On Thu, Apr 28, 2005 at 07:19:31PM +0200, Ivan Popov wrote:
> On Thu, Apr 28, 2005 at 10:05:39AM -0600, Patrick Walsh wrote:
> > >  - the running process's uid
> > >  - the Posix superuser uid (0 aka root)
> > >  - the usual "nobody" uid (may vary on different systems, but still has
> > >    a common semantics)
> > 
> > 	Here's a thought: could the kernel return ownership information
> > differently to different users?  Probably.  Could it then tell the
> 
> Don't know. Jan can tell better, it can be tricky.
> So the first option is in fact uncertain. Possibly not the best either.

The only way to efficiently do it is by hacking the kernel to always put
curent->fs_uid or something in the kstat that is returned from getattr.
Otherwise it would be an futile fight trying to prevent the kernel from
caching the attributes, along with the performance loss of having to
send an upcall to venus on every stat(2) operation.

But modifying all the kernel modules to do this override internally just
seems like a lot of work for something that I don't really think is all
that important.

> It would be a totally another and good thing to have something like
>  cfs ls-l
> which could possibly show a lot of metainformation,

I think AFS was actually using a modified ls which would look up the
usernames assocated with the AFS id's in their equivalent of the pdb
database (pts?) and would show only the parts of the modebits that
mattered.

Jan