Red Hat Bugzilla – Bug 6356
Kudzu seg fault
Last modified: 2014-03-16 22:10:50 EDT
Kudzu segfaults on my system if I run it after loading the lp driver. An strace shows that it dies just after reading /proc/parport/0/autoprobe. My printer (an HJ-400) is too old for probing, and so this file is empty.
/proc/parport/0/autoprobe is empty?? Do you have parport_probe loaded? (This is mainly for informational purposes; it still is a bug. :) ) BTW, bug is fixed in kudzu-0.23-1, which will be in next rawhide.
I'm _almost_ certain I do. (I'm not at home now). Also, the /usr/share/kudzu/printers (I can't remeber the name) file lists the HJ-400 with the epson driver, along with a comment that it only works in 180x360 mode. I don't know what this file does, given that the HJ-400 can't be detected by kudzu, but the epson driver really only works well in 180x180 mode for the HJ-400 The lq850 driver (not shipped with redhat, I'm about to file a bug against ghostscript on that) supports 360x360. (The HJ-400 emulates an epson lq by default. It claims to emulate an IBM PS/2 printer by changing a few settings on the printer, but I've never got that to work, even under windows)
The stuff in printertable currently isn't used by much of anything. It's generated from http://www.picante.com/~gtaylor/pht/printer_list.cgi if you'd like to update the master.
I've now looked, and I _don't_ have parport_probe loaded (or even compiled). I did at one stage, but my printer doesn't report anything, so I took it out of the kernel compile, since I couldn't see any point in it. I can't remember if the file was empty, or if there was just no data (apart from the headers) in it. I've updated the printer database. Thanks for the pointer. What about a note in /usr/doc/kudzu, or at the top of the printertable file?
Will look at adding the note in the next release of kudzu.