> The right solution for unionfs is for it to simply return the
> underlying (st_dev,st_ino) for the actual inodes, in the real file
> systems, that are holding the files.

> This will then have the appearance of a very ugly mess of unusual
> mount points - but at least the semantics would be correct.

Not quite; it still means that a single object can have its
(st_dev,st_ino) pair change during its lifetime (eg, while it has a fd
open on it), which is a fairly drastic semantic change.  (I know of no
precedent for fstat on a given fd returning different st_dev/st_ino
values at different times; if you know of any, I'd be very interested
in hearing details.)

> Overall however, [the changing-values problem] is much less of a
> problem than even considering the possibility that two different
> files may have the same (dev,ino) pair

Entirely agreed.  Of the two evils it is definitely the lesser.

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B