Re: [PATCH] root-hopping for pre-2.3.41-3

From: Horst von Brand (vonbrand@sleipnir.valparaiso.cl)
Date: Wed Jan 26 2000 - 22:18:20 EST


"Richard B. Johnson" <root@chaos.analogic.com> said:
> On Wed, 26 Jan 2000, Horst von Brand wrote:

[...]

> > I'd say go with the proposal of the fake root/cwd directory for kernel
> > threads that don't need access to the filesystem (all of them, I
> > guess). Cleanest way out, unless you allow them to live without root/cwd.

> Most if not all of the use of "root-hopping" occurs upon startup
> where a single statically-linked shell executes statically-linked
> 'insmod' to prepare for the new root. The real init starts with
> the new root and never even knows about any previous.

I know. But gratuitous / and cwd for threads are in the way. If they
aren't needed, get rid of them. Might help in other cases too.

[...]

> Now, if somebody expects to "root-hop" once the system is up
> with mapped shared-libraries, etc., they are in a world of hurt.
> That's not what changing the root was all about. If you intend
> to make root-hopping take on a new meaning and cover all the
> cases of changing the root of a "hot" operating system, it's
> going to be a lot of code to do not much that's useful in the
> end.

I fully understand that. But if it doesn't cost much to round it out
and give it a clean definition (i.e., works only for directories that
aren't in use, no mounts underneath allowed; and then make the rest of
the kernel able to comply for /) is a win IMHV. Might be useful in
other cases too (it would in fact provide atomic umount + mount
another fs) if used elsewhere.

-- 
Horst von Brand                             vonbrand@sleipnir.valparaiso.cl
Casilla 9G, Viņa del Mar, Chile                               +56 32 672616

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Mon Jan 31 2000 - 21:00:17 EST