SUMMARY: Need printing help - fast!

From: <rbabcock_at_heyoka.cc.whecn.edu>
Date: Mon, 27 Nov 2000 13:57:42 -0700 (MST)

Thanks to all who responded, and especially to Ettore Aldrovandi
and Debra Alpert, whose replies are copied below. The solution
was to include an lp=_at_host/service line to the printcap entry.
The curious thing is that the original printcap entry worked just
fine on 4.0f.

From: Ettore Aldrovandi <ettore_at_sg1.hep.fsu.edu>


Hi,

are you trying to use tcp printing? Or just use remote printing?
If the latter is the case, you'll need the the :lp: entry as
well, while the ct flag is probably irrelevant. Also, filtering
should be ignored. If on the other hand you are using tcp
printing, then you should have :ct=tcp: alright, but that needs
the lp flag to be set like :lp=_at_aaa.bbb.ccc.ddd/port: where
"port" is defined in /etc/services (and depends on the printer,
for ex. 9100 for several HPs).

Is this a working printcap entry on 4.0F? If it is, I'm a little
puzzled, because it shouldn't work. Maybe that's the cause. 5.1's
lpd may be less forgiving. I have 5.0A and 5.1 and use tcp
printing more or less as explained above. Here is an entry from
my printcap:

# HP LASERJET 8100DN (B/W Postscript, 2 sided printing)
ps5_2|pspr5_2:\
        :ct=tcp:\
        :lp=_at_xxx.yyy.zzz.ttt/hp8100dn:\
        :mx#0:\
# :rp=raw:\
        :if=/usr/lbin/ppdof +Chplj8000.rpd -K two:\
        :of=/usr/lbin/ppdof +Chplj8000.rpd:\
        :lf=/var/adm/ps5_2err:\
        :sd=/var/spool/ps5_2:

hp8100dn is defined in /etc/services as

        hp8100dn 9100/tcp

Hope this helps,

Ettore

--
Ettore Aldrovandi
ettore_at_sg1.hep.fsu.edu                  Ph:  (850) 644 3289
http://www.hep.fsu.edu                  Fax: (850) 644 6735
From: "Alpert, Debra" <DAlpert_at_lifelinesys.com>
To: "'rbabcock_at_heyoka.cc.whecn.edu'" <rbabcock_at_heyoka.cc.whecn.edu>
Subject: RE: Need printing help - fast
Date: Mon, 27 Nov 2000 13:46:56 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: text/plain
Ron,
We had this same problem with the lpr binary that shipped with TU5.1. By
using truss to trace the system calls made by lpr, I found that the program
parsed the printcap file, assuming that the "lp" field (default output
device) was explicitly set in /etc/printcap for each printer. I had never
set this, as we print to a network device. In any case, since the lpr code
doesn't actually check the return value when it looks for the lp printcap
entry, it proceeds happily along, although a NULL pointer has been returned.
This caused problems down the line, as you can see from the following truss
output:
    Received signal #11, SIGSEGV [default]
      siginfo: SIGSEGV SEGV_MAPERR addr=0x0000000000000000
                                                Err#139 Error 139 occurred.
        *** process killed ***
The OS wasn't happy when access to memory address zero was attempted. It
appears that the lpr code doesn't check the return value of the relevant
call. In any case, I worked around this problem by adding a null lp entry to
our printcap files (that is, ...stuff...:lp=:...stuff). Now the return value
in question is a pointer to the above NULL string. This keeps lpr happy, and
printing works once again.
Hope this helps,
  --Deb
Debra Alpert
Senior Unix Systems Administrator
Lifeline Systems
> -----Original Message-----
> From:	rbabcock_at_heyoka.cc.whecn.edu [SMTP:rbabcock_at_heyoka.cc.whecn.edu]
> Sent:	Monday, November 27, 2000 12:14 PM
> To:	tru64-unix-managers_at_ornl.gov
> Subject:	Need printing help - fast
> 
> Managers,
> 
> Over the weekend I switched from an Alpha AS1000A 5/40 to an ES20D.
> The old alpha was running 4.0f, the new one is running 5.1.  I copied
> the printcap file to the new system but am having a problem with the
> networked printers.  Below is a typical printcap entry:
> 
> 
> lp20|20|LP20|:\
>         :af=/usr/adm/lp20acct:\
>         :lf=/usr/adm/lp20err:\
>         :of=/usr/lbin/lpf -c:\
>         :ct=tcp:\
>         :pw#85:\
>         :rm=aaa.bbb.ccc.ddd:\
>         :rp=text:\
>         :sd=/usr/spool/lpd20:\
>         :sh:
> 
> When a file is sent to this printer I get a segmentation fault and a
> core dump.  E. g., lpr -P20 /etc/motd results in a core dump.  All of
> my LAT printers are working ok.  Any ideas?
> 
> Ron Babcock, Dir. Admin. Computing
> Casper College
> Casper, WY 82601

Received on Mon Nov 27 2000 - 21:00:27 NZDT