At 10:05 AM 7/24/2002 -0700, Bill Studenmund wrote:
>On Wed, 24 Jul 2002, Konrad Schroder wrote:
>> separate value.  (LFSv2 just uses fragments for quotas and di_blocks, so
>> it doesn't need the extra parameter; LFSv1 uses 512 :^)
>
>??
>
>Why does it need to be a separate value? What separate from what? I think
>you're saying it needs to be separate from the (at-creation) sector size,
>which I don't think is right; I think they should be the same.
>
>In the ufs code, it needs to be a mount parameter, for LFS & FFS (using
>different values).

(Light bulb clicks on.)  Of course!  It's so simple.  The ufs code doesn't
need the sector size per se; it only needs the di_blocks unit (which we're
defining as the stored sector size in ffs, but is the fsb size in lfs.)

Okay, I made the adjustments, removed my changes to the di_blocks scale in
lfs, and ran a cvs diff on my changes (thanks Konrad).  The patches can be
retrieved from my ISP:

	http://www.xmission.com/~trevin/NetBSD_patches-2002.07.24

-----------------------
Trevin Beattie          "Do not meddle in the affairs of wizards,
trevin@xmission.com     for you are crunchy and good with ketchup."
      {:->                                     --unknown