Subject: Re: Large inodes for ffs
To: Simon Burge <simonb@telstra.com.au>
From: Jason Thorpe <thorpej@nas.nasa.gov>
List: tech-kern
Date: 03/23/1999 14:01:12
On Wed, 24 Mar 1999 08:51:02 +1100 
 Simon Burge <simonb@telstra.com.au> wrote:

 > Is this intended to replace the 128 byte inode or be a choice at newfs
 > time?  I can think of some situations where the doubling the inode
 > overhead might not be desired - news (and possibly mail) spools, as well
 > as my cddb database partition.

It's optional, at newfs time.  A normal FFS is created unless you explcitly
create a large-inode FFS.

 > Is a byte enough?  Maybe a u_int32_t might give more room for
 > expansion...

...considering the number of u_int32_t's actually avalable for expansion,
you don't need many flags to say which ones are valid :-)

Hm, but I didn't think Bill used a byte, but rather a u_int32_t for the
flags... Bill?

        -- Jason R. Thorpe <thorpej@nas.nasa.gov>