Re: Shell-Script problems

Ion Badulescu (ionut@moisil.wal.rhno.columbia.edu)
Tue, 10 Dec 1996 16:40:00 -0500 (EST)


On Tue, 10 Dec 1996, Michael Ballbach wrote:

> On 10 Dec 1996, Matthias Urlichs wrote:
>
> > In linux.dev.kernel, article <199612091745.MAA05230@ktk.bidmc.harvard.edu>,
> > Kris Karas <ktk@ktk.bidmc.harvard.edu> writes:
> > >
> > > Under 2.0.27, the script identifies itself, via the /proc interface,
> > > as "scum" - somewhat imprecise as the interpreter's identity is
> > > hidden. Under 2.1.14, BOTH the interpreter AND the interpreted
> > > program name are listed, for greater precision. Doing 'ps -ax' will
> >
> > ???
> >
> > Under 2.0.27, ps -ax shows "bash /some/path/scum" (or "(scum) if argv is
> > swapped out) and pstree shows "scum".
>
> I think we need to talk about this, because I like that scheme better. It
> allows you to more accurately determine what a user is doing. Wouldn't it
> be more plausible to change killall to ignore the interpreter? It's just a
> personal opinion, I just think it is more convinient to show all those
> things. :)

Okay, the message you respond to was confusing in the first place. 2.0.27
has the right behavior, i.e. killall works with the name of the script, ps
-c shows the name of the script, etc. 2.1.14 has the broken behavior which
puts the name of the interpreter instead of the name of script and forces
you to use killall interpreter which potentially kills a lot more
processes than you'd like to.

The patch that I posted is for 2.1.14 and makes it do what 2.0.x does,
i.e. the Right Thing (tm) :)

Ionut

--
  It is better to keep your mouth shut and be thought a fool,
            than to open it and remove all doubt.