Re: GDM interprocess communication problems

[George <jirka 5z com>]

On Mon, Jan 19, 2004 at 03:31:56PM +0000, Anton Altaparmakov wrote:
> Note I am happy to invest programming time in this to fix it, I would
> just like to know in which way you would like it fixed so that it can be
> merged into the GDM distribution and everyone can benefit rather than us
> fixing it only for ourselves...

Actually thinking about it more, I suppose that hoping that PAM will be sane
and well defined and nice and all that is maybe too much to expect :)

I have another solution which really solves the problem, but involves quite a
bit of programming (which of course I don't have the time for right now:( )
The idea is that we would fork the session process early on and do all PAM
work from there, instead of from the slave.  But then all communication of
PAM with the greeter must further be proxied through the slave (or a direct
communcation must be opened somehow from the session to the greeter which
doesn't seem as easy).  Further any 'slave related' stuff done from the pam
code must also be proxied.

It's not a huge amount of work, but it's slightly tedious to figure out what
all must be proxied, it adds yet another channel of communication (two way),
though that could be simple and synchroneous so no need for signals and
select calls and such nonsense.  It completely solves the PAM issues since
PAM will have a completely clean environment to play in as that will only
become the session process anyway.  Further it will fix issues where pam
actually thinks this is the case.

Downside is that you'd have to take a plunge into the slave spagetti code :)
Upside is that a lot of similar issues would be solved in one big swoop.

George

-- 
George <jirka 5z com>
   Zivot je kratkej a posranej, jako zebricek do kurniku.
                       -- Neznamy