On Fri, 23 Feb 2001, Stephen C. Tweedie wrote:

> Hi,
> 
> On Fri, Feb 23, 2001 at 04:07:13PM +0100, Andreas Gruenbacher wrote:
> > On Fri, 23 Feb 2001, Stephen C. Tweedie wrote:
> > 
> > In fact, I'm not entirely happy with the acl_posix namespace, and after
> > thinking about it consider the previous classification (user, inode,
> > trusted, system) better. (This is also what Robert Watson suggested.)
> > 
> > The different system attributes (ACLs, CAPs, MAC labels, etc.) could then
> > be identified by attribute name, e.g., "acl.posix.access",
> > "acl.posix.default", "acl.afs", "cap", etc.
> 
> Fine by me: that's a different way of encoding the namespace but
> ultimately still gives us the same flexibility.

Yes. I'm very pleased with that response. I wonder what we've been arging
about all the time...

> > This is a little bit more clear semantically: All EA's in `system' have
> > the semantics of being passed in/out as one chunk, and the kernel needs to
> > check values, convert and dispatch, etc.
> 
> One last question on this: do we still want to bother with
> test-and-set operations on these attributes?  Given that we have cases
> where it's not actually possible (NFS), I'm not sure that it's worth
> the bother.

That is practically the only important reason for having a sequence of EA
operations at all. If we drop that, everything becomes _much_ simpler (in
fact, pretty similar to the current stable code). I could live with that,
especially since the Posix API doesn't support test-and-set anyway. I
doubt it's a good idea to design in such a fundemental weakness from the
start, though.

Robert, what's your position on this?

The fact it's not possible to support this in up to NFSv4 doesn't mean it
won't be possible in the next version. I don't know enough about other
network filesystems. For local filesystems, test-and-set should be cery
cheap to implement using the cookie approach.


--Andreas.

------------------------------------------------------------------------
 Andreas Gruenbacher, [EMAIL PROTECTED]
 Contact information: http://www.bestbits.at/~ag/

-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to [EMAIL PROTECTED]

Reply via email to