[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: pam_env.so: Supposed to work with service xdm?

On Mon, 8 Sep 1997, Dave Kinchlea wrote:

> No it does not, a pain that I hope *sometime* to fix. I haven't looked

Too bad...

> into things in any depth but my suspision is that xdm is not fully pam
> aware (can anyone confirm?). I need to do some research, it is not

Jup. true.

> completely clear to me where xterms are supposed to get their environment

Well from the process that starts them I guess... So in the end from xdm.

> from (for instance, the difference between -ls and +ls, I assume xdm for
> the +ls case and that xterm should be `pamified' to handle the -ls case
> but ....)

I don't think we need to hack xterm, as child processes usually get a copy
of their parents' environment as a start. The -ls option of xterm just prepends
a '-' to the shell invocation to tell the shell this is a login session. This
mostly affects what files are sourced by the shell (.profile vs .bashrc)

> If anyone can shed some more light on xdm and xterm, that would be
> appreciated.

I noticed in the debug.log tha for instance the value that REMOTEHOST gets
assigned is just trash (i.e. random bytes) as if PAM_RHOST contains nothing
or junk. I don't have the log here now (at work) but I will supply it later.

> Note too, I still have intentions to allow pam_env to *set* PAM_
> variables, in case anyone thought I had forgotten ;-) I am going to find
> the time to do this Real Soon Now (tm).

	Ciao Chris
 Home: chris@ds9.au.s.shuttle.de /\ Work: drdisk@tilx01.ti.fht-esslingen.de
  FIDO: 2:246/1416.39 \/ Im Wiesengrund 15, D-71522 Backnang, 07191-980854
     IRC: DrDisk \/ PGP Fingerprint 79AA6C63 84523279 3CA9B679 BDB74B17
|"The only reason Bill Gates is alive today is because the unabomber never|
| used Windows 95" -- Anonymous                                           |

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index] []