Re: Problems with libgs9

[Jonathan Nieder <jrnieder@gmail.com>]

Hi,

Daniel Stender wrote:

> I am not so deep into this, is it a local problem or would be a bug
> report appropriate:
>
> trying to print anything always leads to:
> [ 8561.256399] usblp1: nonzero read bulk status received: -71
> [ 8561.606400] gs[4592]: segfault at 88 ip 00007f3efe12c806 sp 00007fff43433690 error 4 in
> libgs.so.9.01[7f3efdf23000+4fa000]
> or similiar (and nothing happens).
>
> My setup: very up-to-date Linux Mint Debian/Debian Testing, Kernel
> 2.6.32-5, Cups 1.4.6-3, libgs9 9.01~dfsg-2. Printer is Canon MF 4140
> with vendor's Cups driver (cndrvcups-common, cndrvcups-ufr2-uk
> 2.20-2, proper 64bit RPMs converted with Alien).
>
> A couple of days ago it printed fine so I would guess everything is
> installed properly.

Could you install the ghostscript, ghostscript-dbg, libgs9, etc
packages from http://incoming.debian.org/ and try to get a stack
trace?  That would be very interesting.

My hunches are:

 1) This is a regression and a segfault, so it's probably a bug
    (of insufficient validation of input from drivers, if nothing
    else);

 2) Canon's ufr2 driver would be very interesting for inclusion in
    the non-free repository that accompanies Debian, to help others
    who want this driver.  Naturally that would make working on bugs
    like this much easier.  AFAICT the license allows distribution.
    Do you know anyone interested in working on that?

    (For reference, the driver is at
    http://support-asia.canon-asia.com/contents/ASIA/EN/0100270807.html
    for example.)

 3) The UFR II (“ultra-fast rendering”) protocol is a winprinter
    protocol used mostly for multifunction printers in the LBP and MF
    series (imageRunner, etc).  Many printers supporting it also
    accept PCL or postscript.

 4) No one I know of has acquired specifications or even started to
    reverse-engineer it, to develop a free driver.  If you want to
    work on this (please?) a good place to start might be
    http://ww1.pragana.net/gdiprinters.html .

> Any pointers appreciated & greetings,

I'd suggest filing a bug once you get a backtrace and keeping us
posted about getting this printer to work "out of the box".

Thanks and hope that helps,
Jonathan