Re: [PATCH 3/4] integrity: Linux Integrity Module(LIM)

["Serge E. Hallyn" <serue@us.ibm.com>]

Quoting Christoph Hellwig (hch@infradead.org):
> On Mon, Aug 11, 2008 at 12:02:55PM -0500, Serge E. Hallyn wrote:
> > > > Sorry, but I don't think we can bloat the inode even further for this.
> > > 
> > > The original version of IMA was LSM based, using i_security. Based
> > > on discussions on the LSM mailing list, it was decided that the LSM hooks
> > > were meant only for access control.  During the same time frame, there 
> > > was a lot of work done in stacking LSM modules and i_security, but that
> > > approach was dropped. It was suggested that we define a separate set of
> > > hooks for integrity, which this patch set provides. Caching integrity 
> > > results is an important aspect. Any suggestions in lieu of defining 
> > > i_integrity?
> > 
> > The i_integrity is only bloating the inode if LIM is enabled.  Surely
> > that beats having LIM define its own hash table and locking to track
> > integrity labels on inodes?  Do you have another suggestion?
> > 
> > Or is the concern about having more #ifdefs in the struct inode
> > definition?
> 
> No, the concern is over bloating the inode for a rather academic fringe
> feature.  As this comes from IBM I'm pretty sure someone will pressure
> the big distro to turn it on.

By default??  I should hope not...

Note that these are all not loadable modules.  So presumably either it's
in the kernel and enforcing, or it's not there.

> And inode growth is a concern for
> fileserving or other inode heavy workload.  Mimi mentioned this is just
> a cache of information, so consider using something like XFS's mru cache
> which is used for something similar where the xfs_inode was kept small
> despite a very niche feature needing a cache attached to the inode:
> 
> 	fs/xfs/xfs_mru_cache.c

ok, so basically as I said above

> > ... having LIM define its own hash table and locking to track
> > integrity labels on inodes?

:)

But then that is in fact the better way to go if there can be a lot
of inodes with i_integrity=NULL.  It looks like IMA always allocates
something, but if I understand the idea behind templates correctly,
that isn't necessarily always the case.

thanks,
-serge