>> Is there any difference in the output of disklabel between 1.5.3 and 4.0?
>
> Yes.  What do you make of this?
>
> NetBSD 1.5.3
>
> type:  SCSi
> disk:  AXSITS2532A 011
> label:  fictitious
> flags:
> bytes/sector:  512
> sectors/track:  63
> tracks/cylinder:  16
> sectors/cylinder:  1008
> cylinders:  2100
> total sectors:  2116800
> rpm:  3600
> interleave:  1
> trackskew:  0
> cylinderskew:  0
> headswitch:  0			#  microseconds
> track-to-track seek:  0		#microseconds
> drivedata:  0
>
> NetBSD 4.0R5
>
> type:  SCSI
> disk:  AXSITS2532A  011E
> label:  fictitious
> flags:
> bytes/sector:  512
> sectors/track:  32
> tracks/cylinder:  64
> sectors/cylinder:  2028
> cylinders:  2116800
> rpm:  3600
> interleave:  1
> trackskew:  0
> cylinderskew:  0
> headswitch:  0			#  microseconds
> track-to-track seek:  0		#microseconds
> drivedata:  0

All of that stuff is made up by NetBSD. Drives don't actually have a set 
number of sectors per track, for instance, and haven't for two decades. 
What really matters is if there's a difference in the way the partitions 
are reported.

There is a possibility that there's some off-by-one problem in the 
filesystem code which is why deliberate gaps, even though they're tiny, 
appears to make everything happy.

John