> >Yes, you double the number of bits in the allocation map (1 block) and
> >double the amount that each bit refers to (1 fragment each).

> but cpg=42760 for bsize=16k is not a double of the cpg=28456 for bsize=8k...

IIRC your bsize=16k fs had 210MB per cg or 210 * 2000 = 420000 sectors,
the cpg field is only 16 bits so you get massive truncation - ie the
number is failry useless!
To find the second alternate you would to look at 'n * 65536 + cpg'
for increasing 'n'.

Alternate techniques (eg guessing the algorithm that newfs might have
used, or a simple linear scan) are probably better for finding the
later alternate sbs.

	David

-- 
David Laight: david@l8s.co.uk