Bug#932246: Re : Bug#932246: Re : Bug#932246: printer-driver-hpcups: no more printing to a HP LaserJet 1320: stack smashing detected and hpcups crashed on signal 6
On Sat 21 Sep 2019 at 10:53:04 +0200, Francesco Poli wrote:
> On Sat, 21 Sep 2019 09:33:25 +0200 nicolas.patrois@gmail.com wrote:
>
> > Le 21/09/2019 00:02:42, Brian Potkin a écrit :
> >
> > > No idea. I haven't even begun to think about it because I was awaiting
> > > confirmation or otherwise.
> >
> > I checked it again this morning after the morning upgrade.
> > Version 11 crashes printing.
> > Downgrading to version 10.3+deb10u1 makes it work back again.
>
> What's this? Magic?!? :-o
Must be! It has exerted its influence across the English Channel and
affected my machine! A little test, using either the technique I gave
earlier or by printing:
The first line of /usr/lib/os-release has "bullseye/sid" in it. There
are eight characters in "bullseye". Remove any one of them and print.
Also - leave the eight characters there and remove the forward slash.
Now, who is to deal with this? Upstream HPLIP, upstream systemd or the
base-files maintainer?
> I compared the two source packages of base-files (10.3+deb10u1 vs. 11),
> see the attached diff output.
> I really cannot see any relevant difference.
> Should we suppose that printer-driver-hpcups reads /etc/os-release
> (or /etc/issue* or /etc/debian_version) and decides whether it will
> crash, based on this?!?
I do not understand code but, for some reason, /etc/os-release is opened
in common/utils.c.
Regards,
Brian.
Reply to: