Re: [RFC/PATCH] revoke/frevoke system calls

[Pekka Enberg <penberg@cs.helsinki.fi>]

Hi,

On Sat, 22 Jul 2006 09:22:37 +0300 (EEST)
Pekka J Enberg <penberg@cs.Helsinki.FI> wrote:
> > There are alternatives, playing games with ->f_op, creating fake struct 
> > file, and doing IS_REVOKED if-else in the paths, but I think this is by 
> > far the simplest way to do it. So in the Andrew scale of sads, how 
> > sad is it, exactly?-)

On Fri, 2006-07-21 at 23:59 -0700, Andrew Morton wrote:
> Sad enough.  Certainly worth an if-else to fix.

Actually, we can fix it with file->f_light thing Tigran is doing:

http://developer.osdl.org/dev/fumount/kernel2/patches/2.6.12/1/forced-unmount-2.6.12-1.patch

On Fri, 2006-07-21 at 23:59 -0700, Andrew Morton wrote:
> Why is this approach so different from Tigran's, I wonder.

Not so different. I am blocking fork until I can revoke all open file
descriptors (i.e. substitute with NULL) whereas Tigran is dropping
tasklist_lock and retrying. I am not doing get_bad_file() because I
don't think we really need it. Tigran's mmap takedown code looks pretty
much what I want too.

On Fri, 2006-07-21 at 23:59 -0700, Andrew Morton wrote:
> iirc, one of the things we added file.f_mapping for was revokation, but
> this patch doesn't use it.  Please ask Al Viro about this.

I searched fsdevel archives but couldn't find anything on that. Al?

				Pekka