Re: (reiserfs) dynamic stat data (was Re: LVM / Filesystems / High availability)
Hans Reiser (reiser@ricochet.net)
Mon, 22 Jun 1998 23:59:31 -0700
Stephen C. Tweedie wrote:
>
> Hi,
>
> In article <Pine.LNX.3.96.980619205345.20873A-100000@nightshade.ml.org>,
> <greg@nightshade.ml.org> writes:
>
> > Perhaps an intelgent thing to do would be to have the FS support
> > 'extended metadata'. There could be a pointer at the end of the
> > metadata block to some 'extmetadata' node someplace. At the node there
> > could be n where n<256 key=value pairs. This could be used to store
> > ACLs, compression info, and such. (pref things that arn't needed every
> > file access) There could also be some syscalls so that userspace apps
> > could add/remove/view these things..
>
> For kernel-internal data structures such as ACLs and POSIX capability
> masks, 2.3 ext2fs should support such things. However, that will be
> integrated tightly into the filesystem and will not be exported through
> the VFS. Starting to push this higher up the stack just breaks Unix
> semantics in loads of really unhelpful ways. We already have well
> defined APIs for ACLs and capabilities, plus we have a requirement that
> the kernel enforces these features. Together, that justifies a
> kernel-mode solution. For general user-mode metadata, we really do want
> a user-mode solution (eg. dot files and dot directories). Reiserfs's
> efficient small file handling should make a kernel mode solution
> unnecessary. :)
It would be nice to say cat /usr/local/bin/xemacs/version and get the
version of the xemacs installed, say /usr/local/bin/xemacs/ACL and get
the ACLs on xemacs, say /usr/local/bin/xemacs/owner and get the traditional
file stat_data field of owner, but still say /usr/local/bin/xemacs to run
xemacs by referencing the file body (which can be thought of as the default
file of the directory xemacs.)
I can't implement this though with my current resources, however much I'd like to see it.
Hans
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu