Re: Problem with find + AFS + acl="l"

[James Youngman]

On 11/1/06, Daniel Richard G. <address@hidden> wrote:
> On Wed, 2006 Nov 01 18:42:34 +0000, James Youngman wrote:
> >
> > But my question is, does AFS guanrantee that readdir() will return
> > them in that order?
>
> It does appear that "." and ".." come first consistently, but is this
> really necessary? Ext3 itself doesn't do this.

If we see a DT_UNKNOWN before anything else in a directory, we
wouldn't know whether to assume it's a directory or not, unless we had
previously seen a DT_DIR for "." or "..".   And even then we would
only know the DT_UNKNOWN was a nondirectory if we are sure the
directory is in an AFS filesystem, and as we discuss below that may be
nontrivial...

> There isn't a good way of going from some arbitrary path to a filesystem
> type? (I'm not very familiar with this part of the POSIX API, alas.)

Not without stating all filesystem mountpoints, which would cause find
to hang on systems which are clients of unreachable NFS servers.

> If not, it might be sufficient to check if the directory's canonical
> path is under /afs. (You're never going to have AFS mounted on e.g. /usr
> or /mnt/afs.)

It's not always possible to figure out the canonical path (for example
if a parent of the start point has no execute permission).

> > Yes, that's a good idea.  Could you log this as a separate bug on
> > Savannah? (svannah.gnu.org).    The same thing could also be useful
> > for entries in directories which allow r but not x.
>
> Will do!

Thanks.