Subject: Re: More on [r]dump
To: None <torek@horse.ee.lbl.gov>
From: Charles Hannum <mycroft@duality.gnu.ai.mit.edu>
List: netbsd-bugs
Date: 12/01/1993 22:53:06
   Date: Wed, 17 Nov 93 09:15:22 PST
   From: Chris Torek <torek@horse.ee.lbl.gov>

   Inode numbers do indeed start from 1.  Craig Leres and I hacked
   some of this code quite a bit after net.2 and before 4.4BSD---I
   decided I could not stand the inconsitencies between the various
   inode loops, and made them all be of the form:

	   for (ino = 1; ino < maxino; ino++)

   Note that this requires changing the code that sets maxino, which
   currently says something like:

	   maxino = sblock->fs_ipg * sblock->fs_ncg - 1;

   with the `-1' factor existing to make `<=' correct.

I just changed them all to for form `(ino = 1; ino <= maxino;
ino++)'.  I decided that removing the `- 1' required either two more
`- 1's in calls to writeheader() and dumpmap(), or touching a bunch of
other code, which seemed to be entering the Danger Zone.


------------------------------------------------------------------------------