Re: GACI item list - to give some items for discussion

Previous message: Greg KH: "Re: Benchmarks (was Re: Hooking into Linux using the LTT)"
In reply to: Crispin Cowan: "Re: GACI item list - to give some items for discussion"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

neilat_private

Crispin Cowan wrote:
> 
> Casey Schaufler wrote:
> > A useful function, first proposed in literature by
> > W. Olin Sibert, would be one which you pass a bunch
> > of security attributes for the subject and a set for
> > the object along with a proposed access and you get
> > back a best guess answer. It could be implemented
> > strictly in userland for many policies. On the other
> > hand, I've never seens a reasonable specification
> > for the call.
> 
> I don't see a problem with LSM modules providing new system calls (or ioctls,
> or sysctls, or whatever) to provide this kind of functionality.  It's yet
> another module feature, not a feature of LSM per se.
> 

The access() function provides a uniform way for programmers to find out
if a *file system* function would work or not. A function would need to
be implemented as part of the LSM to provide this same feature for other
kinds of access checks such as "Can I bind to this reserved port".
Modules would hook this new function leaving the function itself
security policy agnostic. I would propose re-implementing the access()
call to be a wrapper for the new function so that module writers only
need to hook into one place. Not implementing a generic call leaves us
with a lot of programs that work under one module and not another or
with a number of #ifdefs.

Neil