I already sent email to some other people stating that this is what I want
to do anyway..
> A quick glance suggests that all operations except for
> (read|writ|updatee)page, bmap, truncate, smap and revalidate will have to
> move. If there's no great objection to this, I'll put together a patch
> that does this.
Wait for my pre-patch to show up on ftp.kernel.org. I've changed
"lock_parent()" to move us into this direction (ie the "parent" is no
longer the inode, it's really the parent dentry). That should make your
patches smaller too.
> Sorry if it seems an enourmous change, but it's the only way I can see
> for us to get the usefulness of a dentry-centric system that allows us to
> implement filesystems without underlying indoes. Hey, ideally
> directories wouldn't *need* an inode resident in memory for networks
> filesystems.
It's not an enormous change - the naming code conceptually is a "dentry"
issue, not an inode issue.
Obviously for historical reasons all the operations were inode operations
simply because we didn't _have_ any dentries..
Linus