diff between "include" directive and use of pam_stack.so?

[Tomas Mraz]

On Sat, 2010-10-16 at 08:06 -0400, Robert P. J. Day wrote: 
> perhaps just my inexperience with PAM but i'm reading an online PAM
> tutorial here:
> 
>   http://content.hccfl.edu/pollock/AUnix2/PAM-Help.htm
> 
> and it mentions the use of the pam_stack.so module to support
> recursive(?) processing.  on my centos 5.5 system, "man pam_stack"
> appears to describe it much the same way.
> 
>   but how does this differ from simply "include"ing another PAM config
> file?  on this centos system, the entirety of the use of that module
> is:
> 
> $ grep pam_stack *
> pm-hibernate:#auth       required	pam_stack.so service=system-auth
> pm-powersave:#auth       required	pam_stack.so service=system-auth
> pm-suspend:#auth       required	pam_stack.so service=system-auth
> pm-suspend-hybrid:#auth       required	pam_stack.so service=system-auth
> $
> 
> and, as you can see, all of those uses are commented out.  so is
> pam_stack.so even used anymore?  thanks.

Thorsten already said what is the difference. In RHEL-5 the use of
pam_stack was replaced by include and the pam_stack module was
deprecated. It is no longer present in RHEL-6. Later in the upstream of
Linux-PAM a substack directive was implemented which emulates the
previous pam_stack module behavior although there are still some
differences.
-- 
Tomas Mraz
No matter how far down the wrong road you've gone, turn back.
                                              Turkish proverb