> Theodore Y. Ts'o wrote:
> >
> > B-tree directories can be cleanly added to ext2. The hard part is
> > writing code that implements B-trees cleanly in the kernel (including
> > handling the case of B-tree splits in the SMP case). Once we have this
[...]
> What does Posix say about directory operations if the directory is
> changed (telldir, seekdir and readdir)? It will be difficult to
> guarantee any kind of consistency.
There are non-destructive tree structures commonly used in functional
programming circles that can guarantee that both the old and new
directories trees are valid. By creating the new tree without
destroying the old, marking the new tree as active, waiting until all
read locks on the old tree are released, *then* deleting
the old tree, there should be no consistency problems.
Non-trivial, but certainly possible.
--Scott
@ @
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-oOO-(_)-OOo-=-=-=-=-=
C. Scott Ananian: cananian@lcs.mit.edu / Declare the Truth boldly and
Laboratory for Computer Science/Crypto / without hindrance.
Massachusetts Institute of Technology /META-PARRESIAS AKOLUTOS:Acts 28:31
-.-. .-.. .. ..-. ..-. --- .-. -.. ... -.-. --- - - .- -. .- -. .. .- -.
PGP key available via finger and from http://www.pdos.lcs.mit.edu/~cananian
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu