From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.5)
Description of problem:
In FC1 and FC2 we've been using pbmtext to generate images with Greek
text. Now with FC3, the same command segfaults.
This only happens if pbmtext is given Greek characters as input
(iso-8859-7 or UTF-8 character sets). It works fine with plain english.
$ pbmtext "ole" >text
$ pbmtext "ολε" >text
The first command has "ole" in iso-8859-1 characters and the second
command has them in iso-8859-7 characters.
i'm not sure if the Greek characters will be displayed properly in
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. run pbmtext with Greek as input text.
2. pbmtext segfaults.
Actual Results: Segmentation fault
Expected Results: output is an image file with the Greek characters.
This command always worked fine in FC1 and FC2. Also, the output image
won't contain proper Greek characters by default because we haven't
loaded a Greek font file in bdf format. Though thats irrelevant with
Hello Dimitris, maybe the problem was in the old libpng where I saw
the segfaults, does it work for you if you upgrade to libpng-1.2.8?
I'm unable to reproduce this bug any more after the upgrade.
I've contacted the author of the package and here is his reply:
"There's a bug in the latest release of Netpbm (10.25) that causes
whenever a character of the input text has an ASCII code greater than
He also said that version 10.18.17 does not have the bug, and that the
next release due out this month (maybe its already out?) will have the
I dont think the problem is with libpng. Maybe you updated your netpbm
package at some point?
Created attachment 109614 [details]
Fixed pbmtext.c without the bug for high ascii characters (file came from the author directly)
Dimitris, yes, I build the netpbm-10.26 recently and it resolves your
problem. Please update to it from:
The pbmtext.c source you attached is identical to the one in this release.
Thanks for the bugreport.