the newest version:

ghostscript 8.63.dfsg.1-0ubuntu5

I tested it with -dNOBIND switch and it works so far for me! But I think this is still a bug, because ps2ascii uses this -dDELAYBIND switch as it is written in the man-pages.

Shall I do further tests? If desired, do you have a test-file for me?

Thanks

Thank you very much.

I have reported it upstream

http://bugs.ghostscript.com/show_bug.cgi?id=690124

but I will apply your solution anyway, as we have release freeze tomorrow.

Unfortunately, I cannot apply the "-dNOBIND" solution. It does not work for me.

This bug got fixed upstream and I have verified that "ps2ascii" really works now.

Thanks a lot!

Christian Prim, for getting your problem fixed immediately, apply the patch attached to the upstream bug

http://bugs.ghostscript.com/attachment.cgi?id=4519&action=view

to the file

/usr/share/ghostscript/8.63/lib/gs_init.ps

on your system.

This bug was fixed in the package ghostscript - 8.63.dfsg.1-0ubuntu6

---------------
ghostscript (8.63.dfsg.1-0ubuntu6) intrepid; urgency=low

  * debian/patches/42_gs-init-ps-delaybind-fix.dpatch: Make "ps2ascii" working
    again (LP: #281419, upstream bug #690124).

  * debian/patches/45_bjc600-bjc800-pdf-input.dpatch: Fix setting of the
    "DitheringType" option. With PDF input Ghostscript crashes, with
    PostScript input the "DitheringType" option was probably ignored
    (Upstream bug #690032).

  * debian/patches/48_cups-output-device-pdf-duplex-uninitialized-memory.patch.dpatch:
    Fixed several bugs in the "cups" (CUPS Raster) output device:
     - Ghostscript crashed with PDF input data
     - The "Duplex" and "MediaWeight" options were ignored
     - There was uninitialized memory and wrong usage of pointers, potential
       cause for segmentation faults or even vulnerabilities
     - There were mismatches in data types, leading to possible breakage
       of the "AdvanceDistance", "MediaWeight", and "cupsStringXX" options.
    (Upstream bug #690101).

 -- Till Kamppeter <email address hidden> Sun, 19 Oct 2008 12:48:22 +0200