why fd passing?

[Phil Howard]

On Thu, Jun 27, 2002 at 12:02:32PM +0200, Markus Friedl wrote:

| On Wed, Jun 26, 2002 at 11:46:24AM -0500, Phil Howard wrote:
| > Why not go ahead and have the monitor set one up before it forks
| > the child?
| 
| with protocol 2 multiple pty and multiple login
| shells over one connection are allowed.
| the ssh.com windows clients supports this for
| example.

If a given system can allocate a pty w/o root, would it be possible
for that system to just let the user privilege process do it?  If
so, a system which can't do fd passing but can do pty w/o root could
get around that problem that way.

What about a configurable quota on the maximum number of ptys a
process can get set up for it by the monitor, so that can't be used
as a denial of service exploit from inside a cracked user privilege
process?  Or would that not be worth the trouble?

I like privsep, but I can certainly see a lot of complications in it.


| > I do not see from this illustration how mmap() is involved.
| 
| it's used for passing the internal zlib compression
| state around.

I was assuming something like that.  It just wasn't shown.  I hope
when things settle down that detailed technical documentation can
be prepared, or the current little blub can be expanded.  Not everyone
can figure this out by reading code.

-- 
-----------------------------------------------------------------
| Phil Howard - KA9WGN |   Dallas   | http://linuxhomepage.com/ |
| phil-nospam at ipal.net | Texas, USA | http://phil.ipal.org/     |
-----------------------------------------------------------------